Ecological Studies, Vol. 200 Analysis and Synthesis
Edited by M.M. Caldwell, Washington, USA G. Heldmaier, Marburg, Germany R.B. Jackson, Durham, USA O.L. Lange, Würzburg, Germany H.A. Mooney, Stanford, USA E.-D. Schulze, Jena, Germany U. Sommer, Kiel, Germany
Ecological Studies Further volumes can be found at springer.com Volume 184 Reindeer Management in Northernmost Europe: Linking Practical and Scientific Knowledge in Social-Ecological Systems (2006) B.C. Forbes, M. Bölter, L. Müller-Wille, J. Hukkinen, F. Müller, N. Gunslay, and Y. Konstantinov (Eds.) Volume 185 Ecology and Conservation of Neotropical Montane Oak Forests (2006) M. Kappelle (Ed.) Volume 186 Biological Invasions in New Zealand (2006) R.B. Allen and W.G. Lee (Eds.) Volume 187 Managed Ecosystems and CO2: Case Studies, Processes, and Perspectives (2006) J. Nösberger, S.P. Long, R.J. Norby, M. Stitt, G.R. Hendrey, and H. Blum (Eds.) Volume 188 Boreal Peatland Ecosystem (2006) R.K. Wieder and D.H. Vitt (Eds.)
Volume 192 Geological Approaches to Coral Reef Ecology (2007) R.B. Aronson (Ed.) Volume 193 Biological Invasions (2007) W. Nentwig (Ed.) Volume 194 Clusia: A Woody Neotropical Genus of Remarkable Plasticity and Diversity (2007) U. Lüttge (Ed.) Volume 195 The Ecology of Browsing and Grazing (2008) I.J. Gordon and H.H.T. Prins (Eds.) Volume 196 Western North American Juniperus Communites: A Dynamic Vegetation Type (2008) O. Van Auken (Ed.) Volume 197 Ecology of Baltic Coastal Waters (2008) U. Schiewer (Ed.)
Volume 189 Ecology of Harmful Algae (2006) E. Granéli and J.T. Turner (Eds.)
Volume 198 Gradients in a Tropical Mountain Ecosystem of Ecuador (2008) E. Beck, J. Bendix, I. Kottke, F. Makeschin, R. Mosandl (Eds.)
Volume 190 Wetlands and Natural Resource Management (2006) J.T.A. Verhoeven, B. Beltman, R. Bobbink, and D.F. Whigham (Eds.)
Volume 199 Hydrological and Biological Responses to Forest Practices: The Alsea Watershed Study (2008) J.D. Stednick (Ed.)
Volume 191 Wetlands: Functioning, Biodiversity Conservation, and Restoration (2006) R. Bobbink, B. Beltman, J.T.A. Verhoeven, and D.F. Whigham (Eds.)
Volume 200 Arid Dune Ecosystems: The Nizzana Sands in the Negev Desert (2008) S.-W. Breckle, A. Yair, and M. Veste (Eds.)
Siegmar-W. Breckle • Aaron Yair • Maik Veste Editors
Arid Dune Ecosystems The Nizzana Sands in the Negev Desert
Prof. Dr. Siegmar-W. Breckle University of Bielefeld Department of Ecology Wasserfuhr 24-26 33619 Bielefeld Germany
Prof. Dr. Aaron Yair Hebrew University Department of Geography Mount Scopus Campus Jerusalem 91905 Israel
Dr. Maik Veste University of Hohenheim Institute of Botany Experimental Botany 70599 Stuttgart Germany
Cover illustration: Nizzana Sands: active crest of a linear dune. South-facing slip faces caused by afternoon sea breezes during summer months. (Photograph Axel Allgaier)
ISBN 978-3-540-75497-8
e-ISBN 978-3-540-75498-5
Ecological Studies ISSN 0070-8356 Library of Congress Control Number: 2007941784 © 2008 Springer-Verlag Berlin Heidelberg This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permissions for use must always be obtained from Springer-Verlag. Violations are liable for prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover design: WMXDesign GmbH, Heidelberg, Germany Printed on acid-free paper 9 8 7 6 5 4 3 2 1 springer.com
Preface
“Nobody can claim that poverty and desertification are not linked.” In his keynote speech launching the “Governance and Combating Desertification” Conference, held on 7 September 2006 at gtz-Haus in Berlin, Klaus Töpfer, former Executive Director of the United Nations Environment Programme (UNEP), advocated attaching higher priority to the problem of soil erosion. “Contaminated water can be treated, as the example of the Rhine shows, but eroded soil is lost forever.” Drought and desertification threaten the livelihoods of more than 1.2 billion people in over 110 countries. An area of more than 3.5 times the size of Europe is already affected, and future scenarios predict that desertification worldwide will continue to advance. The resulting economic damage is estimated to be more than 40 billion US dollars per year. To mark the International Year of Deserts and Desertification 2006, the German Federal Ministry for Economic Cooperation and Development hosted a conference to debate the suitability of the UNCCD as a global governance tool. Klaus Töpfer expressed some reservations: “All our conventions are short-term responses to a problem that we have identified. I believe we should stop producing convention after convention, and start looking at them in a more joined-up manner. I am absolutely convinced that this will not only save money and resources, but also significantly improve the quality of governance associated with the conventions. Desertification is not restricted to deserts. But to know and to understand ecosystem processes in desert ecosystems can help to find means of an adequate use of specific desert ecosystems without severe degradation.” We have to distinguish between deserts and desertification. Desertification quite often affects the vulnerable desert margins, which are much more used by human populations than the deserts as such. But even deserts can exhibit signs of desertification and degradation. Sand dune ecosystems are considered to be particularly characteristic of deserts. Sand storms are a threat to people. However, sand dune deserts cover only about 25–30% of the world’s deserts. Sand dunes of the Namib, in Arabian Rub-al-Khali, or in the Mongolian deserts can be higher than 300 m in relative height. They document an arid climate and a large area of eroding stone material or accumulated material from rivers, as in the case of the Negev sand dunes, derived from the Nile. There are also sand dunes in humid areas along many coastlines, where sand is v
vi
Preface
formed by steady wave action and accumulated by strong winds. In this book, however, we deal with arid sand dune systems and their specific ecological processes. Thus, this volume (the “Nizzana” book) gives an overview to many aspects of the very characteristic, predominantly longitudinal sand dune ecosystems, especially of the Northern Negev, as a key case study. It helps to understand a superficially simple-looking ecosystem with a complicated interrelation of climate, vegetation and sandy substrate, and a complex dynamics in space and time. There are many co-authors to bring together these results. This took some time. And it means that rather different aspects have been put together; this also means that a few overlappings may occur giving differing viewpoints and raising open questions. Research on plants under dry field conditions was strongly stimulated by Otto Ludwig Lange. He started measuring photosynthesis of plants at the Avdat runoff farm of M. Evenari in Israel, almost 40 years ago. For his international pioneer work and his stimulation of ecological field research, he received numerous international honours. In 1989, he resumed his fruitful cooperation by establishing the Arid Ecosystems Research Centre (AERC). The Centre was a joint venture of the Minerva foundation in Germany and the Hebrew University in Israel. Otto Ludwig Lange was nominated first Chairman of the Centre. The first initiative of the research centre was the establishment of the Nizzana sandy research site, along the Egyptian-Israeli border, north of the Nizzana settlement. Prof. Lange inspired many of the works included in this book. In 1992, he published (with some colleagues) the first paper dealing with the taxonomy and photosynthetic activity of the biological soil crusts in the Nizzana area (Functional Ecology 6: 519–527). His strong interest in biological topsoil crusts in arid environments is not only very well expressed in the Nizzana book, where several chapters are devoted to the important role that biological soil crusts play in the structure and functioning of a sandy arid ecosystem, but also in volume 150 of the Ecological Studies series (Belnap & Lange 2001: “Biological soil crusts: structure, function, and managment”). Various research groups were involved in joint and international as well as interdisciplinary research. Main results are given in this volume. This volume 200 of the Ecological Studies series is dedicated to Otto Ludwig Lange on the occasion of his 80th birthday. He was born in August 1927 in Dortmund, Germany. Already the volume 100 was dedicated to him on the occasion of his retirement. We take this opportunity to acknowledge O.L. Lange’s great scientific work on desert ecology. We do hope that this special volume on sand dune ecology is useful to biologists, plant and soil ecologists, conservation biologists, desert ecologists and geographers, geomorphologists, climatologists, policy makers, site practitioners, researchers, lecturers, tutors, and many others with an interest in deserts and also in desertification. Several research projects were funded by the BMBF (German Ministry for Education and Research); we thank Hans-Michael Biehl, Hans-Georg Bertram
Preface
vii
and Joachim Kutscher from the Project Management Jülich for their support of the Nizzana projects. The editors gratefully acknowledge Dr. Andrea Schlitzberger at Springer Verlag in Heidelberg for her steady interest, patience and help during preparation of the book, Dr. Dieter Czeschlik for his continuous support, as well as Prof. Otto Lange for his steady interest, his critical reading and valuable comments. November 2007
Siegmar-W. Breckle, Bielefeld Aaron Yair, Jerusalem Maik Veste, Hohenheim
Contents
General Introduction – Desert Sand Dunes and Aims of the Book – Special Characteristics of the Nizzana Research Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S.-W. Breckle, A. Yair, and M. Veste References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Part A 1
2
1 6
The North-Western Negev
Geological Background of the Nizzana Area. . . . . . . . . . . . . . . . . . . . . R. Ben-David and A. Yair
9
1.1 1.2
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Late Geological History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.1 Tertiary Era . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.2 Quaternary Era . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Discussion of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9 9 9 11 13 14 15
Geo-Ecology of the North-Western Negev Sand Field . . . . . . . . . . . . . A. Yair, M. Veste, and S.-W. Breckle
17
2.1 2.2
17 19 20 22 23 23 24
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geo-Ecological Units in the Hallamish Sand Field . . . . . . . . . . . . . 2.2.1 The Sandy Ridge. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 The Interdune Corridor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Geo-Ecology of the Haluza-Agur Sand Field. . . . . . . . . . . . . . . . . . 2.4 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ix
x
3
Contents
Formation and Geomorphology of the North-Western Negev Sand Dunes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H. Tsoar, D.G. Blumberg, and R. Wenkart 3.1 3.2
The Sinai-Negev Dunefield. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aeolian Sand Incursions into the North-Western Negev During the Upper Quaternary . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Period of Aeolian Sand Incursion into the Negev. . . . . . . . . 3.2.2 The Sand Red Colour and its Implications . . . . . . . . . . . . . . 3.3 Wind Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Factors Affecting Mobility and Stability of the Negev Sand Dunes . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 The Negev Dune Forms and Their Evolution . . . . . . . . . . . . . . . . . . 3.4.1 Linear Dunes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 The Effect of Destruction of Vegetation on the Morphology and Dynamics of the Sand Dunes . . . . . . . . . . . . . 3.6 Buried Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.1 Methods of Mapping Buried Drainage Systems . . . . . . . . . . 3.6.2 Nahal Nizzana and Buried Drainage Systems in the Shunra and Haluza Sand Fields . . . . . . . . . . . . . . . . . 3.7 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
25 26 26 27 31 32 33 33 40 42 43 44 46 46
The Regional Climatic Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . T. Littmann and S.M. Berkowicz
49
4.1 4.2
49
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Climate Dynamics in the Eastern Mediterranean and Adjacent Areas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Characteristics of the Northern Negev Climate . . . . . . . . . . . . . . . . 4.3.1 Temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Rainfall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 Spatial and Temporal Patterns of Rainfall . . . . . . . . . . . . . . 4.3.4 Regional Vapour Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.5 Dewfall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
25
50 53 53 54 56 59 60 61 62
Soil Characteristics and Pattern of the Nizzana Research Site . . . . . . H.-P. Blume, L. Beyer, U. Pfisterer, and P. Felix-Henningsen
65
5.1 5.2 5.3
65 65 67 68 68
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Soil Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 Soils of the Sandy Ridges . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2 Soils of the Interdune Playa Surfaces . . . . . . . . . . . . . . . . . .
Contents
6
xi
5.3.3 Soils of the Interdune Area . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.4 Soils at the Haluza Station (N3) . . . . . . . . . . . . . . . . . . . . . . 5.4 Ecological Conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1 Root Penetration Capability . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.2 Water and Oxygen Supply . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.3 Nutrient Supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
71 73 74 74 75 75 76 77
Land Use and its Effect on the Mobilization and Stabilization of the North-Western Negev Sand Dunes . . . . . . . . . . . . H. Tsoar
79
6.1 6.2 6.3 6.4
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Factors Affecting Vegetation Growth on Sand Dunes . . . . . . . . . . . The Effect of the Border on Bedouin Pasture Management . . . . . . . The Effect of Bedouin Pressure on the Negev Sand Dunes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Part B 7
8
79 79 82 83 87 88
Ecosystem Patterns
The Flora of the Nizzana Research Site . . . . . . . . . . . . . . . . . . . . . . . . . K. Tielbörger, R. Prasse, and H. Leschner
93
7.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Conclusions and Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
93 93 94 102 103
The Vegetation of the Nizzana Research Site . . . . . . . . . . . . . . . . . . . . K. Tielbörger, R. Prasse, and R. Bornkamm
105
8.1 8.2 8.3
105 106 107 108 117 120 120 121 121 122 123
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.1 Main Plant Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.2 Annual Plant Distribution Patterns . . . . . . . . . . . . . . . . . . . . 8.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.1 Syntaxonomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.2 Annual Vegetation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.3 Vegetation and Habitat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xii
9
Contents
A Glance on the Fauna of Nizzana . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. Filser and R. Prasse 9.1 9.2
10
11
125
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Study Sites and Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.1 List of Vertebrates and Larger Invertebrates . . . . . . . . . . . 9.2.2 Invertebrate Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.1 Composition of the Fauna . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.2 Ecosystem Engineering. . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.3 Spatial Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4.1 Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4.2 Biotic Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4.3 Ecosystem Engineering. . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4.4 Spatial Heterogeneity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
125 126 126 126 128 128 130 136 139 139 144 144 145 145 146
Biological Crusts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Büdel and M. Veste
149
10.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2 Structure of Biological Soil Crusts . . . . . . . . . . . . . . . . . . . . . . . . 10.3 Crust Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4 Species Diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
149 149 149 151 154 154
Land Cover in the Nizzana Sandy Arid Ecosystem. Mapping Surface Properties with Multi-Spectral Remote Sensing Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. Hill, T. Udelhoven, T. Jarmer, and A. Yair 11.1 11.2 11.3 11.4
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Study Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Field and Remote Sensing Data . . . . . . . . . . . . . . . . . . . . . . . . . . Spatial Variability of Crust Properties Within the Sand Dune Ecosystem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.5 Mapping the Spatial Diversity of Surface Properties with High Spatial-Resolution Aerial Photographs . . . . . . . . . . . . 11.5.1 Domain-Specific Unmixing . . . . . . . . . . . . . . . . . . . . . . 11.5.2 The Vegetated Domain . . . . . . . . . . . . . . . . . . . . . . . . . . 11.5.3 The Substrate Domain . . . . . . . . . . . . . . . . . . . . . . . . . . 11.6 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
157 157 158 159 161 164 165 166 168 170 171
Contents
xiii
Part C Ecosystem Processes 12
Topoclimate and Microclimate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . T. Littmann
175
12.1 12.2
175
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Radiative Energy Budget and Temperatures on Sand Dune Slopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.3 Precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.4 Near-Ground Wind Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
Evapotranspiration, Transpiration and Dewfall. . . . . . . . . . . . . . . . . T. Littmann and M. Veste
183
13.1 13.2 13.3
183 183
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microclimatic Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evapotranspiration Models and Their Application to Dewfall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.4 Zero Plane Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.5 Application of the Zero Plane Model . . . . . . . . . . . . . . . . . . . . . . 13.6 Model Validation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.6.1 Validation of Dewfall . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.6.2 Transpiration and Evapotranspiration. . . . . . . . . . . . . . . 13.6.3 Calculation of Evapotranspiration and its Ecological Implications. . . . . . . . . . . . . . . . . . . . 13.7 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14
175 178 179 182 182
184 189 191 196 196 196 197 198 199
Morphological Changes at Active Dune Crests. . . . . . . . . . . . . . . . . . A. Allgaier
201
14.1 14.2 14.3 14.4
201 201 202 203 203 207 209 209
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Research Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.4.1 Monitored Changes 1997 to 1999. . . . . . . . . . . . . . . . . . 14.4.2 Observed Changes 1993–1999 . . . . . . . . . . . . . . . . . . . . 14.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xiv
15
Contents
Aeolian Sand Transport and Vegetation Cover. . . . . . . . . . . . . . . . . . A. Allgaier
211
15.1 15.2
211 212 213
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Study Sites and Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.2.1 Site A: No Vegetation, No Surface Crust . . . . . . . . . . . . 15.2.2 Site B: Natural and Reduced Vegetation Cover, Without Microphytic Surface Crust . . . . . . . . . . . . . . . . 15.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.3.1 Sand Transport at Site A. . . . . . . . . . . . . . . . . . . . . . . . . 15.3.2 Sand Transport at Site B . . . . . . . . . . . . . . . . . . . . . . . . . 15.3.3 Influence of Vascular Vegetation on Sand Movement . . 15.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
16
17
214 215 215 215 216 220 222 223
Soil Processes and Salt Dynamics in Dune Soils . . . . . . . . . . . . . . . . . P. Felix-Henningsen, B. Rummel, and H.-P. Blume
225
16.1 16.2 16.3
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Soil Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.3.1 Formation of Parent Material . . . . . . . . . . . . . . . . . . . . . 16.3.2 Weathering, Brownification and Redoximorphism. . . . . 16.3.3 Aggregation and Cracking . . . . . . . . . . . . . . . . . . . . . . . 16.3.4 Crust Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.3.5 Humus Accumulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.4 Salt Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.4.1 Salinization of the Playas . . . . . . . . . . . . . . . . . . . . . . . . 16.4.2 Salt Dynamics of Arenosols on Vegetated Linear Dunes . . . . . . . . . . . . . . . . . . . . . . . 16.5 Spatial Variability of Soil Characteristics . . . . . . . . . . . . . . . . . . . 16.6 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
225 226 227 227 227 228 228 229 231 231
Runoff and Erosion Processes Within a Dune System . . . . . . . . . . . . G.J. Kidron and A. Yair
239
17.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.2 Field Instrumentation and Methodology. . . . . . . . . . . . . . . . . . . . 17.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
239 240 242 246 248 249
232 236 237 238
Contents
18
Effects of Surface Runoff and Subsurface Flow on the Spatial Variability of Water Resources in Longitudinal Dunes . . . . . . . . . . . A. Yair 18.1 18.2 18.3
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aim of Study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results of Field Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.3.1 Rainfall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.3.2 Rainfall–Runoff Relationships . . . . . . . . . . . . . . . . . . . . 18.4 The Effect of Subsurface Water Movement on Water Redistribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.4.1 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.4.2 Conditions for Subsurface Lateral Water Flow . . . . . . . 18.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Atmospheric Input of Nutrient Elements and Dust into the Sand Dune Field of the North-Western Negev . . . . . . . . . . . T. Littmann and A. Schultz 19.1 19.2 19.3 19.4
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dimensions of Atmospheric Deposition. . . . . . . . . . . . . . . . . . . . Element Groups and the Boundary Conditions of Atmospheric Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.5 The Role of Vegetation Stands . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.6 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
20
Succession Stages in the Recovery Processes of the Topsoil Crust in a Disturbed Sandy Arid Area . . . . . . . . . . . . . . . . . . A. Yair 20.1 20.2 20.3 20.4
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Case of the Nizzana Research Area. . . . . . . . . . . . . . . . . . . . Aim of Present Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.4.1 Sampling Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.4.2 Laboratory Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.5 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.5.1 Wind Regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.5.2 Rainfall Regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.5.3 Recovery of the Mineral Component of the Crust . . . . . 20.5.4 Recovery of the Biological Components of the Crust . . 20.5.5 Recovery of the Vegetation Cover . . . . . . . . . . . . . . . . . 20.6 Discussion and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xv
251 251 253 253 253 254 261 262 266 266 268
271 271 272 272 276 279 282 283 285 285 287 288 288 288 289 291 292 293 293 294 295 297 301
xvi
21
Contents
Dew Formation and Activity of Biological Soil Crusts. . . . . . . . . . . . M. Veste, B.G. Heusinkveld, S.M. Berkowicz, S.-W. Breckle, T. Littmann, and A.F.G. Jacobs
305
21.1 21.2
305 306 306 307 311
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dew and Fog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2.1 Definition and Measurement. . . . . . . . . . . . . . . . . . . . . . 21.2.2 Dew and Fog in the Northern Negev Desert. . . . . . . . . . 21.3 Physiological Activity of Biological Soil Crusts . . . . . . . . . . . . . 21.3.1 Activation of Soil Lichens After Nocturnal Wetting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3.2 Photosynthetic Activity After Sunrise . . . . . . . . . . . . . . 21.4 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Nitrogen Input Pathways into Sand Dunes: Biological Fixation and Atmospheric Nitrogen Deposition . . . . . . . . R. Russow, M. Veste, S.-W. Breckle, T. Littmann, and F. Böhme 22.1 22.2
23
311 314 316 316
319
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.2.1 Study Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.2.2 Species Investigated . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.2.3 Biological Soil Crusts . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.2.4 Sampling for 15N Determination . . . . . . . . . . . . . . . . . . . 22.2.5 Sampling Atmospheric Deposition. . . . . . . . . . . . . . . . . 22.2.6 15N Methodology and Calculation of Biological N Fixation . . . . . . . . . . . . . . . . . . . . . . . . . 22.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.3.1 Soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.3.2 Biological Soil Crusts . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.3.3 15N Retama raetam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3.4 Estimation of N Input by BNF into the Ecosystem . . . . 22.3.5 Atmospheric Nitrogen Deposition . . . . . . . . . . . . . . . . . 22.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
319 320 320 320 321 321 321
Vascular Plant Response to Microbiotic Soil Surface Crusts . . . . . . R. Prasse and R. Bornkamm
337
23.1 23.2
337 338 338 339 339 339 340
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23.2.1 Plinth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23.2.2 Interdune . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23.2.3 Hard Crust. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23.2.4 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23.2.5 Statistical Analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
322 323 323 324 326 327 329 331 334 334
Contents
24
25
26
xvii
23.3
Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23.3.1 Densities and Species Numbers . . . . . . . . . . . . . . . . . . . 23.3.2 Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23.3.3 Fecundity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23.3.4 Underlying Mechanisms. . . . . . . . . . . . . . . . . . . . . . . . . 23.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
340 340 342 343 345 346 348 349
Ion Relations of Plants and Soil Patterns. . . . . . . . . . . . . . . . . . . . . . . M. Veste, U. Sartorius, and S.-W. Breckle
353
24.1 24.2
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Material and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24.2.1 Plant Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24.2.2 Ion Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24.3.1 Ion Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24.3.2 Salt Accumulation in the Standing Biomass . . . . . . . . . 24.3.3 Salt Accumulation Below the Shrubs . . . . . . . . . . . . . . . 24.4 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
353 354 354 354 355 355 358 358 360 363
Temporal and Spatial Variability of Plant Water Status and Leaf Gas Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . M. Veste
367
25.1 25.2
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Material and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25.2.1 Gas Exchange Measurements . . . . . . . . . . . . . . . . . . . . . 25.2.2 Plant Water Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . 25.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25.3.1 Photosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25.3.2 Water Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
367 368 368 368 368 368 370 371 374 374
Standing Biomass and its Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . M. Veste, C. Sommer, S.-W. Breckle, and T. Littmann
377
26.1 26.2
377 377 377 377 379 379 381
26.3
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Standing Biomass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26.2.1 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26.2.2 Biomass. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modelling Biomass Pattern. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26.3.1 The Meso-Scale Model. . . . . . . . . . . . . . . . . . . . . . . . . . 26.3.2 The Micro-Scale Model . . . . . . . . . . . . . . . . . . . . . . . . .
xviii
27
Contents
26.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
382 382 383
Effects of Shrubs on Annual Plant Populations . . . . . . . . . . . . . . . . . K. Tielbörger and R. Kadmon
385
27.1 27.2
385 386 387
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27.2.1 Sampling Design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27.2.2 Measurements of Seedling Densities and Seedling Survival . . . . . . . . . . . . . . . . . . . . . . . . . . . 27.2.3 Measurements of Reproductive Success. . . . . . . . . . . . . 27.2.4 Measurements of Seed Survival and Germination Rates . . . . . . . . . . . . . . . . . . . . . . . . . . 27.2.5 Statistical Analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27.3.1 Rainfall and Germination . . . . . . . . . . . . . . . . . . . . . . . . 27.3.2 Seedling Densities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27.3.3 Reproductive Success . . . . . . . . . . . . . . . . . . . . . . . . . . . 27.3.4 Probability of Survival and Germination of Newly Produced Seeds . . . . . . . . . . . . . . . . . . . . . . . . 27.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27.5 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . R. Kadmon 28.1 28.2
28.3
28.4
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28.2.1 Experimental Design . . . . . . . . . . . . . . . . . . . . . . . . . . . 28.2.2 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28.3.1 Seedling Survival . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28.3.2 Above-Ground Biomass . . . . . . . . . . . . . . . . . . . . . . . . . 28.3.3 Reproductive Allocation . . . . . . . . . . . . . . . . . . . . . . . . . 28.3.4 Reproductive Biomass . . . . . . . . . . . . . . . . . . . . . . . . . . 28.3.5 Fecundity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28.3.6 Fruit Weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28.4.1 Species-Specific Responses to Habitat Heterogeneity. . . . . . . . . . . . . . . . . . . . . . . . . 28.4.2 Species-Specific Responses to Cover by Sand . . . . . . . . 28.4.3 Species-Specific Responses to Neighbour Removal . . . . . . . . . . . . . . . . . . . . . . . . . .
387 387 388 388 389 389 389 391 391 394 398 398
401 401 403 403 405 405 405 406 410 412 412 413 413 414 415 416
Contents
xix
28.4.4
Interactions Between Neighbour Competition and Habitat Heterogeneity . . . . . . . . . . . . . . . . . . . . . . . 28.5 Conclusions and Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
417 418 419
Part D Research Perspectives / Synthesis and General Conclusions 29
Sensitivity of a Sandy Area to Climate Change Along a Rainfall Gradient at a Desert Fringe . . . . . . . . . . . . . . . . . . . A. Yair, M. Veste, R. Almog, and S.-W. Breckle 29.1 29.2 29.3 29.4
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aim of Study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29.4.1 Rainfall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29.4.2 Vegetation Changes Along the Rainfall Gradient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29.4.3 Hydrological Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . 29.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29.6 Implications for the Sensitivity of the Sandy Area to Changing Climatic Conditions . . . . . . . . . . . . . . . . . . . . . . . . . 29.7 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
General Conclusions – Sand Dune Deserts, Desertification, Rehabilitation and Conservation . . . . . . . . . . . . . . . . S.-W. Breckle, A. Yair, and M. Veste 30.1 30.2
425 425 427 428 429 429 430 432 436 437 438 439
441
Sand Deserts and Sand Dunes. . . . . . . . . . . . . . . . . . . . . . . . . . . . Desertification – the Degradation of Sandy Desert Ecosystems and Threat to Adjacent Areas . . . . . . . . . . . . . . . . . . 30.3 Designing Shelterbelts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30.4 Stabilisation of Sand Dunes in the Aralkum. . . . . . . . . . . . . . . . . 30.5 Stabilisation of Sand Dunes in the Tengger Desert . . . . . . . . . . . 30.6 Restoration of Sand Dunes in Southern Africa. . . . . . . . . . . . . . . 30.7 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
441 442 443 445 450 453 456 457
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
461
Taxonomic Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
473
Contributors
Allgaier, A. Sinclair Knight Merz, P.O. Box 2500, Malvern VIC 3144, Australia,
[email protected] Almog, R. Hebrew University, Department of Geography, Mount Scopus Campus, Jerusalem 91905, Israel,
[email protected] Ben-David, R. Hebrew University, Department of Geography, Mount Scopus Campus, Jerusalem 91905, Israel,
[email protected] Berkowicz, S.M. Arid Ecosystems Research Center, Hebrew University of Jerusalem, Giv’at Ram, Jerusalem 91904, Israel,
[email protected] Beyer, L. Institute of Plant Nutrition & Soil Science, Christian Albrechts Universität, 24105 Kiel, Germany Blumberg, D.G. Ben-Gurion University of the Negev, Department of Geography and Environmental Development, Beer-Sheva, Israel,
[email protected] Blume, H.-P. University of Kiel, Schlieffenallee 28, 24105 Kiel, Germany,
[email protected] Böhme, F. Helmholtz-Center for Environmental Research UFZ Leipzig-Halle, Department of Soil Science, Theodor-Lieser-Str. 4, 06120 Halle, Germany Bornkamm, R. Technical University Berlin, Institute for Ecology and Biology, 12165 Berlin, Germany,
[email protected]
xxi
xxii
Contributors
Breckle, S.-W. University of Bielefeld, Department of Ecology, Wasserfuhr 24-26, 33619 Bielefeld, Germany,
[email protected] Büdel, B. University of Kaiserslautern, Institute of Botany, 67653 Kaiserslautern, Germany,
[email protected] Felix-Henningsen, P. Justus-Liebig-University, Institute of Soil Science and Soil Conservation, 35392 Giessen, Germany,
[email protected] Filser, J. University of Bremen, UFT, Department of General and Theoretical Ecology, Leobener Str., 28359 Bremen, Germany,
[email protected] Heusinkveld, B.G. Wageningen University, Meteorology and Air Quality, 6700 AA Wageningen, The Netherlands,
[email protected] Hill, J. University of Trier, Faculty VI, Geography/Earth Sciences, Department of Remote Sensing, 54286 Trier, Germany,
[email protected] Jacobs, A.F.G. Wageningen University, Meteorology and Air Quality, Droevendaalsesteeg 4, Atlasgebouw, P.O. Box 47, 6700 AA Wageningen, The Netherlands,
[email protected] Jarmer, T. Free University of Berlin, Physical Geography, Malteserstr. 74-100, 12249 Berlin, Germany,
[email protected] Kadmon, R. Hebrew University of Jerusalem, Department of Evolution, Systematics & Ecology, Institute of Life Sciences, Givat-Ram, Jerusalem 91904, Israel,
[email protected] Kidron, G.J. Hebrew University of Jerusalem, Institute of Earth Sciences, Givat Ram Campus, Jerusalem 91904, Israel,
[email protected] Leschner, H. Hebrew University of Jerusalem, Department of Evolution, Systematics & Ecology, The Herbarium, Givat-Ram, Jerusalem 91904, Israel,
[email protected] Littmann, T. Martin-Luther-University of Halle-Wittenberg, Institute for Geoscience, 06120 Halle (Saale), Germany; DLC Dr. Littmann Consulting, Leibnizstr. 33, 58256 Ennepetal, Germany,
[email protected]
Contributors
xxiii
Pfisterer, U. Institute of Plant Nutrition & Soil Science, Christian Albrechts Universität, 24105 Kiel, Germany Prasse, R. Leibnitz-University Hannover, Institute for Environmental Planning, 30419 Hannover,
[email protected] Rummel, B. Justus-Liebig-University, Institute of Soil Science and Soil Conservation, 35392 Giessen, Germany,
[email protected] Russow, R. Helmholtz-Center for Environmental Research UFZ Leipzig-Halle, Department of Soil Science, 06120 Halle, Germany,
[email protected] Sartorius, U. Department of Ecology, University of Bielefeld, 33619 Bielefeld, Germany Schultz, A. Martin-Luther-University of Halle-Wittenberg, Institute for Geoscience, 06120 Halle (Saale), Germany Sommer, C. Department of Ecology, University of Bielefeld, 33619 Bielefeld, Germany Tielbörger, K. University of Tübingen, Botanical Institute, Department of Plant Ecology, 72076 Tübingen, Germany,
[email protected] Tsoar, H. Ben-Gurion University of the Negev, Department of Geography and Environmental Development, Beer-Sheva, Israel,
[email protected] Udelhoven, T. CRP-Gabriel Lippmann, Département ‘Environnement et Agro-biotechnologies’ Geomatic Platform, 41 rue du Brill, 4422 Belvaux, GD Luxembourg,
[email protected] Veste, M. University of Hohenheim, Institute of Botany, Experimental Botany, 70599 Stuttgart, Germany,
[email protected] Wenkart, R. Ben-Gurion University of the Negev, Department of Geography and Environmental Development, Beer-Sheva, Israel,
[email protected] Yair, A. Hebrew University, Department of Geography, Mount Scopus Campus, Jerusalem 91905, Israel,
[email protected]
Abbreviations
Ψmin Ψpd Ψsoil ΨW a.g.l. a.s.l. AERC BMBF BNF b.p. BREB CEC Corg CS DP EC ENSO ET FS Ge GPS GTZ HSS K ka LAI LD LSS Ma MPa MSS N1
minimum diurnal leaf water potential (Chap. 25) pre-dawn water potential (Chap. 25) soil water potential (Chap. 25) plant water potential (Chap. 25) above ground level above sea level Arid Ecosystem Research Centre German Federal Ministry of Education and Research biological nitrogen fixation (Chap. 22) before present Bowen ratio energy balance (Chap. 13) cation exchange capacity (Chap. 5) organic carbon (Chap. 16) coarse sand (0.25–2 mm Ø) (Chap. 5) drift potential (Chap. 3) electric conductivity El Nino/Southern Oscillation (Chap. 4) evapotranspiration (Chap. 13) fine sand (0.063–0.25 mm Ø) (Chap. 5) Gevulot geographical positioning system (Chap.11) German Association for Technological Cooperation highly soluble salts (Chap. 16) Kelvin kilo years (1,000 years) leaf area index (m2 m−2) bulk density (Chap. 5) less soluble salts (Chap. 16) million years mega-Pascal (=10 at) Landsat-multispectral scanner (Chap. 11) Nizzana site (see Fig. 29.1)
xxv
xxvi
N3 NdfA NdfS Nt OSL P PCA PET PPFD PV RI RNE SRL TC TDS TL dating TON UNCCD UTC VLD vpd WC Ye
Abbreviations
Haluza Station nitrogen derived from the atmosphere (Chap. 22) nitrogen derived from the soil (Chap. 22) total nitrogen (Chap. 5) optically stimulated luminescence (Chap. 1) mean annual precipitation principal component analysis (Chaps. 19, 22) potential evapotranspiration photosynthetic photon flux density pore volume (Chap. 5) redness index (Chap. 3) relative neighbour effect (Chap. 27) Spaceborne Radar Laboratory (Chap. 3) total amount of carbon (Chap. 16) total amount of dissolved salts (Chap. 16) thermo-luminescence (Chap. 1) total amount of organic nitrogen (Chap. 16) United Nations Conventions to Combat Desertification (Chap. 30) Coordinated Universal Time vegetated linear dunes (Chap. 3) vapour pressure deficit water capacity (Chap. 5) Yevul
General Introduction – Desert Sand Dunes and Aims of the Book – Special Characteristics of the Nizzana Research Site S.-W. Breckle, A. Yair, and M. Veste
Sand dunes occur in many parts of the world, not only in deserts and other arid regions but also along many coastlines in humid biomes and environments. Sand dune formation requires a large supply of sand, strong winds and limited vegetation cover. This has been the case during glacial times in areas that today are humid regions and exhibit fossil dune systems densely covered with vegetation, only open after human destruction. Active inland sand dunes are commonly widespread in arid and dry regions. Deserts are often misinterpreted as being always of sand dunes. However, about only 25 to 30% of deserts (depending on literature sources) are covered with sand fields (erg). The main deserts of the globe are listed in Table 1, as well as the percentages of sand-covered areas within. A variety of desert types have been described in the literature. On the basis of surface properties, the following main types have been identified: rocky and block surfaces (hamada); gravely surfaces (serir); stone pavement surfaces (reg); clay surfaces (takyr); and saline surfaces (sebkha, playa or salina). All these types commonly occur along catenas, often rather mixed (Breckle 2002). From an ecological point of view, sand deserts offer more favourable conditions for plant cover and species diversity than do other desert types. This is due to the specific characteristics of sand, namely low water absorption; rapid infiltration rate and low evaporation losses (because capillary threads are only 10 to 20 cm long). Thus, all sandy areas represent – even in arid deserts – water-storing bodies. This is rather old ecological knowledge (e.g. Walter 1960). The only limitation for plant establishment and plant cover is a high frequency of extreme wind speeds that limit surface stability. In areas where the frequency of extreme wind speeds is low, the sand surface is relatively stable. Under such conditions, the establishment of crusts and a plant cover takes place, leading to an increased surface stability by further reducing wind speed and sand mobility. The process of surface stabilization is enhanced by the deposition of fine-grained particles and the development of biological topsoil crusts. There is today an increasing awareness of the very important role that should be attributed to biological topsoil biological crusts in the structure and functioning of arid ecosystems (see Belnap and Lange 2001). Biological crusts are composed primarily of cyanobacteria; these are important in nitrogen fixation, nutrient cycling, surface stabilization and germination. Where present, biological topsoil crusts strongly control the soil moisture regime, S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
1
2 Table 1 Major drylands/deserts of the world (extracted from various sources, average values; zonobiomes, ZB, and zono-ecotones, ZE, are according to Breckle 2002) Name
Type of desert
Surface area (106 km2)
Sand desert area (103 km2) Location
Sahara (incl. Egyptian Desert E of Nile)
Subtropical, ZBIII
9.25
2,750
Arabian Desert
Subtropical, ZBIII
2.59
620
Australian deserts (Great Subtropical, ZBIII, ZEII-III 1.38 Victoria, Great Sandy, Gibson, Simpson Sturt, Stewart) Gobi Cold winter, ZBVIIa, ZBVII(rIII) 1.33
400
Patagonia
Cold winter, ZBVIIa
0.67
0
Kalahari
Subtropical, ZBIII, ZBII
0.57
400
Great Basin
Cold winter, ZBVIIa
0.49
15
Thar Chihuahua
Subtropical, ZEII-III Subtropical, ZBIII, ZBII(rIII)
0.45 0.44
45 9
200
Description (including percentage of sand desert)
Northern Africa
S.-W. Breckle et al.
70% gravel, rock plains. Contrary to popular belief, the desert is only less than 30% sand (several erg-fields) Arabian Gravel plains, rocky highlands; one Peninsula quarter is the Rub al-Khali (“Empty Quarter”), the world’s largest expanse of unbroken sand (25%) Australia Sand hills, gravel, rocks, grassland, Simpson parallel sand dunes are the longest in the world: up to 200 km (30%) China, Mongolia Stony, sandy soil (15%), steppes (dry grasslands) Argentina Gravel plains, plateaus, basalt sheets (0%) South Africa, Sand sheets, longitudinal dunes (70%) Botswana, Namibia USA Mountain ridges, valleys, sand dunes (3%) India, Pakistan Rocky sand and sand dunes (10%) Mexico Grassland, cacti savannah (2%)
Cold winter, ZBVII(rIII)
0.36
290
Iranian deserts (Registan) Colorado Plateau
Cold winter, ZBVII, ZBIII Cold winter, ZBVIIa
0.35 0.34
35 0
Sonora Kyzyl-Kum
Subtropical, ZBIII, ZEII-III Cold winter, ZBVII(rIII)
0.31 0.30
15 240
Atacama (Altiplano)
Cool coastal, ZBIII
0.18
20
Mojave
Subtropical, ZBVIIa
0.14
15
Aralkum (new desert)
ZBVII(rIII)
0.055
10
Sinai (part of Sahara) Namib
ZBIII Cool coastal, ZBIII
0.060 0.034
3 15
Negev (part of Sahara)
ZBIII
0.013
1
China
Sand dunes (80%), up to 300 m high; gravel Iran, Afghanistan Salt, gravel, rock, sand fields (10%) USA Sedimentary rock, mesas and plateaus – the Grand Canyon, “Painted Desert” (0%) USA, Mexico Cacti savannah, gravel (5%) Uzbekistan, Sands, rock – name means “red sand” Turkmenistan, (80%) Kazakhstan Chile, Peru, Salt basins, sand (10%), lava; world’s Bolivia driest desert, mountains USA Mountain chains, dry alkaline lake beds, calcium carbonate dunes (12%) Kazakstan, Desiccated seafloor, sand desert (20%) Uzbekistan Egypt Mountains, rocks, gravel, sand (5%) Angola, Namibia, Gravel plains, huge sand dunes (50%), South Africa up to 300 m high Israel Rocks, gravel, sand (5%)
General Introduction – Desert Sand Dunes and Aims of the Book
Taklamakan
3
4
S.-W. Breckle et al.
as they affect infiltration rate, surface runoff generation, spatial redistribution of water resources and depth of water penetration. The Nizzana Research Site is one of the long-term research sites operated by the Arid Ecosystem Research Centre (AERC) established in 1987 by the Minerva Foundation (Germany) and the Hebrew University of Jerusalem. The centre provided technical and partial financial support to many of the studies included in this book. The site is located at the proximity of the Israeli–Egyptian border (Fig. 1) where, at present, human activity is quite limited. The Nizzana Sands site offers an excellent example of the structure and functioning of a sandy desert ecosystem and the importance of biological topsoil crusts. Any sandy area has its own characteristics that derive from its particular geographic location, local rainfall and wind regimes, sand grain mineral composition and geomorphic features of the dune system. The special characteristics of the sandy area where the Nizzana site is located are as follows: ●
●
The sand field on the Israeli side represents the eastern margins of the extensive Sinai erg (see Chaps. 1–3). It is climatologically arid (11 months relatively arid, Fig. 2) with cool, moist winters and hot, dry summers (Fig. 2). The rainfall gradient along the sandy study sites (Fig. 1: sites N1 to N5) varies sharply from less than 90 mm in the south to about 170 mm in the north, over a distance of only 35 km. Annual rainfall fluctuates greatly from year to year all over the area. In the south (N1), annual rainfall is between 28 and 160 mm. Most rainstorms are small (below 5 mm), with a predominance of low rain intensities. Due to the proximity of the area to the Mediterranean Sea, dewfall is quite frequent (see Chaps. 4, 12).
Fig. 1 The Nizzana Research Station (rectangle) and locations of the Nizzana monitoring sites (N1–N5) and other research sites (Yevul, Gevulot). Precipitation was recorded at Qadesh Barnea
General Introduction – Desert Sand Dunes and Aims of the Book
5
Fig. 2 Ecological climate diagram for the Nizzana Research Station (modified after Walter and Lieth 1967, Breckle 2002), showing temperature and rainfall data for the ca. last 19 years
●
●
●
●
●
●
●
The wind regime is characterized by a high frequency of low wind speeds that result in the frequent deposition of fine-grained particles that play an important role in surface stabilization and an extensive development of biological topsoil crusts (see Chaps. 6, 10, 11, 14, 15). The extent of stable areas of topsoil crust cover with specific crust properties increases with increasing annual rainfall from south to north. Despite the arid climatic conditions prevailing in the area, an extensive vegetation cover is observed that, locally at the base of dune slopes, can reach the high values of 80–100% (see Chaps. 6, 8, 18, 20, 23, 26). On a smaller scale, there is a remarkable disturbance gradient between interdunes and dune tops, with high spatial-temporal dynamics of recruitment and mortality, and varying competitive and facilitation processes (Malkinson and Kadmon 2007). It is remarkable that the biodiversity (of spermatophytes) in such a dune mosaic system is relatively high (see Chap. 7); however, this depends on the definition of biodiversity. Area-wise, species numbers may be rather low in comparison with tropical regions but the species numbers related to resources such as water availability (resource-related biodiversity) are in the same range as those in the tropics (Breckle 2006). The development of such a “fertile” ecosystem is explained by the important role that should be attributed to the biological topsoil crusts that increase surface stabilization, improve nutrient cycling, control the water regime and water resources, and exercise a strong influence on the germination and establishment of higher vegetation (see Chaps. 18, 19, 20, 23, 25, 26). The area had been subjected to human activity, mainly grazing by Bedouins, during the period 1967–1982. Human activity ceased after 1982, enabling the investigation of recovery processes of the disturbed topsoil crust and grazed perennial vegetation (see Chaps. 6, 20, 27, 28).
6
S.-W. Breckle et al.
●
The marked difference in average annual rainfall from the south to the north (90–170 mm; cf. above) allows studying the possible effects of the foreseen global climatic change over an area (Fig. 1) where the sandy substratum is almost uniform (see Chap. 29).
The detailed studies conducted in the area at different temporal and spatial scales, as well as the interdisciplinary approach adopted in most studies, represent an interesting holistic case study that can more or less directly be applied to other sandy arid areas, while taking into consideration the specific climatic and geomorphic conditions prevailing there. The book is divided into four main sections. The first section provides an overview of the regional physical characteristics of the area and covers geological, pedological, geomorphological and climatological aspects, as well as desertification processes by land use. The second section focuses on the spatial patterns of the vegetation and topsoil crust covers. The third section covers the numerous studies dealing with ecosystem processes such as sand movement, evaporation and transpiration, runoff generation and water resources, recovery of the vegetation and of the biological crust following disturbance, photosynthesis, dewfall, activity of biological crusts, nitrogen input, demography of annual plants, etc. The last section presents a synthesis of most of the work presented in the book, and focuses on the important issue of specific surface properties, in particular regarding the sensitivity of the area to climate change as well as the rehabilitation measures available to date for desertified sand dune systems.
References Belnap J, Lange OL (eds) (2001) Biological soil crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York Breckle S-W (ed) (2002) Walter’s vegetation of the earth. The Ecological Systems of the Geo-Biosphere. Springer, Berlin Heidelberg New York Breckle S-W (2006) Biodiversity in deserts – is it area- or resource-related? J Arid Land Stud 16:61–74 Malkinson D, Kadmon R (2007) Vegetation dynamics along a disturbance gradient: spatial and temporal perspectives. J Arid Environ 69:127–143 Walter H (1960) Einführung in die Phytologie III, Grundlagen der Pflanzenverbreitung. Teil I. Standortslehre. Ulmer, Stuttgart Walter H, Lieth H (1967) Klimadiagramm-Weltatlas. Fischer, Jena
Part A
The North-Western Negev
Chapter 1
Geological Background of the Nizzana Area R. Ben-David and A. Yair
1.1
Introduction
The main purpose of this chapter is to briefly review the recent geological history of the western Negev desert, focusing on the different stages of sand incursion during the Quaternary era and their effects on the drainage system. The Nizzana watershed represents the major system draining the Negev Highlands towards the Sinai Peninsula in the northwest. It extends over 805 km2. The two main tributaries, in the proximity of the sandy area, are the Lavan and Shunra basins (Fig. 1.1), extending over 329 and 220 km2 respectively. From a physiographic point of view, the area is constituted by three distinct belts trending SW–NE (Fig. 1.1). 1. The rocky belt. This belt forms the upper reaches of the Nizzana basin (Negev Heights, Matred-Avdat Plateau and Qetef Shivta), composed of extensive rocky outcrops of Cenomanian to Eocene formations. 2. The alluvial belt. The western margins of the rocky areas are represented by a large belt of alluvial fans overlying marine sediments attributed to the MiocenePliocene period. The assumed location of the Pliocene shoreline coincides with the western limit of Mt Qeren (Fig. 1.1). 3. The sandy belt. The Haluza-Agur sand field forms the third unit, dated as Upper Pleistocene. The Mt Qeren anticlinal (Fig. 1.1) separates the alluvial belt from the Haluza-Agur sand field, which represents the eastern sector of the extensive Sinai erg in Egypt.
1.2 1.2.1
Late Geological History Tertiary Era
The present landscape is the result of a multistage evolution following the development of an extensive peneplain during the Oligocene-Miocene (Issar 1961; Zilberman 1992). Uplifting of the area took place at the end of the Miocene and during the S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
9
10
R. Ben-David, A. Yair
Fig. 1.1 Location and physiographic map
Pliocene, resulting in the deposition of coarse sediments on top of the peneplain surface. The location of the Pliocene shoreline (Greenbaum and Ben-David 2001) has been derived from sedimentological evidence gathered from the channels of Nahal Besor and Nahal Shunra (Fig. 1.1) as well as from boreholes in the Nizzana and Lavan channels, where late Pleistocene gravely terrestrial sediments overly marine Miocene-Pliocene marls. These latter sediments are indicated as unit 1 in Table 1.1. This sharp unconformity in the geological record marks a sedimentary gap of ~5–2 Ma. The oxygen stage 6 curve, when the Mediterranean sea level decreased by ~120 m, may explain this gap (Imbrie et al. 1984; Chappell and Schakelton 1986; Horowitz 1988).
1 Geological Background of the Nizzana Area
11
Table 1.1 Schematic stratigraphic sequence Period
Stratigraphic unit
Holocene
Unit 5
Late Unit 4 Pleistocene
Thickness (m)
6–30
Unit 3
Pliocene
1.2.2
Unit 2
3–16
Unit 1
>10
Description Erosional phase: gradual incision of drainage network and breaching of sand barriers Main phase of sand penetration and building of the present-day dune system. Blocking of channel network by advancing sand ridges. Age: 25–10 ka B.P. Decrease in flow energy, accompanied by loess and sand deposition. Extensive development of floodplains. Age: 80–50 ka B.P. Fluvial sediments: large to small pebbles. Sometimes well cemented. Age: 100–150 ka B.P. Marine deposits: marls, silty clays, fine gravels, fauna
Quaternary Era
Study of the Quaternary era was based on the description and analysis of numerous stratigraphic sections along the banks of the Nizzana and Lavan channels, on data from 25 boreholes to a depth of 28 m at selected locations, and on 14 isotopic ages of sediment samples. The late Pleistocene sediments are 6 to 30 m thick. Up to four distinct units can be observed, composed of coarse clastic deposits, aeolian sands and loess. Unit 2. This unit, 3 to 16 m thick, was deposited on top of the marine sediments (Table 1.1). It is composed of fluvial sediments, where coarse- and fine-grained particles alternate. Cut and fill features have been detected. The age of this unit (based on TL dating) is 150–100 ka B.P. This depositional phase was followed by an incision phase that ended at about 80 ka B.P. Unit 3. The overlying layer (80–50 ka B.P.) is indicative of a substantial lowering of flow energy. It is characterized by a gradual decrease in the amount of coarse gravely material, accompanied by a parallel increase in the silt and sand components. This phase marks the beginning of loess and sand penetration into the Negev desert. This phase was interpreted by Zilberman (1992) as a period of extensive development of the Nizzana-Lavan floodplains, which preceded that of the presentday dune system. The age of this unit is also indicated by late Palaeolithic prehistoric sites found within the loess deposits (Goring-Morris and Goldberg 1990). This phase ended by the incision of the drainage network, from 30 to 25 ka B.P. Unit 4. The period 25–10 ka B.P. represents the main phase of sand incursion and building of the present-day dune system, characterized by west-east-trending longitudinal dunes separated by large corridors. Boreholes at the crest of a dune and at an adjoining interdune corridor clearly indicate a non-uniform sedimentary sequence. Sand accumulation along the dune axis is up to 30 m thick, deposited on top of the formers units. The depositional sequence in the interdune corridors is thinner and
12
R. Ben-David, A. Yair
characterized by alternating, stratified, fine-grained sediments sometimes mixed with fine gravels. The difference in the sedimentary sequence between the interdune corridors and dune crests is explained by the fact that, during this main phase of sand penetration, all channels of the Nizzana drainage system were blocked by the rapidly advancing sand ridges. The Nizzana channel was blocked at several locations, by each of the advancing sand ridges. Upon reaching a sand barrier, floods in the Nizzana channel were diverted into the interdune corridors, resulting in the deposition of widespread horizontal, fine-grained sediments (Yair 1991; Harrison and Yair 1998). Remnants of a dune barrier are exposed on the east bank of Nahal Nizzana (Fig. 1.2). TL dates imply that the Nizzana channel was blocked before 19 ka B.P., and that the dam across the channel was breached in the early Holocene. The thickness of the sediments deposited behind the sand barriers is not spatially uniform. Deposits of the Nizzana channel in seasonal lakes within the interdune corridors do not exceed 6 m. Much thicker, laminated lake deposits (up to 20 m) are observed behind the sand barrier that blocked the channel of Nahal Lavan. Differences in the thickness of the lake deposits are due mainly to fluctuations in the time span between the blocking of the channel and the incision of the sand barrier. Unit 5. This phase is characterized by the incision of the sand barriers by the drainage network. TL and OSL dating of sediments that have accumulated behind the sand barriers clearly shows that the breaching of the sand barriers by the Nizzana channel and its tributaries did not occur at the same time. As indicated above, the breaching of the sand barrier by the Nizzana channel (Fig. 1.2) took place in the early Holocene.
Fig. 1.2 Lake deposits behind the Nizzana channel dune barrier
1 Geological Background of the Nizzana Area
13
Fig. 1.3 Nahal Lavan lake deposits and isotopic ages
Nahal Lavan, which drains an area of 329 km2, incised the sand barrier only some 1,000 years ago (Fig. 1.3), while Nahal Shunra (220 km2) is still blocked by a sand barrier. The very late incision of Nahal Lavan may explain the thickness of the lake deposits behind the sand barrier (see Chap. 2, this volume).
1.3
Discussion of Results
At least two main factors may explain the non-uniform timing of incision of the sand barriers by the drainage network. The first is the dimensions of the drainage basins. It is obvious that the incision of a sand barrier requires an extreme flow event, or a sequence of extreme events. Under such conditions, the larger the basin, the better are the chances to generate extreme flow discharges capable of incising a sand barrier. The second important factor relates to the surface properties of the drainage basin that control the frequency and magnitude of runoff generation. Several studies conducted in the northern Negev desert (Yair and Lavee 1985; Yair and Kossovsky 2002; Yair and Raz-Yassif 2004) clearly show that the frequency and magnitude of runoff generation in arid areas are strongly controlled by the ratio of rocky surfaces to soil-covered surfaces. Bare rocky areas, devoid of soil cover and with a limited vegetation cover, develop runoff very quickly. The rain threshold for runoff generation over such areas is as low as 2–3 mm. Different conditions characterize soil-covered areas. Here, rain threshold for runoff generation is about 10 mm. Differences in infiltration and absorption rates between rocky and soil-covered areas are also
14
R. Ben-David, A. Yair
Fig. 1.4 Percentages of rocky, sandy and alluvial areas in the Nizzana, Lavan and Shunra watersheds
responsible for frequent discontinuities in runoff. Most of the runoff generated over rocky areas is immediately absorbed on reaching the soil-covered areas. This process is greatly enhanced by the limited duration of most rain showers prevailing in the area (Yair and Kossovsky 2002; Yair and Raz-Yassif 2004). An analysis of the ratio of rocky to soil-covered areas for the three watersheds mentioned above is shown in Fig. 1.4. The data satisfactorily explain the sequence of incision of the sand barrier. Nahal Nizzana, with the highest rock/soil ratio, was the first to overcome the sand barriers, followed by Nahal Lavan. Nahal Shunra, with extensive sandy areas and the lowest rock/soil ratio, is still blocked by a sand barrier.
1.4
Conclusions
The limited number of isotopic dates and the sedimentary data available to date allow us to draw two main conclusions: 1. It is commonly assumed that sand movement and the building of dune systems are correlated with dry climatic phases. The explanation proposed is that under dry conditions, the effect of the factors that limit sand movement, such as topsoil crusts and vegetation cover, is greatly reduced. The bare sand is therefore easy to detach and transport. The transition from the late Quaternary to the Holocene
1 Geological Background of the Nizzana Area
15
period is widely regarded as one from a wet to a dry climatic phase. On the basis of the arguments presented above, such a transition would be expected to increase sand movement and the building of dunes. However, the data obtained show a reversed evolution. The main phase of dune building and obstruction of the drainage network occurred before the Holocene (25–10 ka B.P.). Furthermore, the breaching of most sand barriers took place during the Holocene. 2. The above discussion leads us to the second conclusion. The fact that the breaching of the sand dams can not be attributed to climatic fluctuations draws attention to the important role that should be attributed to the local surface properties of adjoining basins in the process of breaching sand barriers. It appears that under the rainfall regime prevailing in the area, characterized by a high frequency of low-intensity rain events, the temporal sequence of channel incision across the sand dams will be controlled largely by the hydrological regime of each basin. Therefore, under a given regional climatic regime, surface properties may be expected to play a determinant role in the timing of channel incision. Acknowledgements The study was supported by the Minerva Arid Ecosystems Research Centre of the Hebrew University of Jerusalem. We gratefully thank Mrs. M. Kidron of the Department of Geography for drawing the illustrations.
References Chappell J, Shackelton NJ (1986) Oxygen isotopes and sea level. Nature 324:137–140 Goring-Morris N, Goldberg P (1990) Late Quaternary dune migration in the southern Levant: archeology, chronology and paleo-environments. Quat Int 5:115–137 Greenbaum N, Ben-David R (2001) Geological-geomorphological mapping in the Shivta-Rogem site area. Data Rep no 3, Israel Electrical Company Harrison JB, Yair A (1998) Late Pleistocene aeolian and fluvial interactions in the development of the Nizzana dune field. Sedimentology 45:507–518 Horowitz A (1988) The Quaternary environments and paleogeography in Israel. In: Yom-Tov I, Tchernov E (eds) The zoogeography of Israel. Junk, Dordrecht, pp 35–57 Imbrie J, Hays JD, McIntyre A, Mix AC, Morley JJ, Pisia NG, Prell WL, Schackelton NG (1984) The orbital theory of Pleistocene climate: support from a revised chronology of marine D 18 record. In: Berger A, Imbrie J, Hays H, Kukla G, Saltsman B (eds) Milankovitch and climate, part 1. Reidel, Boston, MA, pp 269–305 Issar A, Tsoar H, Gilad I, Zangvil A (1987) A paleoclimatic model to explain depositional environments during the Late Pleistocene in the Negev. In: Bekofsky L, Wurtele MG (eds) Progress in desert research. Rowman and Littlefield, Totowa, NJ, pp 302–309 Yair A (1990) Runoff generation in a sandy area – the Nizzana sands, western Negev, Israel. Earth Surface Processes Landforms 15:597–609 Yair A, Kossovsky A (2002) Climate and surface properties: hydrological response of small arid and semi-arid watersheds. Geomorphology 42:43–57 Yair A, Lavee H (1985) Runoff generation in semiarid and arid areas. In: Anderson MG, Burt TP (eds) Hydrological forecasting. Wiley, New York, 183–220 Yair A, Raz-Yassif N (2004) Hydrological processes in a small arid catchment. Scale effects of rainfall and slope length. Geomorphology 61:155–169 Zilberman E (1992) The Late Pleistocene sequence of the northwestern Negev flood plains: a key to reconstructing the paleoclimate of southern Israel in the last glacial. Israel J Earth Sci 41:155–167
Chapter 2
Geo-Ecology of the North-Western Negev Sand Field A. Yair, M. Veste, and S.-W. Breckle
2.1
Introduction
The north-western Negev sand field represents the eastern extension of the extensive Sinai continental erg. It can be subdivided into several distinct units. The dry riverbed of the Nizzana channel separates the Haluza-Agur sand field north of the channel from the Hallamish-Shunra sand fields south of the channel (Fig. 2.1). The Nizzana research site is located in the proximity of the Egyptian-Israeli borderline in the southern Hallamish sandy area. The sand ridges in the area trend W–E and are considered by Tsoar et al. (see Chap. 3, this volume) as vegetated stabilized linear dunes (Fig. 2.2). The dunes are up to 18 m high, with an average of around 8.5 m (Allgaier 1993). The relative height of the dune increases from north to south. The area is characterized by a sharp rainfall gradient. Average annual rainfall varies from approx. 170 mm in the north to approx. 90 mm in the south along a distance of 35 km. The rainy season is limited to the winter months, extending from October to May. Mean monthly temperatures vary from 9 °C in January to 27 °C in August (see Chap. 4, this volume). The northern sandy area is classified as arid, the southern area as hyper-arid. The whole sandy area is characterized by a low wind energy (see Chaps. 3 and 4, this volume). The prevalence of weak winds explains many of the special and important properties of this sandy ecosystem, such as the high stability of very large areas caused by the extensive development of biological topsoil crusts, the relatively high content of fine-grained particles in the topsoil crust, as well as the high vegetation cover (∼30% over the crusted areas) indicative of a relatively good water regime despite the arid and hyper-arid climatic conditions. The description and analysis of the various geo-ecological units will focus on the units identified in the southern Hallamish sand field where the Nizzana Arid Ecosystems Research Site is located. Numerous studies on various physical and biological aspects have been conducted at this site during the period 1989–2004, enabling an advanced understanding of the structure and functioning of this sandy ecosystem under wet and dry rainfall years.
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
17
18
Fig. 2.1 Location map and physiography of the area
Fig. 2.2 General view of the linear vegetated dunes
A. Yair et al.
2 Geo-Ecology of the North-Western Negev Sand Field
2.2
19
Geo-Ecological Units in the Hallamish Sand Field
The Hallamish sand field occupies the sandy area south of the riverbed of the ephemeral Nizzana channel. This area differs from the Haluza-Agur sandy area in two aspects. 1. The relative elevation of large sections of dune ridges is the highest in the study area, and large parts of the sandy ridges are still active. 2. Flat fluvial sediments deposited by the Nizzana channel occur in the interdune corridors. A transect orthogonal to the dune ridges reveals the following geoecological units (Figs. 2.3, 2.4).
Fig. 2.3 Geo-ecological units across the dune system
Fig. 2.4 View of the active dune crest and of the crusted-vegetated dune slope
20
2.2.1
A. Yair et al.
The Sandy Ridge
The sandy ridge can be divided into two subunits: the active crest and the dune slope (Fig. 2.3). The Active Dune Crest The dune crest is composed of unconsolidated coarse to medium quartzitic sand, with an extremely low amount of fine-grained particles. The sand fraction contains up to 5% carbonates in the form of terrestrial shell fragments from snails. Silt and clay contents increase slightly with depth but never exceed 6%. The silt fraction is composed predominantly of calcite (35–45%), followed by quartz (30–40%) and some dolomite and feldspars. The clay fraction contains primarily montmorillonite with low amounts of kaolinite and illite (Amit and Harrison 1995). Well-developed ripple marks are clearly indicative that this part is very active and unstable under present-day conditions. Cross bedding can be observed. Exceptionally strong winds may reduce the height of the crest by tens of centimetres, while low winds are responsible for sand deposition on the crest (cf. more details in Chaps. 4 and 14, this volume). The crest is often sharp-edged, as in seifs. Blowouts are quite frequent. The active crest is completely devoid of a biological topsoil crust. Total porosity is about 35%. Large pores of 50–100 µm diameter are predominant. Smaller pores, less than 20 µm, account for 10–15% of the voids. Infiltration rates exceed 60 mm h −1 and all rainwater, during all rainstorms, infiltrates into the sand. The salt content is low and does not change significantly with depth. This unit is poor in nutrients. The crests of the dunes are characterized by a sparse vegetation cover (5–15%) dominated by Stipagrostis scoparia, Heliotropium digynum and Cornulaca monocantha, well adapted to burial or exposure of their root system in this very unstable and active area (Danin 1996). The Dune Slopes The dune slopes extend from the bottom of the active crest to the dune base. An important and specific character of the dune slopes in the Agur-Haluza sand field is the extensive occurrence of a topsoil crust that plays a very important role in the structure and functioning of this sandy ecosystem. The topsoil crust is composed of mineral and biological components (see Chap. 10, this volume). It is richer in fine-grained particles, organic matter content and nutrients than is the sterile sand. The crust affects the degree of surface stability, infiltration, and surface runoff rates and, consequently, is associated with
2 Geo-Ecology of the North-Western Negev Sand Field
21
spatial differences in water resources. The biological elements of the crust are mainly cyanobacteria, green algae and fungi (see Chap. 10, this volume). In areas where wet conditions prevail, such as at the base of north-facing slopes or around shrubs, mosses of the genus Bryum and Brachythecium (Lange et al. 1992) are also present. The crust is better developed and more extensive on north- than on southfacing slopes. The crusted area extends over two thirds of the dune slopes on north-facing slopes but is limited to the dune base on south-facing slopes. Differences in crust characteristics are more pronounced on the former slopes. In the upper part of crusted north-facing slopes, the crust covers 40–60% of the area; it is thin, patchy and very friable. Sand deposition on the crust occurs during all significant wind storms. The crust here is therefore very rich in sand particles, with a pale yellowish-greenish colour, of limited stability and with a high infiltration rate. Depth of water infiltration may reach 4–5 m in wet years (Yair et al. 1997). Chlorophyll a content is 15–20 mg m−2. The thickness and spatial continuity of the crust increase gradually towards the dune base. In the lower slope section, the crust is dark-coloured when dry and very greenish when wet. It covers up to 90% of the area, is 2–3 mm thick, quite resistant, and rich in silt and clay (contents of 20–30%). Chlorophyll a content is 20–40 mg m−2. The crust on the south-facing slope is thinner and rich in sand, with a chlorophyll content of 15–20 mg m−2 (Kidron and Yair 1997). The soils are young and weakly developed. However, secondary calcic precipitation of biogenic origin by fungi and bacteria has been observed along roots and borrows in the root zone (Amit and Harrison 1995). The vegetation cover along the crusted north-facing slope is 10–30% but is limited to 10–15% on the opposite, south-facing slope. In the active dune sections, sand mobility and surface instability are major factors determining plant species composition (Kadmon and Leschner 1995; Danin 1996). The dominant perennial plant species along the stabilized north-facing slopes are Moltkiopsis ciliata, Retama raetam, Artemisia monosperma, Convolvulus lanata and Noaea mucronata. Characteristic species for mobile south-facing slopes are S. scoparia, H. digynum and C. monocantha.
The Dune Base The base of north-facing slopes is characterized by a narrow belt with dense vegetation cover, in local concavities reaching up to 100%. The crust is 2–3 mm thick, smooth, and very rich in fine-grained particles (contents up to 45%) and organic matter (contents approx. 4–6%). The high density of the vegetation is due to the good water regime, which results from water concentration at the dune base by three complementary processes: direct rainfall; surface runoff generated over the smooth crusted slope, and shallow subsurface flow (see Chap. 18, this volume).
22
A. Yair et al.
The crust at the slope base is rich in nutrients derived from the mineral and biological components of the crust. Chlorophyll a content is the highest recorded in the whole area, reaching 45–60 mg−2. The crust is also rich in mosses, indicative of a good water regime (Mazor et al. 1996; Veste and Littman 2006). The perennial plant species include A. monosperma, M. ciliata, A. articulata, R. raetam, C. monocantha, Thymelaea hirsuta and N. mucronata.
2.2.2
The Interdune Corridor
Interdune corridors, 100–300 m wide, extend between the parallel dune ridges. The interdune depressions are composed of two main units: nebkhas and flat playa surfaces.
Sandy Nebkhas Scattered coppice dunes or nebkhas cover most of the area. The relative elevation of these sand mounds does not exceed 2–3 m. All are covered with shrubs. The whole area between the shrubs is covered by a thin cyanobacterial crust with a low chlorophyll a content of 15–20 mg m−2, indicative of a dry regime. Runoff does occur on the crusted area. However, due to the rugged micro-topography, flow distances are short, thereby constraining the very positive effect of increased water concentration encountered on the north-facing foot slopes. Water absorption by the mineral and biological components of the crust, as well as the high frequency of rainstorms with low rain amounts reduce the depth of water penetration. Depth of water penetration is limited to 60–90 cm (Yair et al. 1997). At some places, the sandy cover, a few meters thick, is underlain by fined-grained playa sediments. At such sites, local perched water lenses may develop in wet years. Typical perennial shrubs in the interdune depressions include the following species: R. reatam, T. hirsuta, M. ciliata, C. lanata, A. articulata and Echiochilon fruticosum. The latter four and C. monocantha are often found growing on small sandy hills, accumulating sand at their base and thereby forming nebkhas.
Flat Playa Surfaces Playa sediments (takyrs) are scattered in the interdune depressions. They represent fluvial sedimentary surfaces deposited by the Nizzana channel. These depositional
2 Geo-Ecology of the North-Western Negev Sand Field
23
events occurred during the late Quaternary when the north-western course of the channel was blocked by the eastward-advancing sand ridges (Harrison and Yair 1998; Chap 1, this volume). The deposits are characterized by several horizontal sequences of sand–silt and clay layers, indicative of sediment deposition in water bodies. The sediments are well compacted and quite cemented. Rusty mottles at shallow depths of 30–40 cm show that anaerobic conditions with stagnant water occur from time to time. The sediments are relatively rich in carbonates (up to 30%) and saline. At a depth of approx. 40 cm, NaCl is accompanied by CaSO4. EC may exceed the value of 4 mS cm−1 in the fine-grained layers. The soils have been classified as Solonchak and Calcisol by Blume et al. (1995). Due to the compact, fine-grained nature of the sediments, depth of water penetration is limited to 40 cm (Yair et al. 1997). Most of the playa surfaces are devoid of vegetation, except for a few spots where sand has accumulated. The dominant plant species of this unit is the xerohalophyte A. articulata.
2.3
Geo-Ecology of the Haluza-Agur Sand Field
Several important differences exist between the dunes south of the Nizzana channel and north of it. The change from south to north is gradual, except for the occurrence of playas, which are found only south of the Nizzana channel in the interdune corridors. Slope angles and slope lengths decrease from south to north. The extent and continuity of the topsoil crust cover increase northwards, where active crests are quite limited and some dunes are completely covered by the crust. In some interdune corridors, extensive areas are covered by a soil lichen crust composed of Fulgensia fulgens, Squamarina sp., Collema tenax and other cyano-lichens (see Chap. 10, this volume). The thickness of the crust, its organic matter content as well as silt and clay contents increase positively with average annual rainfall, resulting in a very stable sandy area. All these properties lead to a lower spatial variability in edaphic conditions, in contrast to that encountered in the southern Hallamish sandy area. The positive and negative effects of these properties along the rainfall gradient will be discussed in detail in Chapter 30 (this volume).
2.4
Conclusions
This review of geo-ecological units within the sandy study area in the north-western Negev sand field reveals a great variety of edaphic conditions existing at a local scale (from the dune crest into the interdune area) as well as at a regional scale
24
A. Yair et al.
along the rainfall gradient over a distance of 35 km. Many factors contribute to this variability. However, this variability may often be explained by the local predominance of only a single factor. For example, the strong winds, and related surface instability, at the crest of the dunes explain the limited establishment of annual and perennial plants. Differences in the composition of the biological topsoil crust largely control the water regime. Thin cyanobacterial crusts absorb only a limited amount of water, generating high runoff rates leading to water concentration at nearby down-slope positions. Reverse conditions prevail at sites where the crust is thick and rich in mosses and lichens. These elements are able to absorb substantial amounts of water. By so doing, they limit the depth of water penetration and water availability for higher plants. The pronounced spatial variability in edaphic conditions explains the relatively high richness of species, despite the arid to extreme arid conditions. The great diversity in local niches may be regarded as an indication that this sandy ecosystem is quite resilient to climatic changes, due to the fact that local surface properties play an important role in the spatial redistribution of water resources. At the local scale, these properties are sometimes more important than the absolute rain amount.
References Allgaier A (1993) Geomorphologische Untersuchungen an Längsdünen in der westlichen Negev, Israel. Diplome Arbeit Lehrstuhl für Physische Geographie der RWTH, Aachen Amit A, Harrison JBJ (1995) Biogenic calcic horizon development under extremely arid conditions, Nizzana sand dunes, Israel. Adv GeoEcol 28:65–88 Blume HP, Yair A, Yaalon DH (1995) An initial study of pedogenic features along a transect across longitudinal dunes and interdune areas, Nizzana region, Israel. Adv GeoEcol 28:51–64 Danin A (1996) Plants of desert dunes. Springer, Berlin Heidelberg New York Harrison JBJ, Yair A (1998) Late Pleistocene aeolian and fluvial interactions in the development of the Nizzana sand field, Negev desert, Israel. Sedimentology 45:507–518 Kadmon R, Leschner H (1995) Ecology of linear dunes: effects of surface stability on the distribution and abundance of annual plants. Adv GeoEcol 28:125–143 Kidron GJ, Yair A (1997) Rainfall-runoff relationship over encrusted dune surfaces, Nizzana, western Negev, Israel. Earth Surface Processes Landforms 22:1169–1184 Lange OL, Kidron GJ, Budel B, Meyer A, Kilian E, Abeliovitch A (1992) Taxonomic composition and photosynthetic characteristics of the biological soil crust covering sand dunes in the western Negev desert. Funct Ecol 6:519–527 Mazor G, Kidron GJ, Vonshak A, Abeliovitch A (1996) The role of cyanobacterial exopolysaccharides in structuring desert microbial crusts. FEMS Microbial Ecol 21:121–130 Veste M, Littman T (2006) Dewfall and its geo-ecological implication for biological surface crusts in desert sand dunes, North western Negev, Israel. J Arid Land Studies 16(3):139–147 Yair A, Lavee H, Greitzer N (1997) Spatial and temporal variability of water percolation and movement in a system of longitudinal dunes, Western Negev, Israel. Hydrol Processes 11:43–58
Chapter 3
Formation and Geomorphology of the North-Western Negev Sand Dunes H. Tsoar, D.G. Blumberg, and R. Wenkart
3.1
The Sinai-Negev Dunefield
The coastal plain of the northern Sinai Peninsula is a structural depression that borders several anticline mountains in the south. This coastal plain is covered by a dunefield in a wide area of 20–80 km from north to south, and for 260 km from the Nile Delta in the west into the northern Negev Desert, where it terminates south of Beer Sheva (Fig. 3.1). The dunefield covers an area of about 12,000 km2. The Sinai and Negev form one geographical unit subdivided artificially by a political border. The dunefield is located in the northern boundary region of the Eastern Sahara subtropical desert, characterized by a long, hot and dry summer and a cool winter with a mean annual rainfall that is below 200 mm. The political border between the Negev and Sinai has generated two distinctly different landscapes that can be delineated from space-based imagery. The Sinai side of the border tends to be bright and is constituted of bare sand dunes, whereas the Negev side is dark and constituted of vegetated dunes. This political border has thus created a bio-physical border caused by two distinctly different types of land use – grazing and wood-gathering activities in the Sinai, in contrast to almost no human-induced pressure in the Negev (Tsoar, Chap. 6, this volume). The Negev dunefield is triangular in shape, tapering eastwards because of the northern Negev anticline system that stretches from southwest to northeast and delimits the dunes in the southeast, and because of the storm winds blowing in this direction. The anticline of Har Keren is illustrated above the dunefield in Fig. 3.1. Nahal (wadi) Nizzana, which drains the north-western side of the Ramon anticline, reaches to the south-western side of Har Keren but cannot cross the Agur Sands and, therefore, is diverted westwards where it disappears in the Sinai sands and becomes a defunct wadi. A similar defunct wadi is Nahal Shunra that cannot cross the Haluza Sands from south to north, while Nahal Besor does cross the dunes and drains into the Mediterranean south of Gaza (Fig. 3.1). The Negev dunefield can be subdivided into several sections, based on the geological structure, the wadis that cross the area, and the morphology of the sand dunes. The main dunefiled is located in the North-Western part of the area, and is known as the Haluza-Agur dunefield (Fig. 3.1). This dunefield is delimited in the south by
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
25
26
H. Tsoar et al.
Fig. 3.1 Location map of the sand fields in the north-western Negev
Nahal (wadi) Nizzana, which crosses the area from east to west. In the southeast, the Agur Sands are delimited by the slope of the Har Keren anticline. Eastwards, the Haluza-Agur dunefield sand diminishes towards the floodplain of Nahal Besor where the dunes are smaller. A similar diminishment of the dunes occurs in the north where the dune sand gradually transforms into loess. Surrounded by Nahal Nizzana is the Sde Hallamish dunefield, known also as the Nizzana sand field. Another dunefield is located in the Shunra basin (syncline), separated from the Haluza-Agur dunefield by the Har Qeren anticline. Eastwards of Nahal Besor, the sand becomes thicker where the northern Negev dunefield terminates in a triangular shape in the area of Nahal Sekher (Fig. 3.1).
3.2
3.2.1
Aeolian Sand Incursions into the North-Western Negev During the Upper Quaternary Period of Aeolian Sand Incursion into the Negev
The dune sand of the Sinai and Negev is composed mostly of quartz with very few other minerals, mostly calcite, magnetite, hematite and other silicates. The source for this aeolian sand is the Nile Delta, since there is no other source for these minerals within the reach of the wadis that flow through the Sinai-Negev dunefield (Almagor 2002).
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
27
Aeolian sand became geologically significant in the sediments that were deposited in the coastal plains of the Sinai and north-western Negev around 25,000–30,000 b.p. Probably most of the aeolian sands in this period were transported and redeposited in the flood plains of the area, or they were trapped by the vegetation, creating sand sheets (Zilberman 1991). Sand dunes in considerable amount are found associated with the Upper Epipaleolithic sites that are dated to 18,000– 10,000 b.p. (Goring-Morris and Goldberg 1990). Thermoluminescence (TL) dating of the linear sand dunes and the interdune sand of Sde Hallamish reveals that sand was deposited in the interdune area from at least 43,000 until 9,000 b.p., and that there has been little deposition during the Holocene. The linear dune flanks were stabilized during the last 10,000–6,000 b.p. (Rendell et al. 1993). Radiocarbon dates are available for hearths found in two sand quarries; one, northwest of Revivim, where the age of the base of the sand is 3,030±150 b.p. (Zilberman 1991), the other in the most eastern part of the Negev sand invasion, near Nahal Sekher (Fig. 3.1), where the lower sand was mobile around 6,100 b.p. (Tsoar and Goodfriend 1994). These sands may represent the mobilization of formerly fixed Upper Epipaleolithic sand found at a Natufian site (ca. 11,000 b.p.) in the Nahal Sekher area (Goring-Morris and Goldberg 1990).
3.2.2
The Sand Red Colour and its Implications
The Negev aeolian sand shows different intensities of redness, with some variation in the hues because of the content of iron-oxide minerals. If there is a common source of all the sand and the climate is homogenous, we can assume that the different hues of red indicate different ages, as has been argued by many (Norris 1969; Folk 1976b; Walker 1979; Gardner and Pye 1981; Wopfner and Twidale 1988; White et al. 1997). Yellowish sand is younger than redder sand. Redness of the Negev sand ensues from iron oxide-bearing clay-sized particles that adhere to the surface of the quartz sand grains. Scanning electron microscope (SEM) analysis shows that the surface of reddened quartz sand is covered by flakes and granular aggregates of iron oxides (Wopfner and Twidale 1988; Pye and Tsoar 1990).
3.2.2.1
The Redness Index of the Sand
Based on the above assumption, we have mapped the red intensity of the sand by measuring the spectral signature of 63 sand samples taken in the field (Fig. 3.2). The redness index (RI) is determined according to the following spectral ratio (Mathieu et al. 1998): RI =
R2 B ∗ G3
(3.1)
28
H. Tsoar et al.
Fig. 3.2 Landsat TM image (bands 4, 3, 2) of the Negev sand dunes taken in 1987. The blue dots indicate the sand sample locations, and the lines the interpolation of equal values of RI (redness index)
where R is the visible red (640 nm), G the visible green (510 nm) and B the visible blue (460 nm) wave band. The spectral signatures were extracted from the spectral reflectance of the sand samples, measured with an ASD Fieldspec spectrometer. The samples were placed in a black plastic dish and illuminated by a 1,000-W highintensity halogen lamp at an angle of 45° and a distance of 5 cm. In order to obtain the bidirectional reflectance of the samples, each measurement was repeated 20 times from each of four directions. Figure 3.2 presents the redness interpolation map produced from the RI results. Analysis of the redness map shows that there are at least three distinct units of sand, based on their colour. Figure 3.3 shows an RI map of three sand incursions into the north-western Negev. A multiple comparison ANOVA test for the various RI values of the sand samples of the three different sand types (Fig. 3.3) indicates that there are three significantly distinct sand units that can be distinguished in terms of their colour (Fig. 3.4). It seems that there were at least three different sand incursions into the Negev Desert during the Upper Quaternary. Sand type 1, which covers all the low and outspread dunes along the north and east side of the dunefield, is the reddest and probably the oldest of the three types. Sand type 2, which is found in the southern part of the dunefield and includes Sde Hallamish and Shunra Sands, is less red. Sand type 3, which includes the Haluza and Agur Sands, is paler than the two other types and is apparently the youngest sand that penetrated into the Negev.
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
29
Fig. 3.3 Map of the three different sand types in the north-western Negev, based on RI values (Fig. 3.2)
Fig. 3.4 Results of the box & whisker analysis for the three sand types of Fig. 3.3. Based on the results of the ANOVA test, there are three significantly different sand units (P = 2.54×10−9 and F = 44.98)
30
H. Tsoar et al.
Sand type 3 has penetrated from the Sinai into the Negev in a wedge form that tapers towards Har Keren. The RI lines show increasing values in the penetration direction, which indicates that this sand became redder during downwind transport. This fact supports our assumption that the sand becomes redder with time. The distinct sand dune morphology of the Haluza-Agur Sands supports the interpretation that it is a discrete dunefield. According to Fig. 3.3, it is plausible that the HaluzaAgur Sands overlie sand type 1, and that some of sand type 1 has mixed with sand type 3. 3.2.2.2
The Amount of Iron Oxides in the Coated Surface of the Sand
All mafic heavy minerals (with specific gravity greater than 2.8) were separated from quartz grains by submerging the samples in bromoform, which has a specific gravity of 2.89 at 20 °C (Griffiths 1967). The amount of iron in the clay coating of the sand grains was extracted for all the samples by using the sodium dithionite-sodium citrate extraction method (Mehra and Jackson 1960; Smith and Mitchell 1987). The extraction of iron was done for all the coated surfaces of the quartz grains. Iron levels were determined with a Unicam Helios Alpha spectrophotometer. Here, we assume that the cause for the redness of the sand is the amount of iron oxide-bearing, clay-sized particles that adhere to the surface of the quartz sand grains (cf. above). There is a significant regression between the amount of iron (in percent) and the RI (Fig. 3.5).
Fig. 3.5 Regression of dithionite-extractable Fe (in percent) and redness index of the samples collected in the field (P = 1.91×10−7)
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
3.3
31
Wind Climate
The climate in the area of the Negev dunefield is classified as arid in the HaluzaAgur and Nahal Sekher Sands, and becomes hyper-arid in the Sde Hallamish and Shunra Sands. A comprehensive description and analysis of the northern Negev climate is given by Littmann and Berkowicz (Chap. 4, this volume). Wind is a vector characterized by magnitude and direction. The energy of the wind can be calculated from the kinetic energy (KE) equation: KE =
1 rU 3 At 2
(3.2)
where r is the density of air, U the wind speed, A the cross-sectional area of the airflow, and t the time the wind blew at wind speed U. In a similar way, we can calculate the drift potential (DP) of the wind (Fryberger 1979), which is based on Eq. (3.2) and the equation of sand transport (Lettau and Lettau 1978): DP = Σq =
U 2 (U −U t ) t 100
(3.3)
where Σq is related to the total potential sand flux from all wind directions, U is the wind velocity (in knots), measured at a height of 10 m, Ut the threshold wind velocity for sand transport (=12 knots), and t the amount of time the wind blows above the threshold velocity (in annual %). The drift potential (DP) is given in vector units (v.u.). An index of wind direction variability is represented by the ratio between the resultant drift potential and the drift potential (RDP/DP), where values close to 1 indicate a narrow unidirectional drift potential, and values close to zero indicate a multidirectional drift potential. The drift potential of the Nizzana area was calculated from two different wind recorders at the Sde Hallamish research site (one from 1991–1995, the other from 1995–2002), and from another recorder at Qetziot (for 1981–1982), which is 11 km southeast of the Sde Hallamish research site. The data of these three wind recorders were converted into values at 10 m height, based on the von Karman–Prandtl logarithmic velocity profile law (Pye and Tsoar 1990). Figure 3.6 shows sand transport roses for these three stations. The drift potential (DP) of the Nizzana area in the southern part of the North-Western Negev dunefield is between 21 and 108 v.u. (Fig. 3.6). The differences in the DP values ensue from different wind recorders and different periods of records. However, all these values indicate a low-energy wind environment (Fryberger 1979). The drift potential for some active dunes in various humid areas reaches values of 2,000 v.u. and higher (Tsoar 2001, 2005; Yizhaq et al. 2007), which indicate a very high wind energy. The sand-transporting winds in the Nizzana area have a seasonal shift. The strongest winds occur during winter and spring when a depression exists over
32
H. Tsoar et al.
Fig. 3.6 Three sand roses for Sde Hallamish and the nearby Qtziot. DP (drift potential) total vector units (v.u.) for all wind directions, RDP/DP index of wind direction variability where values close to 1 indicate narrow unidirectional drift potential, and values close to zero indicate multidirectional drift potential, RDD direction of the RDP shown by the red arrow in the downwind direction, t percentage of time the wind was above the threshold velocity for sand transport
the north-eastern Mediterranean. Strong southwest to west winds are commonly blowing under this synoptic condition. The winter storm winds are neither constant nor frequent. In contrast to winter and spring, the summer winds ensue from a constant synoptic condition corresponding to a difference in pressure between the Mediterranean and the Negev during the daytime. As a result, a sea breeze is developed that blows regularly everyday from the north–northwest, from noon until the early evening. The summer storm winds are of low magnitude (usually not above 8 m/s) and show a high degree of constancy, while the winter and spring storm winds are of high magnitude (up to 20 m/s) and low constancy.
3.3.1
Factors Affecting Mobility and Stability of the Negev Sand Dunes
For sand dunes worldwide, there is no direct relationship between the amount of rainfall and the vegetation cover. Active dunes with no vegetation cover are found in humid areas – e.g. the Oregon coastal dunes (Hunter et al. 1983), the coastal sand dunes in NE Brazil (Jimenez et al. 1999), or the Alexandria coastal dunes in South Africa (Illenberger and Rust 1988) – while in the Negev Desert, the dunes are fully stabilized by microphytes and macrophytes. Because of the high rate of infiltration
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
33
in dune sand, most of the rain in humid climates is lost to the groundwater and is not available to the plants. Hence, the amount of rainfall is not a decisive factor in sand dune stabilization, whether in humid or in arid climates, except for temporary increases in the cohesiveness of wet sand. The deep infiltration of rainwater into dune sand reduces the effect of evaporation from the ground (Chap. 6, this volume), mostly in arid areas, contrasting with other soils composed of fine particles of silt and clay. The wind energy is thus the most important factor that determines sand dune mobility, because of the noncohesiveness of the sand. High-energy wind has the power to erode sand to such an extent that it prevents seeds from germinating in the sand and stabilizing it (Chap. 6, this volume). A drift potential (DP) above 400 v.u. would be needed in the Negev sand dunes in order to obtain active dunes with little or no vegetation (Tsoar and Illenberger 1998).
3.4
The Negev Dune Forms and Their Evolution
3.4.1
Linear Dunes
The various dunefields in the western Negev are dominated by linear dunes. The linear dunes fall into two type categories, characterized by a simple, longitudinal pattern corresponding to vegetated and unvegetated surfaces. While the former are known as vegetated linear dunes (VLD), the latter are better named seif (sword in Arabic). All linear dunes possess one common characteristic of elongation that differentiates these from transverse and barchan dunes where the whole body of the dune advances.
3.4.1.1
Vegetated Linear Dunes (VLDs)
VLDs are known from many semi-arid and arid regions in Australia, the Kalahari in South Africa, and the Southwest US (Twidale 1981; Wiggs et al. 1996; Wopfner and Twidale 2001). Unlike the seif dunes, the VLDs are straighter and do not meander, are partly or fully vegetated, and have a blunt crest line and round profile (Figs. 3.7, 3.8). An exclusive attribute of VLDs is the tendency for two adjacent dunes to converge and continue as a single ridge. Convergence is in the form of a Y-junction (the tuning fork shape; Fig. 3.8), commonly open to the effect of wind (Folk 1971; Twidale 1972a; Mabbutt and Wooding 1983; Thomas 1986). Y-junctions are a symmetrical or an asymmetrical coalescence of juxtaposed VLDs. This coalescence has been attributed to a deflection by cross wind of the extreme of the ridge during the elongating process (Madigan 1946; Mabbutt and Sullivan 1968; Thomas 1986, 1997). The uniform spacing between the VLDs is a phenomenon common to linear dunes (Twidale 1972a). It is attributed to statistical occurrences (Madigan 1936; Goudie 1969) or to dynamic processes (Folk 1976a). Mabbutt and Wooding (1983) interpret VLD junctions as a response to changes in the dynamic control of
34
H. Tsoar et al.
Fig. 3.7 VLDs on the western Shunra dunefield
Fig. 3.8 Aerial photograph of north-eastern Sde Hallamish taken in 1989. Nahal Nizzana is on the right and upper side. The coalescence of two linear dunes is typical for VLDs. Small dunelets, superimposed on the linear dunes, were formed where vegetation had been removed as a result of human activities (grazing and shrub gathering)
dune pattern seeking its adjustment through equilibrium spacing. When one VLD converges with a dune adjacent to it, a new linear dune is formed downwind in the space that was formed, or where the linear ridges are closest together the constant space is maintained by coalescence of two dunes (Fig. 3.8). Most VLDs worldwide
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
35
comply with the above descriptions, but still there are patterns of linear dunes that are not parallel to each other but nevertheless have Y-junctions (Thomas 1986; Bullard et al. 1995). In many areas, vegetated linear dunes have been reported as aligning approximately parallel to the dominant strong wind direction (Madigan 1936, 1946; Clarke and Priestley 1970; Folk 1971; Higgins et al. 1974; Lancaster 1981, 1982; Hyde and Wasson 1983). This is corroborated by the way they swerve around topographic obstacles (Mainguet 1984). Secondary side winds usually exert a modifying influence on the crest and account for either the symmetry or the asymmetry of the whole dune. Although the elongation process of VLDs is evident (Harrison and Yair 1998), the exact mechanism of elongation is not completely clear because most VLDs worldwide are covered by vegetation and stabilized; there are, however, some hypotheses dealing with this aspect (Tsoar 1989; Tseo 1993). The area shown in Fig. 3.8, located at Sde Hallamish, was under heavy grazing by Bedouins from the Sinai region until 1982 when the current international border between Israel and Egypt was established. Since then, vegetation has recovered on the Israeli side while the Egyptian side has continued to be impacted by animal grazing and woodgathering (Chap. 6, this volume). Since vegetation accompanies all VLDs worldwide, it can be assumed that their formation is related to vegetation. When vegetation is removed from the linear dunes, the pattern changes to the braided form (Figs. 3.8, 3.9). Vegetation is not a solid obstacle to sand-moving winds, and sand tends to penetrate and to be trapped
Fig. 3.9 Aerial photographs of VLDs in the eastern Sinai (immediately west of Sde Hallamish) where all vegetation has been removed as a result of human activities (grazing, agriculture, shrub gathering). The formation of small dunelets superimposed on the VLDs (braided pattern) is due to the removal of vegetation
36
H. Tsoar et al.
by the vegetation. This results in the formation of coppice dunes or nebkhas (Fig. 3.10), which are sand mounds capped and protected by vegetation. The nebkhas are formed when the sand is free of phytogenic crust. Such a condition occurs on the crest of the dune and the upper south-facing slopes (Fig. 3.10). In arid areas with low wind energy, such as the Negev, vegetation thrives more easily on sand than on other, finer soils (Chap. 6, this volume). Therefore, additional vegetation will eventually clutch to the shadow lee dune, formed either from a solid obstacle or a shrub, in a process of self-propagation, thereby forming a vegetated linear dune along the direction of the strongest dominant wind. This phenomenon is known from other areas where vegetation forms nebkhas and lee (shadow) dunes that coalesce into linear ridges (see, for example, Hesp 2004). Crosswinds may and will add sand to the linear dune and impart it, in some cases, with an asymmetric profile. Vegetation, as an element of surface roughness, tends to decrease the impact of wind on sand. Hence, only strong winds are effective for vegetative dune development, and that is the reason why VLDs are aligned along the dominant strongest wind.
3.4.1.2
Seifs
Unlike VLDs, seifs are completely devoid of vegetation on both slopes. This accounts for the formation of a sharp crest, which explains the term seif. Another typical characteristic of the seif is the tortuosity of its crest line. The vegetative
Fig. 3.10 Nebkhas (coppice dunes) formed on the crest of VLDs by shrubs that trap sand. Note that the crest is devoid of phytogenic crust, which promotes sand movement
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
37
cover that flourishes in the Negev, the Israeli side of the border, stabilized all seifs that were active there when the area was under human pressure (Chap. 6, this volume). Large, active seif dunes can be observed on the Sinai side of the border, which is still subjected to anthropogenic pressure (Fig. 3.11). Seif dunes are the only type of linear dune shaped under bidirectional wind regimes impacting the dunes obliquely, the dunes extending parallel to the resultant direction of the wind (Tsoar 1983). Seif dunes are known to run parallel for scores of kilometres but they do not show any tendency for two adjacent dunes to converge into a single dune, as is common with VLDs. Seif dunes, the only elongating dunes devoid of vegetation, have a complicated mechanism of sand transport and deposition. When winds encounter the seif dune crest line at acute angles, the flow at the leeside is deflected to parallel the crest line (Tsoar 1983). Accordingly, the leeward slope is not merely a zone of deposition but also a zone of erosion by the diverted wind flow (Tsoar et al. 1985). The occurrence of erosion or deposition depends upon the angle of incidence between the wind and the crest line. When this angle is < 40°, the deflected wind has the power to erode sand along the lee slope. When the angle is close to 90°, the velocity of the diverted wind decreases and deposition mainly occurs. As a result, the linear seif dunes develop a waveform (Fig. 3.12) that then changes the angle of incidence of the wind. This sand, eroded from the leeside by the deflected wind flow, is deposited on the same side as where the dune meanders, and the angle of incidence changes to around 90°. The consequence is erosion of one side of the winding dune by one wind direction, and deposition on the other side by the other wind direction. The elongation of seif dunes is performed by migration of the waveforms along the dune.
Fig. 3.11 Seif dunes on the Sinai side of the border at Agur Sands, north of Nahal Nizzana. This linear dune meanders with peaks and saddles lengthwise
38
H. Tsoar et al.
Fig. 3.12 Meandering seif dunes formed in north-eastern Sde Hallamish from a VLD that aligned obliquely to the storm wind direction, and of which the vegetation was removed between 1970 and 1982
The two main wind directions that affect and form seif dunes, and cause erosion and deposition along the leeside always differ in their incidence angle and in their total yearly wind power. In winter, most storm winds (from west and southwest) encounter the southern slope of the seif obliquely, while in summer the storm wind is mostly from the northwest (Fig. 3.6). This lack of uniformity in total wind power from both sides consequently brings about a lack of uniformity in the rates of erosion and deposition. The response of the dune to this lack of uniformity is to form peaks and saddles along its length (Figs. 3.11, 3.12). The peaks experience deposition from the more effective wind direction, and erosion from the less effective one. Conversely, the saddles receive deposition from the less effective wind direction, and are eroded by the more effective one (Tsoar 1983). The distinction between VLDs and seif dunes is not widely accepted. Some see similarity between both types, which brought them to conclude that VLDs were originally formed as seifs during the late Pleistocene and have since become stabilized as the climate became more humid and less windy (Lancaster 1994). It is well demonstrated that when vegetation is removed from VLDs, these dunes do not evolve into seif dunes but rather into braided forms (Fig. 3.9). As was stressed above, VLDs are aligned parallel to the direction of the strongest prevailing winds, while seif dunes are formed under bidirectional wind regimes impacting the dunes obliquely from the two sides. However, seif dunes are known to develop from those parts of VLDs, such as the Y-junctions, which deviate from the usual alignment parallel to the strongest wind, and have also experienced removal of vegetation. When such deviation reaches 15–20° from the dominant wind direction, the dune is under the influence of oblique winds that divert on the leeside, and meandering forms develop (Tsoar 1989; Fig. 3.8).
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
3.4.1.3
39
Linear Dunes with Barchans in the Interdune Area
The sand dunes in the Haluza-Agur dunefield are thicker, and composed of VLDs with a braided pattern that extends from west to east. In the interdune areas, there are many stabilized, crescentic slip faces arranged along transverse lines (Figs. 3.13, 3.14). There are two hypotheses for these unique dune forms. Transverse or barchan dunes were formed first by westerly winds when the dunes were completely bare of vegetation. VLDs commenced to form later when the dunes were sparsely covered by vegetation. The arrangement of the transverse dunes along transverse lines (Fig. 3.14) supports this hypothesis. The second hypothesis states that the VLDs were formed first when the dunes were sparsely covered by vegetation. The transverse dunes with crescentic slip faces formed after the dunes were grazed and all the vegetation removed. The wind was funnelled along the interdune trough and there formed barchans. The crescentic slip faces are stabilized today but their form is very obvious, which indicates that activity was very recent (Figs. 3.13, 3.14). This dune morphology is different from the morphology of the linear dunes found at Sde Hallamish and Shunra, where the interdunes are not covered with any bedform. As we postulated above, the Haluza-Agur dunefield probably marks the youngest and thickest invasion of sand into the Negev.
Fig. 3.13 Stabilized slip face in the interdune of two linear dunes at Haluza Sands
40
H. Tsoar et al.
Fig. 3.14 Aerial photograph of Haluza Sands, taken in 2003. Note the VLDs with braided superimposed dunes (braided pattern) and the crescentic slip faces in the interdune areas
3.5
The Effect of Destruction of Vegetation on the Morphology and Dynamics of the Sand Dunes
The sand dunes of the Sinai and Negev have experienced several cycles of vegetation covering and removal (Tsoar 1995; Meir and Tsoar 1996; Chap. 6, this volume). VLDs were formed in the Sinai and Negev when vegetation grew on the dunes in the absence of human disturbance. When vegetation was removed, the shape and the profile of the dunes became subject to change by the creation of secondary, superimposed transverse dunelets with slip faces facing downwind (Figs. 3.8, 3.9). This type, known as linear-braided (Tsoar and Møller 1986), is also known from Australia (Madigan 1936; Twidale 1972b; Mabbutt and Wooding 1983). Destruction of vegetation on VLDs that change their azimuth of alignment in the order of 16 to 25°, which occurs when they converge to form a Y-junction, promotes the formation of seifs (Figs. 3.8, 3.12). Therefore, the transformation takes place after the destruction of vegetation in those areas (such as Y-junctions) where VLDs became aligned obliquely to the strongest dominant winter winds as well as to the frequent summer winds (Tsoar 1989). The occurrence of seif dunes in the Sinai and Negev (Figs. 3.8, 3.11) was initiated after the vegetation was destroyed. According to the rate of elongation of seifs in the eastern part of the Sinai, and their maximum length, it is assumed that seif dunes
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
41
started their formation in the 18th century, after the inhabitation of the Bedouin in the Sinai and the long dry period during most of the 17th and 18th century (Tsoar 1995). Sinai dunes have been bare of vegetation since then, while the vegetation of the Negev dunes recovered in the second half of the 20th century when human pressure ceased (Chap. 6, this volume). The seif dunes of Sde Hallamish had a low shape with a round profile in 1968 when the dunes were covered by vegetation. The removal of vegetation by the Sinai Bedouin between the late 1970s and 1982 resulted in the formation of sharp-edged crest lines in those areas where the linear dunes stretch in an average azimuth of 285 and 290°, which is between the summer’s northerly and north-westerly winds and the winter’s westerly to south-westerly winds. Figure 3.15 shows four corre-
Fig. 3.15 Four cross sections made across VLDs in Sde Hallamish from topographical maps of 1968 and 1982. The VLDs were vegetated in 1968, and devoid of vegetation in 1982 (after Tsoar and Møller 1986)
42
H. Tsoar et al.
Fig. 3.16 Landsat image from 1987 showing the north-western Negev, the main geographic units (azure, blue and green) and the major wadis (red and yellow)
sponding cross sections of nine linear dunes in Sde Hallamish in 1968 and 1982. All the VLDs that were in a position to become seifs with a conspicuous, sharpedged crest have increased in height by up to 6 m. VLDs that were in a position to change to the linear-braided pattern did not increase in height or change their profile (Fig. 3.15).
3.6
Buried Channels
The cycles of dune activity, be they caused by climate or human activity, have resulted in a complex interaction between the drainage patterns and the sand mantle. Many dry channels (wadis) that start from rock outcrops (mostly chalk and limestone) and continue to the sand dunes are nowadays covered by sand and have essentially become defunct wadis. Some of these wadis have thus been covered by the sand dunes and are now buried channels. Mapping this complex drainage system cannot be achieved merely by means of aerial photography or by using existing topographic maps. The drainage systems, including the buried and unburied drainage systems, were mapped by using visible and near-infrared (Landsat) as well as synthetic aperture radar (SLR) imagery (Blumberg et al. 2004),
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
43
with the objective of better understanding this interaction between the aeolian geomorphology and the fluvial one.
3.6.1
Methods of Mapping Buried Drainage Systems
The regional drainage system was mapped using Landsat data and synthetic aperture radar data from the Spaceborne Radar Laboratory (SRL) missions. The Landsat TM data have six visible and near-infrared channels (there is also a thermal channel that was not used here), and the SRL mission acquired data in L-, C- and X-band in several polarizations. In this study, the C- and L-band data were used in HH and HV polarizations. All the data were used to generate a series of drainage maps each showing a different drainage pattern. Each spectral combination was used to generate its own vector map of the drainage pattern, and the four most distinct combinations were used. These amount to a map from visible data, and near-infrared, C-band HV, and L-band HV data. All four maps were then combined (Fig. 3.17) to create a map of the combined drainage patterns.
Fig. 3.17 Map showing the combined set of drainage patterns based on visible, near-infrared, C-band HV, and L-band HV data. The areas in the centre are most strongly suspected of having buried drainage patterns
44
H. Tsoar et al.
Differences between the maps produced from the radar and Landsat imagery are related to the different processes governing the backscatter of radar energy versus reflectance in the visible and near-infrared region. Some instances where such differences occur were attributed to surface penetration and backscattering from subsurface inhomogeneities. Several continuous features were identified where the L-HV backscatter is enhanced, compared to the C-band λ = 5.6 cm, and to some extent compared to the L-HH data (note: the L-band had a wavelength of 24 cm, and HV denotes the polarization of the transmitted and received electromagnetic waveforms, i.e. horizontally polarized transmitted wave and vertically polarized received wave). These continuous features do not appear as vegetated in Landsat data and, therefore, the enhancement of the cross-polarized backscattering is attributed to scattering from inhomogeneities within the sand layer. The difference between the two wavelengths suggests that these inhomogeneities in the subsurface are deeper than the C-band wavelength.
3.6.2
Nahal Nizzana and Buried Drainage Systems in the Shunra and Haluza Sand Fields
Figure 3.18 shows the drainage route of Nahal Nizzana, as can be seen in a Landsat image. Nahal Nizzana meanders from the southeast to the northwest across the
Fig. 3.18 Landsat image showing the northern Sinai, specifically the Wadi El-Arish and Nahal Nizzana. Nahal Nizzana disappears in the sand after crossing from the Negev side into the Sinai, and is mantled by sand dunes
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
45
Fig. 3.19 SRL radar image generated from L-band HV data, showing a different and more complete meandering of Nahal Nizzana all the way to the Mediterranean
Israel–Egypt border. This wadi is one of the largest hydrological systems in the northern Negev, and drains the Negev Highlands before it disappears in the sand mantle of the Haredin dunefield in the Sinai. This is clearly observed in the Landsat 5 TM image of the region (Fig. 3.18). The regional slope in this area is 0.14° towards the northwest. Despite this slope, there is no visible central channel but rather a gradual split into many minor channels, eventually disappearing in the sands. Medial reaches of Nahal Nizzana exhibit a wide channel and suggest that an ancient channel, buried by aeolian sediments, may still be present beneath the sandy mantle of the Haluza-Agur Sands. The area was subsequently studied by means of a radar image from the Spaceborne Radar Laboratory Mission of 1994, using the L- and C-bands. The overlapping radar image of the area using the Spaceborne Radar Laboratory in L-band HV data (Fig. 3.19) enhances the trace of Nahal Nizzana all the way to the Eastern Mediterranean coast. Similarly to Nahal Nizzana, several smaller wadis become better visible in the Shunra and Haluza sand fields. These wadis connect between Har Keren and the Evha Ridge (Figs. 3.16 and 3.17). These can be seen in the centre of Fig. 3.17 as areas that appear only when the L-band HV data are used, explaining the older connection between the southern drainage network and the older, more northerly one.
46
3.7
H. Tsoar et al.
Conclusions
At some of the upper reaches of the wadis in the dune areas of the north-western Negev, “playa”-type sediments composed of silt and clay from the late Quaternary can be found. These sediments were deposited by flooding of the interdune areas of Sde Hallamish, caused by the blocking of Nahal Nizzana by sand dunes (Harrison and Yair 1998). Other areas display disruptions and deterioration of the drainage patterns that can be attributed to the cycles of vegetation removal in the dunefields. Every such cycle causes the dunes to progress to the east and blocks existing drainage features, resulting in more buried fluvial sediments and the meandering of the current fluvial feature. Acknowledgements We wish to thank the Minerva Arid Ecosystems Research Centre (AERC) of the Hebrew University of Jerusalem for supplying the required climate data for the Sde Hallamish sand dunes. We also thank Tali Neta, Nir Margalit and other “remote-sensing” students that contributed to the processing of some of the imagery, and the Jet Propulsion Laboratory of NASA for providing the SIR-C data.
References Almagor G (2002) The Mediterranean coast of Israel. Geological Survey of Israel, Jerusalem Blumberg DG, Neta T, Margalit N, Lazar M, Freilikher V (2004) Mapping exposed and buried drainage systems using remote sensing in the Negev Desert, Israel. Geomorphology 61:239–250 Bullard JE, Thomas DSG, Livingstone I, Wiggs GFS (1995) Analysis of linear sand dune morphological variability southwestern Kalahari desert. Geomorphology 11:189–203 Clarke RH, Priestley CHB (1970) The asymmetry of Australian desert sand ridges. Search 1:77–78 Folk RL (1971) Longitudinal dunes of the northwestern edge of the Simpson desert, Northern Territory, Australia, l. Geomorphology and grain size relationships. Sedimentology 16:5–54 Folk RL (1976a) Rollers and ripples in sand, streams and sky: rhythmic alteration of transverse and longitudinal vortices in three orders. Sedimentology 23:649–669 Folk RL (1976b) Reddening of desert sands: Simpson desert, Northern Territory, Australia. J Sediment Petrol 46:604–615 Fryberger SG (1979) Dune forms and wind regime. In: McKee ED (ed) A study of global sand seas. Washington, DC, US Geol Surv Prof Pap 1052, pp 137–169 Gardner R, Pye K (1981) Nature, origin and palaeoenvironmental significance of red coastal and desert dune sands. Prog Phys Geogr 5:514–534 Goring-Morris AN, Goldberg P (1990) Late quaternary dune incursions in the southern Levant: archaeology, chronology and palaeoenvironments. Quat Int 5:115–137 Goudie A (1969) Statistical laws and dune ridges in South Africa. Geogr J 135:404–406 Griffiths JC (1967) Scientific method in analysis of sediments. McGraw Hill, New York Harrison JBJ, Yair A (1998) Late Pleistocene aeolian and fluvial interactions in the development of the Nizzana dune field, Negev desert, Israel. Sedimentology 45:507–518 Hesp PA (2004) Coastal dunes in the tropics and temperate regions: location, formation, morphology and vegetation processes. In: Martinez ML, Psuty NP (eds) Coastal dunes ecology and conservation. Springer, Berlin Heidelberg New York, pp 29–49
3 Formation and Geomorphology of the North-Western Negev Sand Dunes
47
Higgins GM, Baig S, Brinkman R (1974) The sands of Thal: wind regimes and sand ridge formations. Z Geomorphol 18:272–290 Hunter RE, Richmond BM, Alpha TR (1983) Storm-controlled oblique dunes of the Oregon coast. Bull Geol Soc Am 94:1450–1465 Hyde R, Wasson RJ (1983) Radiative and meteorological control on the movement of sand at Lake Mungo. In: Brookfield ME, Ahlbrandt TS (eds) Eolian sediments and processes. Elsevier, Amsterdam, pp 311–323 Illenberger WK, Rust IC (1988) A sand budget for the Alexandria coastal dunefield, South Africa. Sedimentology 35:513–521 Jimenez JA, Maia LP, Serra J, Morais J (1999) Aeolian dune migration along the Ceara coast, north-eastern Brazil. Sedimentology 46:689–701 Lancaster N (1981) Aspects of the morphometry of linear dunes of the Namib desert. S African J Sci 77:366–368 Lancaster N (1982) Linear dunes. Prog Phys Geogr 6:475–504 Lancaster N (1994) Dune morphology and dynamics. In: Abrahams AD, Parsons AJ (eds) Geomorphology of desert environments. Chapman & Hall, London, pp 474–505 Lettau K, Lettau H (1978) Experimental and micrometeorological field studies of dune migration. In: Lettau HH, Lettau K (eds) Exploring the world’s driest climate. Centre for Climatic Research, University of Wisconsin, Madison, WI, pp 110–147 Mabbutt JA, Sullivan ME (1968) The formation of longitudinal dunes. Evidence from the Simpson Desert. Austr Geogr 10:483–487 Mabbutt JA, Wooding RA (1983) Analysis of longitudinal dune patterns in the northwestern Simpson Desert, central Australia. Z Geomorphol suppl Bd 45:51–69 Madigan CT (1936) The Australian sand-ridge deserts. Geogr Rev 26:205–227 Madigan CT (1946) The Simpson Desert Expedition, 1939 scientific reports: no. 6, geology. The sand formations. Trans R Soc S Austr 70:45–63 Mainguet M (1984) A classification of dunes based on aeolian dynamics and the sand budget. In: El-Baz F (ed) Deserts and arid lands. Martinus Nijhoff, The Hague, pp 31–58 Mathieu R, Pouget M, Cervelle B, Escadafal R (1998) Relationships between satellite-based radiometric indices simulated using laboratory reflectance data and typic soil color of an arid environment. Remote Sensing Environ 66:17–28 Mehra OP, Jackson ML (1960) Iron oxide removal from soils and clays by a dithionite-citrate system buffered with sodium bicarbonate. Clays Clay Miner 7:317–327 Meir A, Tsoar H (1996) International borders and range ecology: the case of Bedouin transborder grazing. Human Ecol 24:39–64 Norris RM (1969) Dune reddening and time. J Sediment Petrol 39:7–11 Pye K, Tsoar H (1990) Aeolian sand and sand dunes. Unwin Hyman, London Rendell HM, Yair A, Tsoar H (1993) Thermoluminescence dating of sand movement in northern Negev, Israel. In: Pye K (ed) The dynamics and environmental context of aeolian sedimentary systems. Geological Society of London, pp 69–74 Smith BFL, Mitchell BD (1987) Characterization of poorly ordered minerals by selective chemical methods. In: Wilson MJ (ed) A handbook of determinative methods in clay mineralogy. Chapman and Hall, New York, pp 275–294 Thomas DSG (1986) Dune pattern statistics applied to the Kalahari dune Desert, Southern Africa. Z Geomorphol 30:231–242 Thomas DSG (1997) Sand seas and aeolian bedforms. In: Thomas DSG (ed) Arid zone geomorphology. Wiley, Chichester, pp 373–412 Tseo G (1993) Two types of longitudinal dune fields and possible mechanisms for their development. Earth Surface Processes Landforms 18:627–643 Tsoar H (1983) Dynamic processes acting on a longitudinal (seif) sand dune. Sedimentology 30:567–578 Tsoar H (1989) Linear dunes – forms and formation. Prog Phys Geogr 13:507–528 Tsoar H (1995) Desertification in Northern Sinai in the eighteenth century. Climate Change 29:429–438
48
H. Tsoar et al.
Tsoar H (2001) Types of aeolian sand dunes and their formation. In: Balmforth NJ, Provenzale A (eds) Geomorphological fluid mechanics. Springer, Berlin Heidelberg New York, pp 403–429 Tsoar H (2005) Sand dunes mobility and stability in relation to climate. Physica A 357 (1):50–56 Tsoar H, Goodfriend GA (1994) Chronology and palaeoenvironmental interpretation of Holocene aeolian sands at the inland edge of the Sinai-Negev erg. The Holocene 4:244–250 Tsoar H, Illenberger W (1998) Reevaluation of sand dunes’ mobility indices. J Arid Land Stud 7S:265–268 Tsoar H, Møller JT (1986) The role of vegetation in the formation of linear sand dunes. In: Nickling WG (ed) Aeolian geomorphology. Allen and Unwin, Boston, MA, pp 75–95 Tsoar H, Rasmussen KR, Sørensen M, Willetts BB (1985) Laboratory studies of flow over dunes. In: Barndorff-Nielsen OE, Møller JT, Rasmussen KR, Willetts BB (eds) Proc Int Worksh Physics of Blown Sand. Department of Theoretical Statistics, University of Aarhus, pp 327–349 Twidale CR (1972a) Evolution of sand dunes in the Simpson Desert, Central Australia. Trans Inst Br Geogs 56:77–109 Twidale CR (1972b) Landform development in the Lake Eyre region, Australia. Geogr Rev 62:40–70 Twidale CR (1981) Age and origin of longitudinal dunes in the Simpson and other sand ridge deserts. Die Erde 112:231–247 Walker TR (1979) Red color in dune sand. In: McKee ED (ed) A study of global sand seas. Washington, DC, US Geol Surv Prof Pap 1052, pp 62–81 White K, Walden J, Drake N, Eckardt F, Settle J (1997) Mapping the iron oxide content of dune sands, Namib Sand Sea, Namibia, using Landsat Thematic Mapper data. Remote Sensing Environ 62:30–39 Wiggs GFS, Livingstone I, Thomas DSG, Bullard JE (1996) Airflow and roughness characteristics over partially vegetated linear dunes in the southwestern Kalahari Desert. Earth Surface Processes Landforms 21:19–34 Wopfner G, Twidale CR (1988) Formation and age of desert dunes in the Lake Eyre depocentres in central Australia. Geol Rund 77:815–834 Wopfner H, Twidale CR (2001) Australian desert dunes: wind rift or depositional origin? Austr J Earth Sci 48:239–244 Yizhaq H, Ashkenazy Y. Tsoar H (2007) Why do active and stabilized dunes coexist under the same climatic conditions? Phys Rev Lett 98(18) no 188001 Zilberman E (1991) Landscape evolution in the Central, Northern and Northwestern Negev during the Neogene and the Quaternary. Geological Survey of Israel, Jerusalem, pp 1–164
Chapter 4
The Regional Climatic Setting T. Littmann and S.M. Berkowicz
4.1
Introduction
Arid climates are generally characterised by a negative water balance (i.e. rainfall amounts are lower than potential evaporation) for at least 10 months per year (see Fig. 2 in Introduction chapter, this volume). Contrary to the most hyper-arid areas (core regions of large continental deserts, e.g. the central Sahara or coastal deserts such as the Atacama), the desert margins are influenced by rainy seasons with a distinctive seasonality. Summer rainfall may show deep intrusions into the arid regions along their equatorward fringe, whereas the poleward margins receive winter rainfall when extratropical atmospheric circulation is intensified and becomes more widespread because of the large thermal and pressure contrast between subtropical and sub-polar latitudes at that time. The sand dune field of the north-western Negev is located at the contact zone of Mediterranean sub-humid to semiarid climates, and the hyper-arid climate of the southern Negev and Sinai, both part of the Saharo-Arabian desert belt. With its northern margin being only about 16 km from the Mediterranean Sea, the area under consideration shows a steep decline in rainfall totals, as interpolated from a few recording stations from north (coastal plain: 200 mm) to south (90 mm) over a horizontal distance of only 50 km. The decrease in rainfall is associated to the frequency of intrusion of Mediterranean cyclonic fronts into the area, depending on the individual track of the pressure system (west–east with more rainfall in the northern part; southwest–northeast with uniform rainfall distribution). In this way, the regional rainfall distribution is the most important macro-scale climatic parameter controlling the environmental gradient along the desert margin, such as in the sand dune field of the north-western Negev, while meso- and micro-scale features may show well-defined interaction with vegetation and relief. However, the dynamics of regional climate have to be understood as the major boundary condition for sub-scale ecological structures and processes, i.e. evapotranspiration.
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
49
50
4.2
T. Littmann, S.M. Berkowicz
Climate Dynamics in the Eastern Mediterranean and Adjacent Areas
Located along the southernmost, subtropical branch of the mid-latitude circumpolar vortex, the climate of the entire Mediterranean Basin and adjacent areas is characterised by large seasonal contrasts resulting from seasonal shifts of zonal circulation, and the occurrence of mid-tropospheric barotropic waves and their corresponding sea-level pressure systems (“centres of action”). However, because of the eastward propagation of barotropic waves with wavelengths of several thousands of kilometres, the frequency of weather types as well as long-term fluctuations of temperature and rainfall may be completely anti-phase in the western and eastern parts of the Mediterranean. Corte-Real et al. (1995) showed non-seasonal temperature anomalies to be much more coherent with atmospheric flow patterns than with rainfall fluctuations. Atlantic blocking and cyclogenesis in the Western Mediterranean result in positive temperature anomalies in the eastern part, and vice versa (“Mediterranean Oscillation”). The seasons are directly associated with the movement and development of the great continental pressure systems over the Atlantic, Eurasia and Africa (Meteorological Office 1962). The overall pattern during the winter half year (October to March) is characterised by the Azores High being in its seasonal southernmost position over the Eastern Atlantic, whereas the Siberian Anticyclone extends to the Black Sea, the Balkans and, at times, into Scandinavia (Meteorological Office 1962; Taha et al. 1981), leading to cold continental air mass influx into the region as well as to a potential deflection of Central to Eastern Mediterranean cyclonic tracks towards southern Greece and Cyprus. Thus, a large meridional thermal contrast over south-eastern Europe is typical in winter (Furlan 1977), which may also lead to local cyclogenesis in the Eastern Mediterranean. Rainfall is most intensive when associated with cut-off lows slowly moving eastwards from the Gulf of Genoa and the Tyrrhenian Sea, sometimes enhanced by cyclogenesis over Cyprus Lows. In the summer half year, the Azores High moves along the frontal systems to a northern position and extends into the Western Mediterranean. To the east, the Siberian Anticyclone collapses in April and gives way to a large thermal depression over the Arabian Peninsula which may act as a large-scale counterpart to high pressure in the Western Mediterranean. The typically calm type of summer weather results from this constellation (Meteorological Office 1962). In the south-eastern parts of the Mediterranean, the seasonal pattern of cyclogenesis associated with central and easterly troughs from September to March, and a weak and shallow summer circulation between the Azores High and the large thermal low over the Arabian Peninsula and the Persian Gulf is controlled by the above-mentioned continental/sub-continental-scale features. In the case of frontal passages, airflow over Israel and southern Turkey may be southwest to northeast, whereas anticyclonic situations in winter (a rare case but enhanced in drought years) show north-easterly flow of continental Asian air
4 The Regional Climatic Setting
51
masses (Taha et al. 1981), interrupted by south-westerly extensions of the Black Sea Trough at times. Depending on the actual position of the barotropic waves, cyclones may follow different tracks over the Eastern Mediterranean (Karmon 1983). For Israel, Aelion (1958) described eight different synoptic types occurring with high-magnitude rainfall events over the area. Applying principal component analysis to a set of surface and aerological parameters, Ronberg and Sharon (1985) identified 18 weather types, whereas Koplowitz (1973) identified 47 surface pressure types and 23 500hPa types for the same area using composite analysis. A relatively small proportion of the types in each study may be associated with high-magnitude rainfall. In a more recent study, Littmann (2000) identified 20 weather types in the Mediterranean from cluster analysis of the daily occurrence of centres of activity over the period 1992–1996. Some of those types were shown to account for high positive rainfall departures in the easternmost Mediterranean (Fig. 4.1). Based on the regional pressure field, the seasonal wind field in the north-western Negev shows distinctive differences between the winter and summer half year (Fig. 4.2). Mean airflow in winter is from south-easterly directions towards the seasonal centre of low pressure located over the Eastern Mediterranean. However, the highest wind speeds occur during rainstorms associated with frontal passages from south-westerly to westerly directions. In summer, airflow is much shallower and directed from north-westerly to northerly directions towards the large thermal low over the Arabian Peninsula and the Persian Gulf. Spring shows the highest seasonal frequency of dust storms in the region, when Khamsinic depressions originating south of the Atlas Mountains in the north-western Sahara, and moving over Libya and Egypt induce high thermal instability and drive dust plumes with westerly to south-westerly components into the southern and central parts of Israel in April and May (Joseph et al. 1973; Middleton 1986; Littmann 1991a, 2006). In
Fig. 4.1 Weather types which may lead to rainfall events in the northern Negev. H High-pressure cell, L low pressure and cyclones with warm and cold frontal systems, arrows prevailing regional flow pattern, hatched or continuous lines without further marking upper tropospheric troughs
NW
20 W
15
30 20 10 0
NE
NW
E
30 20 10 0W
NE
NW
E
W
30 20 10 0
NE
52
25
E
10 SW
5
Summer
SW
SE
0 30 25
NW
20 W
15
N 40 30 20 10 0
SE
SW
S
S
NE
NW
E
N 40 30 20 10 0W
SE S
NW
NE
E
W
N 40 30 20 10 0
NE
E
10 SE
SW
5
Fall
SW
S
0 30 25
NW
20 W
15
N 40 30 20 10 0
SW
SE
NE
NW
E
N 40 30 20 10 0W
SE S
S
NE
NW
E
W
N 40 30 20 10 0
NE
E
10 SW
5
Winter
SE
SW
S
0 30 25
NW
20
SW
NE
NW
E W
N 40 30 20 10 0
SE S
NE
NW
E
W
N 40 30 20 10 0
NE
E
10 SE
SW
5 5
>5
4
4,5
3,5
3
2,5
2
1,5
1
0,5
<0,2
0
S
north
SW
SE S
central
SW
SE S
south
wind speed intervals [m/s] north
central
south
wind direction intervals [%]
Fig. 4.2 Seasonal wind direction and wind speed frequency distributions (both in %) in the sandy area of the north-western Negev (1998–2000)
T. Littmann, S.M. Berkowicz
W
15
N 40 30 20 10 0
SE S
4 The Regional Climatic Setting
53
summer, deepening of the Arabian thermal low (in cases, a “Red Sea trough”) may give rise to intrusions of dusty Saharan air into the region (Dayan 1987). Observed regional dust storm seasonality indeed shows a fairly good correlation with monthly deposition rates at Nizzana (Littmann 1997). The seasonal contrast in wind direction frequencies, however, is most pronounced in the southern part of the sandy area, i. e. under increasing continental conditions (Fig. 4.2). Compared to the northern parts, here we may also find higher wind speeds all over the year (Fig. 4.2).
4.3 4.3.1
Characteristics of the Northern Negev Climate Temperatures
The northern Negev is classified as an arid region – or BW, using the Koeppen classification. As the amount of annual cloud cover is low, the global solar radiation is high. Although largely the summers are hot, spring and autumn have occasional heat-wave conditions (“Sharav”). In spring, lows over NW Africa moving eastwards can bring strong hot and dry winds. Hot, dry winds coming from southerly and easterly directions are referred to as “Khamsin”. Very hot temperatures can also occur in the fall, when the Red Sea Trough can extend into the NE Mediterranean. Air temperature data (at 10 m) for the Nizzana experimental station (approx. 190 m a.m.s.l.) are shown in Fig. 4.3 for the years 1991–1995. The annual mean air temperature is 19.2 °C, with the mean annual minimum and maximum air temperatures being 12.5 and 25.9 °C respectively. The highest maximum air temperatures
40
Temperature (C)
35 30 25 20 15 10
Minimum
5
Maximum
0 1
2
3
4
5
6
7
8
9
10
11
12
Month
Fig. 4.3 Nizzana mean annual minimum and maximum air temperatures at 10 m, 1991–1995
54
T. Littmann, S.M. Berkowicz
were found in April, hovering at about 45 °C. Frost temperatures, as recorded at 10 m, seldom occurred.
4.3.2
Rainfall
Following the basic pattern of cyclonic tracks over the region, rainfall in the northern Negev decreases from north to south and from east to west. The average annual rainfall measured at the Nizzana experimental station from 1988 to 2000 was about 87 mm. Although rain can occur from October even up to May, most of the annual rainfall takes place between December and February. However, mean rainfall totals over a longer observation period mask interannual variability which has an immediate impact on the ecosystems. Figure 4.4 shows rainfall anomalies for five stations from 1960 to 1989 (hydrological years from October to September). All series are highly correlated (r 2 > 0.7), except for 1984–1989, and do not show any overall trend. Further time series analysis gave some insight into the periodicities of dry and wet periods, especially those exceeding mean standard deviation (35–40%). Autocorrelation functions and the corresponding power spectra indicate significant signals at 3.5 and 8 years, with the 8-year signal being coherent in all coherency spectra. Indeed, extremely wet years (1963/1964, 1971, 1979, and a return towards wetter conditions after 1987) show the same 8-year recurrence intervals as do drought cycles (1961/1962, 1969, 1977, 1983/1984). However, the extremely significant bivariate phase spectra signal of 3.5 years indicates more complexity in the anomaly series. In fact, a coupled cosine/sine function model combining both wave-
100,00 80,00
standard deviation
60,00 anomaly (%)
40,00 20,00 0,00 −20,00 −40,00 −60,00 −80,00
Beer Sheva
−100,00 60
62
64
66
68
Avdat 70
72
Sede Boqer 74
76
78
Gevulot 80
year
Fig. 4.4 Rainfall anomalies 1960–1989 in the north-western Negev
82
84
Revivim 86
88
4 The Regional Climatic Setting
55
100 80
Nizzana series
model series
departures (%)
60 40 20 0 −20 −40 −60
Negev mean series
−80 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 0
2
4
6
8 10
Fig. 4.5 Rainfall anomaly model for the north-western Negev. Model equation: M = cos(years • 360/8 years)+sin(years • 360/3.5 years). Note that amplitudes of the shorter wavelength have been modified
140
rain [mm]
120 100 80 60 40 20 0 1990 1992 1994 1996 1998 2000 2002 2004 2006
year
Fig. 4.6 Annual rainfall in the Nizzana area. Between 1989/1990 and 2003/2004, rainfall was recorded at the Nizzana experimental station (black bars) and, since 2004/2005, in the village of Kadesh Barnea (grey bars) about 8 km SSE of the Nizzana station (see Fig. 1 in Introduction chapter, this volume)
lengths (Fig. 4.5) matches 65% of mean series variance and, thus, reveals the basic periodic pattern of regional rainfall variability. Except for the drought years of 1972 and 1981, the model provides a fairly good explanation of the series and even a sufficient match for the Nizzana rainfall of 1989–1998 (Fig. 4.5). While the 8-year wave is a continuous model component, the 3.5-year wave shows frequent changes in amplitude, and distinctive phase shifts in 1972 and 1981, both being ENSO (El Nino/ Southern Oscillation) anomalies as in 1997/1998. Both wavelengths have been reported to occur in indices of the circumpolar vortex (around 4 and 7.6 years respectively; Littmann 1991a), as alternations of zonal and meridional winter circulation – which is fairly consistent with our findings of rainfall fluctuations in a subtropical region – and as the wavelength of monsoonal rainfall anomalies in the
56
T. Littmann, S.M. Berkowicz
tropics (Littmann 1991a, b). Furthermore, the model results imply that the current drought period in the northern Negev will be a characteristic feature of the years to come, with an intermittent wet phase from 2003 to 2006. In terms of regional circulation anomalies, what are the synoptic patterns inducing drought or extremely good rainy seasons? Within the observational period of our field experiments, the rainy season 1994/1995 was well above normal (167 mm at the Nizzana experimental station on the southern margin of the dune field), whereas the following winters of 1995/1996, 1998/1999 and 1999/2000 experienced droughts (38, 29 and 35 mm respectively; Fig. 4.6). According to Littmann’s (2000) classification of weather types, wet and dry years in the northern Negev should be an effect of the Mediterranean Oscillation. November 1994 to March 1995 was characterised by a highly persistent Azores High over the Western Mediterranean, and by the virtual absence of high-pressure systems over the eastern part (cold air outbreaks from the Siberian Anticyclone were confined to eastern and south-eastern Europe). Mid-tropospheric troughs were concentrated over the Central and Eastern Mediterranean or the Black Sea, inducing cyclogenetic activity and southerly cyclonic tracks covering the study area. In addition to bringing rainfall to the region, the same frontal disturbances led to large-scale dust storms in north-eastern Africa and in the Negev (especially 2 November 1994). The drought of the following winter showed a completely different situation. On 70% of all days between November 1995 and March 1996, the Azores High was located far west over the Atlantic and trough systems were restricted to western Europe. This situation favoured the build-up of persistent anticyclones over central Europe, Egypt, and cold air intrusions from the Siberian Anticyclone preventing cyclonic activity in the Eastern Mediterranean. In this way, the dynamics of drought in the northern Negev is closely linked to the persistence in zonal circulation over the Eastern Mediterranean in winter which, in turn, reflects a wave structure in the circumpolar vortex which is strongly influenced by the intensity of cooling in central and north Asia. It is typical that weak rainy seasons show extremely low core winter rainfall in December and January when the Siberian Anticyclone is strongest but some rainfall in late winter (March) and spring (April/May) after the winter high-pressure cells have collapsed.
4.3.3
Spatial and Temporal Patterns of Rainfall
On the meso-scale level, i.e. over the entire sand dune field of the north-western Negev (approx. 100 km), and over larger time spans (at least 10 years), the climatic gradient from the Mediterranean to the hyper-arid climate zone may be well observed in long-term rainfall records. It has been shown that rainy seasons with above-normal rainfall are controlled mainly by the frequency and position of Eastern Mediterranean cyclonic systems and their frontal passages over the study area, whereas belownormal rainy seasons show a vast expansion of high pressure into the Eastern Mediterranean during the winter months. Although our observational period along a transect from the semiarid northern margin of the dune field to its hyper-arid southern
4 The Regional Climatic Setting
57
edge covered only two full rainy seasons which both were below average, we found spatial patterns of rainfall which may explain vegetation patterns better than would the overall rainfall gradient. Rainfall was measured continuously at four measurement stations (1, 3, 5 and Yevul) along a transect of roughly 60 km from the semiarid northwestern part to the hyper-arid south-eastern end of the dune field. The rainy seasons 1998/1999 and 1999/2000 were far below average in all parts of the area, compared to the long-term totals. However, constantly low rainfall was found only at the southern margin at Nizzana (Fig. 4.7). In 1998/1999, the northern half received nearly the same amount of rainfall (around 50 mm) while the southern part was extremely dry. In 1999/2000, the rainfall was much higher at the northern margin (88 mm in Yevul), and rainfall penetrated deeper into the sand dune field (53 mm at the southern station 1 (Nizzana), because of a singular high rainfall event on 27 and 28 January 2000). Thus, although both rainy seasons were bad, we found two differing spatial patterns of rainfall over the area. To investigate the synoptic background of such spatial distributions, we compiled synoptic data from daily surface and 500-hPa weather charts for each individual rain100
annual rainfall [mm]
Fig. 4.7 Annual rainfall 1998/1999 and 1999/2000 along the geo-ecological gradient from Nizzana (1) to Yevul (Y)
80 60 40 20 0
Y
Yevul
20
5
30
3 40
1 site
distance to the coast [km] Nizzana
Fig. 4.8 Spatial rainfall patterns along the ecological gradient, based on cluster analysis of data for two rainy seasons (winter 1998/1999 and 1999/2000). Clusters are sorted from left to right in terms of frequency. Sample total: 40 rainfall events. Spatial sequence from north (Yevul) to south (station 1, Nizzana). For further explanation, see text
58
T. Littmann, S.M. Berkowicz
fall event exceeding 0.2 mm. The resulting data from 40 events between October 1998 and April 2000 were evaluated by means of cluster analysis (Ward algorithm). The results are shown in Fig. 4.8. The most frequent (22.5% each) patterns are represented by clusters 2 and 5. Rainfall patterns of type 2 are characterised by extremely low rainfall decreasing slightly from north to south, with a minor increase at the southern end at Nizzana (site 1, Nizzana). Such rainfall did not originate from genuine frontal depressions but rather from isolated cloud clusters crossing the area from cut-off lows over the Eastern Mediterranean, which is typical for a bad rainy season in the region. Rainfall pattern type 5 shows higher rainfall with a linear decrease from north to south. This is the typical pattern associated to westerly and northerly frontal depressions moving across the area under a trough over the Black Sea. This finding demonstrates the dependence of rainfall on the frequency of Black Sea troughs. Rainfall patterns 4 and 3 represent situations in which individual rainfall events are strongest in the central part of the sand dune area (sites 3 and 5). These situations result from cloud clusters moving over the area from the southwest towards a cut-off low over Cyprus, or coming from dissolving frontal outliers of Tyrrhenian cyclones. In both cases, rainfall is singular and extremely patchy. Clusters 6, 1 and 7 are rare cases. In five rainfall events, rainfall was highest at Nizzana because the remnant of a frontal depression crossed the area from the southwest. In four cases, there was rainfall only at the Yevul site, associated with some clouds moving along
Fig. 4.9 Rainfall patterns along the measurement transect on 28 January 2000. The southern edge of a cold front of a depression over Cyprus reached the northern part of the study area during the night of 27/28 January 2000. Slight rainfall from shallow cumulus clouds reached the northern half of the transect, whereas the southern part at Nizzana (station 4) received some rainfall from a small cloud cluster some hours later. This is a typical rainfall pattern associated with frontal passages from the NW, and may be classified as rainfall cluster 5 (see text). Individual rainfall totals on that day were 20.6 mm (Yevul), 16.4 mm (site 5), 18.5 mm (site 3), and 10.9 mm (site 1, Nizzana). However, the highest individual rainfall per hour was recorded at station 3 in the middle of the dune field
4 The Regional Climatic Setting
59
the Mediterranean coast. In only one case was rainfall very low (but evenly distributed), when a small front reached the area from the west. Even rainfall type 5, which should be very typical for the general rainfall gradient (linear decrease from north to south), shows a very complex temporal and spatial structure when rainfall events are investigated with a high time resolution (Fig. 4.9). Although rainfall may commence in the northern part of the dune area and last for a few hours, the southern part receives rainfall in a clearly patchy and irregular manner. While the band of cumulus clouds progressed further south on 28 January 2000, the northern edge (site 5: 20 mm) and the southern part at site 3 (19 mm) received the highest individual rainfall, while the area in-between (site 5: 15 mm) and the southern edge (Nizzana: 10 mm) remained much drier. We may conclude from these findings that only in case of well above-normal rainy seasons will the regular rainfall gradient be clearly visible. In most years, however, rainfall will be as patchy as described above.
4.3.4
Regional Vapour Flux
With the prevailing wind direction during the summer months being from the northwest, this airflow may be expected to lead to a shallow vapour transport from the Mediterranean across the study area which is associated with a diurnal landsea wind system most significant from June to August (Littmann 1997). More detailed information could be obtained from the GIDEX field experiment (1997– 2000). As Fig. 4.10 shows, near-ground temperatures are nearly identical across the sand dune field of the north-western Negev and, thus, the vapour pressure at saturation also shows no clear difference. However, it increases from winter to summer, by 130%. On the other hand, vapour pressure also follows the same seasonal pattern and rates of increase towards the summer months (100% in the southern and 150% in the northern part). Consequently, the saturation deficit shows an overall high interrelation with vapour pressure at saturation (r2 = 0.81) but seasonal fluctuations which are higher along the northern margin of the dune field (increase from winter to summer by 174%) than on the southern side (135%). Ecologically important, the seasonal increase in vapour pressure in summer could be an effect of significantly increased plant transpiration coinciding with the maximal saturation deficit of the air, but this would imply a continuously sufficient soil water storage available for root uptake, which is not likely. In terms of the seasonal wind field, however, the summer peak could be an effect of vapour advection from the Mediterranean. A detailed investigation of net vapour flux at the northern margin of the sand dune field (Fig. 4.11) indeed confirms the enormous influx of water vapour from the Mediterranean, especially from north-westerly directions, which increases the net vapour flux by a factor of 2.5 above that of winter; a finding corresponding nicely to the seasonal variation in saturation pressure deficit (Fig. 4.10). Nearground studies of actual evapotranspiration, especially of plant transpiration, will
60
T. Littmann, S.M. Berkowicz 45 40
vapor pressure at saturation
35 30 25 20
temperature
15 10 5
saturation deficit
0
30
25
vapor pressure
20
15
10
north central
5
south
0 Jun
Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun
Jul Aug Sep
Fig. 4.10 Temperatures, vapour pressure at saturation, vapour pressure, and the saturation deficit at 2 m above ground as measured at three experimental sites of the GIDEX field experiment across the sand dune field of the north-western Negev in 1998 and 1999
have to consider this circumstance. Consequently, the regional vapour field is fairly homogeneous over the sand dune field at all seasons. Assuming a homogeneous vertical distribution of specific humidity within a theoretical boundary layer of 500 m, we found differences in the liquid water equivalent (in terms of precipitable water) from north to south of only 0.5 mm in summer 1998, 0.7 mm in summer 1999, and 0.5 mm in winter 1998/1999. Thus, there is no overall implication of a decreasing water vapour gradient above canopy level in the north-western Negev.
4.3.5
Dewfall
The near-ground vapour pressure is often high enough to enable considerable dewfall during several late-night situations. Rather often, dewfall reaches 0.1 to 0.2 mm per night. Jacobs et al. (1998) found a dewfall amount of around 5 mm
4 The Regional Climatic Setting
61 Summer (Jun - Sep 1999)
Winter (Nov - Feb 1998/99) N
NW
North
W
70 60 50 40 30 20 10 0 -10
N NE
NW
E
SW
SE
S
W
70 60 50 40 30 20 10 0 -10
NE
E
SW
SE
S
Fig. 4.11 Near-ground mean net water vapour flux in % of total mean flux at the northern margin of the sand dune field (Yevul), showing the individual percentage values per wind direction. Overall mean flux is 19,226 mm of liquid water equivalent in winter, and 49,646 mm in summer. Before computing the individual percentage per wind direction, the non-horizontal flux in calm situations (wind speed < 0.5 m s−1) was subtracted following Nd • ul,d •qd • p w − Nc • ul,c • qc • rw = Fd where N is the number of cases (hours), ul the mean wind length (m), q the specific humidity (kg kg−1), rw water vapour density (kg m−3), Fd the mean net vapour flux at wind direction d, and c calm situations. Closed dots Vapour flux (%), open dots wind direction (%)
for September 1997; since there is no pronounced annual course, dewfall may reach about 20–45 mm per year which, relative to annual precipitation, is an important ecological factor for vegetation, even more so for soil crusts (Chaps. 10 and 21, this volume). More details are given about this aspect in Chapter 13 (this volume).
4.4
Conclusions
Northern Negev climate is characterised by large seasonal contrasts between the rainy season in winter and the long, dry season from April/May to October, which is typical for a desert margin at the contact zone of Mediterranean and arid climates. As rainfall amounts depend on the frequency and tracks of Central to Eastern Mediterranean cyclonic fronts skimming the area, it shows a dramatic decrease from north to south over a very short horizontal distance. However, high interannual rainfall variability is caused by the Mediterranean Oscillation when abovenormal rainy seasons are linked to the high persistence of troughs and their corresponding cyclonic systems over the Eastern Mediterranean, whereas drought years show highly persistent high-pressure cells over eastern Europe and northeast Africa coinciding with an extremely intensified Siberian Anticyclone intruding into the region. Wet and dry fluctuations show recurrence intervals which can be
62
T. Littmann, S.M. Berkowicz
sufficiently explained by a numeric model including periods of 7–8 and 3–4 years, which strongly implies the presence of circumpolar vortex zonal circulation and ENSO signals in the northern Negev rainfall anomaly series. In this way, tropical warming events indicated by SE-anomalies would cause an intensification of zonal mid-latitude circulation in winter, which is in tune with the undisturbed development of high-pressure cells over Asia and the Eastern Mediterranean and, thus, drought. The model also implies drought to be the dominant feature in regional rainfall variability in the forthcoming years. A sequence of drought years (i.e. beginning in 1995/1996) has inhibited the recharge of water storage in deeper soil layers or the groundwater reservoir, and can lead to an immediate decrease in annual plant biomass and, ultimately, perennial vegetation. However, interannual rainfall variability is a dominant environmental factor which, in terms of steady-state equilibrium, should be matched by the ecosystem’s resilience. At the regional scale, dewfall in the northern Negev is influenced largely by proximity to the Mediterranean Sea and a strong sea-breeze circulation supplying moisture. The high frequency of clear evening skies coupled with low evening wind speeds is favourable for efficient nocturnal radiational surface cooling. Hence, though droughts may occur, dewfall can still take place and provide moisture for use by some desert organisms. On the other hand, the regional water vapour field shows a reversed seasonality, with specific humidity being highest during the summer months. Maritime vapour flux reduces the vapour pressure deficit in summer by 50%, which should greatly help to reduce plant transpiration rates in the event of unlimited soil moisture availability. However, overall estimated vapour flux in the boundary layer over the dune field is relatively low (monthly means are around 5 mm in terms of liquid water equivalent in summer, and 2 mm in winter) and fairly homogeneous over the entire area, as maximum divergence from north to south does not exceed 2.5 mm.
References Aelion E (1958) A report of weather types causing marked dust storms in Israel during the cold season. Meteorological Service Israel, Hakirya, Series C Misc Pap 10 Corte-Real J, Zhang X, Wang X (1995) Large-scale circulation regimes and surface climatic anomalies over the Mediterranean. Int J Climatol 15:1135–1150 Dayan U (1987) Sand storms and dust storms in Israel – a review. Israel Atomic Energy Comm Publ 1419 Furlan D (1977) The climate of Southeast Europe. In: Wallen C (ed) Climates of Central and Southern Europe. World Survey of Climatology vol 6. Elsevier, Amsterdam, pp 185–236 Jacobs AFG, Heusinkveld BG, Berkowicz S (1998) Dew deposition in the Negev Desert: the biological crust. In: Proc 1st Int Conf Fog and Fog Collection, Vancouver, pp 261–264 Joseph J, Manes A, Ashbel D (1973) Desert aerosols transported by Khamsinic depressions and their climatic effects. J Appl Meteorol 12:792–797 Karmon Y (1983) Israel. Wissenschaftliche Buchgesellschaft, Darmstadt
4 The Regional Climatic Setting
63
Koplowitz R (1973) An objective classification of synoptic pressure-field patterns of the Eastern Mediterranean Basin for use in synoptic-climatological studies. Thesis, Hebrew University of Jerusalem, Jerusalem Littmann T (1991a) Rainfall, temperature and dust storm anomalies in the African Sahel. Geogr J 157:136–160 Littmann T (1991b) Dust storm frequency in Asia: climatic control and variability. Int J Climatol 11:393–412 Littmann T (1997) Atmospheric input of dust and nitrogen into the Nizzana sand dune ecosystem, northwestern Negev, Israel. J Arid Environ 36:433–457 Littmann T (2000) An empirical classification of weather types in the Mediterranean and their interrelation with rainfall. Theoretical Appl Climatol 66 (3/4):161–171 Littmann T (2006) Dust storms in Asia. Geogr Rundsch Int Edn 2(3):8–12 Meteorological Office (1962) Weather in the Mediterranean. Stationery Office, London Middleton N (1986) Dust storms in the Middle East. J Arid Environ 10:83–96 Ronberg B, Sharon D (1985) An objective weather typing system for Israel: a synoptic climatological study. In: Proc 9th Conf Probability and Statistics in Atmospheric Science, Virginia Beach Taha MF, Harb SA, Nagib MK, Tantawy AH (1981) The climate of the Near East. In: Takahashi K, Arakawa K (eds) Climates of Southern and Western Asia. World Survey of Climatology vol 9. Elsevier, Amsterdam, pp 183–246
Chapter 5
Soil Characteristics and Pattern of the Nizzana Research Site H.-P. Blume, L. Beyer, U. Pfisterer, and P. Felix-Henningsen
5.1
Introduction
Soils of the Nizzana Sandfield were formed under arid climate conditions during the Upper Quaternary (Chap. 1, this volume). They developed from shifting sands of sandy ridges, sandy surfaces in interdune areas, and playa surfaces. Land surfaces in deserts derived from sand are often not regarded as soils because of their instability and, therefore, unsuitability as rooting zone for higher plants (Soil Survey Staff 1994). As young desert soils are less characterized by their state of weathering, processes such as the formation of cracks, crusts and aggregates, and the translocation of salts are steps of pedogenesis (Dan 1981). To understand the pattern of soil distribution, a characteristic part of the Nizzana Sandfield (about 20 km NW of the Nizzana village) was mapped. The observed soil conditions will be discussed with regard to plant growth. In order to study the influence of the rainfall gradient on soil properties, soil observations were carried out at station N3 situated about 13 km north of the main Nizzana Research Site (cf. Fig. 29.1, Chap. 29, this volume). In Chapter 16 (this volume), soil formation and especially the salt dynamics of the soils will be discussed.
5.2
Methods
Mapping was done during the rainy season: 250 profiles were described according to FAO (2006), and sampled down to 1 m with a drilling auger over an area of about 100 ha. Texture, structure, moisture, colour (after the Munsell Color Chart), pH (H2O) and EC (1:2.5) values (conversion to saturation extract), and carbon content at different soil horizons were assessed on the basis of the field study. The soil map (Fig. 5.1) was produced by using a topographic map (1:5,000) and aerial photographs. For selected soil horizons, bulk density and water content were measured gravimetrically. Content of CaCO3 was determined for 60% of all the samples; those of organic carbon were determined by dry combustion at 1,200 °C, coulometric determination of released CO2 and subtraction of carbonate-C for 30% of the samples.
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
65
66 H.-P. Blume et al.
Fig. 5.1 Soil map of a dune area near Nizzana, Negev (34°22′E, 30°58′N), at the Nizzana test site (N1; the road on the left is the border road between Israel and Egypt; the lodge close to the playa is also indicated; A–B line position of Fig. 5.2; after Pfisterer et al. 1996). Legend changed after WRB (2006); definitions of second-level units (shortened): aridic mean of upper 20 cm < 0.2% Corg if sandy, otherwise, < 0.6%; evidence of aeolian activity; moist Munsell colour value > 2.5 and chroma > 1.5; base saturat.> 75%; calcaric with > 2% CaCO3; calcic > 15-cm-thick layer with > 15% CaCO3 at least partly secondary lime; endosalic below 30 cm EC partly > 15 dS m−1 and pH > 8.5; haplic typical; hyposalic EC partly > 4 dS m−1; protic showing no horizon development; takyric thick surface crust with polygonal cracks; yermic aridic and platy surface layer
5 Soil Characteristics and Pattern of the Nizzana Research Site
67
Fig. 5.2 Landscape position of eight soils of Table 5.1 (for position of the catena, see Fig. 5.1)
The soils described here were initially classified according to FAO (1990), but the classification was subsequently changed based on WRB (2006). Plant available water capacity (WC) was calculated according to: WC (vol%) = H2O at pF 2.5–H2O at pF 5.5. These data were estimated from the correlation between soil texture and water contents at field capacity (pF 2.5) and the wilting point (at pF 5.5 for plants of arid climate conditions, rather than pF 4.2 for temperate climate conditions). In case of high electrical conductivity, the osmotic potential is higher than pF 2.5 (for example, pF 2.6, 3.2, 3.8 at 1, 4, 15 mS cm−1 respectively). Cation exchange capacity (CEC) was calculated according to the correlation between soil texture as well as organic matter content, and potential CEC (Schlichting et al. 1995). WC (in l m−2) and CEC (in molc m−2) were calculated for the total profile from 0–100 cm. For nine profiles of a representative catena (Fig. 5.2), organic C, total nitrogen (Nt), and concentrations of cations and anions in the 1:1 water extract were analyzed. Pedogenic iron oxide (Fe0) was analyzed by oxalate extraction. Cations and anions were determined by AAS and colorimetry or titrimetry respectively. Texture was determined by laser granulometrically after removing soluble salts, carbonates and dispersion by NaPO3. Bulk density and pore size distribution (dry mass and water content after saturation at pF 1.8, 2.5 and 4.2 of 100-cm3 core samples) were analyzed, too (for further information concerning the methods, see Schlichting et al. 1995).
5.3
Soil Distribution
Arenosols, Solonchaks, Regosols, Calcisols and Fluvisols were mapped in the area (Fig. 5.1). All soils have aridic properties: low humus contents, light colour, high base saturation, and evidence of aeolian activity (WRB 2006). All soils are at least calcaric. The distribution of the soil units followed a rough scheme. On the dunes, only Haplic Arenosols appeared with scattered vegetation besides Protic Arenosols and shifting dunes without vegetation. Solonchaks were usually situated on playas close to dune slopes associated with Arenosols with salic properties. Yermic Arenosols, Takyric Regosols and Yermic Fluvisols, all with a surface crust, were distributed mostly in interdunal depressions with an even topography. At the topographic
68
H.-P. Blume et al.
lowest positions were found Calcisols, often associated with Arenosols without salic properties. Abandoned arable land consisted of different soil units, mainly of Fluvisols, Calcisols and Regosols with sandy loam as texture. Some chemical and physical features of the profiles are given in Tables 5.1 and 5.2, taken along a soil catena (Fig. 5.2).
5.3.1
Soils of the Sandy Ridges
Haplic and Protic Arenosols (calcaric) of the sandy ridges (e.g. profiles 1 and 8 in Table 5.1) are poorly developed. Nearly no diagnostic horizons could be observed down to a depth of 1 m. The colour of the moist sand is light brown (9YR 6/5). Silt and clay contents increase slightly with depth but never exceed 6%. Organic C contents and free iron oxides are low and somewhat higher near the surface. Carbonates are approximately 5%, with no variation with depth. The alkaline pH (9 to 9.5) and the low salt contents show almost no variation with depth. The structure is mainly single grained, with sedimentary cross-bedding laminae still clearly evident in certain sections. Total porosity is about 30–35 vol%. Large pores of 50– 100 µm diameter are predominant. Pores with a diameter below 10 µm represent no more than 10 vol% (profile 8 in Table 5.2). Such a pore size distribution allows deep vertical infiltration of occasional rains, with possible lateral moisture distribution caused by the cross-bedding structure of the sediment and slight changes in texture (Yair et al. 1997, and Chap. 18, this volume). Many Arenosols in slope positions have a thin platy surface crust, and were therefore classified as Yermic Arenosols (but without vesicular structure).
5.3.2
Soils of the Interdune Playa Surfaces
Most playas without vegetation cover are characterized by Solonchaks. A typical profile (see soil 3 in Tables 5.1 and 5.2) has a takyric surface. The upper 6 mm shows thin, dense layers (1–2 mm thick) above a vesicular layer. The next 3 dm is angular blocky; the ped surfaces are partly mottled and partly covered with dark brown clay skins. Deeper down, single grained sandy layers, with very few tubules and holes, alternate with loamy units, having a massive blocky structure. All layers are nearly horizontal with sharp textural boundaries, suggesting a sedimentary origin (Harrison and Yair 1998). The rusty mottles show that anaerobic conditions with stagnant water occur from time to time. The clayey layers are dense with pore sizes mainly in the range of 10–50 µm, thus retarding water infiltration and movement. Organic C content is higher than in the sandy ridge; it varies irregularly with depth, being higher in the loamy than in the sandy layers; carbonate content is higher than in the sandy ridge, and unevenly distributed. Carbonate content at such places may exceed 30%. The pH is alkaline.
0.45 0.36
0.06 0.03
11 9
4.5 4.6
0.07 0.06
8.9 9.2
43 38
2 Haplic Arenosol (calcaric, aridic) 2.1 0–0.2 3 2.2 −22 2 2.3 −55 1 2.4 −100 3 2.5/6 −155 2 2.7 −180 2
2 8 1 7 2 2
55 45 58 50 61 78
40 47 40 40 35 18
3.03 0.15 0.15 0.42 0.12 0.08
0.15 0.01 0.01 0.04 0.01 0.01
20 15 14 10 9 8
12.3 5.6 5.9 6.4 7.0 4.8
0.38 0.07 0.12 0.26 0.50 0.32
8.0 9.0 9.1 9. 9.4 10.1
149 54 61 87 97 37
3 Calcic Solonchak (aridic, takyric) 3.1 0–0.2 17 3.2 −10 26 3.3 −16 17 3.4/6 −56 7 3.7 −80 28 3.8 −116 10 3.9 −124 28 3.10/2 −200 2
45 70 48 58 64 35 66 4
35 2 23 23 7 5 4 60
3 2 12 12 1 12 2 34
1.22 3.29 1.00 0.77 2.63 1.78 2.20 0.46
0.62 0.38 0.38 0.22 0.40 0.08 0.47 0.10
2 9 3 4 13 22 5 5
32.4 38.5 40.6 23.9 39.4 13.2 39.8 7.2
2.75 23.1 11.6 5.8 17.4 4.40 12.2 1.80
7.7 8.5 8.4 8.2 7.9 8.2 7.9 8.7
903 1,530 1,450 665 130 184 753 112
4 Yermic Arenosol (calcaric, aridic) 4.1 0–5 3 4.2/4 −25 2 4.5/7 −115 2 4.8/11 −190 5
2 1 2 16
64 65 58 49
32 33 38 30
1.45 1.11 0.39 0.86
0.08 0.02 0.08 0.08
18 55 5 10
5.9 6.2 6.8 5.3
0.08 0.07 0.11 0.32
8.5 9.0 9.9 10.0
87 82 63 91
5 Yermic Calcisol (aridic) 5.1 0–0.2 5.2 −15
15 2
54 67
28 29
0.58 0.39
0.16 0.05
4 8
9.8 5.8
0.29 0.07
8.1 9.0
136 71
3 2
(continued)
69
1 Haplic Arenosol (calcaric, aridic): FS: fine sand, CS: coarse sand 1.1 0–10 <1 <1 43 55 1.2/5 −100 <1 <1 56 42
5 Soil Characteristics and Pattern of the Nizzana Research Site
Table 5.1 Properties of soils of the Nizzana dune area (for position, see numbers in Fig. 5.2 and catena in Fig. 5.1; from Blume et al. 1995; Pfisterer et al. 1996) CS Corg Norg Lime Salt pH Fe0 Depth Clay Silt FSa Profile no. (cm) (%) (%) (%) (%) (g kg−1) (g kg−1) C/N (%) (g kg−1) (H2O) (g kg−1)
70
Clay (%)
Silt (%)
FSa (%)
CS (%)
Corg (g kg−1)
Norg (g kg−1)
C/N
Lime (%)
5.3 5.4/5 5.6/8 5.9/10
3 3 27 2
9 12 54 2
52 63 24 67
36 12 5 29
0.25 0.60 2.50 0.50
0.05 0.05 0.39 0.05
4 10 9 8
7.3 8.8 35.2 7.1
6 Yermic Fluvisol (calcaric, aridic) 6.1 0–0.2 2 6.2 −10 2 6.3/4 −60 2 6.5 −68 13 6.6/10 −200 2
15 8 5 17 4
53 62 54 49 62
30 28 39 21 22
1.18 1.27 0.28 2.28 0.79
0.17 0.10 0.03 0.03 0.08
7 13 10 8 10
7 Yermic Fluvisol (calcaric, aridic) 7.1 0–0.2 1 7.2 −8 1 7.3 −32 2 7.4/6 −140 2
5 2 3 3
78 75 83 75
16 22 12 20
0.60 0.33 1.09 0.55
0.06 0.05 0.05 0.05
8 Protic Arenosol (calcaric, aridic) of a shifting dune 8.1/2 0–50 <1 <1 75
24
0.29
9 Yermic Calcisol (aridic) 9.1 0–12 9.2 −50 9.3 −80
23 24 27
0.49 0.70 1.00
a
−25 −95 −140 −200
12 3 3
16 21 12
49 55 58
FS, fine sand (0.063–0.25 mm Ø); CS, coarse sand (0.25–2 mm Ø)
Salt (g kg−1)
pH (H2O)
Fe0 (g kg−1)
0.06 0.25 4.00 0.66
9.0 9.6 8.6 9.7
89 100 773 368
10.5 8.1 5.2 9.1 5.2
0.32 0.13 0.08 0.51 0.48
8.1 8.5 9.3 10.0 10.3
156 110 61 80 112
10 7 22 11
7.4 5.5 5.9 4.9
0.14 0.10 0.09 0.13
8.5 8.5 9.0 9.2
89 70 78 67
0.03
10
4.4
0.08
9.2
54
0.17 0.04 0.17
29 18 7
3.3 3.1 16.0
013 0.19 0.19
8.3 8.5 8.5
78 96 68
H.-P. Blume et al.
Table 5.1 (continued) Depth Profile no. (cm)
5 Soil Characteristics and Pattern of the Nizzana Research Site
71
Table 5.2 Pore volume (PV), bulk density (LD), and the pore size distribution (in volume%) of soils of a dune area near Nizzana (for position, see numbers in Fig. 5.2, and catena in Fig. 5.1); diameter of pores (µm): pF < 0.6: >1,000, 0.6–1.8: 1,000–50, 1.8–2.5: 50–10, 2.5–4.2: 10–0.2, > 4.2: < 0.2 pF: <0.6 −1.8 −2.5 −4.2 > Depth LD
µm: >1,000
1.19 1.40 1.60 1.44 1.62 1.39 1.73
9 1 1 1 0 2 3
4 Yermic Arenosol (calcaric, aridic) 4.3 8–12 34 1.75 4.8 122–126 33 1.78 4.9 135–139 34 1.76 4.10 156–160 39 1.62 4.11 180–184 35 1.71 5 Yermic Calcisol (aridic) 5.2 6–10 36 5.4 50–54 34 5.5/6 93–97 36 5.7/8 128–132 35 5.10 180–184 55
No.
(cm)
PV
3 Calcic Solonchak (aridic) 3.2 2–6 55 3.3 11–15 48 3.5 44–48 40 3.7 68–72 46 3.8 98–102 39 3.9 118–122 47 3.11 146–150 35
1.69 1.75 1.69 1.72 1.19
−10
−0.2
<
8 2 9 3 6 4 15
2 2 17 2 8 2 10
13 21 8 4 8 24 4
24 22 5 36 6 16 3
0 1 2 7 0
12 14 16 17 22
17 8 4 6 5
3 7 2 6 6
2 3 10 2 2
1 0 1 1 5
17 15 2 26 6
11 12 22 2 5
3 4 6 5 6
4 3 6 1 33
19
6
3
3
8 Protic Arenosol (calcaric, aridic) of a shifting dune 8.1 10–14 32 1.82 1
−60
The coarse layers show well-developed laminae, indicating that the pedological overprint has not obliterated the sedimentary structures. Total salt content is high and strongly variable. Salt distribution is highly irregular with depth, with much higher values in the silty-clayey than in the more sandy layers (profile 3 in Table 5.1).
5.3.3
Soils of the Interdune Area
Many soils of the interdune area are stratified with strong texture variations between sandy, silty and loamy layers (Fig. 5.3). Normal soils with a mainly silty to loamy texture were classified as Regosols (Fig. 5.1). Most Solonchaks are usually situated on playas close to dune slopes associated with Arenosols with endosalic properties (Fig. 5.1). Besides Arenosols, soils with fluvic properties (stratification with a change between more sandy and more silty layers: e.g. layers 6.2 and 6.5 in Table 5.1) exist in the interdune depressions with an even topography, and were classified as Fluvisols. At the topographic lowest
72
H.-P. Blume et al.
Fig. 5.3 Mean distribution of texture, pH (H2O) and electrical conductivity (EC of the saturation extract) in soils of the Nizzana dune area N1 (after Beyer et al. 1998). Texture (according to AG Boden 1982): L loam, S sand, U silt, t clay, s sandy, u silty, f fine, m medium, Ltu silty clay loam, mS/Su medium sand above silty sand. pH: mean of three to five typical horizons. EC (in mS cm−1 of the saturation extract): 1 mS cm−1 = 1.7 mg salt g−1 soil (see Table 5.1)
5 Soil Characteristics and Pattern of the Nizzana Research Site
73
positions are found soils with high calcium carbonate contents, which were classified as Calcisols (e.g. soil 5 in Table 5.1). They are often associated with Arenosols without endosalic properties (Fig. 5.1). Soils of the interdune areas have a surface crust, which was classified as a platy surface structure. Beneath shrubs, the crust is covered by a 2–20 cm thick layer of loosely blown sand (for example, layer 4.1 in Table 5.1). The crust consists of two parts. The upper, dense part exhibits a polygonal structure. The diameter of the polygons is up to 20 cm, with shallow cracks 1–2 mm deep filled with windblown sand. The lower part of the crust has a vesicular structure. Similar crusts are typical for many soils of the East Sahara, too (Blume et al. 1984). Another system of cracks can be observed in the sandy material underlying the topsoil crust of some soils (for example, soil 4 in Table 5.1).These cracks, 0.8–1 m apart and 2 mm wide, were followed down to a depth of 100 cm and more. Similar cracks filled with windblown sand have been found in the Sahara, too (Blume 1986). The structure of the soils of the interdune area is single grained within sandy layers but coherent within silty to loamy layers. Compared to the loose sand of the sand ridge, pore size < 50 µm represents 50% of the total porosity (in Table 5.2, see layers 5.2 or 5.4 in relation to layer 8.1), indicating a better coherence, due to small amounts of silt and clay. Contents of organic carbon, carbonates and free iron oxides are low in sandy layers, but high in the crusts close to the surface and within silty to loamy layers. Calcite, gypsum and/or water-soluble salts act as weak cementing agents in most soils. The abandoned arable land consisted of Calcisols (eastern part of the site), Arenosols, Fluvisols, and a small plot with Solonchaks (Fig. 5.1). The fields are not differentiated by soil features but only by the plants aligned in rows, probably along former furrows.
5.3.4
Soils at the Haluza Station (N3)
The dune relief at N3 is more complex, as the linear dunes are intersected with barchanoid-shaped dunes. Therefore, dune basins of up to several 100 s m in diameter, and surrounded by dune ridges were formed. Some evidence exists of agricultural furrows – about 20–30 years old – within the deepest portions of the basins. The soil surface of the basins and the dune ridges, almost up to their crests, are covered by a biological crust consisting of algae, lichens and mosses. The thickness and physical stability of the crust is distinctly stronger than that at the more arid Nizzana study site. There are no ephemeral streams in the area – hence, the absence of playa sediments within the dune valleys. The soil pattern displays a homogeneous development consisting of Yermic Arenosols (calcaric) that are covered by biological crust. Only in small portions of the dune crests susceptible to higher wind erosion are Haplic Arenosols (calcaric) without biological crusts to be found. The lower flanks of the steep dune slopes are lined with a belt of Calcisols, with an indurated calcic horizon occurring below 40–50 cm depth. The topographic
74
H.-P. Blume et al.
position of these Calcisols indicates that they developed by precipitation of calcium carbonate from bicarbonate solutions that must have migrated downslope as surface runoff or sub-surface flow (Chap. 18, this volume). Whereas the topsoil of the Arenosols below the biological crust displays an extremely low organic matter content, even below the canopy of shrubs, the uppermost 2–4 cm of the interspace between shrubs and 5–8 cm below shrubs are stabilized by induration due to the accumulation of coarse silt (mean of about 10%), carbonates (mean of 4%) and salts (mean of 300 mg kg−1). This mineral crust displays a distinct lower limit, and was formed by deposition of aeolian dust. Due to wind-shadow and filter effects, the deposition of dust below perennial shrubs is stronger. Compared to the more arid Nizzana study site, the thickness of the mineral crust due to stronger deposition of silt shows higher values. A denser perennial shrub vegetation, higher growth of annual vegetation, and the surface roughness of the biological crust evidently favour dust deposition. On the other hand, the amounts of soluble salts and carbonates are 50% less compared to the corresponding values recorded in the mineral crust of Arenosols of the southern Nizzana study site. A higher annual precipitation at N3 may have favoured the leaching of soluble elements from the crust to the subsoil. The homogenous brownish-yellow subsoil is only slightly indurated, and contains relatively small amounts of silt (mean of 2%), carbonates (mean of 1.8%) and salts (150 mg kg−1). Despite the higher amounts of carbonates and salts of the mineral crust, the pH values (pH 6.8–8.5) are lower than in the subsoil (pH 8.7– 9.7), indicating on the one hand the effect of mineralization of the biological crust and vegetation detritus and, on the other hand, the leaching of easily soluble salts to the subsoil. Especially highly soluble salts accumulated below 100 cm depth within the subsoil of the Arenosols. This relatively deep leaching and accumulation within the subsoil, compared to Arenosols of the Nizzana station (N1), confirms the effect of higher long-term annual rainfall on soils of N3.
5.4
Ecological Conditions
Plants need a sufficient root penetration capability as well as available water, oxygen and nutrients.
5.4.1
Root Penetration Capability
In principle, the soils near Nizzana should experience no problems in rooting. Nevertheless, the surface crust of many downslope soils and soils of interdune areas may cause problems for the germination of many plant species (see especially all yermic subunits in the soil map of Fig. 5.1) In addition, some of the clay-rich Solonchaks can have a very dense and hard takyric soil
5 Soil Characteristics and Pattern of the Nizzana Research Site
75
structure, like bricks under dry conditions, which can restrict germination and rooting, too. There is a strong and abrupt change in texture with depth in many soils (see profiles 2–6, and 9 in Table 5.1), especially in those of the interdune area (Figs. 5.1 and 5.3). Such boundaries between layers of texture can distinctly influence root penetration. Some plant species tend to form more side roots under such circumstances.
5.4.2
Water and Oxygen Supply
Large differences in plant available water capacity (WC) exist between soils: in some soils, WC100 is lower than 60 mm, in others more than 180 mm down to 1 m (Fig. 5.4). Low WC100 values were found in many sandy Arenosols, due to a low quantity of medium pores (with diameters of 10–0.2 µm), and in most Solonchaks due to high osmotic pressure (see sites with high EC values in Fig. 5.3). Under humid climatic conditions, high WC values normally relate with a better water supply during dry periods. Under warm and arid climates such as in Nizzana with less than 90 mm annual precipitation, however, the reverse can be the case. Thus, in soils with high WC values, precipitation accumulates mainly in the topsoil and generally evaporates rapidly after only a few days, whereas in sandy soils with low WC values rainwater will penetrate the subsoil, too, and is partly protected against evaporation for several months (Blume et al. 1985). This implies that plants with a deep-reaching rooting system will grow better in sandy soils with low WC values, whereas plants with a shallow root system near the soil surface grow better in soils with high WC values. There are no soils with oxygen deficiency in the mapped area.
5.4.3
Nutrient Supply
Alkaline soils exist only near Nizzana. Under these circumstances, the cation exchange capacity (CEC) characterizes the supply of available cationic nutrients (mainly calcium, magnesium and potassium, and under arid conditions, also sodium). Pronounced differences in CEC100 values were documented (Fig. 5.4). The sandy Arenosols are characterized by low CEC values not exceeding 20 molc m−3, whereas loamy Calcisols, Fluvisols and Regosols show higher values of more than 80. The nitrogen reserves of most soils are low due to low organic matter contents (Table 5.1). Some soils may be enriched with nitrates (see Fig. 16.2 in Chap. 16, this volume), especially Solonchaks. All soils are alkaline, so that deficiency of some minor nutrients may exist, especially Fe, Mn and Zn (Marschner 1986). Some soils are extremely alkaline with pH values higher than 9.5 (Fig. 5.3), due to high contents of soda.
76
H.-P. Blume et al.
Fig. 5.4 Plant available water capacity in l m−2 (WC100), and cation exchange capacity in molc m−2 (CEC100), both in 0–100 cm in soils of the Nizzana arid dune area N1 (after Beyer et al. 1998)
Locally, many soils are very saline (see EC values in Fig. 5.3), so that besides a restriction of water availability by high osmotic pressures, plant growth will be confined to those species that can withstand salinity.
5.5
Conclusions
A dune area of 2 km2 north of Nizzana with an annual precipitation of 90 mm was mapped. The soils developed from aeolic and fluviatile sediments are Aridic and Calcaric Arenosols, Solonchaks, Regosols and Fluvisols, as well as Aridic Calcisols. Many of these are Yermic subunits, due to the presence of a surface crust. Soils at site N3 further north with about 150 mm mean annual precipitation are similar, but their salt contents are lower, and their surface crust thicker.
5 Soil Characteristics and Pattern of the Nizzana Research Site
77
The dune area shows very different ecological conditions for plant growth (Beyer et al. 1998). The fine-textured playas with Solonchaks and Calcisols, and a thick, solid surface crust have the lowest plant cover – only 10% during the end of the rainy season. Dune tops with shifting dunes as well as sandy Protic to Haplic Arenosols without a surface crust, and medium available water capacity (mainly 90–120 mm in 100 cm) have a plant cover of 25%, whereas the steep dune slopes (20–30°) with similar sandy soils show a plant cover of 35% (Chap. 8, this volume). There is large variation in plant cover in the valleys – between 25 and 45% – due to different soil conditions. Former fields, situated mainly on levelled Fluvisols, Calcisols and Regosols with sandy loam as texture, show the highest plant cover, with a value of 70% at the end of the rainy season (mainly annuals).
References AG Boden (1982) Bodenkundliche Kartieranleitung, 4. Aufl. Schweizerbart, Stuttgart Beyer L, Tielbörger K, Blume H-P, Pfisterer U, Pingpank K, Podlech D (1998) Geoecological soil features and the vegetation pattern in arid dune area in the Northern Negev, Israel. Z Pflanzenernähr Bodenk 161:347–356 Blume H-P (1986) Bildung sandgefüllter Spalten unter periglaziären und warmariden Bedingungen. Z Geomorphol N F 31:443–448 Blume H-P, Alaily F, Smettan U, Zielinski G (1984) Soil types and associations of Southwest Egypt. Berliner Geowiss Abh 50:293–302 Blume H-P, Vahrson WG, Meshref H (1985) Dynamics of water, temperature and salts in typical aridic soils. Catena 12:343–362 Blume H-P, Yair A, Yaalon DH (1995) An initial study of pedogenic features along a transect across dunes and interdune areas. Nizzana region, Negev Israel. Adv GeoEcol 28:51–64 Dan J (1981) Soils of the sandy region of the Western Negef. In: Dan J, Gerson R, Koyumdjisky H, Yaalon DH (eds) Aridic soils of Israel. Proc Int Conf Aridic Soils, Volcani Center, Bet Dagan, Israel. Inst Soils Water Spec Publ 190:211–222 FAO (1990) Soil map of the world. Revised legend. World Soil Resources Report 60. FAO, Rome FAO (2006) Guidelines for soil description, 4th edn. FAO, Rome Harrison JBJ, Yair A (1998) Late Pleistocene aeolian and fluvial interactions in the development of the Nizzana sand field, Negev desert, Israel. Sedimentology 45:507–508 Marschner H (1986) Mineral nutrition of higher plants, 2nd edn. Academic Press, London Pfisterer U, Blume H-P, Beyer L (1996) Distribution pattern; genesis and classification of soils of an arid dune area in Northern Negev. Z Pflanzenernähr Bodenk 159:419–428 Schlichting E, Blume H-P, Stahr K (1995) Bodenkundliches Praktikum. Blackwell, Berlin Soil Survey Staff (1994) Keys to Soil Taxonomy. USDA Soil Conservation Service, Washington, DC WRB (2006) World Reference Base for Soil Resources 2006. IUSS Working Group, World Soil Resources Reports 103. FAO, Rome Yair A, Lavee H, Greitser N (1997) Spatial and temporal variability of water percolation and movement in a system of longitudinal dunes, Western Negev, Israel. Hydrol Processes 11:43–58
Chapter 6
Land Use and its Effect on the Mobilization and Stabilization of the North-Western Negev Sand Dunes H. Tsoar
6.1
Introduction
Sand dunes are known to be (1) free of vegetation and active, (2) partly vegetated and active or (3) fully vegetated and fixed. All stabilized sand dunes of the world indicate mobility in the past, under different climate regimes. The sand dunes of the north-western Negev Desert are fully vegetated and fixed, in spite of the climate that is between semi-arid and hyper-arid with low amounts of annual average precipitation – from 170 mm in the north to 90 mm in the south (Tsoar and Møller 1986). The relatively dense plant cover and the microphytic crust cover of the dunes (Chap. 10, this volume) make the Negev sand dunes less arid than would appear at first glance. This phenomenon has puzzled many scientists because many models have shown that, under such arid climate, sand dunes should be active with no vegetation (Sarnthein 1978; Talbot 1984; Lancaster 1988). The perennial vegetation covers about 0–17% of the dune area, with higher cover in the interdune area and with decreasing numbers of shrubs towards the linear dunes’ crests. The microphytic crust covers the areas between the shrubs, except for some of the linear dune crests. The crust is resistant to water stress (Chap. 10, this volume) during the hot and dry summer, when it is in a dormant stage (Danin et al. 1989). Annual plants grow during a short period between the beginning of winter and early spring. The number and density of annuals depend on the rain amount and distribution, which determine the moisture in the upper 10 cm of the soil necessary for germination (Chaps. 27 and 28, this volume). Such continuous, shallow moistening of the sand occurs mostly during years with frequent, above-average rainfall, largely during the months of December, January and February, the coldest months of the year (Tsoar and Karnieli 1996).
6.2
Factors Affecting Vegetation Growth on Sand Dunes
Dune sand is generally known as inert soil devoid of almost any favourable characteristics for flora. This is due to the relatively large pore spaces of sand grains, and the lack of clay and silt. Sand has low water holding capacity (the difference between S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
79
80
H. Tsoar
field capacity and wilting point), as well as a high rate of permeability and leaching resulting in the leaching and washing away of nutrient elements necessary for plant growth (Tsoar 1990). The lack of cohesion of dune sand results in easy erodibility of the sand. Sandy soils have the lowest threshold in aeolian erosion of all known soils (Pye and Tsoar 1990). Desert dunes can be free of vegetation and active under one or a combination of the following environmental conditions: 1. Very low average rainfall. The annual average amount of rainfall for Sde Hallamish (Nizzana dunefield; Fig. 6.1) for the last 19 years (1988/1989–2006/2007) was 83 mm. The maximum yearly amount of rainfall during this period was 166.6 mm and the minimum 29.4 mm, with a standard deviation of 37. The first 7 years (1988/1989–1994/1995) had an average rainfall of 114.03 mm while, for the last 12 years (1995/1996–2006/2007), the average rainfall dropped to 65.1 mm. As a result of this long successive drought, about 65% of the shrubs in Sde Hallamish have dried up. Hence, we assume that the average rainfall limit for vegetation in this area (the north-eastern part of the Sahara Desert) is about 50–80 mm on annual average. 2. Very strong and frequent winds, which erode the sand and plants in the seedling stage, expose the plants roots, or bury the vegetation. Sand is very sensitive to erosion because of its lack of cohesiveness. Wind erosion is considered to be the
Fig. 6.1 Map of the sand dune area in the North-Western Negev and NE Sinai, showing the rainfall gradient (200 to 100 mm year−1) and the territories of Bedouin clans (after Meir and Tsoar 1996)
6 Land Use and its Effect on the Mobilization and Stabilization
81
main limiting factor of vegetation growth on dune sand (Bowers 1982; Tsoar 1990; Tsoar and Illenberger 1998). Stipagrostis scoparia, which is the most robust psammophilic plant in the Negev sand dunes, can withstand transport and severe burial by sand, though not erosion and exposure of the roots. 3. Intense human interference. The activities of pastoral nomadic societies of the Sinai have resulted in destruction of vegetation on the sand dunes by grazing, trampling and shrub gathering (Tsoar and Møller 1986; Tsoar and Karnieli 1996). Destruction of vegetation on sand dunes by the more developed societies is caused mostly by trampling and off-road vehicles (Hylgaard and Liddle 1981; Luckenback and Bury 1983). The singularity of the dune sand can explain why rainfall or rain efficiency (P/PET, the quotient of mean annual precipitation (P) to mean annual potential evapotranspiration (PET), which represents the proportion of soil moisture available for the vegetation) is not as decisive a factor in dune stabilization and mobilization as is commonly believed (Muhs and Holliday 1995). The permeability of dune sand is 2,500 times higher than that of loess (Davis and DeWiest 1966). The high rate of water infiltration into dune sand extends the wetting front to depths not reached by other soils. Perennial plants in the Negev sand dunes obtain the necessary moisture from wetting zones at depths of 60 to 180 cm, these values varying from year to year depending on rainfall amount. Even in a year with as little as 85.5 mm rainfall, there was enough moisture in the dunes of Sde Hallamish at the beginning of summer at a depth of 60 cm (Yair et al. 1997). The relatively rich cover of flora in the north-western Negev Desert is the result of the inverse texture effect that was described by Noy-Meir (1973) and supported by Le Houérou (1984), Sala et al. (1988) and Tsoar (1996). The implication of the inverse texture effect is reflected by the higher productivity of sandy soils in deserts, relative to that of fine-textured soils – despite the drawbacks of sandy soils mentioned above. The large pores spaces of sand grains, resulting in a low amount of available water to plants, are also the reason for the high primary productivity of sand in deserts. Deep percolation in sand is an advantage in deserts where minuscule amounts of soil moisture are protected from evaporation during the long dry and hot summer. Water infiltrates to depths of 60–90 cm on the crest and upper slope of the Sde Hallamish linear dunes during years with average rainfall, and to 180 cm and deeper during years with rainfall above average and high rain intensity. The infiltration into a nearby soil composed of silt and clay (known as the “playa”) was limited to 30–40 cm (Yair et al. 1997). Soil moisture at depths below 60 cm or even less is protected from evaporation during the hot, long dry summer months (Sala et al. 1988). In humid areas characterized by a high amount of precipitation, the main cause of soil-moisture loss in sand dunes is the high rate of infiltration beyond the root layer where very little water is available for plants. Hence, dune sand in humid areas has the lowest net primary production. The breakpoint between the advantage of coarse-textured sand (lower evaporation, compared with that of finer-textured soil) and its inherent disadvantages (low water
82
H. Tsoar
availability to plants and low nutrient content) lies somewhere between semi-arid and sub-humid climates (Noy-Meir 1973). Precipitation of 370 mm marks this breakpoint in the central grassland of the USA (Sala et al. 1988). This advantage of coarse sand texture has its most important manifestation in extreme arid regions. The minimum isolated shower required to produce some signs of vegetation on sandy soils is on the order of 10 to 15 mm. Such meagre precipitation has no effect on loam soils (Le Houérou 1986). As pointed out above, wind erosion is the limiting factor for vegetation on sand dunes. The rate of wind erosion depends on the wind power above the threshold velocity for sand transport. Tsoar et al. (Chap. 3, this volume) calculated and found low-energy wind power for the Sde Hallamish sand dunes. It is concluded that the rich flora on the Negev sand dunes, relative to the prevailing climate, can be explained by the inverse texture effect and the low wind power.
6.3
The Effect of the Border on Bedouin Pasture Management
As was mentioned above, the sand dunes of the Negev are richer in vegetation than the loessial soils of the area and, therefore, are more attractive for grazing. However, even within this sand dune environment there are variations in vegetation density, as the interdune low areas are richer than the dunes as such, because of lower wind power in these lower-lying areas. The local sandy vegetation types preferred by the various livestock are Artemisia monosperma, Stipagrostis scoparia, Stipagrostis plumosa, Panicum turgidum and Thymelaea hirsuta. These are the more palatable species for livestock on sand, and are grazed at normal times. Under stressful grazing conditions, however, all types – grassy and shrubby alike – are consumed even though some may be less palatable than those listed above (Bailey and Danin 1974; Bailey 1976). The division between the Negev and Sinai was determined in 1906 by agreement between the Ottoman and British empires. This borderline, which today lies between Egypt and Israel, crosses the dunes artificially from an ecological perspective. The border also crossed the tribal grazing lands of Bedouin nomads but the agreement recognized Bedouin rights of free trans-border access to their water sources, grazing pasture and cultivated lands (Brawer 1970; Meir and Tsoar 1996). Several tribes inhabited the sand dune area until 1948, their territories stretching throughout the north-western Negev and north-eastern Sinai, depending upon the availability of pasture and water (Fig. 6.1). The total population of these tribes in 1947 can be estimated cautiously at about 13,000–14,000 (Meir and Tsoar 1996). Until 1948, Bedouin activities were conducted quite freely, and trans-border movements of livestock were unrestricted. Open access at the border ended in 1949 when the State of Israel was established. Most of the tribes who had lived in the Israeli Negev escaped or were expelled to the Gaza Strip or to their lands in the Sinai. Due to acute military tension and violence between Israel and Egypt, the Israeli border area became a military zone, closed to the Israeli Bedouin who were not allowed to return to their previous habitats, these having been bisected. All Bedouin were evicted from their border habitats inland into a government-designated Bedouin enclave, known as the seig (Marx 1974;
6 Land Use and its Effect on the Mobilization and Stabilization
83
Fig. 6.1), east of Beer Sheva. Thus, the Israeli side of the sand dunes was almost completely emptied of Bedouin, while the Egyptian side contained a presumably larger Bedouin population than previously (Meir and Tsoar 1996). A substantial number of the Bedouin who previously inhabited and exploited a larger, undivided cross-border habitat within the combined Negev and Sinai were now forced into a much smaller area on the Egyptian side of the border. Obviously, they could not disperse further west into the Sinai, either because of the presence of other Egyptian Bedouin tribal territories there, or because of Egyptian government prohibitions – the Bedouin in this area became spatially blocked. The fluctuating nature of the border over the last six decades has been reflected in the subsistence patterns of the Bedouin in the north-eastern Sinai. During the first period (1948–1956), there were numerous cases of Bedouin infiltration eastwards across the border for purposes of grazing, ploughing, sowing and harvesting in their previous habitats between Egypt and Jordan. These infiltration attempts gradually diminished towards the late 1950s, particularly after 1956 when U.N. forces begun to guard the international border (Meir and Tsoar 1996). The fact that the border became largely “impervious” after 1956 completely changed grazing migration patterns. The Bedouin were now able to move only in a west–east direction up to the border on the Egyptian side, and north–south along it. They were isolated not only from their previous grazing and farming lands but also from their major water sources in the Negev (Fig. 6.1). As the carrying capacity of Bedouin territory contracted, spatially stressful conditions had occurred several times to varying degrees during several major droughts, particularly in the late 1950s and early 1960s. The first images of the Middle East taken from space in the early 1960s revealed a sharp contrast in vegetation between the Sinai and Negev, reflected in higher surface albedo in the former (Lowman 1966). During the occupation of the Sinai by Israel (1967–1982), the crossing of the border was eased to a certain degree, enabling free movement of the Sinai Bedouin into the Negev, along with their activities. Satellite images taken in the early 1970s still show the albedo differences between the Negev and Sinai (Fig. 6.2). During the second half of the 1970s and early 1980s, Israel disregarded the invasion of Bedouin from the Sinai into the Negev sand dunes. Figure 6.3 shows a herd of goats and sheep from the Sinai that crossed the Sde Hallamish dunes and reached wadi Nizzana (Nahal Nizzana). A satellite image taken on 19 September 1980 shows only a small difference in albedo between the sand dunes of the Negev and Sinai (Fig. 6.4). In 1982, the political border was re-established, Bedouin activity was stopped on the Negev side, and vegetation started to grow again in the Negev. A satellite image from 1984 shows clearly that there was an intense recovery of vegetation in the Negev (Fig. 6.5).
6.4
The Effect of Bedouin Pressure on the Negev Sand Dunes
Bedouin pressure is reflected by grazing, trampling and shrub gathering (as material for energy and building). Trampling is the main process destroying the crust and increasing sand mobility (Tsoar and Møller 1986). The change in the number of
84
H. Tsoar
Fig. 6.2 The difference in albedo along the border between the Negev and Sinai as shown by a Landsat MSS image from January 1973
Fig. 6.3 A photograph taken in 1981 of a herd of goats and sheep led by Sinai Bedouin who crossed the border and reached wadi Nizzana. In the background are the grazed sand dunes of Sde Hallamish
perennial shrubs per km2 on both sides of the border was measured from aerial photographs taken in 1945, 1956, 1968, 1982 and 1984 for the Haluza-Agur dunes and in 1968, 1976, 1982 and 1984 for the Sde Hallamish dunes (Figs. 6.1 and 6.6).
6 Land Use and its Effect on the Mobilization and Stabilization
85
Fig. 6.4 The difference in albedo along the border between the Negev and Sinai as shown by a Landsat MSS image from September 1980. The dashed line indicates the location of the border between Israel and Egypt. The darker square at the lower left of the image is an exclosure that was fenced in summer 1974 and that became visible in satellite images within only 2 years (Otterman and Tucker 1985)
Fig. 6.5 The difference in albedo along the border between the Negev and Sinai as shown by a Landsat TM image from 1984
86
H. Tsoar
The diachronic analysis reveals that, in 1945, shrub density on both sides of the border in the Haluza-Agur area was indeed almost identical. Until 1956, no significant changes could be detected, although there was a slight increase in the Negev (6% for 1945–1956). If we assume that the change began at the earliest in 1948, then there was an average annual increase of 0.75% in the Negev for the period 1948–1956 and, in the Sinai, a decline of 4%, or 0.5% annually for the same period. It should be recalled that, in these years, the Bedouin were still able (to some extent) to migrate with their flocks across the border. Yet, the increase in shrub cover in the Negev due to reduced grazing intensity, and its decline in the Sinai due to increasing grazing intensity had already begun in these years. The significant changes in the amount of shrub vegetation evidently began in the mid- or late 1950s. Eviction of the remaining Bedouin from the Israeli side of the border area, and more stringent patrolling of the borderline by the Israeli military and U.N. forces generated two trends. On the Israeli side, shrub cover increased significantly by 14% between 1956 and 1968 (1.17% annually, total of 21% since 1945), while on the Egyptian side it declined considerably by 25.2% in this period (2.1% annually, 28% since 1945). Between 1968 and 1982, the Sinai vegetation continued to decline (by 29.6% since 1968 at 2.11% annually, by 50% since 1945). Compared to the previous period (1956–1968), the fact that these rates of decline did not accelerate may be
Fig. 6.6 Changes in the number of perennial shrubs along the Israeli-Egyptian border zone from 1945 to 1989 (after Meir and Tsoar 1996)
6 Land Use and its Effect on the Mobilization and Stabilization
87
attributed to two processes. First, as shown above, other sources of subsistence became available to a significant degree to the Sinai Bedouin in the late 1960s and the 1970s. Second, they were exploiting the recovering vegetation on the stretch along the Israeli side of the border (Fig. 6.3). The extent of exploitation was so high that the previous trend of increasing shrub cover in the Negev was reversed. The amount of shrub cover declined by 21.5% to a level even slightly lower (albeit not significantly different, at 95%) than that of 1945. This decline was quite rapid (1.54% on annual average), approximating the annual rate of decline on the western side of the border. This trend reversed once again after 1982, when the border was again barred to the Sinai Bedouin and, for the first time, became impassable for their livestock due to the installation of fences. The shrub vegetation on the Israeli side increased very rapidly (13%) in only 2 years (6.5% annual average). By contrast, the Sinai Bedouin were confronted with a dwindling shrub vegetation cover on the Egyptian side. In only 2 years, this declined by an additional 30% (15% annual average) to a level only 35% of that recorded in 1945, indicative of the high deterioration rate discernible before 1968 and that has remained essentially unalleviated since then. Figure 6.6 shows very similar trends for the Sde Hallamish area between 1968 and 1984. Shrub cover on the Israeli side in this area continued to recover at a very high rate after 1984 (by 51.8% from 1984 to 1989; 10.4% on annual average), and the same may be asserted for the Haluza-Agur area, although data for the latter are not available for 1989. On the other hand, between 1968 and 1989 shrub cover on the Egyptian side of the Sde Hallamish area declined, initially at a slow rate (2.7% on annual average between 1968 and 1976), which then accelerated (4.6% an annual average between 1976 and 1982 and 9.8% between 1982 and 1984) and eventually slowed again (4.9% on annual average between 1984 and 1989).
6.5
Conclusions
The quick recovery of vegetation on the North-Western Negev sand dunes (Figs. 6.5 and 6.6) indicates that the process of vegetation removal by Bedouin (removal of shrubs for firewood, building shelters, cultivation, and livestock grazing) and the associated destruction of the biogenic crust is reversible in a couple of years. Thus, this is not a process of desertification, contrary to what has been claimed by many researchers (Otterman 1974; Muehlberger and Wilmarth 1977; Warren and Harrison 1984; Tsoar and Møller 1986). Acknowledgements I wish to thank the Minerva Arid Ecosystems Research Centre (AERC) of the Hebrew University of Jerusalem for supplying the required climate data for the Sde Hallamish sand dunes.
88
H. Tsoar
References Bailey C (1976) Elements of nature conservation in Bedouin culture (in Hebrew). Notes Bedouin 5:50–53 Bailey C, Danin A (1974) Wild plants in Bedouin life (in Hebrew). Notes Bedouin 5:1–48 Bowers JE (1982) The plant ecology of inland dunes in western North America. J Arid Environ 5:199–220 Brawer M (1970) Geographical factors in delineating the border between the land of Israel and Egypt (in Hebrew). Studies Geogr Israel 7:125–137 Danin A, Bar-Or Y, Dor I, Yisraeli T (1989) The role of cyanobacteria in stabilization of sand dunes in southern Israel. Ecol Mediterr 15:55–64 Davis SN, DeWiest RJM (1966) Hydrogeology. Wiley, New York Hylgaard T, Liddle MJ (1981) The effect of human trampling on a sand dune ecosystems dominated by Empetrum nigrum. J Appl Ecol 18:559–569 Lancaster N (1988) Development of linear dunes in the southwestern Kalahari, Southern Africa. J Arid Environ 14:233–244 Le Houérou HN (1984) Rain use efficiency: a unifying concept in arid-land ecology. J Arid Environ 7:213–247 Le Houérou HN (1986) The desert and arid zones of Northern Africa. In: Evenari BM, Noy-Meir I, Goodall DW (eds) Hot deserts and arid shrublands. Elsevier, Amsterdam, pp 101–147 Lowman PD (1966) The Earth from orbit. Natl Geogr 130:645–671 Luckenback RA, Bury RB (1983) Effects of off-road vehicles on the biota of the Algodones dunes, Imperial County, California. J Appl Ecol 20:265–286 Marx E(1974) Bedouin of the Negev (in Hebrew). Reshafim, Tel Aviv Meir A, Tsoar H (1996) International borders and range ecology: the case of Bedouin transborder grazing. Human Ecol 24:39–64 Muehlberger WR, Wilmarth VR (1977) The shuttle era: a challenge to the Earth scientist. Am Scientist 65:152–158 Muhs DR, Holliday VT (1995) Evidence of active dune sand on the Great-Plains in the 19thcentury from accounts of early explorers. Quat Res 43:198–208 Noy-Meir I (1973) Desert ecosystems: environment and producers. Annu Rev Ecol Systematics 4:25–51 Otterman J (1974) Baring high-albedo soils by overgrazing: a hypothesized desertification mechanism. Science 186:531–533 Otterman J, Tucker CJ (1985) Satellite measurements of surface albedo and temperatures in semidesert. J Climatol Appl Meteorol 24:228–235 Pye K, Tsoar H.(1990) Aeolian sand and sand dunes. Unwin Hyman, London Sala OE, Parton WJ, Joyce LA, Lauenroth WK (1988) Primary production of the central grassland region of the United States. Ecology 69:40–45 Sarnthein M (1978) Sand deserts during glacial maximum and climatic optimum. Nature 272:43–46 Talbot MR (1984) Late Pleistocene rainfall and dune building in the Sahel. Palaeoecol Africa 16:203–214 Tsoar H (1990) The ecological background, deterioration and reclamation of desert dune sand. Agric Ecosystems Environ 33:147–170 Tsoar H (1996) The effect of soil texture on the biomass in an arid region. In: Gradus Y, Lipshitz G (eds) Ben-Gurion University of the Negev Press, Beer Sheva, pp 385–391 Tsoar H, Illenberger W (1998) Reevaluation of sand dunes’ mobility indices. J Arid Land Studies 7S:265–268 Tsoar H, Karnieli A (1996) What determines the spectral reflectance of the Negev-Sinai sand dunes. Int J Remote Sensing 17:513–525 Tsoar H, Møller JT (1986) The role of vegetation in the formation of linear sand dunes. In: Nickling WG (ed) Allen and Unwin, Boston, MA, pp 75–95
6 Land Use and its Effect on the Mobilization and Stabilization
89
Warren A, Harrison CM (1984) People and the ecosystem: Biogeography as a study of ecology and culture. Geoforum 15:365–381 Yair A, Lavee H, Greitser N (1997) Spatial and temporal variability of water percolation and movement in a system of longitudinal dunes, western Negev, Israel. Hydrol Processes 11:43–58
Part B
Ecosystem Patterns
Chapter 7
The Flora of the Nizzana Research Site K. Tielbörger, R. Prasse, and H. Leschner
7.1
Introduction
The aim of this study was to investigate the composition of the vascular plant flora of the Nizzana research site and to analyse it according to life forms, chorotypes and other species-specific traits. The data should provide basic information about the research site as a prerequisite for ecological research.
7.2
Methods
A detailed field survey on the vascular plant species of the Nizzana research site was conducted from January to August 1992 (Tielbörger 1997). Desert plant species tend to show large yearly fluctuations in abundance and even fail to occur in some years; a few more species were found during subsequent years (until March 2001) and were added to the list of species. The taxonomic concept follows Zohary (1966, 1972), Feinbrun-Dothan (1978, 1986) and Feinbrun-Dothan and Danin (1991), and nomenclature was based on Danin (1998). The classification into life forms and chorotypes followed Feinbrun-Dothan and Danin (1991). In addition, species were characterised according to their overall abundance in Israel, using the classification in Fragman et al. (1999). Two specimens of most plant species were collected during flowering and fruit set. They were dried and labelled, and two plant collections were prepared in 1992. In addition, the most common plant species and those too rare for collecting were photographed during the peak of flowering and fruit set. The collection of colour slides and one plant collection are located at the Arid Ecosystems Research Center; another plant collection was given to the Botanische Staatssammlung München, Germany (index herbar.: M). Species found in years following 1992 are placed in the private herbariums of R. Prasse and/or O. Fragman (Jerusalem).
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
93
94
7.3
K. Tielbörger et al.
Results and Discussion
During the time of study, 197 species belonging to 39 families were found in the research area (Table 7.1). The families with the highest number of species were Asteraceae (24 species), Poaceae (23 species), Fabaceae (18 species) and Brassicaceae (17 species). These families are among the most diverse of the flora of Israel.
Table 7.1 List of plant species found at the Nizzana research site. The nomenclature follows Danin (1998), with reference to synonyms in Feinbrun-Dothan and Danin (1991). Asterisks Species found exclusively on the gravel track leading to the research site. Life forms (according to Feinbrun-Dothan and Danin 1991): T therophyte, H hemicryptophyte, CH chamaephyte, G geophyte, P phanerophyte. Chorotypes: s Saharo-Arabian, i Irano-Turanian, m Mediterranean, n Sudanian, tr tropical, e Euro-Siberian, a Asian. Categories of abundance are according to Fragman et al. (1999) Species name
Abundance
Life form
Chorotype
Adonis dentata Del. Aizoon hispanicum L. Allium papillare Boiss. Allium truncatum (Feinb.) Kollm. & D. Zohary Amberboa crupinoides (Desf.) DC. Ammochloa palaestina Boiss. Anabasis articulata (Forss.) Moq. Anagallis arvensis L. Anthemis melampodina Del. Argyrolobium uniflorum (Dec.) Jaub & Spach Arnebia decumbens (Vent.) Coss. & Kral. Artemisia monosperma Del. Asparagus stipularis Forss. Asphodelus ramosus Miller Asphodelus tenuifolius Cav. Asteriscus hierochunticus (Michon) Wikl. Astragalus annularis Forss. Astragalus asterias Steven Astragalus caprinus L. subsp. lanigerus (Desf.) Maire Astragalus corrugatus Bertol. Astragalus kahiricus DC. Astragalus peregrinus Vahl * Astragalus spinosus (Forss.) Muschl. Astragalus tribuloides Del. Atractylis carduus (Forss.) C. Christ Atractylis prolifera Boiss. Atriplex dimorphostegia Kar. & Kir. * Atriplex halimus L. Avena wiestii Steudel Bassia muricata (L.) Aschers. Bellevalia eigii Feinb.
C C RR F F F CC CC CC F CC CC C CC F CC F F F RR F F C CC F F RR C C C F
T T G G T T CH T T H T CH G G T T T T H T H T CH T H T T P T T G
s, i s s i s s, i, m s i, m, e s s s, i s s, m m s, n s s s, m s, i s, i s s s, i s, i s s s, i s, m s, i s, n s (continued)
7 The Flora of the Nizzana Research Site
95
Table 7.1 (continued) Species name
Abundance
Life form
Chorotype
Biscutella didyma L. Brassica tournefortii Gouan Bromus fasciculatus C. Presl Bupleurum semicompositum L. Calendula arvensis L. Calligonum comosum L’Her. *Callipeltis cucullaria (L.) Stev. *Callipeltis aperta Boiss. & Buhse *Cannabis sativa L. Carduus getulus Pomel Carrichtera annua (L.) DC. Caylusea hexagyna (Forss.) Green Centropodia forsskalii (Vahl) Cope syn. Asthenatherum forsskalii (Vahl) Nevski Centaurea aegyptiaca L. Centaurea pallescens Del. Cistanche salsa (C.A. Meyer) G. Beck Cistanche tubulosa (Schenk) Hooker fil. *Cleome arabica L. Colchicum ritchii R. Br. Convolvulus lanatus Vahl *Conyza bonariensis (L.) Cronquist Cornulaca monacantha Del. Crassula alata (Viv.) A. Berger Crucianella membranacea Boiss. Cutandia memphitica (Sprengel) K. Richter Cutandia dichotoma (Forssk.) Trabut Cynodon dactylon (L.) Pers. Cyperus macrorrhizus Nees syn. Cyperus conglomeratus Rottb. Daucus glaber (Forssk.) Thell. syn. Daucus litoralis Sm. Delphinium peregrinum L. Dipcadi erythraeum Webb & Berth. *Diplotaxis harra (Forss.) Boiss. subsp. harra Echinops philistaeus Feinbrun & Zohary Echiochilon fruticosum Desf. Emex spinosa (L.) Campd. Eminium spiculatum (Blume) Schott subsp. negevense Koach & Feinb. Enarthrocarpus strangulatus Boiss. Ephedra aphylla Forss. Eremobium aegyptiacum (Sprengel) Boiss. Erodium crassifolium L’Her. *Erodium glaucophyllum (L.) L’Her. Erodium laciniatum (Cav.) Willd. subsp. pulverulentum (Boiss.) Batt.
CC C CC F CC F C R C F C C F
T T T T T P T T T T T T H
i, m s, m m s, i, m i, m s, i s, i s, i a s s n s, i
C C R F F C RP CC O F CC F CC F
H T T T H G CH T CH T T T T H H
s s i s s s s tr s m s s, i s, i tr s
CC
T
m
F F CC F C CC F
T G H CH CH T G
i, m s s m s m i, m
C C C CC F C
T P T G H T
s s s s s s (continued)
96
K. Tielbörger et al.
Table 7.1 (continued) Species name
Abundance
Life form
Chorotype
Erodium touchyanum Del. Erucaria microcarpa Boiss. syn. Reboudia pinnata (Viv.) O.E. Schulz Erucaria pinnata (Viv.) Taeckh. & Boulos Erucaria rostrata (Boiss.) Greut. Euphorbia chamaepeplus Boiss. & Gall Euphorbia grossheimii Prokh. Fagonia arabica L. Fagonia glutinosa Del. Fagonia mollis Del. Filago desertorum Pomel Gagea dayana Chod. & Beauv. var. conjungens (Pascer) Heyn & Dafni Gastrocotyle hispida (Forssk.) Bunge *Glaucium corniculatum (L.) J.H. Rudolph Gymnarrhena micrantha Desf. Gymnocarpos decander Forss. Gynandriris sisyrinchium (L.) Parl. Haplophyllum tuberculatum (Forss.) Ad. Juss. *Helianthemum kahiricum Del. Helianthemum sessiliflorum (Desf.) Pers. *Helianthemum ventosum Boiss. Heliotropium digynum (Forss.) C. Chr. *Heliotropium rotundifolium Lehm. Herniaria hemistemon Gay Herniaria hirsuta L. Hippocrepis areolata Desv. Hordeum marinum Hudson Hormuzakia aggregata (Lehm.) Gusuleac Hypecoum littorale Wulfen Ifloga spicata (Forss.) Sch. Bip. *Kickxia floribunda (Boiss.) Taeckh. & Boulos Koelpinia linearis Pallas *Lamarckia aurea (L.) Moench Lappula spinocarpos (Forss.) I.M. Johnst. Launaea mucronata (Forss.) Muschler Launaea fragilis (Asso) Pau Leontice leontopetalum L. Leontodon laciniatus (Bertol.) Weber Leptaleum filifolium (Willd.) DC. Linaria albifrons (Sm.) Sprengel Linaria haelava (Forss.) Del. Linaria tenuis (Viv.) Sprengel Lobularia arabica (Boiss.) Muschler Loeflingia hispanica L. Lomelosia porphyroneura (Blakelock) Greuter & Burdet syn. Scabiosa porphyroneura Blakelock
F CC
T T
s s
F CC RR C F CC CC R
T T T T CH CH CH T G
s, i s s, i s s s s s, i i, m
RP C CC CC C C F F F C C F F C F RP C C C CC CC F F C F F F C F F RR F
T T T CH G H CH CH CH CH CH CH T T T T T T H T T T H H G T T T T T T T T
s, i i, m s, i s m s s s s s i s i, m, e s, m i, m m i, m s s s, i i, m s, i s s, m m s, i s i s s s s i, s (continued)
7 The Flora of the Nizzana Research Site
97
Table 7.1 (continued) Species name
Abundance
Life form
Chorotype
Lotus halophilus Boiss. & Sprun. *Lotus lanuginosus Vent Lycium shawii Roemer & Schultes Malva aegyptia L. Malva parviflora L. Matthiola livida (Delile) DC. Medicago laciniata (L.) Miller Moltkiopsis ciliata (Forss.) I.M. Johnst. *Moricandia nitens (Viv.) E.A. Dur. & Barr. Nasturtiopsis coronopifolia (Desf.) Boiss. subsp. arabica (Boiss.) Greuter & Burde Neotorularia torulosa (Desf.) Hedge & Leonard Neurada procumbens L. var. stellata M. & D. Zohary Nigella arvensis agg. L. Noaea mucronata (Forss.) Aschers. & Schweinf. Oligomeris linifolia (Hornem.) MacBride Ononis serrata Forss. Orobanche aegyptiaca Pers. Orobanche cernua Loefl. Pancratium sickenbergeri C. & W. Barbey Panicum turgidum Forss. Papaver humile Fedde *Paracaryum rugulosum (DC.) Boiss. Paronychia arabica (L.) DC. *Paronychia palaestina Eig Peganum harmala L. *Phagnalon barbeyanum Aschers. & Schweinf. Phalaris minor Retz Picris asplenioides L. Plantago albicans L. Plantago coronopus L. Plantago cylindrica Forss. Plantago ovata Forss. Polycarpon succulentum (Del.) Gay *Polygonum equisetiforme Sm. Pterocephalus brevis Coulter *Pulicaria incisa (Lam.) DC. Reichardia tingitana (L.) Roth Reseda arabica Boiss. Reseda decursiva Forss. Retama raetam (Forss.) Webb *Roemeria hybrida (L.) DC. Rostraria smyrnacea (Trin.) H. Scholz Rumex pictus Forss. *Rumex vesicarius L. Salsola inermis Forss. *Salsola kali L.
C F C F CC CC CC C C C
T H P T T T T CH CH T
s s s, n s i, m s s s s s
C F C C F F F F F C C F C RR F RP CC F F C F CC C CC C CC CC F CC CC F C CC R CC C
T T T CH T T T T G CH T H T H H CH T T H T T T T H T H T T T P T T T T T T
i s i, m, e i n s, m i s, i, m s s, n s s, i s m s, i s i, m s s, m s, i, m s s, i s i, m i, m s, n s, i s s s i, m i, m m s s i, tr (continued)
98
K. Tielbörger et al.
Table 7.1 (continued) Species name
Abundance
Life form
Chorotype
Salvia lanigera Poiret Savignya parviflora (Del.) Webb Schimpera arabica Hochst. & Steudel Schismus arabicus Nees Scorzonera papposa DC. Scrophularia hypericifolia Wydler *Scrophularia xanthoglossa Boiss. Senecio glaucus L. *Silene arabica Boiss. Silene colorata Poiret *Silene decipiens Barc. syn. Silene apetala Willd. Silene villosa Forssk. Sixalix eremophila (Boiss.) Greuter & Burdet syn. Scabiosa eremophila Boiss. *Solanum nigrum Miller Spergula fallax (Lowe) Krause Spergularia diandra (Guss.) Heldr. & Sart. Stipa capensis Thunb. Stipagrostis ciliata (Desf.) de Winter Stipagrostis plumosa (L.) T. Anders. Stipagrostis scoparia (Trin. & Rupr.) de Winter Thymelaea hirsuta (L.) Endl. Trifolium tomentosum L. Trigonella arabica Del. Trigonella stellata Forss. Trisetaria glumacea (Boiss.) Maire Trisetaria linearis Forss. Urginea undulata (Desf.) Steinh. Urginea maritima (L.) Baker Urospermum picroides (L.) F.W. Schmidt Valerianella szovitsiana Fisch. & Meyer *Verbascum fruticulosum Post *Vicia monantha Retz Vulpia brevis Boiss. & Kotschy Vulpia pectinella (Del.) Boiss. syn. Ctenopsis pectinella (Del.) DeNot. *Withania somnifera (L.) Dunal
F F C CC C F C CC F CC C F R
CH T T T H CH CH T T T T T T
s, m s s s, i i s i, m s, i s m i, m s s
C T CC CC F F C C CC CC CC R F F CC CC F C F R R
H s T T H H H CH T T T T T G G T T H T T T
i, m, e s, i, m s, i s s, i s s, m i, m s s s s, m s m i, m i i i, m s, i s
C
P
i, m, tr
In all, 38% of the species are perennial (28 hemicryptophytes, 26 chamaephytes, 15 geophytes, six phanerophytes) and 62% (122) are annual species (Fig. 7.1). The high percentage (84%) of drought-avoiding plants (in the terminology of Orshan 1953), which include therophytes, hemicryptophytes and geophytes, is characteristic for desert areas (Orshan 1953; Danin 1983; Danin and Orshan 1990). The majority of the species are of Saharo-Arabian distribution (46%), and another 27% of the species have a bi-or triregional distribution with Saharo-Arabian contribution (Fig. 7.2). The dominance of Saharo-Arabian elements among the
7 The Flora of the Nizzana Research Site
G 8%
99 P 3%
CH 13%
T 62%
H 14%
Fig. 7.1 Percentage of plant species at the Nizzana research site which belong to a certain life form. T Therophyte, H hemicryptophyte, CH chamaephyte, G geophyte, P phanerophyte
im 10%
others 5%
m 8% s 46%
i 5% sn 3% smi 3% sm 5%
si 15%
Fig. 7.2 Percentage of plant species at the Nizzana research site which belong to a certain chorotype. s Saharo-Arabian, i Irano-Turanian, m Mediterranean, n Sudanian
chorotypes indicates that floristically, the site belongs to the Saharo-Arabian region. Consistent with these findings, Danin and Plitmann (1987) have included the area of the Nizzana site in the Saharo-Arabian phytogeographic territory. Despite the clear connection of the area to the Saharo-Arabian phytogeographic territory, there is a relatively large number of Mediterranean elements. A remarkable 27% of the species occurring in this well-defined desert area have either a
100
K. Tielbörger et al.
Mediterranean or a partially Mediterranean distribution. This may be explained both by the geographical proximity to the Mediterranean Sea and coastal plain, and by the relatively favourable water conditions of sandy soils in deserts. Five species of the observed 197 species (Allium papillare, Anthemis melampodina, Bellevalia eigii, Paronychia palaestina, Phagnalon barbeyanum) are endemic to Israel and the Sinai (Fragman et al. 1999). Despite the dominance of desert plants, there are many plant species at the research site which reflect the edaphic conditions, rather than the phytogeographic region. Most of these are psammophytes, which are common both on the desert inland dunes as well as on the coastal dunes of Israel. These species include Ammochloa palaestina, Atractylis carduus, Argyrolobium uniflorum, Artemisia monosperma, Asthenatherum forsskalii, Cutandia memphitica, Cyperus macrorrhizus, Daucus litoralis, Echinops philistaeus, Echiochilon fruticosum, Hormuzakia aggregata, Ifloga spicata, Launaea tenuiloba, Lotus halophilus, Moltkiopsis ciliata, Neurada procumbens, Picris asplenioides, Polycarpon succulentum, Rumex pictus, Scrophularia hypericifolia, Silene villosa and Vulpia pectinella. Special adaptations are required to cope with mobile sands (Danin 1991). This may explain why the highest absolute and average species numbers are found in the interdune corridors and on the lower dune slopes, where the sand is stabilised by a microbiotic soil surface crust. On upper dune slopes and dune crests, where sand mobility is very high, lower numbers of species occur and large areas are characterised by a few dominant species (e.g. Stipagrostis scoparia, Moltkiopsis ciliata, Heliotropium digynum and Ifloga spicata). Other areas with scarce plant cover are soils with a relatively high content of fines in the interdune corridors (playas). Due to the high evaporative loss of water, drought stress in these areas is relatively large, and they appear almost devoid of plants in years with low rainfall. Perennials are almost absent from the playa centres but the playa edges (where fine-grained soils have a higher sandy fraction) are dominated by Anabasis articulata and, to a lesser extent, by Helianthemum sessiliflorum (Tielbörger 1997; Chap. 8, this volume). In relatively wet years, the highest absolute species number in the area is found on the playas. Several annual plant species may be found nearly exclusively on the playas – for example, Trigonella stellata, Anthemis melampodina, Plantago coronopus, Aizoon hispanicum, Carrichtera annua, Oligomeris liniifolia, Gymnarrhena micrantha, Atriplex dimorphostegia and Stipa capensis. In some parts of the research site, the playa areas have been formerly used for agricultural purposes. Many species which, in Israel, are common at ruderal sites are found almost exclusively in these disturbed parts of the research area (Leontice leontopetalum, Malva parviflora, Urospermum picroides, Leontodon laciniatum, Emex spinosa, Trigonella arabica, Peganum harmala). A different and more recent human impact on the flora of the research site is the construction of a gravel track in 1989, for facilitating access to the research site. Thirty-two plant species (16% of all species of the research area; Table 7.1) were found exclusively on this gravel track. The gravel for the track was taken from nearby Nachal Nizzana, and most of the species found exclusively on the track
7 The Flora of the Nizzana Research Site
101
grow either in Nachal Nizzana or along roadsides in the vicinity of the research area. While the agricultural activity has facilitated the success of mainly annual weeds, the road construction led to the invasion of a relatively high percentage of perennial species (e.g. Astragalus spinosus, Atriplex halimus, Cleome arabica, Glaucium corniculatum, Heliotropium rotundifolium, Moricandia nitens). The occurrence of one species (Cannabis sativa) can be directly tracked back to the activity of Bedouin smugglers, who have regularly utilised the track since 1999 for crossing the research site on their way to the Egyptian border. In February 2000, the first single individual of Cannabis was found near the gate to the site by one of us (RP). By the following season (March 2001), the species had already spread all along the track. The most abundant and frequent perennial plant species in the research area is Cornulaca monacantha (Chenopodiaceae). This shrub occurs along the entire gradient of sand stability. However, the majority of individuals, which grow in areas of cemented sand, either appear to be dead or their physiological activity is restricted to a very small part of their above-ground biomass. The most vital (i.e. fully green and highly reproductive) plants can be found on the mobile dune tops. Similar to other psammophytes, Cornulaca has long taproots, which enable the plant to reach water at large depths (Danin 1991). A thick bark around the roots serves to protect these from desiccation and abrasion by sand, if the root is exposed. Interestingly, despite its abundance at the research site, Fragman et al. (1999) have classified C. monacantha as a species ‘on the verge of extinction’ because the Hallamish sands, where the research site is located, are the only place in Israel where it occurs. However, C. monacantha is a dominant species both at the research site as well as in large parts of the sand dunes of northern Sinai. Therefore, it indicates a floristic relationship between the Hallamish sands and the sandy areas of northern Sinai. The chamaephyte Moltkiopsis ciliata (Boraginaceae) is almost as widespread and abundant at the research site as C. monacantha. Even though this shrub occurs frequently in all habitat types, it reaches highest densities in areas of semi-stable sand along gentle dune slopes (plinths). M. ciliata is well adapted to mobile sands, by being able to produce root-borne shoots (Danin 1991). Stipagrostis scoparia (Poaceae) is a perennial species characteristic for the highly mobile areas on dune tops. According to Danin (1991), S. scoparia does not merely tolerate moving sand but it requires a mobile substrate for growth and establishment. The adult plant produces active nodal roots when covered by sand, and seeds of S. scoparia require a certain sand cover to germinate (Tielbörger and Prasse, unpublished data). Consistent with this finding, only dead individuals of S. scoparia are found in the stabilised interdune dune corridors. This phenomenon indicates that S. scoparia may have established and survived in the interdune corridors in the past, when the sand in these areas was mobile. Recent studies have indicated that S. scoparia may occur mainly on the dune tops, due to competitive exclusion from more stable areas (Tielbörger and Prasse, unpublished data). The most abundant annual plant species at the research site are Senecio glaucus, Erodium laciniatum, Ifloga spicata, Bromus fasciculatus, Cutandia memphitica,
102
K. Tielbörger et al.
Rumex pictus and Picris asplenioides. While B. fasciculatus and P. asplenioides occur mainly on stabilised sand, the other species are found in semi-stable and even mobile sand. For example, I. spicata dominates areas of semi-stable sand along moderate dune slopes. This tiny composite is covered with sand grains, which may protect the plant tissues from abrasion by airborne sand particles and reduce evaporation loss. Of the 197 species recorded at the research site, 18 (9%) have recently been categorised by Fragman et al. (1999) as being rare in Israel. These species and the categories of abundance are: - Species on the verge of extinction (1–3 sites in Israel): Cornulaca monacantha. - Very rare species (4–30 sites in Israel): Allium papillare, Astragalus corrugatus, Atriplex dimorphostegia, Euphorbia grossheimii, Loeflingia hispanica, Paronychia palaestina. - Rare species (31–100 sites): Convolvulus lanatus, Glaucium corniculatum, Hypecoum littorale, Phagnalon barbeyanum. - Somewhat rare species: Callipeltis aperta, Cistanche salsa, Rumex vesicarius, Sixalix eremophila, Trisetaria glumacea, Vulpia brevis, Vulpia pectinella. Of these 18 species, seven species (4% of all observed species) – Allium papillare, Astragalus corrugatus, Atriplex dimorphostegia, Convolvulus lanatus, Euphorbia grossheimii, Loeflingia hispanica, Paronychia palaestina – have been declared ‘red species’ recommended for protection by law (Fragman et al. 1999). An additional six species occurring in Nizzana (3% of all observed species) are already protected by law: Calligonum comosum, Colchicum ritchii, Dipcadi erythraeum, Pancratium sickenbergeri, Retama raetam, Urginea maritima. Interestingly, none of these species has been classified by Fragman et al. (1999) as either rare or endangered in Israel.
7.4
Conclusions and Summary
Floristically, the Nizzana research area belongs to the Saharo-Arabian phytogeographical region, with a strong Mediterranean affiliation. Most species are annuals, and most of the perennial plants shed their above-ground parts during summer. The flora of the Hallamish sands, where the research site is located, is unique in Israel and the dunes represent some of the very rare inland sand dune areas in this country. The importance of the site is reflected by the fact that about 9% of the overall species of the research site are considered rare or very rare in Israel. For a small area such as the research site, this value is relatively high, compared to other areas in Israel. The Hallamish sands are crucial for the survival of the rare Cornulaca monacantha in Israel. A floristic survey of the whole area of the Hallamish sands may reveal an even higher number of rare and endangered vascular plant species. The uniqueness of the Hallamish sands within Israel highlights the necessity for the conservation of the area as a nature reserve. Taking into account the recent rapid
7 The Flora of the Nizzana Research Site
103
and large-scale extensions of agricultural fields in the vicinity of the research site, the need for a protection of this sensitive sand dune area becomes even more pronounced. Acknowledgements A. Danin, D. Podlech and O. Fragman helped to identify species and provided additional findings. Their help is gratefully acknowledged. This study was partially financed by a scholarship to KT by the AERC.
References Danin A (1983) Desert vegetation of Israel and Sinai. Cana, Jerusalem Danin A (1991) Plant adaptions in desert dunes. J Arid Environ 21:193–212 Danin A (1998) Wild plants of Eretz Israel and their distribution. Carta, Jerusalem Danin A, Orshan G (1990) The distribution of Raunkiaer life forms in Israel in relation to environment. J Veget Sci 1:41–46 Danin A, Plitmann U (1987) Revision of the plant geographical territories of Israel and Sinai. Plant Systematics Evol 156:43–53 Feinbrun-Dothan N (1978) Flora Palaestina, vol 3. Israel Academy of Sciences and Humanities, Jerusalem Feinbrun-Dothan N (1986) Flora Palaestina, vol 4. Israel Academy of Sciences and Humanities, Jerusalem Feinbrun-Dothan N, Danin A (1991) Analytical flora of Eretz Israel. Cana, Jerusalem Fragman O, Plitmann U, Heller D, Shmida A (1999) Checklist and ecological data-base of the flora of Israel and its surroundings, including Israel, Jordan, The Palestinian Autonomy, Golan Heights, Mt. Hermon and Sinai. Mifalot ‘Yeffe Nof’ and The Middle East Nature Conservation Promotion Association, Jerusalem Orshan G (1953) Note on the application of Raunkiaer’s system of life forms in arid regions. Palestine J Bot 6:120–122 Tielbörger K (1997) The vegetation of linear desert dunes in the north-western Negev, Israel. Flora 192:261–278 Zohary M (1966) Flora Palaestina, vol 1. Israel Academy of Sciences and Humanities, Jerusalem Zohary M (1972) Flora Palaestina, vol 2. Israel Academy of Sciences and Humanities, Jerusalem
Chapter 8
The Vegetation of the Nizzana Research Site K. Tielbörger, R. Prasse, and R. Bornkamm
8.1
Introduction
The study of the vegetation of the Northern Negev has a long history (see Danin and Orshan 1999 and references cited therein). For example, previous phytosociological studies (Eig 1938; Zohary 1944; Eig 1946; Orshan and Zohary 1963; Danin 1978, Danin 1983; Zohary 1982) have characterized the vegetation of the large northern Sinai and northern Negev sandy areas as one association dominated by Stipagrostis scoparia and Artemisia monosperma. More recently, Danin and Solomeshch (1999) have presented a comprehensive work on the coastal and desert vegetation of Israel, which covers the Nizzana area. While the vegetation of the Nizzana research site has been previously mapped and described in detail (Tielbörger 1993, 1997), no attempt has yet been made to integrate the findings into a larger syntaxonomic system. One aim of this study was to fill that gap and search for ties to the more recent work of Danin and Solomeshch (1999). Relatively little work has been done to classify the vegetation of desert areas, in contrast to the large amount of phytosociological studies in temperate regions. Therefore, there is still a lively discussion about the appropriate methodology to be used. For example, contrary to European tradition, the main criterion used for distinguishing between desert plant communities has been the dominance of perennial species (e.g. Zohary and Orshansky 1949; Danin 1978, 1983; Zohary 1982; Danin and Solomeshch 1999). The dominance criterion is based on the rationale that diagnostic species sensu Braun-Blanquet (1964) are missing in arid environments, and the ‘fitness’ of different species in various habitats is best reflected by their relative abundance (Zohary and Orshansky 1949; Danin et al. 1964). Annual species have mostly been neglected as ‘marker’ species (sensu Danin et al. 1964) for desert communities, since they show large year-to-year fluctuations in relative and absolute abundance (but see Danin and Solomeshch 1999). According to Zohary and Orshansky (1949), only perennial species form a permanent and characterizing framework of desert plant communities, whereas annual species simply mark a highly variable and seasonal aspect. However, Baierle (1993) and Sukopp (personal communication) pointed out that it may be worthwhile to distinguish between communities of different life forms, e.g. annual flora and shrub communities. Yet,
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
105
106
K. Tielbörger et al.
in order to include annuals as a distinguishing criterion, they have to be recorded in separate and smaller sampling units, and over more than one growing season (Baierle 1993). In this study, we could utilize results of a 3-year study which investigated spatial patterns in annual plants at the Nizzana research site (see also Prasse 1999). In particular, we tested whether distribution patterns of annuals correlate with those of perennial plants, and whether the observed patterns are stable over more than a single growing season. The explanation of patterns of distribution and abundance of organisms in space and time is the major goal of empirical ecological research. Using own measurements and results from parallel studies (Beyer et al. 1998), we attempted to investigate which are the factors in the physical environment which may determine the observed distribution patterns of the communities of higher plants. We focused on soil properties (Beyer et al. 1998), exposition, slope, and surface structure as determinants of plant community structure. Thus, the overall results may help in understanding the relationships between biotic and abiotic factors in determining the spatial distribution of certain habitat types within the Nizzana research site.
8.2
Methods
Despite the ongoing methodological discussion (see above), we used the dominance of perennial species as the main criterion for distinguishing between different plant communities. This was done in order to be consistent with the tradition of phytosociology in Israel (Danin and Orshan 1999). However, we subsequently tested the suitability of annual plants as marker species by investigating spatial relationships between the annual and perennial communities. From January to July 1992, 176 plots of 100 m2 each were established at the research site. Location of plots was selected with emphasis on covering the environmental gradient across the dune ridges. The plots were established in apparently homogeneous vegetation stands in different catenary positions on dune ridges with differing geomorphology (i.e. dune top, moderate slopes, steep slopes, dune base, interdune corridors, playas) which had been defined after intensive preliminary observations. Between eight and 37 plots were established in each of the units. The shape of the plots was 10×10 m, except along the dune bases where vegetation grows in narrow strips and where the shape was modified to 5×20 m. In each plot, perennial and annual species were recorded (nomenclature followed FeinbrunDothan and Danin 1991), and the plots were revisited several times throughout the growing season in order to identify and record all emerging annuals. The following data were recorded for each plot (Tielbörger 1993): 1. Percent cover of perennial plants and, relative to this, the cover of each perennial plant species. 2. Percent cover of annual plants and presence of annual plant species. 3. Habitat (catenary position, exposition, surface structure, sand mobility, soil type).
8 The Vegetation of the Nizzana Research Site
107
Soil type was recorded also when any plot was located adjacent to an existing soil profile (for details of soil sampling, see Beyer et al. 1998, and this volume). Relevés were constructed for each plot using the data on relative cover of perennial species only. The relevés were arranged in a table into groups with the same dominating perennial species, and rankless plant communities characterized by one or more perennial species (markers) were distinguished (Tielbörger 1993). A constancy table was constructed to illustrate the differences between the communities. In addition, we estimated average plant cover and species number per plot and, based on perennial species only, the Shannon diversity index and evenness for each plant community. Similar to most European phytosociological studies, the sampling design did not fulfil the assumptions for powerful statistics. Therefore, we did not statistically test for differences among community attributes. The large quadrat size was not appropriate for obtaining reliable estimates of annual plant abundance (see Tielbörger 1993). Therefore, small permanent quadrats of 20×20 cm were established in November 1993, and densities of annual plant species were recorded in these quadrats for three consecutive growing seasons (for detailed description of methods, see Prasse 1999). The question whether annual species distribution patterns match the distribution of perennial plants was tested for four different perennial plant communities: the communities of the playa areas, hard crusts, plinths (gentle dune slopes), and interdune corridors. In all, 45 permanent quadrats were placed randomly within the playa areas and 30 quadrats each in the other three communities, and densities of emerging annual plant species were recorded during 1993–1994, 1994–1995 and 1995–1996. Since the last season was extremely dry and almost no annuals emerged, we excluded the data of 1995–1996 from all subsequent analyses. The effect of plant community on the abundance of annuals was tested for each year and annual species separately using Kruskal-Wallis tests. Mean densities of annual plants were tested for significant differences between all pairs of the four selected plant communities, using Kruskal-Wallis tests. A species was regarded as diagnostic for a certain plant community when it exhibited its highest densities in that plant community in both years, and when between-community differences in densities were statistically significant in at least one season. It should be noted that this criterion is different from the dominance criterion used for perennial plants.
8.3
Results
The most frequent perennial species in the plots was Cornulaca monacantha (Chenopodiaceae). Also very frequent were Moltkiopsis ciliata (Boraginaceae), Anabasis articulata (Chenopodiaceae), Convolvulus lanatus (Convolvulaceae), Cyperus macrorrhizus (Cyperaceae) and Heliotropium digynum (Boraginaceae). The most recorded annual species were Senecio glaucus (Asteraceae), Erodium laciniatum (Geraniaceae), Bromus fasciculatus (Poaceae), Ifloga spicata (Asteraceae) and Rumex pictus (Polygonaceae).
108
8.3.1
K. Tielbörger et al.
Main Plant Communities
The following seven different plant communities were distinguished, each characterized by one or more dominating perennial species.
8.3.1.1
Stipagrostis scoparia-Heliotropium digynum Community
Areas of mobile sand (indicated by ripple marks) on the top or the upper slopes of the dunes are characterized by a plant community which is dominated by the perennial grass Stipagrostis scoparia and the semishrub Heliotropium digynum. Only few other perennial species were regularly recorded here, including Moltkiopsis ciliata and Cornulaca monacantha (Table 8.1). The most frequent annual species found in this plant community are Senecio glaucus, Erodium laciniatum, Cutandia sp. and Rumex pictus. Eremobium aegyptiacum was almost exclusive to the plant community of the dune tops (see Tielbörger 1993, 1997). Mean species number per plot was the lowest detected for all communities, and the absolute number of species found was relatively low, too. Shannon diversity and evenness had intermediate values. The percentage of graminoid perennials reached a maximum in the life-form spectrum here (Table 8.2), indicating the dominance of S. scoparia. The physiognomy of the plant community is characterized by irregularly distributed phytogenic hillocks of the nebkha type formed by S. scoparia and by vegetative runners of H. digynum. The hillocks are between 0.5 and 2.5 m high and between 1 and 10 m in diameter, and may be as far as 50 m apart. Except for the facultative annual plant E. aegyptiacum, annual plant growth is almost exclusively restricted to the hillocks formed by S. scoparia. Mean plant cover per plot was low (Table 8.2). However, no plot was located in the large barren areas with mobile sand, indicating that even this low plant cover overestimates the actual situation. All soil profiles which were located in this plant community were classified as arenosols without crust (Beyer et al. 1998).
8.3.1.2
Moltkiopsis ciliata-Convolvulus lanatus Community
A plant community dominated by Moltkiopsis ciliata and Convolvulus lanatus was found on moderate slopes or in depressions between secondary dunes in the upper dune area (plinths) or in areas in the interdunes which are covered by a sand layer. Furthermore, the perennial species Cornulaca monacantha, Stipagrostis scoparia and Cyperus macrorrhizus frequently appeared in this community (Table 8.1). The most constant annual species were Senecio glaucus, Rumex pictus, Ifloga spicata, Erodium laciniatum, Lotus halophilus and Ononis serrata (Table 8.1). Altogether, 71 higher plant species were recorded for the Moltkiopsis ciliata-Convolvulus lanatus community, and the mean number of species was nearly twice as high as for the community of the mobile dune tops (Table 8.2). Shannon diversity was somewhat higher than on the dune tops, which mainly reflects the higher average species number.
Community COM
SYN
Species
II
III
IV
V
VI
VIIA
V 43 V 26 II 4 IV 11 I0 I0 I1 I0
IV 6 V9 V 57 IV 6 II 3 II 3 II 2
III 2 IV 3 I2 V 48 IV 11 III 3 II 1 I1 I0 II 1 II 1 I2 I0 I0 V 14 II 3
II 7 III 4 III 2 IV 7 III 2 V 38 III 12 II 1 II 0 I0 II 2 I0 II 0 I0 I0 IV 11 III 4
I0 III 3 I0 V 10 V6 IV 6 II 1 III 5 III 1 III 3 III 5 IV 11 III 2 II 1 II 5 V 15 IV 6
II 3 III 2 II 1 IV 7 III 3 III 2 I1 I0 II 0 II 1 II 1 II 2 I0 I0 I0 V 68 II 3
I0 I0 I0 II 0 II 0 II 1 I0 I0 II 0 I1 I1 I0 I0 I1 I2 III 10 V 77
II 2 V 72
I0 IV 3 III 2
II 2 II 0 III 1 III 2
I0 III 1 IV 10 IV 9
I0 I0 III 3 II 2
I1 II 1 II 1 I0
II 0 I1
I0 II 3
III 5 I2
IV 9
VIIB
I0
I4
I5 II 1
IV 4
II 2
(continued)
109
Characteristic perennial species (sensu Tielbörger 1997) I A1 Stipagrostis scoparia I A1 Heliotropium digynum II A1 Echinops philistaeus III C Moltkiopsis ciliata III Convolvulus lanatus IV O Noaea mucronata IV C Artemisia monosperma V O Stipagrostis plumosa V O Pancratium sickenbergeri V Asthenatherum forsskalii V Thymelaea hirsuta V Echiochilon fruticosum V O Argyrolobium uniflorum V Haplophyllum tuberculatum V A2 Helianthemum sessiliflorum VI A2 Cornulaca monacantha VII A2 Anabasis articulata Other perennial species I Scrophularia hypericifolia II Calligonum comosum IV Fagonia mollis V O Salvia lanigera V Launaea tenuiloba V C Retama raetam
I
8 The Vegetation of the Nizzana Research Site
Table 8.1 Constancy table for the plant communities of the Nizzana research site, showing constancy classes and mean relative cover (in % of total perennial cover)a
110
Table 8.1 (continued) Community COM V V V V VIIB VIIA VIIA VIIA
SYN C C
O O
V
A1
C C C
O
Panicum turgidum Cyperus macrorrhizus Atractylis carduus Lycium shawii Erodium crassifolium Astragalus kahiricus Astragalus caprinus Colchicum ritchii Dipcadi erythraeum Gymnocarpos decander Asphodelus ramosus
I
II
III
IV
III 2 I1
II 1 I0
I0 III 3 III 1
III 3 III 1
I0 I0
I0
I1 I0 I0
V
VI
VIIA
VIIB
I0 III 1 II 1
I0 I0 II 0
I0
I0
I
I1 IV 6 IV 1 I1 I0 I0 I0 I0 I0 II 1 I
II 0 II 0 II 0 II 0 I0 II 1 I
I I I I IV III IV III I I V III III
II 0 I0 I0 I0
I0
Eremobium aegyptiacum Silene villosa Ctenopsis pectinella Papaver hybridum Brassica tournefortii Ifloga spicata Rumex pictus Ononis serrata Centaurea pallescens Hypecoum littorale Crucianella membranacea Bromus fasciculatus Linaria haelava
II I I I I III IV I II
I IV II II II IV IV II III
I III I I I V V IV I I
II I
II I
IV II
I II I II II IV V III V II II IV II
Sixalix eremophila
I
I
II
II
I0
I I II I I IV III III III I I IV II I
II 20 I0 I0 I I
II II I III
I I
III II
IV I
I
II
K. Tielbörger et al.
Annuals I II II II, IV II, IV III III, IV III, V IV IV IV V V
Species
O
O A2 O A2 O
O O C C O O O
I
II
I I
III
II
I II I I I
I III II I II I
II III III I I
I IV II I II
III I
III I I
III I I I I I I I I III V IV IV I IV III IV III II II I I I I
III I I I I I I I
I I I
I I
IV V IV II II II I I I
V V V III III II I
IV V V IV II IV III III III I I I I I I
I III V IV IV III III III II II II I I I I I
III V III III III IV III II III I I I
I II V II III II II II III II II I I I I II III III I I III III IV I I II I
I III I II I I IV IV IV III II II II II IV I V IV II I II IV
111
Adonis dentata Matthiola livida Trigonella stellata Bassia muricata Filago desertorum Herniaria hirsuta Atriplex dimorphostegia Carduus getulus Stipa capensis Anthemis melampodina Calendula arvensis Leontodon laciniatus Paronychia arabica Reseda arabica Erucaria rostrata Arnebia decumbens Cutandia sp. Senecio glaucus Erodium laciniatum Lotus halophilus Polycarpon succulentum Picris asplenioides Schismus arabicus Hippocrepis areolata Neurada procumbens Plantago cylindrica Bupleurum semicompositum Nigella arvensis Daucus litoralis Trisetaria linearis Delphinium peregrinum
8 The Vegetation of the Nizzana Research Site
V VI VIIA VIIA VIIA VIIA VIIA VIIB VIIB VIIB VIIB VIIB VIIB VIIB VIIB VIIB
I (continued)
112
Table 8.1 (continued) Community COM
SYN
Species
C O
Hormuzakia aggregata Schimpera arabica Carrichtera annua Urospermum picroides
I
II
III
IV
I I I I
V
VI
I
I
I
VIIA
VIIB
I I I I
I I I
a
Constancy classes: I, <20%; II, 20–40%; III, 41–60%; IV, 61–80%; V, 81–100%. Values in italics indicate community preferences. COM denotes the number of the community for which the species is characteristic or for which it exhibits a preference. Numbers of communities with eponymous species: I, Stipagrostis scoparia-Heliotropium digynum; II, Echinops philistaeus; III, Moltkiopsis ciliata-Convolvulus lanatus; IV, Noaea mucronata-Artemisia monosperma; V, Echiochilon fruticosum-Thymelaea hirsuta; VI, Cornulaca monacantha; VIIA, Anabasis articulata; VIIIB, Anabasis articulata (disturbed). SYN indicates syntaxa (sensu Danin and Solomeshch 1999) for which the species is characteristic: C, Class Retametea raetam; O, Order Erodio laciniatiStipagrostietalia plumosae; A1, Alliance Stipagrostio scopariae-Artemision monospermae; A2, Stipagrostio plumosae-Anabasion articulatae. Only once or twice occurred (with I 0): perennials: P Gynandiris sisyrinchium in IV and VIIA; G Stipagrostis ciliata in V; P Eminium spiculatum in V and VIIB; P Bellevalia eigii, P Leontice leontopetalum, P Moricandia nitens and P Peganum harmala in VIIB; annuals: Astragalus peregrinus in IV; Avena wiestii in V; Cynodon dactylon in V and VIIB; Asphodelus tenuifolius; Aizoon hispanicum and Silene sp. in VIIA; Koelpinia linearis, Oligomeris linifolia, Euphorbia grossheimii in VIIA and VIIB; Vicia monantha, Valerianella szovitsiana and Linaria albifrons in VIIB K. Tielbörger et al.
8 The Vegetation of the Nizzana Research Site
113
Table 8.2 Structural data on the plant communities of Nizzana (for community numbers, see Table 8.1; maximum values for each structural variable are in italics) Community
I
II
III
IV
V
VI
VIIA VIIB
Number of records Structural data Mean total plant cover/record (%) Mean cover perennials/record (%) Mean cover annuals/record (%) Cover perennials/cover annuals Life forms (%) Therophytes Graminoid perennials Perennial herbs and chamaephytes Nanophanerophytes Chorotypes (%) Mediterranean Irano-Turanian Saharo-Arabian Diversity and dominance Total species number in all records Mean species number / record Total no. of perennial species Mean no. of perennial species Shannon diversity, perennials Evenness, perennials Total no. of annual species Mean no. of annual species
21
8
26
19
37
25
13
16
23.7 14.3 9.4 1.52
33.8 26.3 7.5 3.51
23.5 18.2 5.3 3.43
43.6 32.1 11.5 2.79
22.8 16.7 6.1 2.74
33.2 26.4 6.8 3.88
10.9 5.7 5.2 1.10
68.9 1.9 67.5 0.03
55.7 12.8 31.5 0
61.2 7.9 29.5 1.4
58.6 7 30.9 3.5
55 6 37.5 1.5
48.1 7.5 30.7 4.7
56.1 6.5 34.2 3.2
60.5 2.2 35.2 2.1
74.1 1.2 24.2 0.5
19.2 10.8 70
23.7 17.9 58.4
26.5 10.4 63.1
22.2 14.2 61.3
20.5 10.2 67.6
19.1 11.7 67.8
19 12 67
16.1 17.5 66.4
47 12.2 16 4.7 1.675 60.4 31 7.5
42 16.7 14 6.1 1.599 60.6 28 10.6
71 20.7 25 7.8 1.907 59.2 46 12.9
79 23.7 28 8.9 2.266 68 51 14.8
89 27.7 32 12.7 2.781 80.2 57 15
72 20.9 26 7.7 1.428 43.8 46 13.2
93 18.2 32 5.8 0.978 28.2 61 12.4
70 22.7 12 2.6 0.86 34.6 58 20.1
The community is usually located between areas of mobile sand characterized by the Stipagrostis scoparia-Heliotropium digynum community, and areas with stable surface crust which are dominated by other plant communities. The occurrence is independent of the exposition of the slope. The physical environment is characterized by a thin surface crust which is covered by a more or less pronounced layer of mobile sand. The predominant soil types include arenosols both with (35%) and without (65%) a microbiotic crust, indicating the transient nature of the habitat between mobile and partly stabilized sands. The two dominating species form phytogenic hillocks of about 0.5 m in diameter which are regularly spaced. Mean plant cover per plot was 22% (Table 8.2) and, except for I. spicata, densities of annual species were much higher under the canopies of perennial plants than in the areas between these (Tielbörger and Kadmon 1997, and Chap. 27, this volume).
8.3.1.3
Echinops philistaeus Community
In the lower part of steep dune slopes with mobile sand, a plant community dominated by Echinops philistaeus can be found regularly. The perennial species Heliotropium
114
K. Tielbörger et al.
digynum, Cornulaca monacantha, Stipagrostis scoparia and Moltkiopsis ciliata occur regularly within the Echinops philistaeus community (Table 8.1). Most frequent annual species in this community in the year of study were Cutandia memphitica, Erodium laciniatum and Senecio glaucus. Absolute species number was the lowest for all plant communities of the research site, and especially the number of annual species was low (Table 8.2). The physiognomy of the community is characterized by narrow, dune-parallel stands of E. philistaeus. The plots were adjusted to this appearance by changing the shape of the plots from 10×10 m squares to dune-parallel 5×20 m rectangles. Mean plant cover per plot was 31% (Table 8.2). This plant community grows in mobile sand, and borders a plant community which is located at the dune bases and is dominated by Noaea mucronata (IV). Similar to the dune tops, all soil profiles within the Echinops community were classified as arenosols without surface crust.
8.3.1.4
Noaea mucronata-Artemisia monosperma Community
Noaea mucronata (Chenopodiaceae) and Artemisia monosperma (Asteraceae) dominate a plant community which occurs mainly at the dune bases. In addition, the community can be found in other concave-shaped areas in the interdunes, such as small depressions or channels. Other frequent perennial species in the Noaea mucronata-Artemisia monosperma community were Cornulaca monacantha, Moltkiopsis ciliata, Heliotropium digynum, Anabasis articulata, Atractylis carduus (Asteraceae) and Cyperus macrorrhizus (Table 8.1). The annual flora in the year of study was represented mainly by Senecio glaucus, Erodium laciniatum and Rumex pictus, the latter showing higher frequencies in this plant community than in other plant communities of the research site. The total number of species found was high, and the mean species number per plot was the second highest of all plant communities of Nizzana (Table 8.2). Mean plant cover per plot for the Noaea mucronata-Artemisia monosperma community was the highest observed for all plant communities of the study site (Table 8.2), reaching values of 90%. In addition to the high cover of vascular plants, a high cover of mosses and lichens was detected along the dune bases, too. There was an inverse relationship between the inclination of the slope and the width and density of the vegetation belt along the dune base. Along moderate slopes, the belt is less dense and up to 20 m wide whereas, along the base of very steep slopes, the Noaea mucronata-Artemisia monosperma community forms very dense and narrow dune-parallel strips. At the base of steep slopes with neighbouring Echinops philistaeus stands, the Noaea mucronata-Artemisia monosperma community is subdivided from the neighbouring community by a very distinct border. The classification of soils indicates that the Noaea-Artemisia community is located in the transition zone between mobile sands (arenosols without crust, 40%) and crusted sandy soils (40%). Almost 20% of the soil profiles located at the dune bases exhibited either high contents of salt (saline calcisols or solonchaks) and/or carbonate.
8 The Vegetation of the Nizzana Research Site
8.3.1.5
115
Echiochilon fruticosum-Thymelaea hirsuta Community
Large parts of the interdune corridors in Nizzana, which have a more or less stable surface crust, are characterized by a species-rich plant community of shrubs and semishrubs with a highly even species distribution (Table 8.2). Some perennials are exclusive to this community, and the overall combination of species differs considerably from the rest of the study site. Echiochilon fruticosum (Boraginaceae) and Thymelaea hirsuta (Thymelaeaceae) represent the two perennial species with the highest mean cover of those showing highest relative cover and constancy in the interdunes. Other such species are Stipagrostis plumosa, Asthenatherum forsskalii, Haplophyllum tuberculatum, Helianthemum sessiliflorum and Argyrolobium uniflorum (Table 8.1). The most frequent annual species were Bromus fasciculatus, Senecio glaucus and Erodium laciniatum. The total number of perennial species found was the highest for the area, and also the mean species number and Shannon diversity reached a maximum in this plant community (Table 8.2). The perennial plants of the Echiochilon fruticosum-Thymelaea hirsuta community grow at irregular distances from each other, and mean plant cover was estimated as 21%. Several perennials form small phytogenic hillocks, and some species grow up to 3 m high. Nanophanerophytes (Retama raetam, Lycium shawii) were more abundant in this community than in all the others (Table 8.2). Annuals grow mainly under perennial shrubs, in local depressions or out of animal burrows and fissures in the surface crust. The plant community of the interdune corridors is characterized by a patchy distribution of locally dominant perennial species. Also the soil types found here are very diverse. With the exception of the uncrusted arenosols, all soil types recorded for the site were frequently found in the Echiochilon-Thymelaea community. Their proportion of occurrence in the relative abundance was similar to the overall relative abundance of soil types.
8.3.1.6
Cornulaca monacantha Community
Large areas in the interdune corridors and along moderate slopes which are covered with a hard surface crust above cemented sand are dominated by Cornulaca monacantha. This species grows here at its eastern distribution limit within the Sinai sands. The only perennial species occurring regularly along with C. monacantha are Moltkiopsis ciliata and Convolvulus lanatus (Table 8.1). Annual species occurring in considerably high constancy were Senecio glaucus, Ifloga spicata, Picris asplenioides and Bromus fasciculatus. Both absolute and mean species numbers were intermediate in this plant community (Table 8.2). C. monacantha forms relatively homogeneous, dense and almost monospecific stands in areas with cemented sand. Individual plants reach between 2 and 5 m in diameter and 0.5 to 1.5 m in height, and grow at relatively regular distances from each other. Annual plant cover was relatively low (Table 8.2), and annuals grew mainly in the vicinity of the perennial shrubs. The soil types dominating this plant
116
K. Tielbörger et al.
community are arenosols with a prominent surface crust, and a relatively high salt content was found in 16% of the profiles in this community.
8.3.1.7
Anabasis articulata Community
A plant community which is strongly connected to the playa areas in the interdunes is dominated by the chenopod Anabasis articulata. Cornulaca monacantha is the only other perennial species occurring regularly within the playa community. Several annual species had a high constancy on the playas: Trigonella stellata, Stipa capensis, Bupleurum semicompositum, Anthemis melampodina, Carrichtera annua, Atriplex dimorphostegia, Herniaria hirsuta and Oligomeris linifolia. Shannon diversity was very low but the absolute species number was the highest detected for all plant communities of the Nizzana site (Table 8.2). The community may be subdivided into two physiognomically different subunits: the centre of the playas with very low perennial vegetation cover and the edges with a belt of relatively high cover of perennial plants. On and around the playas, A. articulata forms phytogenic hillocks of up to 2-m height and 5-m diameter. Individuals of annual plants were very small and growing mainly out of cracks in the playa surface. Overall annual plant cover was very low but could reach high local values at the playa edges, in depressions and in erosion channels produced by runoff water. Mean overall plant cover in the playa community was the lowest found for all plant communities of Nizzana (Table 8.2). The spatial correlation between the playa areas and the Anabasis articulata community is also reflected by the soil types. The high amount of fines in the playa soils leads to high evaporation and, consequently, high salt contents. Most soils in that community are solonchaks (61%) and saline calcisols (35%). Almost none of the profiles in that community were classified as arenosols (4%). Mainly at the eastern end of the study site, large flat areas were detected in the interdune corridors which are characterized by A. articulata as nearly the only perennial species, and strikingly high annual plant cover (up to 90%, Table 8.2). These areas were therefore treated separately within the playa community (VIIb). The structural data indicate that this community has many dominant features which distinguish it from the other plant communities: absolute and average plant cover was the highest measured, and the annuals dominated both in cover and species number over the perennial species (Table 8.2). The areas were identified as ‘disturbed’ because they show obvious signs of human impact such as traces of ploughing and tracks of heavy vehicles. Consequently, most of the soils in these areas were classified as anthrosols (75%), and only some (25%) as regosols. In the year of study, four perennial species (Bellevalia eigii, Leontice leontopetalum, Moricandia nitens, Peganum harmala) and several annual species (Emex spinosa, Leontodon laciniatum, Linaria albifrons, Malva parviflora, Urospermum picroides,
8 The Vegetation of the Nizzana Research Site
117
Valerianella szovitsiana, Vicia monantha) were found exclusively in the disturbed areas (Table 8.1).
8.3.1.8
Roadside Vegetation
In order to facilitate car access to the research site, in 1989 a track was built leading into the research area from the western end, using gravel from the nearby Wadi Nizzana. The vegetation growing on this track was treated separately because of the unique floristic composition within the research site. In all, 75% of the species recorded for the Nizzana site were detected on the track, indicating a considerably higher species diversity than in any of the plant communities described. As many as 32 species were found exclusively on the gravel road (Tielbörger et al., Chap. 7, this volume). They have either been introduced with the gravel or by vehicles driving on the track, and some have already started to disperse into the sandy areas bordering the track. One prominent example is Cannabis sativa which has been introduced into the research site by smugglers who have regularly made use of the track in order to reach the Egyptian border unnoticed by the Israeli army.
8.3.2
Annual Plant Distribution Patterns
Table 8.3 shows the density counts for all species which were abundant enough to yield useful results in the statistical tests. The data for permanent quadrats indicate that most annuals exhibit a strong preference for specific perennial plant communities. Except for one species (Rumex pictus), all observed annuals were significantly more abundant in specific communities in one or both years of study. Five species (Polycarpon succulentum, Senecio glaucus, Ifloga spicata, Erodium laciniatum, Trisetaria linearis) showed a preference for the plant community of the semi-stable slopes (dominated by Moltkiopsis ciliata and Convolvulus lanatus) in both study seasons. In addition, two of these species (Senecio and Ifloga) were by far the most abundant annuals in that plant community. A comparison with the data from 1992–1993 (Table 8.1) reveals that I. spicata could be identified as a marker species by its constancy values, too. In all, 13 species exhibited highest densities in the interdune areas during one or two seasons. However, there were only three species (Picris asplenioides, Bromus fasciculatus, Neurada procumbens) for which this habitat preference was significant in both seasons. B. fasciculatus had a preference for that plant community in 1992–1993, too (Table 8.1). There was no annual species which occurred preferably in the community dominated by Cornulaca monacantha. Anthemis melampodina, Trigonella stellata and Stipa capensis characterized the playa areas (Table 8.3). The latter species was even exclusive to that habitat type in both seasons.
118
Table 8.3 Average densities (±SE) of the most common annual plant species of Nizzana in four perennial plant communities and two consecutive growing seasons. Values in italics indicate preferences for a certain community. For a given year and species, values with common letters do not differ from each other (Kruskal-Wallis tests, p < 0.05). For community numbers, see Table 8.1 Perennial plant community Annual species Polycarpon succulentum Senecio glaucus Ifloga spicata Erodium laciniatum Trisetaria linearis Rumex pictus Cutandia sp. Centaurea pallescens Plantago cylindrica
Bromus fasciculatus Nigella arvensis
0.43±0.20 a 7.97±2.63 a 46.27±16.49 a 1.03±0.69 ab 0.27±0.13 a 0.10±0.06 a 0.10±0.06 a 0.20±0.20 b 0.00±0.00 b 0.17±0.11 b 1.67±1.27 b 0.00±0.00 a
1994–1995
V 1993–1994
1994–1995
VI 1993–1994
1994–1995
VIIA 1993–1994
1994–1995
1.33±0.54 A 7.13±2.09 A 38.47±11.05 A 0.20±0.09 A 0.97±0.34 A 0.03±0.03 A 0.03±0.03 B 0.00±0.00 B 0.10±0.10 AB 0.03±0.03 B 0.00±0.00 C 0.00±0.00 B
0.00±0.00 b 0.37±0.13 b 2.37±0.91 b 0.03±0.03 bc 0.07±0.05 ab 0.27±0.23 ab 0.10±0.06 a 1.77±0.66 a 0.60±0.27 a 1.13±0.47 a 20.45±7.53 a 0.03±0.03 a
0.03±0.03 B 0.03±0.03 C 1.97±0.96 C 0.00±0.00 B 0.07±0.05 B 0.07±0.05 A 0.17±0.08 A 0.00±0.00 B 0.17±0.07 A 0.93±0.24 A 5.63±1.49 A 0.17±0.11 A
0.00±0.00 b 0.08±0.04 c 0.10±0.07 c 0.03±0.03 bc 0.00±0.00 b 0.00±0.00 b 0.00±0.00 b 0.17±0.17 b 0.00±0.00 a 0.07±0.05 bc 0.13±0.08 b 0.00±0.00 a
0.03±0.03 B 0.08±0.04 C 0.40±0.25 D 0.00±0.00 B 0.00±0.00 B 0.00±0.00 A 0.00±0.00 B 0.00±0.00 B 0.00±0.00 a 0.23±0.15 B 0.57±0.44 B 0.00±0.00 B
0.00±0.00 b 0.03±0.03 c 0.00±0.00 c 0.04±0.03 c 0.00±0.00 b 0.00±0.00 b 0.00±0.00 b 0.26±0.13 b 0.00±0.00 a 0.00±0.00 c 2.98±1.58 b 0.00±0.00 a
0.00±0.00 B 0.10±0.07 C 0.00±0.00 E 0.00±0.00 B 0.00±0.00 B 0.00±0.00 A 0.02±0.02 B 0.36±0.15 A 0.00±0.00 a 0.06±0.04 B 0.54±0.30 B 0.00±0.00 B
K. Tielbörger et al.
Picris asplenioides
III 1993–1994
0.03±0.03 AB 0.00±0.00 A 0.00±0.00 A 0.77±0.51 AB 0.60±0.26 B 0.00±0.00 A 0.00±0.00 B 0.00±0.00 B 0.00±0.00 B 0.03±0.03 B 0.00±0.00 B
0.07±0.05 a 0.23±0.10 a 1.00±0.75 a 0.60±0.22 a 0.60±0.17 a 0.63±0.35 a 3.63±1.49 a 2.80±2.04 a 0.00±0.00 a 1.90±1.90 ab 0.00±0.00 b
0.43±0.27 A 0.00±0.00 A 0.17±0.14 A 0.60±0.25 A 2.07±0.50 A 0.10±0.06 A 3.93±1.59 A 2.63±2.21 AB 0.00±0.00 B 1.33±1.17 B 0.00±0.00 B
0.00±0.00 a 0.00±0.00 b 0.00±0.00 b 0.07±0.07 bc 0.07±0.05 bc 0.00±0.00 b 0.00±0.00 b 0.13±0.08 ab 0.00±0.00 a 0.00±0.00 b 0.00±0.00 b
0.00±0.00 B 0.00±0.00 A 0.00±0.00 A 0.00±0.00 B 0.17±0.14 C 0.00±0.00 A 0.03±0.03 B 0.17±0.11 AB 0.00±0.00 B 0.03±0.03 B 0.00±0.00 B
0.02±0.02 a 0.00±0.00 b 0.02±0.02 b 0.02±0.02 c 0.02±0.02 c 0.00±0.00 b 0.00±0.00 b 1.58±0.68 ab 0.04±0.03 a 2.48±1.06 a 2.54±0.84 a
0.02±0.02 B 0.00±0.00 A 0.00±0.00 A 0.00±0.00 B 0.17±0.14 C 0.00±0.00 A 0.02±0.02 B 0.52±0.25 A 0.20±0.09 A 2.52±0.93 A 3.48±1.16 A
8 The Vegetation of the Nizzana Research Site
0.00±0.00 a 0.03±0.03 Sixalix eremophila ab 0.00±0.00 Filago desertorum b 0.13±0.10 Lotus halophilus bc 0.13±0.06 Schismus arabicus b Crucianella membranacea 0.00±0.00 b 0.00±0.00 Neurada procumbens b Bupleurum semicompositum 0.00±0.00 b 0.03±0.03 Anthemis melampodina a 0.23±0.23 Trigonella stellata ab 0.00±0.00 Stipa capensis b
Hippocrepis areolata
119
120
8.4 8.4.1
K. Tielbörger et al.
Discussion Syntaxonomy
Eig (1938, 1946), Orshan and Zohary (1963) and Zohary (1982) classified the vegetation of the Nizzana Sands as belonging to an association dominated by Stipagrostis scoparia and Artemisia monosperma. Tielbörger (1997) has highlighted the fact that the results from the Nizzana research site suggest a much more differentiated view. More recently, the previously described Aristida scoparia-Artemisia monosperma association of Orshan and Zohary (1963) was given the rank of an alliance and split into six different associations (Danin and Solomeshch 1999). Since the work of Danin and Solomeshch (1999) is based partly on the findings of Tielbörger (1993, 1997), there are (not surprisingly) several similarities between some of the plant communities described. For example, association DSA 01 of Danin and Solomeshch (Heliotropio digyniStipagrostietum scopariae) is identical to the Stipagrostis-Heliotropium community described previously by Tielbörger (1997). Another overlap is between the Echinops philistaeus community of the dune base and the Echinopo philistaeiMoltkiopsietum ciliatae (ass. DSA 03) of Danin and Solomeshch (1999). Despite these similarities, the overall findings indicate that the Nizzana site seems to represent a rather unique vegetation structure, and integration into the novel syntaxonomic system fails in most occasions. Some similarities exist between the Stipagrostio plumosae-Convolvuletum lanati and the community of the slopes (Moltkiopsis ciliata-Convolvulus lanatus) but, at the research site, Stipagrostis plumosa is absent from these areas. Our Cornulaca community correlates with the Atractylo carduiCornulacetum monacanthae association (ass. DSH 02) but Atractylis carduus is uncommon in that community at our site. Despite the absence of Ifloga spicata, which in our study had a preference for the dune slopes, the playa community exhibits some overlap with the Iflogo spicatae-Anabasietum scopariae of Danin and Solomeshch (1999). The Convolvulo lanati-Moltkiopsietum ciliateae association has, despite the name, more in common with the community of the interdunes than with the Moltkiopsis-Colvolvulus community of the slopes. The largest mismatch between Danin and Solomeshch’s classification and our results relates to the fact that Artemisia monosperma is relatively uncommon at Nizzana. Therefore, we could not find any correlates to the associations characterized by A. monosperma (ass. DSA 02 and DSA 04) at the research site. Altogether, the high frequency of Cornulaca monacantha coupled with low densities of Artemisia monosperma appear to be unique within the plant communities of Israel. Since plants in arid regions are limited by essentially the same resources (water and, to a lesser degree, nutrients), the European system of classifying plant communities by diagnostic species is not very useful for desert vegetation. Therefore, our classification scheme is based mainly on dominance. However, the mismatch between our findings and those of Danin and Solomeshch (1999) indicates that random factors such as dispersal limitation and random local resource availability may largely determine
8 The Vegetation of the Nizzana Research Site
121
the occurrence and abundance of plant species in desert ecosystems. Therefore, syntaxonomic systems may have less predictive value than in less variable and more benign environments.
8.4.2
Annual Vegetation
The study was based on the methodology of Zohary and Orshansky (1949) and Danin et al. (1964), who suggested that the main criterion for distinguishing plant communities in arid environments should be the dominance of perennial species. However, the results of this study indicate that also annual species may characterize desert plant communities. Almost all annual species showed considerably higher frequencies and densities in certain habitat types, and some species were even exclusive to only one plant community. However, it has to be emphasized that a classification based on annuals alone would not necessarily lead to the same findings. Our most important finding in this context is that patterns of between-habitat differences in the abundance of annual species were consistent over several years. This contradicts what has been regarded as a major caveat of using annuals for characterizing desert plant communities. More recently, annuals have been utilized for establishing a new syntaxonomic system for Israel (Danin and Solomeshch 1999). Our findings suggest that this approach may indeed be justified and useful for future phytosociological studies in arid regions.
8.4.3
Vegetation and Habitat
The overall results of this study suggest that soil types are a relatively weak predictor for the distribution patterns of plants. This is mainly due to the fact that, except for the playa areas, most of the soils are arenosols with little diagnostic features (Beyer et al. 1998). Therefore, the plant community growing on and around the playas is well distinguished from the other plant communities, while the latter are not easily distinguished by their soil type. However, there is one dominant soil feature which may explain much of the variation in plant distribution patterns. Our overall results suggest that the dynamics and structure of the soil surface are major factors determining the distribution pattern of higher plants at the Nizzana research site (see also Beyer et al. 1998). For example, the occurrence of the plant communities of the dune tops (I) and slopes (III) corresponded to mobile sand in two degrees of stabilization (Allgaier, Prasse and Tielbörger, unpublished data). The fact that sand stability is a major factor influencing the distribution pattern of plants in sandy desert habitats has been repeatedly emphasized (e.g. Danin 1978, 1983, 1996). Another example for the importance of the soil surface in determining vegetation patterns is the community dominated by Cornulaca monacantha (VI), which grows on a unique type of hard surface crust. Also, the habitat of the interdune community, which was the
122
K. Tielbörger et al.
most diverse both in mean species numbers and in evenness, exhibited a large variety of small-scale differences in soil surface structure. This indicates that the high diversity in that community may be the result of a large number of available microsites for plant establishment. Our results therefore corroborate previous findings of Prasse and Bornkamm (see Chap. 23, this volume), who have emphasized the importance of surface structure and crust types in determining distribution patterns in annual vegetation. Surface structure and sand stability are influenced by the catenary position on the dune, the inclination and relief of the substrate, the development and structure of the biological surface crust, and the texture of the soil. Therefore, the overall distribution patterns of higher plants at the Nizzana site can be correlated with the distribution of different combinations of such environmental factors. In the sands of Nizzana, the various combinations of these abiotic factors are not confined to any specific exposition. This finding contradicts results from other studies which describe differences in plant cover and species composition between the two opposite slopes of desert sand dunes (Batanouny and Hilli 1973; Jenny and Smettan 1991). Between plant communities, differences in relative plant cover may be partly determined by differential water and nutrient availability. Plant cover is high along the dune bases, which probably receive higher amounts of water and nutrients by runoff and subsurface flow (Yair and Shachak 1987; Yair 1990; Kidron 1995). On the playas, low vegetation cover may partly be due to high water-holding capacity of the loamy substrate and, on the dune tops, low plant cover may be partly explained by the low nutrient content of the mobile sand (Beyer et al. 1998) and the lack of water due to high infiltration rates. The high species numbers on the playas may be partly explained by dispersal from the higher surrounding areas. The plant communities of the dune tops and the playas are both characterized by a relatively low plant cover and a low ratio of perennial to annual cover. In a previous study, Tielbörger (1993) showed that these two communities are the most dissimilar from each other and from the rest of the plant communities. We therefore suggest that the plant communities of the tops and the playas may reflect the two extremes of an environmental gradient. Environmental conditions at the extremes are unfavourable due, on the one hand, to mobile sand (Allgaier, Prasse and Tielbörger, unpublished data) and, on the other hand, to low water availability and the presence of a sealed surface (Prasse 1999). Stipagrostis scoparia, the dominant species on the mobile tops, has been suggested to function as a pioneer species in a process of dune stabilization (Danin 1996; Bornkamm et al. 1999).
8.5
Conclusions
Our overall findings suggest that the structure and mobility of the soil surface is the overriding factor determining the distribution and abundance of plants at the Nizzana research site. Unlike in other desert ecosystems, water availability, which may be
8 The Vegetation of the Nizzana Research Site
123
affected by exposition or water redistribution processes, plays a secondary role for the vegetation pattern. Water availability is less important because evaporation loss is low, and water redistribution occurs only where a microbiotic crust covers the surface (Chaps. 10 and 20, this volume). These results probably reflect the unique characteristics of sandy substrates in deserts.
References Baierle HU (1993) Vegetation und Flora im südwestlichen Jordanien. Diss Bot vol 200. J. Cramer, Berlin Batanouny KH, Hilli MR (1973) Phytosociological studies of Ghurfa desert, central Iraq. Phytocoenologia 1:223–249 Beyer L, Tielbörger K, Blume HP, Pfisterer U, Pingpank K, Podlech D (1998) Geo-ecological soil features and the vegetation pattern in an arid dune area in the Northern Negev, Israel. Zeitschr Pflanzenernäh Bodenk 161:347–356 Bornkamm R, Darius F, Prasse R (1999) On the life cycle of Stipagrostis scoparia hillocks. J Arid Environ 42:177–186 Braun-Blanquet J (1964) Pflanzensoziologie, 3. Aufl. Springer, Wien New York Danin A (1978) Plant species diversity and plant succession in a sandy area in the Northern Negev. Flora 167:409–422 Danin A (1983) Desert vegetation of Israel and Sinai. Cana, Jerusalem, pp 1–148 Danin A (1996) Plants of desert dunes. In: Cloudsley-Thompson JL (ed) Adaptations of organisms to the desert. Springer, Berlin Heidelberg New York Danin A, Orshan G (eds) (1999) Vegetation of Israel. I. Desert and coastal vegetation. Backhuys, Leiden Danin A, Solomeshch (1999) Synopsis of the vegetation and enumeration of the associations. In: Danin A, Orshan G (eds) Vegetation of Israel. I. Desert and coastal vegetation. Backhuys, Leiden, pp 35–317 Danin A, Orshan G, Zohary M (1964) Vegetation of the neogene sandy areas of the Northern Negev. Israel J Bot 13:208–233 Eig A (1938) On the phytogeographical subdivision of Palestine. Palestine J Bot Jer Ser 1:4–12 Eig A (1946) Synopsis of the phytogeographical units of Palestine. Palestine J Bot Jer Ser 3:183–246 Feinbrun-Dothan N, Danin A (1991) Analytical flora of Eretz Israel. Cana, Jerusalem Jenny M, Smettan U (1991) Distribution patterns of plants on a sand dune and the adjacent playa in the Wadi Araba (Jordan). Flora Vegetatio Mundi 9:155–166 Kidron GJ (1995) The impact of microbial crust upon rainfall-runoff-sediment yield relationships on longitudinal dune slopes, Nizzana, Western Negev Desert, Israel (in Hebrew with English abstract). PhD Thesis, Hebrew University of Jerusalem Orshan G, Zohary M (1963) Vegetation of the sand deserts in the Western Negev of Israel. Vegetatio 11:112–120 Prasse R (1999) Experimentelle Untersuchungen an Gefäβpflanzenpopulationen auf verschiedenen Geländeoberflächen in einem Sandwüstengebiet. Universitätsverlag Rasch, Osnabrück Tielbörger K (1993) Vegetationskundliche Studien im Sandgebiet von Nizzana in der westlichen Negev-Wüste Israels. Diploma Thesis, Institut für Systematische Botanik, LudwigMaximilians-Universität München Tielbörger K (1997) The vegetation of linear desert dunes in the north-western Negev, Israel. Flora 192:261–278 Tielbörger K, Kadmon R (1997) Relationships between shrubs and annual communities in a sandy desert ecosystem: a three-year study. Plant Ecol 130:191–201
124
K. Tielbörger et al.
Yair A (1990) Runoff generation in a sandy area: the Nizzana sands, Western Negev, Israel. Earth Surface Processes Landforms 15:597–609 Yair A, Shachak M (1987) Studies in watershed ecology of an arid area. In: Berkovsy L, Wurtele MG (eds) Progress in desert research. Rowman and Littlefield, Totowa, NJ, pp 209–220 Zohary M (1944) Vegetational transects through the desert of Sinai. Palestine J Bot Jer Ser 3:57–78 Zohary M (1982) Vegetation of Israel and adjacent areas. In: Blume HP, Frey W (eds) Beihefte Tübinger Atlas Vorderer Orient, Reihe A (Naturwissenschaften) 7. Reichert, Wiesbaden Zohary M, Orshansky G (1949) Structure and ecology of the vegetation in the Dead Sea region of Palestine. Palestine J Bot Jer Ser 4:177–206
Chapter 9
A Glance on the Fauna of Nizzana J. Filser and R. Prasse
9.1
Introduction
Desert ecosystems are inhabited by a highly specialised fauna adapted to extremely low water availability, high UV radiation, and temperature extremes. Soil-inhabiting mammals, particularly rodents, constitute large and diverse communities in most desert ecosystems (e.g. Kelt et al. 1996; Whitford and Kay 1999). Besides these and larger mammals, the most obvious animal groups in deserts are reptiles, birds, tenebrionid beetles and ants; yet, the fauna in most arid ecosystems is indeed far more complex (e.g. Wallwork 1982; Cloudsley-Thompson 1996). A review on small mammals revealed a very high habitat diversity of the Negev, compared to other deserts (Kelt et al. 1996). Since Nizzana only partly overlaps with other parts of the Negev, a characteristic and partly endemic fauna can be expected also here. Most zoological studies have been conducted in the rocky parts of the Negev, or at other sites such as arid shrubland (e.g. Wilby et al. 2001). Darkling beetles and granivorous mammals and birds in sandy areas were investigated by Ayal and Merkl (1994), Garb et al. (2000) and Lortie et al. (2000). Thus far, almost nothing has been published on the fauna of Nizzana, with two exceptions: Mahn (1994) surveyed darkling beetles, and Henschel (1998) sand-burrowing spiders. Due to this large deficit in knowledge, we would like to provide some preliminary information on the fauna in Nizzana, with special emphasis on diversity, ecosystem engineering and spatial variability of invertebrates. We present observations of larger animals (vertebrates and invertebrates) made during many years of research activities by Rüdiger Prasse, and invertebrate surveys during two short visits by Juliane Filser. From these data, we derive research needs at Nizzana.
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
125
126
9.2 9.2.1
J. Filser, R. Prasse
Study Sites and Methods List of Vertebrates and Larger Invertebrates
Birds, mammals, reptiles and larger arthropods were recorded non-systematically during several visits to the AERC research site (N1) between 1993 and 2000, usually between November and April. While doing population studies on vascular plants, RP recorded all species of animals seen by chance and which he could identify. Therefore, the data are strongly biased towards less elusive animals such as birds.
9.2.2
Invertebrate Survey
Two invertebrate surveys were conducted during spring at Nizzana 1=N1 and at Nizzana 2=N5 (see Chap. 29, this volume), differing mainly in annual rainfall (90 and 150 mm respectively) and, accordingly, vegetation patterns. The average winter temperature had been higher in 2000/2001 than in 1999/2000, and the total precipitation was almost twice as high (62 mm) in the second winter (Chap. 4, this volume). Therefore, the vegetation period had been in a later stage, and plant cover was less lush in 2000 than in 2001. The 2000 survey, conducted from 4–10 April, aimed for a first general overview on invertebrate diversity at Nizzana, using various non-quantitative methods. Nizzana 1 and 2 were surveyed systematically, trying to cover as many different locations (e.g. dune ridges, slopes, valleys, playa, roads) and microsites (e.g. in the vicinity of various shrubs, among different communities of annual plants or in the vicinity of a bee colony) as possible. Animals were surveyed by direct observation, sweep-netting, soil digging and pitfall traps. For the latter, one trap (Fig. 9.1) was dug in the soil at eight different locations, even with the surface, and sampled after 1 and 4 days. No preservation fluid was used, which caused some visible within-trap predation and certainly also some escaping. Due to the short time available and the enormous diversity of invertebrates1, identification had to be done on a rough systematic level. All animals were assigned to order, where possible to lower systematic units, and additionally separated into morphospecies. This concept has been widely used in assessing biodiversity of unknown or highly diverse areas (e.g. Barratt et al. 2003). A second survey was carried out from 16–20 March 2001, together with students during a field school project (Soil Ecology in a Desert Ecosystem). Here, special attention was drawn to spatial heterogeneity. This survey was restricted to Nizzana 1. Polyethylene flasks connected to plastic funnels served as pitfall traps containing
1 For each single order, often family, of insects, a specialist is required for reliable species identification
9 A Glance on the Fauna of Nizzana
127
Fig. 9.1 A pitfall trap dug under a Retama raetam bush in 2000 (note that all photographs in this chapter are by JF, except for those in Figs. 9.2a, b and 9.4a, b which are by RP)
ca. 20 ml ethylene glycol as collection fluid. The traps were placed at nine locations along a N–S gradient between two dune ridges, with three sub-samples each, and sampled after 3 days. The locations were named as follows: north dune slope: ABN, AIN, RBN, RIN; playa: P; south dune slope: ABS, AIS, RBS, RIS, where A stands for Anabasis articulata, R for Retama raetam, B for ‘under shrub’ and I for ‘inter-space close to shrub’. There was no vegetation in the playa. Due to the limited experience of the students, invertebrates were assigned only to easily identifiable orders. Within each order, three size classes (large, medium, small) were distinguished. In the following text, the resulting entities are referred to as ‘morphotaxa’. Larvae were regarded as one morphotaxon. Additionally, a preliminary survey of the microfauna was done at the same locations (except for the playa), at two soil depths (0–1 and 1–2 cm). A soil extract solution was prepared for each location: 50 g of fresh soil was mixed with 500 ml distilled water, boiled for 10 minutes, filtered and autoclaved. Before counting, the solution was diluted 1:3 with distilled water, and the pH was adjusted to the value of the soil sample using HCl or NaOH. For the investigation of Protozoa, the soil was taken to the laboratory and wetted for 1 day. After that, 0.1–0.5 g of soil was mixed with a few drops of soil extract solution and examined drop by drop at 63–160x magnification. Protozoans were separated from soil by means of a micro-pipette and examined under a cover slip (provided with small wax ‘feet’) at higher magnification. Nematodes were extracted by means of Baermann funnels (Dunger and Fiedler 1989): plastic tubes were connected to the funnels and closed with a clamp. Fresh water was filled into the funnels before inserting a sieve with cheese cloth on which 10 g of soil was carefully spread, securing that the soil was in direct contact with the water but not completely covered by it. A lamp on top of each funnel was
128
J. Filser, R. Prasse
switched on for 3 days. After that, a collecting vessel was held under the tube, the clamp quickly opened and then closed as soon as 1–2 cm of the animal/water suspension was in the vessel. Nematodes were counted under a microscope.
9.3 9.3.1
Results Composition of the Fauna
Between 1993 and 2000, more than 12 mammal taxa2, more than 13 reptile taxa3 (Fig. 9.2a, b), about 15 Orthoptera (Fig. 9.3a, b) and Mantodea species, and over 80 species of birds4 were recorded (Tables 9.1, 9.2, 9.3). While most mammals, reptiles, Orthoptera and Mantodea are well established in the area, the list of birds comprises relatively few residential species, some breeding guests, and a long list of species utilising the area mainly during spring and autumn migration. The most characteristic residential birds are the crested lark (Galerida cristata), scrub warbler (Scotocerca inquieta), great grey shrike (Lanius excubitor) and several species of sandgrouse. Large flocks of the common crane (Grus grus) and white storks (Ciconia ciconia) passed by during migration. Among reptiles, lizards are most commonly observed, most easily when moving on vegetation-free dune slopes. Camouflage occurs not only in lizards and chameleons (Fig. 9.2b) but also in many arthropods: Tmethis pulchripennis (Fig. 9.3a) mimics sand so perfectly that it is almost impossible to detect it when it is immobile. Other species, such as Acinipe zebratus (Fig. 9.3b), hide in the vegetation and display contrast-rich colour patterns. Most carabids (Fig. 9.4a, b) and other large predatory beetles are night-active and, thus, more elusive during daytime. An exception is Adesmia metallica (Fig. 9.4c), a fast-moving tenebrionid which is highly active during daytime in spring. The short invertebrate survey in 2000 also revealed a surprisingly diverse fauna, especially at Nizzana 2 (Table 9.4; also see below and Table 9.5). In all, 56 different morphospecies were recorded, most occurring frequently or at least regularly. Remarkably, most of these were found at only one of the two sites. (In 2001, many additional taxa, particularly grasshoppers, bees and butterflies, were observed but these were not systematically recorded). In terms of abundance, ants and grasshoppers dominated both sites but flies and bees also occurred in large numbers. Concerning within-group diversity, Coleoptera, Chelicerata, Hymenoptera and
2
Sorex, Meriones and Gerbilus not identified to species level Acantodactylus and Stenodactylus not identified to species level. More complete and more systematic data on reptiles of Nizzana have been collected by the workgroup of Prof. Dr. Jehuda Werner, HUJI Jerusalem, E.S.E Department 4 More than 110 species, if the surroundings of the research site would be included 3
9 A Glance on the Fauna of Nizzana
129
Fig. 9.2 Reptiles. A Trapelus savignyi, B Chamaeleon chamaeleon
Diptera had the largest number of morphospecies. Ants differed remarkably in size, habitat preference and behaviour: the large harvester ants Messor sp. were observed mostly foraging solitarily. The small, red ‘ant morphospecies B’ frequently jumped, and the even smaller, black ‘ant morphospecies C’ was usually found on very fine sand – extremely abundant in amberlite containers which had been buried by the soil scientists. We suspect that they had probably fed on the microflora which must have grown on the pellets (due to the accumulated nutrients), since the containers were almost completely emptied as we dug these out. Collembola were found only very rarely, mostly in litter, whereas mites occurred more frequently. Another group of litter decomposers, termites, were not observed at all in 2000, but a small colony of whitish, hardly sclerotised individuals was found under a trapping stone in 2001.
130
J. Filser, R. Prasse
Fig. 9.3 Grasshoppers. a Tmethis pulchripennis asiaticus, b Acinipe zebratus
9.3.2
Ecosystem Engineering
The most impressive observations associated with invertebrates at Nizzana was the large amount and variety of characteristic soil structures shaped by these animals. High densities of invertebrate burrows structure the soil both vertically (Fig. 9.5a) and horizontally (Fig. 9.5b). Larvae of ant lions (Myrmeleonidae) construct perfect funnels (Fig. 9.5c) which locally covered a large percentage of the total area (Fig. 9.5d). The wasp in Fig. 9.5e dug out and moved sand and small stones at an enormous speed. At the entrance of many invertebrate burrows or close to plants, a great diversity of differently shaped ‘Nescafé’ aggregates was found (Fig. 9.6a–c), particularly in spring 2000. The 2001 observations, carried out somewhat earlier in the vegetation period, partly solved this riddle: the fresh faecal pellets found in the vicinity of caterpillars feeding on fresh plant material (Fig. 9.6d) exactly matched some of
9 A Glance on the Fauna of Nizzana
131
Table 9.1 List of birds observed by RP at the Nizzana research station of the AERC between 1993 und 2000 (all own observations by chance, no systematic search for species, observation period always between December and early May) English name Latin name Remarks White stork
Ciconia ciconia
Honey buzzard Black kite Egyptian vulture Hen harrier Pallid harrier Common buzzard Long-legged buzzard Osprey Steppe eagle Booted eagle Kestrel Chukar
Pernis apivorus Milvus migrans Nephron percnopterus Circus cyaneus Circus macrourus Buteo buteo vulpinus Buteo rufinus Pandion haliaetus Aquila nipalensis Hieraaetus pennatus Falco tinnunculus Alectoris chukar
Common crane
Grus grus
Houbara bustard
Chlamydotis undulata
Stone curlew
Burhinus oedicnemus
Cream-coloured coursor
Cursorius cursor
Common snipe
Gallinago gallinago
Green sandpiper
Tringa ochropus
Spotted sandgrouse
Pterocles senegallus
Crowned sandgrouse
Pterocles coronatus
Pin-tailed sandgrouse
Pterocles alchata
Black-bellied sandgrouse Pterocles orientalis
Appears during migration in numerous flocks, single flocks sometimes with up to 5,000 birds During migration Mainly during migration Vagrant During migration During migration Mainly during migration During migration Regular migrant During migration During migration Common hunter in the area Very common in the area and its surroundings, often in large flocks During spring migration, numerous flocks of up to several hundred individuals migrating via the research station, some of the flocks resting at night among the dunes Few in the research area, probably not breeding, disturbance?; several individuals in the vicinity, displaying males and female with chicks near the Nizzana Airfield Frequent at the AERC research site and in its vicinity Few in the immediate surroundings of Nizzana 1 On migration at the AERC research site, frequently at the sewage ponds at K’tziot Few at sewage ponds at K’tziot and after rainfalls on puddles in the vicinity of Nizzana 1 Wadi Passing through the research site, sewage ponds at K’tziot and surroundings Passing through the research site, sewage ponds at K’tziot and surroundings Uncommonly passing through the research site, at least twice at sewage ponds at K’tziot The most common sandgrouse species of the region; passing through the research area sewage ponds at K’tziot and surroundings (continued)
132 Table 9.1 (continued) English name
J. Filser, R. Prasse
Latin name
Remarks
‘Rock dove’
Columba livida and C. livida f. domestica
Turtle dove
Streptopelia turtur
Little owl
Athene noctua
Swift
Apus apus
Alpine swift European bee-eater Hoopoe
Apus melba Merops apiaster Upupa epos
Bar-tailed desert lark Desert lark Short-toed lark
Ammomanes cincturus Ammomanes deserti Calandrella brachydactyla
Lesser short-toed lark
Calandrella rufescens
Crested lark
Galerida cristata
Skylark
Alauda arvensis
Sand martin Barn swallow Red-rumped swallow House martin Tawny pipit Tree pipit Meadow pipit European robin Bluethroat
Riparia riparia Hirundo rustica Hirundo daurica Delichon urbica Anthus campestris Anthus trivialis Anthus pratensis Erithacus rubecula Luscinia svecica
Redstart Black redstart Blackstart Whinchat Stonechat Isabelline wheatear Northern wheatear Black-eared wheatear
Phoenicurus phoenicurus Phoenicurus ochruros Cercomela melanura, Saxicola rubetra Saxicola torquata Oenanthe isabellina Oenanthe oenanthe Oenanthe hispanica
Desert wheatear
Oenanthe deserti
Frequent in and near settlements as well as on the higher banks of Wadi Nizzana Common on the border fence and electricity lines during migration At least two breeding pairs at the AERC research site in old rodent burrows in ‘fossil dunes’, relatively common in the vicinity of Nizzana Common during migration; observed individuals probably also included pallid swifts; no effort was made to distinguish these species Few during migration Large flocks during migration Mainly on lawns in settlements but also at the research site Only once observed at the research site Frequently in the vicinity of the Wadi Sometimes larger flocks at the research site Often mixed flocks at the sewage ponds at K’tziot and surroundings, sometimes also at the AERC research site Very common breeding bird at the research site Large flocks on migration, some also winter guests? Migration Migration Migration Migration Migration Migration Migration Common, winter guest and migration Regular migrant, at the research station and nearby settlements Migrant Migrant Acacias in the Wadi Nizzana Common winter guest und migrant Common winter guest und migrant Common breeding bird Common migrant light-coloured and dark morphs, breeding?, migrant? Regular, breeding? (continued)
9 A Glance on the Fauna of Nizzana
133
Table 9.1 (continued) English name
Latin name
Remarks
Mourning wheatear Scrub warbler
Oenanthe lugens Scotocerca inquieta
Spectacled warbler
Sylvi conspicillata
Rüppels’s warbler Desert warbler
Sylvia rueppelli Sylvia nana
Lesser whitethroat Whitethroat Blackcap Chiffchaff Willow warbler Southern grey shrike
Sylvia curruca Sylvia communis Sylvia atricapilla Phylloscopus collybita Phylloscopus trochilus Lanius meridionalis
Woodchat shrike
Lanius senator
Raven Brown-necked raven
Corvus corax Corvus ruficollis
Chaffinch Goldfinch Linnet
Fringilla coelebs Carduelis carduelis Carduelis cannabina
Desertfinch
Rhodospiza obsoleta
Trumpeter finch
Bucanetes githagineus
Ortolan bunting Cretzschmar’s bunting
Emberiza hortulana Emberiza caesia
Regular breeding bird? Probably the most common breeding bird at the research site Uncommon breeding bird at the AERC research site Rare migrant at the research site Very rare migrant, in 2000 one individual at the AERC research site Common migrant Common migrant Common migrant Migrant Migrant Common breeding bird at the AERC Research site Breeding bird at the AERC research station Rare visitor at the AERC research site Common visitor at the AERC research site Wintering, mainly in settlements Common, migrant? Appears at the AERC research station sometimes in large flocks, migrant? Appears in spring sometimes in large flocks, migrant? breeding bird? Only three or four observations at the AERC research site Migrant Frequent migrant
Table 9.2 List of mammals and reptiles (the latter were observed only at the AERC research site; for observation period and strategy, see Table 9.1) Latin name Remarks Taxon: mammals One road kill near the entrance to Nizzana 1 Canis aureus Once 1 specimen seen on roadside about 20 km south Canis lupus of the area Few cat skulls found in the dunes, not yet identified Felis sp. Frequent in the area, common in fields of the nearby Gazella gazella Moshavim Numerous dead individuals in plastic buckets used as Gerbilus indet. div. runoff catchments Tracks in the area and one individual seen crossing Hyaena hyaena the road near the K’tziot gas station Signs of diggings, spines Hystrix indica Some dead individuals in plastic buckets used as runJaculus jaculus off catchments (continued)
Table 9.2 (continued) Latin name Lepus capensis Meriones indet. Mus musculus Paraechinus aethiopicus Sorex indet. Taxon: reptiles Scincus scincus Chalcides ocellatus Sphenops sepsoides Acanthodactylus div. Varanus griseus Testudo kleinmanni Cerastes vipera Psammophis schokari Ptyodactylus hasselquistii hasselquistii Stenodactylus indet. Hemidactylus turicus Trapelus (Agama) savignyi Chamaeleo chamaeleon
Remarks Common Numerous dead individuals in plastic buckets used as runoff catchments Container of the old police station Few Numerous dead individuals in plastic buckets used as runoff catchments
Container at the old police station Photographs available Runoff buckets Fig. 9.2a Fig. 9.2b
Table 9.3 List of larger insects – a very unsystematic collection of information, there being many more species in the area (for observation period and strategy, see Table 9.1) Latin name Remarks Taxon: Orthoptera, Caelifera Anacridium aegyptium Pygromorpha conica cf. Fig. 9.3a Tmethis pulchripennis asiaticus Fig. 9.3a Acinipe zebratus No effort made to distinguish Truxalis and/or Acrida bicolo Only once, a few specimens Locusta migratoria Hyalorrhipis calcarata Oedipoda indet. Taxon: Mantodea det. R. Ehrmann, Karlsruhe, Germany Eremiaphila rufipennis Rivetina baetica Mantis religiosa Ameles c.f. heldreichii Iris oratoria Empusa fasciata Blepharopsis mendica Taxon: Odonata Common on passage Anax (Hemianax) ephippiger Common on passage Sympetrum fonscolombei Taxon: Coleoptera (without Tenebrionidae) Anthia sexmaculata Fig. 9.4a Graphopterus serrator Fig. 9.4b Scarites sp.
9 A Glance on the Fauna of Nizzana
135
Fig. 9.4 Beetles. a Graphopterus serrator; b Scarites sp., Carabidae; c highly active, long-legged tenebrionid (cf. Adesmia metallica)
136
J. Filser, R. Prasse
Table 9.4 Number of invertebrate morphospecies recorded in April 2000, based on pitfall traps, sweep-netting, direct observation and soil digging at sites N1 (Nizzana) and N5 Number of morphospecies Taxon Subtaxon at site N1 (Nizzana) N5 N1+N5 Gastropoda 1 1 1 Sphincterochila sp. Chelicerata Acari 2 2 Araneae 3 3 6 Pseudoscorpiones 1 1 Scorpiones 1 1 Myriopoda Scolopendridae 1 1 Collembola Entomobryidae 1 1 Zygentoma Lepismatidae 2 1 3 Blattodea 1 1 Homoptera Aphidina 1 1 Heteroptera cf. Nabidae 1 1 Mantodea 1 1 Odonata 1 1 Orthoptera Caelifera 1 2 2 Psocoptera 2 2 Neuroptera 1 1 1 Myrmeleon sp. Lepidoptera Geometridae 1 1 Psychidae 1 Rhopalocera 1 1 Hymenoptera Apidoidea 1 1 Formicoidea 3 2 4 Eumenidae 1 Sphecidae 1 1 Coleoptera Tenebrionidae 2 1 3 Other families 2 7 9 Diptera Bombylidae 4 4 Asilidae 1 1 Syrphidae 2 2 Other Brachycera 2 2 Total 25 33 56
these ‘Nescafé’ pellets in shape (though at first not in colour, due to the higher water content). Moreover, snails deposited small aggregates of slime and sand alongside their tracks when creeping on the sand (Fig. 9.6e).
9.3.3
Spatial Variability
The spatial variability of invertebrates is exemplified with a gradient study on microfauna and epigeal arthropods along a north–south gradient between two dunes, crossing a playa in the centre of the dune valley. The pitfall traps caught a
9 A Glance on the Fauna of Nizzana
137
Table 9.5 Activity density of tenebrionid species along a dune gradient in spring, compiled after Mahn (1994). Total numbers caught during April and May 1993 in ten traps each at three locations (north slope, playa, south slope) at Nizzana 1. The traps were emptied daily during the first 2 weeks, thereafter at weekly intervals Species North slope Playa South slope Adelostoma sulcatum Adesmia carinata Adesmia dilatata Adesmia metallica Anthrodeis cruciatus Erodius opacus Erodius puncticollis Eurycaulus hirsutus Micipsa schaumi Omophlus sp. Oterophloeus haagi Pimelia angulata Pimelia arabica Pimelia barthelemi Pimelia cornata Pimelia jansseni Prionotheca carinata Tentyrina boehmi Zophosis bicarinata cog. Zophosis complanata Total individuals Number of species
7 80 4 189 34 8 217 3 2 1 5 1 6 23 1 6 3 6 39 6 641 20
0 13 9 24 0 4 111 0 0 0 0 0 9 0 1 6 1 0 31 3 212 11
0 56 1 74 4 3 364 0 1 0 0 2 8 14 0 5 3 0 15 5 555 14
total of 474 animals, amounting to an average activity density of 17.6 animals per trap. The students distinguished 23 different ‘morphotaxa’. The overall activity density of animals was highest in the playa where, by contrast, diversity (expressed as number of morphotaxa) was lowest (Fig. 9.7). Diversity was higher on the north slope than on the south slope. No clear preferences for either shrub or intershrub patches, or for either Retama or Anabasis were found (Fig. 9.7). Cicada, Nematocera, medium-sized Brachycera and medium-sized ants were found almost exclusively on the north slope whereas mites were more abundant on the south slope (Figs. 9.8, 9.9). The activity density of ants varied markedly along the dune gradient: small ants were trapped mostly in the playa and showed higher densities on the south slope than on the north slope, whereas large ants showed the reverse pattern (Fig. 9.8). Nematodes and protozoans were concentrated between 1 and 2 cm soil depth (Table 9.6). At the locations RIS and RBN, nematodes were extracted also between 2 and 4 cm, albeit resulting in a total of only two individuals (data not shown), hinting at a strong aggregation between 1 and 2 cm depth. Nematodes tended to be more numerous on the north slope but variability was extremely high.
138
J. Filser, R. Prasse
Fig. 9.5 Examples of soil structures formed by invertebrates at Nizzana. a Soil profile (depth ca. 50 cm) with burrows constructed by small mammals and invertebrates; b a ca. 2-cm-thick piece of crust from Nizzana 2 viewed from below, showing numerous invertebrate burrows;
9 A Glance on the Fauna of Nizzana
139
Fig. 9.5 (continued) c, d ant lion traps, detail and spatial distribution; e digging wasp excavating a burrow
9.4 9.4.1
Discussion Diversity
Our limited faunistic surveys at Nizzana revealed an overall high diversity and, at least during spring, remarkably high abundances. Since Nizzana is situated along a main migratory route, large bird numbers had been expected – yet the high number of species recorded is impressive. Also the range of other vertebrates demonstrates a broad representation of feeding guilds, particularly predators. In March 2000, a total of 56 invertebrate morphotaxa were found, the majority being specific either to site N1 or to N5 (see Fig. 29.1, Chap. 29, this volume). This pronounced differentiation hints at a high habitat diversity, which is typical of the Negev (Kelt et al. 1996). Our survey certainly strongly underestimates the actual invertebrate diversity of Nizzana. A hint of this can be gained by comparing our data to the results of Mahn (1994; cf. Table 9.5): he
140
J. Filser, R. Prasse
Fig. 9.6 Different types of soil aggregates formed by invertebrates. a–c Heaps of aggregates surrounding burrow entrances. d Track of a snail moving on a dune ridge towards its food plant. Note the characteristic, asymmetric shape which is formed by slime/sand aggregates dropped to the right side every few cm. e A nymphalid caterpillar and its faeces
9 A Glance on the Fauna of Nizzana
141
individuals / trap
Fig. 9.6 (continued)
35,0
No. Morphotaxa
30,0
Ind. No.
25,0 20,0 15,0 10,0 5,0 0,0 ABN
AIN
RBN
RIN
P
ABS
AIS
RBS
RIS
Fig. 9.7 Activity density and number of morphotaxa caught in pitfall traps along a dune gradient, mean values ± SE (n=3). A, Anabasis articulata; R, Retama raetam; N north dune slope, S south dune slope, P playa, B under bush, I inter-space close to bush
collected tenebrionids with pitfall traps at weekly intervals at three sites (north slope, playa und south slope) of site N1 (Nizzana) between April and July 1993, and found 20 species with a high degree of endemism (compared to other parts of the Negev) – 10 times the number of tenebrionid morphospecies recorded in 2000. Still, Mahn (1994) considered his survey to be incomplete, due to
142
J. Filser, R. Prasse
25,0 s-ants m-ants l-ants
individuals / trap
20,0 15,0 10,0 5,0 0,0 ABN
AIN RBN RIN
P
ABS
AIS
RBS RIS
Fig. 9.8 Activity density of selected arthropod groups in pitfall traps, mean values ± SE (n=3). s Small, m medium-sized; l large; for other abbreviations, see Fig. 9.7
8,0
individuals / trap
7,0 6,0
s-midges m-flies mites
s-flies s-Cicada
AIN
RIN
5,0 4,0 3,0 2,0 1,0 0,0 ABN
RBN
P
ABS
AIS
RBS
RIS
Fig. 9.9 Activity density of ants in pitfall traps, mean values ± SE (n=3). s Small, m mediumsized, l large; for other abbreviations, see Fig. 9.7
the comparatively short observation period. Evidently, Nizzana is possibly a distinct, particularly diverse ‘hotspot’: indeed, during a 1-year survey, Ayal and Merkl (1994) found only 19 tenebrionid species in the neighbouring Mashabim Sands nature reserve. In addition to beetles, dipterans appeared to be rather diverse, too. Remarkably, five of nine morphotaxa belonged to the Bombylidae (often parasitic) and
9 A Glance on the Fauna of Nizzana
143
Table 9.6 Preliminary survey of Protozoa and Nematoda on south- and north-facing slopes at Nizzana (abbreviations as in Fig. 9.7). Nematodes: numbers per 10 g dry soil; protozoan abundance classes: -, zero; +, few; ++, many Depth South North Protozoa Nematoda
(cm)
AB
AI
RB
RI
AB
AI
RB
RI
0–1 1–2 0–1 1–2
− ++ 0 0
++ + 0 12
+ + 0 0
− + 0 4
− ++ 5 3
− ++ 1 60
++ + 2 16
− + 0 0
Asilidae (predacious). With six morphotaxa distinguished, spiders were among the most diverse groups. During a short-term survey in August and September 1991, Henschel (1998) found nine species of spiders at Nizzana, within 1 km2. Evidently, especially higher trophic levels are very diverse – which, however, may be a somewhat skewed pattern. The survey of 2000 revealed only two Orthopteran taxa, probably due to the rather late stage of the vegetation period. Both in 2001 and before 2000, more taxa of this mainly herbivorous group were recorded. Hymenoptera and Diptera have a large proportion of pollinators, the diversity of which peaked in March and April in another region of the Negev (Wolf and Shmida 1995) – yet in Nizzana in 2000, only one bee taxon was recorded. Most desert invertebrates live in the soil (Wallwork 1982). Ants showed large numbers of both individuals and morphotaxa. Termites are probably close to their tolerance limits at Nizzana, since they decreased in diversity and abundance along a gradient from humid rainforest to semiarid Sahel savannah in Africa (Wallwork 1982). The desert cockroach, Arenivaga sp., and all Lepismatidae were found only by digging (the latter sometimes in pitfall traps), and would disappear immediately once exposed to light. Collembola (only one taxon in both 2000 and 2001) are poorly represented in the Negev, compared to other parts of Israel (Gruia et al. 1999), and were found only in litter. Mites were recorded more frequently, pointing to their importance in arid and semiarid ecosystems (Whitford 1996). Nematode abundances were in the same range as in a loessial sierozem in the Negev (Alon and Steinberger 1999). A locally limited accumulation of organic matter in nutrient-poor soils leads to a high abundance of Protozoa (Verhoeven 2001) and nematodes. The protozoans recorded here contained different ciliate groups typical of nutrient-poor sandy soils: Colpodea are abundant in the algal layer, and have the potential for a very rapid reproduction (Verhoeven, personal communication) and transformation from active stages to drought-resistant cysts. Hypotrichs are represented by worm-like species of 50 up to several hundred micrometers. By means of their bundled cilia groups, they move through the water film on the surface of soil particles and graze on attached bacteria (Verhoeven 2002).
144
9.4.2
J. Filser, R. Prasse
Biotic Interactions
The diverse invertebrate fauna at Nizzana support a countless number of biotic interactions shaping the structure and functioning of the ecosystem. Here, we sketch only a few examples within the aboveground food web. In spring, grasshoppers, caterpillars and snails feed on annual plants, and pollinators exploit nectar and pollen. Later on, harvester ants, small mammals and birds shape the next plant generation by selectively feeding on seeds (Garb et al. 2000; Lortie et al. 2000; Wilby et al. 2001). The large abundance and diversity of predators (e.g. beetles, spiders, birds and lizards) signify a high predation rate – which may sometimes tremendously affect arthropod communities. Especially the white stork preys intensively on tenebrionid beetles during spring migration (personal observation at Nizzana). At another site, storks removed up to 95% of large-sized adults, thereby devastating regional tenebrionid populations (Ayal and Merkl 1994).
9.4.3
Ecosystem Engineering
Bioturbation, the burrowing and digging activity of animals, is of vital importance for the functioning of desert ecosystems – from water infiltration to element distribution, plant community composition and productivity (Boeken et al. 1998; Whitford and Kay 1999). Not only large animals, such as porcupines (Hystrix indica), but also invertebrates such as ants or snails can have a strong impact on these processes (Zaady et al. 1996, 2001; Wilby and Shachak 2000; Ayal et al. 2004). In a greenhouse experiment, Zhang (1996) found that burial of seeds and nutrient availability had a higher positive impact on plant performance than did water availability. Also tenebrionids dig actively, e.g. Pimelia arabica at Nizzana (Mahn 1994). There is perhaps less awareness that many other arthropods, e.g. bees, wasps and spiders, are also very active diggers. For example, four psammophilous spider species found at Nizzana excavate vertical tunnels (13–16 cm deep) by carrying silk-sand bundles (Henschel 1998). These bundles were certainly part of the variety of ‘Nescafé’ aggregate formations which we found. Likewise, some ant species from Israel, such as Messor semirufus, produce characteristic soil aggregates, similar to those in Fig. 9.6b (Amitai 1998). Based on our direct observations of structure formations (Fig. 9.6d, e), we conclude that aboveground herbivores and carnivores at Nizzana also significantly contribute to soil microstructure and small-scale heterogeneity. H. indica and M. arenarius, both herbivorous animals, had much more impact on plants by their burrowing and digging activity than by consumption (Wilby et al. 2001). On ant mounds and in porcupine diggings, plant density was up to 9 times, biomass up to 8 times and species richness up to 3 times higher than in similar areas of undisturbed soil. This study convincingly demonstrated the necessity of integrating ecosystem engineering for a fundamental understanding of ecosystem
9 A Glance on the Fauna of Nizzana
145
functioning in deserts. Still, the amount and quality of invertebrate-mediated soil modifications and their significance for ecosystem functioning in sandy deserts remains largely unexplored and strongly needs further research.
9.4.4
Spatial Heterogeneity
Spatial heterogeneity in deserts can be affected by abiotic conditions, in particular soil properties, and by biotic interactions. The burrowing activity of small mammals is a major source of patchiness in environmental conditions (Whitford and Kay 1999; Shenbrot et al. 2002). Our findings point to the high importance of invertebrates in structuring the soil at Nizzana, both horizontally and vertically. The gradient study by Mahn (1994) and our own data revealed the highest overall diversity on the dune north slope, followed by the south slope and playa. Individual numbers did not always reflect this pattern: small flies were most abundant on the south slope, and small ants in the playa. However, pitfall trap catches are biased towards activity and, thus, do not reflect actual individual numbers (Mommertz et al. 1996). In the Sayeret Shaked park, Messor arenarius foraged preferably in inter-shrub patches (Wilby and Shachak 2000). Our results do not support such a differentiation for any of the ant morphotaxa distinguished. In summary, our data demonstrate a large spatial variability for Nizzana, exemplified by pronounced differences between site N1 and N2 (Chap. 29, this volume), a high structural diversity within each site, and a vertical differentiation of microsites within a few centimetres.
9.5
Conclusions
We are very well aware that our glance on the fauna of Nizzana is by no means complete but the site is – as are many desert ecosystems (e.g. Ayal et al. 2004) – evidently very diverse in morphology, substrate use, behaviour, spatial distribution and influence on ecosystem functioning. Nizzana is an excellent site for studying any kind of ecological interactions in a largely undisturbed arid sand dune ecosystem. The data gathered by us hint at a particularly high faunal diversity, possibly with particular differentiation at higher trophic levels, which should attract the attention of ecologists interested in food web interactions and evolutionary processes. The remarkable variation in effects of invertebrates on structure and nutrient distribution of the soil is a true challenge for soil ecologists, particularly with respect to spatial variability. We hope that our studies will stimulate much future research at Nizzana within the AERC. Acknowledgements First of all, JF would like to thank Peter Felix-Henningsen who invited her to Nizzana and subsequently created and conducted the Minerva School with her. The Minerva Foundation is acknowledged for its generous support of this school. Our hearty thanks go to the
146
J. Filser, R. Prasse
Minerva Arid Ecosystems Research Centre of the Hebrew University of Jerusalem, in particular to Simon Berkowicz and Eyal Sachs for their invaluable help in all field and laboratory work, and for providing data and literature. Bodo Rummel contributed much expertise and aid, Yossi Steinberger supervised the nematode analyses, and Richard Verhoeven – who also provided some literature on the subject – those of the protozoans. Thanks go to Joh Henschel for supporting the literature search so quickly, and to Iris Burfeind who compiled the list of references. The students (seven from Israel and ten from Germany) had their share in the 2001 survey. RP would like to thank all staff and students of the E.S.E. Department of the Hebrew University who helped him and participated in enjoying the beauty of the region’s wildlife. Last but not least, we cordially thank all co-workers, especially Aaron Kaplan, and the students from Israel for their wonderful cooperation and hospitality.
References Alon A, Steinberger Y (1999) Response of the soil microbial biomass and nematode population to a wetting event in nitrogen-amended Negev desert plots. Biol Fertil Soils 30:147–152 Amitai P (1998) Handbook of insects of Israel and other arthropods. Keter, Jerusalem Ayal Y, Merkl O (1994) Spatial and temporal distribution of tenebrionid species (Coleoptera) in the Negev Highlands, Israel. J Arid Environ 27:347–361 Ayal Y, Polis GA, Lubin Y, Goldberg DE (2004) How can high animal biodiversity be supported in low-productivity deserts? The role of macro-detritivory and habitat physiognomy. In: Shachak M, Gosz JR, Pickett STA, Perevolotsky A (eds) Biodiversity in drylands: towards a unified framework and identification of research needs. Oxford University Press, pp 15–29 Barratt BIP, Derraik JGB, Rufaut CG, Goodman AJ, Dickinson KJM (2003) Morphospecies as a substitute for Coleoptera species identification, and the value of experience in improving accuracy. J R Soc N Z 33:583–590 Boeken B, Lipchin C, Guttermann Y, Rooyen Nv (1998) Annual plant community responses to density of small-scale soil disturbances in the Negev Desert of Israel. Oecologia 114:106–117 Cloudsley-Thompson JL (1996) Biotic interactions in arid lands (Adaptations of desert organisms). Springer, Berlin Heidelberg New York Dunger W, Fiedler HJ (1989) Methoden der Bodenbiologie. Gustav Fischer, Stuttgart Garb J, Kotler BP, Brown JS (2000) Foraging and community consequences of seed size for coexisting Negev Desert granivores. Oikos 88:291–300 Gruia M, Poliakov D, Broza M (1999) Collembola of Northern Israel, I. Israel J Zool 45:175–198 Henschel JR (1998) Dune spiders of the Negev Desert with notes on Cerbalus psammodes (Heteropodidae). Israel J Zool 44:243–251 Kelt DA, Brown JH, Heske EJ, Marquet PA, Morton SR, Reid JRW, Rogovin KA, Shenbrot G (1996) Community structure of desert small mammals: comparisons across four continents. Ecology 77:746–761 Lortie CJ, Ganey DT, Kotler BP (2000) The effects of gerbil foraging on the natural seedbank and consequences on the annual plant community. Oikos 90:399–407 Mahn M (1994) A short-term survey of darkling beetles (Col.: Tenebrionidae) in a sand dune system, Nizzana, Negev Desert. Institute of Ecology, Technical University of Berlin Mommertz S, Schauer C, Kösters N, Lang A, Filser J (1996) A comparison of D-Vac suction, fenced and unfenced pitfall trap sampling of epigeal arthropods in agroecosystems. Ann Zool Fennici 33:117–124 Shenbrot G, Krasnov B, Khokhlova I, Demidova T, Fielden L (2002) Habitat-dependent differences in architecture and microclimate of the burrows of Sundevall’s jird (Meriones crassus) (Rodentia: Gerbillinae) in the Negev Desert, Israel. J Arid Environ 51:265–279
9 A Glance on the Fauna of Nizzana
147
Verhoeven R (2001) Response of soil microfauna to organic fertilization in sandy virgin soils of coastal dunes. Biol Fertil Soils 34:390–396 Verhoeven R (2002) Ciliates in coastal dune soils of different stages of development. Eur J Soil Biol 38:187–191 Wallwork JA (1982) Desert soil fauna. Praeger, New York Whitford WG (1996) The importance of the biodiversity of soil biota in arid ecosystems. Biodiversity Conserv 5:185–195 Whitford WG, Kay FR (1999) Bioturbation by mammals in deserts: a review. J Arid Environ 41:203–230 Wilby A, Shachak M (2000) Harvester ant response to spatial and temporal heterogeneity in seed availability: pattern in the process of granivory. Oecologia 125:495–503 Wilby A, Shachak M, Boeken B (2001) Integration of ecosystem engineering and trophic effects of herbivores. Oikos 92:436–444 Wolf M, Shmida A (1995) Association of flower and pollinator activity in the Negev desert, Israel. Adv GeoEcol 28:173–192 Zaady E, Groffman P, Shachak M (1996) Release and consumption of snail feces in Negev Desert soils. Biol Fertil Soils 23:399–404 Zaady E, Offer ZY, Shachak M (2001) The content and contributions of deposited aeolian organic matter in a dry land ecosystem of the Negev Desert, Israel. Atmospheric Environ 35:769–776 Zhang J (1996) Interactive effects of soil nutrients, moisture and sand burial on the development, physiology, biomass and fitness of Cakile edentula. Ann Bot 78:591–598
Chapter 10
Biological Crusts B. Büdel and M. Veste
10.1
Introduction
Biological soil crusts (also called microbiotic, microphytic or cryptogamic crusts) are common cryptogamic communities in various arid and semi-arid regions of the world. Apparently only a few soil crusts worldwide are as well investigated as those of the Nizzana/Sinai-Negev sand fields (Galun and Garty 2001; also see Belnap et al. 2001). Concerning their species diversity, ecology and ecophysiology, several studies have been published and the current knowledge, including some new information, will be presented in the following review. Nevertheless, even though these crusts have been the subject of several studies, we are far from knowing their complete inventory of organisms, since almost all investigations dealing with species composition were restricted to oxygenic-photoautotrophic organisms. Fungi and bacteria, important components in soil crusts also in the north-western Negev Desert (unpublished data), have been widely neglected so far.
10.2 Structure of Biological Soil Crusts The crusts are composed of cyanobacteria, green algae, mosses, fungi as well as lichens. Fine material accumulates within the crust, contrasting with the sand underneath or mobile sand. In the crust in Nizzana, the silt and clay fractions with diameters less than 20 µm represent 23% of the grain-size population (Verrechia et al. 1995). Filaments of cyanobacteria which exude mucilaginous material stick together the sand grains of the topsoil, forming a soil crust with a thickness 1–17 mm (Fig. 10.1, Table 10.1).
10.3
Crust Types
During the last two decades, several approaches have been made to distinguish different crust types, which have been recognized by several means. Danin et al. (1989) distinguished three types of crusts in the Nizzana area, based on their S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
149
B. Büdel, M. Veste 8 100
chlorophyll a crust thickness
80
6
60
4
40 2 20 0
SFS
ID
NFS-M NFS SFS
Nizzana (site 1)
ID-L
NFS
crust thickness [mm]
chlorophyll a-content [mg m−2]
150
0
Haluza (site 3)
Fig. 10.1 The different types of soil crusts, according to their topology, chlorophyll a content and thickness at Nizzana (site 1) and Haluza (site 3). SFS South-facing slope, ID interdune, NFS northfacing slope, NFS-M north-facing slope with mosses, ID-L soil lichen crusts with Fulgensia fulgens in the interdune
Table 10.1 Crust types of the north-western Negev Desert Authors
Described crust types
Location
Danin et al. (1989)
Three main floristic types: (1) species-rich (nine spe- Nizzana, northwest cies) in the northern part; (2) Trichocoleus-Nostocdominated crust (seven species); (3) Trichocoleusdominated with additional Schizothrix, Phormidium and Chroococcidiopsis species
Kidron (1995), Karnieli et al. (1999)
Five different types: types 1–4 are of cyanobacteria Nizzana mainly with low moos cover; chlorophyll content 17–41 mg/m2; sugar content 4.7–14.3 g/m2; protein content 3.7–9.3 g/m2. Type 5 has a high moss cover; chlorophyll content 51 mg/m2; polysaccharide content 30.6 g/m2; protein content 28.4 g/m2
Veste et al. (2001a)
Two main types, differentiated according to topology, Nizzana, thickness, chlorophyll a content and colour: (1) light Haluza brown crusts with an average thickness of 1–3 mm in the interdune and south-facing slope, chlorophyll a content 17–22 mg/m2; (2) dark crusts on the northern slopes and the north-exposed dune base, thickness of 3–5 mm, chlorophyll a content 41–52 mg/m2
Veste et al. (2001a, b) Lichen soil crusts: Fulgensia fulgens, Squamarina Haluza sand fields cartilaginea, Squamarina lentigera, Diploschistes diacapsis, Collema spp., and other cyanobacterial lichens, thickness of 1–17 mm, chlorophyll a content 77–136 mg/m2
different floristic settings (Table 10.1). Kidron (1995), and later Karnieli et al. (1999), applied a combination of floristic, physiological and soil parameters to distinguish between five types of crusts, and were able to visualize and distinguish
10 Biological Crusts
151
all five by their spectral response (% reflectance; cf. also Karnieli 1997). Floristics did not play a direct role in the soil-crust classification system of Veste et al. (2001a); rather, the authors applied a combination of topology plus physiological and soil parameters (Table 10.1, Fig. 10.1). They distinguished mainly between (1) light brown crusts in the interdune and south-facing slopes and (2) darker crusts along north-facing slopes of the dunes. Mosses usually occur in the interdune area and especially at the dune base, whereas lichens may occur in the interdune area and on stable flat, north- and northwest-facing slopes (e.g. Haluza sand fields). According to the terminology introduced by Belnap (2001), all crusts found in the north-western Negev/Sinai region belong to the smooth and/or rugose crust types, typical for regions without soil freezing.
10.4
Species Diversity
A complete list of the photoautotrophic organisms found so far in the Nizzana crusts is given in Table 10.2. This species inventory is, however, still rather incomplete because (1) many of the organisms occurring in such crusts can be identified only by using pure cultures and (2) the taxonomy of the highly specialized soilcrust organisms is a difficult field, as several systematic groups are insufficiently known for correct identification on the species level (cf. preface in Belnap and Lange 2001). Cyanobacteria are clearly the dominating organisms of the crusts in the northwestern Negev/Sinai region. Among these, the filamentous genera Microcoleus and Trichocoleus seem to be the most important in all types of crusts (Danin et al. 1989; Karnieli et al. 1999), and also play a key role as pioneering organisms in disturbed areas (Kidron 1995). Trichocoleus sociatus (Fig. 10.2A, B) was found to be the major species represented in the crusts (Lange et al. 1992). This species contributes much, if not most, to the formation of the biological crust by adhering firmly to sand grains with its extracellular sheath, thereby binding the grains together
Table 10.2 Species diversity Authors
Species Cyanobacteria Trichocoleus sociatus (W. et G.S. West) Anagnostidis M. vaginatus (Vauch.) Gom.
Danin et al. (1989), Danin (1991)
+
Lange et al. (1992)
Kidron (1995), Karnieli et al. (1999)
Veste et al. (2001a, b), Galun and Garty (2001)
+
+
+
+ (continued)
152
B. Büdel, M. Veste
Table 10.2 (continued) Authors
Species Microcoleus sp. Nostoc microscopicum (Carm.) Harv. et Hock. Nostoc spp. Calothrix parietina Thuret Calothrix spp. Oscillatoria spp. Phormidium spp. Schizothrix friesii (Ag.) Gom. Schizothrix sp. Scytonema spp. Chroococcidiopsis spp. Gloeocapsa sp.
Danin et al. (1989), Danin (1991)
Lange et al. (1992)
Kidron (1995), Karnieli et al. (1999)
Veste et al. (2001a, b), Galun and Garty (2001)
+ + +
+ +
+ +
+ + + +
+
+ + +
+
+ +
+ + +
+
+
+
+
+
+
+
Green algae Macrochloris multinucleata (Reisigl) Ettl et Gärtner a Stichococcus sp.
Mosses Brachymenium exile (Doz. et Molk) Bosch et Lac. Bryum bicolor aggr. B. dunnense Tortula brevissima T. muralis Hedw. Pterygoneurum sp. Aloinia sp.
+ + + + + + +
Lichens Collema tenax var. vulgare (Schaer.) Degel. Diploschistes diacapsis (Ach.) Lumbsch Fulgensia fulgens (Sw.) Elenkin Squamarina cartilaginea (With.) P. James S. lentigera (Weber) Poelt a
As Chlorococcum sp. in Lange et al. (1992)
+ + + + +
10 Biological Crusts
153
Fig. 10.2 Light microscope images of cyanobacteria and algae of the Nizzana soil-crust system. A Image of a wet soil crust from Nizzana. Note the filaments of Trichocoleus sociatus and the dark, globose colonies of Nostoc sp. B Filament of Trichocoleus sociatus with numerous trichomes unified within one common sheath. C Scytonema sp., filament with sheath and trichome extruded from the sheath (bottom left). D Nostoc sp., young colony within sheath. E Macrochloris multinucleata, a few minutes after wetting of the natural crust. F M. multinucleata, pure culture from the Nizzana crust, 3 months after transfer to fresh medium; at the onset of N starvation, the algal cell turns red (not shown)
(Belnap 2001). The taxonomy of the genus is still partly uncertain (Yeager et al. 2004). It seems that the genus Microcoleus might be polyphylous in nature, and it may be subdivided into several genera in the future. A number of species formerly attributed to the genus Microcoleus have been transferred to the genus Trichocoleus recently (for a more detailed discussion, see Komárek and Anagnostidis 2005).
154
B. Büdel, M. Veste
Among the more frequent cyanobacteria in the Nizzana crusts are Scytonema sp. (Fig. 10.2C), Nostoc microscopicum and Nostoc sp. (Fig. 10.2A, D), Phormidium sp. and Oscillatoria sp. The green alga Macrochloris multinucleata (Fig. 10.2E, F) appeared to be also very common in the Nizzana crusts (Lange et al. 1992). The M. multinucleata populations were quite large in individual numbers and were situated in the upper parts of the crust. They could be easily recognized in the crust samples by their typical orange colour in the dry state, turning green a few hours after wetting. The orange-coloured masking of the M. multinucleata cells is caused by a haematochrome present in almost all dry cells. However, this change from green to an orange-red colour was also observed in older pure cultures of M. multinucleata, when nutrient shortage occurred in the culture tubes 3 to 4 months after transfer to fresh medium (cf. Fig. 10.2F). Mosses are common on north-facing slopes and along the dune base (Karnieli et al. 1999). Individual moss plants (Tortula spp., Bryum spp.) tend to occur aggregated and, thus, probably play an important role in the stabilization of the soil surface, through the presence of rhizoids and protonemata (Lange et al. 1992). Lichens reach a high cover especially in interdune areas and on flat north- and northwest-facing slopes in the northern dune fields of the Haluza sand fields. Here, cyanobacterial lichens dominate the soil crusts. The green-algae lichens Fulgensia fulgens, Squamarina cartilaginea, S. lentigera and Diploschistes diacapsis have been observed locally on small soil mounds (Veste et al. 2001a).
10.5
Conclusions
The species inventory of crusts of the north-western Negev/Sinai region has its closest relations to Africa and Asia, when concentrating on lichens and algae/ cyanobacteria. Taking the moss flora into account, the closest relations exist to African crust communities (cf. also Büdel 2001). The activity of the microbiotic crusts and their role in various ecosystem processes is dealt with in Chapters 21–23 of this volume.
References Belnap J (2001) Comparative structure of physical and biological soil crusts. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York, pp 177–191 Belnap J, Lange OL (2001) Biological soil crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York Belnap J, Büdel B, Lange OL (2001) Biological soil crusts: characteristics and distribution. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York, pp 3–30
10 Biological Crusts
155
Büdel B (2001) Synopsis: comparative biogeography of soil crust biota. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York, pp 141–152 Danin A (1991) Plant adaptation in desert dunes. J Arid Environ 21:193–212 Danin A, Bar-Or Y, Dor I, Yisraeli T (1989) The role of cyanobacteria in stabilization of sand dunes in Southern Israel. Ecol Mediterr 15:55–64 Galun M, Garty J (2001) Biological soil crusts of the Middle East. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York, pp 95–106 Karnieli A (1997) Development and implication pf spectral crust index over dune sands. Int J Remote Sensing 18:1207–1220 Karnieli A, Kidron GJ, Glaesser C, Ben-Dor E (1999) Spectral characteristics of cyanobacteria soil crust in semiarid environment. Remote Sensing Environ 69:67–75 Kidron GJ (1995) The impact of microbial crust upon rainfall-runoff-sediment yield relationships on longitudinal dune slopes, Nizzana, Western Negev Desert, Israel. PhD Thesis, Hebrew University, Jerusalem Komárek J, Anagnostidis K (2005) Cyanoprokaryota. 2. Teil: Oscillatoriales. In: Büdel B, Gärtner G, Krienitz L, Schagerl M (eds) Süßwasserflora von Mitteleuropa vol 19/2. Elsevier, München Lange OL, Kidron GJ, Büdel B, Meyer A, Kilian E, Abeliovich A (1992) Taxonomic composition and photosynthetic characteristics of the “biological soil crusts” covering sand dunes in the western Negev Desert. Funct Ecol 6:519–527 Verrechia E, Yair A, Kidron G, Verrechia K (1995) Physical properties of the psammophile cryptogamic crust and their consequences to the water regime of sandy soils, north-western Negev, Israel. J Arid Environ 29:427–437 Veste M, Littmann T, Breckle S-W, Yair A (2001a) The role of biological soil crusts on desert sand dunes in the Northwestern Negev, Israel. In: Breckle S-W, Veste M, Wucherer W (eds) Sustainable land-use in deserts. Springer, Berlin Heidelberg New York, pp 357–367 Veste M, Littmann T, Friedrich H, Breckle S-W (2001b) Microclimatic boundary conditions for activity of soil lichen crusts in sand dunes of the north-western Negev Desert, Israel. Flora 6:465–474 Yeager CM, Kornosky JL, Housman DC, Grote EE, Belnap J, Kuske CR (2004) Diazotrophic community structure and function in two successional stages of biological soil crusts from the Colorado Plateau and Chihuahuan Desert. Appl Environ Microbiol 70:973–983
Chapter 11
Land Cover in the Nizzana Sandy Arid Ecosystem. Mapping Surface Properties with Multi-Spectral Remote Sensing Data J. Hill, T. Udelhoven, T. Jarmer, and A. Yair
11.1
Introduction
From the time of early Landsat satellite observations during the 1970s, the nature of the sharp contrast along the Israeli-Egypt border in the northern Negev (Fig. 11.1) has been a matter of scientific interest and debate. This spatial feature also cuts through the Nizzana Experimental Station of the Arid Ecosystems Research Centre (Berkowicz et al. 1995), which is located in the north-west part of the Negev Desert at the border between Israel and Egypt, i.e. in the eastern extension of the Sinai sand fields (see Chaps. 1, 2 and 6, this volume). Although already visible on early space photographs taken from the manned Gemini spacecrafts during the 1960s, studies on the phenomenon started only when first Landsat-MSS images over the border region had been collected in 1972. Based on field evidence, Otterman et al. (1974) were the first to explain this contrast by the presence of dark desert plants (both photosynthetically active and senescent/ dormant) on the Israeli side, while the Egyptian area is almost devoid of vegetation. More recently, Karnieli and Tsoar (1995) re-opened the discussion and rejected the predominant role of plants in producing such spectral contrast. Following findings of Danin (1996), they concluded “that the well-known contrast between Sinai and the Negev, that has drawn the attention of many scientists, is not a direct result of vegetation cover but is caused by an almost complete cover of biogenic crust”. In his reply to this paper, Otterman (1996) insists that the conclusion of Karnieli and Tsoar (1995), namely “that higher vegetation cover is less than 30% and therefore the soil itself is believed to be the principal contributor to the overall spectral response”, is difficult to accept because, even with sparse plant cover, the interception of solar irradiance by desert plants as well as the rays reflected from the soil (i.e. absorption and shadowing) must be accounted for. However, this debate is not only of purely scientific interest with regard to light interaction processes on natural surfaces, but has gained specific importance due to the role microbiotic crusts have been attributed within the ecological functioning of deserts.
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
157
158
J. Hill et al.
Fig. 11.1 The Nizzana Research Site and the Israel-Egypt border, seen from NASA’s Space Shuttle on 10 June 1991
11.2
Study Objectives
Approximately more than 20% of the world’s arid zones are covered by aeolian (windborne) sand (Table 1, p. 2) (e.g. Pye and Tsoar 1990). While Yair (1994) has suggested that sandy arid ecosystems provide superior conditions for plant growth owing to an optimum water availability, which is common knowledge since decades (Walter 1960), Danin (1996) insists that mobile sands constitute a rather poor habitat because they contain only small quantities of fine-grained particles, organic matter and nutrients. In these environments, the ecological importance of microbiotic soil crusts (such as those occurring in the Nizzana sand field of the Western Negev, Israel) is related mainly to their resistance against erosion caused by wind or surface runoff. They develop on the soil surface by the growth of various poikilohydric organisms (see Chap. 10, this volume). When dry, their thalli are dormant but in a latent reversible state; when wetted, they rapidly become physiologically active (Danin 1996). Unlike vascular plants, their cover is not reduced in drought years, they do not dissolve when wet and, thus, contribute substantially to the protection of unconsolidated and highly erodible substrates in drylands, which are frequently prone to degradation and desertification processes (e.g. Belnap 1995;
11 Land Cover in the Nizzana Sandy Arid Ecosystem
159
Belnap and Gillette 1998). The crusts are semi-permeable and, depending on the parent substrate, tend to concentrate runoff and quite efficiently protect infiltrated water against excessive evaporation (Verrecchia et al. 1995). Garcia-Pichel and Belnap (1996) showed that biomass and activity in biological soil crusts are usually concentrated within 3 mm of the soil surface. This suggests that, if it is possible to identify relations between optically active key variables and crust properties, then optical remote sensing systems can be used not only to assess the presence of biological soils crusts but also to differentiate between various crust types, thereby providing important information on areas of stability and disturbance in arid ecosystems. Mapping the spatial extension and diversity of these areas is also highly relevant for sustainable land management and development projects in drylands. The key issues followed in this study are: ● ● ●
●
mapping bio- and geophysical surface properties, determining the ratio of photosynthetically active and senescent plants, assessing to which extent crusting is due to ‘biotic’ components (vs. mineral crusts) and investigating the scale dependency of surface remote sensing.
11.3
Field and Remote Sensing Data
A major problem in the description of microbiotic crusts and their properties in the Nizzana Dune Field results from the fact that most crust data published so far (e.g. Karnieli and Tsoar 1995; Danin 1996; Karnieli 1997) have not been presented in relation to the explicit spatial position from where the corresponding samples have been collected. Therefore, it is difficult to understand whether this information is only indicative for generic characteristics of microbiotic crusts, or whether it can be considered representative for soil crusts in the area. Since it can be easily perceived in the field that the microbiotic crusts exhibit a high degree of spatial variability (e.g. initial forms close to the active dune crest, well-established, robust and sometimes even abundant crusts on northerly exposed dune plinth and slopes, more fragile crusts on older dune surfaces in the interdune corridors, and others), it was felt that a more differentiated assessment of the spatial diversity of soil crusts was needed. The data collected for this purpose comprise spectral field and laboratory measurements as well as the analyses of soil and crust samples taken during several field campaigns (April 1997, March 1998, October 1998, March 1999). To understand the relation between catenary position, crust composition and optical properties to be linked with remote sensing observations, about 60 crust and non-crust samples were collected during wet (March 1998) and dry periods (October 1998). Positioned by means of GPS measurements, they were taken along three selected transects at the Nizzana Research Site which extended over several dune–interdune sequences (Fig. 11.2). Soil crust samples from transect 1 and 2 were collected some days after
160
J. Hill et al.
Fig. 11.2 Geo-referenced land cover map, derived from multi-spectral classification of a digitised aerial photograph from 1992; the sampling transects for analysing the variability of surface crusts are superimposed
a rainfall event in March 1998, the samples from the third transect in October 1998 after the long summer drought. Laboratory analyses conducted at the University of Trier concentrated on the contents of organic and inorganic carbon, carbohydrates, proteins and chlorophyll; additionally, pH and electronic conductivity were measured. For a subset of 21 crust samples from transect 2, grain-size analysis was carried out at the Hebrew University of Jerusalem. Surface reflectance spectra were measured using an ASD FieldSpec-II spectroradiometer, which covers the wavelength region between 350 and 2,500 nm with a
11 Land Cover in the Nizzana Sandy Arid Ecosystem
161
spectral sampling interval of one nanometre. All measurements were fully calibrated with reference to a Spectralon standard with certified reflectance properties, archived and documented in specific readers (e.g. Hill et al. 1997). The remote sensing data used so far include true-colour aerial photographs from the years 1992, 1996 and 1997, Landsat Thematic Mapper data from spring and late summer 1992, and DAIS 7915 hyperspectral imagery from a flight campaign carried out in August 1997. The DAIS-7915 instrument has 79 bands in the visible, near-infrared, mid-infrared, and thermal-infrared wavelength domains. For this work, only the reflective portion of the wavelength spectrum, i.e. DAIS bands 1–72, was considered. For the atmospheric correction of the DAIS images, we have used adjusted in-flight calibration coefficients based on specifically designed reference measurements collected at Nizzana and an additional study site in Spain (Hill et al. 1999). Standard pre-processing of all remote sensing data involved rigorous atmospheric corrections (Hill and Mehl 2003), such that all datasets could be analysed on the level of bi-directional reflectance. Only the digitised true-colour aerial photographs (1992) were calibrated to reflectance with an empirical line approach (Hill et al. 2000).
11.4
Spatial Variability of Crust Properties Within the Sand Dune Ecosystem
It is common knowledge that periods of photosynthetic activity of microphytes are always restricted to short periods of sufficient hydration (e.g. Lange et al. 1992; Danin 1996). Depending on the physiological status (age, nutrient supply, illumination, etc.), the microbiotic crusts can then develop a photosynthetically active layer with up to 10% of the chlorophyll concentration of assimilatory organs of typical arido-active phanerogamous plants of the Negev Desert (Lange et al. 1992). Since the chlorophyll-a content is sensitive to photochemical destruction, its signal will largely disappear during dry and hot periods. However, the disappearance of the signal could also indicate the movement of the poikilohydrous Microcoleus threads to lower, more protected soil levels. Other studies have emphasised that microbiotic crusts tend to trap fine particles and, based on samples from dunes at various stages of higher plant succession, Danin (1991) claims that the proportion of fines (silt and clay) and the amount of organic matter in the crust are in general highly correlated (r > 0.85). He suggests that positive feedback processes start as soon as a certain threshold value of fine particles in the soil has been achieved; then, the cyanobacteria begin to develop and fix airborne sand and fines by gluing their filaments with polysaccharide mucilage to the surrounding grains. The high moisture-holding capacity of the polysaccharide sheaths of the cyanobacteria and the increasing proportions of fine-grained material produce better moisture conditions and enable better growth conditions (Danin 1996). However, since microbiotic crusts (quite similar to mineral crusts) also increase surface runoff (e.g. Yair 1990), better conditions for plant growth tend
162
J. Hill et al.
to occur only in morphological sinks where water is concentrating; we find several indications that the relationships between crust constituents might indeed be more complicated (Fig. 11.3). Although similar trends could also be identified across a wider range of catenary positions along transect 2, the correlation between fine particles (silt and clay) and organic matter is much weaker. The so-called playa sample1 indicates that the relationship between fines and organic matter is disturbed as soon as other process determinants are involved. One can find clear evidence that sediment transport by surface runoff is substantially involved in establishing soil crusts in other catenary positions (primarily the interdune depressions), such that the amount of fines in surface crusts is quite variable without microbiotic activities necessarily being considered the primary driving force. Moreover, it is seen that other components, such as the proportion of inorganic carbon, may correlate with the proportion of fine particles in the crust substrate, thereby providing further indications that it is difficult to use such relationships as an indicator for crust-specific positive feedback mechanisms. To understand whether the overall differences between crust samples from specific catenary positions (i.e. north- and south-oriented slopes, old dunes in the interdune corridor) were indicative for specific ecosystem compartments, the data from transect 1 and 2 underwent a thorough statistical analysis (univariate analysis of variance, cluster and discriminant analyses). The crust parameters used were carbohydrates, proteins and chlorophyll as well as organic and inorganic carbon
Fig. 11.3 Relationships between the proportion of fine particles (silt and clay) and organic (left) and inorganic carbon (right), based on 21 samples from transect 2
1 Here, the proportion of fine particles is due predominantly to fluvial deposition, rather than aeolian transport mechanisms (e.g. Rendell et al. 1993; Pfisterer et al. 1996).
11 Land Cover in the Nizzana Sandy Arid Ecosystem
163
(Hill et al. 2000). The conclusions drawn from this analysis support the hypothesis that source–sink relations between topographic compartments appear more important than microclimatic conditions expressed in terms of specific topographic variables such as slope and aspect. Furthermore, discriminant analysis unveils the differentiating characteristics of the chemical properties2 in relation to four identified surface types (derived from cluster analysis; Fig. 11.4). ●
●
The first group (cluster 1) encompasses all samples which were taken from sandy substrates. Here included are active dunes, areas with initial crust formation, and dune slopes with a mosaic of crusted and non-crusted patches (including signs of crust destruction by wind erosion). These samples have very low chlorophyll, protein and carbohydrate concentrations; some samples from the active dune crest (crossed boxes in Fig. 11.4) are even completely free of chlorophyll (Fig. 11.4). Clusters 2 and 3 exhibit progressively higher contents of carbohydrates, proteins and chlorophyll. The corresponding samples come from various catenary positions within the interdune area, where microbiotic and mineral crusts alternate (Fig. 11.4).
0 Discr. Function
1 Cinorg 0.781 −0.119 Carbohydrates −0.279 Chlorophyll −0.224 Protein Corg −0.063
0.1
0.2 km
2 3 0.626 0.003 0.929 −0.358 0.733 0.620 0.614 0.027 n.u. 0.570 0.092 n.u.
Canonical Discriminant Function 6 4
Function 2
2
2
3
4 Cluster Centers
0 −2
Cluster 4 Cluster 3
1
Cluster 2
−4 −10 0 Function 1
Cluster 1
10
20
Fig. 11.4 Results of a discriminant analysis applied to four major surface types (clusters, here mapped to the corresponding sampling positions), which can be differentiated in terms of corresponding concentrations of Corg, Cinorg, proteins, carbohydrates and chlorophyll (from Hill et al. 2000); red squares with cross indicate positions in active dune areas
2 Independently from the number of classes, it was found that the ‘playa’ samples are always clearly separated from crust- and sand-dominated samples, primarily due to their much higher carbonate content. Carbohydrates and, to a lesser extent, the protein and chlorophyll concentrations are primary variables to structure the remaining samples.
164 ●
J. Hill et al.
Cluster 4 includes only three samples which come from the interdune ‘playa’; these represent typical examples of fine-grained (fluvial) deposits with exceptionally high carbonate concentrations. The surface at these locations is covered by a typical mineral crust.
Since chlorophyll (highly correlated with carbohydrates and protein during and shortly after wet periods), as a major differentiating component of surface crusts, is an optically active substance, it is expected that optical imagery can be used to map this type of spatial diversity. A secondary feature of crusted surfaces which may also trigger specific variations of spectral surface reflectance is associated with the grain-size composition of the mineral matrix (e.g. Elachi 1987). Mineral crusts contain elevated proportions of fines and, since cyanobacteria not only tend to agglomerate sand grains from the parent material but also trap aeolian dust particles (Verecchia et al. 1995; Danin 1996) as well as silt and clay transported by surface runoff, it is suggested that the proportion of fines in microbiotic crusts is increasing with time, too. Grain-size analysis, applied to a subset of crust samples from transect 2, supports these findings; the proportion of fines appears related to specific catenary positions. It is thus concluded that the crusted surfaces within the dune–interdune sequences at Nizzana must be differentiated according to both microbiotic and mineral components. Since chlorophyll concentration and grain-size composition influence the reflectance properties of surfaces, they might be used as primary optical tracers to identify the type and abundance of microbiotic soil crusts by means of remote sensing approaches.
11.5
Mapping the Spatial Diversity of Surface Properties with High Spatial-Resolution Aerial Photographs
When one considers the use of spectroradiometers and, potentially, imaging systems for monitoring the regeneration of microbiotic crusts, one needs to carefully review the boundary conditions under which chlorophyll might be used as a primary indicator. Cyanobacteria have only one form of chlorophyll (i.e. chlorophylla). Additionally, they have characteristic biliprotein pigments, e.g. phycobilins, which function as accessory pigments in photosynthesis. While there is no doubt that chlorophyll-a is an optically active substance with strong indicator function, the chlorophyll content of microbiotic crusts is variable over time. During relatively wet periods (i.e. winter season with occasional rain events, frequent dew fall and, at the same time, reduced solar irradiance), the crust tends to exhibit elevated chlorophyll contents. During dry and hot periods, however, the micro-organisms will be less active. The chlorophyll-a content is dependant on the physiological status of the crust (age, nutrient supply, illumination, etc.), possible photochemical destruction and/or the varying depth of Microcoleus threads during dry periods.
11 Land Cover in the Nizzana Sandy Arid Ecosystem
165
However, degradation products such as phaeophytin, chlorophyllite, phaeophorbite and other pigments also interfere with spectrophotometric and fluorometric measurements of chlorophyll, and they may in fact cause a two- to eightfold overestimation of chlorophyll under laboratory conditions (e.g. Nusch 1980). Hot and dry periods (i.e. summer) are therefore less suited for mapping the types and abundance of microbiotic crusts3 with optical imaging systems, although the accumulation of optically active degradation products could maintain the microbiotic crusts detectable at least for some time during inactive periods. One of the most promising approaches for estimating the proportional abundance of material components which occur within a specific surface area (i.e. pixel) is based on computationally decomposing multi-spectral measurements with reference to a finite number of pure spectral components, so-called endmembers. The method has become known as Spectral Mixture Analysis (e.g. Adams et al. 1989), and it assumes that most of the spectral variation in multi-spectral images is due to mixtures of a limited number of surface materials, and that these mixtures can be approximated as resulting from additive (linear) spectral mixing processes (i.e. where each photon contacts only one type of surface material). The computation of proportional abundances is solved with a system of linear equations, where a unique solution is possible as long as the number of spectral endmembers corresponds to the number of spectral bands. If the problem is underdetermined, i.e. the number of unknown material abundances exceeds the number of available spectral bands by a value of one, then a solution can still be obtained by assuming that the set of endmembers is exhaustive (i.e. the sum of the computed endmember proportions is equal to 1). The method has been widely used for mapping vegetation abundance under variable background conditions (Smith et al. 1990; Ray and Murray 1996). However, based on earlier studies with Landsat-TM (Hill et al. 1995), it was also considered a promising approach for mapping surface properties from true-colour aerial photographs.
11.5.1
Domain-Specific Unmixing
Since the theoretical maximum of endmembers to be used in spectral unmixing is limited to the number of bands plus one, it would not be possible to use more than four endmembers at a time for analysing the three-channel digital air photographs. This problem could be solved by exploiting the high spatial detail of the aerial imagery: based on a simplified foreground-background mixing model (involving
3 This bears an important implication with regard to the dispute about the nature of the sharp reflectance contrast along the Israeli-Egyptian border, since this feature can be observed from space during the whole year (e.g. Otterman et al. 1974; Karnieli and Tsoar 1995; Otterman 1996).
166
J. Hill et al. Nizzana Dune Field: Compenent Spectra 250.00
Dune Sand (98% Sand, 2% Fins) Silt/Clay (>80% Fines)
300.00
200.00
Photosynthetic Vegetation
Scaled Reflectance
Component Spectra, resampled to the spectra bandpass of the true-colour aerial photographs
Biotic Crust (Concentrated Algae, Cyanobacteria etc.)
Woody Vegetation
200.00
150.00
100.00 100.00 50.00
0.00
0.00 0.50
1.00 1.50 Wavelength (mm)
2.00
2.50
0.40
0.45
0.55
0.50 0.60 Wavelength (mm)
0.65
0.70
Fig. 11.5 Spectral endmember spectra for domain-specific unmixing in the spectral resolution of the DAIS-7915 and the corresponding digital aerial photographs
only a bright substrate and a dark vegetation spectrum), a mask was produced which differentiates between vegetated (both photosynthetically active and woody components) and non-vegetated image areas. Then, specifically adapted endmember combinations (Fig. 11.5) could be used for a dedicated analysis of either plant- or substrate-dominated portions of the image, an approach which was termed ‘domainspecific spectral unmixing’. For the vegetated domain, we used two vegetation spectra (photosynthetic and woody vegetation) and one substrate component (a 50:50 additive mixture of sand and silt/clay) to represent the background signal; for the substrate domain, three spectral components (sand, silt/clay, biotic crust) were chosen to model the composition of the soil surfaces characterised by changing proportions of sand, fines and biological crust (Fig. 11.5).
11.5.2
The Vegetated Domain
Spectral unmixing of the vegetated domain is able to unveil the spatial distribution of the perennial vegetation in the Nizzana Dune Field (Fig. 11.6, left). The most important aspect of the methodological approach is that it provides estimates for both photosynthetically active and senescent (i.e. woody) vegetation components, against a substrate background (here: sand). This is a clear advantage over conventional vegetation indices, which are sensitive only to green vegetation, and which have frequently been criticised for the bias introduced by the reflectance properties of the plant background (Price 1993). Accordingly, Fig. 11.6 demonstrates how different plant components are discriminated in the domain-specific unmixing approach, thereby providing separate estimates of live, dead and total plant cover (Figs. 11.6 and 11.7).
11 Land Cover in the Nizzana Sandy Arid Ecosystem
167
Fig. 11.6 Material abundance images for the vegetated (left) and substrate domain (right), derived from domain-specific spectral unmixing of reflectance-calibrated digital aerial photographs. Pure colours in the vegetation RGB colour composite (left) would correspond to 100% of sand/silt (red), photosynthetic vegetation (green) and woody vegetation components (blue), in the substrate RGB colour composite (right) to 100% of sand (red), biological crust material (green) and silt/clay (blue). Colour mixtures represent corresponding proportions of pure material components
Sand Photosythetic Vegetation Wood
Proportional Abundance (%)
100.00
Proportional Abundance
80.00 60.00 Vegetation Abundance Estimate Cross Section
40.00
Total Vegetative Cover Photosynthetic Vegetation
20.00
Woody Vegetation Components Sand Background Proportion
0.00 0.00
2.00
4.00
6.00
8.00 10.00
Relative Pixel Position (0.5 m Size)
Proportional Abundance (%)
100.00 80.00 60.00 Vegetation Abundance Estimate Cross Section
40.00
Total Vegetative Cover Photosynthetic Vegetation
20.00
Woody Vegetation Components Sand Background Proportion
0.00 0.00
4.00
8.00
12.00
Relative Pixel Position (0.5 m Size)
Fig. 11.7 Cross section of two shrubs on the dune crest and the corresponding estimates of dead, live and total vegetation cover, derived from the spectral mixture analysis of digital aerial photographs. Pure colours in the RGB colour composites (left) would correspond to 100% of sand (red), photosynthetic vegetation (green) and woody vegetation components (blue); colour mixtures represent corresponding proportions of pure material components
168
J. Hill et al.
Considering the spatial distribution of perennial plants, it becomes obvious that shrubs with substantial amounts of photosynthetically active leaves tend to cluster along the dune base where the infiltrated rainfall from elevated parts of the dune can be tapped by the plants. In the interdune depressions, smaller shrubs prevail which, as emphasised by the bluish colour tones, are dominated by dry woody components (Figs. 11.6 and 11.7).
11.5.3
The Substrate Domain
The unmixing results prove that it is not only possible to differentiate between active and dormant perennial plants in the vegetation domain, but that one can also derive spatially explicit maps of varying surface composition in the substrate domain (Fig. 11.6). In addition to the active dune (red, due to the dominance of the sand fraction) and ‘playa’ surfaces (blue, dominance of fines), one can distinguish a full range of tonal differences, where the predominance of greenish tones indicates areas with strongly developed biological crusts; cyan and yellow represent areas where the biological crust incorporates more sand and more fines respectively. Disturbed areas, caused either by car access and trampling (here due to the installation and maintenance of measurement equipment) or by wind erosion in the old dune complex of the interdune area, become visible through their stronger reddish colouring (Fig. 11.6). Although it is not yet possible to interpret these fractions in a fully quantitative way (as non-linear mixing effects are not accounted for, which is a basic requirement for dealing with complex mixtures), we are already able to identify and map crust types which are indicative of specific ecosystem characteristics and process domains (Fig. 11.2). It is also difficult to validate these results in a fully quantitative way. One possibility for controlling at least qualitatively the image-retrieved crust composition is to compare the proportional abundance for the three endmembers (sand, cyanobacteria, silt/clay) with the cluster groupings of the crust samples taken along the transects. For transect 1, it is found that the retrieved abundance estimates correspond quite well to the identified spatial pattern of surface types (Figs. 11.6 and 11.8). On the active dune sands, for example, the retrieved abundance estimates are dominated by the sand fraction (red colour of the pie charts) while, downwards to the dune base, we find increasing proportions of the cyanobacteria and silt/clay endmembers. This corresponds to the catenary sequences we identified in the field (see above). Accordingly, it is seen that increasing amounts of fines are detected in the interdune depression. While most of these sampling positions are correctly characterised by associated proportions of the cyanobacteria endmember, the samples from the so-called interdune playa are exclusively interpreted as a mixture between fines (silt/clay) and minor contributions of sandy components, a result which is fully supported by the grain-size analyses.
11 Land Cover in the Nizzana Sandy Arid Ecosystem
169
Fig. 11.8 Comparison between the cluster results for sampling transect 1 and the corresponding abundance estimates for sand (red), silt/clay (blue) and microbiotic crust material (green) for transect 1 at Nizzana, derived from the spectral mixture analysis of the substrate domain (aerial photographs, May 1992); red squares with cross indicate positions in active dune areas with complete absence of chlorophyll
Additionally, the results from the grain-size analyses (available only for transect 2) were compared with the abundance estimates retrieved by spectrally unmixing the aerial photographs. In order to do so, the abundance estimates were normalised in terms of the proportion of the cyanobacteria endmember. This implies that the ‘cyanobacteria’ fraction (Fcya) is removed from the abundance estimate by re-scaling the remaining fractions with the normalisation factor
(
f = 1 1 − Fcya
)
such that the retained proportional estimates (fractions) sum again to unity, i.e. ( n − 1)
∑ F ⋅ f =1 i =1
i
For example, a pixel having 0.33 sand, 0.33 cyanobacteria and 0.33 fines would convert to 0.5 sand and 0.5 fines (Adams et al. 1989). We have thus eliminated the signal due to the influence of the microphytes, thereby retaining only that part of the signal which relates to the mineral crust components, i.e. sand and fines (silt and clay). The comparison between the laboratory measurements and the aerial photography-based estimates of the proportion of fines along transect 2 reveals quite a remarkable correspondence (Fig. 11.9). With a coefficient of determination of
170
J. Hill et al.
Fig. 11.9 Relationship between the laboratory-derived percentage of fines (silt/clay) in the samples from transect 2, and the corresponding proportions retrieved by spectrally unmixing the reflectance-calibrated aerial photograph from May 1992
0.768, the image-based estimate tends to slightly overestimate the true proportion of silt and clay but can reproduce the existing grain-size variability along the sampling transect.
11.6
Conclusions
In combination with a considerable amount of geo-referenced sampling positions, the results from a carefully conducted spectral analysis of reflectance-calibrated true-colour aerial photographs clearly prove that a unique and, therefore, to some extent representative type of microbiotic crust does not exist in the Nizzana region. However, it could be demonstrated that various specific expressions of microbiotic crusts exist which are highly dependent on their catenary position within the dune– interdune sequences. The results from this ‘domain-specific unmixing’ approach, as it was termed here, exhibit the high spatial variability and diversity of crust types in this sandy ecosystem. The analysis of aerial photographs has also revealed that the dark desert plants can easily be detected against the bright background signal (of diverse substrates), and that the microbiotic crust is frequently, in large parts of the area, replaced by much brighter surface types. Additionally, the analysis of band residuals which result from hyperspectral imagery (Hill et al. 1998) clearly shows that the reflectance contribution of the dominantly woody desert plants can not be neglected. These findings resolve some ambiguities about understanding the nature of the sharp contrast along the Israeli-Egyptian border observed from space, which was
11 Land Cover in the Nizzana Sandy Arid Ecosystem
171
first noted and described by Otterman et al. (1974), and more recently put under debate by Karnieli and Tsoar (1995). The results presented here, as well as the analysis of additional remote sensing datasets, clearly prove that the spectral imprint of the microbiotic and mineral crusts alone is not sufficient to explain the well-known contrast along the Israeli-Egypt border, but needs the spectral contribution of woody desert plants, a perspective which quite recently was again underlined by Otterman (1996). Based on the understanding of the spatial-spectral features in Nizzana, it could also be shown that information on important surface (i.e. substrate) properties can be derived also from operational satellite data. This is an important contribution to the recent debate about the global monitoring of desert fringes. With regard to the regeneration of destroyed soil crusts, it can even be expected that the mixed signal which is produced by both chlorophyll-a and its degradation products might enable us to spectrally monitor the formation of microbiotic crusts (see Chap. 20, this volume) over time, as well as human impacts (see Chap. 6, this volume) on such crusts over time. The results reported above may be applicable to other areas where a sharp contrast has been observed along political borders. Wellknown examples are the Mexican-US border and the Uzbekistan-Afghanistan border.
References Adams JB, Smith MO, Gillespie AR (1989) Simple models for complex natural surfaces: a strategy for the hyperspectral era of remote sensing. In: Proc IGARSS Symp, 10–14 July 1989, Vancouver, pp 16–21 Belnap J (1995) Surface disturbances: their role in accelerating desertification. Environ Monitoring Assessments 37:39–57 Belnap J, Gillette DA (1998) Vulnerability of desert biological crusts to wind erosion: the influences of crust development, soil texture, and disturbance. J Arid Environ 39:133–142 Berkowicz SM, Blume HP, Yair A (1995) The Arid Ecosystem Research Centre of the Hebrew University of Jerusalem. Adv GeoEcol 28:1–11 Danin A (1991) Plant adaptation to desert dunes. J Arid Environ 21:193–212 Danin A (1996) Plants of desert dunes. Springer, Berlin Heidelberg New York Elachi C (1987) Introduction to the physics and techniques of remote sensing. Wiley, New York Garcia-Pichel F, Belnap J (1996) The microenvironments and microscale productivity of cyanobacterial desert crusts. J Phycol 32:774–782 Hill J, Mehl W (2003) Geo- und radiometrische Aufbereitung multi- und hyperspektraler Daten zur Erzeugung langjähriger kalibrierter Zeitreihen. Photogrammetrie–Fernerkundung– Geoinformation 1:7–14 Hill J, Mégier J, Mehl W (1995) Land degradation, soil erosion and desertification monitoring in Mediterranean ecosystems. Remote Sensing Rev 12:107–130 Hill J, Jarmer T, Lilienthal H, Yair A (1997) Laboratory spectroscopy: time series of differentially moistened biogenic crusts. Remote Sensing Department, University of Trier, Trier, pp 1–66 Hill J, Udelhoven T, Schütt B, Yair A (1998) Differentiating biological soil crusts in a sandy arid ecosystem based on hyperspectra data acquired with DAIS-7915. In: Schaepman M, Schläpfer D, Itten K (eds) Proc 1st EARSeL Worksh Imaging Spectrometry, RSL, University of Zurich, Zurich, 6–8 October 1998. EARSeL, Paris, pp 427–436
172
J. Hill et al.
Hill J, Hostert P, Atzberger C (1999) Calibration, atmospheric correction and validation of DAIS7915 hyperspectral data. In: Schaepman M, Schläpfer D, Itten K (eds) Proc 1st EARSeL Worksh Imaging Spectrometry, RSL, University of Zurich, Zurich, 6–8 October 1998. EARSeL, Paris, pp 137–146 Hill J, Udelhoven T, Jarmer T (2000) Monitoring recovery processes and rates of a disturbed sandy arid ecosystem, Nizzana (Negev, Israel). Internal Project Rep Part 1, Remote Sensing Department, University of Trier, Trier Karnieli A (1997) Development and implementation of spectral crust index over dune sands. Int J Remote Sensing 18(6):1207–1220 Karnieli A, Tsoar H (1995) Spectral reflectance of biogenic crust developed on desert dune sand along the Israel-Egypt border. Int J Remote Sensing 16(2):369–374 Lange OL, Kidron GJ, Büdel B, Meyer A, Kilian E, Abeliovich A (1992) Taxonomic composition and photosynthetic characteristics of the ‘biological soil crusts’ covering sand dunes in the western Negev Desert. Funct Ecol 6:519–527 Nusch E (1980) Comparison of different methods for chlorophyll and phaeopigment determination. Arch Hydrobiol Beih Ergebn Limnol 14:14–36 Otterman J (1996) Desert shrub as the cause of reduced reflectance in protected versus impacted sandy arid areas. Int J Remote Sensing 17(3):615–619 Otterman J, Ohring G, Ginzburg A (1974) Results of the Israeli multidisciplinary data analysis of ERTS-1 imagery. Remote Sensing Environ 3:133–148 Pfisterer U, Blume H-P, Beyer L (1996) Distribution pattern, genesis and classification of soils of an arid dune area in northern Negev. Zeitsch Pflanzenernäh Bodenk 159:419–428 Price JC (1993) Estimating leaf area index from satellite data. IEEE Trans Geosci Remote Sensing 31(3):727–734 Pye K, Tsoar H (1990) Aeolian sand and sand deserts. Unwin Hyman, London Ray TW, Murray BC (1996) Nonlinear spectral mixing in desert mixing. Remote Sensing Environ 55:59–64 Rendell HM, Yair A, Tsoar H (1993) Thermoluminescence dating of periods of sand movement and linear dune formation in the northern Negev, Israel. In: Pye K (ed) The dynamics and environmental context of aeolian sedimentary systems. Geol Soc Spec Publ 72:69–74 Smith MO, Ustin SL, Adams JB, Gillespie AR (1990) Vegetation in deserts: a regional measure of abundance from multispectral images. Remote Sensing Environ 31:1–26 Verrecchia E, Yair A, Kidron GJ, Verrecchia K (1995) Physical properties of the psammophile cryptogamic crust and their consequences to the water regime of sandy soils, north-western Negev Desert, Israel. J Arid Environ 29:427–437 Walter H (1960) Einführung in die Phytologie. III. Grundlagen der Pflanzenverbreitung, 1. Teil: Standortslehre. Ulmer, Stuttgart Yair A (1990) Runoff generation in a sandy area – The Nizzana sands, western Negev, Israel. Earth Surface Processes Landforms 15:597–609 Yair A (1994) The ambiguous impact of climate change at the desert fringe: Northern Negev, Israel. In: Millington AC, Pye K (eds) Environmental change in drylands: biogeographical and geomorphological perspectives. Wiley, Chichester, pp 199–227
Part C
Ecosystem Processes
Chapter 12
Topoclimate and Microclimate T. Littmann
12.1
Introduction
Unlike macro-scale climatic processes, topoclimate (on a spatial scale, the climate of a terrain of several km2) and microclimate (the climate of relief units comprising the 10- to 100-m2 scale) are strongly interrelated with relief and surface properties such as aspect, slope angle, surface albedo and roughness. Such terrain properties create boundary conditions for specific microclimates which, in turn, will counteract on the associated microhabitats and modify their surface properties until equilibrium state is reached. While there is considerable knowledge of topo- and microclimates in the humid macroclimatic context, little is known about the situation in an arid environment.
12.2
The Radiative Energy Budget and Temperatures on Sand Dune Slopes
Micro-scale ecological processes – e.g. evaporation from the surface, dewfall, light availability for higher plants and biogenic crusts – depend largely on the radiative energy budget of a given relief unit, especially on dune slopes with different aspect. Energy input into the system consists mainly of global radiation, i.e. the sum of beam and scatter solar radiation on site. Figure 12.1 shows the results of model calculations of global radiation for the Nizzana experimental site. Modelling follows a procedure described in Littmann and Kalek (1998), using hourly mean values of global radiation as measured in the horizontal plane, as well as aspect and slope angle of each grid point of a 10 × 10 m grid. In summer, radiative energy input is as much as three times higher than that of winter, but without any significant difference between the north- and south-facing slopes because of the sun’s high elevation. In winter, however, global radiation on the north-facing slopes is about three times lower than on the south-facing slopes, implying winter to be the season with the largest micro-scale differences in the radiation budget. In terms of net radiation of microhabitats within the sand dune relief, however, the actual radiative energy budget depends largely on surface properties (albedo, absorption coefficient), following the law of continuity. As net radiation, temperature and S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
175
176
T. Littmann
39300
440
39200
400 39100 SH
39000
360 Mast
320 38900
W/m2 38800
280
NH
240
DT
38700 200 38600
160
38500
120 89900
90000
90100
90200
90300
90400
90500
90600
90700
90800
Fig. 12.1 Model calculations of global radiation for the Nizzana experimental site (N1) in a winter situation (December)
relative humidity were measured on north- and south-facing dune slopes, we were able to parameterize the terms of the radiative energy budget equation Rnet = Rg • (1−a) − Rt + Ra↓ where Rg is global radiation, α the surface albedo, Rt terrestrial radiation, and Ra↓ atmospheric long-wave radiation as emission from water vapour towards the surface. Rg and Rnet were measured, Rt and Ra↓ calculated following the Stefan-Boltzmann Law and Idso’s (in Oke 1981) approximation of the Angström equation for atmospheric long-wave radiation (equal on both slopes because of no vapour pressure differences at 2 m above ground) respectively, and determined iteratively for hourly values during daytime until the monthly model output showed coincidence with measured Rnet at r2 > 0.9 (Littmann and Kalek 1998). Figure 12.2 shows the corresponding model results for December 1995. The dominant factor effecting differences in the radiative budget of the sand dune slopes is thus simply geometric: at high angles of the sun’s elevation in summer, we do not detect significant differences of any parameter (because of their lower albedo, north-facing slopes may absorb even higher energy) but, in winter, it is the north-facing slopes which show a considerably more negative radiation balance. Negative energy budgets imply a weak potential for water (dewfall) evaporation, an effect discussed in detail below. Furthermore, Littmann and Kalek (1998) showed that the albedo of these opposite slopes follows a distinctive annual course: it is highest during the summer months (south-facing slope 0.28–0.3; northfacing slope 0.17–0.19) and decreases towards the winter rainy season (south-facing slope 0.19–0.23; north-facing slope 0.12–0.14).
12 Topoclimate and Microclimate
177
December1995 Rextraterr = 100% (192 W/m²) Absorption = -26,5%
Absorption = -26,5%
Geometr. Reduction :-35,6 % (Rglobal = 91 W/m2)
Rglobal = 73,4% (141 W/m²)
Rshortwave = 37,8% (73 W/m²)
Rshortwave = 14,0% (27 W/m²)
(plane)
Rlongwave
Albedo = -19,3% (36 W/m²) Rnet: -5,1 % (-9 W/m²)
Geometr. Reduction :-58,4 % (Rglobal = 58 W/m2)
Albedo = -13,1% (25 W/m²)
(Atmosphere)
141,1% (270 W/m²)
140% (260 W/m²)
Rterrestrial = -157,5% (301 W/m²)
Rterrestrial = -164,8% (316 W/m²)
south-facing slope
Rnet: -15,6% (-30 W/m²)
north-facing slope
Fig. 12.2 Model results of global radiation for south- and north-facing slopes for the Nizzana experimental site (N1) for December 1995 Fig. 12.3 Terrestrial radiation (computed from the radiative energy balance equation) and near-ground temperatures for south- and north-facing slopes in A January and B June
450
A: January
400 350 300
W / m²
250 200 450
B: June
400 350 300
North-facing slope South-facing slope
250 200 0
5
10
15
20
25
30
35
40
45
°C
Actually, the complete energy budget of dune slopes would have to include the corresponding fluxes of latent and sensible heat. The flux of latent heat depends on surface evaporation and plant transpiration, and should vary with standing biomass in desert environments. It is difficult to assess in how far near-surface temperatures of the dune slopes are a function of energy transformation processes at (terrestrial radiation) and below the surface (soil heat flux). Figure 12.3 shows that, in June, the ratio of terrestrial radiation (computed from the radiative energy balance equation which, however, would include the fluxes of sensible and latent heat) and near-ground temperatures is equal for both slopes, whereas in January
178
T. Littmann
terrestrial radiation is about 8% lower on the north-facing slope (cf. Fig. 12.2) but temperatures at 20 cm above ground are much higher at the same level of terrestrial radiation, possibly an effect of a much reduced flux of latent heat into the air, implying less evapotranspiration than on the south-facing slope.
12.3
Precipitation
Micro-scale precipitation patterns of rainfall and dewfall may play an important role in the spatial arrangement and quality of cross-dune habitats for higher and cryptogamic vegetation, especially when relief-induced water redistribution is considered (cf. Yair et al. 1997). However, rainfall distribution over the dune–interdune transects could be considered one major boundary condition for the formation of different habitats. In fact, our results from perpendicular bulk samplers (Chap. 19, this volume) over two rainy seasons clearly imply that there is no coincidence of higher rainfall and ecological parameters in various environmental units (Fig. 12.4). It is even the south-facing units which may receive most rainfall, either because of the direction of rain-bearing frontal passages from the southwest or because of the three-dimensionally incident rainfall brought about by the locally disturbed near-ground wind field (Sharon and Berkowicz 1993). Thus, the actual spatial distribution of rainfall over the dune transect is not a controlling agent for the apparent habitat differentiation. More vital in terms of the microclimatic boundary conditions of biogenic crusts are dewfall. Veste et al. (2001) showed the close interrelationship of dewfall and physiological activity of cryptogamic crusts from various in situ field experiments in the Negev sand dunes. Previous studies at the Nizzana experimental station (Littmann and Kalek 1998), based on leaf wetness sensor measurements, showed a very clear difference in the formation and, especially, in the duration of dewfall wetting of the crustal surface on north- and south-facing slopes. While in summer nocturnal dewfall dries up immediately on both slopes immediately after sunrise, it North
South
160 140
200
120 190
100
180
80 60
170
rainfall (mm)
dune height (m a.s.l)
210
40 160
20
150
0 rainy season 1994/95
rainy season 1995/96
Fig. 12.4 Rainfall variations across linear dunes at the Nizzana experimental site (N1) in 1994/1995 and 1995/1996
12 Topoclimate and Microclimate
179 7:00
6:30
6:00
5:55
N
Fig. 12.5 Shading in March from sunrise (5:55 a.m.) until early morning (7:00 a.m.) at the Nizzana test site (SD). At 7:00 a.m., only the concave parts on the north-facing slopes and some parts of the dune base are shaded
may last up to 3 h longer on the north-facing slope in winter (cf. Chap. 21, this volume). These results are well in tune with computations of the diurnal course of the dewpoint temperature difference, and with water content measurements of the crusts after sunrise (Littmann and Kalek 1998). As for the radiative energy balance in winter, it can be also shown here that radiation geometric differences are the main reason for such differences in dewfall duration. A modelling of the surface shadowing in the first hour after sunrise (Fig. 12.5) clearly reveals that the longer lasting shadow “pockets”, where we find the crusts with greatest thickness and quality, are the concave sinuous parts of the north-facing slopes; this enables dewfall to wet the crustal surfaces longer before final desiccation about 3–4 h after sunrise in winter. Furthermore, the interrelationship of dewpoint temperature differences<1 K and the formation of dew on leaf wetness sensors, corresponding to physiological activity of the crusts as described in Veste et al. (2001), seems to hold for all parts of the sand dune field, irrespective of water vapour pressure (Chap. 21, this volume) or vegetation cover.
12.4
Near-Ground Wind Conditions
Except for occasional storm events, the entire study area is a weak wind energy area. Figure 12.6 shows seasonal characteristics of wind speed and wind direction distributions at 2 m a.g.l. in the northern, central and southern sand dune field. Because of the
180
T. Littmann 30 25 20
A: spring
15 10 5 0 30 25
B: summer
20
Frequency (%)
15 10 5 0 30 25
C: fall
20 15 10 5 0 30 25
D: winter
20 15 10 5 0 <0.2 0.5
1.0 1.5 2.0 2.5
3.0
3.5 4.0 4.5 5.0 >5.0
Wind speed intervals (m s−1) North
Central
South
Fig. 12.6 a Seasonal wind speed frequency in the southern (N1), central (N3) and northern areas (N5). b Seasonal characteristics of wind speed and wind direction distributions at 2 m a.g.l. in the northern and southern sand dune field
seasonal occurrence of Sharav dust storm events and the associated passage of Saharan depressions, wind speeds in the southern part are slightly higher in spring and summer, while turbulence leads to quite untypical frequency distributions of wind speed in all parts over the summer half year. It is only in winter that all distributions follow the typical Weibull function. However, Weibull A and k factors remain well
12 Topoclimate and Microclimate
181
below 5 m s−1 and 2.0 respectively. Thus, it is not the regional near-ground wind field which controls sand movement and dune mobility in a regular seasonal pattern – this occurs only in the case of rare storm events (cf. Littmann and Gintz 1998). Seasonal wind directions follow the main seasonal synoptic patterns over the area, with north-westerly Mediterranean winds dominating in summer and continental south-easterlies in winter. In the northern part of the dune field, however, the south-easterly component is much weaker, which is not plausible in terms of the overall regional pressure patterns. When comparing monthly means of surface roughness length, computed from wind speed measurements (hourly means) and the friction velocity proportional to wind speed (Fig. 12.7), friction velocity should show an exponential increase
0.5 0.4 0.3 0.2 0.1
north
0 0.5 0.4 0.3
u* [m/s]
0.2
north central
0.1
R2 = 0,61
0 0.5 0.4 0.3 0.2 0.1
south central
0 0.5 0.4 0.3 0.2
south
0.1
R2 = 0,80
0 0
0.1
0.2
0.3
0.4
z0 [cm] Fig. 12.7 Roughness length z0 and friction velocity u* at the measurement stations along the transect
182
T. Littmann
with the roughness parameter. In fact, this interrelation is observed only in the southern sector of the sand dunes at Nizzana, whereas in the northern parts the correlation is weak or virtually absent. As monthly differences in the roughness parameter are an effect of changes in vegetation cover and density (annuals, rainy season canopies, etc.), our wind measurements imply that such changes are insignificant in the northern and central dune field with small seasonal roughness variation. Probably also the smaller frequency of south-easterlies is an effect of more effective surface shielding by higher vegetational stands.
12.5
Conclusions
The examples briefly discussed above clearly show that major ecological processes and structures such as biogenic crusts and standing biomass distribution are well defined by microclimatic boundary conditions which, in turn, are a function of sand dune morphology and geometry. In this way, partially stabilized sand dunes do not show microclimatic characteristics significantly different from those of other relief units with slopes of different aspect. The specific property of linear sand dunes in an arid area, however, is that their geomorphologically defined microclimatic ecotopes trigger the formation of habitats for biogenic crusts which, in turn, act as a regulative factor for other processes, thereby resulting in typical spatial patterns of vegetation and of stable and unstable environmental units.
References Littmann T, Gintz D (1998) Eolian transport and deposition in a partially vegetated linear sand dune area (northwestern Negev, Israel). Zeitsch Geomorphol suppl Bd 121:77–90 Littmann T, Kalek J (1998) Mikroklimatische Strukturen als Steuergröße für Ökosystemprozesse in einem ariden Dünengebiet (nordwestlicher Negev, Israel). Hallesches Jahrb Geowiss Reihe A Geogr Geoökol 20:77–92 Oke T (1981) Boundary layer climates. Routledge, London Veste M, Littmann T, Friedrich H, Breckle S-W (2001) Microclimatic boundary conditions for activity of soil lichens crusts in sand dunes of the north-western Negev Desert, Israel. Flora 6:465–474 Yair A, Lavee H, Greitser N (1997) Spatial and temporal variability of water percolation and movement in a system of longitudinal dunes, western Negev, Israel. Hydrol Processes 11:43–58
Chapter 13
Evapotranspiration, Transpiration and Dewfall T. Littmann and M. Veste
13.1
Introduction
Actual evapotranspiration is the most critical parameter in hydrological water balance models. The knowledge of atmospheric, plant and soil interactions in drylands is limited. Many evaporation models have been developed, but mainly for agricultural crops. The best known is the single-source Penman-Montheith evaporation model (Monteith 1965). This model assumes that canopies can be regarded as one uniform surface or big, single leaf. Most of the models have been largely successfully used for estimating evapotranspiration from vegetation which is not drought-stressed and relatively uniform, such as in agricultural fields. However, arid and semi-arid regions are characterized by patchy vegetation and larger open spaces. In an arid environment, especially in sandy areas where surface runoff is of no practical importance in the hydrological budget, it is rainfall, dewfall, and evaporation and plant transpiration which constitute the most relevant parameters. Dewfall may become an input variable at least seasonally even more important than rainfall (Zangvil 1996). In the Negev sand dunes, especially dew has an important ecological implication for the activity and distribution of biological soil crusts and lichens (Chap. 21, this volume). In this paper, we will apply and compare some of the most common approaches to compute actual evapotranspiration from field measurements in the sandy north-western Negev Desert in Israel, and introduce a practical model specifically applicable to arid environments. In this context, dewfall will be inferred as a reversal of the evaporative process.
13.2
Microclimatic Measurements
Microclimatological measurements were carried out at two recording stations in the study site Nizzana on the encrusted parts of dune slopes facing north and south, from July 1995 to June 1996. Measured parameters (hourly means) relevant for this investigation include net radiation (2 m above ground), soil heat flux (−0.15 m), air S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
183
184
T. Littmann, M. Veste
temperature (2 and 0.2 m), relative humidity (2 m), and horizontal wind speed (2 and 0.5 m). As relative humidity was measured only on the north-facing slope, we will present results for this site except for the eddy correlation technique which was applied to the opposite slope where also vertical wind speed data were available. This conventional arrangement of field measurement necessarily leads to some inaccuracy which has influence on the model results once gradients within an air column are introduced, e.g. critical temperature and vapour gradients between 0.2 m and the actual surface have to be neglected. However, most of the methods applied here derive data from standard measurement heights (Schrödter 1985). As we had to compute relative humidity at 0.2 m from humidity data at 2 m without an additional measurement at the lower height, all results including gradients will possibly underestimate the actual situation to a certain extent. Dewfall amounts were additionally measured by means of an acrylic condensation plate (20×20 cm) attached to a pressure-transforming load cell calibrated daily.
13.3
Evapotranspiration Models and Their Application to Dewfall
Roughly, most approaches to approximate actual evapotranspiration are either vapour flux models, energy balance models or physical-empirical models (Schrödter 1985). A vapour flux model may be based on gradients of properties of a nearground air column following Fick’s law of diffusion – such as wind speed and specific humidity. Introducing an explicit form of the turbulent diffusion coefficient of momentum, Thornthwaite and Holzman’s (1942) gradient-flux model considers evapotranspiration (ET) as the upward (positive) flux of specific humidity between two heights above ground enforced by the vertical wind shear: ET = r k
(qz1 − qz 2 ) (uz 2 − uz1 ) ⎛z ⎞ ln ⎜ 2 ⎟ ⎝z ⎠
2
(13.1)
1
where u1 and u2 are the horizontal wind speed at heights z1 and z2 respectively, q1 and q2 the corresponding specific humidities, r the air density, and k the Karman constant. As this method requires a determination of the logarithmic wind profile above the roughness height z0 which, in turn, is the aerodynamic surface where turbulent vapour flux may start, we computed mean monthly z0 from best-fit hourly adiabatic and diabatic wind profiles between 2 and 0.5 m, after Littmann (1994), including the Richardson number as stability parameter. The determination of neutral, stable or labile situations followed near-ground temperature profiles. In our case, z0 remained constantly at 0.19 m on the north-facing dune slope, which coincides with the lower height of conventional temperature measurements. The actual wind profile for each hour was then computed from 2 m down to z0.
13 Evapotranspiration, Transpiration and Dewfall
185
As no pressure measurements were available, humid air density was computed using standard pressure. However, we found that even strong pressure fluctuations will not have large effects on the corresponding values. The same applies to the calculation of specific humidity from relative humidity and saturation pressure. For the 0.2-m level, air temperature at this height was used for the corresponding computations. The Thornthwaite-Holzman method expresses upward flux (i.e. evapotranspiration from a surface) in terms of positive values, while negative values denote a downward vapour flux. It is this turbulent downward flux which leads water vapour to the surface where it may condensate under appropriate boundary conditions. In this way, turbulent downward movements may be considered potential dewfall. As dew is a phenomenon whereby water vapour condenses on a substrate and transforms into liquid water, once the saturation pressure at the temperature of the substrate is lower than the saturation pressure at air temperature (Beysens 1995), the main source of vapour in a natural environment is the air itself (Jacobs et al. 1998). Once vapour is deposited on the surface by downward flux, it will be thermodynamically stable only in the case of droplet formation, which requires large thermal fluctuations to overcome the cost in free energy of forming a liquid–vapour interface, i.e. to overcome interfacial tension (Beysens 1995). Following Thompson’s theorem, the nucleated droplet is stable when its radius exceeds the critical radius given by the ratio of interfacial tension and the gain in droplet volume energy (Roedel 1992). Such homogeneous nucleation, however, requires oversaturation exponentially equivalent to 1/droplet radius, and very small droplets would be expected. On the other hand, natural surfaces tend to lower the barrier of free energy by their physical and chemical properties, such as hygroscopy, resulting in lower saturation pressure and small dewpoint temperature differences (Roedel 1992), and specific wetting conditions which depend on the contact angle of the droplet on the surface (Beysens 1995) or on the vapour pressure difference of the near-surface soil pores and the diffusive layer above the surface (Jacobs et al. 1998). However, any approach to infer dew deposition from evapotranspiration models depends on the way condensation at the surface is considered potentially possible. In an empirical approach, the critical dewpoint temperature difference can be determined as (Häckel 1990) ⎛ 234.67 log e − 184.2 ⎞ dtd = ta − ⎜ ⎝ 8.233 − log e ⎟⎠
(13.2)
where ta is the air temperature in °C, and the term in brackets is the dewpoint temperature with e as water vapour pressure (hPa). Dewpoint temperature difference provides a criterion to evaluate vapour flux towards the surface as a condensation event once it becomes smaller than the critical threshold for pre-condensation. We iterated the ratio of hours with downward vapour flux after the ThornthwaiteHolzman method, and the number of hours with a dtd lower than a given value over a wide range of dtd values. As Fig. 13.1 shows, this ratio remains stable in a range of 1 K < dtd < 3 K, approximately equivalent to relative humidities > 80% and, thus, within the range of pre-condensation (Roedel 1992).
186
T. Littmann, M. Veste 1
0.8
ratio
0.6
0.4
0.2
0
0
1
2
3
4
5
6
8
10
12
14
dtk[K] Fig. 13.1 Limits of dew formation: definition of the critical dewpoint filter. x-axis Dewpoint temperature difference, y-axis ratio of hours with negative vapour flux gradients and total number of hours per day
Janssen and Römer (1991) introduced a method to determine dew hours and the overall duration of dewfall by means of a parameterization of the influence of net radiation, and sensible and latent heat on a dew plate, under the assumption that dewfall may occur when the flux of latent heat Hl is < 0. From the relations of Hl and the friction velocity u* for several classes of night-time net radiation equalling cloud cover classes, and for some values of the specific humidity deficit (calculated as the difference of specific humidity saturation values at air and dewpoint temperatures), they provide three empirical limit values for u* for three corresponding net radiation classes. If the measured wind speed is <10 × u*crit, then dewfall hours will occur. Figure 13.2 compares the frequency of cases remaining under the two corresponding criteria versus the frequency of cases where leaf wetness sensors installed at 0.2 m indicated condensation events. Both methods to define dewfall events show 88% correlation to the measured data. As the interrelation of both approaches is very good, we apply the more practical dewpoint filter suggested here to all subsequent model outputs. A direct method to estimate both evapotranspiration and dewfall is the eddy correlation technique (Frankenberger 1955; Swinbank 1955; Jacobs et al. 1998). Based on the fact that turbulent exchange of properties in the boundary layer is triggered by the friction velocity, surface roughness, and gustiness in terms of Reynold’s covariance, which considers the actual wind speed as the sum of mean wind speed over a period of time and the momentary variance, the approach correlates the vertical fluctuations of wind speed and specific humidity over a certain period:
13 Evapotranspiration, Transpiration and Dewfall
187
condensation frequency [%]
100
80
60
40
20
dew point filter
r2 = 0,87
Janssen - Römer
r2 = 0,88
0 0
10
20
30
40
50
60
70
80
90
leaf wetness condensation frequency [%]
Fig. 13.2 Dew formation: condensation frequencies for leaf wetness sensors versus dewfall frequencies, based on the dewpoint filter and the Janssen-Römer criterion
( rw ) ET = ( r w )’q ’
(13.3)
where ( rw ) is the mean vertical flux of mass, and ( rw ) and q ’ are the corresponding mean deviations of mass flux and specific humidity, based on their means over the observation period. Physically sound models should include the energy budget at the surface: Rn − G = Hs + Lv ET
(13.4)
where Rn is the net radiation, G the soil heat flux, Hs the flux of sensible heat, Lv the latent heat of vaporization (2.45×106 J kg−1), ET the evapotranspiration, and LvET the flux of latent heat (latent enthalpy). When defining the evapotranspiration ET as ET =
( Rn − G − Hs )
(13.5)
Lv
measurements of the flux of sensible heat are required. Using this definition, Hicks (1983) found a good interrelation of negative values of ET at night and dewfall rates in an arid region in Australia. Substituting ET in Eq. (13.4) by Lvq would result in a calculation of the evapotranspiration in terms of q at a given time, which will strongly underestimate the actual situation.
188
T. Littmann, M. Veste
Including the energy budget and a parameterization of turbulent flux, the Bowen Ratio Energy Balance (BREB) technique originally presented for the Sverdrup (1936) method is widely accepted as the physically most sophisticated approach (cf. Schrödter 1985): Lv ET = −
Rn + G ∂t / ∂z 1+ g ∂e / ∂z
(13.6)
where g is the psychrometric constant (0.66 hPa K−1), ∂t/∂z the temperature difference between two heights of measurement, and ∂e/∂z the corresponding vapour pressure difference. LvET/Lv converts the flux of latent heat into actual evapotranspiration values (mm), which become negative at night in the case of negative net radiation. Such cases, however, represent potential dewfall, as the method does not consider a downward vapour flux but rather the water equivalent of latent heat in the air column between the two measurement heights. Monteith’s (1957) definition of the potential rate of dew formation, Dp = Rn
s Lv ( s + g )
(13.7)
where s is the slope of the saturation vapour pressure function, explains this phenomenon explicitly for night-time conditions with negative Rn (cf. Garrat and Segal 1988). For the purpose of assessing the plausibility of the different model outputs, it is feasible to use the potential evapotranspiration as an upper limit value. In the context of potential evaporation from open water surfaces, the combined method after Penman (1948) has been applied to a wide range of world climates (Henning and Henning 1984). Strictly physical, the approach includes a parameterization of the energy balance and the influence of wind: ETpot =
u ⎞ ⎛ s s ⎞ ⎛ Rn + ⎜ 1 − ( E − e) 0.27 ⎜⎝ 1 + l ⎟⎠ ⎟ 100 ⎝ s+g ⎠ s+g
(13.8)
where E is the saturation vapour pressure and ul is the mean daily wind length. Also developed for non-arid climates, Penman’s rather simplified wind function will lead to a larger distortion of results in the case of high surface roughness or advective processes which obliterate the diurnal course of wind speed and water vapour in the boundary layer (Schrödter 1985). Using a daily mean sum of positive net radiation, the Penman approach applies only to daily means of evapotranspiration and is not suitable for a higher time resolution in its present form. There have been efforts to develop correction factors for various climates (Doorenbos and Pruitt 1977) or to define monthly minimum thresholds for humid areas (Penman 1948). However, an application of the combined Penman formula to irrigated land in an arid region (Ohlmeyer and Hoyningen-Huene 1975) showed that the results for potential evapotranspiration even underestimated actual values by 27%.
13 Evapotranspiration, Transpiration and Dewfall
13.4
189
Zero Plane Model
One of the main problems of approaches using vapour gradients between two heights of measurement originates from the fact that vapour fluctuations resulting from evaporative or condensation processes will be much larger immediately above the active surface, compared to a larger height. Using the mean specific humidity of the air column under consideration, or its gradient between two heights without including air movement as a purely empirical boundary condition will underestimate actual evapotranspiration and, especially, dewfall; including wind speed or wind shear, as in the Thornthwaite-Holzman Eq. (13.1), will lead to overestimations. On the other hand, gradient-free approaches such as the eddy correlation technique will fail in high-relief terrain when turbulent vapour flux is enforced by topographic obstacles. Alternatively, the passage of vapour through a two-dimensional plane near the active surface – which may be the canopy layer in the case of vegetational stands, the bare surface or, most likely, a height in-between these – reflects (depending on its direction) either evapotranspiration or potential condensation in terms of a fluctuation of properties (i.e. sensible and latent heat) as a function of time. We assume that all processes (stability, turbulence, advection, evapotranspiration, dewfall) will affect fluctuations of specific humidity on such a near-surface plane at the relative bottom of the air column involved in the processes, irrespective of the actual height or volume of the column. However, any such approach requires vertical fluxes which are relatively free of advective disturbance. Advection may be excluded only in the case of a measurement area large enough to minimize divergence of the humidity field, as in most humid regions (Schrödter 1985). Especially in arid regions, the influence of advection cannot easily be neglected, as it may lead to either an increase in actual evapotranspiration relative to the water equivalent of the radiative budget, because of a much larger saturation deficit (cf. Ohlmeyer and Hoyningen-Huene 1975), or to a decrease when moist air is advected – e.g. in the case of sea breezes, as present in our study area during the summer half year (Littmann 1997). Although wind speeds at 2 m show a close interrelation with wind directions (66% explained variance in one-way analysis of variance), which explains the diurnal course of wind speed in terms of the land and sea breezes, specific humidity does not (5% explained variance by both wind direction and wind speed). This implies that periodic daily advection by the regional wind systems does not play a major role. In an arid region of China, Niu et al. (1997) found, by means of a coupled soil–atmosphere model, the vertical flux of latent heat to control specific humidity up to a height of 100 m. Naturally, vapour or dry air advection is coupled to fluctuations in horizontal and vertical wind speed on different time scales. Paw et al. (1995) reported that the advective term in energy exchange within a canopy layer is represented by highfrequency scalar traces such as temperature change within the 100 to 10−1 s time range. However, advective effects will be largest in the case of frontal passages or
190
T. Littmann, M. Veste
khamsinic depressions over the area. To exclude part of such disturbance, we did not consider cases which showed anomalous wind speeds relative to the mean monthly diurnal course of wind speed (roughly 10% of the overall series), following the anomaly filter:
(uTn − uTn ) > σ n
(13.9)
where uTn is the actual data n at time T of the series, uTn the mean at time T, and sn the standard deviation. As fluctuations occur in terms of variance around a mean value, we chose to split the monthly mean of specific humidity at the measuring height of 0.2 m into 24-h means in order to observe vapour fluctuations in terms of explained variance inbetween these group means. The following model was developed in three consecutive steps (Fig. 13.3):
1.0
7
0.8
6
0.6
5
0.4
4
0.2
22
20
18
16
14
−1
12
−0.6 10
−0.4
0 8
1 6
−0.2
4
0.0
2
2
3
dq [g*kg−1]
8
0
q, cum. dq [g*kg−1]
1. Vapour fluctuations within the plane of observation (strictly controlled by evapotranspiration and condensation processes) are expressed as a differential series of the mean hourly values of specific humidity dq. They may be positive in the case of an upward vapour flux from the active surface of evapotranspiration (soil and canopy) through the plane, while an opposite flux from above will not be detectable as positive dq because it may occur in condensation situations only (except for rare cases of strong turbulent advective downward mixing). Decreasing dq reflects either the condensational removal of water vapour from the near-ground air column or constantly decreasing vapour input from the active surface. In fact, we found that dq values at 0.2 and 2 m are closely interrelated (r2=0.78) with no time lag but, at 0.2 m, i.e. close to the active surface, dq is
−0.8
local time cumulated dq
q
dq
Fig. 13.3 Evolution of the zero plane model. Discrete mean hourly values of specific humidity (q) over 1 month in the plane of measurement are differentiated (dq), and dq is expressed as a cumulative series with a zero start value. The shaded area is used for integration of mean monthspecific humidity fluctuations in the zero plane (here, the abscissa is taken as zero line)
13 Evapotranspiration, Transpiration and Dewfall
191
much more variable than at 2 m (σ=0.9 vs. 0.7). The use of hourly means excludes high-frequency turbulent signals which would result in no vapour fluctuation on the 100 to 10−1 s time scale, in many cases because of equalling upward and downward passages (A.F.G. Jacobs, personal communication). 2. Actual evapotranspiration should be the total amount of water vapour passing through the plane at time intervals ∆T in terms of an increase in specific humidity, and dewfall the total amount removed from the plane. However, removal or increase within the mean time series sets is relative to the prior fluctuation. This is represented by the cumulated dq series and is consistent with the principle of continuity, as cumulated dq also represents the flux of latent heat through the plane. A similar approach, including the introduction of finite differences and summation of such derivatives with reference to heat exchange in canopies, was introduced by Paw et al. (1995). 3. Finally, the cumulated series is integrated following the determination of precipitable water (Peixoto 1973): ET = z rw
T = 24
∫
f (q(1...24 ) − q1 ) > 0
(13.10)
f (q(1...24 ) − q1 ) < 0
(13.11)
T =1
D = z rw
T = 24
∫
T =1
where z is the unit height of the plane (1 m) and ρw is the density of water vapour. We use the functions of the differential specific humidity series (which show a start value of zero at 0:00 local time, thus “zero plane”) because it is identical with the cumulated dq series. Equation (13.10) includes the value of the integrated functions, as it may occur that the differential series leads to negative values during daytime when the specific humidity decreases relative to the zero start value. This is the case when either dry air is advected during khamsinic situations in spring (March, April, May), not detectable by the anomaly filter (Eq. 13.9), or transpiration is greatly reduced in the afternoon (summer). Night-time negative values following Eq. (13.11) were subjected to the critical dewpoint filter, and we found a perfect coincidence of hours where condensation is physically possible (i.e. where the mean frequency of cases with dtd<3 K is high until it decreases to zero during the first hour after sunrise when net radiation values become positive) and those where the model output indicated a decrease in specific humidity.
13.5
Application of the Zero Plane Model
Figures 13.4 and 13.5 show the zero plane model results for January 1996 and July 1996. Specific humidity decreases from the zero start value until sunrise and the onset of labilization during the first hour thereafter, which is also matched by the dewpoint filter frequencies and by the frequency of near-ground stability. In
192
T. Littmann, M. Veste 2.5
14 12
2.0 m*s−1;10−2 %;g*kg−1
10 8
hPa
1.5
6
1.0
4 2
0.5 0 0
−2
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
0.0
local time
cumulated dq
dew point filter
stable layer frequency
wind speed 0,5m
saturation deficit
Fig. 13.4 Zero plane model output for January 1996 (24 discrete values for mean daily courses). Note that cumulated dq (g kg−1), dewpoint filter, and stable layer frequencies (in 10−2%) and wind speed at 0.5 m (m s−1) are on the left y-axis, saturation deficit (hPa) on the right y-axis 5
40 35
4
−1 −2
30 3
2
hPa
20
−1
m*s ;10
%;g*kg
25
15 1 10 0 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
−1
19
20
21
22
23 5 0
local time cumulated dq
dew point filter
wind speed 0,5m
stable layer frequency
saturation deficit
Fig. 13.5 Zero plane model output for July 1996 (24 discrete values for mean daily courses). Cumulated dq (g kg−1), dewpoint filter, and stable layer frequencies (in 10−2%) and wind speed at 0.5 m (m s−1) are on the left y-axis, saturation deficit (hPa) on the right y-axis. (Note: values for summer are several times higher than for winter; cf. Fig. 13.4)
both examples as well as in all individual months, specific humidity shows a sharp increase for the first 3 h after sunrise towards the daily maximum at 8:00 to 10:00 or 10:00 to 12:00, depending on the season. This maximum is well before the saturation deficit maximum, which indicates that actual evapotranspiration is not
13 Evapotranspiration, Transpiration and Dewfall
193
simply controlled by the saturation deficit (analysis of variance results in 13% overall explanation of specific humidity fluctuations in terms of the saturation deficit). However, in the afternoon when the saturation deficit is highest, and also the near-ground wind speed reaches its maximum, water vapour decreases to values near or well below the zero start value, except for a small secondary peak in July. After sunset, specific humidity increases again to a level well above the start value, paralleling the development of a stable layer. As wind speed, wind direction and the saturation deficit all explain only 20% of total specific humidity variance in the zero plane model, it is apparent that the daytime fluctuations are controlled mainly by the diurnal cycle of transpiration typical for most desert plants. Coupled with stomata opening and increasing VPD, transpiration increased after sunrise. However, drought-stressed plants actively reduced the transpiration rate by closing the stomata when the air to leaf vapour pressure difference became higher. The importance of the stomatal control of transpiration for the estimation of total evapotranspiration has been emphasized by various ecophysiologists (cf. Jarvis and McNaughton 1986; Baldocchi 1993). In the sand dunes of Nizzana during and after a good rainy season, the shrub Thymelaea hirsuta showed no pronounced stomata closure at midday, contrary to what could be expected for desert plants. Transpiration rates decreased constantly after 12:00 local time (Chap. 19, this volume; Veste and Breckle 1996a). In this case, and also for other deep-rooting shrubs such as Retama raetam and Anabasis articulata, water uptake from the lower root zone of the plants enabled higher transpiration rates. After a long-term drought and a bad rainy season and during the dry season, however, the shrubs showed stomata closure as well as a gradual reopening in the late afternoon (Veste and Breckle 1996b). In an experiment, the measured diurnal courses of transpiration of Artemisia monosperma match water vapour fluctuations on the zero plane to a high degree. However, water vapour decreases on the zero plane in terms of a downward flux towards the surface of condensation only after the near-ground air temperature is low enough to reduce the dewpoint temperature difference below the critical value, i.e. increasingly during the early morning hours before sunrise. Comparing the results of all different models (Fig. 13.6 and Table 13.1) shows fairly similar values of evapotranspiration for the winter months (November to February). Over the summer months, however, deviations become extreme. It is especially the Thornthwaite-Holzman method which models monthly totals close to potential evapotranspiration after Penman. Such exaggeration is unavoidable when a method multiplies unweighted gradients of specific humidity and wind speed; both are high in summer and low in winter, a situation resembling humid environments where at least the specific humidity gradient is small throughout the year. Based primarily on the energy budget, the Sverdrup (BREB) model shows a very similar annual course with an extremely low value in December, but also fairly high summer totals because it will yield higher water equivalents when both net radiation and vapour gradients are high, as in summer. Except for winter, the second model group shows not much similarity to these two methods. Both
194
T. Littmann, M. Veste
Evapotranspiration [mm]
160 140 120 100 80 60 40 20 0 Jul 95
Aug 95
Sep 95
Oct 95
zero plane
Nov 95
Dec 95
Jan 96
Thornthwaite
Feb 96
Mar 96
Sverdrup
Apr 96
May 96
Jun 96
Eddy
Fig. 13.6 Monthly evapotranspiration in Nizzana (1995/1996) using different model approaches
Table 13.1 Model data for dewfall and evapotranspiration (ET) at the Nizzana site (N1) between July 1995 and June 1996 Monteith or ThornthwaiteEddy Penman (pot.) Holzman Sverdrup correlation Zero plane Load cell Dewfall Jul. 95 Aug. 95 Sep. 95 Oct. 95 Nov. 95 Dec. 95 Jan. 96 Feb. 96 Mar. 96 Apr. 96 May 96 Jun. 96 Total ET Jul. 95 Aug. 95 Sep. 95 Oct. 95 Nov. 95 Dec. 95 Jan. 96 Feb. 96 Mar. 96 Apr. 96 May 96 Jun. 96 Total
Monteith (pot.) 14.04 15.51 19.14 17.67 18.79 17.94 12.33 14.65 18.56 18.60 20.09 18.77 192.11 Penman (pot.) 139.17 151.98 119.87 88.12 84.22 41.68 61.48 80.90 121.82 151.49 179.02 160.78 1,380.54
3.57 5.48 5.85 4.44 2.73 6.02 5.18 3.33 3.23 2.70 1.94 4.22 48.70
18.60 11.36 13.03 11.32 6.39 12.76 10.05 8.87 10.90 13.12 10.38 12.76 139.53
0.77 0.64 0.66 2.34 0.56 0.81 0.63 0.61 0.70 0.46 1.01 1.52 10.72
1.88 3.67 2.59 1.13 1.10 2.25 1.20 3.40 1.85 2,32 1.44 3.12 25.95
107.73 146.47 107.44 62.85 25.90 16.58 15.94 25.24 44.26 61.11 94.39 128.74 836.65
24.04 85.26 61.45 33.79 9.13 5.99 12.16 17.76 37.55 65.98 76.22 89.01 518.35
29.74 30.56 27.56 33.56 27.55 12.32 18.07 21.41 22.01 12.64 33.94 48.19 317.53
12.15 15.02 20.77 22.01 8.86 12.92 8.27 10.05 9.12 4.93 14.01 15.21 153.32
2.99 4.16 3.14 1.93 3.37 2.31 2.95 2.17 2.54 2.62 2.90 2.70 33.77
13 Evapotranspiration, Transpiration and Dewfall
195
the Eddy correlation technique model and the zero plane model are not dependent on the height of the air column considered in the gradient approaches, and both methods use mean values over the observational period. Although three values (July to September) were generated by regression due to lack of vertical wind speed data, except for December the Eddy correlation model consistently shows higher results than does the zero plane model, as it equally includes vertical wind speed fluctuations in terms of turbulent transport whereas the zero plane approach considers only the actual specific humidity fluctuations. Thus, there is a wide range in the model results for annual evapotranspiration: 836 mm for Thornthwaite-Holzman, 518 mm for Sverdrup, approx. 317 mm for the eddy correlation, and only 153 mm for the zero plane model. We will consider these results in our conclusions. For dewfall, the Sverdrup model results in monthly totals which are close to the potential dewfall after Monteith and, in one case (July), even higher (Fig. 13.7). Also the Thornthwaite-Holzman model output is comparatively high. Both model results should be considered quasi-potential dewfall, as they imply a homogeneous condensational flux of vapour over a given air column, irrespective of the application of the dewpoint filter. Only in the case of small specific humidity gradients, as in February to May, may the ThornthwaiteHolzman model give results comparable to those of the gradient-free approaches. In our case, the eddy correlation technique modelled extremely low dewfall, which (different to its results for ET) is an effect of very low night-time vertical wind speed in a stable layer near the surface. It is only for months with night-time convection induced by khamsinic depressions (May, October) that this model results in dewfall amounts comparable to those produced by the zero plane model.
20 18
Dewfall [mm]
16 14 12 10 8 6 4 2 0 Jul 95
Aug 95
Sep 95 zero plane
Oct 95
Nov 95
Dec 95
Thornthwaite
Jan 96
Feb 96
Sverdrup
Mar 96 Eddy
Apr 96
May 96
Jun 96
load cell
Fig. 13.7 Monthly dewfall in Nizzana (1995/1996) using different model approaches
196
13.6 13.6.1
T. Littmann, M. Veste
Model Validation Validation of Dewfall
As confirmatory measurement for the dewfall, we used the mean load cell data from two acrylic plates, one of which was covered with styrofoam underneath to prevent condensation on the lower face of the plate. Nightly dewfall was calculated as the individual daily range, which excludes any influence of contamination of the plate by sand or dust. Indeed, the use of hourly means removes the oscillating noise of plate movements induced by wind. The annual total (2 months had to be interpolated, due to lack of data) of approximately 33 mm is well within the range of the zero plane model output (26 mm), although the series do not show good correlation over the winter months (Fig. 13.7). Both results are comparable to those obtained by Evenari et al. (1982) at Avdat, using Duvdevani wooden blocks (30 mm), and by Zangvil (1996) using a modified Hiltner dew balance at Sede Boqer, 40 km southeast of our experimental site (18 mm). Evidently, our model and measurement results are quite consistent with other data from the region, while the other methods used here provide much too high (139 mm for the Sverdrup method), fairly high (48 mm for Thornthwaite-Holzman) or fairly low (10 mm for the eddy correlation technique) results. However, we did not find a pronounced annual course of dewfall (Fig. 13.8), as indicated by Zangvil (1996) with maxima in August to October and December and January, and minima in April and November. August, September and December show also high values in our series but also do February, April and June. It seems that even during frequent night-time khamsinic situations, as in April 1996, near-ground vapour pressure was high enough to enable considerable dewfall (Veste et al. 2001). Extrapolating results from an extensive short-term field campaign at the Nizzana test site in September 1997, Jacobs et al. (1998) found dewfall of around 5 mm for this month. Their modelled results of 0.1 to 0.2 mm of dewfall per night are similar to our findings (Jacobs et al. 1999; Chap. 21, this volume).
13.6.2
Transpiration and Evapotranspiration
Transpiration rates of dominant plants should also be an indicator for model evaluation. Because no long-term measurements exist for our experimental sites, we conducted an experiment in March 1999 in the most northern experimental site at Yevul. At this site, the dominant shrub is A. monosperma (Veste et al. 2005). The average plant cover in the vicinity of the micrometeorological station is 35%. Diurnal courses of transpiration were measured at 8-day intervals using a portable porometer system HCM-1000 (Walz GmbH, Effeltrich, Germany). Water vapour exchange was analyzed with an infra-gas analyser (BINOS 100, Rosemount, Hanau, Germany) and the fluxes calculated after von Willert et al. (1995). The porometer head is climatized with Peltier elements, and air temperature and water
13 Evapotranspiration, Transpiration and Dewfall
wind speed ( m s−1) temperature (°C)
0
B
20
air temp.
30 20
10
10
vpd 0
0
transpiration
C
12
2
spec. humidity
8
1
transpiration (mmol m−2s−1)
spec. humidity (g kg−1)
500
0
VPD
1000
1
(mPa Pa−1)
1500
wind
2
PPFD
2000
PPFD
A
3
(µmol m−2s−1)
4
197
0 0
8
16
26 March
0
8
16
27 March
Fig. 13.8 Daily courses of microclimatic parameters (A wind speed, PPFD, B air temperature and VPD, C specific humidity) from the microclimate station, and transpiration (C) of Artemisia monosperma measured with a porometer system at Yevul (26–27 March 1999)
vapour deficit do not differ between ambient air and the cuvette (Fig. 13.8B, Midgley et al. 1997). An example of the transpiration and microclimatic parameters are shown in Fig. 13.8. For the extrapolation of the leaf area-based transpiration rates to soil area-based fluxes, all leaves of A. monosperma in a representative 5×5 m plot were harvested and the leaf area index was calculated as LAI=0.12 m2 m−2. The mean daily water loss of A. monosperma calculated from the gas exchange data was 1.6 kg m−2 leaf area day−1. The estimated monthly water loss of the A. monosperma stand is (in March) approx. 5.8 mm per month. In the northern Negev near Beer Sheva, Zohary and Orshan (1954), in a Zygophyletum dumosi association on loessic soil, found mean transpirational water losses of 70–80 mm year−1. The transpirational water loss of the Artemisia shrub community is around 35% of the total evapotranspiration calculated with the zero plan model. In other arid and semi-arid shrub communities, the percentage of transpiration attributable to total evapotranspiration varies from 34 to 54% (Caldwell et al. 1977; Smith et al. 1995; Domingo et al. 1999; Reynolds et al. 2000).
13.6.3
Calculation of Evapotranspiration and its Ecological Implications
Valuable data for the evaluation of our model results are also provided in the study of rainwater percolation in the Nizzana dunes by Yair et al. (1997). They found that on a north-facing dune slope in a year following an average rainy season, as in 1993, water percolation is limited to a depth of approximately 0.6 m whereas, after
198
T. Littmann, M. Veste
a very good rainy season (1992), percolation reached down to 4 m with lateral subsurface flow on the north-facing slope adjacent to our recording station. Water content of the upper layer was around 3% in the drier year, in the lower layer 1%, while it increased in the wetter year to 6 and 2% respectively. These data imply a water storage of 18 mm in the upper root zone (−0.6 m) and of 35 mm in the lower one in a dry year, but of 106 mm in the total rooted soil column in a wet year. Recharge of soil water in the lower root zone occurs only after a good rainy season (Yair et al. 1997). If we calculate (from these data) the time the soil water accumulated within the total root depth of the dominant plants (at least 3.5 m; cf. Veste and Breckle 1996a) over 3 years in-between two good rainy seasons would last, applying the different model outputs (evapotranspiration in Table 13.1 minus approx. 30 mm of dewfall equalling surface evaporation), then the vegetation surrounding the recording station (A. monosperma shrubs) should have been dead after 3 months (Thornthwaite-Holzman), 5 months (Sverdrup), 8.6 months (Eddy correlation technique), and 20 months (zero plane). However, this did not occur. This allows us to estimate the necessary root area over the rooted soil column and the corresponding vegetation densities which would be in equilibrium with the rainfall regime over the 3-year period under consideration, because root competition of the plants will lead to a projected root area which enables water uptake by a plant sufficiently high to endure the rechargeless period between two good rainy seasons, whereas competitively inferior individuals would not survive. The ThornthwaiteHolzman model would imply a necessary projected root area of 12 m2 equivalent to a vegetation density of 8%, the Sverdrup model 7.2 m2 (14%), the Eddy correlation technique 4.2 m2 (24%), and the zero plane model 1.8 m2 (55%). In fact, the observed vegetation density on the north-facing slope at the Nizzana site ranges from 21 to 30% and, in spots on the upper slope, can reach 40% (Chap. 26, this volume). This shows that in an arid environment, the zero plane model should provide the most plausible results, compared to all other approaches evaluated here.
13.7
Conclusions
The various explicit models applied to the study of actual evapotranspiration and dewfall in an arid environment provide a wide range of results. Without comparing such outputs to actual measurements – e.g. by lysimeters, in the case of evapotranspiration – the model-specific results cannot be easily validated. However, as we showed that evapotranspiration is controlled largely by plant transpiration whereas surface evaporation occurs only after rainfall or, secondarily, after dewfall in the study area, the water potential in the soil which can be used for transpiration is limited by annual rainfall totals. At the Nizzana experimental site, the rainfall record since 1991 indicates a period of 3 years between two above-normal rainy seasons, i.e. > 90 mm in 1991/1992 (131 mm) and in 1994/1995 (148 mm), while the period in-between was average or below average (87 mm in 1992/1993 and 50 mm in 1993/1994). Following the very good rainy season of 1994/1995, in 1995/1996 rainfall was very low (38 mm) and was only average in 1997/1998 (78 mm).
13 Evapotranspiration, Transpiration and Dewfall
199
It should be clearly stated that the zero plane model will work only in an arid environment where the diurnal and hourly differences of specific humidity close to the active surface are large, especially when controlled by transpiration. In humid climates, such a diurnal course is regularly obliterated by strong advective mixing, and the model does not provide reasonable results. This was found when applying the model to comparable data from measurements in northern Germany, where dq may be zero over long periods. On the other hand, the model is rather conservative, as advective influences on the near-ground plane are quite limited and are largely removed from the series by the anomaly filter, compared to larger heights or other approaches using gradients. Acknowledgements We would like to thank the Arid Ecosystem Research Center of the Hebrew University of Jerusalem, namely Mr. Simon Berkowicz and Mr. Eyal Sachs for scientific cooperation, and logistic and technical support at all stages of the investigation; Dr. Adrie Jacobs and Mr. Bert Heusinkveld (Agricultural University of Wageningen) for their critical comments on the manuscript. This study was supported by grants of the German Ministry of Education and Research, BMBF (no. 0339635, University of Halle, no. 0339495A, University of Bielefeld).
References Baldocchi DD (1993) Scaling water vapor and carbon dioxide exchange from leaves to a canopy: rules and tools. In: Ehleringer JR, Field CB (eds) Scaling physiological processes leaf to globe. Academic Press, San Diego, CA, pp 77–116 Beysens D (1995) The formation of dew. Atmospheric Res 39:215–237 Domingo F, Villagracia L, Brenner AJ, Puigdefábregas J (1999), Evapotranspiration model for semi-arid shrub-lands test against data from SE Spain. Agric Forest Meteorol 95:67–84 Doorenbos J, Pruitt WO (1977) Guidelines for predicting crop water requirements. FAO Irrigation and Drainage Pap no 24 (revised), Rome Evenari M, Shanan L, Tadmor A (1982) The Negev – The challenge of a desert. Harvard University Press, Cambridge, MA Frankenberger E (1955) Über Strahlung und Verdunstung. Ann Meteorol 6:5–13 Garrat JR, Segal M (1988) On the contribution of dew formation. Boundary Layer Meteorol 45:209–236 Häckel H (1990) Meteorologie. Ulmer, Stuttgart Henning I, Henning D (1984) Die klimatologische Wasserbilanz der Kontinente. Münstersche Geographische Arbeiten 19, Münster Hicks BB (1983) A study of dewfall in an arid region: an analysis of Wangara data. Q J R Meteorol Soc 109:900–904 Jacobs AFG, Heusinkveld BG, Berkowicz S (1998) Dew deposition in the Negev Desert: the biological crust. In: Proc 1st Int Conf Fog and Fog Collection, Vancouver, pp 261–264 Jacobs AFG, Heusinkveld BG, Berkowicz S (1999) Dew deposition and drying in a desert system: a simple simulation model. J Arid Environ 42:211–222 Janssen LHJM, Römer FG (1991) The frequency and duration of dew occurrence over a year. Tellus 43B:408–419 Jarvis PG, McNaughton KG (1986) Stomatal control of transpiration: scaling up from leaf to region. Adv Ecol Res 15:1–48 Littmann T (1994) Immissionsbelastung durch Schwebstaub und Spurenstoffe im ländlichen Raum Nordwestdeutschlands. Bochumer Geographische Arbeiten 59, Bochum Littmann T (1997) Atmospheric input of dust and nitrogen into the Nizzana sand dune ecosystem, northwestern Negev, Israel. J Arid Environ 36:433–457
200
T. Littmann, M. Veste
Midgley GF, Veste M, von Willert DJ, Davis GW, Steinberg M, Powrie LW (1997) Comparative field performance of three different gas exchange systems. Bothalia 27(1):83–89 Monteith JL (1957) Dew. Q J R Meteorol Soc 83:322–341 Monteith JL (1965) Evaporation and the environments. Proc Soc Exp Biol 19:205–234 Niu GY, Sun SF, Hong ZX (1997) Water and heat transport in the desert soil and atmospheric boundary layer in western China. Boundary Layer Meteorol 85:179–195 Ohlmeyer P, Hoyningen-Huene Jv (1975) Die Probleme bei der Diagnose des Wasserverbrauchs eines Pflanzenbestandes, dargestellt am Beispiel der extrem ariden Klimabedingungen der Oase Al Hassa/Saudi Arabien. Mitt Leichtweiß-Institut für Wasserbau TU Braunschweig 46:1–117 Paw UKT, Qiu J, Su HB, Watanabe T, Brunet Y (1995) Surface renewal analysis: a new method to obtain scalar fluxes without velocity data. Agric Forest Meteorol 74:119–137 Peixoto JP (1973) Atmospheric vapor flux computations for hydrological purposes. WMO contribution to the International Hydrological Decade (IHD) 20, Geneva Penman HL (1948) Evaporation in nature. Rep Progr Phys 11:366–388 Reynolds JF, Kemp PR, Tenhunen JD (2000) Effects of long-term rainfall variability on evapotranspiration and soil water distribution in the Chihuahuan Desert: a modeling analysis. Plant Ecol 150:145–159 Roedel W (1992) Physik unserer Umwelt, Die Atmosphäre. Springer, Berlin Heidelberg New York Schrödter H (1985) Verdunstung. Springer, Berlin Heidelberg New York Smith SD, Herr CA, Leary KL, Piorkowski J (1995) Soil-plant water relations in a Mojave Desert mixed shrub community: a comparison of three geomorphic surfaces. J Arid Environ 29:339–351 Sverdrup HU (1936) Das maritime Verdunstungsproblem. Ann Hydrogr Maritim Meteorol 32:41–47 Swinbank WC (1955) An experimental study of eddy transports in the lower atmosphere. CSIRO, Sydney, Tech Pap 2 Thornthwaite N, Holzman B (1942) Measurement of evaporation from land and water surfaces. USDA Tech Bull no 817 Veste M, Breckle S-W (1996a) Root growth and water uptake in a desert sand dune ecosystem. Acta Phytogeogr Suec 81:59–64 Veste M, Breckle S-W (1996b) Gaswechsel und Wasserpotential von Thymelea hirsuta in verschiedenen Habitaten in der Negev-Wüste. Verhandl Gesell Ökol 25:97–103 Veste M, Littmann T, Friedrich H, Breckle S-W (2001) Microclimatic boundary conditions for activity of soil lichens crusts in sand dunes of the north-western Negev Desert, Israel. Flora 196(6):465–476 Veste M, Eggert K, Breckle S-W, Littmann T (2005) Vegetationsänderungen entlang eines geoökologischen Gradienten im Sinai-Negev-Sandfeld (nordwestlicher Negev, Israel). In: Veste M, Wissel C (Hrsg) Beiträge zur Vegetationsökologie der Trockengebiete und Desertifikation. UFZ Bericht 1/2005, Leipzig, pp 65–81 von Willert DJ, Mattysek R, Herppich WB (1995) Experimentelle Ökologie. Thieme, Stuttgart Yair A, Lavee H, Greitser N (1997) Spatial and temporal variability of water percolation and movement in a system of longitudinal dunes, western Negev, Israel. Hydrol Processes 11:43–58 Zangvil A (1996) Six years of dew observations in the Negev Desert, Israel. J Arid Environ 32:361–371 Zohary M, Orshan O (1954) The Zygophylletum dumosi and its hydroecology in the Negev of Israel. Vegetatio 5/6:341–350
Chapter 14
Morphological Changes at Active Dune Crests A. Allgaier
14.1
Introduction
The active parts of the linear dunes at Nizzana are known to undergo periodic changes due to the seasonal change of the major sand-moving wind direction. In addition to the seasonal changes, an ongoing reduction of vegetation free surfaces at the dune crests has been identified on aerial photographs taken in the period 1982 up to 1999. This development is likely to be caused by the change in land use at the site following the peace treaty between Egypt and Israel in 1982 (Meir and Tsoar 1996; Chap. 6, this volume). The opening of the border in the aftermath of the war of June 1967 had led to a rapid decrease of vegetation (Meir and Tsoar 1996) and an increase of dune height (Tsoar and Møller 1986). In order to quantify both the seasonal and permanent changes, an active dune part has been monitored in detail during 33 months, from February 1997 until October 1999.
14.2
The Research Site
The location for the experimental setup within the research site of Nizzana has been selected based on aerial photography and field investigation (Fig. 14.1). The chosen dune ridge extends from the border road eastwards up to Nahal Nizzana. Its condition changes from the border road towards the east. Vegetation cover becomes sparser while the dune height and the area of open sand surfaces increase. The cross section of the ridge is asymmetrical. Its steep, north-facing slope extends from the crest to the adjacent interdune corridor. The angle of the slope is close to the angle of repose of dry sand. The south-facing slope is not as high because the interdune corridor slopes gently towards the dune ridge, meeting it at a sharp knick point. From here, the slope of the crest area continues at steep angles. The lower slopes on both sides are covered by a microbial surface crust and vascular vegetation. Open sand surfaces are not found in these parts of the ridge. The crest area is separated from the foot slopes by a sharp change of slope. Apart from active sand accumulations, the crest area is flat, with interspersed nebkhas.
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
201
202
A. Allgaier
study site 3181-b
90.000
90.500
91.000
91.500
92.000
92.500
Fig. 14.1 Location of the research site within the dunefield. Point 3181-b is located immediately south of the entrance from the road to site N1
The monitored area is a local termination of the ridge, eastwards of which the dune continues at a lower height and increased vegetation cover. Vegetation cover of the site is very low, the plants being concentrated in isolated spots. Ripple marks, blowouts and sand accumulations with slip faces were signs of ongoing aeolian activity.
14.3
Methods
An area of 90×30 m of the most active part has been equipped with erosion pins, set in a rectangular grid at a distance of 3 m from each other. In total, 310 pins have been staked out in 11 rows of up to 33 pins roughly parallel to the ridge. The pins were 900 mm long and 2 mm in diameter, their initial length above the surface being set to 400 mm. Based on fixed reference points in the interdune corridor, three-dimensional positions of the pins were determined using a theodolite (EDM). The coordinates of the reference points were determined with a differential GPS based on the known coordinates of the trigonometric point 3181-b situated near the border road. To determine surface changes, the length of the pins above the surface was determined with a tape measure at least once a week during winter. After storms, additional measurements were taken as soon as possible to determine the change caused by single events. A total of 91 datasets have been acquired (20 measurements in the period from 22 February 1997 to 7 June 1997, 40 measurements from 7 November 1997 to 10 May 1998 and 31 measurements from 24 November 1998 to 13 October 1999). Similar techniques have been used successfully to evaluate dune mobility on a large Namib dune (Livingstone 1989), on transverse coastal dunes (Burkinshaw and Rust 1993) and on vegetated linear ridges in the Kalahari (Wiggs et al. 1995). The absolute height of the surface at the pin was calculated through continuous addition of the differences of the length of the pins based on the EDM survey. The calculated heights were used to interpolate surfaces and to generate maps of zones of erosion and deposition. The grid width used for the interpolation is 1 m. To compute the volume changes of the dune ridge, the calculated surfaces have been subtracted from each other. Wind speed and direction measurements conducted at the monitored site showed high-magnitude/low-frequency SW storms during winter and low-magnitude/highfrequency NW winds during summer to be mainly responsible for sand movement (Chap. 6, this volume). The SW winds are connected with low-pressure zones passing
14 Morphological Changes at Active Dune Crests
203
over the Eastern Mediterranean, and often with rainfall (Chap. 4, this volume). The NW winds are caused by the temperature difference between the water body of the Mediterranean and the land surface of the Negev during daytime in summer. This diurnal sea breeze reaches the threshold velocity for aeolian sand movement at the dune crests in the afternoons between early spring and late autumn (Chap. 3, this volume). A more detailed description of the climatic conditions of Nizzana is given in Chapters 3 and 4 (this volume).
14.4
Results
14.4.1
Monitored Changes 1997 to 1999
14.4.1.1
Seasonal Changes
Aeolian activity in summer is dominated by the diurnal NW sea breeze. Sand transport occurs regularly at the uppermost part of the dune crest. The area affected depends on the strength of the sea breeze. During calm weather in winter, the sea breeze does not reach threshold velocity for sand transport, as the heating of the desert surface is not sufficient. The dominant sand-transporting winds in winter are connected to episodic storms. To associate changes of morphology based on erosion pin measurements with the wind regime, seasons were determined based on wind recordings. The beginning of summer is thus marked by the last recorded SW storm in spring, while the first SW storm in autumn determines the beginning of winter. The zones of erosion and deposition are shown for six seasons in Figs. 14.2 and 14.3. The first season is not complete, as the measurements were started in late February, when several SW storms had already passed the site. The measurements were also stopped before the end of summer 1999, and therefore this season is also included only in part. The changes observed display a regular pattern but differing in intensity. The crest as a whole is moved towards the north in winter and towards the south during summer. Erosion is most intense on the upper windward slopes and in the blowout between the nebkhas, which are located mainly on the northern part of the crest. That is because the SW winter winds are of high magnitude and prevent shrubs from growing where erosion is very strong. The winter storms affect lower areas of the crest, while the sea breeze is restricted to the uppermost parts. Deposition occurs immediately downwind of the crest line or in the lee of plant patches. Early winter storms erode sand at the steep south-facing slip faces built up during summer (Fig. 14.4). Due to the specific loose layering of the deposited sand, erosion is eased, compensating partly for the steep gradient against which the sand particles have to be transported. This northward transport leads to a build up of a sharp crest line, as shown in Fig. 14.5.
204
A. Allgaier 90780
90790
90800
90810
90820
90830
90840
90850
90860
90870
22.02.-18.04.1997 38940
38930
38920
18.04.-17.11.1997 38940
38930
38920
17.11.1997-19.04.1998 38940
38930
38920 90780
90790
90800
90820
90830
90840
90850
90860
90870
+/0.0
+/-0.2
+/-0.4
+/-0.6
+/-0.8
+/-1.0
+/-1.2
+/-1.4
+/-1.6
+/-1.8
+/-2.0
+/-2.2
deposition
90810
Fig. 14.2 Seasonal changes of surface level (in metres) between 22 February 1997 and 19 April 1998. The areas with ongoing deposition are additionally marked (hatched)
Beginning in spring, the sea breeze intensifies and the lack of SW storms leads to a reversal of the dune crest. Figure 14.6 shows the situation in late spring, when south-facing slip faces are already well developed. In the final stage, the former slip faces have transformed into rounded deposition areas, as shown in Fig. 14.4. At the end of summer, the surface is adjusted to a steady state (Tsoar 1985) with regard to the NW winds of the sea breeze.
14.4.1.2
Total Changes
The volumes of the total and seasonal changes are shown in Fig. 14.7. The calculation of the changes during the measurement period is based on volumes calculated with the pin-coordinates of 20 February 1997 and 13 October 1999 (965 days). Between these two dates, a loss of 1,803 m3 was observed over a surface area of 2,668 m2.
14 Morphological Changes at Active Dune Crests 90780
90790
90800
90810
205
90820
90830
90840
90850
90860
90870
19.04.-14.12.1998 38940
38930
38920
14.12.-19.02.1999 38940
38930
38920
19.02.-13.10.1999 38940
38930
38920
+/0.0
+/-0.2
+/-0.4
+/-0.6
+/-0.8
+/-1.0
+/-1.2
+/-1.4
+/-1.6
+/-1.8
+/-2.0
+/-2.2
deposition
Fig. 14.3 Seasonal changes of surface level (in metres) between 19 April 1998 and 13 October 1999. The areas with ongoing deposition are additionally marked (hatched)
Fig. 14.4 Situation at the end of summer 1998 (23 November 1998), view westwards. Southfacing slip faces caused by northerly winds
206
A. Allgaier
The total loss per unit area was 0.67 m3/m2. This corresponds to an average loss of 0.26 m3/m2 per year, or a reduction of average dune height of 0.26 m/year. Average elevation of the monitored area was reduced from 207.37 m asl on 20 February 1997 to 206.67 m asl on 13 October 1999 (Fig. 14.8). Main losses occurred during summer when northerly winds prevailed. During the winters of 1997/1998 and 1998/1999, the sediment budget had been balanced. The negative linear trend for the total volume of the monitored dune crest area can be written as V = −1.87 * d where V is the volume in m3, and d the number of days. Correlation is strong (r2 = 0.97).
Fig. 14.5 Test site during winter, view westwards, 12 January 1998. A sharp crest line with north-facing slip faces has been established by southerly winds
Fig. 14.6 Part of the monitored area at the beginning of summer 1999 (5 May 1999). South-facing slip faces indicate transport by northerly sea breeze (cf. Fig. 14.4)
14 Morphological Changes at Active Dune Crests
0
207
-93m³ −410m³
−500
−434m³ summer 1998
[m³]
summer 1997
−1000
?
Y = -1.87 * X data points: 6 r² = 0.98
−1500
summer 1999
−1478m³
−1465m³ −1803m³
−2000 20-Feb-97
20-Feb-98
20-Feb-99
Fig. 14.7 Volume balance of the monitored area between February 1997 and October 1999 (abscissa: months). The different lengths of season can be explained by the wind regime. In summer, only local winds prevail. Winter is defined as the season from the first storm in fall until the last storm in spring. The precise start of the winter season in October 1999 is uncertain
38940
38930
38920 90800
90830
90840
90850
90860
90870
+/0.0
+/−0.2
+/−0.4
+/−0.6
+/−0.8
+/−1.0
+/−1.2
+/−1.4
+/−1.6
90820
+/−2.0
deposition
90810
+/−1.8
90790
+/−2.2
90780
Fig. 14.8 Total changes of surface level (in metres) between 22 February 1997 and 13 October 1999
14.4.2
Observed Changes 1993–1999
To verify the recognized trend of dune volume reduction, ground photographs taken from a similar perspective and at the same time of the year in 1993 and 1999 (Fig. 14.9) have been compared, focusing on changes. The main changes since 1993 were an increase of vegetation cover at the dune crest and a reduction of dune height. The crest of the dune monitored in this study was, in 1993,
208
A. Allgaier
Fig. 14.9 Development of the dune ridge monitored between 1993 and 1999. Note the increase of vegetation cover and the decrease of dune height in the lower photograph. Dashed lines Extent of the monitored area, N3 a distinct Nebkha located on the northern part of the crest
a continuous sand body with few nebkhas on it. Hummocks topped by vegetation, protruding from the main sand body, are visible at the dune crest only in the western part of the 1993 photograph. The other vegetation appears to be only slightly elevated above the surrounding sand surface. In 1999, this area appears dissected by large and deep blowout areas which can be recognized between the hummocks. The former continuous crest line with isolated hummocks was replaced by a chaotic conglomeration of hummocks. The significant reduction of dune height is most clearly visible in the eastern part. In this area, hummocks stand out clearly in 1999, while in 1993 only small vegetation patches integrated into the main sand body are visible. In 1999, the monitored site resembles the western area in the photograph of 1993. In contrast to the crest area, the foot slopes do not appear to have undergone any significant change during this period. The aerial photograph in Fig. 14.10 shows the same part of the dune ridge as do the ground photographs in Fig. 14.9. It was taken in August 1998, i.e. during the second half of summer. Parts of the dune crests show signs of activity: South of vegetation patches, accumulation areas protrude. The monitored part of the ridge has the lowest visible vegetation cover, resembling the status of the whole ridge crest in 1989. The interpretation of the aerial photograph was confirmed through ground checks. Transect walks along the dune ridges revealed that open areas with ripple marks and slip faces, i.e. signs of aeolian activity, increased towards the east as relative dune height increased. This longitudinal development repeated itself east of the test site. Three more areas similar to the test site were identified on the aerial photograph and in the field up to termination of the ridge at Nahal Nizzana.
14 Morphological Changes at Active Dune Crests
209
Fig. 14.10 Aerial view of the research site, August 1998 (same area as in Fig. 14.9)
14.5
Conclusions
The direct monitoring of an active dune crest area has shown that, 15 years after the last change of land use within the dune area, adjustment to the new conditions is still ongoing. In addition to regular seasonal changes caused by the regional bidirectional wind regime, a clear linear trend shows the continuous reduction of dune height. This is a reversal of the development observed by Tsoar and Møller (1986) in their study on the effect of intensive grazing at Nizzana. Acknowledgements The study was carried out as part of DISUM 042, a joint project of Ben-Gurion University of the Negev, Beer-Sheva and Universität Trier. I wish to thank the Arid Ecosystems Research Centre (AERC) at the Hebrew University of Jerusalem for logistic support during the study.
References Burkinshaw JR, Rust IC (1993) Aeolian dynamics on the windward slope of a reversing dune. Alexandria coastal dune field, South Africa. In: Pye K, Lancaster N (eds) Aeolian sediments. Ancient and modern. International Association of Sedimentologists Special Publication vol 16. Blackwell, Oxford, pp 13–21 Livingstone I (1989) Monitoring surface change on a Namib linear dune. Earth Surface Processes Landforms 14:317–332 Meir A, Tsoar H (1996) International borders and range ecology: the case of Bedouin transborder grazing. Human Ecol 24:39–64 Tsoar H (1985) Profiles analysis of sand dunes and their steady state signification. Geografiska Annaler A 67:47–59 Tsoar H, Møller JT (1986) The role of vegetation in the formation of linear sand dunes. In: Nickling WG (ed) Aeolian geomorphology. Allen & Unwin, Boston, MA, pp 75–95 Wiggs GFS, Thomas DSG, Bullard JE, Livingstone I (1995) Dune mobility and vegetation cover in the southwestern Kalahari Desert. Earth Surface Processes Landforms 20:515–529
Chapter 15
Aeolian Sand Transport and Vegetation Cover A. Allgaier
15.1
Introduction
The Nizzana area has experienced periods of heavy grazing which resulted in the destruction of a cryptogamic soil crust and vegetation cover which, in turn, led to increased aeolian sand movement (Tsoar and Møller 1986). The influence of cryptogamic soil crusts on deflation has been the focus of several studies mainly in drylands used or suitable for grazing (Harper and Marble 1988; West 1990; Eldridge and Greene 1994; Belnap 1995; Belnap and Gillette 1998; Leys and Eldridge 1998). The resistance of such crusts against wind erosion has been tested in wind tunnel studies (McKenna Neuman et al. 1996; McKenna Neuman and Maxwell 1999, 2002). The results of these studies show that intact crusts inhibit deflation of sandy soils, while a destruction (e.g. trampling by grazing animals) leads to a sharp increase of aeolian transport rates (Leys and Eldridge 1998; Allgaier 2005). Aeolian sand transport rates over crust-covered areas at the Nizzana site have been found to be negligible, and experimental studies showed that current wind energy at the site is not sufficient to destroy the crust (Allgaier 2005). Where no soil crust is present, vascular vegetation offers an increasing protection of the surface against wind action. It increases the aerodynamic roughness of a surface, thereby extracting energy from the airflow and reducing shear stress at the soil surface. Various studies have investigated the influence of vegetation on aeolian sand movement. Stockton and Gillette (1990) and Musick and Gillette (1990) have dealt with the relationship between plant cover and erodibility of surfaces, concentrating on the influence of vegetation density on the partition of shear stress between the vegetation and soil surface. Lancaster and Baas (1998) undertook field studies over areas with different vegetation cover. Wiggs (1993) reports a possible threshold cover of 14% for certain areas of the Kalahari Desert, above which no noteworthy aeolian sand transport occurs. However, he does not claim this to be a universal value. The value is consistent with the findings of Marshall (1973), who concluded that wind erosion will increase rapidly if vegetation cover decreases below 15% on level alluvial sand surfaces. However, Ash and Wasson (1983, p. 20) report sand movement on dune crests with a ground cover of up to 35%.
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
211
212
A. Allgaier
Wolfe and Nickling (1993) focused on the general effect of sparse vegetation within an environment prone to wind erosion. They distinguished three main effects: (1) cover of the surface, (2) momentum extraction from the airflow and (3) trapping of soil particles already in motion. In their work, they stress that dead vegetation also plays an important role in the protection of a surface against wind erosion, a point also made by others (Ash and Wasson 1983; Wiggs et al. 1995). The influence of a single obstacle and of different roughness element concentrations has been discussed by Ash and Wasson (1983) and Wolfe and Nickling (1993, p. 56f). An isolated obstacle may lead to increased velocities – and thereby to increased erosion – as the flow is streamlined around it (Ash and Wasson 1983; Thomas and Tsoar 1990; Wolfe and Nickling 1993). This effect has been observed for vegetation cover of up to 25% when plant diameter is approximately equal to height. The effect is most pronounced for a cover of 5–10% (Ash and Wasson 1983, p. 19). The same obstacle may act as an accretion focus for moving sand on its leeside, due to the reduced wind velocity in that area (Hesp 1981). Deposition of material might also occur on the windward side, depending on the size of the plant and its porosity to the wind (Thomas and Tsoar 1990). The influence of single obstacles is only local, resulting in only small coppice dunes of nebkas. However, increasing sand supply might lead to the development of ‘real’ dunes from such focus points, reducing the influence of the plants as the dune body grows (Thomas and Tsoar 1990). Wolfe and Nickling (1993, p. 56f) state that a surface cover of 40% will lead to ‘skimming flow’, which means that the surface is not influenced by the flow above the roughness elements, i.e. vegetation. Lee (1991) also notes that high densities of roughness elements lead to a smooth aerodynamic surface above the elements. Apart from the percentage of surface cover by vegetation, its nature and distribution, and the actual wind regime in the area under observation have to be considered (Thomas and Tsoar 1990, p. 478). While the majority of geomorphological studies have focused on the effect of a given vegetation cover on the processes discussed above, these processes do have an effect on the plants. Several studies at the research site have dealt with this aspect (Kadmon 1994; Kadmon and Leschner 1995; Tielbörger and Kadmon 1995; Tielbörger 1997; Prasse 1999). Vulnerability is highest in the early stage of the life of a plant. After germination has taken place, the survival of the seedlings depends to a large extent on the mobility of the surface, because shifting sand can cover vegetation, deflation may expose roots, and saltating grains can damage the plants (Kadmon 1994; Kadmon and Leschner 1995; Tielbörger 1997).
15.2
Study Sites and Methods
This study focuses on the influence of properties of vascular vegetation on aeolian sand transport in the interdune corridors of the Nizzana research site. To conduct the study, two test sites were set up which represent different possible states of
15 Aeolian Sand Transport and Vegetation Cover
213
surface condition. Both sites were set up in the centre of a corridor to minimise the influence of adjacent dune ridges. Site A resembled the condition as found prior to 1982, when the surface crust had been destroyed by the trampling of grazing animals and the edible vegetation had been consumed. Aeolian sand transport is not hindered by surface obstacles or cementing, and therefore available wind energy is the restricting factor. Site B represented the interdune surface condition after 15 years of undisturbed plant development, but without the surface crust, as this has a major influence on sand transport. In a later stage, vegetation density at site B was reduced to half the natural value, in an attempt to simulate degradation of vegetation. Wind speed was measured at each site using Porton A100 cup anemometers. Three instruments were installed at 0.24, 0.65 and 1.77 m above ground. This spacing enables the direct estimation of shear velocity u* (Bagnold 1941, p. 51), and provided information about the influence of roughness elements on near-surface wind speed. Wind direction was determined 2.3 m above the ground. A data logger was used to record and store data at 2-minute intervals. Acoustic saltation sensors (saltiphones, supplied by Eijkelkamp; for technical details, see Spaan and van den Abeele 1991) were used to detect sand movement at 4.5 cm height. Saltation data were logged in conjunction with wind data. Sand traps of the type described by Leatherman (1978) and Rosen (1979) were used to determine the amount of sediment moved by wind. The material caught was checked on a weekly basis or following major storm events. The actual sand transport was calculated by using saltation data, and information on wind direction and the amount of sand caught in the sand traps (for a detailed description of the procedure, see Allgaier 2005). The sites were situated in an interdune corridor 90–120 m wide, bordered to the north by a linear dune ridge (height 10 m) with a steep south-facing slope, and to the south by a subdued dune ridge (height 6 m). The surface, if undisturbed, is covered by a microphytic crust with an average thickness of approximately 1 mm, on which thin layers of loose sand appear in patches. The vegetation has been described as Echiochilon fruticosum-Thymelaea hirsuta community (Tielbörger 1997). Annual vegetation is concentrated under the canopy of perennials or in their immediate vicinity (Chap. 8, this volume). Patches of annuals appear mainly in areas where the crust has been disturbed previously (Prasse 1999).
15.2.1
Site A: No Vegetation, No Surface Crust
All vegetation and the microphytic crust were removed from the surface at site A. The cleared area was of rectangular shape, with a total area of 1,027 m2; its long axis oriented along the expected wind directions of the major winter storms (SW); the layout is shown in Fig. 15.1. Six sand traps facing SW were deployed: four (A/1–A/4) were located in the central part of the rectangle, close to the anemometer mast and the saltiphone. One trap (A/5) was placed at the upwind end of the test area, to
214
A. Allgaier
A/1 220° A/2 240° A/5 240°
saltiphon
A/6 240°
mast A/3 240° A/4 260°
0m
5m
10m
15m
20m
Fig. 15.1 Setup of site A
record material coming into the area during westerly storms, and one trap (A/6) was set up at the downwind end to measure material transported out of the site.
15.2.2
Site B: Natural and Reduced Vegetation Cover, Without Microphytic Surface Crust
Site B was used to determine sand movement on surfaces with various degrees of vegetative cover. Four sand traps were installed to face the high-magnitude winter winds from the western sector. The surface crust was destroyed within a radius of 25 m from the anemometer mast, by raking the surface. Vegetation patch positions and sizes at site B (Fig. 15.2) were determined in January 1998, using a total station. The plants removed in January 1999 were chosen randomly from the list created during the survey. Initial vegetation cover was determined to be 17.6%, mean vegetation height was 0.37 m and the vertical projected area A′ was 59 m2. While values of A′ are usually determined as the product of the average element height and diameter, assuming a cylindrical shape, observation of the plant patches of site B led to the conclusion that the use of a rounded shape would better represent their actual appearance. Therefore, A′ has been determined as
(
A′ = sum (p * r 2 ) / 2
)
where r is the mean of patch height and patch radius. Based on the above information, the roughness element concentration Lc can be calculated. It is defined by Wolfe and Nickling (1993, p. 57) as
15 Aeolian Sand Transport and Vegetation Cover
215
Lc = n * A′ / SB where SB is the total surface area of the site. At site B, Lc of natural vegetation was 0.096 before and 0.052 after the removal of half of the vegetation patches. Based on published values for flow regimes and associated roughness element concentrations (Wolfe and Nickling 1993, p. 57), the undisturbed vegetation at site B should result in wake interference flow, where the wake zones overlap, thereby reducing the area prone to be influenced by the airflow. After the removal of selected plants, site B could be interpreted as belonging to the isolated-roughness-flow class, where each vegetation patch develops its own ‘wake and separation region’ (Wolfe and Nickling 1993).
15.3
Results
By comparing the results for aeolian sand transport at sites A and B, the amount of protection offered by vegetation properties could be determined. Sand transport in the interdune corridors was restricted to dune-parallel winds for the majority of storm events (Allgaier 2005).
15.3.1
Sand Transport at Site A
As the distance separating the traps from the fringes of the cleared area differed, their position within the plot (Fig. 15.1) had a significant influence on the amount of sand transport recorded. The amount of material transported increased from the edges of the cleared area towards the centre and towards the downwind (NE) end. However, traps A/3 in the middle and A/6 at the downwind end of the plot consistently recorded similar amounts, which indicates that the upwind fetch of A/3 was long enough to enable saturation of the air stream during saltation events. Thus, the results of traps A/3 and A/6 are regarded as representative for non-vegetated surfaces. Trap A/5, at the western (upwind) end of the plot, collected only isolated grains during all detected transport events. This showed that no upwind source outside the cleared area contributed significantly to the amount of sand transport measured within the plot.
15.3.2
Sand Transport at Site B
The differences between the traps at site B are more pronounced than at site A. The reason is seen in the position of the traps relative to vegetation patches, as their
216
A. Allgaier
saltiphon
B/1
mast
B/2 B/3 B/4
0
2
4 6 metre
8 10
Fig. 15.2 Setup of site B. Shaded circles show the planar area of the vegetation patches. Plants marked with x were removed on 23 January 1999
distribution (Fig. 15.2) plays an important role for small-scale variations in sand flux. As the scope of the study did not enable a detailed assessment of flow conditions within the vegetation layer, the mean value calculated from the results of all four traps at site B is regarded as the representative sand flux at vegetated interdune corridor sites.
15.3.3
Influence of Vascular Vegetation on Sand Movement
Various recent studies have shown that aeolian sand transport and near-surface horizontal wind speed are closely correlated (McKenna Neuman et al. 1997; Sterk et al. 1998; Walker 1999). This has been verified in this study by relating u0.24 during transport events to sand flux q at sites A and B (Fig. 15.3). The results show that, if wind speed is measured within the vegetation canopy, the relation is independent of vegetation density.
15.3.3.1
Wind Speed
The reduction of sand transport over vegetated surfaces is caused by the extraction of momentum from the airflow. To quantify this effect, the wind speed at different
15 Aeolian Sand Transport and Vegetation Cover
217
100
A, no vegetation B2, Lc 0.054
q [g m−1s−1]
10
B2, Lc 0.096 1 q = 8.6 * 10−5 u0.247.07 R2 = 0.75
0.1
0.01 1
u0,24 [m s−1]
10
Fig. 15.3 Relation of measured mean sediment flux q and mean wind speed at 0.24 m height (u0.24) during saltation events at sites A and B
heights over the non-vegetated surface of site A was related to the corresponding values of site B. The results are shown in Fig. 15.4 for different vegetation densities during saltation. Mean speed has been calculated at each site for the actual period during which saltation was recorded. Above the vegetation canopy, at heights of 1.18 and 1.77 m above ground, the relation between sites A and B is close to 1, independent of vegetation cover. Within the vegetation canopy at 0.24 m above the surface, the relation is highly dependent on vegetation cover. Natural vegetation cover of 17% (Lc 0.096) reduced average speed to 75% of wind speed over non-vegetated surfaces. When average u0.24 at site A was below 5 m s−1, no sand movement was recorded at the vegetated site B. After the vegetation cover at site B had been reduced to 9% (Lc 0.052), average u0.24 of 3 m s−1 at site A was sufficient to cause saltation at site B. Wind speed within the vegetation canopy was 84% of the velocity measured at site A.
15.3.3.2
Transported Mass
As flux q is dependent on horizontal speed u, lower wind speeds lead to lower flux and, thus, to lower total transport. Figure 15.5 shows the effect of different
218
A. Allgaier
cover 17% Lc 0.096
9
u0.24 uB = 0.75 * uA r2 = 0.99 u1.18 uB = 1.02 * uA r2 = 0.99 u 1.77 uB = 1.05 * uA r2 = 0.99
8
uB [m s−1]
7 6 5
cover 9% Lc 0.052 u0.24 uB = 0.84 * uA r2 = 0.98 u1.18 uB = 1.02 * uA r2 = 0.98 u1.77 uB = 1.07 * uA r2 = 0.99
4
u0.24 17%
3
u0.24 9% 2 2
3
4
5
6
uA [m s−1]
7
8
9
Fig. 15.4 Relation of average wind speed at various heights for sites A and B. Vegetation cover at site B was 9 and 17%
10000
17% cover mB = 0.010 * mA r² 0.80 17% cover, high magnitude storm mB = 0.085 * mA r2 = 0.96
mass site B [g cm−1]
1000
100
9% cover mB = 0.077 * mA r2 = 0.43
10
1
0.1
0.01 0.01
0.1
1 10 100 mass site A [g cm−1]
Fig. 15.5 Relation of transported mass at sites A and B
1000
10000
15 Aeolian Sand Transport and Vegetation Cover
219
vegetation cover on the transported mass per unit width during selected events. With 17% cover, transported mass at site B is about 1.0% of the transported mass at site A during average winter storms. The value is significantly higher during exceptional events (15–17 March 1998, 19 March 1998, 16 April 1998; see Allgaier 2005, p. 62) when transported mass at B was 8.5% of that at site A. The reduction of vegetation cover to 9% led to an increase in the average transported mass at site B, to 7.7% of the transported mass at site A. As no high-magnitude events such as those of March 1998 were recorded after the reduction of vegetation cover, no measured flux data of such events under reduced cover are available. Assuming the same relation for flux between average storms and exceptional storms as with 17% cover, transport at site B would reach above 65% of site A values with a reduced cover of 9%. In general, events at site A were longer than at site B. The increase of transported material at site B after reduction of vegetation cover was partly caused by a change in the duration of each event. The decreased vegetation cover led to an increase of the duration of saltation events at site B, compared to the duration of events at site A (Fig. 15.6). The ratio during the exceptional events in March 1998, under 17% cover, is similar to the values for average cyclonic storms under reduced vegetation cover. Using the results of site A (0% cover, 100% transport) and site B during ‘normal’ cyclonic storms, the effect of vegetation cover on sand transport in the interdune corridor is well described by an exponential function, as shown in Fig. 15.7. This shows that the current natural vegetation cover of 17% reduces transported mass to less than 1%, compared to open surfaces. This is in accordance with the results of Lancaster and Baas (1998). Although absolute values differ, most likely caused by different sediment and vegetation properties, the general trend is similar.
duration site B [min]
600 480 360 240 120
17% 9%
0 0
120
240 360 480 600 duration site A [min]
720
840
Fig. 15.6 Duration of saltation events at sites A and B for different vegetation cover at site B
220
A. Allgaier
normalised flux
1 0.1
90% reduction
0.01
ln(Y) = -0.27 * X − 0.04 r2 = 0.99 99% reduction
0.001 0.0001 0
5
10 15 20 vegetation cover [%]
25
30
Fig. 15.7 Relation between plant cover and sand transport at Nizzana
15.3.3.3
Deflation
The average deflation at both sites was calculated using the volume trapped by the sand traps. For site A, the amount was calculated using the results of traps A/3 and A/6. Based on the length of the cleared rectangle of 47 m, and the measured volume of approximately 13.400 cm3 cm−1 at trap A/6 during a period of 521 days, an average deflation of 1.96 cm year−1 was determined for non-vegetated, non-crusted interdune corridor surfaces. Considering that trap A/3, with a smaller fetch of approximately 32 m, yielded essentially the same volume as trap A/6, the estimated average deflation rate is 2.86 cm year−1. These values include the loss during high-magnitude events in March 1998, which were responsible for approximately 60% of the transported material at site A. Corresponding values for site B have been calculated by relating mean values of site B to the average of traps A/3 and A/6. Average transported mass during average storms at site B prior to 23 January 1999 was 0.9% of that at traps A3 and A6. Thus, a deflation rate of 0.025 cm year−1 can be expected for 17% vegetation cover. When including the high-magnitude events of spring 1998 into the calculation, deflation at site B rises to 0.18 cm year−1, i.e. 6.2% of that recorded at site A. As these storms were isolated and exceptional events, however, their influence is most likely overestimated. The reduction of vegetation density to 9% cover increased average transport at site B to 12.2% of that at site A, resulting in a deflation rate of 0.35 cm year−1.
15.4
Discussion
Aeolian sand transport requires open sand surfaces as a source of transportable material. Any vegetation cover reduces the open area and, thus, the amount of available material. Vascular vegetation also reduces near-surface wind speed, thereby leading to a reduction of sand transport rates and total transported mass.
15 Aeolian Sand Transport and Vegetation Cover
221
In the interdune corridors at Nizzana, a natural vegetation cover of 17% at site B has been shown to be insufficient to completely inhibit sand transport if the underlying surface crust is removed. Considering only the transported mass per unit width, the presence of vascular vegetation at site B led to a reduction to 6.9% of bare surface values. A more detailed view revealed that two exceptional events were responsible for the bulk (94% at site B) of the transported mass between November 1997 and January 1999. During average winter storms, the vegetation at site B reduced transported mass to about 1.0%, compared to the non-vegetated interdune surface of site A. The storms of exceptional magnitude increased the transported mass to 8.5% of that recorded at site A. This exceeds the increase during average storms caused by a reduction in vegetation cover. The reduction in vegetation cover at site B to 9% led to the transport of 7.7% of the mass at site A during average winter storm transport. The steep increase caused by the removal of about 50% of the vegetation is attributed to a change in near-surface flow regime. According to Wolfe and Nickling (1993), the roughness concentration Lc determines the flow regime. A vegetation cover of 17%, corresponding to a value of 0.096, results in ‘wake interference flow’ in which the wakes caused by the roughness elements are superimposed, leaving only isolated areas where the near-surface flow may act unrestrictedly onto the surface. ‘Isolated roughness flow’ is caused by a cover of less than 16%, or Lc 0.082, leading to fully developed, isolated wakes in the lee of obstacles, and leaving a large percentage of the surface open to unobstructed flow. Thus, the switch from one flow regime to another caused the steep increase of sand transport. In addition, Ash and Wasson (1983) claim that a plant cover of 5 to 10% with plants having a diameter equal to height causes acceleration of near-surface wind speed as flow is diverted around the obstacle. As the plant geometry at site B fulfils this requirement, increased flux may in part be a result of such accelerated flow. As the natural cover of 17% was close to the threshold for different flow regimes, it appears likely that during exceptional events, as in March 1998 and February 1999, ‘isolated roughness flow’ developed at the experimental site. This could explain the exceptional amounts of transported mass during these storms, which exceeded values expected for 9% cover. The general results of the study are consistent with other published data: a rapid increase of sand transport was found by Marshall (1973) when cover decreased below 15%, while Wiggs (1993) claimed 14% as a threshold, above which only negligible transport occurs. Near-surface horizontal flow velocity u0.24 has been used as the variable determining the intensity of sand movement. Differences in vegetation cover did not lead to differences of threshold speed u0.24t above which saltation starts over noncrusted surfaces, yet actual values of u0.24 are dependent on vegetation cover. Values of mean sand flux during a saltation event and u0.24, measured within the vegetation canopy, show a good correlation (Fig. 15.3) which is independent of vegetation cover; at least up to a vegetation cover of 17%. If the relation between u0.24 over non-vegetated surfaces and different vegetation cover for the same event is considered, it becomes evident that the gradient of the linear relation depends on
222
A. Allgaier
vegetation cover (Fig. 15.4). An influence on u1.77 has not been found, which stresses the importance of measurements within the vegetation canopy. This also confirms the results of Lee (1991), who concluded that evenly spaced vegetation creates an aerodynamically smooth ‘surface’ above the tops of the vegetation. The increase of transported mass as vegetation cover decreases is partly due to the increase of transport duration. Saltation threshold velocity is reached earlier during an event, and events of lower magnitude do cause sand transport as cover decreases. Although no strong relationship has been found, a trend is recognisable. Under 17% cover, no event shorter than 5 h at site A caused sand movement at site B, while under 9% cover, saltation was recorded during events of less than 1 h duration at site A. The absolute flux rates determined for the interdune corridors have to be viewed with caution, as all non-crusted sand surfaces were results of repeated artificial disturbance. No upwind sand source other than the surface within the limited disturbed experimental area exists, as the results at site A (trap A/5) and an undisturbed control site showed (Allgaier 2005). Thus, a depletion of transportable material must be expected, resulting in decreasing flux rates due to a lack of material. It is most likely that this phenomenon is more pronounced at site A than at site B because of the generally higher flux at site A. Despite regular use of a harrow to equalise the surface of site A, several storms uncovered underlying, weakly cemented sand layers, so that parts of the area were not covered by loose sand and, therefore, did not contribute to transport. Vegetation cover and surface stability have been shown to be interdependent (Kadmon and Leschner 1995; Wiggs et al. 1995). A dense vegetation cover will protect a surface from the influence of wind and, in turn, this surface will be more stable than a surface of the same material under identical wind conditions but less vegetation cover. Despite considerable aeolian activity at the non-crusted surfaces, regular manual disturbance was necessary to prevent the re-establishment (A) or an increase (B) of vegetation cover. In addition to germination and growth of vascular plants, microphytic crust growth was fast (see Allgaier 2005, p. 144). Artificially disturbed surfaces at sites A and B showed initial stages of crust growth within a short period of time, unless further manual disturbance was caused. Undisturbed crust-covered surfaces experienced only sporadic sand movement during exceptional events in winter. Amounts were too low for a calculation of sediment flux.
15.5
Conclusions
The results of the study show the importance of vascular vegetation for the reduction of aeolian sand transport in the interdune corridor. At the Nizzana research site, it reduces sand transport in interdune corridors to less than 1%, compared to bare sand surfaces when cover is 17%. A decrease in cover leads to an exponential increase in transported mass caused by higher near-surface wind speeds, increase of exposed area, and prolonged duration of events.
15 Aeolian Sand Transport and Vegetation Cover
223
Acknowledgements The study was carried out as part of DISUM 042, a joint project of Ben-Gurion University of the Negev, Beer-Sheva and Universität Trier. I wish to thank the Arid Ecosystems Research Centre (AERC) at the Hebrew University of Jerusalem for logistic support during the study.
References Allgaier A (2005) Aeolian sand movement in an arid linear dune ecosystem, Nizzana, Western Negev, Israel. PhD Thesis, Universität Würzburg Ash JE, Wasson RJ (1983) Vegetation and sand mobility in the Australian desert dune field. Zeitsch Geomorphol N F suppl 45:7–25 Bagnold RA (1941) The physics of blown sand and desert dunes. Methuen, London Belnap J (1995) Surface disturbances: their role in accelerating desertification. Environ Monitoring Assessment 37:39–57 Belnap J, Gillette DA (1998) Vulnerability of desert biological soil crusts to wind erosion: the influences of crust development, soil texture, and disturbance. J Arid Environ 39:133–142, 165–174 Eldridge DJ, Greene RSB (1994) Microbiotic soil crusts: a review of their roles in soil and ecological processes in the rangelands of Australia. Austr J Soil Res 32:389–415 Harper K, Marble J (1988) A role for nonvascular plants in management of arid and semiarid rangelands. In: Tueller P (ed) Vegetation science applications for rangeland analysis and management, vol 14. Kluwer, Dordrecht, pp 135–169 Hesp PA (1981) The formation of shadow dunes. J Sediment Petrol 51:101–112 Kadmon R (1994) Ecology of linear dunes. II. Differential demographic responses of annual plants to local scale variation in sand stability. Israel J Plant Sci 42:275–284 Kadmon R, Leschner H (1995) Ecology of linear dunes: effect of surface stability on the distribution and abundance of annual plants. In: Blume HP, Berkowicz SM (eds) Arid ecosystems. Adv GeoEcol 28:125–143 Lancaster N, Baas A (1998) Influence of vegetation cover on sand transport by wind: field studies at Owens Lake, California. Earth Surface Processes Landforms 23:69–82 Leatherman SP (1978) A new aeolian trap design. Sedimentology 25:303–306 Lee JA (1991) The role of desert shrub size and spacing on wind profile parameters. Phys Geogr 12:72–89 Leys JF, Eldridge DJ (1998) Influence of cryptogamic crust disturbance to wind erosion on sand and loam rangeland soils. Earth Surface Processes Landforms 23:963–974 Marshall JK (1973) Drought, land use and soil erosion. In: Lovett JV (ed) Drought. Angus and Robertson, Sydney, pp 55–80 McKenna Neuman C, Maxwell C (1999) A wind tunnel study of the resilience of three fungal crusts to particle abrasion during aeolian sediment transport. Catena 38:151–173 McKenna Neuman C, Maxwell C (2002) Temporal aspects of abrasion of microphytic crusts under grain impact. Earth Surface Processes Landforms 27:891–908 McKenna Neuman C, Maxwell C, Boulton JW (1996) Wind transport of sand surfaces crusted with photoautotrophic microorganisms. Catena 27:229–247 McKenna Neuman C, Lancaster N, Nickling WG (1997) Relationships between dune morphology, air flow, and sediment flux on reversing dunes, Silver Peak, Nevada. Sedimentology 44: 1103–1113 Musick HB, Gillette DA (1990) Field evaluation of relationships between a vegetation structural parameter and sheltering against wind erosion. Land Degradation Rehabil 2:87–94 Prasse R (1999) Experimentelle Untersuchungen an Gefäßpflanzenpopulationen auf verschiedenen Geländeoberflächen in einem Sandwüstengebiet. Universitätsverlag Rasch, Osnabrück
224
A. Allgaier
Rosen PS (1979) An efficient, low cost, aeolian sampling system. Current Research Part A, Geological Survey of Canada, pp 531–532 Spaan WP, van den Abeele GD (1991) Wind borne particle measurements with acoustic sensors. Soil Technol 4:51–63 Sterk G, Jacobs AFG, van Boxel JH (1998) The effect of turbulent flow structures on saltation sand transport in the atmospheric boundary layer. Earth Surface Processes Landforms 23:877–887 Stockton PH, Gillette DA (1990) Field measurement of the sheltering effect of vegetation on erodible land surfaces. Land Degradation Rehabil 2:77–85 Thomas DSG, Tsoar H (1990) The geomorphological role of vegetation in desert dune systems. In: Thornes JB (ed) Vegetation and erosion: processes and environments. British Geomorphological Research Group Symposia Series, pp 471–489 Tielbörger K (1997) The vegetation of linear desert dunes in the northwestern Negev, Israel. Flora 192:261–278 Tielbörger K, Kadmon R (1995) Effect of shrubs on emergence, survival and fecundity of four coexisting annual species in a sandy desert ecosystem. Ecoscience 2:141–147 Tsoar H, Møller JT (1986) The role of vegetation in the formation of linear sand dunes. In: Nickling WG (ed) Aeolian geomorphology. Allen & Unwin, Boston, MA, pp 75–95 Walker IJ (1999) Secondary airflow and sediment transport in the lee of a reversing dune. Earth Surface Processes Landforms 24:437–448 West NE (1990) Structure and function of microphytic soil crusts in wildland ecosystems of arid to semi-arid regions. In: Begon M, Fitter AH, MacFadyen A (eds) Advances in Ecological Research, vol 20. Academic Press, London, pp 179–223 Wiggs GFS (1993) Desert dune dynamics and the evaluation of shear velocity: an integrated approach. In: Pye K (ed) The dynamics and environmental context of aeolian sedimentary systems. Geological Society of London Special Publication, vol 72, pp 37–46 Wiggs GFS, Thomas DSG, Bullard JE, Livingstone I (1995) Dune mobility and vegetation cover in the southwestern Kalahari Desert. Earth Surface Processes Landforms 20:515–529 Wolfe SA, Nickling WG (1993) The protective role of sparse vegetation in wind erosion. Progr Phys Geogr 17:50–68
Chapter 16
Soil Processes and Salt Dynamics in Dune Soils P. Felix-Henningsen, B. Rummel, and H.-P. Blume
16.1
Introduction
In Chapter 5 (this volume), we described the soil pattern and characteristics of the sand dune ecosystem of Nizzana. In the following, we will reconstruct the formation of the main soils, their weathering and brownification, their aggregation and crust formation, their humus accumulation and carbonate accumulation being some of the main soil processes. However, our special interest will be on the salt dynamics of the soils. The sand dune ecosystem of Nizzana is influenced by the atmospheric deposition of soluble salts and carbonates (see Chap. 19, this volume). While carbonates derive mainly from exposed limestones of the adjacent mountainous areas, dissolved ions from sea spray are transported by north-westerly winds from the Mediterranean and precipitated through rainfall (Eriksson 1958). Yaalon (1964) estimated annual deposition of up to 100 kg km−2 of marine-borne salts, with a decreasing gradient from the coast inlands. For the Nizzana ecosystem, Littmann and Gintz (2000) reported annual deposition rates of dissolved ions in the range of 800–1,000 kg km−2, calculated from deposition experiments in 1997 and 1998, while Yair et al. (1991) determined an annual salt deposition at Sede Boqer in the range of 800 kg km−2. During evaporation of the soil solution, different salt species are precipitated according to the concentration and composition of ions and the solubility product of the salts. In Nizzana, the limited rainfall of the arid climate is insufficient to leach soluble salts and carbonates from the soils, hence leading to their accumulation. Deposition of carbonates and soluble salts by dust and rainfall, as well as the redistribution of soluble elements by surface runoff and migration of soil solutions influence the spatial distribution of soil types and vegetation cover by affecting salinization and carbonatization. The distribution of vegetation reveals that the perennial Chenopodiaceous species, Anabasis articulata, is found with a high density around the playas, on dune soils of the interdune valley, and on the lower to middle slopes of the longitudinal dune ridges (see Chap. 18, this volume). Because many Chenopodiaceae are halophytes adapted to moderately saline soils (see Chap. 24, this volume), the spatial distribution of A. articulata could be indicative of higher soil salinity. The ion accumulation of
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
225
226
P. Felix-Henningsen et al.
Na+, K+ and Cl− in A. articulata is genetically fixed. It does not correlate with the salt contents of the rooted topsoil (see Chap. 24, this volume). The accumulation of ions results in a lower osmotic potential, which enables water uptake under conditions of higher soil salinity. Field investigations by Ebeling (1996) revealed that the amount of soluble salts is distinctively higher in the upper root zone of A. articulata than in the adjacent, bare interspace soil. One hypothesis is that, apart from the relief influencing the rates of deposition (see Chap. 19, this volume) and redistribution of carbonates and soluble salts, the perennial shrub vegetation favours the concentration of elements within the upper root zone. In order to identify these processes of salt distribution within the ecosystem, normative salt concentrations, water-soluble ions, carbonates and fines were investigated at a local scale, accounting for the vegetation pattern.
16.2
Methods
Methods of soil description and laboratory analysis of representative soils are given in Chapter 5 (this volume). For our special studies of salt dynamics, soil sampling of the top and subsoil horizons was carried out on Haplic and Yermic Arenosols (FAO 2006) on both south- and north-facing slopes of a linear dune. Both slopes have an angle of 7–8°. At all sites, soil pits of up to 120 cm depth were excavated to expose the root zone of A. articulata shrubs (canopy about 1–1.5 m in diameter), with similar pits for the adjacent interspace where a vegetation cover was absent. In order to ascertain the ranges and spatial variability of important soil characteristics, investigations were carried out on shallow soil profiles of Arenosols below A. articulata and Retama raetam shrubs and the interspace between shrubs along transects following the north and south exposed slopes of the linear dunes. The lateral distance between the 10 sampling sites of each transect was at least 50 m, in order to cover a large area. Samples were taken from the indurated crust, from which the loose sand was blown off (0.5–1 cm thick in the interspace, 2–5 cm thick below shrubs), the topsoil below the crust down to 10 cm depth, and the subsoil between 25 and 35 cm depth. The laboratory methods applied here are described in detail by Schlichting et al. (1995). Due to the low content of fines, the sand and coarse silt fractions of most samples were determined only by sieving, after removal of salts and carbonates and dispersion by sodium pyrophosphate. Fine material<20 µm was calculated as the difference between the sieve fractions and the carbonate-free sample weights. The pH values of suspensions with a ratio soil:distilled water of 1:2.5 were determined potentiometrically by means of a glass electrode. Contents of carbonates were determined gas volumetrically. A rather quick, initial (normative) approximation of the contents of soluble salts was gained by the determination of the electrical conductivity (EC) of clear suspensions with soil:distilled water ratios of 1:2.5
16 Soil Processes and Salt Dynamics in Dune Soils
227
(EC2.5) and 1:5 (EC5). The EC5 was used for the calculation of the normative total amount of dissolved salts (TDS), based on Simon et al. (1994). The proportions of easily soluble salts (ESS) were calculated according to Ruck and Stahr (1987), on the basis of differences between EC2.5 and EC5; the proportion of less soluble salts (LSS) is given by the difference between TDS and ESS. Further detailed information about concentrations and species of salts of selected profiles was gained from the concentrations of soluble cations and anions. Cations and anions, dissolved in the clear extracts of the EC5 suspensions, were determined by using an atomic adsorption spectrometer for the cations, and a Dionex Anion Chromatograph for the anions. On the basis of the concentrations of cations and anions, the salt species were modelled according to Smettan and Blume (1987). The determination of total carbon (TC) and total organic nitrogen (TON) was carried out by dry combustion at 950 °C of 20–100 mg of oven-dry (105 °C), finely ground samples. The combustion gases CO2 and N2 are separated by gas chromatography and measured with a TCD. The proportion of carbon bound in carbonates was subtracted from the total amount of C in order to calculate the amount of organic carbon (Corg).
16.3 16.3.1
Soil Formation Formation of Parent Material
The dunes are aeolian sediments. Soils of the interdunal depressions are strongly stratified with sandy, silty and loamy layers (profiles 3–7, 9 in Table 5.1, Chap. 5, this volume). According to Yair (1990), silt will be accumulated by lateral movement after heavy rainfall in the lower topographical positions; Savat (1982) assumes that fine-textured layers are the result of accumulation after texture fractionation due to outwash after heavy rain. A part of the silt is imported by wind and episodic rainfall: Littmann (1997) calculated a yearly import of about 15 g m−2 of dust, although the deposition may vary due to the uneven topography. By contrast, the silty and loamy layers at larger distances from the foot slopes seem to be accumulated by episodic flooding of the interdune area by wadi waters, mainly Wadi Nizzana, which borders the dune area in the east (Pfisterer et al. 1996).
16.3.2
Weathering, Brownification and Redoximorphism
Soils of arid climates are relatively stable in their pedogenic features, because chemical processes are limited to the short time when water is present. Colours were in the range of 8–10 YR, indicating goethite as the dominating iron oxide. Visually, there was no change in brownification within the profiles. The quantity
228
P. Felix-Henningsen et al.
of amorphous iron oxide (Feo) is generally low in all analyzed profiles (Table 5.1, Chap. 5, this volume). Higher amounts of free iron oxides were found in the finetextured layers of the Solonchaks and Calcisols: here, the free oxides are rather lithogenic, as a result of in situ weathering. The redoximorphic features with mottling of iron oxides in fine-textured layers could be derived from various events such as flooding or infiltration after rainstorms. Soils of such a state of chemical immaturity must therefore be defined by their physical state, such as the development of aggregates and crusts, and by their salt dynamics.
16.3.3
Aggregation and Cracking
The Arenosols are single-grained with laminar structures due to aeolian sedimentation. The Solonchaks, Fluvisols and Calcisols show prismatic and angular aggregates, and buried cracks with depths sometimes exceeding 10 cm. The cracks are up to 5 cm wide and sand-filled. Loamy soils tend to shrink and swell in response to changing water content in the soils. As it was observed, the formation of cracks was weak even after a period of high rainfall. Deep cracks of the soils are, therefore, a relict feature created by deep penetration of water in the past. Cracks cause rapid drainage and the removal of salt and nutrients, along with rapid desiccation and diminution of available water. Dark cutans on some of the aggregates show that rapidly percolating water translocated particles of organic matter and clay.
16.3.4
Crust Formation
The topsoil of the Yermic Arenosols is covered by a greenish-brown biological crust. Microscopically hyphae-like filaments can be seen connecting the sand particles, thereby consolidating the surface, as Evenari (1985) has described (see Chap. 10, this volume). This biological crust consists of an up to 2-mm-thick, friable crust of cyanobacteria, algae, lichens, bacteria or fungi (see Chap. 10, this volume), which covers a 0.5–5 cm thick inorganic crust of more or less cemented dune sand. Cementation results from the accumulation of fines (< 63 µm), soluble salts and carbonates (Ebeling 1996). Also the contents of windblown silt as well as of organic carbon within the crust were mostly higher than in the horizon below (see, e.g. layers 5.1, 6.1 and 7.1 in Table 5.1, Chap. 5, this volume), reflecting the increased biomass production near the surface. The thickness of the biological crust varies systematically in function of slope aspect, altitude and shrub vegetation. Generally, crust thickness increases continuously down-slope but is spatially highly variable due to crust destruction and subsequent partial regeneration, resulting from both human impact (research activities) and
16 Soil Processes and Salt Dynamics in Dune Soils
229
bioturbation. In all relief positions, the crust displays a significant increase in thickness from <1–1 cm in the interspace to >2 cm below the canopy of perennial shrubs. Generally, the crust under A. articulata is significantly thicker (4–5 cm) and more strongly indurated than the crust under R. raetam (1–3 cm) or other shrub species. The induration of the crust increases with their thickness. Furthermore, the biological crust in the interspace between shrubs of the north-exposed slope is thicker, more stable and covers the slope to a higher level than is the case for the south-exposed slope (see Chap. 10, this volume). Compared to the surficial biological crust, the ACk and Ck horizons of the Yermic Arenosols of both slope expositions are only slightly indurated, due to a weaker accumulation of salts and carbonates. Nevertheless, the exposed walls of pits in the dry soil remain stable. The soil structure is coherent without shrinkage cracks, since these occur in the more strongly indurated Yermic Arenosols and Calcisols on ancient dunes, where the inorganic crust is much stronger. Although the crust is extremely sensitive to mechanical stress, the dunes become rather well stabilized against wind erosion (see also Danin et al. 1989; Amit and Harrison 1995). Disturbed crusts probably regenerate rather quickly. On a sample of loose sand from the dune, a crust formed on the surface within a few days of having sprayed a thin film of water onto the sand (Pfisterer et al. 1996). The regeneration of crusts is dealt with in detail in Chapter 20 (this volume). Pure inorganic crusts may be formed by sedimentary processes where aeolian silt or clay have been washed into the soil and accumulated at the depth of water penetration. After erosion, the silty layer appears at the surface. Because of the lower water conductivity of the silty layers, salts become enriched on the surface, and this inorganic crust becomes hardened by solution and recrystallization of soluble salts (Evenari 1985).
16.3.5
Humus Accumulation
The content of organic matter in all soils is very low, reflecting the low biomass production in this arid environment, except in some silty layers in the interdune areas. The latter may have various sources: lithogenic origin of wadi sediments, and higher plant production when exposed at the surface due to higher water capacity. In some layers/horizons, the C/N ratio was lower than 10 but it ranged between 2 and 5.5. It can be stated that the lower the quantity of C, the lower the C/N ratio (Table 16.1). There is no relation between texture and C/N ratio. A very narrow ratio indicates a high rate of decomposition, similar to that reported for tropical forests (Steinberger and Whitford 1988), although some narrow ratios may be the consequence of N accumulated by precipitation. Littmann (1997) calculated an input of N at the Nizzana site of 0.4 kg ha−1 year−1. Besides that, presumably the abundance of N-fixing microbes in the surface crust (Le Houerou 1986, and others) led to a higher concentration of N near the soil surface.
230
P. Felix-Henningsen et al.
Table 16.1 Nizzana (South): characteristics of Calcaric Arenosols, medians and spatial variation (c.v. coefficient of variance, n.d. not determined) Crust Topsoil Subsoil Median Anabasis articulata, n = 20 Carbonate (g kg−1) pH (H2O) Corg (g kg−1) Nt (g kg−1) C/N TDSnorm (mg kg−1)a HSSnorm (mg kg−1) LSSnorm (mg kg−1) Ca2+ (mg kg−1) Mg2+ (mg kg−1) K+ (mg kg−1) Na+ (mg kg-1) SO42− (mg kg−1) NO3− (mg kg−1) Cl− (mg kg−1) C. silt 63–20 µm Fines < 20 µm
Crust 0–5 cm 85 8.34 4.9 0.5 9.9 760 343 355 42 39 260 87 40 33 34 68 88
Retama raetam, n = 20 Carbonate (g kg−1) pH (H2O) Corg (g kg−1) Nt (g kg−1) C/N TDSnorm (mg kg−1) HSSnorm (mg kg−1) LSSnorm (mg kg−1) Ca2+ (mg kg−1) Mg2+ (mg kg−1) K+ (mg kg−1) Na+ (mg kg−1) SO42− (mg kg−1) NO3− (mg kg−1) Cl− (mg kg−1) C. silt 63–20 µm Fines < 20 µm Interspace, n = 30 Carbonate (g kg−1) pH (H2O) Corg (g kg−1) Nt (g kg−1) C/N TDSnorm (mg kg−1) HSSnorm (mg kg−1) LSSnorm (mg kg−1) Ca2+ (mg kg−1) Mg2+ (mg kg−1) K+ (mg kg−1)
c.v. (%) Median 22 37 27 38 24 35 56 30 85 79 27 87 149 143 114 85 39
ACk 5–10 cm 64 9.02 2.0 0.2 10.0 404 172 220 44 29 128 50 19 15 21 23 77
Crust 0–3 cm 95 7.85 3.4 0.3 11.0 466 196 270 131 28 75 34 42 4 35 93 115
16 33 27 34 27 23 52 16 23 17 49 31 44 147 57 34 15
Crust 0–3 cm 86 8.00 1.9 0.2 12.5 392 135 245 84 20 30
21 52 58 39 29 22 34 20 62 42 46
c.v. (%) Median
c.v. (%)
25 56 42 52 60 37 63 24 41 55 47 153 120 106 116 57 35
Ck 10–30 cm 60 9.08 n.d. n.d. n.d. 326 175 151 48 33 70 49 30 9 21 23 68
17 58 n.d. n.d. n.d. 38 66 20 67 48 67 81 107 98 109 62 24
ACk 3–8 cm 66 8.28 1.5 0.1 13 233 61 174 72 22 41 18 8 1 11 49 90
16 40 57 66 33 24 67 13 50 50 72 33 115 130 99 44 17
Ck 8–30 cm 52 8.72 n.d. n.d. n.d. 172 74 123 78 30 23 18 13 1 15 27 76
13 26 n.d. n.d. n.d. 20 52 26 24 44 64 30 79 82 60 51 12
ACk 3–8 cm 67 8.75 n.d. 0.1 12.5 147 25 123 71 23 25
18 53 n.d. 82 33 19 47 27 60 30 50
Ck 8–30 cm 56 8.86 n.d. n.d. n.d. 147 37 123 73 33 24
16 98 n.d. n.d. n.d. 17 60 11 32 41 41
(continued)
16 Soil Processes and Salt Dynamics in Dune Soils
231
Table 16.1 (continued) Crust Median c.v. (%)
Topsoil Median c.v. (%)
Na+ (mg kg−1) 29 53 14 54 30 73 6 99 SO42− (mg kg−1) 1 94 n.d. n.d. NO3− (mg kg−1) 34 84 16 66 Cl− (mg kg−1) C. silt 63–20 µm 66 52 38 67 Fines<20 µm 107 27 91 27 a TDS, total dissolvable salts; HSS, highly soluble salts; LSS, less soluble salts
16.4 16.4.1
Subsoil Median c.v. (%) 14 4 n.d. 8 26 74
158 73 n.d. 65 70 22
Salt Dynamics Salinization of the Playas
Playas, which are distributed in the western edge of the dune valleys of Nizzana, developed from the sedimentation of fines transported by the inflow of Nahal Nizzana during the Pleistocene (Harrison and Yair 1998). High concentrations of soluble salts and carbonates of the playas caused the development of a Calcic Solonchak within shallow depressions (profile 3 at depths of 0–16 and 56–80 cm in Table 5.1, Chap. 5, this volume). Further investigations of Ebeling (1996) and Grenz (2000) show that the maximum concentration of salts and carbonates is clearly bound to layers rich in clay and silt, while interstratified layers of aeolian sands are much less saline (see Arenosol profile 4 at the depth of 95–140 cm in Table 5.1, Chap. 5, this volume). The main salt species of the clay-rich layers is NaCl, with minor admixtures of gypsum, both nearly completely absent in the Arenosols of the dunes. The pH ranged from 7.7 to 10.5. In the Solonchaks where neutral salts such as NaCl are dominant with proportions of 40–90% (Fig. 16.1), pH is around 8. In the less saline layers of the Arenosols with relatively more Na2CO3, pH rises to 10.5 and the EC values vary between 0.7 and 115 dS m−1. The distribution of the salts provides hints about their origin (Smettan and Blume 1987): in soils under the influence of groundwater, the highly soluble salts (chlorides and nitrates) should precipitate near the surface above carbonates and sulphates. Salt-forming ions from infiltrating rain or in drying ponds precipitate in reverse. In Fig. 16.1, the position of the various salts within the profiles are shown as calculated by their solubility, and give a rough estimate of the probable distribution of the salts. As the highly saline playa sediments also occur below younger aeolian sand sheets and shallow dune ridges (apart from the slopes of the linear dunes), it can be assumed that the salinity of the playa developed mainly by sedimentary precipitation of salts in drying ponds or small lakes which persisted following the regression of Nahal Nizzana from the interdune valleys. Furthermore, deposition of airborne salts and their leaching to deeper layers caused the formation of salt
232
P. Felix-Henningsen et al.
Fig. 16.1 Calculated locations of different types of salt (after Rauscher et al. 1982, similar to Smettan and Blume 1987, and Pfisterer et al. 1996). Profile 1 Haplic Arenosol, 3 Calcic Solonchak, 4 Yermic Arenosol, 5 Yermic Calcisol, 6 Yermic Fluvisol, 8 Protic Arenosol, 9 Yermic Calcisol (for further properties, see Table 5.1 in Chap. 5, this volume)
enrichment horizons. The soils on dunes adjacent to the playas show no evidence (apart from local transport of sand and fines by rill erosion at foot slopes of the linear dunes and ancient dunes) of a catenary translocation of salts by runoff or subsurface migration of soil solutions from the Arenosols on linear dune ridges towards the interdune valleys. In the vicinity of slightly elevated areas of the flat playa surface, Endosali-yermic Calcisols are developed, displaying carbonate contents of 20–35 mass% and salinities (EC2.5) of 1–4 mS cm−1 at pH 8.5–9.
16.4.2
Salt Dynamics of Arenosols on Vegetated Linear Dunes
The Arenosols on the slopes of the linear dune ridges as well as those of shallow dune ridges and sand sheets of the interdunes are vegetated by perennial shrub communities and annual plants (see Chap. 8, this volume). The horizons of the Arenosols are generally weakly developed, due to the young soil age of about 20 years. Heavy nomadic grazing in the area (see Chap. 6, this volume) ceased in 1982, and enabled the regeneration of the vegetation cover and the stabilization of the dunes. Nevertheless, leaching of salts and carbonates over long periods have led
16 Soil Processes and Salt Dynamics in Dune Soils
233
to the formation of calcic enrichment layers deep below the rooted soil (Amit and Harrison 1995). Calcareous Ah horizons with low humus contents (0.3–0.5 mass%) occur only below the canopies of shrubs. Compared to saline soils (EC2.5>1 mS cm−1; see Alaily 1996), the salt content of the Calcaric Arenosols is low (EC2.5<0.5 mS cm−1 in all horizons, even below A. articulata) and obviously without any physiological effect on the annual and perennial vegetation (see Chap. 24, this volume). Thus, the Calcaric Arenosols can be designated as “weakly saline”. Sali-calcaric Arenosols occur on the shallow dune ridges and aeolian sand sheets adjacent to the playa (cf. Pfisterer et al. 1996). Their higher salinity can be traced back to deposition of saline dust blown from the surface of the playa, and the occurrence of saline playa sediments below the dune sand. The uniform texture of the unweathered dune sand (Ck horizon) dominated by fine sand, and the uniform mineralogy with the predominance of quartz and low contents of carbonates (5–6 mass% CaCO3) and soluble salts (about 150 mg kg−1) provide us with the possibility of identifying systematic distribution patterns of fines, carbonates, soil acidity and alkalinity, and water-soluble elements in the biological crust and upper soil horizons. The variation in concentration (and masses) between horizons and between sites is a result of spatial variations in deposition by dust and rainfall, and a consequence of the short-range redistribution of mobile elements by surface runoff (cf. Kidron and Yair 1997) and migrating soil solutions. Figure 16.2 shows the depth distribution of salts (normative calculation) of sites on the lower mid-slopes of a north- and south-facing slope below A. articulata, compared to the bare interspace between shrubs. In Yermic Arenosols of the interspace, and below A. articulata, the maximum concentration of soluble salts occurs in the biological crust, decreasing more or less erratically with depth (depending on relative solubility). The total amount of salts (HSS+LSS) of all soil horizons under A. articulata is distinctively higher than in adjacent interspaces between the shrubs. The soil on the north-facing slope shows a higher salt concentration in the interspace and below A. articulata than do the soils of the south-facing slope. It is thus inferred that at least the amount of rainfall and wet deposition is higher on the north-exposed slope, while Littmann (1997) confirms a higher dry deposition of dust due to turbulent airflow on the south-exposed slope. The biological crust and soil horizons on the south-facing slope show a higher proportion of highly soluble salts, while the Arenosols of the north-facing slope display a higher proportion of less soluble salts. This would coincide with a higher amount of rainfall causing preferential leaching of highly soluble salts. Enrichment of leached soluble salts in deeper horizons is visible only in the Arenosol of the south-exposed interspace. Leaching of deposited soluble salts to the deeper root zone is impeded below the shrubs, due to a high interceptive evaporation of rainfall from the dense canopy of A. articulata. The interspace profile on the north-exposed slope, on the other hand, shows no deep enrichment horizon. As argued above, the higher amount of rain could have led to a leaching of soluble salts and accumulation in enrichment horizons below the depth of the profile. It can be assumed that leaching and accumulation of soluble salts show a high spatial variability, due to relief and vegetation, but to date no systematic pattern has
234
P. Felix-Henningsen et al. under Anabasis articulata [cm]
[cm]
N-facing slope
0
S-facing slope
0 1350
5
5
10
10
15
15
20
20
25
25
30
30 0
200
400
600
0
800
200
[mg*kg−1]
400
600
800
[mg*kg−1]
interspace between shrubs [cm] 0
N-facing slope
[cm]
S-facing slope
0
5
5
10
10
15
15
20
20
25
25 30
30 0
200
400
600
[mg*kg−1]
800
Highly Soluble Salts
0
200
400
600
[mg*kg−1]
800
Less Soluble Salts
Fig. 16.2 Nizzana (South): normative salt contents (calculated from EC in 1:2.5 and 1:5 soil: water extracts) in Calcaric Arenosols of vegetated middle slopes of linear dunes below Anabasis articulata (top) and the interspace between shrubs (bottom)
been confirmed. As the highest concentrations of fines, carbonates and soluble salts are generally found in the crust and the underlying soil horizons, decreasing down to a depth of 40 cm (Fig. 16.2), the transect investigations of shallow profiles below the canopies of A. ariculata and R. raetam and the adjacent interspace reveal a clearer picture of the influences of perennial shrubs on the distribution of fines, carbonates and soluble salts in Calcaric Arenosols of the linear dunes. The results are presented in Table 16.1.
16 Soil Processes and Salt Dynamics in Dune Soils
235
The deposition and redistribution of airborne dust (coarse silt, fines<20 µm, carbonates, Table 16.1) as well as soluble cations and anions, which lead to the precipitation of highly soluble (HSS) and less soluble salts (LSS), show a clear relationship to vegetation cover and slope aspect (Table 16.1). The concentrations and masses of coarse silt, fines and carbonates, which together result from the dry deposition of dust, are highest in the biological crust below both shrub species and in the interspace, significantly higher than in the topsoil or subsoil. While coarse silt shows a sharp decrease below the crust, the amount of fines < 20 µm remains in the same range in the topsoil, too, and decreases gradually towards the subsoil (Table 16.1). The median values of the amounts of carbonates, coarse silt and fines < 20 µm (Table 16.1) display higher concentrations below R. raetam than below A. articulata. This may reflect higher dust deposition below R. raetam, a consequence of larger size and the more permeable canopy of the shrub (cf. Littmann 1997; see Chap. 19, this volume). Nevertheless, statistical median tests show no significance in the differences in carbonate content, coarse silt and fines < 20 µm of the same horizons below the shrub species and the interspace, due to high spatial variability. Rather narrow (r2>0.65) and highly significant correlations between carbonates, coarse silt and fines in the soil horizons under R. raetam indicate dust deposition as the source, while in soil horizons under A. articulata a significant inter-element correlation cannot be proven. The median values of the normative salt contents, calculated from the EC, as well as of the pH, water-soluble cations and anions (Table 16.1) reveal higher concentrations in the biological crust than in the underlying soil horizons, as well as differences between concentrations under both shrub species and the interspace. The pH value is generally lower in the biological crust, and increases with depth in the underlying soil horizons, due to the formation of protons as a consequence of respiration of the cyanobacteria and the decomposition of bacterial organic matter with a C/N ratio of 10. As HCO3− and CO32− are the dominant anions, the modelled salt species show a high proportion of NaHCO3, which causes pH values above 8.2. The formation of NaHCO3 results from the reaction of deposited sea-spray NaCl with Ca(HCO3)2, which is co-precipitated or formed within the soil horizons during moist periods due to dissolution of carbonates. As a consequence, CaCl2 is also formed, which has a slightly higher solubility than NaCl and is easily leached from the upper soil horizons (cf. Pfisterer et al. 1996), while the solubility of NaHCO3 is six to seven times lower than that of NaCl or CaCl2. Based on the results of statistical median tests, the differences in many of the medians are not significant, due to high spatial variability (see below). Compared to the interspace, the pH of the biological crust below Anabasis is significantly higher but significantly lower under Retama. Consequently, the pH of the biological crust below both shrubs differs significantly, which coincides with the higher Na+ content in the crust and underlying soil horizons below Anabasis. Compared to the interspace, the crust below both shrubs shows significantly higher concentrations and, due to the greater thickness of the crust, also much higher masses of Cl−, NO3−, SO42−, Ca2+, Mg2+, Na+, K+ and normative salts (TDS, HSS and LSS).
236
P. Felix-Henningsen et al.
Apart from the pH, Anabasis had significantly higher concentrations of K+, Na+ and higher normative salt contents (HSS, LSS) in the biological crust as well as in the underlying soil horizons. Although the medians of the other ion concentrations show clear differences between the shrubs, these are not significant. The generally higher amounts of salts and carbonates within the biological crust results from the retention of dust and ions due to adsorption, and the reduction of infiltration as a consequence of higher capillarity than in the underlying unconsolidated sand. Several reasons must be considered to account for the higher amounts of fines, carbonates and soluble salts below perennial shrubs, compared to the interspaces free of vegetation: 1. Soluble elements can migrate, even over short distances, by surface runoff from the crust of the interspaces down-slope to the vegetated islands during strong rainfall events. 2. Shrub canopies can intercept dust, carrying fines and carbonates, especially under R. raetam (cf. Littmann 1997; see Chap. 19, this volume). 3. The high surface area, especially of the leaved branches of A. articulata, filters rainfall. The drops evaporate, leading to a precipitation of salts on the leaves and branches. These salts are then washed off by subsequent rainfall as well as by dew or fog. 4. The active excretion of excessive salts by enhanced leaf shedding in A. articulata can be explained by the higher Na+ concentrations in the leaves of this shrub (Veste and Breckle 2000; see Chap. 24, this volume). 5. The deposition and decomposition of litter below the shrubs lead to a release of organically bound elements, which explains the relatively high concentrations and masses of Na+, K+ and NO3−, especially in the soil horizons below A. articulata. Differences in C/N ratios of the litter of A. articulata, which displays ratios of 23–39 with a median of 35, and litter of R. raetam, with ratios of 40–52 and a median of 42, account for the higher rate of litter decomposition of Anabasis. The impeded litter decomposition of Retama leads to an accumulation of a thick layer on the soil surface.
16.5
Spatial Variability of Soil Characteristics
The analytical data of the transect profiles in all horizons show a wide range reflecting high spatial variability along the transects, indicated by the coefficient of variance (c.v.) based on the mean values (not displayed in Table 16.1). All soil parameters display coefficients above 25%. Most soil characteristics of the crust, especially below the shrubs, display a higher variability than is the case for the topsoil and the subsoil. Here, the concentrations of carbonates, fines and soluble salts are strongly modified by the activity of the cyanobacteria, due to favourable light intensity and moisture conditions. Furthermore, burrowing and bioturbation of soil animals as
16 Soil Processes and Salt Dynamics in Dune Soils
237
well as sedimentation of aeolian sand, which is very pronounced below shrubs, lead to the deposition of material from the topsoil and subsoil on the surface. The coefficients of variance show a clear differentiation in terms of the relative mobility of the constituents and ions. Na+ and all anions, which are highly mobile in the soil solution, display much higher variability with c.v. values reaching 150%, especially below shrubs, than do the other cations. Differences in deposition due to slope aspect and shrub species, leaching and mineralization, or in rates of N fixation due to differences in the activity of the biological crust account for the high spatial variability of these elements. It is clear that only a statistical evaluation of data from a large number of systematically taken profiles can identify valid spatial patterns and trends of element fluxes within the ecosystem. Assessments of ecosystem processes or mass balances which are based on local soil investigations may not be representative and, hence, lead to incorrect interpretation or conclusions.
16.6
Conclusions
In Nizzana, the main differences in soil salinity between the playas, ancient dunes and younger aeolian sands, forming linear dunes as well as sand sheets and low dune ridges within the interdunes, result from the past geological, geomorphological and palaeopedological development of the landscape. Catenary element fluxes via migration of soil solutions and surface runoff are visible only in the soil association of the playa. Haplic and Yermic Arenosols on linear dunes as well as sand sheets and low dune ridges, which cover most of the area, display low salinity within the upper 1 m. Fines and carbonates from dry deposition, and ions, forming soluble salts resulting predominately from wet deposition are accumulated mainly in a slightly indurated biological soil crust at the surface and in the underlying 10 cm of the topsoil. Deposited elements are systematically distributed according to variations in microclimatic parameters influenced by slope angle and orientation. The perennial shrub vegetation (depending on the species) accumulates fines, carbonates and soluble salts due to canopy interception, salt excretion and mineralization of litter, with time leading to increasing salinity of the root zone. Small-scale differences in deposition and leaching of elements, movement and sedimentation of aeolian sand, burrowing and bioturbation as well as type, age and morphology of shrub species can cause high spatial variability of accumulated elements. Accordingly, results of local soil investigations alone should not be used to explain soil processes and their relationship to large-scale desert ecosystem functioning. Acknowledgements The authors thank the Bundesministerium für Bildung und Forschung (BMBF) for funding this project (BEO 0339702). We are also grateful to the Hebrew University of Jerusalem Minerva Arid Ecosystems Research Centre for providing long-term technical and logistical assistance, especially the efforts of E. Sachs and S.M. Berkowicz.
238
P. Felix-Henningsen et al.
References Alaily F (1996) Carbonate, sulfate, chloride, phosphate. In: Blume H-P, Felix-Henningsen P, Frede H-G, Horn R, Stahr K (eds) Handbuch der Bodenkunde vol I, Chap. 2.1.1.3, pp 1–12. Ecomed, München Amit R, Harrison JBJ (1995) Biogenic calcic horizon development. Adv GeoEcol 28:65–88 Danin A, Bar-Or Y, Dor I, Yisraeli T (1989) The role of cyanobacteria in stabilization of sand dunes in Southern Israel. Ecologia Med 9:55–64 Ebeling D (1996) Salzdynamik in Böden des Dünengebietes von Nizzana. Diploma Thesis, Westfälische Wilhelms-Universität, Münster Eriksson E (1958) The chemical climate and saline soils in the arid zone. In: Climatology, Reviews of Research, Arid Zone Research vol 10, pp 147–188. UNESCO, Paris Evenari M (1985) The desert environment. In: Evenari M, Goodall DW, Noy-Meir EM (eds) Ecosystems of the World vol 12A. Elsevier, Amsterdam, pp 1–19 FAO (2006) Guidelines for soil description, 4th edn. FAO, Rome Grenz JH (2000) Bodenentwicklung auf Altdünen in einem Dünengebiet des nordwestlichen Negev, Israel. Diploma Thesis, FB 09, Justus-Liebig-Universität Giessen Harrison JBJ, Yair A (1998) Late Pleistocene aeolian and fluvial interactions in the development of the Nizzana dune field, Negev desert, Israel. Sedimentology 45:507–518 Kidron GJ, Yair A (1997) Rainfall-runoff relationship over encrusted dune surfaces, Nizzana, Western Negev, Israel. Earth Surface Processes Landforms 22:1169–1184 Le Houerou HN (1986) The deserts and arid zones in Northern Africa. In: Evenari M, Goodall DW, Noy-Meir EM (eds) Ecosystems of the World vol 12B. Elsevier, Amsterdam, pp 101–143 Littmann T (1997) Atmospheric input of dust and nitrogen into the Nizzana sand dune ecosystem, northwestern Negev, Israel. J Arid Environ 36:433–457 Littmann T, Gintz D (2000) Eolian transport and deposition in a partially vegetated linear sand dune area (northwestern Negev, Israel). Z Geomorphol Suppl Bd 121:77–90 Pfisterer U, Blume H-P, Beyer L (1996) Distribution patterns of soils of an arid dune area in northern Negev. Z Pflanzenernähr Bodenk 159:419–428 Rauscher KJ, Voigt J, Wilke J, Wilke Th (1982) Chemische Tabellen und Rechentafeln für die analytische Praxis. VEB Verlag für Grundstoffchemie, Leipzig Ruck A, Stahr K (1987) Wasser- und Salzhaushalt aus der Leitfähigkeit verschiedener Boden-zuWasser-Extrakte. Mitt Dtsch Bodenk Ges 55(1):233–238 Savat J (1982) Common and uncommon selectivity in the process of fluid transportation. In: Yaalon DH (ed) Aridic soils and geomorphic processes. Catena suppl 1:139–160 Schlichting E, Blume H-P, Stahr K (1995) Bodenkundliches Praktikum, 2nd edn. Blackwell, Berlin Simon M, Cabezas O, Garcia I, Martinez P (1994) A new method for the estimation of total dissolved salts in saturation extracts of soils from electrical conductivity. Eur J Soil Sci 45:153–157 Smettan U, Blume H-P (1987) Salts in sandy desert soils in South-western Egypt. Catena 14:333–343 Steinberger J, Whitford WG (1988) Decomposition process in Negev ecosystems. Oecologia 75:61–66 Veste M, Breckle S-W (2000) Ionen- und Wasserhaushalt von Anabasis articulata in Sanddünen der nördlichen Negev-Sinai-Wüste. In: Breckle S-W, Schweizer B, Arndt U (eds) Ergebnisse weltweiter ökologischer Forschung. Günter Heimbach, Stuttgart-Hohenheim, pp 481–485 Yaalon DH (1964) Airborn salts as an active agent in pedogenetic processes. In: Ext Abstract Vol 8th Int Congr Soil Science, Bucharest, no 5, pp 997–1000 Yair A (1990) Climate change and desert environment, Northern Negev, Israel. Earth Surface Processes Landforms 15:597–609 Yair A, Karnieli A, Issar A (1991) The chemical composition of precipitation and run-off water on an arid limestone hillside, northern Negev. Israeli J Hydrol 129:371–388
Chapter 17
Runoff and Erosion Processes Within a Dune System G.J. Kidron and A. Yair
17.1
Introduction
The occurrence of surface runoff processes is often regarded by geomorphologists or hydrologists as irrelevant within sandy areas and dune systems, due to the very high infiltration rates of sand. The presumed lack of runoff processes has been mentioned several times in the case of the northwestern Negev sand field (Hillel and Tadmor 1962; Tsoar and Zohar 1985; Tsoar and Møller 1986). The high infiltration rates of over 100 mm h−1 for loose sand, far above the common range of rain intensities in the Negev, may explain the assumed unlikelihood of runoff generation within the Negev dune systems. However, field observations have drawn attention to the existence of widespread biological topsoil crusts (Danin 1978; Tsoar 1990; Yair 1990), rich in fine-grained particles that are expected to reduce infiltration rate (Chaps. 10 and 18, this volume). In order to obtain an initial insight into the issue of runoff generation in the Nizzana sandy area, a sprinkling experiment, with an intensity of 18.4 mm h−1, was conducted over a small plot covering 1.5 m2 (Fig. 17.1). The experiment was performed in the winter time, under wet surface conditions. Runoff developed within 3 minutes, after only 1 mm of rain. Final infiltration rate was reached quickly, and amounted at ~12 mm h−1 (Yair 1990). Shortly after this run, a second sprinkling experiment was performed over the same plot after the topsoil crust had been removed. Rain intensity in the second run was increased to 53 mm h−1. Despite the antecedent wet conditions, and the extremely high rain intensity applied, no runoff was observed for 42 minutes, by which time the accumulated rain amount reached the value of 37 mm. The results of the two experiments clearly demonstrated the importance of the biological crust in limiting infiltration and enhancing runoff generation. It was therefore postulated that the microbiotic crust may affect the hydrological behavior of the surface under natural rain conditions, and the spatial distribution of water resources in the study area. The results of the sprinkling experiments cannot adequately represent processes under natural rainfall conditions, for the following reasons:
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
239
240
G.J. Kidron, A. Yair
Fig. 17.1 Results of sprinkling experiment
1. The sprinkled area covers only 1.5 m2, where the topsoil crust is relatively uniform, while under field conditions crust thickness and composition may vary along the dune slope as well as laterally. 2. The sprinkling was conducted under wet surface conditions, making it impossible to verify whether the crust possesses water-repellent properties, usually very pronounced under dry conditions (Rutin 1983). Water repellency would be expected to limit infiltration rate, and enhance surface runoff generation. 3. Finally, rainfall was applied at a uniform intensity, while natural rainstorms in the area are intermittent, and characterized by high temporal variability in rain intensities.
17.2
Field Instrumentation and Methodology
For the reasons listed above, the research site was equipped with devices for the measurement of rainfall and runoff. Runoff plots were constructed on north (N)and south (S)-facing slopes in the upper, middle, and lower parts of the dune slopes (Fig. 17.2). The location of the plots, and their characteristics are given in Fig. 17.3 and Table 17.1. Runoff was measured either with a V-notch weir and a pressure gauge transducer or by collection in large containers. Rainfall was recorded with an electronic tipping-bucket recorder. Ten large plots were constructed, using 0.5mm-thick metal sheets, 20 cm high, inserted 10 cm into the sand. Four of the large plots were constructed on the north-facing slope (N1–N4) and four on the southfacing slope (S1–S4). Two additional large plots (C1, C4) were constructed on the north-facing slope of a low, stabilized dune. Of the four plots constructed on each of the active (i.e., having a mobile crest) dunes, two plots were subdivided into three subplots: N1.1 and N2.1 included the upper active section of the north-facing
17 Runoff and Erosion Processes Within a Dune System
241
Fig. 17.2 View of runoff plots
Fig. 17.3 Layout of the experimental site
exposure, while S1.1 and S3.1 drained the upper active area of the south-facing exposure. Likewise, N1.2 and N2.2 on the northern exposure, and S1.2 and S3.2 on the southern exposure drained the semi-stabilized mid-slopes. N1.3, N2.3, and S1.3, S3.3 drained the lower flanks of the northern and southern exposure, respectively; stabilized by a continuous topsoil biological crust. Strips 0.6–1.2 m wide
242
G.J. Kidron, A. Yair
Table 17.1 Main characteristics of runoff plots Plot number
Plot characteristics
Subdivided large plots N1.1, N2.1, S1.1, S3.1 No crust, < 5% vegetation N1.2, N2.2, S1.2, S3.2 40–75% patchy crust, 10–30% vegetation N1.3, N2.3, S1.3, S3.3 > 95% crust, 10–30% vegetation Whole-slope plots N3, N4, S2, S4 0–97% crust, 10–40% vegetation C1,C4 > 95% crust, 10–20% vegetation
Plot location Dune crest, active dune Mid-dune, active dune Dune bottom, active dune Entire slope, active dune Entire slope, stabilized dune
separated the subunits to enable plot construction and passage without destroying the crusts within the plots. All other plots drained the entire slope length. Four plots and subplots (N2.3, N3, S2, and S3.3) were equipped with V-notch weirs, and pressure transducers connected to data loggers. All other plots were equipped with large containers. Where large runoff volumes were expected, a splitting device was used. The splitting device enabled the capture of 10% runoff water and suspended sediment. Runoff and sediment collected are considered representative of runoff water and sediment allowed to flow away. Rainfall was measured using an electronic tipping-bucket rain recorder (Texas Electronics, USA; accuracy 0.1 mm), connected to a CR-10 Campbell data logger. Runoff and sediment were measured following each rainstorm. Sediments collected were oven-dried at 105 °C until reaching a constant weight, and then weighed. A representative sample of approximately 30 g was wet-sieved with 0.5% sodium hexametaphosphate (to ensure clay separation) through a 62-µm mesh, and the amounts of sand, silt, and clay were determined. Rainfall, runoff, and sediment yield measurements were carried out during 1990–1994. Owing to differences in plot dimensions, as well as to differences in surface characteristics along the slope, the unit chosen herein for the comparison of runoff from large plots is based on runoff volume per plot width. Likewise, sediment yield will be also presented per plot width, following Rutin (1983). This presentation enables comparison without any a priori knowledge of the actual plot length, and area involved in runoff and sediment contribution. Needless to say, differences in surface properties are more pronounced in large than in medium-sized and small plots.
17.3
Results
Annual precipitation exhibits a high variability (46.9–131.4 mm), with a limited number of rainstorms (10–20). Most storms were small, over 60% yielding less than 5 mm (Fig. 17.4A). High rain intensities of over 30 mm h−1 lasted for up to 8 minutes.
17 Runoff and Erosion Processes Within a Dune System
243
Fig. 17.4 Frequency distributions of rainstorm depths (A) and rain intensities (B) during the period 1989–1994
The maximum rain intensity recorded was 72 mm h−1, lasting for 1 minute only. Medium and high rain intensities of ≥ 9 mm h−1 accounted for approximately 20% of the precipitation (Fig. 17.4B). The storms were characterized by intermittent rain spells of variable intensities (Fig. 17.5). Runoff generation was intermittent, with 1–8 independent flows during a single rainstorm. Whereas medium and high rain intensities were capable of runoff generation, low rain intensities of < 9 mm h−1 were not (Fig. 17.5). Nevertheless, not all medium- and high-intensity rain spells resulted in runoff. Runoff was not generated during the onset of most rainstorms, during which the infiltration capacity of the dry crusts usually exceeded medium and high rain intensities. High infiltration rates under dry surface conditions are indicative of the fact that the biological topsoil crusts in the study area are not water-repellent. Rather, runoff generation resulted from pore clogging (Kidron et al. 1999). Infiltration was impeded as a result of sheaths and slime swelling following water absorption (Verrecchia et al. 1995;
244
G.J. Kidron, A. Yair
Fig. 17.5 Rainstorm of 1–3 January 1992
Mazor et al. 1996; Kidron and Yair 1997). Infiltration, and consequently runoff were not uniform throughout the area, and showed high spatial and temporal variability (Chap. 18, this volume). Runoff, and consequently sediment flow were not generated at the mobile crest devoid of crust. Runoff and sediment flow were, however, generated at the mid- and foot-slope plots. Runoff volumes and sediment yields recorded at the mid- and foot-slope subplots of north- and south-facing plots are shown in Fig. 17.6. In both cases, runoff and sediment yield were significantly higher at the foot-slope than at the mid-slope plots (paired t-test; p < 0.05). The absence of runoff from the mobile dune sections, the low amounts obtained from the mid-slopes, and the much higher amounts obtained from the foot-slope plots clearly indicate that runoff yield, and consequently sediment yield are controlled by the crust cover (Yair 1990; Kidron and Yair 1997; Yair 2001). The runoff volumes recorded were significantly higher at north-facing than at south-facing plots (Fig. 17.7). On average, the north-facing semi-active and stabilized dune sections yielded 16.1 and 15.7 l m−1, respectively, compared to 5.0 l m−1 for the south-facing slope (Kidron 1999). This is due to the better development and spatial continuity of the biological crust at the former than at the latter plots. Plot N1.3 generated the highest runoff, explained by its very smooth surface, and relatively low vegetation cover. For sediment yields, the trends are similar to those recorded for runoff. The highest amounts of sediments were obtained at the north-facing plots. This result may be explained by the higher capacity of runoff to carry sediment. Plot N3 often yielded the highest values (Fig. 17.8). This plot drains a whole plot in which the upper part is characterized by a blowout. Loose sand particles, blown by the prevalent southwesterly winter winds, were deposited over the crusted area, and washed out by runoff into the sediment collector. Average annual sediment yield
17 Runoff and Erosion Processes Within a Dune System
245
Fig. 17.6 Average annual runoff volumes at mid- and foot-slope plots of north- and south-facing slopes
Fig. 17.7 Average annual runoff volumes for all plots (1990–1994)
on the active and stabilized north-facing slopes were 795 and 431 g m−1, respectively, compared to 82 g m−1 on the south-facing slopes (Kidron and Yair 2001). This process, combined with the limited runoff characteristic of the mid-slope sections, explains why higher sediment concentrations were obtained at the mid-slope
246
G.J. Kidron, A. Yair
Fig. 17.8 Relationships between sediment yield and plot area
than at the foot-slope plots (Fig. 17.9). Sediment concentrations obtained in the mid-slope section sometimes reached the value of 500 g l−1. This is due to the fact that a higher proportion of the sediment collected at the mid-slope plots is composed of heavy sand particles; while sediment collected at the foot-slope plots is richer in fines (up to 30%) derived from the topsoil biological crust, rich in finegrained particles.
17.4
Discussion
The notion that the development of arid sand dunes is solely a result of eolian activity is widespread in the literature. Many researchers have assumed that runoff does not take place in arid sand dunes, and consequently no erosion or sedimentation by overland flow can be expected. However, microbiotic crusts are widespread in many arid and semiarid parts of the world (see West 1990, and Belnap and Lange 2001 for reviews). Once present, the crusts may significantly alter the hydrological behavior of the surface. Data obtained in the Nizzana area show that runoff yield, and consequently sediment yield are positively linked to crust cover and biomass,
17 Runoff and Erosion Processes Within a Dune System
247
Fig. 17.9 Average sediment concentrations at a north- and b south-facing plots
which in turn are linked to exposure and topography (Yair 1990; Kidron and Yair 1997, 2001). Whereas the crest of the active dunes lacked microbiotic crusts, and as a result did not generate runoff, intermediate runoff and sediment yields characterized the mid-slope sections having a thin and patchy crust cover. Furthermore, differences in crust biomass and spatial continuity are responsible for the differential behavior of north- and south-facing slopes. Runoff and sediment yields are higher on the former. North-facing crusts are indeed characterized by 2 to 3 times higher chlorophyll a contents than are south-facing crusts (Kidron 1995). Higher crust biomass, characterized by a dense network of sheaths and filaments, promotes pore clogging responsible for higher runoff yield on the bottom north-facing slopes. The larger runoff amounts, and the supply of loose sand by the prevailing southwesterly winds may also explain the higher average annual sediment yield on the north-facing slopes, this being on average one order of magnitude higher than that of the southfacing plots. Owing to the abrupt change in slope angle in the interface dune–interdune section, most runoff, and hence sediments will concentrate along a narrow (usually
248
G.J. Kidron, A. Yair
3–5 m) belt in the north-facing foot-slope area. In many cases, runoff and sediments will be concentrated within small local depressions usually 1–4 m large, characterized by a dense vegetation cover. The dune–interdune interface along the north-facing foot slopes can be regarded as the most mesic habitat within the dune field. It is inhabited by a moss-dominated crust. This crust has a chlorophyll a content of 50–60 mg m−2, compared to only 30–40 mg m−2 for the other northfacing slopes, and to only 15–20 mg m−2 for the south-facing slopes and interdune corridors. It also has the highest variety of cyanobacteria and green algae (Kidron et al. 2000).
17.5
Conclusions
The notion that runoff generation, and hence runoff-transported sediments do not occur in arid dune fields is not supported at the Hallamish dune field. The complex interrelationships between abiotic and biotic factors, and between eolian and fluvial factors that contribute to the development of the crust, and to its role within the Hallamish dune-field ecosystem can be described as follows: 1. Erosion and deposition by wind may form and alter the topography, creating longitudinal dunes and blowouts. The wind may also affect the establishment and cover of microbiotic crusts, preventing crust establishment in places subjected to high erosion or deposition, and facilitating crust establishment in areas having a relatively low eolian activity. Wind activity, topography, and slope aspect will affect the length of time during which the surface is wet, and consequently will affect crust development, infiltration rates, and runoff generation. 2. Once established, the crust may impede wind erosion. The crust may also alter the hydrological properties of the surface, affecting water redistribution and wetness duration. 3. Runoff will in turn affect crust and plant growth, as well as sediment production. Sediment production will also be influenced by topography, i.e., slope angle, and the presence of blowouts. 4. Erosion and sedimentation by runoff will alter the topography, resulting in local sedimentation along the foot slopes. Although runoff and erosion may be very low in this case, compared to that of sand dunes in wet climates (Rutin 1983), or of other types of surfaces in the Negev desert (Yair 1974), their role in shaping the dune-field ecosystem cannot not be ignored. Both runoff and sediment yield will affect plant density and biomass, as well as animal distribution. Acknowledgements The research was supported by the Arid Ecosystem Research Centre (AERC) of the Hebrew University of Jerusalem and the MINERVA foundation, and by a grant from DISUM. We would like to thank E. Sachs for his assistance in the field, M. Kidron for the drawings, and C.A. Kidron for editing the manuscript.
17 Runoff and Erosion Processes Within a Dune System
249
References Belnap J, Lange OL (eds) (2001) Biological soil crusts: structure, function, and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York Danin A (1978) Plant species diversity and plant succession in a sandy area in the Northern Negev. Flora 167:409–422 Hillel D, Tadmor N (1962) Water regime and vegetation in the Central Negev highlands of Israel. Ecology 43:33–41 Kidron GJ (1995) The impact of microbial crust upon rainfall-runoff-sediment yield relationships on longitudinal dune slopes, Nizzana, western Negev Desert, Israel (in Hebrew with English summary). PhD Thesis, The Hebrew University of Jerusalem Kidron GJ (1999) Differential water distribution over dune slopes as affected by slope position and microbiotic crust, Negev Desert, Israel. Hydrol Processes 13:1665–1682 Kidron GJ, Yair A (1997) Rainfall-runoff relationships over encrusted dune surfaces, Nizzana, western Negev, Israel. Earth Surface Processes Landforms 22:1169–1184 Kidron GJ, Yair A (2001) Runoff-induced sediment yield over dune slopes in the Negev Desert. 1. Quantity and variability. Earth Surface Processes Landforms 26:461–474 Kidron GJ, Yaalon DH, Vonshak A (1999) Two causes for runoff initiation on microbiotic crusts: hydrophobicity and pore clogging. Soil Sci 164:18–27 Kidron GJ, Barzilay E, Sachs E (2000) Microclimate control upon sand microbiotic crust, western Negev Desert, Israel. Geomorphology 36:1–18 Mazor G, Kidron GJ, Vonshak A, Abeliovich A (1996) The role of cyanobacterial exopolysaccharides in structuring desert microbial crusts. FEMS Microbiol Ecol 21:121–130 Rutin J (1983) Erosional processes on a coastal sand dune, De Blink, Noordwijkerhout. Physical Geography and Soils Laboratory, University of Amsterdam, Publ no 35, pp 1–144 Tsoar H (1990) The ecological background, deterioration and reclamation of desert dune sand. Agric Ecosystem Environ 33:147–170 Tsoar H, Møller JT (1986) The role of vegetation in the formation of linear sand dunes. In: Nickling WG (ed) Aeolian geomorphology. Proc 17th Annual Binghamton Geomorphology Symp, Allen and Unwin, Boston, MA, pp 75–95 Tsoar H, Zohar Y (1985) Desert dune sand and its potential for modern agricultural development. In: Gradus Y (ed) Desert development. Reidel, Boston, MA, pp 184–200 Verrecchia E, Yair A, Kidron GJ, Verrecchia K (1995) Physical properties of the psammophile cryptogamic crust and their consequences to the water balance of sandy soils, Northwestern Negev Desert, Israel. J Arid Environ 29:427–437 West NE (1990) Structure and function of microphytic soil crusts in wildland ecosystems of arid to semi-arid regions. Adv Ecol Res 20:180–223 Yair A (1974) Sources of runoff and sediment supplied by the slopes of a first order drainage basin in an arid environment. Report on present day geomorphological processes. Abhandlungen der Akademie der Wissenshaften Göttingen, Mathematische-Physikalische Klasse no 29, pp 403–407 Yair A (1990) Runoff generation in a sandy area – the Nizzana sands, Western Negev, Israel. Earth Surface Processes Landforms 15:597–609 Yair A (2001) Effects of biological soil crusts on water redistribution in the Negev desert, Israel. A case study in longitudinal dunes. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function, and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York, pp 303–314
Chapter 18
Effects of Surface Runoff and Subsurface Flow on the Spatial Variability of Water Resources in Longitudinal Dunes A. Yair
18.1
Introduction
The relative textural homogeneity of sand particles in desert dunes, coupled with their high porosity and low water-holding capacity, have led to the widespread idea that water movement in unsaturated sand is fast and predominantly vertical. Under such conditions, one would expect a quite uniform depth of water penetration, of soil moisture and vegetation cover in dune areas. However, several studies have cast doubt on the vertical water movement in dunes. Miles et al. (1988) showed that slight differences in the compaction of texturally homogeneous sand layers affected the hydraulic conductivity and controlled water movement under unsaturated conditions. Yeh and Harvey (1990) drew attention to heterogeneities between layers as factors controlling the direction of flow. Zaslavsky and Sinai (1981) suggest that lateral flow can be expected over a sloping surface, even in the absence of an impeding layer at a shallow depth. Water concentrates especially in slope concavities. Dunes are characterized by a layered cross-bedded structure and lateral changes in slope angles. In view of the arguments presented above, a lateral flow component would be expected for steep dunes, especially if slight differences in compaction and texture exist between adjoining layers. Such a phenomenon should lead to a water redistribution process of infiltrated waters, a process by which water infiltrated in the upper part of a dune ridge moves laterally towards the base of the dune slope at a shallow depth. Local slope concavities would be expected to increase the effect of local water concentration. The actual existence of a lateral flow component in dune areas under field conditions was demonstrated in several studies conducted in the desert of New Mexico (USA), where average annual rainfall is 220 mm (Stephens and Knowlton 1986; McCord and Stephens 1987; Stephens 1994). Similar results have been reported in a study conducted in the Nizzana sand field in Israel (Yair et al. 1997). Lateral water movement at a shallow depth is supported by field observations. Vascular plant cover in the Nizzana area increases from the upper to the lower part of the dune slope, with high concentrations in slope concavities, especially at the base of steep dune ridges (Fig. 18.1). The discussion above refers to the process of water redistribution in dunes as related to water movement following water infiltration. However, water redistribution
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
251
252
A. Yair
Fig. 18.1 Dense vegetation cover at the dune base
in sandy arid areas may be caused also by runoff generation over the dune surface. On many arid and semi-arid surfaces, vegetation cover is sparse but the space between the shrubs is not bare (Belnap et al. 2001). This space is often covered by a topsoil biological crust composed of a community of many organisms (such as cyanobacteria, algae, mosses, lichens). Some of these organisms excrete exopolysaccharides which bind the fine-grained inorganic particles forming the matrix. The important role of this tangled mesh is well known in the case of nitrogen and carbon fixation, germination, as well as the stabilization of dune surfaces. By contrast, less is known about the effects on infiltration, runoff, water redistribution and soil moisture. Topsoil biological crusts can affect water redistribution in two different ways. The first is via hydrophobicity, which seals the surface and thereby prevents rainwater infiltration, resulting in runoff generation on sloping surfaces is long as hydrophobicity persists (Bond and Harris 1964; Roberts and Carson 1971; Burch et al. 1989; Dekker and Jungerius 1990; Wessell 1998). The second is via pore clogging. Unlike the main sand body, almost completely devoid of fine-grained particles, the topsoil crust is relatively rich in silt and clay (20–40%). The presence of fines reduces the total porosity and pore size (Verrecchia et al. 1995) of the topsoil crust, thereby limiting its infiltration rate. Porosity and infiltration rate are further reduced by clogging of the pores in the topsoil layer, caused by the combined swelling of microorganisms and fine-grained crust particles when wetted (Avnimelech and Nevo 1964; Campbell 1979; Wang et al. 1981; Mazor et al. 1996).
18 Effects of Surface Runoff and Subsurface Flow
18.2
253
Aim of Study
A biological topsoil crust partly covers the longitudinal dunes in the Nizzana area (Fig. 18.2). The crust is absent on the active crest of the sandy ridges. It is better developed on north than on south-facing slopes (Chap. 10, this volume). The extent and spatial continuity of the crust increase from the upper to the lower part of the dune slope (Kidron 1995; Kidron and Yair 1997). The crust contains up to 50% silt and clay, compared to only 3–7% in the underlying sand (Yair 1990). The principal aim of the present chapter is to present data dealing with factors affecting the non-uniform spatial distribution of water resources in the longitudinal dunes of the Nizzana dune field, complementing the experiments performed at the site (see Chap. 17, this volume).
18.3 18.3.1
Results of Field Studies Rainfall
The monitoring period covered 4 rainy years (Table 18.1). Above-average rain fell in the first 2 years (∼120 and 131 mm respectively), while the following years were normal to dry (∼85 and 47 mm respectively). In all, 70% of the rainstorms recorded during 1990–1994 had rain amounts below 5 mm, and 85% had rain amounts below
Fig. 18.2 View of the biological topsoil crust
254
A. Yair
Table 18.1 Distribution of rain intensities (1990–1994) Max. intensity (mm h−1) for 1-minute duration Annual rain amount Rain amount ≤ 6 mm h−1 6–12 mm h−1 12–18 mm h−1 18–24 mm h−1 24–30 mm h−1 ≥ 30 mm h−1 Total rain amount ≥12 mm h−1
Rainfall year 1990–1991 1991–1992 119.7 131.4 83.1 101.5 21.6 20.5 2.1 2.9 6.6 2.0 1/2 2.0 5.1 2.5 15.0 9.4
1992–1993 85.3 65.0 10.0 1.0 0.4 2.0 6.9 10.3
1993–1994 46.9 28.1 10.1 3.9 1.6 0.1 2.2 7.8
15 mm. However, extreme rainstorms in the range of 38–50 mm were recorded in the first years. All rainstorms were of intermittent nature. Rain intensities were generally low: 70–85% of annual rains fell at intensities below 12 mm h−1. Annual rain amounts at intensities higher than 12 mm h−1 varied from 7 to 14 mm. These amounts are low, especially if one considers that they were recorded at different rainstorms or at several bursts during a given rainstorm.
18.3.2
Rainfall–Runoff Relationships
The data obtained highlight the role of the biological crust in runoff generation. No runoff was collected from the plots draining the upper and middle parts of the slopes, where crusts are absent. Negligible runoff was collected from plots draining the mid-slope, where crusts are patchy, thin and often covered by sand during sandstorms. Runoff generation was limited to the crusted and densely vegetated areas, located on the lower flanks of the north- and south-facing slopes. The runoff data collected are presented in Table 18.2. One to five annual rainstorms yielded runoff during the research period. As expected, runoff frequency and magnitude were higher in the rainy years (1990–1991 and 1991–1992) than in the two following dry years. Runoff frequency and magnitude were higher on the north- than on the south-facing slopes. Typical hydrographs recorded are presented in Fig. 18.3. Data analysis will focus on three issues: (1) the conditions for runoff generation, (2) the extent of the runoff-generating area and (3) implications for the spatial redistribution of runoff.
18.3.2.1
Conditions for Runoff Generation
The topsoil biological crust in the Nizzana area did no show any signs of water repellence. Water repellence is a temporary property occurring under dry surface conditions, and disappears gradually upon wetting. Runoff never developed during
Rainfall year
Date
(mm)
per minute
Plot N2
Plot N3
Plot S2
Plot S3
Rain ≥ 12 mm h−1
1990–1991 (119.7 mm)
24–26 Jan. 1991 30 Jan. 1991 7–8 Feb. 1991 5–6 Mar. 1991 22–23 Mar. 1991
38.5 2.9 7.1 22.3 38.8
36.0 12.0 24.0 18.0 72.0
1–3 Jan. 1992 30 Jan.–2 Feb. 1992 6–11 Feb. 1992 17 Feb. 1992 24–26 Feb. 1992
49.3 9.8 35.5 2.1 13.3
42.0 18.0 12.0 48.0 18.0
11–13 Jan. 1993 12 May 1993
24.2 9.7
12.0 72.0
21–23 Dec. 1993
20.7
54.0
10.8 10.7 3.8 4.8 243.4 273.5 40.0 5.0 68.6 61.9 90.8 267.3 3.5 23.5 27.0 35.1 35.1
18.4 3.9 10.8 8.7 178.3 220.1 29.6 3.5 64.0 46.9 40.0* 184.0 0.8 8.4 9.2 3.8 3.8
0 0 0.7 3.1 38.8 42.6 1.4 0 9.4 23.5 26.0 60.3 1.7 33.1 34.8 0 0
0 0 1.8 3.7 74.4 79.9 36.5 0 36.2 37.3 36.2 146.2 28.3 51.3 79.6 12.0 12.0
1.7 0 1.5 0.6 10.3 14.1 3.4 0.8 0.2 1.7 0.3 6.4 0.1 8.6 8.7 7.1 7.1
Annual runoff volume (l) 1991–1992 (131.4 mm)
Annual runoff volume (l) 1992–1993 (85.3 mm) Annual runoff volume (l) 1993–1994 (46.9 mm) Annual runoff volume (l)
18 Effects of Surface Runoff and Subsurface Flow
Table 18.2 Rainfall–runoff relationships, 1990–1994: N north-facing slope, S south-facing slope, asterisk minimum value, due to overflow Runoff volume (l) Rain I. max. (mm h−1
255
256
A. Yair
Fig. 18.3 Typical hydrographs at plots N2 and N3 during stormy weather: a 24–26 January 1991, b 9–11 February 1992
the first rainstorms of the rainy season, recorded after a long, hot and dry summer. Most runoff events occurred in January–March when the surface was wet. Furthermore, the hydrographs presented in Fig. 18.3 show that rain intensities in
18 Effects of Surface Runoff and Subsurface Flow
257
Fig. 18.4 Pore size distribution of the topsoil crust and of mobile sand
excess of ∼12 mm h−1 at the beginning of a rainstorm, when the surface is relatively dry, are not sufficient for runoff generation. Rather, runoff begins when such intensity occurs after the topsoil crust is wet, in the middle or at the end of the storm. This also implies that surface sealing due to raindrop impact does not take place in the sandy area. The cohesive and flexible biological elements of the crust absorb raindrop energy and prevent the rapid development of a rain crust conducive to runoff generation. The biological topsoil crust in the Nizzana area behaves as a normal soil. Once wetted, infiltration decreases with time, due to pore clogging caused by the swelling of fine-grained soil particles and cyanobacterial sheath material. Campbell (1979) and Wang et al. (1981) report that cyanobacterial sheaths may absorb up to 12 times their dry weight, increasing their volume up to 10 times. Once the crust is saturated, runoff occurs when rain intensities are higher than ∼10–12 mm h−1. The relatively high final infiltration rate of the crust can be explained in terms of pore size distribution (Fig. 18.4). Swelling of the biological elements and of the fine-grained particles is sufficient to fill up most of the small voids, limiting water infiltration. However, water can still move along the larger pores, which remain unclogged.
18.3.2.2
The Extent of the Runoff-Contributing Area
Figure 18.5 shows that the runoff volumes collected cannot be explained by total storm rain amount or by storm rain amount in excess of ∼12 mm h−1 (considered as the final infiltration rate of the topsoil crust). Furthermore, the runoff volumes
258
Fig. 18.5 Relationships between storm rainfall and runoff
A. Yair
18 Effects of Surface Runoff and Subsurface Flow
259
collected are not positively correlated with drained area. Runoff per unit area is highest at the mini-plots (which drain an area smaller than 4 m2) than on the larger plots. Also, no correlation exists between the encrusted area of the large plots and specific runoff (Table 18.3). Such results point to a phenomenon of partial area contribution, i.e. areas contributing to runoff were probably limited to the lower parts of the encrusted plots. The limitation of the contributing area to the lower part of the encrusted slope can be attributed to differences in infiltration rates within the large plots. It has already been mentioned that the thickness and spatial continuity of the crust decrease upslope, enhancing infiltration rate in the upper part of the encrusted slope. A thin, patchy crust with local sandy depressions would impede the process of flow continuity in the downslope direction. Such losses are minimized over a small area, where spatial variability of crust properties is low, thus explaining their high runoff rate per unit area. Another factor limiting the extent of the contributing areas is related to rainfall properties. As shown on Fig. 18.6, rainstorms in the area are characterized by a pronounced intermittent pattern, coupled with high temporal variability in rain intensity. In addition, annual rain amount at an intensity in excess of ∼12 mm h−1, capable of producing runoff under wet surface conditions, is quite low (Table 18.1); indeed, we should keep in mind that these amounts were recorded during several storms or different rain bursts during a given rainstorm. What is more important is that the duration of the effective rain bursts is usually extremely short, as clearly
Table 18.3 Runoff for encrusted areas Rainfall year Total rain (mm) Runoff (l m−2) Plots with a recording device Plot name Encrusted area (m2) 1990–1991 119.7 1991–1992 131.4 1992–1993 85.3 1993–1994 46.9 Plots with a runoff collector
N3 478 0.50 0.40 0.02 0.04
N2.2 307 0.91 0.91 0.07 0.11
S2 411 0.05 0.07 0.07 0.01
S3.3 198 0.25 0.76 0.33 0.05
Plot name Encrusted area (m2) 1990–1991 1991–1992 1992–1993 1993–1994 Mini-plots
119.7 131.4 85.3 46.9
N4 1,093 0.71 0.60 0.11 0.03
N1.3 74 7.03 7.70 3.11 1.22
S4 799 0.25 0.50 0.16 0.01
S1.4 30 0.16 0.16 0.82 0.49
119.7 131.4 85.3 46.9
NM 3.6 12.80 9.00 7.00 3.20
SM 3.9 1.95 2.30 1.80 0.82
Plot name Encrusted area (m2) 1990–1991 1991–1992 1992–1993 1993–1994
C1 188 0.90 1.86 0.05 0.02
C4 238 0.92 2.20 0.15 0.02
260
A. Yair
Fig. 18.6 Rain storm of 22–23 March 1991 (plot 2)
indicated by the very steep rising and falling limbs of hydrographs recorded during the most extreme rain events (Fig. 18.6). Once effective rainfall stops, runoff generated on the upper encrusted slope section has enough time to infiltrate before reaching the slope base. In other words, given the high final infiltration rate, the concentration time required for a continuous flow from the upper to the lower part of the encrusted area is longer than the short duration of most effective rain bursts. Infiltration loss, in a downslope direction, is enhanced by the uneven micro-topography and by shrubs acting as water sinks. One can obtain a good idea of the actual contributing area in the large plots by analyzing the ratio between peak flow rate and peak rain intensity in a given storm. For example, peak flow during the most extreme runoff event, in terms of rain intensity and rain-shower duration (Fig. 18.6) at the most responsive plot (N2), was only 14.2% of the causative rain burst (with an intensity of 72 mm h−1 lasting for 1 minute). This rain burst occurred at the very end of a wet rainy season and at the end of the most extreme storm recorded during the study period (Table 18.2), when the biological crust had already reached saturation conditions. With a final infiltration rate of ∼12 mm h−1, and assuming that the whole crusted area of the plot had contributed runoff, peak flow values should have been 1.19 mm. The value recorded, however, was only 0.17 mm, indicating that only part of the drained area actually contributed to runoff. The contributing area at peak flow amounted to about 40–50 m2, with an estimated slope length of 7–10 m (Yair 2001). The very steep rising and falling limbs of all hydrographs recorded indicate that most runoff is provided during the short peak flow, supporting the idea that only the lower portions of the plots were contributing runoff.
18 Effects of Surface Runoff and Subsurface Flow
18.3.2.3
261
Implications for Water Redistribution
The data collected show that runoff generation in the study area plays an important role in water redistribution over the encrusted area. This area is composed of runoffcontributing bare areas and run-on vegetated patches. Each shrub receives runoff from a limited contributing area extending for a short distance upslope. As each shrub can be regarded as a sink interrupting the continuity of downslope water flow, the extent of the potential contributing area for each shrub is determined by the location of the neighbouring shrubs. In view of the short flow distances and short flow duration, data collected at the mini-plots provide the best assessment of water yield for estimating runoff water redistribution in the crusted surfaces. Accounting for the density of the shrubs in the runoff-generating areas, each shrub has a potential contributing area of 4–6 m2. Using the mini-plot data (Table 18.3), each of the shrubs growing on the north-facing slopes would have, in rainy years, on the order of 40–80 Liters of water available for collection. This amount is approximately equivalent to 40–90% of average annual rainfall, but with a much higher efficiency, as water concentration by runoff enables deeper infiltration and better water preservation. Lower amounts would be expected in drier years. The importance of runoff as a source of soil moisture is more pronounced on north- than on south-facing slopes, due to the fact that the frequency and magnitude of runoff events are more limited on the latter slopes.
18.4
The Effect of Subsurface Water Movement on Water Redistribution
The depth of water percolation, water movement, and sand water content were measured with a neutron probe. Sixteen boreholes were dug to a depth of 6 m. Their locations are shown on Fig. 18.7. Water content measurements were conducted during two consecutive rainy seasons (1991–1993), shortly after each storm and between storms, at vertical intervals of 30 cm. In addition, sand samples from the
Fig. 18.7 Locations of the neutron access tubes
262
A. Yair
boreholes were analyzed for their electrical conductivity and particle size composition. Values of the latter variables are regarded as indicative of long-term water movement and processes.
18.4.1
Results
Rain amount in the first year was above average (131 mm), whereas in the following year rain was slightly below average (85.5 mm). Figures 18.8 and 18.9 display the results obtained during the first rainy season. Figure 18.10 deals with two distinct environments representing the relatively flat interdune area, where water percolation is vertical. Borehole K51 is located in a playa surface, characterized by compacted fine-grained deposits of the Nizzana channel (Yair 1990; Harrison and Yair 1998). Water penetration depth was limited to 30–60 cm. This shallow depth is attributed to the low porosity and high water-holding capacity of the fine-grained material. Borehole K55 represents a sandy environment in the interdune area. This borehole shows a higher sensitivity to the temporal variations in rainfall. Maximum water penetration depth is 90 cm. Following a rainy period in January 1992, soil water content was 8.5%, a value close to the field capacity of the sand. A rapid drying process was observed towards the end of the rainy season. A completely different water movement pattern was detected along the steep dune slopes (Fig. 18.9). Borehole K04 is located at the middle of the slope. Depth of water infiltration here is down to 200 cm, much deeper than in borehole K55 located in the sandy interdune area. Two peaks of water content can be observed during most of the season. The two peaks are far more pronounced in borehole K12, where the depth of water penetration reached 420 cm. This borehole is located at a short distance downslope, below a local topographic depression. Water which concentrates in the depression flows laterally, rather than vertically. The actual existence of subsurface water flow at a shallow depth, parallel to the sloping surface, was also detected by Kutsishin (2002) who used the same boreholes and a neutron probe device within the framework of a study of groundwater recharge in 1998–1999. He showed the occurrence of local water lenses, parallel to the steep slope at depths of 3–4 m. The response of the area differed completely in the second year (Fig. 18.10). Annual rainfall was slightly below average (85.5 mm), with numerous small rainstorms. Water percolation depth on the dune ridge was limited to 60–90 cm, being slightly higher in the upper slope section (borehole K02) devoid of a microphytic crust than in the lower encrusted slope segment (borehole K05). Depth of water percolation over the playa borehole (K51) was limited to 30–40 cm. Subsurface lateral flow was not detected at any of the boreholes. The long-term effect of subsurface lateral water flow is supported by data on changes in electrical conductivity and amount of fine-grained material at the base of the dune slopes. Wedges of increased salinity and content of fine-grained particles, parallel to the sloping surfaces, were detected at the slope bases (Fig. 18.11).
18 Effects of Surface Runoff and Subsurface Flow
Fig. 18.8 Temporal variations in water content: interdune area (1991–1992)
263
264
Fig. 18.9 Temporal variations in water content: dune slope (1991–1992)
A. Yair
18 Effects of Surface Runoff and Subsurface Flow
Fig. 18.10 Water contents at selected boreholes (1992–1993)
265
266
A. Yair
Fig. 18.11 Saline wedges at the base of dune slopes
18.4.2
Conditions for Subsurface Lateral Water Flow
Studies dealing with groundwater recharge have shown that annual rainfall and potential evaporation are not suitable for estimating groundwater recharge rate in sandy arid areas (Nixon and Lawless 1960; Stephens and Knowlton 1986; Nichols 1987). All these authors stress the importance of short rainy periods which result in deep downward pulses of water movement. The data presented here are in full agreement with the pulse, or piston hypothesis. The rainy season of 1991–1992 was one of the wettest on record. An extreme rainstorm recorded in January 1992 (49.3 mm) did not develop a lateral flow. However, when a second cold and rainy period occurred in early February (altogether, 114.5 mm), favourable conditions for deep percolation developed. The new infiltrated water probably displaced part of the moisture accumulated in January. The lack of a prolonged rainy period, together with the limited rain amounts recorded during most storms are considered responsible for the lack of subsurface flow in 1992–1993.
18.5
Conclusions
Arid sandy areas are usually perceived as being characterized by very high infiltration rates, in excess of prevailing storm rain amounts and rain intensities. Data collected at the Nizzana research station highlight the complexity of the processes involved in the spatial distribution of water resources in a system of longitudinal dunes. Although the data for a short period of 2–4 years is insufficient to evaluate the water regime over a longer period of time, they shed light on the processes and trends of water
18 Effects of Surface Runoff and Subsurface Flow
267
movement and spatial water redistribution prevailing in the study area. The data obtained point to high spatial variability of water resources over short distances. Two main factors are responsible for the non-uniform distribution of water resources. The first is the occurrence of runoff, and the second a process of subsurface lateral flow parallel to the dune slope. The importance of these factors is more important in rainy than in dry years. The importance of runoff is limited to the lower part of the dune slopes where the topsoil biological crust is best developed and contiguous. On the north-facing slopes, runoff may increase water input to plants by up to 80%, compared to direct rainfall. Runoff contribution is more limited on south-facing slopes where, due to limited crust development, the frequency and magnitude of runoff events are low. The dune base is also a site receiving water by subsurface flow. The dune base therefore represents the best water regime, as it receives water from three sources: direct rainfall, runoff, and subsurface flow. This site is further favoured by a high concentration of nutrients related to the high content of fine-grained particles, and to nitrogen fixation by the well-developed biological topsoil crust. This is why a dense, narrow vegetated belt is observed at such sites. Subsurface lateral flow seems to be important at additional sites along the steep dune slopes, especially at slope concavities close to the dune crest. In the absence of a topsoil biological crust, runoff does not develop here and all rainwater infiltrates. Such sites represent the second-best site in terms of water resources. Depth of water infiltration here is high due to the loose structure of the sand in the vicinity of the active crest, and lateral water movement concentrates water in local concavities. However, the availability of nutrients is limited. At the crest, despite deep water infiltration, surface instability (due to strong wind activity) introduces a factor which limits water use (Kadmon and Leschner 1995). The third site in terms of water resources is represented by sandy areas within the interdune corridor. Depth of water infiltration here is more limited than in the sandy areas along the dune crests, for two reasons. The first is the occurrence of a thin compacted, topsoil mineral crust which limits infiltration. The second is that the sand underlying the topsoil crust is far more compacted than the sand of the dune ridges. In terms of water regime, the worst site is represented by the playas surfaces, specific to the Hallamish sandy area and located south of the Nizzana channel. The very high contents of compacted fine-grained sediments, with up to 80% silt and clay, seriously limit infiltration depth. An additional factor which constrains or even eliminates vegetation growth on the playa surfaces is the high salinity of these sediments (Harrison and Yair 1998). Acknowledgements The studies reported here were supported by the Arid Ecosystems Research Centre of the Hebrew University of Jerusalem, and a research grant by DISUM (German-Israeli Research Program). Financial support for the study of subsurface water movement was provided by the Forestry Service of the Jewish National Fund. I am grateful to Mrs. M. Kidron, Department of Geography, for drawing the illustrations.
268
A. Yair
References Avnimelech Y, Nevo Z (1964) Biological clogging of sands. Soil Sci 98:222–226 Belnap J, Burghard B, Lange OL (2001) Biological soil crusts: characteristics and distribution. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York, pp 3–31 Bond RD, Harris JR (1964) The influence of the microflora on physical properties of sand. Effects associated with filamentous algae and fungi. Austr J Soil Res 2:111–122 Burch GJ, Moore DI, Burns J (1989) Soil hydrophobic effects on infiltration and catchment runoff. Hydrol Processes 3:211–222 Campbell SE (1979) Soil stabilization by prokaryotic desert crust. Implication for Precambrian land biota. Origins Life 9:335–348 Danin A (1996) Plants of desert dunes. Springer, Berlin Heidelberg New York Dekker LW, Jungerius PD (1990) Water repellency in the dunes with special reference to the Netherlands dunes of the European coasts. Catena suppl 18:173–183 Harrison JBJ, Yair A (1998) Late Pleistocene aeolian and fluvial interactions in the development of the Nizzana dune field, Negev Desert, Israel. Sedimentology 45:307–518 Kadmon R, Leschner H (1995) Ecology of linear dunes: effect of surface stability on the distribution and abundance of annual plants. Adv GeoEcol 28:125–143 Kidron GJ (1995) The impact of a microbial crust upon rainfall-runoff-sediment yield relationship on longitudinal dune slopes, Nizzana, Western Negev Desert (in Hebrew). PhD Thesis, The Hebrew University, Jerusalem Kidron GJ, Yair A (1997) Rainfall-runoff relationships over encrusted dune surfaces, Nizzana, Western Negev, Israel. Earth Surface Processes Landforms 22:169–1184 Kutsishin L (2002) Assessment of deep percolation processes in a sand dunes terrain. MSc Thesis, Ben Gurion University of the Negev Mazor G, Kidron GJ, Vonshak A, Abeliovitch A (1996) The role of cyanobacterial exopolysaccharides in structuring desert microbial crusts. FEMS Microbiol Ecol 21:121–130 McCord JT, Stephens DB (1987) Lateral moisture flow beneath a sandy hillslope without an apparent impeding layer. Hydrol Processes 1:225–238 Miles JC, Thomas HR, Abrishami J (1988) The effect of small density changes on the movement of water through an un-saturated sand. J Hydrol 104:93–110 Nichols WD (1987) Geohydrology of the unsaturated zone at the burial site for low level radioactive waste near Beaty, Nevada. US Geological Survey, Water Supply pap no 2312 Nixon PA, Lawless GP (1960) Detecting of deeply penetrating rain water with neutron scattering moisture meter. Trans Am Soc Agric Eng 3:5–6 Roberts FG, Carson BA (1971) Water repellence in sandy soils of southwestern Australia. Austr J Soil Sci 10:35–42 Stephens DB (1994) A perspective on diffuse natural recharge in areas of low precipitation. Soil Sci Soc Am J 58:40–48 Stephens DB, Knowlton R (1986) Soil water movement and recharge through sand at a semi-arid site in New Mexico. Water Resources Res 22:881–889 Verrecchia E, Yair A, Kidron GJ, Verrecchia K (1995) Physical properties of the psammophile cryptogamic crust and their consequences to the water regime of sandy soils, north western Negev, Israel. J Arid Environ 29:427–437 Wang F, Zung Z, Hu Z (1981) Nitrogen fixation by an edible terrestrial blue-green algae. In: Gibson AH, Newton E (eds) Current perspectives in nitrogen fixation. Elsevier, Amsterdam, p 455 Wessel AT (1998) On using the effective contact angle and the water drop penetration time for classification of water repellency in dune soils. Earth Surface Processes Landforms 13:555–561 Yair A (1990) Runoff generation in a sandy area – the Nizzana sands, Western Negev, Israel. Earth Surface Processes Landforms 15:597–609
18 Effects of Surface Runoff and Subsurface Flow
269
Yair A (2001) Effects of biological soil crusts on water redistribution in the Negev desert, Israel: a case study in longitudinal dunes. In: Belnap J, Lange OL (eds) Biological crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York, pp 303–314 Yair A, Lavee H, Greitser N (1997) Spatial and temporal variability of water percolation and movement in a system of longitudinal dunes, Western Negev, Israel. Hydrol Processes 11:43–58 Yeh TC, Harvey DJ (1990) Effective unsaturated hydraulic conductivity of layered sand. Water Resources Res 26:1271–1279 Zaslavsky D, Sinai G (1981) Surface hydrology: causes of lateral flow. Hydrol Division ASCE 107:37–52
Chapter 19
Atmospheric Input of Nutrient Elements and Dust into the Sand Dune Field of the North-Western Negev T. Littmann and A. Schultz
19.1
Introduction
In arid regions, primary productivity is limited mainly by the spatial and temporal availability of water, followed by soil salinity (West 1990) and the generally poor nutrient pool (Offer et al. 1992), except for singular cryptogamic (blue algae) or leguminous species (e.g. Retama raetam) which have the capability of atmospheric nitrogen fixation (Esser 1989; West 1990; Chap. 22, this volume). If lateral matter inputs, as in the case of fluvial processes, are not involved, then atmospheric deposition is the main diffuse source of nutrients and salts in an arid ecosystem. Point sources with a very local control of the nutrient pools may be litter accumulations beneath shrubs, or deflation hollows where organic matter is deposited by the wind (Kadmon and Leschner 1995). This overall situation is especially true for arid sand dune areas, such as the sand dune field of the north-western Negev, with 11 arid months (Fig. 2 in Introduction chapter, this volume). From an ecological point of view, sandy areas within an arid environment are favourable habitats because of less evaporative soil water losses, compared to rocky deserts or desert soils rich in clay. Sand dunes may enhance rainwater infiltration, show subsurface flow along the slopes after good rainy seasons (Yair et al. 1997, Chap. 18, this volume) and, thus, have a soil moisture reservoir which enables considerable establishment of higher vegetation capable of water uptake from greater depths of the unsaturated soil column (Veste and Breckle 1996). Even under such favourable conditions, however, high local soil salinity and extremely low nutrient availability may aggravate the patchy character of vegetation establishment, biomass production and surface stability. With regard to regional climatic change over shorter periods, the atmospheric contribution to local soil salinity is a key factor in understanding the climatic impact on arid ecosystems (Yair 1994). Thus, it is absolutely necessary to have information on the regional atmospheric input of critical compounds such as nitrogen, phosphorus, potassium and salts, a field where our knowledge is still quite limited (Offer et al. 1992).
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
271
272
19.2
T. Littmann, A. Schultz
Methods
Samples of atmospheric bulk deposition were taken at five locations across the loessic coastal plain and the entire sand dune field, as indicated in Fig. 1 (Introduction chapter, this volume), from September 1998 to June 1999 (10 months). Each site was equipped with two bulk deposition samplers 5 m apart at 1 m above ground. The samplers were made of two 5,000-ml receptor and collector polyethylene bottles (exposed area of receptor 200 cm2, collector aperture covered by a plastic mesh and a layer of 5-mm glass beads) mounted in PVC pipes topped by a metal bird crouch ring. In 2-week intervals (and immediately after rainfall events), the receptor sides were rinsed and wiped with 150 ml distilled water, and the samples were emptied from the collector bottle. The total series thus consists of 36 samples per site. Each bulk sample was resuspended in an ultrasonic bath and filtered through a diaphragm filter (0.45-µm pore size). Elemental analysis was carried out for the water-soluble elements within the filtrate by means of ion chromatography and atom absorption spectrometry. Parameters measured are the ions Na+, Mg2+, Ca+, K+, NH4+, Cl−, SO42−, NO3− and PO43−. The deposited elemental mass was then calculated in mg m−2 day−1 and kg ha−1 year−1. Before subjecting the data to further analysis, we computed the ion balance for each individual sample, which resulted in an addition of the remaining soluble carbonate components (CO32−, HCO3−, CaO−, MgO−) until the error percentage was < 5% (equivalent to pH 7). For each sampling interval, the corresponding mean values from the two samplers per site were used for further interpretation.
19.3
Dimensions of Atmospheric Deposition
Bulk deposition of water-soluble elements was fairly high in the north-western Negev over the observational period (September 1998 to June 1999). The highest overall mass was recorded at site Gevulot (Fig. 19.1) in the north-eastern part of the study area, decreasing towards site Yevul (Fig. 19.1) in the southernmost coastal plain. Further south, across the sand dune field, depositional mass increased from minimum values at the northern margin (site N5) to values equalling those of the coastal plain at the dune field’s southern end (site N1, Nizzana). Such differences in depositional mass are mainly an effect of carbonate input, which generally is the major constituent in bulk deposition (76% at site Gevulot = Ge, 70% at site Yevul = Ye, 67% at site N5, 69% at site N3 and 75% at site N1) and would be even higher if the depositional mass of non-soluble carbonatic residues were included. With carbonate (CO32−), hydrogenic carbonate (HCO3−) and calcium oxide (CaO−) being the main anionic components, as inferred from ion balancing, bulk deposition is weakly acid and pH values do not show significant differences between the sampling locations (6.4 to 6.6). The second largest individual deposition mass was determined for chloride (8.5% at site Gevulot = Ge, 10% at site Yevul = Ye and site
18 Atmospheric Input of Nutrient Elements
273
180 170
171 4,27 3,76
160 150
26,36
145
142
4,14 3,28
140
136 132
130
22,10
3,44 3,12
120
2,17 3,15
1,86 2,65
Ntotal
-
MgO
-
24,15
CaO
22,13
deposition (kg / ha*a)
19,96
110
54,15
100 41,10
90
34,28
35,56
41,55
-
HCO3
80 70
26,63 20,21
16,86
17,49
60 50
9,98 0,4
40 30 20 10
14,49 2,50
11,25 0,19 15,49 2,14
14,29
14,17
0,39
0,23
14,21
14,92
20,43
CO3
12,30
SO4 F
11,75
Cl
1,55
1,09
15,80
14,27
15,82
17,14
1,98
1,88
1,90
7,22
7,92
7,70
7,18
1,59 5,88
sites: Ge
Ye
N5
N3
N1
2,26
0
2-
-
0,38
2,07
18,84
2-
K
-
+
Ca
2+
Mg
2+
+
Na
Fig. 19.1 Mean elemental deposition in kg ha−1 year−1
N5, 11% at N3 and 8% at site N1) and sulphate (6, 8, 11, 10 and 9% respectively). In contrast to the carbonatic components, both these elements show an increase from northeast to southwest in the northern coastal plain, and a further linear decrease from north to south across the sand dune field. The deposition of nitrogen components will be discussed in Chapter 22 (this volume).
274
T. Littmann, A. Schultz
As we could not cover an entire summer period (cf. September 1998 and June 1999 only), the data available show that bulk deposition mass was highest in winter (November to March) and spring (April, May; Table 19.1) at all stations. However, the percentage of mass deposited in the winter months varies from 50% (station site Ye and site N3) and 45% (site N5 and site N1) to 40% (site Ge). Springtime deposition, on the other hand, increased from the northwest (18% at site Ye) to the southeast (31% at site N1) and northeast (26% at site Ge). For an earlier monitoring period at site N1 (18 months in 1995 and 1996), Littmann and Gintz (2000) reported a very similar seasonal structure with highest input in winter and spring and minimum deposition in summer. Generally, the depositional mass at a given location depends largely on the deposition mode, be it wet (rain-out of solution droplets from above the cloud basis or wash-out of gaseous components or particulates along raindrop trajectories), or dry Stokian movements in stable and turbulent deposition (cf. Littmann 1994). Liquidphase reactions, such as sea-spray salts and sulphates, should show a quasi-linear increase with rainfall whereas terrestrial components (geogenic background aerosols) will not. On the other hand, along the desert margin rainfall occurrence is patchy and will coincide mainly with atmospheric instability, turbulence and, thus, wind-induced erosive suspension of particulates, be it local or in long-range transport, as in the case of dust storms (Ganor and Mamane 1982; Dayan 1987; Littmann 1991). In this context, it is not surprising that bulk mass deposition in the study area occurred increasingly as dry deposition (i.e. in sampling intervals without any rainfall event) towards the drier southern part (percentages of absolute mass as wet/dry deposition: site Ge 60/40%, site Ye: 56/44%, site N5: 36/64%, site N3: 36/64%, site N1 (Nizzana): 35/65%), as the frequency of rainfall events showed a very similar spatial pattern in winter 1998/1999 (Chap. 29, this volume). However, following a definition of the relative contribution of sampling intervals with rainfall (wet deposition) and without rainfall (dry deposition) given by Littmann and Gintz (2000), it is apparent that rainfall always leads to the highest relative deposition rates (Table 19.1), except for F− and NO3−-N which occurred both with a 100% relative frequency in dry deposition in the southernmost part at station site N3 and site N1. Also in terms of relative deposition, the relevance of overall wet deposition decreases from north (site Ge: 80%, site Ye: 77%) to south (site N5: 60%, site N3: 68%, site N1: 59%) but generally remains the most effective deposition mode. It may be concluded that the efficiency of rain-out and wash-out processes decreases with decreasing rainfall frequency and rainfall totals in the study area. Atmospheric deposition of dust (mean: 37 kg ha−1 year−1) showed a spatial pattern comparable to that of carbonate bulk deposition. However, the dust mass decreased from site Ge (+23%, compared to the overall mean) towards site Ye (+10%) and site N5 (+12%), whereas the southern part is characterized by fairly low dust deposition (−15 to −20%). On the other hand, the northern and central parts of the sand dune field are significantly interrelated in their depositional behaviour (station site Ye, site N5 and site N3 show r2 coefficients of 0.55 to 0.60), while dust deposition at site Ge and station site N1 does not correlate with any other site, possibly an effect of local disturbance (cf. arable soils at site Ge) and different dust
Element
Bulka
Bulkb
Dryb
Wetb
Dry (rel.)c
Wet (rel.)c
Springb
Summerb
Fallb
Winterb
Na+ Mg2+ Ca+ K+ Cl− SO42− CO32− HCO3− CaO− MgO− F− NH4+-N NO3−-N N total PO43− Total % Dust a mg m−2 day−1 b kg ha−1 year−1 c %
1.97 0.53 4.49 0.51 3.88 3.40 5.57 11.32 6.28 0.88 0.09 0.12 0.06 0.87 0.00 39.79 100.00 125.57
7.19 1.93 16.39 1.86 14.16 12.41 20.33 41.32 22.92 3.21 0.33 0.44 0.22 3.18 0.00 145.23 100.00 458.34
4.12 1.07 8.49 0.99 7.37 7.21 10.63 21.62 11.82 1.76 0.29 0.18 0.18 1.24 0.00 76.61 52.75 271.25
3.07 0.86 7.90 0.87 6.79 5.20 9.70 19.70 11.10 1.45 0.04 0.26 0.04 1.94 0.00 68.62 47.25 187.11
36.68 34.20 30.44 33.60 31.28 35.06 31.06 31.04 30.06 33.80 88.34 20.64 67.46 24.30 x x 37.71 x
63.32 65.80 69.56 66.40 68.72 64.94 68.94 68.96 69.94 66.20 11.66 79.36 32.54 75.70 x x 62.29 x
1.10 0.47 4.32 0.26 2.64 2.99 5.21 10.58 6.07 0.78 0.01 0.08 0.00 0.54 0.00 34.97 24.08 7.67
1.30 0.33 2.51 0.36 2.28 1.99 3.46 7.02 3.55 0.56 0.13 0.03 0.01 0.38 0.00 23.87 16.44 77.96
1.52 0.27 1.88 0.30 1.70 1.25 3.06 6.23 2.41 0.43 0.19 0.02 0.01 0.12 0.00 19.36 13.33 57.77
3.27 0.86 7.68 0.94 7.54 6.18 8.60 17.49 10.89 1.44 0.00 0.31 0.20 2.14 0.00 67.03 46.15 314.96
18 Atmospheric Input of Nutrient Elements
Table 19.1 Bulk, dry, wet and seasonal atmospheric deposition of water-soluble elements. Mean values from five monitoring stations in the sand dune field, September 1998–June 1999 (x traces)
275
276
T. Littmann, A. Schultz
storm synoptics (cf. south-westerly Sharav conditions at Nizzana). This finding may also hold for the interrelation of dust and sand deposition at a given site because a significant correlation can be found only in the central sand dune field (site N5 and site N3 have sand–dust r2 coefficients of 0.83 and 0.33 respectively) where easterly dust storms with higher wind speeds lead simultaneously to the blow-off of sand from the dune ridges. The effect of disturbed arable soils on wind erosion processes is also revealed in the very high depositional sand mass (mean: 62 kg ha−1 year−1) at site Ge (+52%). However, the mass of sand blown into the interdunes does not necessarily depend on dune height or the area of mobile dune crests, as the second highest mass was recorded at station site N3 (+15%) whereas it was lowest at site N1 (−42%), despite this being the most arid and windy part of the sand dune field (station site Ye: −28%, station site N5: −37%). Interdune width seems to be more important in controlling the amount of driven sand at our monitoring stations. However, statistical testing did not reveal overall significant differences in the spatial pattern of atmospheric deposition. All depositional series are highly correlated (slightly higher in the southern part (r2 between site N3 and site N1 is 0.84) and northern part (r2 between station site Ge and site Ye is 0.7) than between the north and south, with r2 around 0.65). T-testing showed only site Ge (highest depositional mass) to be significantly different from all other stations. Although a chi2test also revealed overall homogeneity in the depositional series (chi2=9.7 with 52 degrees of freedom), there is an indication of increased SO42− deposition at station site N5 and site N3, while it is lower than expected at site Ge. At this location, it is the HCO3− input which is higher than expected.
19.4
Element Groups and the Boundary Conditions of Atmospheric Input
When investigating the boundary conditions of atmospheric matter input, one prerequisite is to reduce the variables under consideration to a minimum (Littmann 1994) by means of principal component analysis (PCA). As all series are highly correlated, we applied PCA to nutrient element mean series, which resulted in the solution shown in Table 19.2 (77% explained variance, all communalities>0.6, varimax rotation including the Kaiser criterion). Principal component (PC) 1 includes the entire carbonate group and may be interpreted as a genuine terrestrial component (water-soluble fractions of Ca+ and Mg2+ carbonates; cf. Littmann 1994), typical for the regional terrestrial background aerosols from fine-grained loessic sediments in the Hovav area near Beer Sheva or fluvial fine material in wadi channels originating from the limestone Negev Highlands (Littmann 1997; Littmann and Gintz 2000). Apparently, acidity in atmospheric deposition is effectively buffered by carbonate input, as pH also loads positively on PC1. However, there is no simple interrelation of carbonate deposition and dust storm or dust fallout events, as the chemical composition of dust deposition
18 Atmospheric Input of Nutrient Elements
277
Table 19.2 PCA solution for the mean depositional series (all stations) PC1 PC2 PC3 2+
Ca , 0.96 CaO−, 0.96 CO32−, 0.96 HCO3−, 0.96 Mg2+, 0.85 MgO−, 0.85 pH, 0.57
+
Na , 0.94 K+, 0.71 F−, 0.71 Cl−, 0.65
+
NH4 -N, 0.76 Ntotal, 0.70 K+, 0.57
PC4 NO3−-N, 0.74 SO42−, −0.69
(Mg2+, 0.41) (MgO−, 0.41)
in the area depends largely on its source area characteristics (Littmann 1997). Figure 19.2 shows PC1 deposition to occur in singular events (for further analysis, we refer to factor values as obtained from hierarchical regression analysis) such as in October 1998 (a regional input event), February (a long-range dust transport event originating in Libya), and April to May 1999 (dust storms by khamsinic depressions which, unfortunately, could not be verified by dust mass input because determination had to be stopped in mid-April). On the other hand, the highest dust mass input events in September and November 1998 had other sources than the regional loessic background component. PC2 refers to chlorides (NaCl, MgCl, KCl) of marine origin, as these components are typical for sea-spray salts (cf. Littmann 1994). Thus, PCA does not provide a straight solution for Mg2+, as it may occur as a secondary sea-spray component (the ratio of NaCl and MgCl in sea salt is typically 8:1) as well as in carbonate compounds of terrestrial origin. The PC2 factor value series (Fig. 19.2) does not show an overall interrelation with either rainfall or dust, which implies that sea salt deposition in the study area is predominantly a dry mode case, although more effective when combined with rain-out processes (Table 19.1). PCs 3 and 4 include the input of nitrogen compounds, and will be discussed in Chapter 20 (this volume). However, sulphate deposition loads negatively on nitrate deposition and is thus not interrelated with any of the major PCs. Generally, the PCA solution for the entire region presented here corresponds to earlier results obtained by Littmann and Gintz (2000) for the southernmost part at Nizzana (site N1). However, their solution showed a simpler structure (three PCs as terrestrial, sea-spray and nitrogen compounds), and sulphate input as a coastal urban emission (Levin and Lindberg 1979) was related to the sea-spray component. Depending on the more or less singular character of PC deposition, the terrestrial compound shows a seasonal maximum in winter and spring, which complies with the overall seasonal frequency of dust storms of regional and long-range origin (Littmann 1997); the sea-spray component in autumn (October) and early winter shows minor peaks in the series coinciding with rainfall (January and February 1999). Summer deposition cannot be interpreted further due to the lack of data for July and August. It is interesting to note that, in the northern part of the study area (station site Ge and site Ye), bulk deposition mass per month (in kg ha−1 year−1; Table 19.1) does not show any clear seasonal difference between winter (12–15)
278
T. Littmann, A. Schultz 100
2,50
PC 1 model
2,00
90 80
1,50
70
1,00
60
0,50
50
0,00
40
−0,50
30
−1,00
20
−1,50
10
−2,00
0 16
2,50
rainfall
2,00
PC 2 model
14
0,50
8
0,00
6
−0,50
4
−1,00
2
−1,50
0 99 06 07
05
99 17
04
99 26
04 05
03 15
02
99
99 22
02 01
01
98 11
12
98 21
11 30
11
98 09
10 19
09 28
99
10
99
1,00
99
12
98
1,50
98
PC 2 and model (factor values)
dust (kg/ha*a)
dust
rain (mm/d)
PC 1 and model (factor values)
3,00
Fig. 19.2 Modelling the depositional series. PC1 model series: input=0.307 NE+0.158 WS 3−0.263 W+0.128 WS 2−3.735, where NE and W denote the percentage of north-easterly and westerly wind directions per sampling interval respectively, and WS 3 and WS 2 the percentage of wind speeds in the intervals 4–5.9 m s−1 and 2–3.9 m s−1 respectively. PC2 model series: input=0.047 N+0.009 NE+0.1 E+0.03 SE−0.09 S+0.05 SW−0.152 W+0.01 North-Western +0.05 WS 2+0.23 WS 3−0.749 WS 4; this includes all wind directions and additionally the wind speed intervals > 6.0 m s−1. For further explanation, see text
and spring (12–14). In the southern sand dune field (site N3 and site N1), bulk deposition (highlighting the terrestrial component) becomes increasingly higher in spring (16–21, compared to 13 in winter), which is an effect of khamsinic intrusions from the Sinai, confined to the southernmost part of the study area. As neither dust input nor rainfall provide sufficient explanation for the mean depositional series of PC1 and PC2, we derived multiple regression equations arranged stepwise and including dust input, rainfall, and the relative frequency of wind directions and wind speed intervals per sample interval as independent variables.
18 Atmospheric Input of Nutrient Elements
279
Figure 19.2 shows the corresponding best-fit model output. The PC1 model series primarily includes the relative frequency of north-easterly wind directions (explained variance at this step: 40%) combined with higher wind speeds (interval 4–5.9 m s−1; explained variance 58%). A further consideration of the negative interrelation with westerly wind directions increased explained series variance to 66%; including the frequency of low wind speeds (2–3.9 m s−1) resulted in a best-fit solution with 88% of explained PC1 variance. Thus, the model output (Fig. 19.2) offers a rather good explanation for the deposition mechanisms of the terrestrial component in the entire area. Carbonatic components are deposited mainly as particulates during dry intervals from north-easterly directions, i.e. the loessic Hovav Plateau rich in carbonates, mostly at light wind speeds but occasionally at high wind speeds in the case of regional dust storms. The most relevant explanatory variable in the PC2 model series is the relative frequency of north-westerly wind directions (explained variance: 52%). However, the second-step model run (80%) included all wind directions with southerly and westerly directions being negatively interrelated, and higher but not the highest wind speeds>5.9 m s−1. It may be implied from this model that the deposition of sea-spray salts occurs primarily in the case of north-westerly winds at higher wind speeds, i.e. in stronger sea breeze situations which reach their highest frequency in summer from May to October (cf. Littmann 1997 and Chap. 4, this volume). Rainfall events may lead to minor secondary peaks.
19.5
The Role of Vegetation Stands
It has been found earlier that shrubs in the interdune areas at the Nizzana experimental site show a dust retention about 68% higher than is the case for bare interdune surfaces, as measured by net deposition rates on artificial surfaces (Littmann and Gintz 2000). However, the efficiency of desert shrubs in retaining dust and sand and, thus, contributing to the evolution of underlying stable crustal patches depends largely on the aerodynamic porosity of the plant’s cross section (Littmann 1997). In this context, it is poorly known whether desert shrubs show a depositional pattern differing from bare surface deposition in terms of soluble nutrients and trace elements. Over the observational period, we operated bulk samplers even with the surface in a flat interdune area at the Nizzana experimental site (site N1), one being located in an open, bare space and the other beneath a large (1.5-m-high) Retama shrub. The samplers were additionally covered by a metal mesh to prevent plant litter entering the receptor bottle. The samples were recovered and treated in the same way as described in Section 19.2. Generally, the overall mass of deposited elements beneath the shrub was 17% higher than that recorded at the open control site, especially for F−, carbonates and potassium (Table 19.3). However, chlorides including Na+ and Mg2+ showed only minor enrichment in the vegetation stand sample. As was found for general bulk deposition, atmospheric input during wet sampling intervals (five cases or 28%
280
T. Littmann, A. Schultz
Table 19.3 Elemental deposition beneath an interdune Retama shrub and at an open interdune surface (control); Nizzana experimental site (station site N1=Nizzana)
Element Na+ Mg2+ Ca2+ K+ Cl− SO42− CO32− HCO3− CaO− MgO− F− NH4+-N Total/ mean (%)
Shrub (kg ha year−1)
(kg ha % dry % wet year−1)
% dry
Difference Shrub– Control % wet Total (%) Dry (%) Wet (%)
6.35 3.29 57.38 1.28 13.94 9.13 86.03 174.91 80.30 5.48 0.62 0.17 438.86
67.3 64.3 59.3 61.0 54.5 52.1 61.7 61.7 59.3 64.3 100.0 47.1 63.5
59.6 59.5 51.4 48.2 43.4 48.3 52.9 52.9 50.0 58.0 100.0 77.0 57.8
40.4 40.5 48.6 51.8 56.6 51.7 47.1 47.1 50.0 42.0 0.0 23.0 42.2
32.7 35.7 4.7 39.0 45.5 47.9 38.3 38.3 40.7 35.7 0.0 52.9 36.5
5.88 3.29 45.92 1.02 14.71 7.71 70.08 142.46 66.21 5.55 0.26 0.15 363.22
Control
7.5 0.0 20.0 20.2 −5.5 15.5 18.5 18.5 17.5 −1.3 58.8 11.8 17.2
7.7 4.8 7.9 12.8 11.2 3.8 8.9 8.9 9.2 6.3 0.0 −29.9 5.7
−7.7 −4.8 −7.9 −12.8 −11.2 −3.8 −8.9 −8.9 −9.2 −6.3 0.0 29.9 −5.7
with cumulated rainfall>2 mm) is significantly higher (42% of overall deposition occurred within the wet intervals in the control samples, 36% in the shrub samples) but, when comparing control and shrub series, it is apparent that deposition beneath the shrub was increased by 6% during dry intervals (especially NO3−, K+, Cl−; Table 19.3), with the exception of NH4+ (deposition rate 30% higher beneath the shrub, compared to the open space in wet intervals). Thus, we may imply an individual contribution by the plants to nitrogen, potassium and carbonate input at the underlying soil surface in terms of either particle retention in dry deposition or leaf washout and droplet enrichment by leading during rainfall events. Element grouping of the control series by means of PCA resulted in a much simpler solution than that recorded along the larger-scale transect (Table 19.4). K+ is now included in PC1, SO42− and NO3− in PC2, and NH4+ forms an individual PC. A similar solution was found by Littmann and Gintz (2000) for bulk deposition at 1 m above ground at Nizzana for a longer monitoring period (1994–1996). The PCA solution for the shrub series reveals an interesting difference: nitrogen compounds are included in PC2 but Na+ and K+ form an individual PC3 and, thus, correlate with neither terrestrial nor sea-spray component deposition; this result points to the conclusion drawn from Table 19.3. Except for the third PC in both solutions, elemental deposition in the shrub and control series is fairly interrelated (Fig. 19.3), much better for the PC1 group because (except for K+) the same elements load on PC3 whereas, in PC2, Na+ and NH4+ show different loads. Finally, it is interesting to note that neither the terrestrial PC1 nor PC3 (Na+, K+) of the shrub series show any correlative interrelation with rainfall. Thus, leaf washout in the case of rainfall events is not likely to contribute to the deposition of Na+
18 Atmospheric Input of Nutrient Elements
281
Table 19.4 PCA solutions for the Nizzana interdune depositional series: part A, control surface; part B, Retama shrub PC1 PC2 PC3 A Ca2+, 0.97 CaO−, 0.97 CO32−, 0.97 HCO3−, 0.97 Mg2+, 0.89 MgO−, 0.89 K+, 0.72 B Ca+, 0.97 CaO−, 0.97 CO32−, 0.97 HCO3−, 0.97 Mg2+, 0.91 MgO−, 0.91
Na+, 0.83 Cl−, 0.87 SO42−, 0.76 NO3−, 0.71
NH4+-N, 0.98
Cl−, 0.82 SO42−, 0.77 NH4+-N, 0.72 NO3−, 0.53
Na+, 0.74 K+, 0.80
PC 1 (vegetation)
4,00 3,00
y = 0,8373x + 0,0389 r2 = 0,78
2,00 1,00 0,00 −1,00 −2,00
−2,00
−1,00
0,00
1,00
2,00
3,00
4,00
PC 1 (control)
PC 2 (vegetation)
3,00 2,00
y = 0,6671x - 0,0148 r2 = 0,42
1,00 0,00 −1,00 −2,00 −2,00
−1,00
0,00
1,00
2,00
3,00
PC 2 (control)
Fig. 19.3 Interrelations between surface PC1 (terrestrial carbonates, top) and PC2 (sea-spray salts, bottom) deposition beneath a large Retama shrub and a nearby open space (control) at an interdune site within station site N1 (Nizzana)
282
T. Littmann, A. Schultz 3,00
PC 2 factor values
2,50
r2 = 0,55
2,00 1,50 1,00 0,50 0,00 −0,50 0 −1,00
5
10
15
rainfall (mm)
−1,50
Fig. 19.4 Sea-spray salt (PC2) deposition beneath a Retama shrub at Nizzana (site N1) versus rainfall
and K+, which seems to occur at random. On the other hand, 55% of chloride, sulphate and nitrogen (PC2) deposition beneath the shrub are explained by rainfall events (Fig. 19.4), a result pointing strongly to wet phase solution of these elements associated to plant organic matter, dust and raindrops. However, none of the elemental groups is correlated with dust deposition at the experimental site. This holds also for the terrestrial PC1, as the mineralogical composition of dust does not depend on the overall mass but rather on the origin of the material, which will result in a higher carbonatic content when derived from the loessic area of the Hovav Plateau (Littmann 1997).
19.6
Conclusions
It was found that the overall mass of water-soluble elements deposited by atmospheric processes in the sand dune field of the north-western Negev depends largely on the amount of carbonatic compounds, and shows a decrease in the northern coastal plain from the north-eastern to the south-western margin, consistent with regional loess transport from the Hovav Plateau in the Be’er Sheva region. Within the western parts of the sand dune field, bulk deposition increases from north to south, likely an effect of the Sharav-controlled depositional regime at Nizzana and local carbonate sources (Nahal Nizzana sediments), as the deposition of dust decreases from north to south. Chloride and sulphate input, however, decreases from the coastal plain towards the southern dune field. There is a clear seasonal maximum of deposition mass in winter (i.e. the rainy season) and spring (the dust storm season). Because of this, we may imply a much more effective element input during intervals with rainfall, especially along the northern fringe of the sand dune
18 Atmospheric Input of Nutrient Elements
283
field. However, the spatial pattern indicated here proved to be statistically non-significant. Principal component analysis confirmed two major deposition modes of elemental groups. One component includes the terrestrial carbonatic group, which is modelled to occur at higher wind speeds from north-easterly directions (i.e. from the loessic areas); the other component comprises sea-spray salts from the Mediterranean deposited from north-westerly wind directions. Evidently, the input of salts does not necessarily depend on the amount or frequency of rainfall but rather on the strength and continuity of the seasonal wind field with sea breezes dominating during the summer half year. Vegetation stands may enhance overall deposition mass by 17%, mainly in terms of dry mode particle retention. However, sea-spray salts plus potassium and nitrogen compounds are elemental groups which show increased input at the surface beneath shrubs during and after rainfall events, pointing strongly to leaf washout processes. It may be implied that vegetational stands create an individual depositional environment characterized by enhanced elemental deposition in both dry and wet modes. To conclude, atmospheric deposition is definitely a major contribution to the local nutrient and trace element budgets and cycles within the arid sandy environment of the north-western Negev. The main climatic boundary conditions for dominant transport modes, dust storms and the seasonal wind field may vary in space and time, and will lead to an increased loessification of the northern part of the sand dune field (a positive effect in terms of soil fertility but negative when considering the soil water budget; cf. Yair 1994; Verrecchia et al. 1995) and increasing salinization in the same area during phases of stable zonal circulation over the summer half year. In the southern part, processes may be much slower. However, the vegetation cover plays an important role in accelerating dust and carbonate enrichment as well as in creating accelerated nutrient cycles.
References Dayan U (1987) Sand storms and dust storms in Israel – a review. Israel Atomic Energy Comm Publ IA-1419, Yavne Esser U (1989) Zum Stickstoffhaushalt arider Hangökosysteme im nördlichen Negev-Hochland, Israel und den Auswirkungen der Hang-Minicatchment-Technologie auf Stickstoffumsätze und Vorräte. PhD Thesis, University of Münster Ganor E, Mamane Y (1982) Transport of Saharan dust across the eastern Mediterranean. Atmospheric Environ 16:581–587 Kadmon R, Leschner H (1995) Ecology of linear dunes: effect of surface stability on the distribution and abundance of annual plants. Adv GeoEcol 28:125–143 Levin Z, Lindberg JD (1979) Size distribution, chemical composition and optical properties of urban aerosols in Israel. J Geophys Res 84:6941–6950 Littmann T (1994) Immisssionsbelastung durch Schwebstaub und Spurenstoffe im ländlichen Raum Nordwestdeutschlands. Bochumer Geogr Arb 59, Bochum Littmann T (1991) Dust storm frequency in Asia: climatic control and variability. Int J Climatol 11:393–412
284
T. Littmann, A. Schultz
Littmann T (1997) Atmospheric input of dust and nitrogen into the Nizzana sand dune ecosystem, north-western Negev, Israel. J Arid Environ 36:433–457 Littmann T, Gintz D (2000) Eolian transport and deposition in a partially vegetated linear sand dune area (northwestern Negev, Israel). Z Geomorphol suppl Bd 121:77–90 Offer Z, Goossens D, Shachak M (1992) Aeolian deposition of nitrogen to sandy and loessial ecosystems in the Negev desert. J Arid Environ 23:355–363 Verrechia E, Yair A, Kidron G, Verrechia K (1995) Physical properties of the psammophile cryptogamic crust and their consequences to the water regime of sandy soils, north-western Negev desert, Israel. J Arid Environ 29:427–437 Veste M, Breckle S-W (1996) Root growth and water uptake in a desert sand dune ecosystem. Acta Phytogeogr Suec 81:59–64 West N (1990) Nutrient cycling in soils of semiarid and arid regions. In: Skujins J (ed) Semiarid lands and deserts: soil resources and rehabilitation. Macmillan, New York, pp 180–223 Yair A (1994) The ambiguous impact of climate change at a desert fringe, Northern Negev, Israel. In: Millington A, Pye K (eds) Environmental change in drylands: biogeographical and geomorphological perspectives. Wiley, Chichester, pp 199–227 Yair A, Lavee H, Greitser N (1997) Spatial and temporal variability of water percolation and movement in a system of longitudinal dunes, western Negev, Israel. Hydrol Processes 11:43–58
Chapter 20
Succession Stages in the Recovery Processes of the Topsoil Crust in a Disturbed Sandy Arid Area A. Yair
20.1
Introduction
Biological crusts (called also microphytic, microbiotic or microbial crusts) are topsoil crusts, 1–5 mm thick. They are composed of a variety of biological elements (cyanobacteria, lichens, mosses, fungi, yeast and algae) and of fine-grained material (Chap. 10, this volume). They are highly resistant to extreme temperatures, and are widespread in arid and semi-arid areas. The important role of microphytic crusts in arid ecosystems is now widely recognized (see reviews by Isichei 1990; West 1990; Belnap 1994; Belnap and Lange 2001). Many studies stress their role in the stabilization of sandy areas, protecting the soil against wind or runoff erosion (Booth 1941; Fletcher and Martin 1948; Metting 1981; Campbell et al. 1989; Danin et al. 1989; Pluis and de Winder 1990; Pye and Tsoar 1990; Danin 1996; Belnap and Gillette 1997). Other studies focus on their important role in the cycling of nutrients, especially in nitrogen fixation (Stewart 1967; Lange 1974; Skujins and Klubek 1978; Isichei 1980; Wang et al. 1981; Harper and Pendleton 1993). Their role in seed germination is controversial. Some authors contend that biological crusts, acting as traps of fine material, improve water and nutrient regimes, thereby enhancing seed germination, survival and growth of higher plants (Shields and Durell 1964; Belnap 1995; Kadmon and Leshner 1995; Eldridge and Tozer 1997). Dulieu et al. (1977) adopt a reverse position. They consider that compacted algal crusts limit germination and productivity in grazing lands. A similar controversy exists regarding their hydrological role. Much of the literature reports conflicting information on the relationship between crust cover, infiltration, runoff and soil water regime. Some authors claim that the crusts tend to increase infiltration, thereby enhancing the depth of water infiltration and soil moisture content (Booth 1941; Fletcher and Martin 1948; Loope and Gifford 1972; Brotherson et al. 1983; Belnap and Gardner 1993; Perez 1997; Eldridge and Tozer 1997). The studies discussed above relate to areas where the substratum underlying the biological crust is mainly fine-grained and of aeolian origin. Bare loamy soils are very sensitive to raindrop impact, responsible for surface sealing. Under such conditions, the cohesive flexible biological elements, in addition to the binding effect of the fine-grained soil particles, absorb raindrop energy and prevent the rapid
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
285
286
A. Yair
development of a topsoil rain crust conducive to runoff generation. Infiltration is thereby increased, resulting in limited runoff. Quite different conditions prevail in sandy areas characterized by extremely high infiltration rates. Here, biological crusts, regardless of the components constituting the microbiotic community, tend to reduce infiltration and to generate runoff (Bond 1964; Roberts and Carbon 1971; Yair 1990, 2001; Dekker and Jungerius 1990; Bisdom et al. 1993; Kidron and Yair 1997). The reduced infiltration can have numerous explanations, such as clogging processes linked to the accumulation of polysaccharides produced by the microorganisms (Avnimelech and Nevo 1964), the water-repellent properties of some crusts (Bond 1964; Van den Acker and Jungerius 1985), as well as the sealing effect of the topsoil layer, due to wetting causing the combined swelling of the biological elements and fine-grained material included in the crust (Campbell 1979; Wang et al. 1981; Verrecchia et al. 1995). The destruction of biological crusts, caused mainly by overgrazing, is often regarded as leading to desertification processes (Otterman 1974; Metting 1981; Tsoar and Møller 1986; Belnap 1995). As biological crusts represent an essential component of soil stability and productivity in arid and semi-arid areas, the issue of crust recovery subsequent to crust destruction has attracted the attention of many desert ecologists. Two essential aspects are discussed in the literature. The first aspect is the rate of recovery of disturbed crusted areas, and the second the stages of crust recovery. Anderson et al. (1982) estimate full crust recovery to occur within 14–18 years at sites disturbed by livestock, while Belnap (1993) indicates a longer time frame in the order of 30–40 years for cyanobacteria, 45–85 years for lichens and as much as 250 years for mosses. Johansen (1986) reports, for sites disturbed by fire or grazing in the western USA, a recovery rate of 3–8 years for the algal components, and 10–15 years for the moss–lichen component. The variability in recovery rates of cyanobacteria–lichen crusts has been found to depend on the type and extent of the disturbance, the availability of nearby inoculation material, and soil properties (Belnap 1995). Recovery rates such as given above are often assessed visually. However, Belnap (1993) showed that many components of crust recovery cannot be assessed visually, and West (1990) states that “carefully designed field experiments are needed to obtain definitive answers”. The issue of succession stages in the recovery process of biological crusts is also a matter of controversy. Van den Acker and Jungerius (1985) contend that algae are the first to colonize blowouts along the Dutch sandy coast. The algae are followed by annual plants, for which they appear to provide the nutrients. The above authors also estimate that even an initial crust dramatically diminishes the vulnerability of the sand to deflation. The pioneering role of cryptogams in primary and secondary plant succession is also supported by several authors (Booth 1941; Metting 1981; Rayburn et al. 1982; Mazor et al. 1996). A different position is adopted by Danin et al. (1989), who propose the following stages for stabilization: (1) decreasing sand mobility on rainy days, and germination of pioneer grass; (2) accumulation of fine-grained particles, due to decreasing wind speed near the ground among plant tussocks; (3) establishment of cyanobacteria, and formation of crust; (4) further accumulation of airborne silt and clay, and improvement of water-holding capacity;
20 Succession Stages in the Recovery Processes
287
and (5) establishment of mosses and higher plants. Tsoar et al. (1995) adopt a more extreme position and contend that “Biogenic crusts are formed when the vegetation cover reaches about 20–30% of the area. Vegetation and the biogenic crust then act as a trap of eolian material”. Most studies quoted above were conducted by ecologists or botanists. They focused their attention on purely biological aspects, indicated by the recovery of the perennial vegetation, the occurrence of cyanobacteria, lichens or mosses (Anderson et al. 1982; Johansen 1986; Belnap 1993), and chlorophyll and sugar contents (Mazor et al. 1996). However, the possibly important role of the mineral component in the stabilization process of biological topsoil crusts has been overlooked. It is obvious that this role requires fast deposition of fine-grained particles which stick to the sand grains, and promote the development of a well-cemented and compacted topsoil crust. As the deposition rate of fine-grained aeolian particles is highly dependant on the availability of fines, and on the local wind regime, the importance of the mineral component in surface stabilization can be expected to vary greatly from one area to another.
20.2
The Case of the Nizzana Research Area
A well-known example of surface disturbance by human activity is found along the Israeli-Egyptian border, where a sharp contrast exists between the two sides of the border. The strong difference in spectral reflectance between the Negev and Sinai has been ascribed to grazing and trampling by Bedouins on the Egyptian side, such activities having been absent for long periods of time, or limited, on the Israeli side (Otterman 1974; Warren and Harrison 1984; Tsoar and Møller 1986; Tsoar 1990; Tsoar and Karnieli 1996). According to these authors, overgrazing on the Egyptian side destroyed the vegetation cover, and large areas lost their value as grazing land. However, the study of temporal changes in vegetation cover on the Israeli side, where overgrazing and surface trampling were intermittent, shows that the natural recovery process of the perennial vegetation is relatively rapid when overgrazing is stopped (Thieberger 2001; Chap. 6, this volume). The studies mentioned above were conducted at the regional scale, using satellite imagery or aerial photographs at long time intervals. They focused on the role of the reestablishment of the woody vegetation. Although of great interest, these regional studies do not attempt to provide an understanding of all factors controlling such recovery processes and rates. Field observations in the sandy area along the Israeli-Egyptian border drew attention to the important role which should be attributed to a widespread biological topsoil crust in the functioning of this sandy ecosystem (Yair 1990; Kidron and Yair 1997; Chaps. 10, 17 and 18, this volume). An aspect overlooked in previous studies conducted in the area is the specific role played by the different elements of the biological crust in the stabilization processes occurring in disturbed areas, and its possible effect on the recovery of the
288
A. Yair
perennial vegetation. An additional aspect to be considered is the recovery process in terms of the type and degree of surface disturbance.
20.3
Aim of Present Study
The aim of the present study is to monitor succession stages in the recovery of a heavily disturbed sandy area, considering the recovery of the following ecosystem elements: (1) the mineral component of the topsoil biological crust, (2) the biological component of the biological crust and (3) the perennial vegetation cover.
20.4
Methodology
The study was based on 21 plots from where the topsoil biological crust had been removed. As the microbial elements forming the crust move vertically upon changes in temperature and soil moisture, and although the crust is 1–3 mm thick, a layer 10 cm thick was removed, thereby creating an extremely disturbed bare, unconsolidated sandy surface. The disturbance applied is far more severe than that related to trampling by grazing, which leaves the crumbled crust at the surface. It is more similar to the disturbance caused by the wheels of heavy machinery. The plots were located within the interdune area and on north- and south-facing slopes (Fig. 20.1). They were bordered by metal sheets inserted 10 cm into the sand, preventing inoculation from the adjoining crusted areas. The study was usually based on pairs of plots at each site. From one of the plots, the topsoil crust as well as the perennial vegetation was removed. Removal of the vegetation was done by uprooting the shrubs. From the other plot, the biological crust was removed but the perennial vegetation left untouched (Fig. 20.2). A larger-scale operation (a strip 60 wide and 300 m long) of removal of the perennial vegetation was conducted by means of a bulldozer, in the eastern part of the area (Fig. 20.3). In this case, the thickness of the topsoil layer removed was 10–20 cm.
20.4.1
Sampling Program
The monitoring period lasted 8 years (1996 to 2004). Prior to the removal of the topsoil layer, a representative sample of the undisturbed crust was taken from all plots. The plots were sampled again immediately after the removal of the topsoil crust. During the years 1996–1999, sampling was conducted twice a year, at the beginning and end of the rainy season. Thereafter, sampling was conducted two more times, by the end of the hot summer season in the years 2000 and 2004. Sampling of the topsoil layer
20 Succession Stages in the Recovery Processes
289
Fig. 20.1 Location of monitoring plots
was limited to the top 1–2 mm. Before analyzing the samples in the laboratory, the thin and fragile topsoil crust was carefully separated from the underlying sand. This operation proved to be very delicate at the beginning of the research program, when the crust was still friable. With time, the separation procedure became much easier.
20.4.2
Laboratory Analysis
Previous studies have shown that the topsoil crust is composed of mineral and biological components (Yair 1990; Kidron and Yair 1997; Yair and Verrecchia 2002). Therefore, both elements were considered in this study.
20.4.2.1
Recovery of the Mineral Component
Assessment of the recovery of the mineral component of the topsoil crust was based on the analysis of particle size distribution. This was conducted on wet samples dispersed with sodium pyrophosphate. The sand fraction was analyzed by means of a visual accumulation tube and the silty-clayey fractions by means of a Sedigraph. In addition, a scanning microscope was used in order to record the degree of development of the biotic and mineral components of the crust. The magnification varied in the range of
290
A. Yair
Fig. 20.2 A Non-vegetated and B vegetated plots
1:350 to 1:650. The thickness of the scanned samples was 1–2 mm. The laboratory work was accompanied by direct field observations during each sampling campaign.
20.4.2.2
Recovery of the Perennial Vegetation
The monitoring of the recovery of the perennial vegetation was based on field observations during the period 1996–2004.
20 Succession Stages in the Recovery Processes
291
Fig. 20.3 Large-scale clearing of the topsoil crust and perennial vegetation (1996). Views of the area before (a) and after clearing (b)
20.5
Results
Although the trends of crust recovery are quite similar throughout the study area, the data presented here will focus on the recovery of the crust on the north-facing slope where recovery processes were fastest, most pronounced and, therefore, easier to detect.
292
20.5.1
A. Yair
Wind Regime
It is evident that crust regeneration rates will depend greatly on the wind regime. High to very high winds are expected to limit, or even prevent crust regeneration over a disturbed sandy surface, as shown by the total absence of crust on the crest of sandy ridges, where strong winds prevail. Velocities above 9 m/s are high enough to detach and move the medium sand forming the dunes in the area. An example of the wind regime prevailing in the area is displayed in Fig. 20.4. Wind velocities in
Fig. 20.4 The wind regime in the period 1996–1999
20 Succession Stages in the Recovery Processes
293
excess of 9 m/s occur mainly during the rainy season and early spring. They rarely exceed 20 m/s and their consistency is quite low. The drift potential (representative of the total vector units of all wind directions; Chap. 3, this volume) is between 21–108, indicative of a low-energy wind environment. The drift potential for active dunes may exceed the value of 2,000 (Tsoar 2001).
20.5.2
Rainfall Regime
The monitoring period was characterized by a high frequency of dry to very dry years (Table 20.1). Annual rainfall varied in the range of 27 to 108 mm, and was below the long-term average of ~90 mm for 7 of the 8 years considered. Average annual rainfall for the monitoring period (1996–2004) was 62 mm.
20.5.3
Recovery of the Mineral Component of the Crust
Following the removal of the topsoil crust in September 1996, the silt and clay content in the disturbed surface samples was less than 5% in all cases. An earlier study (Yair and Verrecchia 2002) had shown a fast recovery of the mineral component of the crust. In March 1999, 30 months after the complete removal of the initial topsoil crust, a crusted surface with no ripple marks developed (Fig. 20.5), and a sharp increase in silt and clay content was recorded (Table 20.2; also see Fig. 20.6). The fast recovery of the mineral elements of the crust is corroborated by SEM analysis of topsoil samples taken in March 1999 (Fig. 20.7A). The sand grains are embedded in a dense and compacted, smooth matrix of fine-grained particles. A further increase in silt and clay content was found for samples taken in August 2004. At this stage, silt and clay levels are even higher than in the original topsoil crust (Table 20.2), and the surface appears highly compacted.
Table 20.1 Annual rainfall (1996–2004) Rainfall year Rain amount (mm) 1996–1997 1997–1998 1998–1999 1999–2000 2000–2001 2001–2002 2002–2003 2003–2004 Average
73.2 71.8 26.8 28.7 35.0 69.2 108.0 83.0 62.0
294
A. Yair
Fig. 20.5 Crusted and stable surface with no ripple marks (May 1999)
20.5.4
Table 20.2 Temporal changes in silt and clay content (%) September 1996 March 1999
August 2004
North-facing slope South-facing slope Interdune
42.5 30.1 32.5
32.4 22.9 18.2
28.5 22.5 20.1
Recovery of the Biological Components of the Crust
SEM photographs of the initial topsoil surface, prior to the disturbance applied, show a dense network of filaments composed of cyanobacteria and lichens (Fig. 20.6). SEM photographs taken in March 1999 (Yair and Verrecchia 2002) showed no biological elements at the surface of the crust, and a very loose network of filaments at the bottom of a thin crust from a north-facing plot (Fig. 20.7). In the field, all plots had a yellowish colour, indicative of a limited development of the biological elements. Figures 20.8 and 20.9 show crust samples taken 5 years later, in August 2004. A clear increase in the amount of biological elements at the surface and at the bottom of the crust can be observed. Most biotic elements represent filamentous structures, probably of Cyanobacteriae. Lichens, present in the original and undisturbed surface (Yair and Verrecchia 2002), were not observed in later samples. As could be expected, the increase in the biotic elements is more pronounced for samples from the north-facing plots than for those from the south-facing plots, while samples representing the interdune area show an intermediate development. These findings are supported by field observations. At this stage, the topsoil crust which had developed in the north-facing plots had a light-grey colour, contrasting with the still more yellowish colour of the surface in the interdune and south-facing plots. Upon wetting, the grey colour turned into green.
20 Succession Stages in the Recovery Processes
295
Fig. 20.6 SEM images of the initial crust: a Cyanobacteria, b Lichens (after Yair and Verrecchia 2002)
20.5.5
Recovery of the Vegetation Cover
Annuals were already observed in all plots during the first winter season after disturbance. The density of annuals varied from year to year, in function of the annual rain amount. A different trend was observed for perennials. Figure 20.10 shows the large strip from where the perennial vegetation had been completely removed in
296
A. Yair
Fig. 20.7 SEM images of the topsoil crust after 3 years. a Sand grains embedded in a compacted matrix of fine-grained particles. b Very loose network of filaments with an open structure after 3 years (after Yair and Verrecchia 2002)
September 1996 (Fig. 20.3). The photograph was taken 8 years after topsoil layer removal, in April 2004 at the end of the rainfall season 2003–2004. No change in the perennial vegetation cover had occurred since the clearing of the perennial
20 Succession Stages in the Recovery Processes
297
Fig. 20.8 Views of the crusted surface after 8 years (August 2004). a View of a whole plot. b Close-up view of the new, compacted topsoil crust
shrubs. However, an extensive cover of annuals can be seen. The same observation applies to the smaller plots (Fig. 20.8).
20.6
Discussion and Conclusions
Sandy areas are often regarded as poor habitats because they usually contain a limited amount of nutrients and organic matter. This is especially correct in arid areas where the low rain amounts are an additional limiting factor. Furthermore, due to the low cohesion of sand grains, sandy areas are highly sensitive to disturbance. Once disturbed, sand mobility increases strongly and the recovery process may become irreversible, or would require specific climatic conditions, such as a
298
A. Yair
Fig. 20.9 A mat of biological elements at the surface of the crust after 8 years (August 2004)
Fig. 20.10 Extensive area devoid of perennial vegetation 8 years after clearing
number of consecutive wet years. Quite often, the damages are not limited to the disturbed area. The easily blown sand encroaches on arable lands at the desert fringe, thereby limiting their potential productivity. In view of the important role
20 Succession Stages in the Recovery Processes
299
played by biological topsoil crusts in arid and semi-arid ecosystems, the knowledge of the processes and rates of crust recovery following surface disturbance is of importance in the management policy of areas regarded as being highly sensitive to human disturbance. The data obtained in the present study provide an insight into these recovery processes and rates for three important elements of a heavily disturbed sandy arid area: the mineral component of the topsoil crust, the biological components of the topsoil crust, and the perennial vegetation. The data obtained clearly show that the very first element to recover is represented by the annual vegetation (Fig. 20.10). As would be expected, its density varies from year to year, depending on the annual rain amount. The second element to recover is represented by the mineral component of the biological topsoil crust. This conclusion is supported by analysis of scanning electron micrographs. The samples taken in March 1999 (3 years after disturbance) show a dense and well-compacted, thin topsoil crust. The rapid deposition of finegrained particles may be explained by the fact that wind speeds in the study area are quite low (Chap. 6, this volume), resulting in a low mobility index enabling fast deposition of fine airborne particles. The rapid surface stabilization by the dense and compacted mineral component of the topsoil crust does not support the views advanced by several authors that the early establishment of the living elements of the topsoil crust and/or of the higher vegetation preconditions the deposition of fine-grained particles. Surface stabilization can therefore be achieved without, and independently of the biological elements of the ecosystem considered. Furthermore, the rapid recovery of the mineral component of the crust, over a very short period of less than 3 years, does not support the view expressed by Littman (1997) that such a process would require a time span of 100–300 years in the Nizzana area. As the deposition rate of fine-grained aeolian particles is highly dependant on the availability of fines, and on the local wind regime, the importance of the mineral component in surface stabilization can be expected to vary greatly from one area to another. The third element to recover is represented by the biological elements of the crust. The initial recovery of the biological elements was observed within 2–3 years. After 8 years, the mat of microbial elements was better developed (Fig. 20.9), with a predominance of many filamentous structures. These are most probably Cyanobacteriae (Chap. 10, this volume), which usually appear first and are the dominant organisms. Lichens and algae had not yet appeared. The time required for the full recovery of the crust under surface disturbance by grazing is estimated at 15–20 years. This estimation is based on the fact that lichens and algae were already observed in crust samples collected in 1996 from the area which had been under severe grazing pressure from 1967 to 1982, when the border between Israel and the Sinai was open. Heavy trampling by goats and sheep destroyed large parts of the topsoil biological crust. Following the peace agreement with Egypt, the border was closed in 1982. Grazing practices ceased on the Israeli side, enabling the recovery of the crust. When the Nizzana Research Station was established in 1989, extensive areas were already covered by a topsoil crust. The fast recovery of the crust may be explained by the fact that surface
300
A. Yair
disturbance was limited to the crumbling of the crust by trampling. One may therefore conclude that the time span required for the full recovery of the topsoil crust within the experimental plots (from where the topsoil crust had been completely removed) would exceed 30 years. At this stage, it is not possible to advance any sounder evaluation. However, the data obtained show that the type of disturbance is an important factor for the sequence and speed of crust recovery, and that the recovery time is not the same for the mineral and biological components of the topsoil crust. The last element to recover is represented by the perennial vegetation. In fact, 8 years after the removal of the topsoil crust from a large area (Fig. 20.10), the perennial vegetation had not yet recovered. This finding is quite different from statements in previous studies (Thieberger 2001, and Chap. 6, this volume). Two possible explanations are proposed for the results obtained in the present study. The first explanation is based on the type and severity of surface disturbance. In the case of studies dealing with the recovery process after the cessation of grazing, surface disturbance was limited to the crumbling of the topsoil crust, by trampling. Grazing did not lead to the uprooting of the perennial shrubs. Upon the cessation of grazing pressure, living dwarf shrubs as well as the topsoil crust had the possibility to quickly recover. In the present study, all perennial plants had been completely uprooted from the experimental plots. The regeneration of the perennial vegetation requires, in this case, the germination of completely new shrubs. This process is far more complex and problematic, especially in view of the fact that most years following the removal of the topsoil crust had been dry to very dry (Table 20.1). The process would probably need several consecutive wet years. The second explanation is based on the new surface properties developed after strong surface disturbance. Due to the fast deposition of fine-grained aeolian particles, a compacted and resistant topsoil crust developed quickly (Figs. 20.7 and 20.8). A possible negative effect of the compacted crust is to reduce infiltration depth and, consequently, water resources for perennial shrubs. The new, very stable surface may represent an ecosystem less fertile than the original one. Under such conditions, it is possible that the strong disturbance applied will have long-term, irreversible effects. The only possibility for the reversal of the negative trend would be the occurrence of several consecutive wet to very wet years. Finally, the present results suggest that any study of topsoil crust regeneration should consider both the physical and the biotic elements of the crust. Topsoil crust regeneration may be expected to vary widely, depending on numerous factors such as the local wind and rainfall regimes, the deposition rate of fine-grained particles, and the type and degree of disturbance. All these factors will determine the properties of a new topsoil surface which exercises strong control on the water regime of the study area. Acknowledgements The study was supported by the Arid Ecosystem Research Centre of the Hebrew University of Jerusalem, and a research grant from the GIF (German-Israeli Foundation). I am grateful to Prof. E. Verrecchia of the University of Neuchâtel, Switzerland, for his assistance with the SEM photography, and to Mrs. M. Kidron of the Department of Geography for drawing the illustrations.
20 Succession Stages in the Recovery Processes
301
References Anderson DC, Harper KT, Rushforth SB (1982) Recovery of cryptogamic soil crusts from grazing on Utah winter ranges. J Range Manage 35:355–359 Avnimelech Y, Nevo Z (1964) Biological clogging of sands. Soil Sci 98:222–226 Belnap J (1993) Recovery rates of cryptobiotic soil crusts: assessment of artificial inoculant and methods of evaluation. Great Basin Naturalist 53:89–95 Belnap J (1994) Potential role of cryptobiotic soil crusts in semiarid rangelands. In: Proc Ecology and Management of Annual Range Lands. US Department of Agriculture Forest Service, Ogden, UT, pp 179–185 Belnap J (1995) Surface disturbances: their role in accelerating desertification. Environ Monit Assess 37:39–57 Belnap J, Gardner JS (1993) Soil microstructure in soils of the Colorado plateau: the role of the cyanobacterium Microcoleus vaginatus. Great Basin Naturalist 53:40–47 Belnap J, Gillette DA (1997) Disturbance of biological soil crusts: impacts on potential wind erodibility of sandy desert soils in Southeastern Utah. Land Degrad Reclamation 8:355–354 Belnap J, Lange OL (eds) (2001) Biological soil crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York Bisdom EBA, Dekker LW, Schoute JFT (1993) Water repellency of sieve fractions from sandy soils and relationships with organic material and soil structure. Geoderma 56: 105–118 Bond RD (1964) The influence of the microflora on physical properties of sand. Effects associated with filamentous algae and fungi. Austr J Soil Res 2:111–122 Booth WE (1941) Algae as pioneers in plant succession and their importance in erosion control. Ecology 22:38–46 Brotherson JD, Rushforth SB, Johansen JR (1983) Influence of cryptogamic crusts on moisture relationships of soils in Navajo National Monument, Arizona. Great Basin Naturalist 43:73–78 Campbell SE (1979) Soil stabilization by prokaryotic desert crusts. Implication for Precambrian land biota. Origin Life 9:335–348 Campbell SE, Seeler JS, Golubic S (1989) Desert crust formation and soil stabilization. Arid Soil Res Rehabil 3:217–228 Danin A (1996) Plants of desert dunes. Springer, Berlin Heidelberg New York Danin A, Bar-Or Y, Dor I, Yisraeli T (1989) The role of cyanobacteria in stabilization of sand dunes in southern Israel. Ecol Mediterr 15:55–64 Dekker LW, Jungerius PD (1990) Water repellency in the dunes with special reference to the Netherlands dunes of the European coast. Catena suppl 18:173–183 Dulieu D, Gaston A, Darley J (1977) La dégradation des pâturages de la région de N’Djamena (Tchad) en relation avec la présence de cyanophycées psammophiles. Rev Elevage Méd Vét Pays Trop 30:181–190 Eldridge D, Tozer ME (1997) A practical guide to soil lichens and bryophytes of Australia’s dry country. Department of Land and Water Conservation, Sydney Fletcher JE, Martin WP (1948) Some effect of algae and molds in the rain-crust of desert soils. Ecology 29:95–100 Harper KT, Pendleton RI (1993) Cyanobacteria and cyanolichens: can they enhance availability of essential mineral for higher plants. Great Basin Naturalist 53:59–72 Isichei AO (1980) Nitrogen fixation by blue-green algae soil crusts in Nigerian savanna. In: Rosswall T (ed) Nitrogen cycling in West African ecosystems. NFR, Stockholm, pp 191–199 Isichei AO (1990) The role of algae and cyanobacteria in arid lands. A review. Arid Soil Res Rehabil 4:1–17 Johansen JR (1986) Soil algae and range management. Appl Phycol Forum 3:1–2
302
A. Yair
Kadmon R, Leshner H (1995) Ecology of linear dunes: effect of surface stability on the distribution and abundance of annual plants. Adv GeoEcol 28:125–143 Kidron GJ, Yair A (1997) Rainfall-runoff relationship over encrusted dune surfaces, Nizzana, western Negev, Israel. Earth Surface Processes Landforms 22:1169–1184 Lange W (1974) Chelating agents and blue-green algae. Can J Microbiol 20:1311–1321 Littman T (1997) Atmospheric input of dust and nitrogen into the Nizzana sand dune ecosystem, north-western Negev, Israel. J Arid Environ 36:433–457 Loope WL, Gifford GF (1972) Influence of a soil microfloral crust on select properties of soils under pinyon-juniper in southwestern Utah. J Soil Water Conserv 27:164–167 Mazor G, Kidron GJ, Vonshak A, Abeliovitch A (1996) The role of cyanobacterial exopolysaccharides in structuring desert microbial crusts. FEMS Microbiol Ecol 21:121–130 Metting B (1981) The systematics and ecology of soil algae. Bot Rev 47:195–312 Otterman J (1974) Baring high albedo soils by overgrazing: a hypothesized desertification mechanism. Science 186:153–533 Perez FL (1997) Microbiotic crusts in the high equatorial Andes, and their influence on Paramo soils. Catena 31:173–198 Pluis JLA, de Winder B (1990) Natural stabilization. Catena suppl 18:195–208 Pye K, Tsoar H (eds) (1990) Aeolian sand and sand dunes. Unwin Hyman, London Rayburn WR, Mack RN, Metting B (1982) Conspicuous algal colonization of the ash from Mount St. Helens. J Phycol 18:537–543 Roberts FJ, Carbon BA (1971) Water repellence in sandy soils of south-western Australia. Austr J Soil Res 10:35–42 Shields LM, Durell LW (1964) Algae in relation to soil fertility. Bot Rev 30:92–128 Skujins J, Klubek B (1978) Nitrogen fixation and cycling by blue-green algae-lichen crusts in arid rangeland soils. Ecol Bull (Stockholm) 26:164–171 Stewart WDP (1967) Transfer of biologically fixed nitrogen in a sand dune slack region. Nature 214:603–604 Thieberger Y (2001) Spatial and temporal recovery patterns of perennial plants on desert sand dunes. PhD Thesis, The Hebrew University, Jerusalem Tsoar H (1990) The ecological background, deterioration and reclamation of desert dunes. Agric Ecosystems Environ 33:147–170 Tsoar H (2001) Types of aeolian sand dunes and their formation. In: Balmforth NJ, Provenzale A (eds) Geomorphological fluid mechanics. Lecture Notes in Physics vol 582. Springer, Berlin Heidelberg New York, pp 403–429 Tsoar H, Karnieli A (1996) What determines the spectral reflectance of the Negev-Sinai sand dunes. Int J Remote Sensing 17:513–525 Tsoar H, Møller JT (1986) The role of vegetation in the formation of linear sand dunes. In: Nickling WG (ed) Aeolian geomorphology. Allen & Unwin, Boston, MA, pp 75–95 Tsoar H, Goldsmith V, Schoenhaus S, Clarke K, Karnieli A (1995) Reversed desertification on sand dunes along the Sinai-Negev border. In: Tchakerian VP (ed) Desert aeolian processes. Chapman & Hall, London, pp 251–266 Van den Acker JAM, Jungerius PD (1985) The role of algae in the stabilization of coastal dune blowouts. Earth Surface Processes Landforms 10:189–192 Verrecchia E, Yair A, Kidron GJ, Verrecchia K (1995) Physical properties of the psammophile cryptogamic crust and their consequences to the water regime of sandy soils, north-western Negev, Israel. J Arid Environ 29:427–437 Wang F, Zhung Z, Hu Z (1981) Nitrogen fixation by an edible terrestrial blue-green algae. In: Gibson AH, Newton WE (eds) Current perspectives in nitrogen fixation. Elsevier, Amsterdam, p 455 Warren A, Harrison CM (1984) People and the ecosystem: biogeography as a study of ecology and culture. Geoforum 15:365–381 West NE (1990) Structure and function of microphytic soil crusts in wildland ecosystems of arid and semi-arid regions. Adv Ecol Res 20:179–223
20 Succession Stages in the Recovery Processes
303
Yair A (1990) Runoff generation in a sandy area – the Nizzana sands, western Negev, Israel. Earth Surface Processes Landforms 15:597–609 Yair A (2001) Effects of biological soil crusts on water redistribution in the Negev desert, Israel: a case in longitudinal dunes. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management. Ecological Studies vol 150. Springer, Berlin Heidelberg New York, pp 303–314 Yair A, Verrecchia E (2002) The role of the mineral component in surface stabilization processes of a disturbed desert sandy area. Land Degrad Develop 13:295–306
Chapter 21
Dew Formation and Activity of Biological Soil Crusts M. Veste, B.G. Heusinkveld, S.M. Berkowicz, S.-W. Breckle, T. Littmann, and A.F.G. Jacobs
21.1
Introduction
Biological soil crusts are prominent in many drylands and can be found in diverse parts of the globe including the Atacama desert, Chile, the Namib desert, Namibia, the Succulent-Karoo desert, South Africa, and the Negev desert, Israel. Because precipitation can be negligible in deserts – the Atacama desert being almost rainfree – or restricted to infrequent rains during short rainfall seasons, atmospheric moisture in the form of dew and/or fog can be a major, regular supplier of water for cryptogams. Dew and fog have received less attention in ecology, primarily because of their far-smaller quantitative output relative to rainfall. Dew and fog research, however, demands an interdisciplinary and multidisciplinary perspective because of the multiple roles and complex contributions that dew and fog can play as ecosystem “engine”. Desert invertebrates, such as isopods, ants, beetles and desert snails, are well known to rely on dew and fog as reliable moisture sources (Broza 1979; Moffett 1985; Degen et al. 1992), and desert soil fauna such as nematodes are also sensitive to dew deposition on soil surfaces (Steinberger et al. 1989). In contrast to lichens, there has been some controversy as to whether dew/fog can serve as water sources for higher desert plants. In earlier studies (for example, Waisel 1958), it was believed that the majority of desert plants were not capable of utilizing water deposited or appearing on a leaf surface. However, succulents in the South African Karoo showed a clear cooling of the leaf surface below the dewpoint and an increase in fresh weight after dewy nights (von Willert et al. 1992). Recent experimental work revealed that Crassula species from the Succulent Karoo were able to absorb liquid water from the leaf surface by means of hydratodes (Martin and von Willert 2000), whereas in the coastal zone of the Namib, Arthraerua leubnitziae was able to use fog as water source (Loris et al. 2004). Munne-Bosch et al. (1999) and Munne-Bosch and Alegre (1999) found that some Mediterranean plants/shrubs could absorb dew and thereby restore plant water status. There have been few well-instrumented studies investigating surface moisture formation/evaporation coupled with physiological experiments. Only a handful of long-term experiments have been conducted on the nocturnal hydration of lichens
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
305
306
M. Veste et al.
by dew and fog. Examples include Lange et al. (1994, 2006), and Loris et al. (2004) in the Namib desert, Lange et al. (1970), Kappen et al. (1979) and Evenari et al. (1982) in the Negev Highlands. Detailed information on microclimatic processes relevant to biological crust activity is relatively rare. The reasons for this are common to most field-based research, namely excessive isolation of the desired research area, logistics, manpower and equipment costs, power supply for instruments, and the fact that technological/electronic developments have only relatively recently miniaturized sensors and data logging devices and the remote transmission of data. To study in situ microclimatic boundary conditions of dew formation and/or influence on biological crust activity in a hot desert, a variety of intensive field experiments were conducted by the authors in the Haluza sand dune region, NorthWestern Negev desert. Microclimatic parameters such as the radiative energy budget, specific humidity, or difference between air temperature and dewpoint are needed to determine the onset and termination of lichen photosynthetic activity. In the present paper, the physiological activation of soil lichens was measured by chlorophyll fluorescence (as used by Schroeter et al. 1992; Leisner et al. 1997). For the biological sand crusts, general meteorological stations were established on a dune slope or along a transect, in addition to intensive field campaigns where a variety of meteorological sensors were operated in parallel with manual and automatic microlysimeter dew measurements of both physical and biological crusts. The purpose focused on acquiring detailed information on the dew formation and drying process and dew quantities that could condense overnight. Full details regarding the experiments and instrumentation may be found in Jacobs et al. (1999, 2000a), Veste et al. (2001), Heusinkveld et al. (2006) and Littmann and Veste (2006).
21.2 21.2.1
Dew and Fog Definition and Measurement
Dew refers to atmospheric liquid water that condenses on a substrate that has reached the dewpoint. Radiative cooling of the near-ground air layer is the basic process involved, and can start about 1 h before sunset. Terrestrial radiation losses from the surface then lead to large negative net radiation values over 2 to 3 h, before levelling off. During the evening, free liquid water appearing on a natural surface can originate from three separate sources (Monteith 1957; Garratt and Segal 1988): the air (so-called “dewfall”), the soil (“dew-rise”), and plants (guttation). As deserts are characterized by very low soil moisture and scant perennial vegetation, “dewfall” is the predominant source of such surface moisture. In contrast, fog consists of tiny water droplets suspended in the atmosphere, and is defined in terms of visibility, officially as limited to less than 1 km. A fog layer forms when a moist air mass is cooled to saturation, i.e. the dewpoint. The diameters of fog droplets range from about 1 to 40 µm; in comparison, drizzle ranges from 40 to 500 µm and raindrops from about 0.5 to 5 mm. The very low fog droplet settling
21 Dew Formation and Activity of Biological Soil Crusts
307
velocity is thus conducive to wind transport. Deposition of fog droplets on a given surface is caused mainly by interception. Fog can provide a significant moisture input in some deserts, notably the coastal desert of Namibia (Armstrong 1990) and the Atacama desert in Chile (Cereceda et al. 2002). A major limitation in assessing the ecological role of dew has been the difficulty in accurately measuring moisture input to a given substrate. The reader is referred to von Rönsch (1990), Berkowicz et al. (2001); Richards (2004) and Agam and Berliner (2006) for succinct reviews on this problem. In brief, a variety of direct approaches have been tried to date, based on, e.g. moisture-absorbing material, dewdrop size calibrations, recording balances, and electrical surface wetness circuits. These are greatly limited in that they rely on artificial measuring surfaces and will not register dew below specific thresholds. One of the problems in dew observations is the need to have information on the rate of dew accumulation, its duration, and subsequent evaporation. Spot observations immediately after sunrise can provide information only on the maximum amount of dew on the observation surface at that time. Equally problematic is fog input measurements. Since the droplets are carried easily by wind, Schemenauer and Cereceda (1994) developed a simple method to intercept and collect these for drinking purposes. The approach consists of a standard mesh that is installed vertical to the ground and orientated to the predominant direction of incoming fog. In a field setting, however, intercepted fog will vary greatly according to the dimensions of the vegetation, exposure, leaf area index, and wind speed at the height of the vegetation or substrate in question.
21.2.2
Dew and Fog in the Northern Negev Desert
Because of the proximity of the northern Negev to the Mediterranean Sea, this region can be described as a coastal desert. The northern Negev desert experiences frequent dew occurrences throughout the year, especially during the long, hot summer. In order for dew to form on a given surface, the humidity near that surface must be high. Zangvil (1996) outlined the synoptic conditions promoting dew formation in this region. Sharp land–sea temperature differences create a sea-breeze effect, bringing moist air inland to the Negev. Subsidence inversions caused by high pressure in the mid-troposphere helps contain such moisture. The accompanying clear skies allow for radiational cooling. Light winds enhance the cooling process, since stronger winds would lead to mixing of air. In general, fog occurrences in Israel range between 10–50 nights per year, depending on location from the coast, elevation and season, with radiation fogs being the most common form (Goldreich 2003). In the northern Negev, there are about 40 fogs per year in the Beer Sheva Valley, and about 20–25 fogs in Sede Boker (Negev Highlands; Bitan and Rubin 1991). Fog occurrence in the Nizzana sand dune region is similar to that of Sede Boker (Israel Meteorological Service, personal communication). Fog interception measurements using a 1-m2 collector were carried out in Nizzana, at the southern limit of the Haluza sand dunes. During one summer measurement period, the amount of fog water collected overnight
308
M. Veste et al.
reached almost 5 kg on two successive nights (Berkowicz, unpublished). This is the equivalent of 10 mm of rain. In the northern Negev, dew contributes small amounts of water on a very regular basis. Such apparently insignificant but regular moisture contributions play an important role in desert ecosystems. There has been limited research on dew in Israel, however, and this has tended to focus on simple measurements using proxy surfaces (cf. Duvdevani wood block, Evenari et al. 1982; Hiltner balance, Zangvil 1996; absorbent cleaning cloth, Kidron 1998). Kidron (1999) attempted a shortterm study of dew and fog deposition in the Negev using cut squares of cleaning cloth resting on a larger, flat glass plate insulated by wood and set horizontal on the ground. Apart from the general problem of proxy measurements, such adsorbent cloth brands vary considerably in fibre composition, thickness and threading, and the exposed glass surface surrounding the cloth can serve as a dew condenser. The resulting dewdrops bordering the cloth can be absorbed, thereby leading to overestimates in terms of the measuring surface as such. In addition, fog droplets are very fine and windborne; deposition takes place mainly through interception, and settling is hard to quantify. The average long-term dew amount per event measured in Avdat and Sede Boker (both about 50 km SE from the Haluza sand sites, approx. 600 and 480 m a.s.l. respectively) varies between 0.06–0.14 mm but dew can reach higher values of between 0.2–0.30 (Evenari et al. 1982; Zangvil 1996). In Avdat, up to 33 mm of dew per year was measured (Kappen et al. 1979; Evenari et al. 1982). Evenari et al. (1982) relied on a Duvdevani wood block that allows the observer to convert observed dewdrop size to mm equivalent depth of precipitation, while Zangvil (1996) used a Hiltner balance. The limitation is that both sensors are offset from the ground and have different radiative properties. Recent studies in the northern Negev by Jacobs et al. (1999, 2000a, b, 2002), Heusinkveld et al. (2006) and Littmann and Veste (see Chap. 13, this volume) have carried out highly sophisticated measurements on dew formation and evaporation in the Nizzana sand dune region, while Ninari and Berliner (2002) and Agam and Berliner (2004, 2006) have done similar research on a bare loess soil in an area about 20 km south of Beer Sheba. Heusinkveld et al. (2006) used manual and recording microlysimeters in the Nizzana region (190 m a.s.l.) of the Haluza sands, demonstrating that dew values of between 0.2–0.3 mm per night are, in fact, common during the hot summer. According to Littmann and Veste (Chap. 13, this volume), the annual dew amount obtained from a zeroplane model was 26 mm year−1 and their attempts with a load cell gave 33 mm year−1, whereas other models mostly overestimated the amount of dew. These values are not extraordinarily high, given that dew occurs on about 200 nights per year in the northern Negev. Jacobs et al. (1999, 2000a, b, 2002) and Heusinkveld et al. (2006) used several dew measurement techniques and instruments, such as manual microlysimeters, surface sampling and oven-drying, leaf wetness sensors, Duvdevani wood block, a recording load-cell microlysimeter, and fiber-optic wetness sensing devices in field campaigns carried out in the autumn of 1997 and 2000 in the Nizzana sand dunes. Two important questions were posed: when does dew accu-
21 Dew Formation and Activity of Biological Soil Crusts
309
mulation begin in a desert ecosystem, and how long can dew persist before evaporating? In 1997a 200-m dew transect was set up from a south- to a north-facing slope using sets of microlysimeters. The microlysimeters were weighed at intervals throughout the night and also throughout the morning until no further change in weight was detected. These measurements were repeated in autumn 2000 for the same north-facing dune slope, this time also using a high-performance recording microlysimeter. It was found that both slopes received surprisingly similar amounts of dew, ranging from 0.1–0.2 mm. The south-facing slope had a marginally higher dew input than the north-facing slope. Although south-facing slopes receive more solar radiation, this leads to faster and greater long-wave cooling at the surface at night, compared to more sun-sheltered north-facing slopes. According to Jacobs et al. (1999, 2000a, b), dry soil pores could already begin absorbing atmospheric moisture about 1 h before sunset. The drying process, however, was found to proceed very slowly. In the early morning, the difference in vapour pressure between the atmosphere and the air in the soil pores at the interface is the driving force for the evaporation process, while the reverse temperature gradient in the upper soil tends to block this process. Evaporation from the soil will continue until no water is available at the interface. The measurements also revealed that dew could persist in the soil until 13:00, especially on the more sun-protected north-facing slope. This has important implications. The slow drying process following sunrise means that, within the uppermost few millimetres, moisture remains available longer to the biological crust. The data also suggest that, in winter, dew input may in fact persist throughout the day. For the autumn 2000 field campaign, two high-performance recording microlysimeters were installed: one on the north-facing dune midslope and the other near the footslope in a playa. Both manual and recording microlysimeters highlighted that the midslope location consistently averaged 34% less dew than on the playa soil. To assess whether spatial variations in microclimate could be a factor, additional manual microlysimeters were transposed between the midslope and the playa. The result was the same, thus highlighting soil properties as likely factors. Chemical analyses of the samples pointed to the high salinity of the playa surface. Salinity increases the vapour pressure deficit in soil pores, thereby enhancing dew yield. Differences in dew input and in the penetration of dew into a soil, along with relevant meteorological data, are presented in Fig. 21.1. For this experiment, dew data from a recording microlysimeter on the midslope of the encrusted north-facing dune were compared with those from crust samples collected near the footslope (0–4 mm) and soil samples taken immediately below the crust (5–35 mm depth). Dew differences between the crusts may be explained, in part, by physical and chemical properties. The graph highlights that dew does not penetrate beneath the crust, except for a slight increase occurring at peak dew accumulation. The dewpoint and wind speeds increase sharply in Fig. 21.1a on the afternoon of 18 October, reflecting the sea-breeze effect. Of interest is that the dewpoint reached the surface
M. Veste et al.
T1.5 Td 3m Ts u Q*
(a)
30
1200
900
-1
20
600
10
300
0
0
-10 0.20
Lysimeter (mm)
θ dune crust 4 mm (-)
-300 0.05
θ dune -5 to -35 mm (-)
accumulated dew (mm)
(b)
0.04
0.15
0.03 0.10 0.02 0.05
θ (m3 m-3)
Windspeed [m s ], T [oC]
40
Q* [W m-2]
310
0.01
0.00
0 06:00
12:00
1 8: 00
00:00
06: 00
12:00
18:00
00:00
Time [EET]
Fig. 21.1 a Meteorological data for 18–19 October 2000 in Nizzana: T1.5, air temperature at 1.5 m; Td 3 m, dewpoint at 3 m; Ts, surface temperature; u, wind speed at 3 m; Q*, net radiation. b Dew amounts. Arrows indicate sunrise and sunset. Data from a recording microlysimeter on the dune midslope are compared to those from 4 mm thick crust samples (closed triangles) at the footslopes and soil samples taken immediately below the exposed crust samples (–5 to –35 mm depth, open triangles). θ, volumetric moisture
air temperature (Fig. 21.1b) at about 20:30 local time, and yet dew formation had already started at about 16:00 local time, with sunset at about 17:00 local time. The explanation is that at very low soil moisture content, the relative humidity decreases sharply in the soil pores, which causes a high vapour pressure difference at the soil–atmosphere interface. Dew becomes bound in the soil capillaries and adsorbed by the soil. Once the dewpoint reached the surface air temperature, larger soil pores filled with dew and the dew formation rate increased.
21 Dew Formation and Activity of Biological Soil Crusts
21.3
311
Physiological Activity of Biological Soil Crusts
21.3.1
Activation of Soil Lichens After Nocturnal Wetting
Figures 21.2 and 21.3 provide examples of an extremely stable layer at a height of 0.2 m around 20:00 local time, coinciding with the onset of crust activity (8 and 26 March 1999) or preceding it (9 and 22 March 1999), and reaching 2 m a few hours later. Wind speeds decreased (<1 m s−1) to below the instrumentation threshold detection of calm conditions (approx. 0.2–0.4 m s −1). Under such conditions, the difference between the air and dewpoint temperature decreased to 2.5
D
20
2.0 1.5
15
1.0
10
0.5
5
0.0
0
11
E
300 200
9
100 7
0
B 2 1 0 -1 -2 -3
F
March 07/08
4
March 08/09 0
2
0
6
8
1200
wet
G
Leaf wetness
Local time EET [hours]
4
Fluorescence
16 18 20 22
12 8
.
.
C
Dewpoint temp. diff [°C]
−100 16
5
Net radiation [W m−2]
Temp. Gradient [K] Specific humidity [g g−1]
A
25
Wind speed [m s −1]
Air temp. [°C]
30
800 400
dry
0 16
18 20
22
0
2
4
6
8
Local time EET [hours]
Fig. 21.2 Early March 1999: microclimatic boundary conditions (A air temperature, B specific humidity, C temperature gradient between 20 cm and 2 m height, D wind speed at 20 cm height, E net radiation, F dewpoint temperature difference and G leaf wetness) during nights either with (8–9 March 1999) or without lichen activity (7–8 March 1999). Activity (expressed as minimal fluorescence F0 during the night and minimal fluorescence yield during the day) of Squamarina lentigera is also shown (solid line in G; after Veste et al. 2001)
312
M. Veste et al. 2.5
A
25
D
20
2.0 1.5
15 1.0
10
0.5
Specific humidity [g g−1]
300 200
9
100 7
0 −100 16
5
C
1
F
12
0 8
−2
March 22/23
−3
March 26/27 16
18 20
22
0
2
6
0 wet
8 1200
G
Leaf wetness
Local time EET [hours]
4
4
Fluorescence
−1
800 400
dry
0 16
Dewpoint temp. diff [°C]
2
0.0
E
B
Net radiation [W m−2 ]
0 11
Temp. gradient [K]
5
Wind speed [m s −1 ]
Air temp. [ °C ]
30
18
20
22
0
2
4
6
8
Local time EET [hours]
Fig. 21.3 Late March 1999: microclimatic boundary conditions (A air temperature, B specific humidity, C temperature gradient between 20 cm and 2 m height, D wind speed at 20 cm height, E net radiation, F dewpoint temperature difference and G leaf wetness) during nights either with (26–27 March 1999) or without lichen activity (22–23 March 1999). Activity (expressed as minimal fluorescence F0 during the night and minimal fluorescence yield during the day) of Squamarina lentigera is also shown (solid line in G; after Veste et al. 2001)
below 1.0 K, and the leaf wetness sensors simultaneously indicated dew condensation. This phenomenon is consistent with observations that specific humidity increased slightly after sunset, which is caused by a downward vapour flux in the near-ground boundary layer. Specific humidity then decreased immediately after the onset of stable conditions, indicating vapour removal from the air layer in the form of condensation at the surface. In general, maximum lichen activity always occurred when the dewpoint temperature difference reached 0 K (Fig. 21.4). Activity was recorded only during nights in which the dewfall was ≥0.1 mm. The in situ measurements indicated that maximum F0-values were attained 2.5–5.5 h after the initiation of dew (Veste et al. 2001). Kappen et al. (1979) found, for the fruticose lichen Ramalina maciformis at Avdat (Negev Highlands), that the maximum water content of the thalli was reached 0.5 to 6.5 h after sunset. The annual mean period of nocturnal dew imbibition sufficient to induce respiration was 8.8 h.
21 Dew Formation and Activity of Biological Soil Crusts
313
1.0
Relative activity
0.8
0.6
0.4
0.2
0.0 0.0
0.4
0.8
1.2
1.6
Dewpoint temp. difference [K] Fig. 21.4 Crust activity measured as minimal fluorescence yield (in relative units) versus difference between air and dewpoint temperatures in March 1999 (redrawn after Veste et al. 2001)
During nights when chlorophyll fluorescence measurements did not detect any crust activity, a different microclimatic situation was observed. During the night of 7–8 March 1999 (Fig. 21.2), a few hours of stability were followed by an advective increase in temperature and specific humidity, and subsequent evaporation of dew from the leaf wetness sensors. An exclusively advective situation with a warm south-westerly wind flow inhibited dewfall on 24–25 March 1999, while on 22–23 March 1999 (Fig. 21.3) the dewpoint temperature difference did not drop below the critical threshold. As a result, there was no leaf wetness sensor signal. On the other hand, 23–24 March showed 5 h of condensation, but crust activity was likely restricted by low dewfall amounts. According to a model approximating monthly dewfall in the area by Littmann and Veste (Chap. 13, this volume), dewfall at the experimental site in March 1999 was only 1.4 mm. This was for the 22 nights that showed hours with a critical dewpoint temperature difference of <1.0 K. Assuming a more or less constant condensation rate, average dewfall should have been around 0.06 mm per night. However, there were only seven nights when dewfall reached approximately 0.1 mm, of which three coincided with observed activity of the crust (8–9, 9–10 and 26–27 March 1999). This coincided with the occurrence of a leaf wetness sensor signal and the development of a near-ground stable air layer with light or calm airflow. The data presented here demonstrate consistent interrelations of activity and microclimatic parameters. Onset of lichen activity occurred only after maximum radiative cooling of this stable layer. Crust activity after nocturnal wetting reached maximum values only if the dewpoint temperature difference was exactly 0 K, i.e.
314
M. Veste et al.
when a few hours of cumulative dew within an extremely stable layer sufficiently hydrated the lichens. In general, the in situ dewfall measurements performed on different crusts and lichens indicate that activity may differ considerably because of variable hygroscopy and growth form. Once vapour is deposited on the surface by downward flux, it will be thermodynamically stable only in the case of droplet formation that requires large thermal fluctuations to overcome the cost in free energy of forming a liquid–vapour interface, i.e. to overcome interfacial tension (Beysens 1995). Natural surfaces tend to lower the barrier of free energy availability by their physical and chemical properties, such as hygroscopy that results in lower saturation pressure and small dewpoint temperature differences (Roedel 1992), and specific wetting conditions that depend on the contact angle of the droplet on the surface (Beysens 1995).
21.3.2
Photosynthetic Activity After Sunrise
Under field conditions, the nocturnal wetting of soil crusts will permit photosynthesis for about 2 h after sunrise (Fig. 21.5). This is because such activity is determined through a combination of light, temperature and humidity conditions. No significant difference in the drying process could be observed between the cyanobacterial lichen Collema tenax, Fulgensia fulgens and Squamarina lentigeria. However, cloudy weather after sunrise, especially after nocturnal rainfall, can reduce evaporation and extend the duration of crust activity by several hours.
March 25, 1996 Fulgensia Fluorescence F [rel. units]
800
rain
March 26, 1996
March 31, 1996
A
D
G
B
E
H
C
F
I
400 0
Squamarina
200 100 0
Collema
200 100
no activity
0 6
8
10
12
14
6
8
10
12
14
6
8
10
12
14
Local time EET [hours] Fig. 21.5 Measurements of minimal fluorescence yield after sunrise in Fulgensia fulgens (A, D, G), Squamarina lentigera (B, E, H) and Collema tenax (C, F, I) after rainfall (A to F) and nocturnal dewfall (G to I; after Veste et al. 2001)
21 Dew Formation and Activity of Biological Soil Crusts
315
Lichen wetting and, therefore, activity time for photosynthesis depends also on the slope aspect and shading of biological crusts. However, wetting of the crusts by rainfall does not necessarily mean that photosynthesis will occur for all soil-crust lichen species (Lange 2001). At very high thallus water contents, CO2 uptake by F. fulgens is reduced due to an increase of the thalli diffusion resistance (Lange et al. 1995) but this effect is less pronounced in S. lentigera (Lange et al. 1997a). Diploschistes diacapsis, like other Diploschistes species, exhibits no reduction in net photosynthesis under high water content (Lange et al. 1997a). The optimal water contents for net photosynthesis in F. fulgens and S. lentigeria are 0.25 and 0.42 mm respectively, in C. tenax the corresponding values are 1.1–1.3 mm, while in Diploschistes there is no distinct optimum (Lange et al. 1997a, 1998). The suppression of CO2 uptake by high water contents in some lichen species clearly demonstrates that CO2 exchange and photosynthesis cannot be related directly to the quantum yield measured by chlorophyll fluorescence. The chlorophyll fluorescence measurements give information only on physiological activity, and not on the photosynthesis rates of the lichen species investigated in this paper. A combination of CO2 uptake measurements and chlorophyll fluorescence 20
1.0
A
18 16
Leaf wetness
14 12
0.6
10 8
0.4
6
Dew point diff. [ K ]
0.8
4
0.2
2 0 18 16 14 12 10 8 6 4 2 0
B
0.8 0.6 0.4 0.2 0.0 12
16
20
0
4
8
Dew point diff. [ K ]
Leaf wetness
0.0 1.0
12
Local time EET [h] Fig. 21.6 Average of dewpoint temperature differences (squares) and leaf wetness (circles) on south- (closed symbols) and north-facing (open symbols) slopes in summer 1995 (A) and winter 1995/1996 (B; after Veste and Littmann 2006)
316
M. Veste et al.
(Lange et al. 1997b) is required to obtain unequivocal information on the photosynthesis and C gain of those lichens under natural conditions. Surface wetting duration of biological sand crusts may influence photosynthetic activity more than actual dew amounts. In winter, dew persists for longer periods in the study region on north-facing slopes of the linear sand dunes than on south-facing slopes, because of shading (Fig. 21.6), and in fact may not fully evaporate. This situation does not occur during the remainder of the year, when higher temperatures prevail. In the study sites presented here, biological soil crusts on the south-facing slopes are about 1.7 mm thick, whereas they are about 4.2 mm thick on the northfacing dune slopes. Shading around shrubs and bushes shields the topsoil somewhat, thereby delaying evaporation. This inhibits lichen desiccation, compared to surfaces exposed to sunlight immediately after sunrise. This may correspond to the development of different crusts within the sand dune ecosystem and along the geo-ecological gradient (Veste and Littmann 2006). The different photosynthetic response to rainfall and dew results in micro-scale variations in soil lichens. In general, green-algal soilcrust lichens have a lower moisture compensation point than do cyanobacterial lichens such as Collema. On the other hand, a high water content and water film on the thallus reduces net photosynthesis. In contrast, cyanobacterial lichens need greater amounts of liquid water for the activation of photosynthesis (Lange et al. 1993, 1998), and their water-holding capacity is also higher.
21.4
Conclusions
Dew is an important, regular supplier of water for soil lichens and soil crusts in the Haluza sand dunes. Lange et al. (1992) demonstrated (under laboratory conditions) that biological crusts composed of cyanobacteria and green algae were active at moisture levels > 0.1 mm. This is similar to the findings reported here, under field conditions for soil lichen crusts in the northern Negev dune field. Calculation of lichen activity based only on the difference between air temperature and dewpoint will be inadequate. Appropriate information about microclimatic boundary conditions and moisture fluxes are necessary for the modelling of the annual course of hydration/dehydration of biological crusts and lichens. The need to develop non-destructive approaches, such as remote sensing, is important to allow repetitive dew measurements on a given substrate. Spectral reflectance spectroscopy, which does not interfere with the surface wetting and drying process, is a step in this direction (Heusinkveld et al., 2008).
References Agam N, Berliner PR (2004) Diurnal water content changes in the bare soil of a coastal desert. J Hydrometeorol 5:922–933 Agam N, Berliner PR (2006) Dew formation and water vapor adsorption in semi-arid environments – A review. J Arid Environ 65:572–590
21 Dew Formation and Activity of Biological Soil Crusts
317
Armstrong S (1990) Fog, wind and heat: life in the Namib desert. New Scientist 127(1725):46–50 Berkowicz SM, Heusinkveld BG, Jacobs AFG (2001) Dew in an arid ecosystem: ecological aspects and problems in dew measurement. In: Schemenauer RS, Puxbaum H (eds) Proc 2nd Int Conf Fog and Fog Collection, 15–20 July 2001, St. John’s, Newfoundland, Canada, pp 301–304 Beysens D (1995) The formation of dew. Atmospheric Res 39:215–237 Bitan A, Rubin S (1991) Climatic atlas of Israel for physical and environmental planning and design. Ramot, Tel Aviv University, Tel Aviv Broza M (1979) Dew, fog and hygroscopic food as a source of water for desert arthropods. J Arid Environ 2:43–49 Cereceda P, Osses P, Larrain H, Fari M, Lagos M, Pinto R, Schemenauer RS (2002) Advective, orographic and radiation fog in the Tarapaca region, Chile. Atmospheric Res 64:261–271 Degen A, Leeper A, Shachak M (1992) The effect of slope direction and population density on water influx in a desert snail, Trochoidea seetzenii. Funct Ecol 6:160–166 Evenari M, Shanan L, Tadmor N (1982) The Negev: the challenge of a desert, 2nd edn. Harvard University Press, Cambridge, MA Garratt JR, Segal M (1988) On the contribution of atmospheric moisture to dew formation. Boundary-Layer Meterol 45:209–236 Goldreich Y (2003) The climate of Israel: observation, research and application. Kluwer/Plenum, New York, Dordrecht, London Heusinkveld BG, Berkowicz SM, Jacobs AFG, Holtslag AAM, Hillen WCAM (2006) An automated microlysimeter to study dew formation and evaporation in arid and semi-arid regions. J Hydrometeorol 7:825–832 Heusinkveld BG, Berkowicz SM, Jacobs AFG, Hillen WCAM, Holtslag AAM (2008) A remote optical wetness sensor using spectral reflectance spectroscopy. Agric Forest Meteorol (in press) Jacobs AFG, Heusinkveld BG, Berkowicz S (1999) Dew deposition and drying in a desert system: a simple simulation model. J Arid Environ 42:211–222 Jacobs AFG, Heusinkveld BG, Berkowicz S (2000a) Dew measurements along a longitudinal sand dune transect, Negev desert, Israel. Int J Biometeorol 43:184–190 Jacobs AFG, Heusinkveld BG, Berkowicz S (2000b) Force restore technique for surface temperature and surface moisture in a dry desert system. Water Resources Res 36:1261–1268 Jacobs AFG, Heusinkveld BG, Berkowicz S (2002) A simple model for potential dew-fall in an arid region. Atmospheric Res 64:285–295 Kappen L, Lange OL, Schulze E-D, Evenari M, Buschbom U (1979) Ecophysiological investigations on lichens of the Negev desert. Flora 168:85–108 Kidron GJ (1998) A simple weighing method for dew and fog measurements. Weather 53:428–433 Kidron GJ (1999) Altitude dependent dew and fog in the Negev Desert, Israel. Agric Forest Meteorol 96:1–8 Lange OL (2001) Photosynthesis of soil crust-biota as dependent on environmental factors. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management. Ecological Studies vol 150, Springer, Berlin Heidelberg New York, pp 217–240 Lange OL, Schulze E-D, Koch W (1970) Experimentell-ökologische Untersuchungen an Flechten der Negev-Wüste. III. CO2-Gaswechsel und Wassergehalt von Krusten- und Blattflechten am natürlichen Standort während der sommerlichen Trockenperiode. Flora 159:525–528 Lange OL, Kidron GJ, Büdel B, Meyer A, Kilian E, Abeliovich A (1992) Taxonomic composition and photosynthetic characteristics of the biological crusts covering sand dunes in the western Negev. Funct Ecol 6:519–527 Lange OL, Büdel B, Meyer A, Kilian E (1993) Further evidence that activation of net photosynthesis by dry cyanobacterial lichens requires liquid water. Lichenologist 25:175–189 Lange OL, Meyer A, Zellner H, Heber U (1994) Photosynthesis and water relations of lichen soil crusts: field measurements in the coastal fog zone of the Namib Desert. Funct Ecol 8:253–264 Lange OL, Reichenberger H, Meyer A (1995) High thallus water content and photosynthetic CO2 exchange of lichens. Laboratory experiments with soil crust species from local xerothermic
318
M. Veste et al.
steppe formations in Franconia, Germany. In: Daniels FJA, Schulz M, Peine J (eds) Flechten Follmann. Contributions to lichenology in honor of Gerhard Follmann. Geobotanical and Phytotaxonomical Study Group, Universität Köln, pp 139–153 Lange OL, Belnap J, Reichenberger H, Meyer A (1997a) Photosynthesis of green algal soil crust lichens from arid lands in southern Utah, USA: role of water content on light and temperature responses of CO2 exchange. Flora 192:1–15 Lange OL, Reichenberger H, Walz H (1997b) Continuous monitoring of CO2 exchange of lichens in the field: short-term enclosure with an automatically operating cuvette. Lichenologist 29:259–274 Lange OL, Belnap J, Reichenberger H (1998) Photosynthesis of the cyanobacterial soil-crust lichen Collema tenax from arid lands in southern Utah, USA: role of water content on light and temperature responses of CO2 exchange. Funct Ecol 12:519–527 Lange OL, Green TGA, Melzer B, Meyer A, Zellner H (2006) Water relations and CO2 exchange of the terrestrial lichen Teloschistes capensis in the Namib fog desert: measurements during two seasons in the field and under controlled conditions. Flora 201(4):268–280 Leisner JMR, Green TG, Lange OL (1997) Photobiont activity of a temperate crustose lichen: long-term chlorophyll fluorescence and CO2 exchange measurements in the field. Symbiosis 23:165–182 Littmann T, Veste M (2006) Determination of actual evapotranspiration and transpiration in desert sand dunes (Negev Desert) using different approaches. Forestry Stud China 8(1):1–9 Loris K, Jürgens N, Veste M (2004) Zonobiom III. Die Namib-Wüste im südwestlichen Afrika (Namibia, Südafrika, Angola). In: Walter H, Breckle S-W (eds) Ökologie der Erde, Band 2. Spezielle Ökologie der tropischen und subtropischen Zonen, 3. Aufl. Elsevier, Amsterdam, pp 441–513 Martin CE, von Willert DJ (2000) Leaf epidermal hydathodes and the ecophysiological consequences of foliar water uptake in species of Crassula from the Namib Desert in Southern Africa. Plant Biol 2:229–242 Moffett MW (1985) An Indian ant’s novel method for obtaining water. Natl Geogr Res 1:146–149 Monteith JL (1957) Dew. Q J R Meteorol Soc 83:322–341 Munne-Bosch S, Alegre L (1999) Role of dew on the recovery of water-stressed Melissa officinalis L. plants. J Plant Physiol 154:759–766 Munne-Bosch S, Nogues S, Alegre L (1999) Diurnal variations of photosynthesis and dew absorption by leaves in two evergreen shrubs growing in Mediterranean field conditions. New Phytol 144:109–119 Ninari N, Berliner PR (2002) The role of dew in the water and heat balance of bare Loess soil in the Negev Desert: quantifying the actual dew deposition on the soil surface. Atmospheric Res 64:325–336 Richards K (2004) Observation and simulation of dew in rural and urban environments. Progr Phys Geogr 28:76–94 Roedel W (1992) Physik unserer Umwelt, die Atmosphäre. Springer, Berlin Heidelberg New York Schemenauer RS, Cereceda P (1994) A proposed standard fog collector for use in high-elevation regions. J Appl Meteorol 33:1313–1322 Schroeter B, Green TGA, Seppelt RD, Kappen L (1992) Monitoring photosynthetic activity of crustose lichens using a PAM-2000 fluorescence system. Oecologia 92:457–462 Steinberger Y, Loboda I, Garner W (1989) The influence of autumn dewfall on spatial and temporal distribution of nematodes in the desert ecosystem. J Arid Environ 16:177–183 Veste M, Littmann T (2006) Dewfall and its geo-ecological implication for biological surface crusts in desert sand dunes (north-western Negev, Israel). J Arid Land Stud 16(3):139–147 Veste M, Littmann T, Friedrich H, Breckle S-W (2001) Microclimatic boundary conditions for activity of soil lichen crusts in sand dunes of the north-western Negev desert, Israel. Flora 196:465–476 von Rönsch H (1990) Tau und Reif in Harzgerode. Z Meteorol 40:197–204 von Willert DJ, Eller BM, Werger MJA, Brinckmann E, Ihlenfeldt H-D (1992) Life strategies of succulents in deserts with special reference to the Namib desert. Cambridge University Press, New York Waisel Y (1958) Dew absorption by plants of arid zones. Bull Res Council Israel D6:180–186 Zangvil A (1996) Six years of dew observations in the Negev Desert, Israel. J Arid Environ 32:361–371
Chapter 22
Nitrogen Input Pathways into Sand Dunes: Biological Fixation and Atmospheric Nitrogen Deposition R. Russow, M. Veste, S.-W. Breckle, T. Littmann, and F. Böhme
22.1
Introduction
In arid and semiarid regions, water availability is considered to be the controlling factor for the productivity and pattern of vegetation. The total biotic and abiotic N pool size of desert ecosystems is lower than in most other ecosystems (Skujins 1981). Several studies have found that even in arid lands, nutrients are critical for plant growth and successions (McLendon and Redente 1992). After good rainy years, nitrogen can become the limiting factor (Trumble and Woodroofe 1954) whereas added nitrogen increased productivity in several experiments in dry areas (Ettershank et al. 1978; Ludwig 1987). The main N input pathways into the ecosystems are atmospheric deposition in wet, dry and gaseous forms, and the biological fixation of atmospheric nitrogen N2. Biological fixation is carried out by free-living bacteria, Fabaceae–Rhizobium symbiosis and associative symbiontic free-living cyanobacteria, as well as by cyanobacteria in lichens. Another N source is by non-leguminous nitrogen-fixing species; particularly shrubs and trees play a major role in these ecosystems (Schulze et al. 1991; Valladares et al. 2002). In most drylands, the ‘biological soil crust’ influences ecosystem processes (West 1990; Belnap and Lange 2001; Veste et al. 2001a). Nitrogen-fixing cyanobacteria of the genera Nostoc, Microcoleus, Chroococcus and Calothrix are common in such soil crusts, and have been reported in places such as arid and semiarid regions of Australia, North America and the Negev (Lange et al. 1992; Zaady et al. 1998; Belnap 2001; Chap. 10, this volume) Soil lichens with cyanobacterial phytobionts are also able to fix nitrogen. The importance of biological N fixation by soil crusts has been emphasised by several authors (e.g. Shields et al. 1957; Rychert and Skujins 1974; West 1990; Evans and Ehleringer 1993; Zaady et al. 1998), although determining N fixation under field conditions has several methodological problems (West 1990). As a result, high variation has been revealed among different drylands investigated under simulated field conditions. An increase in total nitrogen has been observed beneath the soil lichens and crusts of the sand dunes of the north-western Negev (Veste et al. 2001a). Most of the information about their contribution to N input in different dry ecosystems results only from laboratory investigations or simple estimates based on crust development
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
319
320
R. Russow et al.
and N content. Mainly known are acetylene reduction and 15N2 incubation assays performed in laboratory investigations simulating field conditions (Mayland et al. 1966; Rychert and Skujins 1974; Rai et al. 1983; Zaady et al. 1998; Belnap 2002; Arnibar et al. 2003). Measurements of N fixation by cyanobacteria in the biological crust of the Negev desert are unknown, apart from a laboratory study using an acetylene reduction assay (Zaady et al. 1998). A common way of determining the biological N2 fixation of wild legume– Rhizobium symbioses is the natural 15N abundance method (Shearer et al. 1983; Shearer and Kohl 1989; Virginia et al. 1989; Högberg 1997; Boddey et al. 2000; Dawson et al. 2002). The advantage of the natural 15N abundance method is that it can be easily applied at any site without the need for additional 15N labelling. The disadvantages causing some limitations are the relatively low natural 15N enrichment of nitrogen derived from soil (NdfS), compared to nitrogen derived from the atmosphere (NdfA), and some additional isotopic effects (e.g. Högberg 1997; Roth 1997; Dawson et al. 2002; Russow et al. 2004). However, as far as Fabaceae are concerned, especially woody plants and trees in a natural environment, this method is the only practical way to assess N2 fixation in situ (Binkley et al. 1985; Boddey et al. 2000), which is why it was used here to determine the biological N fixation (BNF) of the Fabaceae Retama raetam, a common shrub in the Negev and beyond (cf. below). R. raetam could be an important nitrogen fixer in such dry ecosystems but so far, nodulation and, therefore, nitrogen fixation have been observed only in the greenhouse, not under field conditions (Farnsworth 1975). In this paper, we present field measurements of biological N2 fixation (BNF) obtained by the natural 15N abundance method, and use these to estimate the annual nitrogen input by the soil crusts and R. raetam. We follow a novel approach for the natural 15N abundance technique, by using the non-N2-fixing lichens Squamarina lentigeria and S. cartilaginea (=S. crassa) as reference in order to determine N2 fixation by the biological crust in situ in the Negev desert. N input by BNF of atmospheric nitrogen is compared with atmospheric nitrogen deposition.
22.2 22.2.1
Materials and Methods Study Sites
The samples were collected along the geo-ecological gradient described in Chapter 29 (this volume).
22.2.2
Species Investigated
R. raetam (Fabaceae) is a stem-assimilating shrub with a height of up to 2.5 m. It has a wide ecological range, from the Mediterranean coastal dunes to the dry Saharo-Arabian
22 Nitrogen Input Pathways into Sand Dunes
321
deserts of North Africa. In the Negev, it is a common species of the loessial northern Negev, the central Negev Highlands, the stony southern Negev and the arid Arava Valley. In the sand dunes of the north-western Negev, R. raetam grows in interdune areas and on stable dune slopes and crests. It is a deep-rooting shrub, its roots sometimes reaching a length of more than 11 m. Artemisia monosperma (Compositae) is a semishrub up to 1.5 m in height, and characteristic of desert and coastal sand dunes in the Middle East (Danin 1996). Artemisia inhabits semi-stable and stable sands. The stemsucculent Anabasis articulata (Chenopodiaceae) is a semi-shrub. It is common in the study area in the interdunes and dune base.
22.2.3
Biological Soil Crusts
Biological soil crusts consist of cyanobacteria, green algae and local outcrops of mosses and soil crust lichens (Lange et al. 1992; Evans and Lange 2001; Veste et al. 2001a, b). The most common cyanobacteria in biological crusts include Nostoc sp., Scytonema sp., Phormidium sp. and Oscillatoria sp. (Chap. 10, this volume). The interdunes in the Haluza sand field (site N3) are completely covered with soil lichen crusts. This lichen community is composed of Fulgensia fulgens (Teloschistaceae), S. cartilaginea (Lecanoraceae), S. lentigera and Diploschistes diacapsis syn. D. steppicus (Diploschistaceae). The nitrogen-fixing soil lichens are Collema tenax and other unidentified cyanobacterial lichens. In most of the interdunes, the cyanobacterial lichens dominate the soil crusts.
22.2.4
Sampling for 15N Determination
Samples of biological crusts and the underlying sand to a depth of 90 cm, lichens (C. tenax, F. fulgens, S. lentigeria and S. cartilaginea) and shrubs (R. raetam, A. articulata, A. monosperma) were collected in March 1998, 1999 and 2000 at different locations. Samples of the soil, crusts and R. raetama were taken in three replicates from two different plots at each location. Shrub samples were cut from two different shrubs at each location. In 2000, additional samples were taken from plants with the aim to confirm the measurements of 1998 and 1999. The material was oven-dried at 65 °C and then ground to determine N content and also natural 15N abundance.
22.2.5
Sampling Atmospheric Deposition
Total nitrogen in the bulk samples were analysed by Kjedahl digestion following steam distillation to expel the ammonium. 15N of the isolated ammonium was measured by means of a coupled element analyser-IRMS (Scharf 1988; Mulvaney 1993; Russow et al. 2004).
322
22.2.6
R. Russow et al. 15
N Methodology and Calculation of Biological N Fixation
15
N abundance in the pretreated samples was measured on a DELTA S isotope mass spectrometer (Finnigan MAT Bremen, Germany) coupled online to a CHN EA 1108 elemental analyser (Carlo Erba, Italy; see details in Russow et al. 2004). The natural 15N abundances measured are given in δ ‰ units (Ehleringer and Rundel 1989): δ
‰ = (Rs–R0) /R0 × 1,000
(22.1)
where Rs is the 15N/14N ratio of the sample, R0 that of atmospheric N2. The 15N method to determine the biological N2 fixation of legume–Rhizobium symbioses introduced by McAuliffe et al. (1958) was originally a 15N isotope dilution technique based on artificially tracing the plant-available N pool with 15N. The different variations of this technique are explained in Russow and Faust (1990). Because the δ15N signature of the nitrogen derived from the soil (NdfS) is commonly enriched relative to nitrogen derived from the atmosphere (NdfA), this 15 N soil enrichment can be used as a natural tracer. This approach is known as the natural 15N abundance method, and is commonly used in natural and semi-natural ecosystems (Shearer et al. 1983; Shearer and Kohl 1989; Virginia et al. 1989; Boddey et al. 2000; Dawson et al. 2002). NdfA can be calculated using a two-pool model from the quotient of the natural 15N abundance of the N2-fixing plant and the plant-available soil N. Because representatively determining the plant-available soil N is difficult, a non-N2-fixing reference plant growing at the same site with a similar root system and temporal N uptake pattern to that of the N2-fixing plant is often used. However, frequently these assumptions are not met. Hence, the 15N abundance of the reference plant should differ from the available soil N by up to ± 3 ‰ (Evans 2001). Using both a reference plant and the soil nitrogen pool, the N2 fixation expressed as NdfA can be calculated as NdfA = (δr – δf / δr – δb ) × 100
(22.2)
where δr is the δ15N of the reference sample in ‰, δf that of the N2-fixing plant in ‰, and δb the isotopic shift relative to the air’s δ15N signature by the BNF as such. BNF is known to create a slight isotopic 15N shift to about −2‰ (Boddey et al. 2000) but this is not known for R. raetam. Therefore, in Eq. (22.2) for δb we used a very small value of −0.5‰ for BNF calculations, to avoid an excessive overestimate of BNF for R. raetam. The precision of this approach depends primarily on the difference between the 15N in the soil and the atmosphere. Therefore, estimates based on the natural 15N abundance method are often only semi-quantitative (Dawson et al. 2002), because of the additional uncertainty in determining the true δ15N value of plant-available soil N (NdfS), as mentioned above. This calculation
22 Nitrogen Input Pathways into Sand Dunes
323
is explained in more detail in Shearer and Kohl (1989), and a more recent review for woody plants is given in Boddey et al. (2000). The NdfA of the biological crust can be calculated using an equation similar to Eq. (22.2). The parameter δb in Eq. (22.2), i.e. the isotopic shift relative to the air’s δ15N signature (0 ‰) caused by the BNF as such, is also unknown for biological crusts and N2-fixing cyanobacteria. Therefore, to avoid overestimating the crust’s BNF, here δb was set to ± 0 ‰, and Eq. (22.2) becomes Eq. (22.3): NdfA = (δ L – δC / δL) × 100
(22.3)
where δC is the δ15N of N2-fixing cyanobacteria in ‰, δL that of non-N2-fixing lichens in ‰.
22.3 22.3.1
Results Soil
Figure 22.1 shows the interrelation between N contents and 15N abundances at various soil depths. Like other investigators (e.g. Skujins 1981; Evans and Ehleringer 1993; 15
N - abundance [%] 4 8 12 16
0
20
0 −10
soil depth [cm]
−20 −30 −40 −50 −60 −70 −80 −90 0.00
0.02
0.04
0.06
0.08
0.10
0.12
N-content [%]
•
Fig. 22.1 N contents ( ) and 15N abundances (n) at various soil depths in the Haluza sands (site N3)
324
R. Russow et al.
Nadelhoffer and Fray 1994), we found a decrease in N content with soil depth. In contrast, 15N abundance shows an increase with soil depth. As we measured 15N in the surface crusts separately, there is a marked hiatus in 15N abundances from about −2 ‰ in the crusts to values of +4 to +7 ‰ in the underlying sand.
22.3.2
Biological Soil Crusts
The δ15N values of the biological soil crusts are all negative but differ significantly among the various species sampled (Fig. 22.2). The non-N-fixing soil crust lichens of the genus Squamarina showed a very low δ15N value of −11‰. Because the Squamarina types cannot fix N2, we used these as reference in Eq. (22.3), to estimate the BNF of the various soil crust types in the Haluza sands. Crust NWS Crust ID
Coll ID
Ful ID
Squa ID
N content [kg N kg−1 dw]
Natural 15 N abundance [‰]
0 −2
.
−4 −6 −8 −10
A
0.6
1998
B
1999
0.5 0.4 0.3 0.2 0.1 0.0 Crust NWS Crust ID
Coll ID
Ful ID
Squa ID
Fig. 22.2 Natural 15N abundances (A) and N contents (B) for biological crusts and lichens at the Haluza sands. Crust NWS Biological crust on north-west-facing slope, Crust ID biological crust in the interdune, Coll ID Collema tenax sp. in the interdune, Ful ID Fulgensia fulgens in the interdune, Squa ID Squamarina sp. in the interdune
22 Nitrogen Input Pathways into Sand Dunes
325
The thickness and density of the crust and lichens sampled are listed in Table 22.1. These values and the N content of the samples (Fig. 22.2) can be used to calculate the stock of nitrogen, which is shown in Fig. 22.3a. As shown, the biological crusts and soil lichens contain considerable amounts of N which increase with increasing thickness, from 5–7 g N m−2 in cyanobacterial crusts at the dune base and interdune to 27–40 g N m−2 in the soil lichen crusts. The thickness of the crust varies greatly from one location to the other – from 1 to 4 mm for the biological soil crusts and from 4 to 7 mm for the lichen crusts. Therefore, the N amounts calculated (Fig. 22.3a) are valid only for the particular sampling spots of the present study, and generally exhibit high variation. The BNF of the biological soil crusts and soil lichens, calculated as relative NdfA values based on Eq. (22.3), are shown in Fig. 22.3b. Table 22.1 Thickness (x) and density (ϕ) of biological crust types at different locations within the Haluza sands (ID interdune, NWS north-west-facing slope) Species (location)
Thickness x (mm)
Density ϕ (g×cm−3)
Remarks
Squamarina species (ID) Fulgensia fulgens(ID)
4.0
1.1
4.5
1.1
Collema tenax (ID)
5.0
1.1
Biological soil crust (ID) Biological soil crust (NWS)
3.0
1.3
3.0
1.3
Photobiont: green algae Photobiont: green algae contaminated by cyanobacteria Photobiont: cyanobacteria Cyanobacteria and green algae Cyanobacteria and green algae
Fig. 22.3 N stock in different crusts and soil lichens (A) and relative biological N2 fixation (NdfA) (B) by soil crusts and lichens in the Haluza sand field. Crust NWS Biological crust on north-west-facing slope, Crust ID biological crust in the interdune, Coll ID Collema tenax in the interdune, Ful ID Fulgensia fulgens in the interdune, Squa ID Squamarina sp. in the interdune
rel. biological N fixation [%]
N stock [g m−2]
40
1998 1999
A
30 20 10 0 100
B
80 60 40 20 0 Crust NWS Crust ID
Coll ID
Ful ID
Squa ID
326
R. Russow et al.
22.3.3
15
N Retama raetam
in ‰
28
15N
Figure 22.4a shows the δ15N values for the N2-fixing shrub R. raetam, compared to those of the non-fixing plant A. articulata and the soil itself. Retama shows significantly less 15N enrichment than the non-N2-fixing reference plant A. articulata and the soil, proving the uptake of atmospheric nitrogen by R. raetam via nodulation with Rhizobia. The only references available were the non-N2-fixing plant A. articulata and the soil as such. The deep-rooted shrub A. monosperma has a strikingly lower, positive 15N abundance in the study area, probably due to
24
A
Retama Anabasis
20
Soil 10-60 cm
16 12 8 4 0 NdfA(Anabasis)
NdfA(Soil)
100 %
B 80 60 40 20
99 s/ B D
4
sw
/9
9
98 ID 4
4
D
B
sw
s/
/9
8
99 4
ID
ID 3
3
ID
e/
n/
99
98 n/
98
ID 3
ID
e/
s/ 3
ID 2
2
ID
n/
99
99
0
Sampling location / Year
Fig. 22.4 Natural 15N abundance for Retama raetam, Anabasis articulata and soil samples (A) and relative biological N2 fixation (NdfA) (B) at different locations. ID Interdune, DB dune base
22 Nitrogen Input Pathways into Sand Dunes
327
phyllospheric N2 fixation from the atmosphere (Farnsworth 1975), and so could not be used as a reference plant (Russow et al. 2004). NdfA values, i.e. the relative contribution of BNF to biomass production by R. raetam, were calculated according to Eq. (22.2) (Fig. 22.4b). Using the soil as reference entails additional uncertainties for the determination of BNF with the natural 15N abundance method. This results from an increase in δ15N values with soil depth (Russow et al. 2004). To improve the soil-based calculation of BNF, an average of three soil depths (5–10, up to 30, up to 60 cm) was used (Fig. 22.4a). The error bars for the soil shown in Fig. 22.4a reflect the mean range of 15N abundance at a soil depth of 5–60 cm. The standard deviation of 15N abundance for R. raetam and this 15N range in the soil allows us to calculate the uncertainties in the determination of NdfA for the soil-based NdfA values, shown in Fig. 22.4b as error bars, using the rules of error propagation. Compared to the standard deviation of the Anabasis-based values, this uncertainty is very large. On the other hand, the reference values determined with A. articulata may also be distorted by airborne N deposition with a strongly negative δ15N value. This can be concluded from the low δ15N values of Squamarina. The fixation rate NdfA calculated depends on the reference used (Fig. 22.4b). However, since the two NdfA values coincide with the standard deviation of the Anabasis-based values, the mean of the two is used for further analysis and discussion (see below).
21.3.4
Estimation of N Input by BNF into the Ecosystem
As mentioned above, the NdfA provides only relative values and does not reflect absolute N input. N input can be calculated only by additionally using data on total N contents and annual growth rates of the pools considered. No growth data are available for undisturbed crusts. Therefore, we started by using the regeneration time for the soil crust from an experiment in which large areas of the crust in the interdunes at the Nizzana test site had been completely removed. The subsequent increase in chlorophyll content with time after removal was measured (Veste et al. 2001a). The average NdfA values and N pools were taken from Fig. 22.3, the growth rate was estimated from the experiment mentioned above, and the absolute biological N2 fixation calculated from these values is shown in Table 22.2 (also see below and Table 22.3). The latter ranges from 0.9 to 1.3 g m−2 year−1 for the cyanobacterial crusts and up to 4 g m−2 year−1 for C. tenax. According to Eq. (22.4), calculating absolute nitrogen input by biological nitrogen fixation from NdfA for R. raetam requires the production of biomass (m) per area and year based on the relative fraction of R. raetam (pR) and mean N content of R. raetam (cR): BNF = NdfA × m × pR × cR
(22.4)
Standing biomass (BM), the fraction of R. raetam (pR) and the average N content (cR) were determined at location 4, Nizzana (Table 22.4). For the estimation, we
328
R. Russow et al.
Table 22.2 Biological N2 fixation by biological soil crusts and soil lichens in the Haluza sands (ID interdune, NWS north-west-facing slope) Crust type (location)
N stock (g N m−2)
NdfA (%)
N growth (g N m−2 Abs. BNF (g N m−2 year−1) year−1)
Fulgensia fulgens(ID)a Collema tenax (ID) Biological crust (ID) Biological crust (NWS)
17.7 34.0 5.4 6.5
68 88 91 84
2.5 4.9 1.4 1.1
a
1.7 4.3 1.2 0.9
Fulgensia fulgens contaminated by N-fixing cyanobacteria
Table 22.3 Calculated nitrogen fixation (NdfA) of Retama raetam at different locations and years with Anabasis articulata and soil (5–60 cm) as a reference (ID interdune, DB dune base, SP span of uncertainty) Location Year NdfA in % Average SP N5/North (ID) N3/Haluza (ID) N3/Haluza (ID) N1/Nizzana (ID) N1/Nizzana (ID) N1/Nizzana (DB)
1999 1998 1999 1998 1999 1998
87 75 63 88 71 46
29–100 ±33 ±28 ±64 ±53 ±36
Table 22.4 Estimation of N amounts fixed from the atmosphere by Retama raetam in the interdunes and dune base at Nizzana Parameter Interdune Dune base Standing biomass (BM, kg ha−1)a Estimated biomass production (kg ha−1 year−1)b Relative fraction (pR, %) Mean N content (cR, % of dry weight)c Increase of N (g ha−1 year−1) Mean NdfA (%) Estimated BNF (g ha−1 year−1) a Veste et al. (2000; also see Chap. 25, this volume) b Growth rate Retama raetam 4% (after Tenbergen 1991) c Stratmann (1996; for details, see text)
1,850 74 1.7 1.62 20.4 79 16
3,800 152 9.5 1.62 234 46 108
assumed an annual growth rate of 4% for R. raetam, following investigations in the Negev Highlands by Tenbergen (1991). This relative growth rate results in a biomass production per year (m) for R. raetam of 74 kg ha−1 year−1 in the interdune and 152 kg ha−1 year−1 at the dune base of the Nizzana test site. From these biomass production rates and the relative BFN determined, an N input of 16 and 108 g N ha−1 year−1 respectively was estimated (Table 22.4).
22 Nitrogen Input Pathways into Sand Dunes
22.3.5
329
Atmospheric Nitrogen Deposition
The general solution for atmospheric element deposition derived from principal component analysis (PCA) in Chap. 19 (this volume; cf. Table 22.2) indicated that nitrogen input during the September 1998 to June 1999 study period (for further details and sampling methods, refer to Chap. 15, this volume) may be grouped into two principal components: PC3 (NH4+-N, total N deposition, and with K+ loading secondarily on this factor), and PC4 (NO3- -N, with SO42− loading negatively). In this way, the deposition of nitrogen compounds in the study area is not correlated with atmospheric input of other mineral elements. At the five monitoring stations in the northern coastal plain and the sand dune areas (Chap. 15, this volume), there is a sharp decrease in the deposition of Ntotal (determined by means of the Kjeldahl method) from north to south (cf. Fig. 15.1 in Chap. 15, this volume). For the northern coastal plain, we observed a mass input of Ntotal of 4.27 and 4.14 kg ha−1 year−1 at the Gevulot and Yevul stations respectively, whereas in the sandy area the corresponding values decreased from 3.44 kg ha−1 year−1 at site N5 to 2.17 kg ha−1 year−1 at site N3, and 1.86 kg ha−1 year−1 at site N1 (Nizzana; Fig. 22.5). On the other hand, the decrease in overall N deposition is accompanied by a decrease in the organic fractions. In the cultivated coastal plain with predominantly agricultural activities, NH4+-N and NO3−-N constitute 25–10% of Ntotal, and in the sand dune field the values increase from 16% along the northern margin to 20% in the central part, and 40% in the southern part. The increase in the fraction of inorganic N deposition is due mostly to NO3−-N input which, in turn, is primarily of dry mode (cf. Table 19.1, Chap. 19, this volume). However, NH4+ is most effectively deposited during wet intervals. Both principal components of nitrogen deposition show a clear maximum in winter at all stations (cf. Table 15.1, Chap. 15, this volume), followed by the spring months. Due to the limited monitoring time series, we cannot infer seasonal characteristics for the summer period. However, Littmann (1997) found summer values
12
6 .
8
4 4
2
15
N - abundance [o/oo]
8
0
N-Deposition [kg N kg−1 a−1]
16
0 1
10
2
20
3
30
4
40
50
Distance from coast [km] Fig. 22.5 Bulk N deposition (bars) and mean 15N abundance (line) along the geo-ecological gradient
330
R. Russow et al. 1,50
PC 3 model
factor values
1,00
0,50
0,00
-0,50
99
99
19 6.
07
.0
99
19 5.
17
.0
99
19 4.
26
.0
99
19 4.
05
.0
99
19 3.
15
.0
99
19 2.
22
.0
99
19 2.
01
.0
98
19 1.
11
.0
98
19
19
2.
21
.1
98 1.
30
.1
98
19
19
1. .1
09
0. .1
.0 28
19
9.
19
98
-1,00
Fig. 21.6 Modelling the mean depositional series for ammonium (PC3) and nitrate (PC4) in the north-western Negev. Model regression equations: NH4+−N=0.067 N−0.14 NE+0.012 E−0.03 SE+0.009 S+0.075 SW+0.155 W−0.0095 NW−0.001 WS2−0.21 WS3+0.24 WS4−0.363, where wind directions refer to the percentage of a given direction within a sampling interval, and WS2, WS3 and WS4 are the percentage of wind speeds in the 2.0–3.9, 4.0–5.9 and >6.0 m s−1 intervals per sampling interval respectively; NO3−−N=0.008 N+0.163 NE+0.006 E+0.017 SE−0.147 S+0.251 SW−0.243 W+0.029 NW−0.084 WS2−0.007 WS3−0.123 WS4+3.256
of total nitrogen input to be very low over a longer observational period for the southern sand dune field, and the seasonal maximum occurred in spring (April, May). Also in our depositional time series, in spring we can observe a gradual increase in the mass of total nitrogen deposited per month from north to south. As was found earlier (Littmann 1997), the deposition of nitrogen compounds is not interrelated with dust input. Thus, dust does not appear as an independent variable in multiple regression models for the factor value series of PCs 3 and 4. PC3, on the other hand, is most effectively deposited under wet conditions and, therefore, rainfall was included in a first-step model run as a controlling variable (explained variance of PC3 factor value series: 37%; Fig. 22.6). However, upon introducing wind parameters into the stepwise multiple regression model, rainfall was replaced by the entire set of wind directions. The best-fit model for PC3 (explained series variance: 82%) includes N, E, S, SW and W as positively interrelated variables, and NE, SE and NW with negative interrelation. Only the highest wind speed interval (the relative frequency of wind speeds>6.0 m s−1 per sampling interval) was included in the model equation. The deposition of nitrate (almost entirely of dry mode, following Table 19.1 in Chap. 19, this volume) seems to follow boundary conditions other than climatic,
22 Nitrogen Input Pathways into Sand Dunes
331
compared to the deposition of total nitrogen and ammonia. In a stepwise multiple regression model, all wind parameters were introduced in the first step, with only wind from westerly directions being negatively interrelated. However, all wind speed intervals > 2 m s−1 are also negatively interrelated. Thus, the overall significance of the regression model (Fig. 22.6) is weak (explained variance: 47%).
22.4
Discussion
As pointed out in the introduction, measurements of N2 fixation by cyanobacteria in the biological soil crusts of the Negev desert are unknown, with the exception of laboratory studies using the acetylene reduction assay (Zaady et al. 1998). Our results with the natural 15N abundance method showed clearly that the different biological crusts are able to fix a significant amount of nitrogen. Measuring the natural 15N abundance of the non-fixing crustal lichens S. lentigeria and S. cartilaginea, compared to that of the cyanobacterial soil crusts and cyanolichens, has enabled us to estimate fixation under field conditions. This novel approach for the in situ determination of N2 fixation by biological crusts in the field is explained in the following. Assuming that no soil N moves upwards from the crust into the soil, the crusts obtain their nitrogen solely from two N pools: (1) atmospheric N2 fixed by the cyanobacteria present in the biological crust (BNF); (2) airborne N deposition. The lichens S. lentigeria and S. cartilaginea – which do not contain cyanobacteria – cannot fix N2 and, consequently, obtain their nitrogen only from airborne N deposition. This assumption is supported by other investigations in which lichens were used to monitor air pollution (Hawksworth and Rose 1976; Ahmadjian 1993; Stolte 1993). As shown in Fig. 22.2, the δ15N of these non-N2-fixing lichens is strongly negative down to an average of −11‰, probably caused by the absorption of strongly negative airborne nitrogen in either gaseous form (NH4+, NOx) or from rainwater uptake (NH4+, NO3−; Heaton 1990; Freyer 1991; Garten 1992). Although natural 15N values for lichens in arid regions are unknown, measurements in other regions support our assumption. δ15N values of the lichens Hypogymnia physodes and Pseudevernia furfuracea from the eastern central Alps range from −4 to −7‰ , depending on altitude (Schlee et al. 1996), even reaching −14‰ for H. physodes in a forest in central Germany (Jung, personal communication). Comstock (2001) found values of between −4 and −8‰ in epiphytic lichens in the Kings Canyon and Glacier National (USA). Because of the relatively large difference between the δ15N value of the two pools (atmospheric nitrogen 0 ‰), the natural 15N abundance technique should also be applicable to determine the fixation of N2 from air by biological crusts and cyanobacteria-containing lichens. In this sense, the negative 15N value of airborne N deposition is used as an N tracer, the non-N2-fixing lichens being used as reference. In contrast to Squamarina, the relatively high δ15N values of −7‰ in F. fulgens (also unable to fix N2) living at the same spot indicate contamination by free-living cyanobacteria, resulting in a significant N2 fixation of 68%. We have no proof that
332
R. Russow et al.
there is no N2 fixation in Squamarina. Because its δ15N is very negative, however, there is evidently no or merely negligible contamination by free-living cyanobacteria, and thus this species can be used as reference for the natural 15N abundance of airborne N deposition. The calculated relative contribution of atmospheric N2 (NdfA) by cyanobacteria to the total nitrogen content is very high, 84–91%. The cyanolichens C. tenax investigated here have a similar fixation rate (88% NdfA). Our estimated absolute biological N2 fixation based on 15N values amounts to between 0.9 and 1.2 g m−2 year−1 (9 and 12 kg N ha−1 year−1) for the cyanobacterial crusts, and up to 4.3 g m−2 year−1 (43 kg N ha−1 year−1) for the C. tenax this is in the range of 1–10 g m−2 year−1 published by other authors (Rychert and Skujins 1974; West and Skujins 1977; West 1990, 1991). Under optimum light and moisture conditions in a laboratory experiment, Zaady et al. (1998) determined an N2 fixation rate of max. 95 g N cm−2 h−1 for biological crusts from the Negev. From our own measurements of the activity of biological crusts in function of dewfall and rain, an average effective fixation period of 45 and 150 minutes can be concluded after a dew or rain event respectively (Veste et al. 2001b). For the study site, we can assume approx. 195 days with sufficient dewfall and an average of 15 rainy days, resulting in roughly 190 h year−1 with nearly optimum fixation conditions. Based on the above potential fixation rate of max. 9.5 mg N m−2 h−1, this effective time would enable total fixation of 1.8 g N m−2 year−1 by the crusts. This value is in the lower fifth of the range of BNF estimated for biological crusts in other desert ecosystems, i.e. 1 to 10 g N m−2 year−1. However, our results amount to 50–67% of the above-estimated potential N2 fixation by biological crusts (without C. tenax) of the Negev. To assess this result, the high variation between the different drylands must be taken into account. Investigations under simulated field conditions have found values ranging from up to 3.5 g N m−2 year−1 for general drylands (McGregor and Johnson 1971) and 4.0–6.6 g N m−2 year−1 for semiarid rangeland in Arizona (Mayland et al. 1966) to 1–10 g N m−2 year−1 in the Great Basin (Rychert and Skujins 1974; West and Skujins 1977). West (1990, 1991) reported that, in the Great Basin, nitrogen fixation by cryptogamic crusts can reach 4.1 g N m−2 year−1. In a more recent publication, Belnap (2002) reported for dark cyanobacterial crusts, and a crust with 20% cover of C. tenax, an annual N input by BNF of 9 and 13 kg N ha−1 year−1 respectively in the Canyonlands National Park, south-eastern Utah, USA. These values were estimated from results of acetylene reduction measurements in the laboratory, similarly to our approach above with the laboratory measurements of Zaady et al. (1998). This approach has two main uncertainties: 1. a high variability in the factor used to convert acetylene values into real N2 fixation levels; Belnap (2001) herself reported a range of 0.06–0.38 for this factor; 2. extrapolation from this laboratory value, obtained under optimum fixation conditions, to N2 fixation in the field, accounting for real temperature, precipitation and moisture characteristics for the crusts measured in the field for an observation period of 2 years.
22 Nitrogen Input Pathways into Sand Dunes
333
Long-term measurements for annual crust activity and its relation to microclimate conditions are rare. Therefore, exact estimation of physiological activity from microclimate conditions still has uncertainties. Despite these, the different approaches used by Belnap (2002) and in our investigations result in the same range of values. Biological crusts cover, in the interdunes, 55% of the area at Nizzana (site N1), 75% in the Haluza sands (site N3) and 93% at site N5. From the crust covers, we can calculate a nitrogen input of approx. 13 kg ha−1 year−1 for the crusts investigated at the Halzua sands. These are average values for nitrogen input by biological crusts. It needs to be considered that, on a smaller scale, crust distribution is very patchy and, therefore, up-scaling from smaller spot measurements to ecosystems is another problem. In contrast to the biological soil crusts, which cover most of the sand dune, nitrogen fixation by R. raetam has a smaller, more localised impact. Retama contributes only 1.7% of the standing biomass in the interdune, and the highest density can be found at the dune base (9.5%). Accordingly, total N input via BNF by this shrub is only 16 and 108 g N ha−1 year−1 (0.002 and 0.011 g N m−2 year−1) respectively, although the relative contribution of BNF to the total nitrogen of R. raetam is relatively high, 46–79%. Shearer et al. (1983) estimated a similar relative contribution of BNF to the total nitrogen of Prosopis in various desert ecosystems (43–65%). Nevertheless, the nitrogen content of the detritus under such shrubs may be elevated, leading to heterogenic N distribution within the ecosystems. In combination with a more suitable microclimate, a denser cover of annuals can very often be observed and, thus, more biomass can be found in these ‘fertile islands’ or ‘fertile patches’ than in their surroundings (Gardener and Steinberger 1989; Pugnaire et al. 1996; Xie and Steinberger 2001, 2002). The overall atmospheric deposition of total nitrogen is comparatively low but typical for semiarid to arid desert margins. It decreases, in parallel with biomass and agricultural activity, from around 4 kg ha−1 year−1 in the northern coastal plain to 2 kg ha−1 year−1 in the southern sand dune field; simultaneously, the percentage of inorganic compounds increases. Furthermore, both PC3 and PC4 show a seasonal maximum during the winter months from November to March. Although peaks in the factor value series in winter and spring do show some coincidence with either rainfall or dust storm events, only rainfall may be considered a controlling variable for PC3 (Ntotal and NH4+) input. More significant in multiple regression models are wind directions and wind speeds which, however, do not indicate any clear source area or preferred wind direction to which regional nitrogen deposition may be assigned. These findings imply that different levels of atmospheric deposition of total nitrogen, and of ammonia as the first product of the mineralization of organic compounds, are confined to specific local environments: they are higher in the northern area, with a relatively high biomass and agricultural land use, and lower in the arid southern parts of our transect. In the event of rainstorms with high wind speeds and atmospheric instability, organic and primarily mineralized plant and soil particles may be blown off and redeposited within the affected area. No long-range transport may occur. The deposition of nitrates, however, is constrained to calm conditions which may also include rain and dust storm events
334
R. Russow et al.
during a sampling interval. As no specific wind direction may predominate, this implies that NO3−-N deposition is largely in a gaseous deposition mode also constrained to local environments.
22.5
Conclusions
From our field investigations, we conclude that the novel approach using the nonfixing lichens Squamarina lentigeria and S. cartilaginea as reference for the natural 15 N abundance method results in reasonable values for the fixation of atmospheric N2 by the biological crusts and cyanobacteria-containing lichens. On the ecosystem level, biological fixation by biological crusts and Collema tenax lichens of approx. 13 kg N ha−1 year−1 is a very important nitrogen input pathway, whereas N input by dust can be considered a minor pathway with only 2–4 kg N ha−1 year−1 in the sand dunes of the north-western Negev (Littmann 1997; Russow et al. 2004). In contrast, the BNF of Retama raetam determined in the present study leads only to local N input which creates fertile islands surrounding these shrubs. Calculated even on a hectare scale, this N input is very low at up to 0.11 kg N ha−1 year−1. Acknowledgements We would like to thank the German–Israeli Arid Ecosystems Research Centre (Hebrew University of Jerusalem), especially Aaron Yair, Eyal Sachs and Simon Berkowicz, for their technical and organisational assistance. Many thanks go to Ms. Flügel at the Department of Analytic Chemistry of the UFZ Leipzig for analysing many samples. This research was funded by the German Federal Ministry for Education and Research (DISUM 23, BMBF grants BEO 0339495A [University of Bielefeld; SWB, MV], BEO 0339692J [University of Halle/UFZ, RR, TL]).
References Ahmadjian V (1993) The lichen symbiosis. Wiley, New York Arnibar JN, Anderson IC, Ringrose S, Macko SA (2003) Importance of nitrogen fixation in soil crusts of southern African arid ecosystems: acetylene reduction and stable isotope studies. J Arid Environ 54:345–358 Belnap J (2001) Factors influencing nitrogen fixation and nitrogen release in biological soil crusts. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management. Ecological Studies vol 150, Springer, Berlin Heidelberg New York, pp 241–261 Belnap J (2002): Nitrogen fixation in biological soil crusts from southeast Utah, USA. Biol Fertil Soils 35:128–135 Belnap J, Lange OL (eds) (2001) Biological soil crusts: structure, function and management. Ecological Studies vol 150, Springer, Berlin Heidelberg New York Binkley D, Sollins P, McGill WB (1985) Natural abundance of nitrogen-15 as a tool for tracing alder-fixed nitrogen. Soil Sci Soc Am J 49:444–447 Boddey RM, Peoples MB, Palmer B, Dart PJ (2000) Use of the 15N natural abundance technique to quantify biological nitrogen fixation by woody perennials. Nutrient Cycles Agroecosystems 57:235–270 Comstock JP (2001) Steady-state isotopic fractionation in branched pathways using plant uptake of nitrate as an example. Planta 214:220–234 Danin A (1996) Plants of the desert dunes. Adaptation of desert organisms. Springer, Berlin Heidelberg New York
22 Nitrogen Input Pathways into Sand Dunes
335
Dawson TE, Mambelli S, Plamboeck AH, Templer PH, Tu K (2002) Stable isotope in plant ecology. Annu Rev Ecol Systematics 33:507–559 Ehrlinger JR, Rundel PW (1989) Stable isotope: history, units, and instrumentation. In: Rundel PW, Ehrlinger JR, Nagy KA (eds) Stable isotope in ecological research. Ecological Studies 68, Springer, Berlin Heidelberg New York, pp 342–374 Ettershank G, Ettershank JA, Bryant M, Whitford WG (1978) Effects of nitrogen fertilization on primary production in a Chihuahuan Desert ecosystem. J Arid Environ 1:135–139 Evans RD (2001) Physiological mechanisms influencing plant nitrogen isotope composition. Trends Plant Sci 6:121–126 Evans RD, Ehleringer JR (1993) A break in the nitrogen cycle in arid lands? Evidence from δ15 of soils. Oecologia 94:314–317 Evans RD, Lange OL (2001) Biological soil crusts and ecosystems nitrogen and carbon dynamics. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management. Ecological Studies vol 150, Springer, Berlin Heidelberg New York, pp 263–279 Farnsworth RB (1975) Nodulation and nitrogen fixation in shrubs. In: Stutz HC (ed) Proc Symb Worksh Wildland Shrubs. Bringhan Young University Press, Provo, UT, pp 32–71 Freyer HD (1991): Seasonal variation of 15N/14N ratios in atmospheric nitrate species. Tellus 43B:30–44 Gardener W, Steinberger Y (1989) A proposed mechanism for the formation of “fertile islands” in the desert ecosystems. J Arid Environ 16:257–26 Garten CT (1992) Nitrogen isotope composition of ammonium and nitrate in bulk precipitation and forest throughfall. Int J Environ Anal Chem 47:33–45 Hawksworth DL, Rose F (1976) Lichens as pollution monitors. Edward Arnold, London Heaton THE (1990) 15N/14N ratios of NOx from vehicles and coal-fired power stations. Tellus 42B:304–30 Högberg P (1997) Tansley Review No. 95. 15N natural abundance in soil-plant systems. New Phytol 137:179–203 Lange OL, Kidron GJ, Büdel B, Meyer A, Killian E, Abeliovich A (1992) Taxonomic composition and photosynthetic characteristics of the “biological soil crusts” covering sand dunes in the western Negev Desert. Funct Ecol 6:519–527 Littmann T (1997) Atmospheric input of dust and nitrogen into the Nizzana sand dune ecosystems, northwestern Negev Desert, Israel. J Arid Environ 36:433–457 Ludwig JA (1987) Primary productivity in arid lands: myths and realistic. J Arid Environ 13:1–7 Mayland HF, McIntosh TH, Fuller WH (1966) Fixation of isotopic nitrogen on a semiarid soil by algal crust organism. Soil Sci Soc Am Proc 30:56–60 McAuliffe C, Chamblee BS, Uribe-Arongo H, Woodhouse WW (1958) Influence of inorganic nitrogen on nitrogen fixation by legumes as revealed by N-15. Agronomy J 50:334–337 McGregeor AN, Johnson DE (1971) Capacity of desert algal crusts to fix atmospheric nitrogen. Soil Sci Soc Am Proc 35:843–844 McLendon T, Redente EF (1992) Effects of nitrogen limitation on species replacement dynamics during early succession on a semiarid sagebrush site. Oecologia 91:312–317 Mulvaney RL (1993) Mass spectrometry. In: Knowles R, Blackburn TH (eds) Nitrogen isotope techniques. Academic Press, New York, pp 11–57 Nadelhoffer KJ, Fry B (1994) Nitrogen isotopes in forest ecosystems. In: Laitha K, Michener RH (eds) Stable isotopes in ecology and environmental science. Blackwell, Oxford, pp 22–44 Pugnaire FL, Haase P, Puigdefábregas J (1996) Facilitation between higher plant species in a semiarid environment. Ecology 77:1420–1426 Rai AN, Rowell P, Stewart WDP (1983) Interactions between cyanobacterium and fungus during 15 N2-incorporation and metabolism in the lichens Peltigera canina. Arch Microbiol 134:136–142 Roth E (1997): Critical evaluation of the use and analysis of stable isotopes. Pure Appl Chem 69:1753–1828 Russow R, Faust H (1990) Vergleichende Betrachtung zur Bestimmung der biologischen Stickstoff-Fixierung aus der 15N-Isotopenverdünnung. Zentralb Mikrobiol 145:605–613 Russow R, Veste M, Littmann T (2004) Using the natural 15N-abundance to assess the major nitrogen inputs into the sand dune area of the north-western Negev Desert (Israel). Isotopes Environ Health Stud 40:57–67
336
R. Russow et al.
Rychert RC, Skujins J (1974) Nitrogen fixation by blue-green algae-lichen crusts in the Great Basin Desert. Soil Sci Soc Am Proc 38:768–771 Scharf H (1988) 100 Jahre Kjeldahl-Aufschluss zur N-Bestimmung. Arch Acker-Pflanzenb Bodenkd 32:321–332 Schlee D, Jung K, Türk R, Gehre M (1996) Natural isotopic variation in species of lichens on an altitude gradient in the eastern central Alps. Ber Nat-med Verhandl Salzburg (Austria) 11:25–34 Schulze ED, Gebauer G, Ziegler H, Lange OL(1991) Estimates of nitrogen fixation by trees on an aridity gradient in Namibia. Oecologia 88:451–455 Shearer G, Kohl DH (1989) Estimates of N2 fixation in ecosystems. The need for and basis of the 15N natural abundance method. In: Rundel PW, Ehrlinger JR, Nagy KA (eds) Stable isotope in ecological research. Ecological Studies vol 68, Springer, Berlin Heidelberg New York, pp 342–374 Shearer G, Kohl DH, Virginia RA, Bryan BA, Skeeters JL, Nilsen ET, Sharifi MR, Rundel PW (1983) Estimation of N2-fixation from variation in the natural abundance of 15N in Sonoran desert ecosystem. Oecologia 56:365–373 Shields LM, Mitchell C, Drouet F (1957) Alga- and lichens-stabilized surface crusts as soil nitrogen sources. Am J Bot 44:489–498 Skujins J (1981) Nitrogen cycling in arid ecosystems. In: Clark FE, Rosswall T (eds) Terrestrial nitrogen cycles. Ecol Bull (Stockholm) 33:477–491 Stolte K (1993) Lichens as bioindicators of air quality. General Tech Rep RM-224. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO Stratmann A (1996) Untersuchungen zur Verteilung von Stickstoff in Vegetation und Boden eines Längsdünensystems in der Negev-Wüste, Israel. Diplomarbeit, Universität Bielefeld Tenbergen B (1991) Vergleichende Landschaftsökologische Untersuchungen im nördlichen Negev-Hochland von Israel. Arbeitsber Lehrstuhl Landschaftsökologie Münster 12 Trumble HC, Woodroffe K (1954) The influence of climatic factors on the reaction of desert shrubs to grazing by sheep. In: Cloudsley-Thompson JL (ed) Biology of deserts. Institute of Biology, London, pp 129–147 Valladares F, Villar-Salvador P, Domínguez S, Fernandez-Pascual M, Penuelas JL, Pugnaire FI (2002) Enhancing the early performance of the leguminous shrub Retama sphaerocarpa (L.) Boiss.: fertilisation versus Rhizobium inoculation. Plant Soils 240:253–262 Veste M, Littmann T, Schultz A, Eggert K, Sommer C, Breckle S-W (2000) Biomasseverteilung und deren räumliche Modellierung in Sanddünen der Negev-Wüste (Israel).Verhandl Gesell Ökol 30:85 Veste M, Littmann T, Breckle S-W, Yair A (2001a) The role of biological soil crusts on desert sand dunes of the north-western Negev (Israel). In: Breckle S-W, Veste M, Wucherer W (eds) Sustainable land-use in deserts. Springer, Berlin Heidelberg New York, pp 357–367 Veste M, Littmann T, Friedrich H, Breckle S-W (2001b) Microclimatic boundary conditions for activity of soil lichen crusts in sand dunes of the north-western Negev desert, Israel. Flora 196:465–476 Virginia RA, Jarrell WM, Rundel PW, Shearer G, Kohl DH (1989) The use of variation in the natural abundance of 15N to assess symbiotic nitrogen fixation by woody plants. In: Rundel PW, Ehleringer JR, Nagy KA (eds) Stable isotope in ecological research. Ecological Studies vol 68, Springer, Berlin Heidelberg New York, pp 375–394 West N (1990) Structure and function of microphytic soil crusts in wildland ecosystems of arid and semi-arid regions. Adv Ecol Res 20:179–223 West N (1991) Nutrient cycling in semi-arid and arid regions. In: Skujins J (ed) Semiarid lands and deserts: soil resources and reclamation. Marcel Dekker, New York, pp 295–332 West NE, Skujins J (1977) The nitrogen cycle in North America cold winter semiarid ecosystems. Oecologia 12:45–53 Xie G, Steinberger Y (2001) Temporal patterns of C and N under shrub canopy in a loessial soil desert ecosystem. Soil Biol Biochem 33:1371–1379 Xie G, Steinberger Y (2002) Dynamics of the nitrogen-efficient guild and its relationship to nitrogen and carbon patterns in two desert soil ecosystems. Arid Land Res Manage 16:69–81 Zaady E, Groffman P, Shachak M (1998) Nitrogen fixation in macro- and microphytic patches in the Negev Desert. Soil Biol Biochem 30:449–454
Chapter 23
Vascular Plant Response to Microbiotic Soil Surface Crusts R. Prasse and R. Bornkamm
23.1
Introduction
Microbiotic soil surface crusts are a common and widespread feature of the world’s arid and semiarid lands (e.g. West 1990; Eldridge and Greene 1994). The structure of the microbiotic crusts of Nizzana, their role in the different ecotopes and their physiological activity have been addressed in other chapters in this volume (Chaps. 10, 17, 20, 21). This chapter will focus on the effects of those crusts on vascular plants. Several reviews on microbiotic crusts (e.g. Harper and Marble 1988; Johansen 1986; Isichei 1990; West 1990; Metting 1991; Johansen 1993; Eldridge and Greene 1994; Belnap et al. 2001) have concluded that crusts have either no effect or positive effects on vascular plants. Negative effects are usually stressed much less (but see Eldrigde and Greene 1994). Yet, the abovementioned authors had to base their conclusions on previous studies which were mainly observational (Kleiner and Harper 1972; Nebecker and St. Clair 1980; Anderson et al. 1982a, b; Graetz and Tongway 1986; Johansen and St. Clair 1986; Marble and Harper 1989; Tongway and Smith 1989; Beymer and Klopatek 1992; Eldridge 1993). These studies usually compared one particular measure of vascular plant abundance (e.g. cover) at grazed sites with disturbed crusts and at undisturbed sites with well-developed crusts. A shortcoming of such an approach is the lack of distinction between the direct impact of the source of the disturbance (i.e. trampling and grazing) on vascular plants, and the effect of the crust itself. Experimental studies on the effect of crusts on vascular plant success are very rare and limited to sowing experiments (McIlvanie 1942; St. Clair et al. 1984; Harper and St. Clair 1985; Sylla 1987; Zaady et al. 1997). Using sowing experiments in the field and the greenhouse, Sylla (1987) found that, compared to disturbed soil crusts, both germination densities and seedling survival were lower on undisturbed crusts. Similarly, McIlvanie (1942) observed no germination on undisturbed crusts on which seeds were sown in the greenhouse. On the other hand, St. Clair et al. (1984), who sowed three grass species on undisturbed and experimentally trampled crusts, found a tendency towards higher densities on untrampled plots. Harper and St. Clair S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
337
338
R. Prasse, R. Bornkamm
(1985) observed more seedlings and a higher seedling survival on undisturbed crusts, compared to plots where the crust had been removed before applying seeds. Zaady et al. (1997) found positive as well as negative effects of microbiotic crusts on germination probabilities of vascular plants. In order to draw a conclusive picture about “crust effects”, it may be necessary to address that surface morphology and differences in underlying soils may alter the effect of microbiotic crusts on vascular plants. For example, most previous studies, which were conducted on rough crusts and on loamy or loessy soils, found positive or no influence of microbiotic soil surface crusts on infiltration rates (e.g. reviews of Harper and Marble 1988; West 1990; Eldridge and Greene 1994). Kidron and Yair (1997), however, showed that on the sandy soils of the Nizzana research site, infiltration rates are negatively influenced by microbiotic soil surface crusts. Therefore, it may be expected that on coarse-grained soils with high infiltration rates, crusts tend to decrease infiltration while the reverse is the case for finer-grained soils with low infiltration rates. In addition, Johansen (1986) has suggested that differences in crust surface topography (i.e. rough surface vs. smooth surface) may account for differences in responses of vascular plants to microbiotic crusts. Compared to rather smooth crusts, the rough surfaces of microbiotic crusts, which are rich in lichen and mosses, may not only increase infiltration rates but also enhance lodgement probabilities of vascular plant diaspores. Unfortunately, virtually nothing is known about the effects of smooth microbiotic crusts above sand on vascular plant success. The microbiotic crusts at the Nizzana research site provide an excellent opportunity to study the effects of such smooth crusts on vascular plants, and to evaluate the mechanisms by which crusts and higher plants interact. In particular, the following questions were studied: 1. is vascular plant success (e.g. densities and reproduction) reduced by the presence of the smooth microbiotic crusts of Nizzana? 2. can vascular plant response to microbiotic crusts be explained by negative crust effects on lodgement probabilities for diaspores?
23.2
Methods
The study was conducted in three of the ecotopes described in Chap. 8 (see also Chap. 2, this volume).
23.2.1
Plinth
These are areas with a very brittle and thin (<1 mm) microbiotic crust which is frequently covered by a more or less pronounced layer of mobile sand. This ecotope is found mainly on the moderate upper slopes of dune flanks (i.e. plinths, in the
23 Vascular Plant Response to Microbiotic Soil Surface Crusts
339
terminology of, e.g. Cooke et al. 1993). Total plant cover is about 30%. Annual plants are more or less evenly distributed in the interspaces between perennials but tend to concentrate under shrub canopies (Tielbörger and Kadmon 1997). The most dominant perennial species are Moltkiopsis ciliata, Heliotropium digynum and Convolvulus lanatus, while the most abundant annual species is the tiny composite Ifloga spicata (Chap. 7, this volume).
23.2.2
Interdune
In interdunes with a relatively “flat” cross section (“Ebene Gassen”, according to Allgaier 1993), large areas are covered by a thin (∼1 mm), light-coloured microbiotic crust. The crust surface is smooth and usually not covered with sand (Chap. 10, this volume). Perennial plant cover is about 20% and the dominating perennial species are Echiochilon fruticosum, Thymelaea hirsuta, Retama raetam, Stipagrostis plumosa and Helianthemum sessiliflorum. Annual plants are abundant under shrubs but are very patchily distributed in the crusted, open areas between the perennials. Large parts of the crusts are completely devoid of vascular plants.
23.2.3
Hard Crust
This ecotope is characterized by a very smooth soil surface which is formed by a thin (∼1 mm) microbiotic crust. The surface is extremely stable, due to the presence of cemented underlying sand. These crusts are located mainly along lower, moderate southern slopes adjacent to interdune areas with round cross sections (“Runde Gassen”, according to Allgaier 1993). The perennial plant cover is about 30% and is dominated by a single species (Cornulaca monacantha). Annual plant cover is extremely sparse and restricted largely to areas under shrub canopies (Tielbörger 1997). The crusted interspaces between the perennials are almost devoid of vascular plants.
23.2.4
Experiments
The influence of the crusts on vascular plant success was tested by means of removal and destruction experiments in the following setup. In March 1994, before the main seed dispersal of vascular plants, on each of the crust types eight groups of quadrats (size: 20×20 cm) were established at randomly chosen locations. In each group, the following treatments were randomly assigned to an equal number of quadrats: control, removal of the crust before (March 1994) and after (October 1994) main seed dispersal, and destruction of the crust (crushing the crust but otherwise retaining it on the quadrat) before and after main seed dispersal. Densities of all emerging vascular plant species were counted several
340
R. Prasse, R. Bornkamm
times during the 1994/1995 season, starting after the first rainfall in November 1994. Mortality was estimated as the percentage of all individuals (per quadrat) which died before fruitset. Average fecundity (numbers of capitules or achenes) was measured for two common and widespread species (Ifloga spicata and Senecio glaucus) on all crust types and for all treatments. For each plot, five randomly selected individuals of each species were measured for a morphological feature (e.g. diameter of flowerhead) which was highly correlated with a measure of fecundity (e.g. number of achenes). In plots with less than five individuals, all available specimens were measured. For a detailed description of the fecundity measurements, the reader is referred to Prasse (1999). In order to investigate whether the crust affects vascular plant emergence by preventing seeds from being deposited, or whether it prevents an existing seed bank from coming into contact with the mineral soil, a second set of experiments was established in March 1995. These were similar to those of the first set and used the following treatments: control, and removal or destruction of the crust (both in March 1995, before main seed dispersal; n=12). Half of the quadrats of each treatment were subsequently covered with cotton sheets to prevent seed dispersal into the quadrats. The sheets were removed during the first rain in the 1995/1996 season (December 1995). All seedlings emerging from the covered quadrats recruited from the seed bank, while seedlings emerging in the uncovered quadrats recruited from the seed bank plus net dispersal into the quadrats. Preliminary experiments with the cotton sheets indicated that there was no influence of the sheet itself on soil surface structure and on the plants.
23.2.5
Statistical Analyses
The dependent variables for analyses of the field experiments were mean emergence densities, mean species number, mean percentage mortality and mean fecundity per quadrat (400 cm2), as well as mean fecundity per individual. The independent variable for the statistical analyses of the data was treatment type. Due to a large number of zero values, assumptions for t-tests and ANOVA were not met by the data, which were therefore analysed using nonparametric tests. The effect of the treatments on the dependent variables was tested for each year and habitat type separately by performing Mann-Whitney U-tests.
23.3 23.3.1
Results Densities and Species Numbers
On the interdune crust and on the hard crusts, both removal and destruction of the soil surface led to increased overall densities and species numbers (Figs. 23.1 and 23.2). On the crust at the dune slopes (plinth crust), however, the same manipulations
23 Vascular Plant Response to Microbiotic Soil Surface Crusts 120
341
Hard Crust
Mean Density/ 400 sqcm
100 80 60 40
*
*
1* 2*
20 0 1
2
120
3
4
5
4
5
Interdune
Mean Density/ 400 sqcm
100 80
*
60 40
*
* *
20 0 1
Mean Density/ 400 sqcm
120 100
2
3 Plinth
1*
1994/95 1995/96
80 60
2
40 20 0
1
2
3
4
5
Treatment
Fig. 23.1 Mean (+SE) densities of vascular plants in three different ecotopes, under five different treatments during two consecutive growing seasons. Numbers for treatments (X axis) denote removal of surface crust before (1) and after (2) seed dispersal, destruction of the soil surface before (3) and after (4) seed dispersal, and control (5). Asterisks Significant differences between treatment and control within a given year, numbers significant differences between before and after dispersal, within a given treatment and year (p < 0.05, Mann-Whitney U-tests)
caused either a slight (mostly non-significant) increase in densities when disturbances were applied before seed dispersal, or they led to decreased densities and species numbers when the disturbance was applied after main seed dispersal (Figs. 23.1 and 23.2). There was a general trend towards higher densities and, for the hard crust and interdune crust, higher species numbers when the disturbances were applied before, rather than after the seed dispersal period.
342
R. Prasse, R. Bornkamm Hard Crust
Mean No.of Species/ 400sqcm
8
* 6 4
1* *
2*
2 0 1
Mean No.of Species/ 400sqcm
8
2
3
Interdune
*
*
4
5
4
5
*
6
* 4 2 0 1
2
Plinth
8 Mean No.of Species/ 400sqcm
3
1994/95 1995/96
6 4 2 0
1
2
3
4
5
Treatment
Fig. 23.2 Mean (+SE) numbers of vascular plant species in three different ecotopes, under five different treatments during two consecutive growing seasons (cf. Fig. 23.1 for more information)
During the second growing period, densities were too low to allow meaningful statistical analyses.
23.3.2
Mortality
In none of our experiments was a significant effect of the disturbances on vascular plant mortality found (Fig. 23.3).
23 Vascular Plant Response to Microbiotic Soil Surface Crusts
343
Hard Crust
1994/95
Mean Mortality (%)
100
1995/96
80 60
*
40 20 0 1
2
3
4
5
4
5
4
5
Interdune Mean Mortality (%)
100 80 60 40 20 0 1
2
3 Plinth
Mean Mortality (%)
100 80 60 40 20 0 1
2
3 Treatment
Fig. 23.3 Mean (+SE) percentage mortality of vascular plants in three different ecotopes, under five different treatments during two consecutive growing seasons (cf. Fig. 23.1 for more information)
23.3.3
Fecundity
On each of the three studied crust types, fecundity per individual and/or per unit area of the two study species increased in response to disturbance. In some cases, however, the numbers of individuals found were too low to allow a statistically significant detection of treatment effects. In no case was a decreased fecundity observed as response to the treatments. In a few cases, effects of timing of the treatment
R. Prasse, R. Bornkamm Mean No. of Capitules per Plant
344 12
1* 2*
8
*
4 0 1
Mean No. of Capitules per 400sqcm
*
2
3 Treatment
4
5
300
*
250 200 150 100
* *
50
*
0
1
2
3
4
5
Treatment
Mean No.of Capitules per Plant
Fig. 23.4 Mean (+SE) number of capitules per plant and per unit area for Ifloga spicata in the interdune ecotope. Numbers for treatments (X axis) denote removal of surface crust before (1) and after (2) seed dispersal, destruction of the soil surface before (3) and after (4) seed dispersal, and control (5). Asterisks Significant differences between treatment and control, numbers significant differences between before and after dispersal for a given treatment (p < 0.05, Mann-Whitney U-tests) 8
*
1*
*
6
2*
4 2 0 1
2
3
4
5
Mean No. of Capitules per 400sqcm
Treatment 500
1*
400
1*
300 200
2
2
100 0 1
2
3 Treatment
4
5
Fig. 23.5 Mean (+SE) number of capitules per plant and per unit area for Ifloga spicata in the plinth ecotope (cf. Fig. 23.4 for more information)
Mean No. of Seeds per Plant
23 Vascular Plant Response to Microbiotic Soil Surface Crusts 200
*
345
* *
150
*
100 50 0 1
2
3
4
5
Treatment
Mean No. of seeds per 400sqcm
1400
*
1200 1000
(*)
*
800 600 400 200 0 1
2
3 Treatment
4
5
Fig. 23.6 Mean (+SE) number of achenes per plant and per unit area for Senecio glaucus in the plinth ecotope (cf. Fig. 23.4 for more information)
were recorded, with lower fecundity per unit area when the treatment was applied after main seed dispersal. Figures 23.4, 23.5 and 23.6 show results for I. spicata and S. glaucus on plinth and interdune crusts.
23.3.4
Underlying Mechanisms
Similar to the first set of experiments, significantly higher densities and species numbers were observed in manipulated quadrats of the second set of experiments on all three crust types (Fig. 23.7). However, no such effect was observed when seed dispersal into the quadrats was inhibited by covering these during the dispersal period. This implies that no germinable seeds were available on intact crusts and that intact crusts reduced lodgement probabilities.
R. Prasse, R. Bornkamm
Mean Density / 400 sqcm
346 45 40 35 30 25 20 15 10 5 0
1 2 3 4 5 6
*
*
* *
Mean No. of Species / 400 sqcm
Hard Crust 6 5 4 3 2
1 2 3 4 5 6
*
Interdune
*
Plinth
* *
* *
* *
1 0 Hard Crust
Interdune
Plinth
Fig. 23.7 Mean (+SE) densities and mean (+SE) species numbers of vascular plants under different treatments in 1995/1996. Numbers for treatments denote control covered (1) and uncovered (2), destruction of surface crust, covered (3) and uncovered (4), and removal of surface crust, covered (5) and uncovered (6). Asterisks Significant differences between treatment and control (p < 0.05, Mann-Whitney U-tests). Note that there was no germination for treatments 1, 2, 3 and 5
23.4
Discussion
In both sets of experiments, all measured parameters and variables (density, species number, mortality, fecundity) were affected by the different treatments. The disturbances were usually followed by enhanced densities, species number and reproductive success of vascular plants. However, the direction and magnitude of vascular plant response to the experimental manipulation varied among species, treatment type and timing, crust type and year of study. Our results indicate that intact microbiotic crusts may have a strong inhibiting effect on vascular plant densities and species number at the Nizzana research site. The fact that the increase in densities and species number was higher when the disturbances were applied before seed dispersal highlights that the crusts affected seed dispersal of the plants. When dispersal of vascular plant diaspores was inhibited, there was germination neither on intact nor on disturbed crusts in the following growing period.
23 Vascular Plant Response to Microbiotic Soil Surface Crusts
347
From this, we conclude that the smooth surface of undisturbed microbiotic crusts decreased the probability for a seed to come to rest. The experimentally applied disturbances roughened the soil surface, which then acted as a seed trap. Additional data consistent with this interpretation come from experiments with samples of undisturbed crusts watered in a greenhouse, where nearly no germination was observed on intact microbiotic crusts (see Prasse and Bornkamm 2000). This finding is the first experimental evidence supporting the suggestion of Johansen (1986) that microbiotic crusts with relatively smooth surfaces may provide a lower number of safe sites for plant establishment than do “rough” crusts (e.g. crusts of North America). Such a mechanism would also explain our finding that the seedtrapping effect was less pronounced when a disturbance was applied after main seed dispersal. Most seeds would already be lodged by then, and remaining time for seed trapping may not be sufficient to enhance densities and species numbers. Also, the crusts of the research site regenerate rapidly after the first rainfalls, which smoothens the surfaces again. Sylla (1987) concluded from his study that, in arid areas, microbiotic crusts constitute a barrier to seedling emergence. His greenhouse experiments indicated that the seedling’s radicles could not penetrate the crust to reach the mineral soil. A similar conclusion was drawn earlier by McIlvanie (1942) who observed that seeds on encrusted soils did not come into contact with the mineral soils, and dried out. Zaady et al. (1997) inferred an inhibition of germination by “components” leaching from cyanobacterial filaments. Our results indicate an additional mechanism of crust effects on plants. If leaching of inhibiting substances would have played an important role in the research area, we should have found lower densities and/or species number where the crust was only destroyed (and retained on the plot), compared to places where the crust was completely removed. Also, additional observations showed that lodged seeds are capable of penetrating through wet crust (but crusts dry rapidly in the research area). Therefore, we conclude that crusts with smooth surface topography influence vascular plant success mainly by reducing lodgement probabilities for diaspores. In addition to crust effects on plant abundance and species diversity, we showed that crusts may negatively affect fecundity. It is interesting that a higher fecundity per individual was found on disturbed plots even though densities were enhanced by the same disturbance. This indicates that negative effects of crusts on vascular plant fecundity may be stronger than competitive effects among vascular plants. Kidron and Yair (Chap. 17, this volume) have found that the crusts of the research area initiate run-off after being saturated. This redirecting of water probably deprives higher plants growing on intact crusts of some resources (water and nutrients). Therefore, we suggest that the small-scale disturbances applied in the present work may have acted as mini-catchments for the run-off from surrounding areas. Additional water caught by the disturbed surface is most likely the reason for the observed higher fecundity of the study species on disturbed crusts. The fact that the increase in fecundity per unit area was less pronounced when the disturbance was applied after main seed dispersal is most likely due to the lower densities on late-disturbance plots.
348
R. Prasse, R. Bornkamm
The finding that an intact crust reduced fecundity (apparently by depriving plants of resources) is contradictory to the results of Lesica and Shelly (1992). They found no influence of crusts on the fecundity of a species from the mustard family (Arabis fecunda). Since this negative influence of the crust in the Nizzana research area may partly be related to its run-off-generating features, the fact that such an effect was not revealed by Lesica and Shelly (1992) may be due to the rather rough surface of the North American crusts, which may enhance infiltration (cf. compilations by West 1990 and Eldridge and Greene 1994). From our overall results, we propose for the research area a positive feedback process of vascular plant development after an initial moderate disturbance of the microbiotic crust. In the first year after a disturbance, vascular plant densities and species numbers are higher due to seed trapping by disturbed crusts. Because of the higher reproductive success and a seed trapping by dry remains of plants, higher vascular plant densities may also be expected in the following years. Thus, smallscale disturbances may have long-term consequences for vascular plant dynamics. Still, the process described above may be counteracted by competition among vascular plants, so that it is interrupted if more than a certain amount of space is occupied. Furthermore, it is important to stress that there is probably an upper limit to the spatial extent of a disturbance which causes “positive” vascular plant response. Seed availability is not unlimited and, at a certain spatial scale, the available amount of seeds would be insufficient to enhance densities in the whole disturbed area. Also, the positive effect of additional water from run-off probably depends on the existence of undisturbed crust surfaces in the direct vicinity of a disturbance. Therefore, in large-scale disturbances, only the edges may receive additional water while no supplementary water reaches the centre of the disturbance. In such cases, vascular plant response will probably be different from our observations. More knowledge about the mechanisms underlying the interactions between crusts and vascular plants is required to determine the upper limit of disturbance size above which there are no positive density responses of vascular plants. In addition, a positive feedback process probably requires a well-timed disturbance. Disturbances after main seed dispersal may not lead to increased seed trapping, since most diaspores are already lodged by then. During the next growing period, the crusts may re-grow rapidly, and thereby smoothen the surface again. This would reduce the seed-trapping function of the disturbance during subsequent years and, in such cases, no long-term effect may be observed.
23.5
Conclusions
The results from Nizzana are particularly interesting, as they present experimental evidence for negative crust effects on vascular plants. This adds important knowledge to a subject which has rarely been studied experimentally and is still controversially discussed (e.g. reviews of Harper and Marble 1988; Johansen 1986; West 1990; Johansen 1993; Eldridge and Greene 1994; Belnap et al. 2001). Indeed, the nature of the interactions between different crust types and vascular plants still requires
23 Vascular Plant Response to Microbiotic Soil Surface Crusts
349
more research, since the effects of crusts on vascular plants are most likely very diverse. They may include a trade-off between the negative effects described above and the more beneficial effects of crusts on vascular plants. The latter are, for example, enhanced soil stability (e.g. Booth 1941; Belnap and Gillette 1998; Leys and Eldridge 1998) and increased nutrient availability (Harper and Pendleton 1993; Belnap and Harper 1995) on crusted surfaces. Also, while plants growing in run-off areas are deprived of water, conspecific plants in run-on areas (e.g. depressions and dune bases) may benefit from the presence of crusts by receiving additional water from surrounding encrusted areas. We suggest that crust topography (smooth vs. rough) and soil type (coarsegrained vs. fine-grained) play a major role in determining the response of vascular plants to the presence of crusts. Therefore, future attempts to generalise effects of microbiotic crusts on vascular plant success should account for differences in crust topography and underlying soil types. Acknowledgements We are indebted to R. Kadmon for hosting R.P. in his laboratory at the Hebrew University of Jerusalem and for helpful discussions during the critical beginning of the study. We appreciate that K. Tielbörger and S. Lahav-Ginott read the manuscript and offered valuable comments. A. Stratmann gave invaluable assistance in the field and laboratory during the 1994/1995 growing period. We are also grateful to C. Holzapfel and H. Parag for their efforts in supplying us with much of the “cryptic” literature from the USA. The staff of the Arid Ecosystems Research Centre at the Hebrew University in Jerusalem helped greatly by providing facilities and, despite all security problems, access to the study site. The study was part of a project funded by the Bundesministerium für Bildung und Forschung in Germany (Förderkennzeichen 0339498A).
References Allgaier A (1993) Geomorphologische Untersuchungen an Längsdünen in der westlichen Negev, Israel. Magisterarbeit, Rheinisch-Westfälische Technische Hochschule, Aachen Anderson CA, Harper KT, Rushforth SR (1982a) Recovery of cryptogamic soil crusts from grazing on Utah winter ranges. J Range Manage 35:355–359 Anderson DC, Harper KT, Holmgren RC (1982b) Factors influencing development of cryptogamic soil crusts in Utah deserts. J Range Manage 35:180–185 Belnap J, Gilette DA (1998) Vulnerability of desert biological crusts to wind erosion: the influences of crust development, soil texture, and disturbance. J Arid Environ 39:133–142 Belnap J, Harper KT (1995) Influence of cryptobiotic soil crusts on elemental content of tissue of two desert seed plants. Arid Soil Res Rehabil 9:107–115 Belnap J, Prasse R, Harper KT (2001) Influence of biological soil crusts on soil environments and vascular plants. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function and management (2nd edn, 2003). Ecological Studies vol 150. Springer, Berlin Heidelberg New York, pp 281–300 Beymer RJ, Klopatek JM (1992) Effects of grazing on cryptogamic crusts in pinyon-juniper woodlands in Grand Canyon National Park. Am Midl Naturalist 127:139–148 Booth WE (1941) Algae as pioneers in plant succession and their importance in erosion control. Ecology 22:38–46 Cooke R, Warren A, Goudie A (1993) Desert geomorphology. UCL Press, London Danin A, Bar-Or Y, Dor I, Yisraeli T (1989) The role of cyanobacteria in stabilization of sand dunes in southern Israel. Ecol Mediterr XV:55–64
350
R. Prasse, R. Bornkamm
Eldridge DJ (1993) Cryptogams, vascular plants, and soil hydrological relations: some preliminary results from the semiarid woodlands of eastern Australia. Great Basin Naturalist 53:48–58 Eldridge DJ, Greene RSB (1994) Microbiotic soil crusts: a review of their roles in soil and ecological processes in the rangelands of Australia. Austr J Soil Res 32:389–415 Graetz RD, Tongway DJ (1986) Influence of grazing management on vegetation, soil structure and nutrient distribution and the infiltration of applied rainfall in a semi-arid chenopod shrubland. Austr J Ecol 11:347–360 Harper KT, Marble JR (1988) A role for nonvascular plants in management of arid and semiarid rangelands. In: Tueller PT (ed) Vegetation science applications for rangeland analysis and management. Kluwer, Dordrecht, pp 133–169 Harper KT, Pendleton RL (1993) Cyanobacteria and Cyanolichens: Can they enhance availability of essential minerals for higher plants? Great Basin Naturalist 53:59–72 Harper KT, St. Clair LL (1985) Cryptogamic soil crusts on arid and semiarid rangelands in Utah: Effects on seedling establishment and soil stability. Department of Botany and Range Science, Brigham Young University, Provo. Final report BLM contract no. BLM AA 851-CTI-48, Bureau of Land Management, Utah State Office, Salt Lake City, UT Isichei AO (1990) The role of algae and cyanobacteria in arid lands. A review. Arid Soil Res Rehabil 4:1–17 Johansen JR (1986) Importance of cryptogamic soil crusts to arid rangelands: Implications for short duration grazing. In: Tiedeman JA (ed) Short Duration Grazing and Current Issues in Grazing Management, Shortcourse. Washington State University, Kennewick, WA, pp 127–136 Johansen JR (1993) Cryptogamic crusts of semiarid and arid lands of North America. J Phycol 29:140–147 Johansen JR, St. Clair LL (1986) Cryptogamic soil crusts: Recovery from grazing near camp Floyd State Park, Utah, USA. Great Basin Naturalist 46:632–640 Kidron GJ, Yair A (1997) Rainfall-runoff relationship over encrusted dune surfaces, Nizzana, western Negev, Israel. Earth Surface Processes Landforms 22:1169–1184 Kleiner EF, Harper KT (1972) Environment and community organisation in grasslands of Canyonlands National Park. Ecology 53:299–309 Lesica P, Shelly JS (1992) Effects of cryptogamic soil crust on the population dynamics of Arabis fecunda (Brassicaceae). Am Midl Naturalist 128:53–60 Leys JF, Eldridge DJ (1998) Influence of cryptogamic crust disturbance to wind erosion on sand and loam rangeland soils. Earth Surface Processes Landforms 23:963–974 Marble RJ, Harper KT (1989) Effect of timing of grazing on soil-surface cryptogamic communities in a Great Basin low-shrub desert: A preliminary report. Great Basin Naturalist 49:104–107 McIlvanie SK (1942) Grass seedling establishment and productivity – overgrazed vs. protected range soils. Ecology 23:228–231 Metting B (1991) Biological surface features of semiarid lands and deserts. In: Skujin J (ed) Semiarid lands and deserts: soil resource and reclamation. Marcel Dekker, New York, pp 257–293 Nebecker GT, St. Clair LL (1980) Enhancement of seed germination and seedling development by cryptogamic soil crusts in an Atriplex confertifolia-Sarcobatus vermiculatus community. Bot Soc Am Misc Ser Publ 158:81 Prasse R (1999) Experimentelle Untersuchungen an Gefäßpflanzenpopulationen auf verschiedenen Geländeoberflächen eines Sandwüstengebietes. Universitätsverlag Rasch, Osnabrück Prasse R, Bornkamm R (2000) Effect of microbiotic soil surface crusts on emergence of vascular plants. Plant Ecol 150:65–75 St. Clair LL, Webb BL, Johansen JR, Nebeker GT (1984) Cryptogamic soil crusts: Enhancement of seedling establishment in disturbed and undisturbed areas. Reclamation Reveget Res 3:129–136 Sylla D (1987) Effect of microphytic crust on emergence of range grasses. MSc Thesis, School of Renewable Natural Resources, University of Arizona, Tucson, AR
23 Vascular Plant Response to Microbiotic Soil Surface Crusts
351
Tielbörger K (1997) The vegetation of linear desert dunes in the north-western Negev, Israel. Flora 192:261–278 Tielbörger K, Kadmon R (1997) Relationships between shrubs and annual communities in a sandy desert ecosystem: A three year study. Plant Ecol 130:191–201 Tongway DJ, Smith EL (1989) Soil surface features as indicators of rangeland site productivity. Austr Rangelands 11:15–20 West NE (1990) Structure and function of microphytic soil crusts in wildland ecosystems of arid to semi-arid regions. In: Begon M, Fitter AH, MacFadyen A (eds) Advances in Ecological Research, vol 20. Academic Press, London, pp 179–223 Zaady E, Gutterman Y, Boeken B (1997) The germination of mucilaginous seeds of Plantago coronopus, Reboudia pinnata, and Carrichtera annua on cyanobacterial soil crust from the Negev Desert. Plant Soil 190:247–252
Chapter 24
Ion Relations of Plants and Soil Patterns M. Veste, U. Sartorius, and S.-W. Breckle
24.1
Introduction
Climatic conditions govern the water cycle and balance and, thereby, not only the availability of water during the seasons, but also the presence of soluble ions in upper soil horizons and, thus, in an ecosystem. In humid regions, the landscape geomorphology is characterized by a typical drainage system which starts at springs and wells and then continues along water-collecting creeks, rivers and streams, eventually reaching the ocean. This water always contains more or less small amounts of water-soluble ions leached from rocks and soils during capillary movement. A rather small proportion of the ions, however, is always to be found also in rainwater (Walter and Breckle 1983, 1985). In humid regions, water transport and capillary threads of soil water are directed mainly downstream. In arid regions, capillary movement of soil water is, if present at all, mainly upstream to the soil surface. Here, various ions transported by the water are precipitated and can cause the formation of salt crusts (Breckle 2002b). In general, in arid regions the input of water by precipitation (rain, snow, dew) over the year to a distinct ecosystem or part of landscape is less than the possible output by potential evapo-transpiration. In humid regions, this is reversed. This has the consequence that, in arid areas, salinization of soils is always a serious danger, especially if leaching of salts is possible in deeper soil horizons or adjacent parent rocks (mainly of NaCl and, to a smaller extent, other water-soluble ions, too). Marine aerosols in coastal areas can be another source of salt accumulation, if prevailing winds carry these onshore in the form of spray from breaking waves. Salts can thus be blown many km inland and steadily deposited on soils, causing salinization (Teakle 1937). The fact that salinity is a common feature in arid climates, and that NaCl (according to Kinzel 1982) must be regarded as the most widespread chemical compound in ecosystems which restricts plant growth (Wucherer and Breckle 2005), makes it an important question whether ion conditions in the arid, longitudinal dune system of the Negev are also influenced by salinity.
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
353
354
M. Veste et al.
The occurrence of some Chenopodiaceae and some local accumulations forming small salt crust spots close to the dune basis with salt accumulations were checked. Salinity is not very evident in the system as a whole, unlike what has generally been observed in sand dune ecosystems. Analyses of ion contents (Na+, K+, Cl−) of aboveground plant organs of perennial phytomass, K+/Na+ ratios, local variability and the specific mosaic of salt-accumulating plants (cf. heterogeneous distribution of halophytes), a possible feedback to subsurface soil salinity, as well as an effect on the small-scale distribution of salts in soil before (January) and after (April) the short rainy season reveal results on ion conditions in the area (cf. Chap. 16, this volume). On the other hand, the analysis of several plant species and comparison among different stands allows conclusions on taxon-specific ion patterns and salt accumulation in plant tissues, and the dependence on or independence of soil salinity. This is despite the fact that there are only very few halophytic species present, in contrast to, e.g. Iran (Breckle 1986, 2002a; Akhani 2006) or the Aralkum (Breckle et al. 2001).
24.2 24.2.1
Material and Methods Plant Material
Leaves were collected from Artemisia monosperma, Convolvulus lanatus, Heliotropium digynum, Moltkiopsis ciliata, Noaea mucronata, Retama raetam, Thymelaea hirsuta and Stipagrostis ciliata. Green stems, and stems and leaves were selected from Anabasis articulata and Cornulaca monacantha respectively. Plant collections were made in September 1993, and January and April 1994. Additional A. articulata material was taken in November 1994, April and October 1995, and April 1996 in various habitats. The plants were collected in the characteristic ecotopes of the sand dune system at the Nizzana experimental site (N1): the dune crest, dune slope with sand cover, dune slope with biological crust, dune base, interdune, old dune with a hard crust, and playa (cf. Chaps. 2 and 3, this volume).
24.2.2
Ion Analysis
The plant material was oven-dried at 80 °C until constant weight. The fraction of water-soluble cations (Na, K, Ca and Mg) was determined by atomic absorption spectrophotometry (Perkin Elmer 5100), and chloride by using a Micro-Chloro-Counter (Marius) on hot water extracts (based on Breckle 1976).
24 Ion Relations of Plants and Soil Patterns
24.3 24.3.1
355
Results and Discussion Ion Patterns
Mean sodium (Na), potassium (K) and chloride (Cl) contents in the dominant species of the sand dune area are shown in Fig. 24.1. Accumulation patterns of inorganic solutes are characteristic for certain taxa, which represent different physiotypes (Breckle 1990, 1995, 2000; Albert et al. 2000). Most of the psammophytes accumulated mainly K > Na > Cl and, therefore, in those species the K/ Na quotient is >1 (Fig. 24.2). Within the sand dune area, the ion patterns of the species investigated are nearly independent of soil type. Sodiophilic plants are the Chenopodiaceae A. articulata and C. monacantha. In both species, the K/Na quotient is <1 in all habitats (Fig. 24.2). Generally, the salt content of the sand dunes is very low (Chap. 5, this volume). Therefore, most of the psammophytic plants show relatively low contents of ions in their leaves. The ion patterns of the species investigated are similar to those reported in other investigations in the region (e.g. Shaltout 1992; Bornkamm et al. 1998), and is largely independent of ecotope. A. monosperma is able to adapt to higher salinity conditions while growing on coastal sand dunes near the Mediterranean Sea (Fig. 24.3, Winter et al. 1976). In terms of ion relations, the Chenopodiaceae (Anabasis, Cornulaca and Noaea) behave distinctly different from the other psammophytes of the sand dunes. Notably, A. articulata accumulates the highest amounts of Na and Cl. As is characteristic for many Chenopodiaceae, ion accumulation is genetically fixed and often almost independent of soil ion content (e.g. Breckle 1976; Reimann 2003, 2005). Even on sandy soils, A. articulata showed the same ion pattern as in the playa area (Fig. 24.4), where the highest salt accumulation can be found (Chap. 4, this volume). The Na/K and Na/Cl relations in Anabasis are very variable (Fig. 24.5). Ion contents in the assimilating stems of A. articulata are relative stable between the rainy and dry seasons (Fig. 24.6). A higher accumulation of Na > K > Cl has been found also in other investigations from the Negev and northern Egypt (Winter et al. 1976; El-Ghonemy et al. 1977; Bornkamm et al. 1998). A higher accumulation of Na in A. articulata was found in the Central Negev Highlands. The Na content, with an average of 1.2 mol kg−1 dw, was higher than in the sand dunes, while K and Cl showed similar values (Veste 2004) to those of the southern Negev (Winter et al. 1976). Also other Anabasis species from Afghanistan (Breckle 1986) and the Negev (Rodin and Bazilevich 1967) preferably accumulate sodium. The higher sodium accumulation is also related with a higher succulence (Butnik et al. 2001). In some halophytic Chenopodiaceae such as Suaeda aegyptiaca, salt contents can reach 60% of the dry weight of the plant (Eshel 1985). Succulence in (xero-) halophytes is a mechanism of salt regulation, because the increasing vacuole volume allows a compartmentation of salt (Waisel 1972; Breckle et al. 2001; Breckle 2002b). This has been shown also in species of the Aizoaceae and Mesembryanthemaceae growing on desert sand dunes in the Namib desert. These also accumulate high amounts of sodium and chloride (Albert et al. 2000; Veste 2004; Veste and Mohr 2005).
356
M. Veste et al.
A
Na K Cl
600 500 400 300 200 100
700 ions content [mmol kg−1dw]
ions content [mmol kg−1dw]
700
0 Mo
400 300 200 100
Cor
700
C
Na K Cl
500 400 300 200 100
Con
Mo
Na K Cl
D
600 500 400 300 200 100 0
0
Ana No
Con
Mo
Ret
Na K Cl
E
600
Cor
No
Art
Mo
Ret
Thy
Art
500 400 300 200 100
700 ions content [mmol kg−1dw]
Cor
ions content [mmol kg−1dw]
500
Sti
ions content [mmol kg−1dw]
ions content [mmol kg−1dw]
He
600
700
600
0 Cor
700
Na K Cl
B
G
F
600 500 400
Na K Cl
300 200 100 0
0 Ana
Cor
No
Con
Ret
Thy
Cor
Ana
Fig. 24.1 Mean sodium (Na), potassium (K) and chloride (Cl) contents in leaves of dominant plant species in different habitats of the Nizzana experimental site (site N1). A Dune crest, B dune slope with sand cover, C dune slope with biological crust, D dune base, E interdune, F old dune with a hard crust, G playa. Plant species: Ana, Anabasis articulata; Art, Artemisia monosperma; Con, Convolvulus lanatus; Cor, Cornulaca monacantha; He, Heliotropium digynum; Mo, Moltkiopsis ciliata; No, Noaea mucronata; Ret, Retama raetam; Thy, Thymelaea hirsuta; Sti, Stipagrostis ciliata (data from Sartorius 1996)
24 Ion Relations of Plants and Soil Patterns
357 14
14
A
10
10
8
8
6
6
4
4
2
2 0
0 Cor
He
Mo
Sti
Cor
10
8
8
6
6
4
4
2
2 0
0 Cor No
Con Mo
Ret
Art
Ana Cor No Art Mo Ret Thy
14
14
E
12
10
10
8
8
K/Na
K/Na
Mo
D
12
K/Na
K/Na
C
10
12
Con
14
14 12
B
12
K/Na
K/Na
12
6
G
F
6
4
4
2
2 0
0 Ana Cor No Con Ret Thy
Cor
Ana
Fig. 24.2 Median of the K+/Na+ relation in leaves of dominant plant species in different habitats of the Nizzana experimental site (site N1). A Dune crest, B dune slope with sand cover, C dune slope with biological crust, D dune base, E interdune, F old dune with a hard crust, G playa. For abbreviations, see Fig. 24.1 (data from Sartorius 1996)
Compared to other sand dune plants, the high accumulation of inorganic ions, and the biosynthesis of organic acids in A. articulata and C. monacantha leaves and shoots lower the osmotic water potential, thereby strongly decreasing the leaf water potential. Measured pre-dawn water potentials in A. articulata were between −1.62 and −2.5 MPa and, in C. monacantha, −1.8 MPa during the rainy season (Chap. 26, this volume). Due to the higher salt content and soil water availability on the playas, the salt content of Anabasis is here lower than in the sandy soils (Fig. 24.7). Also the growth of Anabasis is strongly reduced in the playa area (Veste and Breckle 2000). The osmotic adaptation of A. articulata is a precondition for the invasion of
358
M. Veste et al.
Artemisia monosperma
K content [mmol kg−1dw]
1200 1000 800 600 400 200 0 0
200
400
600
800 1000 1200
Na content [mmol kg−1dw] Fig. 24.3 Correlation between Na and K of Artemisia monosperma growing in the desert sand dunes of the north-western Negev (circles) and in coastal dunes north of Tel Aviv (squares)
the shrubs into the playa area. The larger osmotic gradient between the soil and plant can improve water uptake by the evergreen Chenopodiaceae from the sands during the dry season (Chap. 26, this volume).
24.3.2
Salt Accumulation in the Standing Biomass
Salt accumulation in the standing biomass was estimated from the mean salt content of the species investigated (Fig. 24.8) and their biomass in the different ecotypes (Chap. 26, this volume). The highest salt accumulation can be found at the dune base, where the highest standing biomass was measured (Fig. 24.8). The Chenopodiaceae N. mucronata, with a high K accumulation, is here the dominant species (48% of the standing biomass). The high Na contents recorded in the various habitats are associated mainly to the occurrence of A. articulata and C. monacantha. In terms of total biomass, Cornulaca contributes 53% on the dune crest and 59% on the old dunes. Anabasis is the dominant species on the playa and along the playa margins (100% of the biomass), and contributes 51% to the biomass in the interdunes.
24.3.3
Salt Accumulation Below the Shrubs
The accumulation of ions in the standing biomass also has a feedback on soil patterns. Mineralization of the detritus leads to small-scale soil patterns, so-called salty islands. Compared to the inter-shrub areas, higher salt contents were found beneath
24 Ion Relations of Plants and Soil Patterns
359
C
1200 1000
800
800
600
600
.
400
400
200
200
0 1200
0 1200
D
B
1000
1000
800
800
600
600
400
400
200
200
0
Mg content [mmol kg−1 dw]
K content [mmol kg−1 dw]
A
1000
Cl content [mmol kg−1 dw]
Na content [mmol kg-1 dw]
Anabasis articulata 1200
0 ND P14 P15 P16 P17 P18 P19
Playa
Sand
ND P14 P15 P16 P17 P18 P19
Sand
Playa
sand
P14
P15
P16 P17
Playa
P18
P19
Fig. 24.4 Spatial variations of ion contents (A Na, B K, C Cl, D Mg) in Anabasis articulata (top) along a soil catena (bottom). P14–P19: Nizzana; site 1 (n=5–10, ND: Haluza, site N3). ND, P14, P15: sand, P16–P18: playa, P19: old dune. For soil types, see Chap. 16 (this volume), Blume et al. (1995) and Ebeling (1996). The horizontal length of the image is 270 m
A. articulata dwarf shrubs (Fig. 24.9) and, to some extent, also under C. monacantha (Fig. 24.10). Salt accumulation below the shrubs and in the open areas between the shrubs is significant for both species (Table 24.1), as has also been reported for perennials, e.g. Atriplex confertifolia in Utah (Breckle 1976; Jackson and Caldwell 1993). Potassium accumulates under the shrubs, while Na is washed into deeper soil layers (Table 24.2). This concentration of potassium ions under shrubs has been shown also in other ecosystems rich in Chenopodiaceae (Macdonald et al. 1999).
M. Veste et al.
Cl content [mmol kg−1 dw]
K content [mmol kg−1 dw]
360 Anabasis articulata
1000
Na / K
A
800 600 400 200
0 1000
Na / Cl
B
800 600 400 200 0 0
200
400
600
800 1000 1200 1400 −1
Na content [mmol kg dw]
Fig. 24.5 Correlation between Na and K (A) and Na and Cl (B) growing on sandy soils (open symbols) and playa (closed symbols) in Nizzana and the dunes of Haluza
A
1200
B
1000
1000
800
800
600
600
400
400
200
200
Month
Dec-95
Jun-95
Dec-94
Jun-94
Dec-93
Dec-95
Jun-95
Dec-94
Jun-94
Dec-93 Na K Cl
Jun-96
0
0
Ions content [mmol kg−1 dw]
Ions content [mmol kg−1dw]
Anabasis articulata 1200
Month
Fig. 24.6 Temporal variations of ion contents (Na, K, Cl) of Anabasis articulata (n=10) growing on playa (A) and sand (B) in Nizzana (site N1)
Infiltration rates are lower under Anabasis shrubs than in the interspaces, as a result of interception losses (Rummel and Felix-Henningsen 2004).
24.4
Conclusions
Compared to other desert ecosystems, salt accumulation in the sand dunes of the northern Negev is low. In most habitats of the dune systems, salt is not a limiting factor for plant growth. The highest salt soil content can be found in the playa,
24 Ion Relations of Plants and Soil Patterns
361
Water content [kg kg−1dw]
Anabasis articulata
1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 AS
P18
P17
P16
P14
Playa
ND
Sand
ions amount standing biomass [g / 100 m2]
Fig. 24.7 Spatial variations of water content of Anabasis articulata (n=10) growing on playa (AS, P16–P18) and sand (P14: sand at Nizzana; site N1, ND: Haluza, site N3). For location, see Fig. 24.5
Na
400
K Cl
300
200
100
0 1
2
3
4
5
6
8
9
Fig. 24.8 Total sodium (Na), potassium (K) and chloride (Cl) contents in the standing biomass of different habitats at the Nizzana experimental site (site N1). 1 Dune crest, 2 dune slope with sand cover, 3 dune slope with biological crust, 4 dune base, 5 interdune, 6 old dune with a hard crust, 7 centre of the playa, 8 playa margin (data from Sartorius 1996)
where also water availability is reduced for the vegetation. In this area, the xerohalophyte Anabasis articulata dominates due to its superior osmotic adaptation. This is a specific feature of the N1-Nizzana site. In most species, ion relations exhibit a typical pattern (Reimann and Breckle 1993) which is genetically fixed (Munns 2005).
362
M. Veste et al. A: under Anabasis
Na
Na
40
30
30
20
20
10
10
0 15
0
K
K
15
10
10
5
5
0 10
0
Cl
Cl
10
5
0
5
surface
-10 cm
-20 cm
surface
-10 cm
-20 cm
ions content [mmol kg−1 soil dw]
ions content [mmol kg−1 soil dw]
40
B: between shrubs
0
Fig. 24.9 Mean sodium (Na), potassium (K) and chloride (Cl) contents on the soil surface, 10 and 20 cm soil depths under Anabasis articulata (A) and between shrubs (B) at the Nizzana experimental site (site N1; data from Sartorius 1996)
A: under Cornulaca
Na
Na
40
30
30
20
20
10
10
0 15
0
K
K
15
10
10
5
5
0 10
0
Cl
Cl
10
5
0
ions content [mmol kg−1 soil dw]
ions content [mmol kg−1 soil dw]
40
B: between shrubs
5
surface
-10 cm
-20 cm
surface
-10 cm
-20 cm
0
Fig. 24.10 Mean sodium (Na), potassium (K) and chloride (Cl) contents on the soil surface, 10 and 20 cm soil depths under Cornulaca monacantha (A) and between shrubs (B) at the Nizzana experimental site (site N1; data from Sartorius 1996)
24 Ion Relations of Plants and Soil Patterns
363
Table 24.1 Comparison of ion contents (Na, K, Cl) at different soil depths under shrubs (Sh) and in the open spaces between shrubs (OP). n.s. Not significant, * p ≤ 0.05, ** p ≤ 0.01 Soil depth Anabasis articulata Cornulaca monacantha Na
Surface 10 cm 20 cm Surface 10 cm 20 cm Surface 10 cm 20 cm
K
Cl
Sh > OP Apr.**, Sep.* Sh = OP Sh = OP Sh > OP Jan.**, Apr.** Sh > OP Jan.*, Apr. n.s. Sh = OP Sh > OP Jan.*, Apr.**, Sep.** Sh = OP Sh = OP
Sh > OP Jan.**, Apr.* Sh = OP Sh = OP Sh > OP Jan.**, Apr.** Sh > OP Jan.**, Apr.** Sh > OP Jan.**, Apr.** Sh > OP Jan.*, Apr. n.s. Sh = OP Sh = OP
Table 24.2 Accumulation of Na, K, Cl at different soil depths under Anabasis articulata and Cornulaca monacantha and between shrubs during different sampling periods (January and April 1994) at the Nizzana experimental site. n.s. Not significant, * p ≤ 0.05, ** p ≤ 0.01 Cornulaca monacantha
Anabasis articulata Na
Under shrubs
Between shrubs
Sep.*
Sep.*
Under shrubs
Between shrubs
Jan.*
Jan.*
Jan.**
Jan.*
Apr.*
Apr.**
Apr.**
Apr. n.s.
Jan.*
Jan.*
Jan.*
Jan. n.s.
Apr.*
Apr.*
Apr.*
Apr. n.s.
Sep. n.s.
Sep. ns
Jan. n.s.
Jan. n.s.
Jan.*
Jan.*
Apr. n.s.
Apr. n.s.
Apr. n.s.
Apr. n.s.
K
Cl
Ion content in the soil decreased with depth Ion content in the soil increased with depth Ion content in the soil showed no differences with depth
Acknowledgements We thank Irmingard Meier and Anja Scheffer for their help with ion analyses. The project was funded by the German Ministry of Education and Science (BMBF).
References Akhani H (2006) Biodiversity of halophytic and sabkha ecosystems in Iran. In: Khan A, Böer B, Kust GS (eds) Sabkha ecosystems, vol II. West and Central Asia. Tasks for Vegetation Science vol 42. Kluwer, Dordrecht, pp 71–88
364
M. Veste et al.
Albert R, Pfundner G, Hertenhagen G, Kästenbauer T, Watzka M (2000) The physiotype approach to understanding halophytes and xerophytes. In: Breckle S-W, Schweizer B, Arndt U (Hrsg) Ergebnisse weltweiter ökologischer Forschung. Heimbach, Stuttgart, pp 69–87 Blume H-P, Yair A, Yaalon DH (1995) An initial study of pedogenic features along a transect across dunes and interdune areas. Nizzana region, Negev Israel. Adv GeoEcol 28:51–64 Bornkamm R, Darius F, Prasse R (1998) Element content of perennial plant species in the sand desert near Nizzana. J Plant Nutr Soil Sci 161:189–195 Breckle S-W (1976) Zur Ökologie und zu den Mineralstoffverhältnissen absalzender und nichtabsalzender Xerohalophyten (unter besonderer Berücksichtigung von Untersuchungen an Atriplex confertifolia und Ceratoides lanata in Utah/USA). Cramer, Berlin, Dissertationes Botanicae 35, pp 1–169 Breckle S-W (1986) Studies of halophytes from Iran and Afghanistan. II. Ecology of halophytes along salt gradients. Proc R Soc Edinburgh 89B:203–215 Breckle S-W (1990) Salinity tolerance of different halophyte types. In: El Bassam N, Dambroth M, Loughman BC (eds) Genetic aspects of plant nutrition. Proc 3rd Int Symp Genetic Aspects of Plant Mineral Nutrition (Developments in Plant and Soil Sciences). Springer, Amsterdam, pp 167–175 Breckle S-W (1995) How do plants cope with salinity? In: Khan MA, Ungar IA (eds) Biology of salt tolerant plants. Proc Int Symp, Department of Botany, University of Karachi, Pakistan, pp 199–221 Breckle S-W (2000) Wann ist eine Pflanze ein Halophyt? Untersuchungen an Salzpflanzen in Zentralasien und anderen Salzwüsten. In: Breckle S-W, Schweizer B, Arndt U (Hrsg) Ergebnisse weltweiter ökologischer Forschungen. Proc 1st Symp A.F.W. Schimper-Foundation, establ. by H. and E. Walter, Hohenheim. Heimbach, Stuttgart, pp 91–106 Breckle S-W (2002a) Salt deserts in Iran and Afghanistan. In: Barth H-J, Böer B (eds) Sabkha ecosystems, vol. I. The Arabian Peninsula and adjacent countries. Tasks for Vegetation Science vol 36. Kluwer, Dordrecht, pp 71–88 Breckle S-W (2002b) Salinity, halophytes and salt affected natural ecosystems. In: Läuchli A, Lüttge U (eds) Salinity. Environment – Plants – Molecules. Kluwer, Dordrecht, pp 53–77 Breckle S-W, Scheffer A, Wucherer W (2001) Halophytes on the dry seafloor of the Aral Sea. In: Breckle S-W, Veste M, Wucherer W (eds) Sustainable land use in deserts. Springer, Berlin Heidelberg New York, pp 139–146 Butnik AA, Japakova UN, Begbaeva GF (2001): Halophytes: structure and function. In: Breckle S-W, Veste M, Wucherer W (eds) Sustainable land use in deserts. Springer, Berlin Heidelberg New York, pp 147–153 Ebeling D (1996) Salzdynamik in Böden des Dünengebeites von Nizzana (Israel). Diplomarbeit, Institut für Geographie, Westf.-Wilhelms Universität Münster El-Ghonemy AA, El-Gazar A, Wallace A, Kish F, Rommel EM (1977) Mineral element composition of perennial vegetation in relation to soil types in the Northeastern corner of the Western desert of Egypt. Bot Gaz 138:192–205 Eshel A (1985) Response of Suaeda aegyptiaca to KCl, NaCl, Na2SO4 treatments. Physiol Plant 64:308–315 Jackson RB, Caldwell MM (1993) Geostatistical patterns of soil heterogeneity around individual perennial plants. J Ecol 81:683–692 Kinzel H (1982) Pflanzenökologie und Mineralstoffwechsel. Ulmer, Stuttgart Macdonald BCT, Melville MD, White I (1999) The distribution of soluble cations within chenopod-patterned ground, arid western New South Wales, Australia. Catena 37:89–105 Munns R (2005) Genes and salt tolerance: bringing them together. New Phytol 167:645–663 Reimann C (2003) Vergleichende Untersuchungen zum Salzhaushalt der Chenopodiaceae, unter besonderer Berücksichtigung der Kalium-Natrium-Verhältnisse. Cramer, Berlin, Dissertationes Botanicae 372, pp 1–303 Reimann C (2005) Die Kalium-Natrium-Verhältnisse der Chenopodiaceae in ihrer Beziehung zu taxonomischen und ökophysiologischen Charakteristika der verschiedenen Arten. In: Veste M,
24 Ion Relations of Plants and Soil Patterns
365
Wucherer W, Homeier J (eds) Ökologische Forschung im globalen Kontext. Festschrift Siegmar-Walter Breckle, Cuvillier, Göttingen, pp 25–43 Reimann C, Breckle S-W (1993) Sodium relations in Chenopodiaceae: a comparative approach. Plant Cell Environ 16:323–328 Rodin LE, Bazilevich I (1967) Production and mineral cycling in terrestrial vegetation. Oliver and Boyd, Edinburgh Rummel B, Felix-Henningsen P (2004) Soil water balance of an arid linear sand dune. Int Agrophys 18:333–337 Sartorius U (1996) Untersuchungen zur Verteilung von Na, K, Cl auf die oberirdische Biomasse und deren kleinräumige Dynamik in einem Längsdünensystem in Nizzana, Israel. Diploma Thesis, University of Bielefeld Shaltout KH (1992) Nutrient status of Thymelaea hirsuta (L.) Endl. in Egypt. J Arid Environ 23:423–432 Teakle JH (1937) The salt (sodium chloride) content of rain water. J Agric West Austr 14:115–133 Veste M (2004) Zonobiom III: Sinai-Halbinsel und Negev-Wüste. In: Walter H, Breckle S-W (Hrsg) Ökologie der Erde, Band 2. Spezielle Ökologie der tropischen und subtropischen Zonen. Elsevier, Spektrum Akademischer, Amsterdam, pp 629–659 Veste M, Breckle S-W (2000) Ionen- und Wasserhaushalt von Anabasis articulata in Sanddünen der nördlichen Negev-Sinai-Wüste. In: Breckle S-W, Schweizer B, Arndt U (Hrsg) Ergebnisse weltweiter Forschung. Heimbach, Stuttgart, pp 481–485 Veste M, Mohr M (2005) Vegetation der Lineardünen der zentralen Namib und deren Ionenhaushalt. In: Veste M, Wucherer W, HomeierJ (eds) Ökologische Forschung im globalen Kontext. Festschrift Siegmar-Walter Breckle, Cuvillier, Göttingen, pp 93–107 Waisel Y (1972) Biology of halophytes. Academic Press, New York Walter H, Breckle S-W (1983) Ökologie der Erde. Band 1. Ökologische Grundlagen in globaler Sicht. UTB-Große Reihe, pp 1–238. Fischer, Stuttgart Walter H, Breckle S-W (1985) Ecological systems of the geobiosphere, vol 1. Ecological Principles in Global Perspective. Springer, Berlin Heidelberg New York Winter K, Troughton JH, Evenari M, Läuchli A, Lüttge U (1976) Mineral ion composition and occurrence of CAM-like diurnal malate fluctuations in plants of coastal and desert habitats of Israel and Sinai. Oecologia 25:125–143 Wucherer W, Breckle S-W (2005) Desertifikationsbekämpfung und Sanierung der Salzwüsten am Aralsee. Sukzession und Phytomelioration, Naturschutz und nachhaltige Entwicklung. Bielefelder Ökologische Beiträge (BÖB) 19:1–94
Chapter 25
Temporal and Spatial Variability of Plant Water Status and Leaf Gas Exchange M. Veste
25.1
Introduction
Arid and semi-arid regions are characterised by low rainfall as well as high potential and actual evaporative demand. Consequentially, water is the major limiting factor for plant growth and productivity. Besides precipitation, hydrological soil properties are most important for soil water availability in the Nizzana sand dunes (Chap. 18, this volume). The vegetation pattern in these sand dunes reflects the spatial differences in soil water availability (Chap. 26, this volume). Detection of spatial heterogeneity requires a high number of soil sensors to evaluate water availability on the landscape level. Unfortunately, the use of tensiometers is limited mainly by excessively low soil water contents in the upper layers. As alternative, phanerophytes are good indicators of water resources in these heterogeneous ecosystems. Desert perennials develop extensive root systems and are able to exploit soil water from deeper horizons (Evenari et al. 1982; Adar et al. 1995; Batanouny 2001; Groom 2003, 2004). Especially shrubs and trees depend on sufficient water resources during the entire year. Water uptake by roots depends on gradients of water potential in the soil–plant–atmosphere continuum. The leaf water potential can be easily and rapidly determined by means of pressure chambers (Scholander et al. 1965) or by thermocouple psychrometers (e.g. von Willert et al. 1995). Commonly used parameters for plant water stress characterisation are the minimum water potential (Ψmin) and the predawn water potential (Ψpd). During the night, the water potential of a nontranspiring plant will equilibrate with the “wettest” water potential of the substrate around the roots, and Ψsoil becomes Ψpd of (Ritchie and Hinkley 1975; Hinckley et al. 1978; Richter 1997). Therefore, Ψpd in many cases will be a good estimate of the soil’s water availability (e.g. Verotec et al. 2001). Information on the spatial and temporal accessibility of soil water in desert ecosystems is limited. The purpose of the present study is to evaluate spatial and temporal changes of soil water availability and their impact on the gas exchange and plant water status of characteristic shrubs in the sand dune system of Nizzana. In this context, measurements of Ψpd are used as diagnostic tool to rapidly determine water availability on the landscape level (Veste and Staudinger 2005).
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
367
368
25.2 25.2.1
M. Veste
Material and Methods Gas Exchange Measurements
Net CO2 exchange and transpiration of fully expanded leaves were measured with an open minicuvette system (CMS 400, Walz GmbH, Effeltrich, Germany), based on Midgley et al. (1997) and Veste and Herppich (1995), and transpiration and CO2 exchange with a differential infrared gas analyser (BINOS 1004P, Rosemont GmbH, Hanau, Germany). Ambient CO2 concentrations were determined by an absolute IRGA (BINOS 1004P), and found to be nearly constant at 350 ppm throughout the day. All data were continuously recorded by means of a personal computer at 5-minute intervals. Gas exchange parameters were calculated after von Cammerer and Farquhar (1981), and based on single leaf area.
25.2.2
Plant Water Potential
Plant water potential (ΨW) was determined by a Scholander-type pressure chamber (Plant Water Status Console 3000, Soilmoisture Inc., Santa Barbara, CA), after Scholander et al. (1965) and Turner (1988). Terminal shoots were covered with aluminium foil to prevent transpirational loss, and cut with a razor blade. Predawn water potentials (Ψpd) were measured starting 1 h before sunrise. Studies in Nizzana (site 1) were carried out between October 1994 and October 1997.
25.3 25.3.1
Results Photosynthesis
Net CO2 exchange and transpiration of Thymelaea hirsuta were measured in the interdunes. Characteristic diurnal courses of CO2 uptake are shown in Fig. 25.1. The maximum net CO2 exchange rates are rather high and, in wet years, can reach more than 20 µmol m−2 s−1. Even in the dry period, however, maximum CO2 uptake reaches up to 20 µmol m−2 s−1. During continuation of the drought period in the following winter, CO2 uptake is substantially reduced due to stomatal closure (Fig. 25.1a). CO2 uptake of the C4-plant Anabasis articulata reached 11 mol m−2 s−1 on sandy soils, and showed no midday stomatal closure typical for desert plants. On the playa, soil water availability is strongly limited, resulting in a closure of the stomata in the morning hours (Fig. 25.2).
25 Temporal and Spatial Variability of Plant Water Status and Leaf Gas Exchange
24
Thymelaea hirsuta A
20
net photosynthesis [µmol m−2s−1]
369
21. Apr. 1993 26. Oct. 1993
16
25. Feb. 1994
12 8 4 0
B
20 16 12 8
14. Mar 1995
4
15. Mar. 1995
0
16. Mar. 1995
6
8
10
12
14
16
18
local time [hours]
net photosynthesis [µmol m−2 s−1]
Fig. 25.1 Diurnal courses of net CO2 exchange of Thymelaea hirsuta. A At the end of the good rainy period with 145 mm rainfall (21 April 1994), at the end of the dry period (26 October 1994), after only 27 mm winter rainfall by 23 February 1995, and B in March 1995
12
Anabasis articulata
10 8 6 4 2 0 −2
5
7
9
11 13 15 local time [hours]
sand
17
19
playa
Fig. 25.2 Diurnal courses of net CO2 exchange of Anabasis articulata growing on sandy soils (18 January 1994) and playa (14 April 1994)
370
25.3.2
M. Veste
Water Potential
Diurnal changes in leaf water potential in the shrub species investigated in the interdune area and on the dune slope after the end of the rainy season near Nizzana are shown in Fig. 25.3. After sunrise, leaf water potential decreases due to increasing plant transpiration but rapidly starts to recover after sunset. No significant difference in leaf water potential between the different species as well as between the plants growing on the dune and in the interdune could be detected after winter rainfall. However, seasonal variations in predawn water potentials (Ψpd) were more pronounced in the interdune area than on the dune slope (Fig. 25.4). At the end of the dry season, Ψpd of shrubs in the interdune was between −2.42 MPa in Thymelaea and −1.95 MPa in Retama raetam. In the same dry season, predawn water potentials of plants growing on the slope were only −0.7 MPa (Retama) and −0.8 MPa (Artemisia) (Fig. 25.4). After rain, predawn water potential increased within a few days in all species investigated. Similar spatial and seasonal changes in predawn water potential were observed in the following years (Fig. 25.4). In A. articulata, spatial differences in Ψpd could be observed (Fig. 25.5). At the end of the rainy season in April, shrubs growing on the fine-grained soils of the interdunes showed lower water potential (Ψpd=−2.53 MPa) than those growing on sandy soils (Ψpd=−1.62 MPa). This difference in water potential clearly reflects differences in water availability between the two soil types. At the end of the dry season, predawn water potential decreased to −4.2 MPa on the playa and −3.88 MPa on sandy soils.
Ψpd [MPa]
−0.4 −0.8 −1.2 −1.6 −2.0 −2.4
6
8
10 12 14 local time [hours]
16
18
Artemisia (interdune)
Thymelaea (interdune)
Artemisia (slope)
Retama (interdune)
Fig. 25.3 Diurnal courses of plant water potential of Thymelaea hirsuta, Retama raetam and Artemisia monosperma growing in the interdune (closed symbols) and on the slope (open symbols) at the end of the rainy season (4 April 1995)
25 Temporal and Spatial Variability of Plant Water Status and Leaf Gas Exchange
371
−0.4
Ψpd [MPa]
−0.8 −1.2 −1.6 −2.0
March 1997
Sep 1996
April 1995
Nov 1994
Oct 1994
−2.4
Retama (slope)
Retama (interdune)
Artemisia (slope)
Thymelaea (interdune)
Fig. 25.4 Seasonal variations of predawn leaf water potential (Ψpd) of Thymelaea hirsuta, Retama raetam and Artemisia monosperma growing in the interdune (closed symbols) and on the slope (open symbols) of the Nizzana sand dunes
0.0 Thymelaea (sand)
Ψpd [MPa]
−1.0 −2.0 −3.0 Anabasis (sand)
−4.0 Oct 1994
Anabasis (playa)
Nov 1994
April 1995
Fig. 25.5 Seasonal variations of predawn leaf water potential (Ψpd) of Anabasis articulata growing on sand and on playa, compared to that of Thymelaea hirsuta on sand
25.4
Discussion
The water potential measurements and gas exchange measurements in Nizzana clearly indicate that the shrubs on the dune have access to water flow in the sand dunes (Chap. 18, this volume). Studies by Pavlik (1980) on sand dune plants in the Eureka Valley of California already showed that the water relations of the plants are more comparable to those of mesic habitats. The root system of desert perennials
372
M. Veste
have characteristics and adaptations that enable water uptake from a large soil volume. In the sand dunes of Nizzana, water can infiltrate in sand layers deeper than 3 m, and lateral water flows are observed (Yair et al. 1997). After a 2-day rainstorm that yielded 42 mm of precipitation, direct infiltration into the dune was limited to the upper soil layers (to 0.6 m). During summer, the upper sand layers are completely dry and water is available only in deeper layers. Detailed information about the rooting depths of the shrubs investigated in the sand dunes is still missing. The taproot system of A. articulata (Fig. 25.6) has been excavated in the sand dunes to a depth of 2.5 m (Veste and Breckle 1996b, 2000). The rooting depth of T. hirsuta is more than 3.5 m in wadis (Evenari et al. 1982), that of R. raetam may exceed 10 m (Zohary and Fahn 1952) and, for Acacia tortilis from the northern Negev, water uptake from 12 m depth has been found by Adar et al. (1995). All species mentioned above develop an additional surface root system (“T-root system”) enabling the plants to use water from both layers discussed above. However, the ratio of the water-collecting efficiency of both rooting systems is unclear. Banksia tree species in Mediterranean sand dunes in Australia develop T-root systems, too. The deep root system of Banksia species reaches 8–9 m (Groom 2004). In this case, investigations showed a rapid switch in water uptake from upper soil water to groundwater and deeper soil water sources with the onset of the dry summer period (Downson and Pate 1996; Zencich et al. 2002). In the winter rainy season, the shallow root system collects high amounts of water from the upper soil. In the summer months, predawn water potential was lower for dune crest plants than for plants in the interdune, due to the larger distance (up to 30 m)
Fig. 25.6 Taproot system of Anabasis articulata
25 Temporal and Spatial Variability of Plant Water Status and Leaf Gas Exchange
373
to groundwater sources. Based on a relatively high shoot water potential maintained during the dry season, Zencich et al. (2002) concluded that subsurface water storage within Australian dunes is a sufficient water source for the trees characterising this environment. These findings from Australian dunes support our results based on water potential measurements, and soil water investigations showing that water sources within deeper layers of the dunes play an important role for the survival of dune plants in the Negev (Yair et al. 1997). Hydraulic lift can also contribute to soil water heterogeneity. Several phreatophytes from very different deserts (e.g. Artemisia tridentata, Banksia prionotes) absorb water with their deep-reaching root system and release it from the root into the dry upper soil layer at night, taking it up again at daytime (Caldwell et al. 1998; Burgess et al. 2000). Still controversial is the question of whether the predawn water potential is in equilibrium with the soil water potential. Studies by Donovan et al. (1999, 2001) with the cold-desert shrubs Chrysothamnus nauseosus and Sarcobatus vermiculatus showed that predawn plant water potential was significantly lower than soil water potential. Also the Chenopodiaceae A. articulata and Cornulaca monacantha growing in the sand dunes of the Negev showed lower water potential than the shrubs investigated in this study (Veste and Breckle 1995, 2000). This discrepancy is consistent with the accumulation of salts and other organic solutes in the leaves of halophytes and xerohalophytes. Accumulation of inorganic and organic solutes results in lowering the osmotic potential. In xerohalophytes, accumulation mainly of NaCl is a common feature and the ion patterns are genetically fixed (Breckle 1990; Albert et al. 2000; Chap. 24, this volume). Even on non-saline soils, such plants accumulate high amounts of salts. Therefore, it is not surprising that in (xero-) halophytes the water potential is substantially lower than the soil water potential. Another factor potentially leading to a plant–soil disequilibrium is night-time transpirational loss. Its contribution was approx. 0.1 MPa in Chrysothamnus and approx. 0.6 MPa in Sarcobatus (Donovan et al. 1999). In our study, the contribution of night-time transpiration cannot be fully excluded, but gas exchange measurements of Artemisia monosperma and T. hirsuta showed a minimal transpiration rate during the night (Veste and Breckle 1996a). In fact, the aims of our investigations are the relative temporal and spatial differences, rather than the exact measurement of the soil water potential of the wettest soil. In a split root experiment with walnut trees, Améglio et al. (1999) showed that the predawn water potential equilibrated with the wet soil conditions – it does not reflect the mean water status of the soil. An important role for the recovery of leaf water potential after sunset is also soil hydraulic conductivity. In A. monosperma, R. raetam and T. hirsuta, leaf water potential regained its predawn values within 1 h after sunset. This implies good water accessibility. Other studies showed a time lag in the recovery of water potential. In these cases, soil hydraulic conductivity declines and water uptake is limited by a reduced water flow rate at the soil–root interface (Kutilek and Nielsen 1994; Schmidthalter 1997). However, even when the predawn water potential does not exactly reflect the water potential of the wettest soil, it is nevertheless a good estimation of a mean value for the nearest root zone soil water potential. Measuring water potential as done in the present study has the advantage that the instrument selected is very easy
374
M. Veste
to use. In contrast, other methods to determine soil water content and soil water potential are often at their physical limits in dry soils. Thermocouple psychrometers and especially tensiometers correspond to direct methods but they work only at moderate or high soil water potential, and salinity further limits their application. Even TDR systems have limits due to the low water content of sandy soils (e.g. Rummel and Felix-Henningsen 2004). In addition, instrumental investigations in most cases cover only the upper soil layers, and not the entire rooting soil volume.
25.5
Conclusions
Predawn water potential shows spatial differences in plant water availability in xeric ecosystems, and can be used as a diagnostic tool to measure and map changes in water availability in desert sand dunes in an integrative manner. Thus, plant water relations can be easily related to hydrological processes in sand dunes. Acknowledgements Thanks go to Werner B. Herppich (Potsdam) for valuable comments on the manuscript. The project in the Negev desert was funded by the Federal Ministry of Education and Research (BMBF) by a grant to the Department of Ecology, University of Bielefeld (BEO 0339495A).
References Adar E, Gev I, Berliner P, Knol-Paz I (1995) Water recharge and percolation in sand dune terrain. In: Field Guidebook Int Conf Geomorphic Response of Mediterranean and Arid Areas to Climate Change (GERTEC), Field Trip B. Hebrew University of Jerusalem, pp 1–12 Albert R, Pfundner G, Hertenhagen G, Kästenbauer T, Watzka M (2000) The physiotype approach to understanding halophytes and xerophytes. In: Breckle S-W, Schweizer B, Arndt U (eds) Ergebnisse weltweiter ökologischer Forschung. Günter Heimbach, Stuttgart, pp 69–87 Amélio T, Archer P, Cohen M, Valancogne C, Daudet FA, Dayau S, Cruiziat P (1999) Significance and limits in the use of predawn water potential for tree irrigation. Plant Soil 207:155–167 Batanouny KH (2001) Plants in the deserts of the Middle East. Springer, Berlin Heidelberg New York Breckle S-W (1990) Salinity tolerance of different halophyte types. In: Bassam N El (ed) Genetic aspects of plant mineral nutrition. Kluwer, Amsterdam, pp 167–175 Burgess SSO, Pate JS, Adams MA, Dawson TE (2000) Seasonal water acquisition and redistribution in the Australian woody phreatophyte, Banksia prionotes. Ann Bot 85:215–224 Caldwell MM, Dawson TE, Richards JH (1998) Hydraulic lift: consequences of water efflux from roots of plants. Oecologia 113:151–161 Dawson T, Pate J (1996) Seasonal water uptake and movement in root systems of Australian phreatophytic plants with a dimorphic root morphology: a stable isotope investigations. Oecologica 107:13–21 Donovan LA, Grisé DJ, West JB, Pappert RA, Alder NN, Richards JH (1999) Predawn disequilibrium between plant and soil water potentials in two cold-desert shrubs. Oecologia 120:209–217 Donovan LA, Linton MJ, Richards JH (2001) Predawn plant water potential does not necessarily equilibrate with soil water potential under well-watered conditions. Oecologia 129:328–335 Evenari M, Shanan L, Tadmor W (1982) The Negev – The challenge of a desert. Harvard University Press, Cambridge, MA
25 Temporal and Spatial Variability of Plant Water Status and Leaf Gas Exchange
375
Groom PK (2003) Groundwater-dependency and water relations of four Myrtaceae shrub species during a prolonged summer drought. J R Soc Western Austr 86:31–40 Groom PK (2004) Rooting depth and plant water relations explain species distribution patterns within a sandplain landscape. Funct Plant Biol 31:423–428 Hinckley TM, Lassoie JP, Running SW (1978) Temporal and spatial variations in the water status of forest trees. Foren Sci Monogr 20:1–72 Kutilek M, Nielsen DR (1994) Soil hydrology. GeoEcology textbook. Catena, Cremlingen Midgley G, Veste M, von Willert DJ, Davis GW, Steinberg M, Powrie LW (1997) Comparative field performance of three different gas exchange systems. Bothalia 27(1):83–89 Pavlik BM (1980) Patterns of water potential and photosynthesis of desert sand dune plants, Eureka Valley, California. Oecologica 46:147–154 Richter H (1997) Water relations of plants in the field: some comments on the measurement of selected parameters. J Exp Bot 48:1–7 Ritchie GA, Hinckley TM (1975) The pressure chamber as an instrument for ecological research. Adv Ecol Res 9:165–254 Rummel B, Felix-Henningsen P (2004) Soil water balance of an arid linear sand dune. Int Agrophys 18:333–337 Schmidthalter U (1997) The gradient between pre-dawn rhizoplane and bulk soil matric potentials, and its relation to the pre-dawn root and leaf water potentials of four species. Plant Cell Environ 20:953–960 Scholander PF, Hammel HT, Bradstreet ED, Hemmingsen EA (1965) Sap pressure in vascular plants. Science 148(3668):339–346 Slayter RO (1968) Plant-water-relationships, 2nd edn. Academic Press, London Turner NC (1988) Measurements of plant water status by pressure chamber technique. Irrig Sci 9:289–308 Vertovec M, Sakcali S, Ozturk M, Salleo S, Giacomich P, Feoli E, Nardini A (2001) Diagnosing plant water status as a tool for quantifying water stress on a regional basis in Mediterranean drylands. Ann Forest Sci 88:113–125 Veste M, Breckle S-W (1995) Xerohalophytes in a sandy desert ecosystem. In: Khan MA, Ungar IA (eds) Biology of salt tolerant plants. University of Karachi, Pakistan, pp 161–165 Veste M, Breckle S-W (1996a) Gaswechsel und Wasserpotential von Thymelaea hirsuta in verschiedenen Habitaten der Negev-Wüste. Verhandl Gesell Ökol 25:97–103 Veste M, Breckle S-W (1996b) Root growth and water uptake in a desert sand dune ecosystem. Acta Phytogeogr Suec 81:59–64 Veste M, Breckle S-W (2000) Ionen- und Wasserhaushalt von Anabasis articulata in Sanddünen der nördlichen Negev-Sinai-Wüste. In: Breckle S-W, Schweizer B, Arndt U (eds) Ergebnisse weltweiter Forschung. Günter Heimbach, Stuttgart, pp 481–485 Veste M, Herppich W (1995) Diurnal and seasonal fluctuations in the atmospheric CO2 concentration and their influence on the photosynthesis of Populus tremula. Photosynthetica 31(3):371–378 Veste M, Staudinger M (2005) Räumliche Variabilität der pflanzlichen Wasserversorgung an Trockenstandorten in Südmarokko. In: Veste M, Wissel C (Hrsg) Beiträge zur Vegetationsökologie der Trockengebiete und Desertifikation. UFZ Bericht 1/2005:55–64 von Cammerer S, Farquhar GD (1981) Some relationship between the biochemistry of photosynthesis and the gas exchange of leaves. Planta 153:376–387 von Willert DJ, Mattysek R, Herppich WB (1995) Experimentelle Pflanzenökologie – Grundlagen und Anwendungen. Thieme, Stuttgart Yair A, Lavee H, Greitser N (1997) Spatial and temporal variability of water percolation and movement in a system of longitudinal dunes, western Negev, Israel. Hydrol Processes 11:43–58 Zencich SJ, Froend RH, Turner JV, Gailitis V (2002) Influence of groundwater depth on the seasonal sources of water accessed by Banksia tree species on a shallow, sandy coastal aquifer. Oecologia 131:8–19 Zohary M, Fahn A (1952) Ecological studies on East Mediterranean dune plants. Bull Res Council Israel Sect D1:38–53
Chapter 26
Standing Biomass and its Modelling M. Veste, C. Sommer, S.-W. Breckle, and T. Littmann
26.1
Introduction
The sand dunes of the north-western Negev are characterized by a small-scale vegetation pattern (cf. Chap. 8, this volume). Aim of this investigation is to distinguish the standing biomass in the major ecotopes resulting from long-term ecological processes controlling ecotope pattern. Furthermore, we will present a simple numerical approach for modelling the actual standing biomass distribution in the sand dune mosaic.
26.2 26.2.1
Standing Biomass Methods
The standing biomass was investigated at the Nizzana test site (site N1) in spring 1994 by means of non-destructive measurements (Sommer 1996). For eight species, the relation between dry biomass and size index was calculated using either a regression model 1 (y = a + bx) or model 2 (y = a + b1x + b2x2), where y is the dry biomass, x the size index, and b the species, all specific constants (Fig. 26.1). The regression equation for each shrub species is listed in Table 26.1. The biomass of perennials was estimated on five 5 × 5 m plots in different geo-ecological units using the non-destructive method, or it was completely harvested.
26.2.2
Biomass
The standing biomass of the perennial vegetation is shown in Table 26.2. The highest biomass density and vegetation cover can be found in a small belt along the dune bases (Fig. 26.2). In general, the biomass values in the different ecotopes correspond well with the mean values of the vegetation cover of the perennials. The playa area S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
377
378
M. Veste et al. 16000 14000
Biomass [g]
12000
Anabasis
10000 8000 6000 4000
Noaea
2000 0
0
10000 20000 30000 40000 50000 60000
Size index Fig. 26.1 Correlation between size index and dry biomass using a linear regression model (Noaea mucronata) or an exponential regression model (Anabasis articulata) (data from Sommer 1996)
Table 26.1 Regression functions for the calculation of aboveground biomass (data used from Sommer 1996) Plant species Function R2 Anabasis articulata Artemisia monosperma Cornulaca monacantha Noaea mucronata Convolvulus lanata Moltkiopsis ciliata Retama raetam Thymelaea hirsuta
y = 4.5 10−8x2 + 0.005867x + 41.581040 y = 8.71 10−2x−24.776 y = 7.482 10−3x−391.565075 y = 3.384 10−3x−41.825269 y = 3.171 10−3x−1.011180 y = 2.996 10−3x−20.464226 y = 6.902 10−2x−108.340199 y = 8.977 10−2x−5.228655
0.999 0.996 0.855 0.984 0.962 0.896 0.784 0.982
Table 26.2 Dry biomass (kg 100 m−2) of dominating perennials at the Nizzana test site in spring 1994 Crest, Crest, N Dune Shrub mobile sand semi-mobile slope base Interdune Anabasis articulata Artemisia monosperma Convolvulus lanatus Cornulaca monacantha Heliotropium digynum Moltkiopsis ciliata Noaea mucronata Retama raetam Stipagrostis scoparia Thymelaea hirsuta Others Total biomass
6.41 0.92 4.88 12.21
0.59 2.06 0.14 8.43 0.67 7.84 0.09 19.82
0.71 0.44 2.23 0.10 0.10 2.25 8.44 3.23 2.48 1.89 21.87
1.93 1.10 0.68 2.78 0.61 0.59 17.75 3.60 3.00 3.08 2.67 37.79
6.88 0.55 2.67 0.03 1.00 3.16 0.31 1.19 0.25 2.46 18.50
379
biomass
40
40
cover
30
30
20
20
10
10
0
crest
slope
dune base interdune
vegetation cover [%]
biomass [kg 100 m−2]
26 Standing Biomass and its Modelling
0
Fig. 26.2 Mean biomass and vegetation cover in various ecotopes in the linear dunes of the Nizzana research site (site N1)
is dominated by the Chenopodiaceae Anabasis articulata. On the playa, total biomass is the lowest of the entire Nizzana test site, reaching 1.1 kg 100 m−2 (Anabasis 99%); at the playa edges, the corresponding value is 25.4 kg 100 m−2 (Anabasis 71.5%). The old dunes are dominated by Cornulaca monacantha and the biomass is here 31.1 kg 100m−2 (Cornulaca 67.2%).
26.3 26.3.1
Modelling Biomass Pattern The Meso-Scale Model
A stochastic model for biomass distribution over the entire sand dune field (mesoscale) was developed to simulate the biomass in the entire dune field (Littmann and Veste 2005). The VEGDUNE model is based on input data (mean values over the 1998–2000 observational period) for the following abiotic parameters: distance from the sea, relief energy, percentage of mobile sand per unit, infiltration rate and infiltration depth, radiation balance, rainfall, dewfall, evapotranspiration, frequency of stable layers, and dewpoint temperature difference. The distance to the sea was computed trigonometrically for each grid point. The relief energy is a measure of the complexity of the terrain, and is expressed as the cross product of the elevations of the four edges of a grid cell. The width of the geometric grid was 100 m for the meso-scale simulation. These parameters were used in a stepwise multiple regression analysis, and those parameters showing a correlation of at least 95% with the actual biomass index at the four stations were selected for the formulation of the following regression equation: Biomass index= – 0.4* distance from sea – 2.2* relief energy – 8.9* 10– 4 * rainfall + 23.44 (26.1)
380
M. Veste et al.
With Eq. (26.1), it was possible to compute and interpolate a biomass index value for each grid point on the meso-scale (Fig. 26.3). The meso-scale model revealed patches of high standing biomass in small depressions between steep sand dune slopes in many parts of the southern dune field, irrespective of rainfall totals. It is evident from the equation that the geo-ecological gradient from north to south is still a dominant feature in the overall spatial distribution, as biomass is negatively linked to the distance from the sea and relief energy; both indicate higher overall biomass in the northern, flatter part of the sand dune field. However, biomass is also negatively linked to rainfall. Confirmatory assessments of biomass at 16 locations selected randomly across the entire transect revealed a very good correlation with the values modelled for the corresponding sites (Fig. 26.4).
44000.00
43000.00
42000.00
30.00
25.00 41000.00 20.00 40000.00 15.00 39000.00
38000.00
10.00
5.00
0.00 37000.00
36000.00 88000.00 89000.00 90000.00 91000.00 92000.00 93000.00
Fig. 26.3 Modelled vegetation pattern on the meso-scale in the southern sand dunes of Nizzana using the VEGDUNE model. Scale of biomass index, relative units
26 Standing Biomass and its Modelling
381
28 Y = 0.937 * X + 0.011 R2= 0.95
Cover [%] model
24 20 16
.
12 8 4 0
0
4
8
12
16
20
24
28
Cover [%] observed Fig. 26.4 Evaluation of the modelled biomass cover of perennials
30.00
25.00
20.00
15.00
10.00
5.00
0.00
Fig. 26.5 Modelled biomass pattern in the Nizzana linear dunes, using the VEGDUNE model. Scale of biomass index, relative units
26.3.2
The Micro-Scale Model
The meso-scale model equation was applied to the micro-scale data for the Nizzana test site (N1). The width of the geometric grids was here 10 m. The output of the micro-scale model shows a very realistic simulation of the spatial pattern of standing biomass in the sand dunes at Nizzana (Fig. 26.5). Even the differences between the upper and lower slopes towards the dune base could be modelled satisfactorily.
382
M. Veste et al.
Correlations between the modelled vegetation pattern and measurements of the vegetation cover using aerial photographs confirmed the three-dimensional biomass model for Nizzana (Littmann and Veste 2005). However, the biomass on the southfacing slopes was overestimated. At the Nizzana site, the steep dune slopes are not covered by biological soil crusts, and the mobile sand prevents the establishment of more vegetation.
26.4
Discussion
Sandy dunes are favourable habitats for desert plants due to higher water availability than for other desert types. The standing biomass in Nizzana, at an average annual rainfall of 90 mm year−1, is higher than in the rocky Central Negev desert with similar rainfall amounts. The biomass in the stable sand dunes ranges between 1,850 and 3,750 kg ha−1, whereas on the rocky slopes of Sede Boqer (mean annual rainfall 97 mm year−1) values of 759–1,464 kg ha−1 were recorded, and of 460–940 kg ha−1 in Avdat (88 mm year−1; Evenari et al. 1976; Esser 1989). These differences show that surface hydrology properties (e.g. field capacity, infiltration rates, runoff) effectively control vegetation development, a fact which is aggravated in arid environments; sand dunes show much better water budgets than do rocky or loessial areas (Yair and Berkowicz 1989; Yair 1994; Chaps. 17 and 18, this volume). On the playas, the low water availability and hard soil structures limit plant growth. Only A. articulata is able to grow here (Veste and Breckle 2000). However, the individuals are smaller than in the adjacent sandy areas. The vegetation pattern along the geo-ecological gradient is the result of a complex interrelation of contrasting process gradients on the meso- and micro-scale level. The complexity of the terrain (e.g. narrow and crossed sand dune ridges with steep slopes and small interdunes) with high relief energy favours micro-scale habitats, irrespective of annual rainfall amounts. Sand mobility is one major factor controlling the vegetation pattern in a sand dune ecosystem. Surface properties of, e.g. crusts and fine material cover primarily control water redistribution on the micro-scale level (Chap. 17, this volume) and, thus, vegetation pattern.
26.5
Conclusions
Numerous large-scale comparisons have shown positive relationships between standing biomass and rainfall (Walter 1939; Shmida 1985; Kutiel and Lavee 1999). Vegetation cover and biomass show the same decrease from humid to arid climates as does mean annual rainfall, supposedly the main controlling factor along climatic gradients. However, in our study on the meso-scale level, standing biomass is negatively correlated with annual rainfall. This is a most interesting result which points to patches of higher standing biomass in favourable habitats throughout the sand
26 Standing Biomass and its Modelling
383
dune field – especially in the higher sand dunes of the southern, drier part. This finding emphasizes the importance of surface properties controlling water availability along the climatic gradient and, eventually, the standing biomass and vegetation cover in the sand dunes (Chap. 18, this volume). With increasing rainfall towards the north, soil crust thickness increases and limits deep water infiltration (Littmann et al. 2000; Chap. 29, this volume). Reduced soil water availability counteracts the positive effects of rainfall and limits the biomass production of higher plants (Veste and Littmann, unpublished data). Acknowledgements We thank Kerstin Eggert and Anja Scheffer for their help with the field work. The project was funded by the German Ministry of Education and Science (BMBF).
References Esser U (1989) Zum Stickstoff-Haushalt arider Hangökosysteme im nördlichen Negev-Hochland, Israel und den Auswirkungen der “Hang-Minicatchment-Technologie” auf Stickstoffumsätze und -vorräte. Arbeitsber Lehrstuhl Landschaftsökologie Münster 9 Evenari M, Schulze E-D, Lange OL, Kappen L, Buschbom U (1976) Plant production in arid and semi-arid areas. Springer, Berlin Heidelberg New York, Ecological Studies 19, pp 19–35 Kutiel P, Lavee H (1999) Effect of slope aspects on soil and vegetation properties along an aridity transect. Israel J Plant Sci 47:169–178 Littmann T, Veste M (2005) Modelling spatial pattern of vegetation in desert sand dunes. Forestry Stud China 7(4):24–28 Littmann T, Hering E, Koch S (2000) What happens to rainfall at the desert margin? Water infiltration experiments in a sandy arid area. Hallesches Jahrb Geowiss 22:49–58 Shmida A (1985) Biogeography of desert flora. Ecosystems of the World, vol 12A. Elsevier, Amsterdam Sommer C (1996) Untersuchungen zur oberirdischen Biomasse der perennen Pflanzenarten verschiedener Ökosystemeinheiten in einem Längsdünensystem bei Nizzana, Israel. Diploma Thesis, Department of Ecology, University of Bielefeld Veste M, Breckle SW (2000) Ionen- und Wasserhaushalt von Anabasis articulata in Sanddünen der nördlichen Negev-Sinai-Wüste. In: Breckle SW, Schweizer B, Arndt U (Hrsg) Ergebnisse weltweiter ökologischer Forschung. Günter Heimbach, Stuttgart, pp 481–485 Walter H (1939) Grasland, Savanne und Busch der ariden Teilen Afrikas in ihrer ökologischen Bedingtheit. Jahrb Wiss Bot 87:750–860 Yair A (1994) The ambiguous impact of climate change at a desert fringe: northern Negev. In: Millington AC, Pye K (eds) Environmental change in drylands: biogeographical and geomorphological perspectives. Wiley, New York, pp 183–220 Yair A, Berkowicz SM (1989) Climatic and non-climatic controls of aridity: the case of the Northern Negev of Israel. Catena suppl 14:145–158
Chapter 27
Effects of Shrubs on Annual Plant Populations K. Tielbörger and R. Kadmon
27.1
Introduction
Most plant species may be found in many different habitat types, even within relatively small geographic regions. Consequently, probabilities of survival and reproduction may vary for local subpopulations of the same species, depending on which habitat they occupy (e.g. Wiens 1976; Mack and Pyke 1983; Fowler 1988; Weiss et al. 1988; Kadmon 1993). In arid and semi-arid environments, local-scale gradients in habitat conditions are often produced by the presence of shrubs. For example, shrubs may positively affect the success of subcanopy plants by providing shade and reducing evaporation loss and extreme temperatures (Shreve 1931; Nobel 1980; Franco and Nobel 1989; Valiente-Banuet and Ezcurra 1991; Turner et al. 1996), and by enhancing soil nutrient content (Garcia-Moya and McKell 1970; Weinstein 1975; Charley and West 1975; Franco and Nobel 1989; Rostagno et al. 1991; Gutiérrez et al. 1993; Sarig et al. 1994; Pugnaire et al. 1996). Negative effects of shrubs include water interception (Pressland 1976; Tromble 1988; Martinez-Meza and Whitford 1996), allelopathy (Muller and Muller 1956; Friedman et al. 1977; Jackson and Caldwell 1993), competitive effects (Holzapfel and Mahall 1999), and the reduction of light (Franco and Nobel 1989). It has been found that plant seedlings emerging below shrubs experience different probabilities of survival and/or reproduction than do conspecific seedlings germinating in open areas between shrubs (Weinstein 1975; Halvorson and Patten 1975; Keeley and Johnson 1977; Nelson and Chew 1977; Friedman et al. 1977; Jaksic and Fuentes 1980; Shmida and Whittaker 1981; Gutiérrez and Whitford 1987; Dean and Milton 1991; Gutiérrez et al. 1993; Tielbörger and Kadmon 1995; Pugnaire et al. 1996; Holzapfel and Mahall 1999; Tielbörger and Kadmon 1996, 2000). Most of the existing literature has been dominated by the perception of desert shrubs as ‘nurse plants’ (e.g. Niering et al. 1963; Turner et al. 1966; Franco and Nobel 1989) or as ‘fertile islands’ in a nutrient-poor matrix (e.g. Schlesinger et al. 1996). This reflects the general conception that desert shrubs have predominantly positive effects on their understory. However, it has been found that the response of desert annuals to the presence of shrubs may be highly species-specific and that negative shrub effects are not uncommon (Weinstein 1975; Shmida and Whittaker 1981; Tielbörger and Kadmon 1997, 2000).
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
385
386
K. Tielbörger, R. Kadmon
While there is no question that patterns of desert annual plants are influenced by biologically induced patchiness caused by the presence of shrubs, the demographic processes responsible for the observed patterns have rarely been investigated. Only a few studies have measured fitness parameters (e.g. survival or reproduction) of desert annual plants (but see Tielbörger and Kadmon 1995; Tielbörger 1997b; Holzapfel and Mahall 1999). Most studies have assessed densities of annual plants and have inferred, from the detected patterns of abundance, whether the plants performed better under shrubs or in open areas. Yet, plant distribution patterns do not enable one to identify the demographic processes responsible for the observed patterns. Additionally, such patterns can not be automatically related to differences in habitat quality, since densities are affected by a variety of factors including competition, dispersal, etc. The lack of detailed demographic data in desert shrublands highlights the importance of studying both densities and fitness parameters of annual plants. In arid environments, the availability of water, the main factor limiting plant growth, is highly unpredictable and variable from year to year. Yet, most studies on the effect of desert shrubs on annual vegetation have been based on a single year of observation (but see Weinstein 1975; Nelson and Chew 1977; Tielbörger 1997a; Tielbörger and Kadmon 2000). The interpretation of results obtained from such short-term studies is questionable because desert annuals may show considerable year-to-year variations in their patterns of germination, survival, growth and reproduction (Juhren et al. 1956; Beatley 1967, 1969; Nelson and Chew 1977; Loria and Noy-Meir 1979/1980; Bowers 1987; Inouye 1991; Kadmon 1993; Tielbörger and Kadmon 2000; Shmida et al., unpublished data). Also, the relative favourability of habitat types may differ or even be reverted in years with variable rainfall conditions (Tielbörger and Kadmon 2000). The above findings point to the importance of studying demographic responses of desert plants over several growing seasons. In the present article, we present the main findings of an extensive, 4-year study designed to investigate the effect of shrubs on annual plants in the Nizzana sand dunes. The main purposes of the study were to document the effect of perennial plants on patterns of annual plant distribution, to identify the main demographic processes responsible for the observed patterns, to test whether demographic responses of annuals to the presence of shrubs differ between species, and to evaluate the degree to which the responses are stable over consecutive years.
27.2
Methods
The study was conducted in an area of partly stabilized sand along a lower, northfacing dune slope during four successive growing seasons, from September 1993 to May 1997. Total cover of perennial plants in the selected type of habitat was 30%, and the dominant perennial species were the dwarf shrubs Moltkiopsis ciliata (Forss.) I.M. Johnst. (Boraginaceae) and Heliotropium digynum (Forss.) C. Chr.
27 Effects of Shrubs on Annual Plant Populations
387
(Boraginaceae), and the grass Stipagrostis scoparia (Trin. and Rupr.) De Winter. Results from previous studies (Tielbörger 1997a; Tielbörger and Kadmon 1997) have indicated that densities and biomass of most annual plant species are usually higher under the canopy of perennial shrubs than in the open areas between these. Four annual species were selected for the study: Senecio glaucus L. (Asteraceae), Ifloga spicata (Forss.) Sch. Bip. (Asteraceae), Rumex pictus Forss. (Polygonaceae) and Erodium laciniatum (Cav.) Willd. (Geraniaceae). These species were chosen because they were among the most abundant annual species in the study area (Tielbörger 1997a), and they occurred both under shrubs and in open areas.
27.2.1
Sampling Design
All measurements were conducted in a randomized block design with repeated measures. In September 1993, four permanent blocks (replicates) of approximately 50×70 m were marked in the selected experimental area. The area of each block was subdivided into two habitat types: shrubs and openings. The ‘shrub’ habitat was defined as the zone beneath the canopies of perennial plants, and the remaining area was defined as openings. All subsequent measurements were done in these two types of habitat.
27.2.2 Measurements of Seedling Densities and Seedling Survival Before the first rains in 1993, 16 permanent quadrats of 25×25 cm were placed randomly in each habitat type and block. Seedlings of the four selected annual species emerging in the 64 quadrats were counted several times after the first rainfall of each season, depending on total amount and within-year distribution of rainstorms. Seedlings emerged in one cohort in the second and third year (1994–1995, and 1995–1996), and in two distinct cohorts in the first (1993–1994) and last year (1996–1997) of study. Densities of emerging plants were defined as the sum of maximum densities recorded for each cohort. The last count in each season recorded the densities of annual plants which survived to seed production. Seedling survival was defined as the percentage of emerging plants surviving to seed set, and was estimated for each block separately.
27.2.3
Measurements of Reproductive Success
During the time of seed set of each season, individuals of the four focal species were collected randomly and the number of seeds produced per plant was determined. In the first, second and fourth study season, 25 individuals were collected for each species, habitat type and block. Due to low densities and high mortality in
388
K. Tielbörger, R. Kadmon
the third year of study (1996), no individuals could be collected for S. glaucus and R. pictus and, for I. spicata, only 48 individuals (12 per block) were collected in the openings. Reproductive success was defined as the number of seeds produced per emerging plant. This parameter was estimated by multiplying per-block means of seed production per reproductive plant and mean seedling survival.
27.2.4
Measurements of Seed Survival and Germination Rates
The fraction of newly produced seeds surviving to and emerging in the next season was determined using sowing experiments. During the time of seed set (in spring of 1994, 1995 and 1996), 32 matched pairs of 10×10 cm quadrats were established in each habitat type (eight per block), and all plants were removed from these prior to seed set to reduce the input of fresh seeds. The quadrats were surrounded with 10-cm-high fences of aluminum to prevent secondary seed dispersal. One quadrat of each pair served as control, whereas a fixed number of fresh seeds, collected from nearby individuals of the focal species, was added to the other. The number of seeds supplemented to each sowing quadrat was 50 for S. glaucus and I. spicata, 40 for R. pictus, and 25 for E. laciniatum. In the following year, the number of seedlings emerging in each quadrat was counted, and the difference in densities between control and sowing quadrat was determined. For all combinations of species, habitat type and year, densities in the sowing quadrats were higher than in the control quadrats, and differences between control and sowing quadrats were highly significant in paired t-tests (p < 0.001). Therefore, the combined value of survival and emergence of seeds was calculated by dividing the differences in densities between sowing and control quadrats by the number of seeds sown. In the third season (1995–1996), seed production was extremely low, and a reasonable number of seeds could be retrieved only for E. laciniatum. Eleven seeds were supplemented to the sowing quadrats in that season. In the following year, only six seedlings emerged from all 64 sowing quadrats (no seedlings were found in the control quadrats) and, therefore, the data from that season were excluded from the analyses.
27.2.5
Statistical Analyses
The demographic data were analyzed using repeated-measures analyses of variance (ANOVAR). The dependent variables in the analyses were the per-block means of the demographic variables (n = 4). Repeated-measure ANOVA models were constructed with per-block means of the demographic variables as separate dependent variables, habitat type and year as within-subject factors, species as between-subject factor, and among-block variation as error term (n = 4, Table 27.1). This design allowed testing for the interactions between the effect of year and the effects of habitat type and species (for detailed explanation of statistical models, see Tielbörger and Kadmon 2000).
27 Effects of Shrubs on Annual Plant Populations
389
In addition to ANOVAR, Tukey’s honestly significant differences were estimated for all pairwise comparisons between species, habitats and years. Densities and seed production data were log (x+1)-transformed, and the data of the sowing experiments were square root-transformed for the ANOVARs and multiple range tests to improve the linearity of the models (Sokal and Rohlf 1995).
27.3 27.3.1
Results Rainfall and Germination
Both absolute amount and distribution of rainfall events differed considerably between the 4 years of study. Accordingly, timing and intensity of germination events varied among the 4 years. The total precipitation of the first and third season of study was far below the long-term average, with a total of 50 mm in the 1993–1994 season and 38 mm in the 1995–1996 season. The rainfall of the second year of study (1994–1995) was the highest ever measured since the establishment of the research site in 1989, with a total of 167 mm (Arid Ecosystems Research Center, Climate Data Base, see Chaps. 5 and 29, this volume). Rainfall in the last season (1996–1997) was below average (74 mm).
27.3.2
Seedling Densities
Seedling densities varied significantly with species, years and their interactions (Table 27.1). In addition, the two-way interactions of habitat type with year and species, and the three-way interaction of all effects were highly significant in the repeated-measures ANOVA. Table 27.1 Results of the repeated-measures ANOVAs (F values) constructed to test for the within-subject effects of year and habitat type (shrubs vs. openings) and the between-subject effect of annual plant species on seedling densities and seed production per plant in the 4 years of study (1993–1994, 1994–1995, 1995–1996, 1996–1997) Source of variation (df) Densities Seed production Year (3) Habitat (1) Species (3) Year x habitat (3) Year x species (9) Habitat x species (3) Year x habitat x species (9) Error (12)
171.33*** 0.03 96.21*** 96.94*** 31.68*** 46.11*** 16.17***
616.33*** 60.17*** 0.63 53.26*** 2.97* 89.84*** 4.42***
390
K. Tielbörger, R. Kadmon
There were large between-year differences in densities of emerging seedlings (Fig. 27.1) which were not consistently related to total annual rainfall. Seedling densities were similarly high in the first (dry) and second (very wet) year, and the severe drought in the third season was accompanied by extremely low seedling densities (Fig. 27.1). Low densities were observed also in the 1996–1997 season, which received slightly below-average rainfall. There were large differences between annual plant species in overall densities of emerging seedlings. I. spicata was the most abundant and E. laciniatum the rarest of all species throughout the 4 years of study (Fig. 27.1). 1993-94
100 80 60 *
40 20
*
0 1994-95
100 80
Density per 625 sqcm
60 40
*
20
*
0 Shrubs Openings
1995-96
100 80 60 40 20 0
0
*
1996-97
100 80 60 40 20
*
*
0 Senecio
Ifloga
Rumex
Erodium
Fig. 27.1 Mean (+SE) densities of emerging plants in the four seasons. Asterisks indicate significant differences between shrubs and openings for a given species in a given season (p < 0.05, Tukey’s honestly significant difference). Zeros indicate zero values
27 Effects of Shrubs on Annual Plant Populations
391
An interaction of the effects of habitat and year could be demonstrated for each species separately, with several examples: In the last season, densities of I. spicata were larger under shrubs whereas, in the other years, they tended to be larger in open areas (Fig. 27.1). R. pictus and S. glaucus were more significantly abundant under shrubs in the first 2 years and the last year, whereas no between-habitat differences were observed during the drought year. In none of the years were there any differences between habitat types in the density of E. laciniatum.
27.3.3
Reproductive Success
The results of the ANOVAR indicate that reproductive success was significantly affected by year and habitat type, as well as by the two-way and three-way interactions of the three main effects (Table 27.1). Annual variation of seed production corresponded well with variation in yearly rainfall. Overall reproduction was by far lowest in the drought year (1995–1996), when two species (S. glaucus and R. pictus) did not survive to seed production (Fig. 27.2). Mean seed production was also relatively low during the first dry season. However, low reproduction was observed more frequently for plants growing beneath shrubs canopies than for individuals in open areas. In the last year, which received slightly below-average rainfall, seed production was much higher than in the first year, and overall seedling reproductive success was by far the highest in the wet second year. Between-year differences in seed production were much larger for the shrub habitat than for the open areas. There were differences between habitat types in seed production, but these differences were dependent on the year of study. In general, reproductive success was either similar in both habitat types or was significantly higher in the open habitat (Fig. 27.2). Between-habitat differences were particularly pronounced during the dry first year of study (1993–1994). In addition to the above patterns, reproductive success differed between the four annual plant species (Fig. 27.2). Except for the wet second year, when seed production was similarly high for all species, E. laciniatum had the highest reproduction.
27.3.4
Probability of Survival and Germination of Newly Produced Seeds
The results of the ANOVAR indicate that percentage emergence of newly produced seeds varied with year, habitat and species, as well as with the interaction between year and habitat type and between year and species (Table 27.2). In particular, the results of the sowing experiments show very large differences in emergence of seeds between the 2 years of study (Fig. 27.3). Emergence was
392
K. Tielbörger, R. Kadmon 1000
1993-94
100 *
10 1 0.1 1000
1994-95 *
Number of seeds per plant
100
*
*
10 1 0.1 1000
1995-96
Shrubs Openings
100 10 1 0.1
0
0
m
0
1000
0
1996-97
100
*
*
10 1 0.1 Senecio
Ifloga
Rumex
Erodium
Fig. 27.2 Mean (+SE) reproductive success in the four seasons. Asterisks indicate significant differences between shrubs and openings for a given species in a given season (p < 0.05, Tukey’s honestly significant difference). Note that the scale is log10. Zeros indicate zero values, m missing values
considerably smaller in the dry 1995–1996 season than in the preceding wet year. However, the effect of the year was also dependent on the identity of the annual species and on habitat type. For example, between-year differences were not significant in the case of E. laciniatum (Fig. 27.3) whereas, for Rumex pictus, differences were the largest of all four species and significant for both habitat types. Yearly variation in emergence of newly produced seeds was larger in the shrub habitat than in the open.
27 Effects of Shrubs on Annual Plant Populations
393
Table 27.2 Results of the repeated-measures ANOVAs (F values) constructed to test for the within-subject effects of year (1994–1995 and 1995–1996) and habitat type (shrubs vs. openings) and the between-subject effect of annual plant species on emergence rates of newly produced seeds in the two study years Source of variation (df) Seedling emergence Year (1) Habitat (1) Species (3) Year x habitat (1) Year x species (3) Habitat x species (3) Year x habitat x species (3) Error (12)
206.03*** 13.21** 9.78** 7.01* 21.82*** 0.30 2.34
1994-95
Shrubs Openings
Emergence of seeds produced in the previous year [%]
60 50 40 30 20 10 0
1995-96 60 50 40 30 20
*
10 0 Senecio
Ifloga
Rumex
Erodium
Fig. 27.3 Mean (+SE) emergence of newly produced seeds in two seasons. Seeds were collected in the spring of 1994 and 1995, and emerged in the winter of 1995–1996 and 1996–1997 respectively. Asterisks indicate significant differences between shrubs and openings for a given species in a given season (p < 0.05, Tukey’s honestly significant difference)
Except for S. glaucus in 1994–1995, emergence was higher in the open areas for both years (Fig. 27.3). However, these between-habitat differences were significant only in one case (S. glaucus, 1995–1996). In the wet 1994–1995 season, there were large differences between the four species in emergence probability. In particular, percentage emergence of R. pictus was
394
K. Tielbörger, R. Kadmon
more than twice as high as that of all other species (Fig. 27.3). In the drought year of 1995–1996, emergence of newly produced seeds was similar for all species in both habitats.
27.4
Discussion
The overall results of this study indicate that desert annuals growing beneath the canopies of perennial plants experience different demographic rates than do conspecific plants growing in open areas between shrubs. This finding is consistent with the hypothesis that local-scale gradients in habitat conditions caused by the presence of shrubs are important in determining the population dynamics of desert annual plants (Went 1942; Muller 1953; Weinstein 1975; Keeley and Johnson 1977; Nelson and Chew 1977; Jaksic and Fuentes 1980; Shmida and Whittaker 1981; Dean and Milton 1991; Silvertown and Wilson 1994; Sarig et al. 1994; Tielbörger and Kadmon 1995; Pugnaire et al. 1996; Tielbörger 1997b; Holzapfel and Mahall 1999). However, the degree and direction in which the shrubs affected the annual plants differed considerably between the 4 years of study. Generally, the effects of shrubs on demographic success (reproduction and emergence) were positive during relatively wet seasons whereas a reverse pattern was detected in years of drought. These findings are important, considering that nearly all previous studies dealing with the effects of shrubs on desert annual plant populations were based on measurements taken within a single year only (Shreve 1931; Went 1942; Halvorson and Patten 1975; Keeley and Johnson 1977; Friedman et al. 1977; Jaksic and Fuentes 1980; Shmida and Whittaker 1981; Gutiérrez and Whitford 1987; Dean and Milton 1991; Rostagno et al. 1991; Gutiérrez et al. 1993; Silvertown and Wilson 1994; Sarig et al. 1994; Tielbörger and Kadmon 1995, 1996; Pugnaire et al. 1996). Some of these studies assumed that annual species in shrub–opening systems can be classified into distinct subgroups based on whether they exhibit a ‘preference’ for either shrubs or open areas, or do not have any preference for either habitat (Went 1942; Silvertown and Wilson 1994). Shmida and Whittaker (1981) concluded, from studies of shrub-structured annual communities in southern California, that “… for annual plants, the shrubs produce a strong and relatively stable pattern of microsite differentiation”. The overall results of the present study contradict the view of these authors, and show that shrubs may facilitate the establishment of understory species during a given year but inhibit or even prevent the establishment of the same species during another year. These findings suggest that conclusions about the dynamics and structure of desert annual populations derived from short-term studies (e.g. Went 1942; Shmida and Whittaker 1981) are highly questionable. There were interesting patterns of annual variations in emergence densities of the annual populations studied. Overall, densities did not correlate very well with total annual precipitation. For example, densities in the extremely wet year (1994–1995) were very similar to those of the preceding drought year (1993–1994), whereas densities in that drought year were much larger than those observed in the last year
27 Effects of Shrubs on Annual Plant Populations
395
(1996–1997), in a season which received roughly twice as much rain. In contrast to results from previous studies which report ‘mass germination’ of desert annuals in exceptionally wet years (e.g. Evenari et al. 1982; Holzapfel 1994), overall emergence densities were highest during the first, relatively dry season, and not in the wet second year. Massive germination of annual desert plants has usually been attributed to the ability of desert annuals to maintain a persistent seed bank from which a large number of seeds are recruited during wet years. In a previous study, Tielbörger (1997b) found that recruitment from the between-year seed bank was, in fact, larger in the wet year. However, the relative contribution of the seed bank to natural densities was relatively low even in the wet year (between 8 and 23%, depending on the species), which may explain why no mass germination of annuals was observed. In the second year of study, densities were positively correlated with habitatspecific seed production in the first year. Also, the drought year of 1995–1996 (with extremely low reproduction) was followed by a year with very low densities. These findings indicate that seedling densities of a given year are largely determined by the amount of seeds produced in the previous year. Seed production is a direct function of annual precipitation. Therefore, densities observed in a given year may be at least partially explained by the annual rainfall of the preceding year, rather than by the rainfall in that year itself. Storage effects of the between-year seed bank may result even in a ‘carry over’ from high-precipitation years to years long beyond that immediately following. Another example shows how densities may nevertheless be a function of present annual rainfall: densities in the drought season of 1995–1996 were extremely low despite the high seed production in the preceding season. In that case, the bottleneck was the low emergence rates of the seeds, which was a function of habitat- and year-specific annual rainfall. Our results of the annual switches in habitat quality are particularly interesting when viewed in terms of previous empirical and theoretical findings dealing with the effect of environmental variation on plant–plant interactions. Such studies suggest that positive interactions among plants predominate under harsh environmental conditions, when one species may benefit from the amelioration of environmental conditions by another species. In contrast, negative interactions predominate under relatively favourable conditions (the stress gradient hypothesis; see Bertness and Callaway 1994; Bertness and Hacker 1994; Bertness and Leonard 1997; Callaway and Walker 1997; Brooker and Callaghan 1998). Based on these studies, we would expect shrubs to have positive effects on the understory annuals during dry years but negative effects during wet years. However, the results obtained in our study were exactly the reverse: positive effects of shrubs on their understory annual plants were observed in wet years but negative effects in dry years. This mismatch between our results and previous theoretical and empirical evidence is illustrated in Fig. 27.4 by depicting the relationship between annual rainfall and the relative neighbour effect ‘RNE’ (sensu Markham and Chanway 1996). The relative neighbour effect (in our case, the relative effect of shrubs on annuals) is estimated as RNE = P–N – P + N / max =(P–N ; P + N )
396
K. Tielbörger, R. Kadmon
where P is the performance of annuals with neighbours (+N, i.e. under shrubs) and without neighbours (−N, i.e. in open areas). RNE is −1 when performance is zero without neighbours (maximum facilitation), zero, when shrubs have no effect, and +1 when there is maximum interference (Markham and Chanway 1996). For a better illustration, we estimated RNE*, which is –RNE and attains positive values when facilitation prevails and negative values when interference is more intense. The stress gradient hypothesis predicts a monotonically decreasing RNE* with increasing rainfall; in our study, a monotonically increasing function was detected for the fitness parameters seed production and emergence rates of all studied species (Fig. 27.4). Tielbörger and Kadmon (2000) presented a conceptual model which may explain these apparently unexpected results. According to this model (Fig. 27.5, relationships a–e), a certain threshold of precipitation is necessary for the successful
1
SEED PRODUCTION
0.5
RELATIVE NEIGHBOR EFFECT (RNE)
0 30
60
90
120
150
−0.5
−1
EMERGENCE FRACTION
1
Senecio Ifloga Rumex
0.5
Erodium MODEL
0 30
60
90
120
150
−0.5
−1 TOTAL ANNUAL RAINFALL [MM]
Fig. 27.4 Relative neighbour effect RNE* (corresponds to –RNE sensu Markham and Chanway 1996) for seed production (top) and emergence probabilities (bottom) versus predictions of the Bertness-Callaway model. RNE* = P+N – P−N / max(P−N;P+N) where P is the performance of annuals under shrubs (+N) and in open areas (−N)
REPRODUCTIVE SUCCESS
27 Effects of Shrubs on Annual Plant Populations
397
Shrubs Openings
e-
d-
a
b
c AMOUNT OF RAINFALL
Fig. 27.5 Conceptual model depicting the qualitative relationships between the amount of rainfall and reproductive success of annual plants under shrub canopies and in open areas (after Tielbörger and Kadmon 2000). The model predicts negative effects of shrubs on plant success in relatively dry years (rainfall
c). a Threshold value of rainfall allowing reproduction in the open habitat. b Threshold value of rainfall allowing reproduction in the shrub habitat (higher than a due to rainfall interception by shrub canopies). c Critical rainfall threshold above which positive effects of shrubs on plant reproduction predominate over negative effects; below c, water is the main limiting resource under shrubs and, above c, nutrients are limiting in open areas. d Reproduction in the open habitat when water is not limiting. e Reproduction under shrubs when water is not limiting
reproduction of annual plants in open areas (a). This threshold is higher beneath shrubs (b), due to rainfall interception by the shrub canopies. Above the threshold, reproductive success increases monotonically with the amount of rainfall until a saturation point is reached, which is due to a second limiting factor. The upper limit of reproductive success of plants in open areas (d) is lower than that of conspecifics beneath shrubs (e), due to nutrient limitation which is more severe in open sandy areas. This simple model is also consistent with the results obtained for emergence rates: water limits emergence under shrubs during dry years whereas, in wet years, there is no negative effect of shrubs on emergence rates. An explicit validation of the model proposed by Tielbörger and Kadmon (2000) requires experimental manipulation of environmental severity (both water and nutrients). However, the model does find some empirical support by measurements done at the Nizzana research site in 1999 and 2000. Soil water content was found to be significantly lower under shrubs, irrespective of the amount and intensity of a particular rainfall event (Prasse and Tielbörger, unpublished data). On the other hand, soil nutrient content was higher under shrubs (Felix-Henningsen, Prasse and Tielbörger, unpublished data), indicating that the interactions between shrubs and annuals in the study system may indeed operate through the mechanisms suggested by the model of Tielbörger and Kadmon (2000).
398
27.5
K. Tielbörger, R. Kadmon
Conclusions
Much of our results may be explained by the negative effect of shrubs on soil water content. Interestingly, most previous studies have stressed positive effects of shrubs on soil water content, mostly because of reduced evaporation loss in the shade of the shrub canopy (Shreve 1931; Turner et al. 1966; Nobel 1980; Tromble 1988; Franco and Nobel 1989; Valiente-Banuet and Ezcurra 1991 but see Pressland 1976; Martinez-Meza and Whitford 1996). However, almost all studies dealing with nurse-plant effects have been conducted on non-sandy soils where evaporation is important in decreasing soil water content. In sands, however, evaporation loss is relatively low, and the negative effects of shrubs on soil water may dominate over the potentially positive effects. We are aware of only two previous studies which were conducted in a sandy environment: Casper (1996) has investigated the effect of a drought year on interactions between shrubs and their understory, and Holzapfel and Mahall (personal communication) have conducted a 3-year study of annuals in a shrubland of the Mojave Desert. Interestingly, in both studies the pattern of annual variation in interactions was consistent with those obtained in our study: facilitation of understory annuals by shrubs was more pronounced during wet years, whereas negative or neutral interactions were detected only in relatively dry years. We therefore suggest that our results may be representative for sandy desert environments where interference of annuals by shrubs predominates when water is scarce. Acknowledgements We would like to thank R. Prasse, A. Höhn, and A. Stratmann for their help in the field. The extraordinary support of I. Künne and H. Künne is gratefully acknowledged. The research was funded by the MINERVA foundation, the Lady Davis Fellowship Trust, the Israel Academy of Sciences and Humanities, and the German Ministry for Education and Research (BMBF). Fieldwork was done at the Nizzana research site of the Arid Ecosystems Research Center of the Hebrew University of Jerusalem and the MINERVA foundation.
References Beatley JC (1967) Biomass of desert winter annual plant populations in southern Nevada. Oikos 20:261–273 Beatley JC (1969) Survival of winter annuals in the northern Mojave desert. Ecology 48:745–750 Bertness MD, Callaway RM (1994) Positive interactions in communities. Trends Ecol Evol 9:191–193 Bertness MD, Hacker SD (1994) Physical stress and positive interactions among marsh plants. Am Naturalist 144:363–372 Bertness MD, Leonard GH (1997) The role of positive interactions in communities: lessons from intertidal habitats. Ecology 78:1976–1989 Bowers MA (1987) Precipitation and the relative abundance of desert winter annuals: a 6-year study in the northern Mohave Desert. J Arid Environ 12:141–149 Brooker RW, Callaghan TV (1998) The balance between positive and negative plant interactions and its relationship to environmental gradients: a model. Oikos 81:196–207
27 Effects of Shrubs on Annual Plant Populations
399
Callaway RM, Walker LR (1997) Competition and facilitation: a synthetic approach to interactions in plant communities. Ecology 78:1958–1965 Casper BB (1996) Demographic consequences of drought in the herbaceous perennial Cryptantha flava: effects of density, associations with shrubs, and plant size. Oecologia 106:144–152 Charley JL, West NE (1975) Plant-induced soil chemical patterns in some shrub-dominated semidesert ecosystems of Utah. J Ecol 63:945–963 Dean WR, Milton JS (1991) Patch disturbance in arid grassy dunes: antelopes, rodents and annual plants. J Arid Environ 20:231–237 Evenari M, Shanan L, Tadmor N (1982) The Negev, the challenge of a desert, 2nd edn. Harvard University Press, Cambridge Fowler NL (1988) The effect of environmental heterogeneity in space and time on the regulation of populations and communities. In: Davy AJ, Hutchings MJ, Watkinson AR (eds) Plant population ecology. Blackwell, Cambridge, pp 249–269 Franco AL, Nobel PS (1989) Effect of nurse plants on the microhabitat and growth of cacti. J Ecol 76:870–886 Friedman J, Orshan G, Ziger-Cfir Y (1977) Suppression of annuals by Artemisia herba-alba in the Negev desert of Israel. J Ecol 65:413–426 Garcia-Moya E, McKell CM (1970) Contribution of shrubs to the nitrogen economy of a desertwash plant community. Ecology 51:81–88 Gutiérrez JR, Whitford WG (1987) Chihuahuan desert annuals: importance of water and nitrogen. Ecology 68:2032–2045 Gutiérrez JR, Meserve PL, Contreras LC, Vàsquez H, Jaksic FM (1993) Spatial distribution of soil nutrients and ephemeral plants underneath and outside Porlieria chilensis (Zygophyllaceae) in arid coastal Chile. Oecologia 95:347–352 Halvorson WL, Patten DT (1975) Productivity and flowering of winter ephemerals in relation to Sonoran desert shrubs. Am Midl Naturalist 93:311–319 Holzapfel C (1994) Einfluß menschlicher Störungen auf die Vegetation entlang eines ökologischen Gradienten von der mediterranen Zone bis zur Wüste. Cuvillier, Göttingen Holzapfel C, Mahall BE (1999) Bidirectional facilitation and interference between shrubs and associated annuals in the Mojave Desert. Ecology 80:1747–1761 Inouye RS (1991) Population biology of desert annual plants. In: Polis GA (ed) The ecology of desert communities. University of Arizona Press, Tucson, AZ, pp 27–54 Jackson RB, Caldwell MM (1993) Geostatistical pattern of soil heterogeneity around individual perennial plants. J Ecol 81:683–692 Jaksic FM, Fuentes ER (1980) Why are native herbs in the Chilean Matorral more abundant beneath bushes: microclimate or grazing? J Ecol 68:665–669 Juhren M, Went FW, Phillips E (1956) Ecology of desert plants. IV. Combined field and laboratory work on germination of annuals in the Joshua Tree National Monument, California. Ecology 37:318–330 Kadmon R (1993) Population dynamic consequences of habitat heterogeneity: an experimental study. Ecology 74:816–825 Keeley SC, Johnson AW (1977) A comparison of the pattern of herb and shrub growth in comparable sites in Chile and California. Am Midl Naturalist 97:120–132 Loria M, Noy-Meir I (1979/1980) Dynamics of some annual populations in a desert loess plain. Israel J Bot 28:211–225 Mack RN, Pyke DA (1983) The demography of Bromus tectorum: variation in time and space. J Ecol 71:69–93 Markham JH, Chanway CP (1996): Measuring plant neighbor effects. Funct Ecol 10:548–549 Martinez-Meza E, Whitford WG (1996) Stemflow, throughfall and channelization of stemflow by roots of three Chihuahuan desert shrubs. J Arid Environ 32:271–287 Muller CH (1953) The association of desert annuals with shrubs. Am J Bot 40:53–60 Muller WH, Muller CH (1956) Association patterns involving desert plants that contain toxic products. Am J Bot 43:354–361
400
K. Tielbörger, R. Kadmon
Nelson JF, Chew RM (1977) Factors affecting seed reserves in the soil of a Mojave Desert ecosystem, Rock Valley, Nye County, Nevada. Am Midl Naturalist 97:300–320 Niering JF, Whittaker RH, Lowe CH (1963) The saguaro: a population in relation to environment. Science 142:15–23 Nobel PS (1980) Morphology, nurse-plants and minimum apical temperatures for young Carnegia gigantea. Bot Gaz 141:188–191 Pressland AJ (1976) Soil moisture redistribution as affected by throughfall and stemflow in an arid shrub community. Austr J Bot 24:641–649 Pugnaire FI, Haase P, Puigdefábregas J, Cueto M, Clark SC, Incoll LD (1996) Facilitation and succession under the canopy of a leguminous shrub, Retama sphaerocarpa, in a semi-arid environment in south-east Spain. Oikos 76:455–464 Rostagno CM, del Valle HF, Videla L (1991) The influence of shrubs on some chemical and physical properties of an aridic soil in north-eastern Patagonia, Argentina. J Arid Environ 20:179–188 Sarig S, Barness G, Steinberger Y (1994) Annual plant growth and soil characteristics under desert halophyte canopy. Acta Oecol 15:521–527 Schlesinger WH, Raikes JA, Hartley AE, Cross AF (1996) On the spatial pattern of soil nutrients in desert ecosystems. Ecology 77:364–374 Shmida A, Whittaker RH (1981) Pattern and biological microsite effects in two shrub communities, Southern California. Ecology 62:234–251 Shreve F (1931) Physical conditions in sun and shade. Ecology 12:96–104 Silvertown JW, Wilson JB (1994) Community structure in a desert perennial community. Ecology 75:409–417 Sokal RR, Rohlf PJ (1995) Biometry, 3rd edn. W.H. Freeman, New York Tielbörger K (1997a) The vegetation of linear desert dunes in the north-western Negev, Israel. Flora 192:261–278 Tielbörger K (1997b) Effect of shrubs on population dynamics of annual plants in a sandy desert ecosystem. PhD Thesis, Ludwig-Maximilians Universität, München Tielbörger K, Kadmon R (1995) The effect of shrubs on the emergence, survival and fecundity of four coexisting annual species in a sandy desert ecosystem. Ecoscience 2:141–147 Tielbörger K, Kadmon R (1996) The effect of shrubs on community structure of annual plants in a sandy desert ecosystem. Verhandl Gesell Ökol 25:59–64 Tielbörger K, Kadmon R (1997) Relationships between shrubs and annual communities in a sandy desert ecosystem: a three-year study. Plant Ecol:130:191–201 Tielbörger K, Kadmon R (2000) Temporal environmental variation tips the balance between facilitation and interference in desert plants. Ecology 81:1544–1553 Tromble JM (1988) Water interception by two arid land shrubs. J Arid Environ 15:65–70 Turner RM, Alcorn SM, Olin G, Booth JA (1966): The influence of shade, soil, and water on saguaro seedling establishment. Bot Gaz 127:95–102 Valiente-Banuet A, Ezcurra E (1991) Shade as a cause of the association between the cactus Neobuxbaumia tetezo and the nurse plant Mimosa luisana in the Tehuacàn Valley, Mexico. J Ecol 79:961–971 Weinstein N (1975) The effect of a desert shrub on its micro-environment and on herbaceous plants. MSc Thesis, Department of Botany, Hebrew University of Jerusalem Weiss SB, Murphy DD, White RR (1988) Sun, slope, and butterflies: topographic determinants of habitat quality for Euphydrias editha. Ecology 69:1486–1496 Went FW (1942) The dependence of certain annual plants on shrubs in Southern California deserts. Bull Torrey Bot Club 69:100–114 Wiens JA (1976) Population responses to patchy environments. Annu Rev Ecol Systematics 7:81–120
Chapter 28
Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition R. Kadmon
28.1
Introduction
There is abundant evidence that local heterogeneity in habitat conditions may influence the demography of plant populations (Fowler and Antonovics 1981; Mack and Pyke 1983; van Tienderen 1992; Tielbörger and Kadmon 1995). It is also well evident that competition for resources may be important in determining the survival, growth and reproduction of individual plants (Weiner 1988; Keddy 1989a; Pantastico-Caldas and Venable 1993; Tremmel and Bazzaz 1993; Kadmon 1995; Grace and Platt 1995). Much less is known about the degree to which, and the manner by which, competition interacts with local variation in habitat conditions in determining the demography of plant populations. Interactions between the effects of habitat conditions and competition are expected to be common in natural plant communities because factors affecting the intensity of competition, such as population density (Keddy 1982; Condit et al. 1994), standing crop (Wilson and Keddy 1986a; Reader et al. 1994), productivity (Turkington et al. 1993; Kadmon 1995), abiotic stress (Keddy 1981; Stadt et al. 1994) and herbivory (Swank and Oechel 1991; Burger and Louda 1994; Shabel and Peart 1994), may vary considerably between one habitat and another, even at very small spatial scales. Under such circumstances, any attempt to interpret observed patterns of demographic variation must take into account the interactive effects of spatial heterogeneity in habitat conditions and competition. Moreover, since plant species differ from each other in their tolerance to both abiotic factors (Chapin et al. 1993; Aerts and de Caluwe 1994) and competition (Goldberg and Fleetwood 1987; Keddy 1989b; Silvertown and Dale 1991), coexisting species may respond differentially to the same changes in habitat conditions and competition, depending on their life-history traits. Such species-specific responses may be important in determining opportunities for the coexistence of competing species, as well as community-level patterns of species composition and diversity. Ecological theory allows making some predictions on expected interactions between habitat heterogeneity, competition, and species life-history traits. For example, it has long been claimed that highly stable and relatively favourable habitats are characterized by more intense competitive interactions than are unstable
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
401
402
R. Kadmon
habitats or habitats characterized by strong abiotic stress (Gause 1934; MacArthur and Wilson 1967; Grime 1979). This idea has led to the prediction that species inhabiting stable habitats should have stronger competitive ability than do species inhabiting unstable or abiotically stressed ones. Both hypotheses have been supported by empirical evidence (Connell 1972; Abrahmson 1975; Solbrig and Simpson 1977; Lubchenco 1980; Keddy 1989b). It has also been suggested that community patterns of competitive effects and responses can be predicted from information on plant size in the absence of competition, with relatively large species showing greater competitive effects but smaller competitive responses than do relatively small ones (Goldberg and Fleetwood 1987; Gaudet and Keddy 1988). Assuming a trade-off between the life-history characteristics required for coping with competition and abiotic stress (Southwood et al. 1974; Grime 1977; Grace 1990), it can further be expected that species dominating stable and relatively favourable habitats should be less sensitive to competitive effects but more sensitive to habitat instability or abiotic stress than is the case for species occurring in unstable habitats and abiotically stressed ones. In desert dune ecosystems, sand instability is a major factor affecting the establishment, growth and reproduction of plants. Accordingly, patterns of plant species distribution on desert dunes have usually been related to variation in the stability of the underlying sand (e.g. Yeaton 1988; Abbas et al. 1991). Yet, most studies of desert dune vegetation have focused on perennial plants, and the degree to which sand instability affects the distribution of annual plants has largely remained untested (though see Kadmon and Leschner 1995). Even less is known about the role of competitive interactions in determining the performance of desert dune annuals (though see Tielbörger and Kadmon 2000). Considering this lack in knowledge, a field experiment was designed to investigate whether and how spatial variation in sand stability interacts with competitive effects in determining the demography of desert dune annuals. The work was carried out in the Nizzana sandfield. Two coexisting species of annual plants, Neurada procumbens L. and Bromus fasciculatus C. Presl., were chosen for the study. These species differ from each other in their local-scale patterns of distribution (Kadmon and Leschner 1995): N. procumbens is limited to relatively stabilized habitats, mainly interdune corridors, whereas B. fasciculatus occurs in both stabilized and unstabilized habitats (e.g. dune slopes). Considering these observed, local-scale patterns of distribution, it was expected that N. procumbens would be more sensitive to local variation in the stability of sand than was the case for B. fasciculatus. Also, based on ecological theory (MacArthur and Wilson 1967; Grime 1979) and previous measurements indicating that dune slopes are much less stable than interdune corridors (Kadmon and Leschner 1995), it was expected that competition for resources would be more intense in the interdune areas than on the slopes. Finally, based on previous evidence indicating that individuals of N. procumbens attain larger size than do individuals of B. fasciculatus when growing in the absence of competition (Kadmon 1994), it was expected that N. procumbens would show weaker demographic responses to neighbour
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
403
competition than was the case for B. fasciculatus. In this paper, I present and discuss the results of experiments designed to test these predictions.
28.2
Methods
An area of about 200 × 200 m representing a typical dune gradient was selected for the study. Three types of habitats, differing from each other in the stability of the sand, have previously been distinguished along this gradient (Kadmon and Leschner 1995): an active dune crest, where sand movement is most intense, a slope area characterized by intermediate sand stability, and a highly stable interdune area. The vegetation cover was less than 10% on the crest of the dune and increased to 20–50% in the interdune corridor. The dune crest was dominated by the perennial grass Stipagrostis scoparia and the slope was dominated by Moltkiopsis ciliata, Heliotropium digynum and Convolvolus lanatus. The stable interdune corridor was characterized by a combination of several woody species including Retama raetam, Cornulaca monacantha, Echiochilon fruticosum and Anabasis articulata. Annual plants were limited to the two lower parts of the gradient (the slope and the interdune corridor). Direct measurements of sand stability (Kadmon and Leschner 1995) have shown that weekly changes in surface height are of the order of 1–10 cm on the slope but only 0.1 cm or less in the interdune corridor. Differences between the two habitats in water availability, organic matter content and sand salinity were found to be extremely small (Kadmon and Leschner 1995), suggesting that spatial heterogeneity in the stability of the sand would be the most important factor structuring annual communities along the gradient. This conclusion was supported by multivariate analyses of relationships between community composition and environmental factors (Kadmon and Leschner 1995).
28.2.1
Experimental Design
Two types of experiments were carried out: a neighbour removal experiment and a sand cover experiment. The former experiment was designed to test the combined effects of competition and sand instability (as expressed by topographic position) on the demography of the two study species. The latter experiment was designed to directly test for differences between the two species in their tolerance to sand instability.
28.2.1.1
The Neighbour Removal Experiment
Neighbour manipulations were conducted in two different habitats along the dune gradient: one set of experiments was established on the slope of the dune and the other set was established in the interdune area. The performance of plants growing
404
R. Kadmon
under natural density conditions was determined by monitoring the survival, growth and reproductive success of 36 randomly chosen seedlings of each species in each habitat type (see below). The potential performance of plants growing in the absence of competitive effects was determined using transplanting experiments. Prior to the first rainfall, fruits of both species collected from dense populations were transferred into randomly located quadrats in each type of habitat. Each quadrat was 25 cm in diameter, and 36 quadrats were used for each species in each habitat type. Following the main germination event, a single, randomly chosen seedling of the target species was marked in each quadrat and all of its neighbours (including naturally occurring seedlings) were removed from the quadrat. Preliminary observations of root distribution and shoot sizes had shown that a radius of 25 cm is sufficient to prevent any root or shoot interactions with neighbouring plants. Thus, a total of 288 individual plants organized in a full three-way factorial design (two habitats x two species x two neighbour treatments with 36 replicates) were included in this experiment. Plants surviving to reproduction were harvested during the period of fruit set for measurements of biomass (as dry weight) and fruit production (number and dry biomass of fruits). Based on these measurements, seven variables were calculated for each of the eight habitat/species/treatment combinations: the percentage of seedlings which survived to reproduction, the above-ground biomass of surviving plants, the percentage of above-ground biomass allocated to fruit production, the fruit biomass, the number of fruits per surviving plant, the number of fruits per germinating plant, and the average fruit weight.
28.2.1.2
The Sand Cover Experiment
This experiment was conducted in the interdune corridor, its main purpose being to directly test for differences between the two species in their tolerance to sand instability. Sand movement was simulated by artificially covering seedlings emerging in experimental quadrats with sand to a depth of 1 cm. Previous measurements (Kadmon and Leschner 1995) had indicated that such local events of cover by sand are common in the unstabilized parts of the dunes. The experimental design was based on a system of ten blocks of paired quadrats 25 cm in diameter, located in patches of high density of the study species. One randomly chosen quadrat in each block was covered by sand to a depth of 1 cm about 3 weeks after the main germination event, the other quadrat serving as experimental control. All of the quadrats were thinned into constant densities of the study species (ten individuals of each species, a total of 20 individuals per quadrat) to control for within- and between-treatment variation in density conditions. Individuals surviving in these quadrats were harvested during fruit set and were measured for the same variables described for the neighbour removal experiment. Based on the observed differences between the two species in their patterns of distribution along gradients of sand stability, it was expected that N. procumbens would show stronger negative demographic responses to cover by sand than did B. fasciculatus. The control quadrats of the sand cover experiment were also used to evaluate the performance of the two species under controlled high-density conditions. Thus, while the neighbour removal experiment enabled me to evaluate the average effect
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
405
of neighbour competition on the performance of plants growing under natural density conditions, the comparison of experimentally isolated plants with plants growing under controlled neighbour conditions allowed me to test for differences in competitive responses between the two species.
28.2.2
Data Analysis
Differences in seedling survival between habitat types, neighbour removal treatments and species were tested using Chi-square tests. Corresponding differences in above-ground biomass, reproductive allocation, fruit biomass, fruit number per seedling, fruit number per surviving plant and fruit weight were tested using three-way ANOVAs with habitat, removal treatment and species as fixed effects. Biomass data were log-transformed prior to the analysis, to reduce heteroscedasticity (Sokal and Rohlf 1981). In all models, sums of squares were decomposed with each effect being adjusted for all other effects. A similar approach was used to analyze the results of the sand cover experiment. No mortality was observed following thinning of the quadrats and, therefore, the analysis of this experiment was restricted to five variables: total above-ground biomass per plant, reproductive allocation, fruit biomass, fruit number per surviving plant and average fruit weight. Each of these variables was analyzed separately using a three-way ANOVA with the mean value per quadrat as the dependent variable (to avoid pseudo-replication), treatment and species as fixed effects and block as a random effect (see Sokal and Rohlf 1981, p. 312 for details about the computation of F values in such a mixed, non-replicated three-way design). In interpreting the results of the statistical analyses, it should be taken into account that some of the response variables were not independent. For example, reproductive biomass represents the multiplication of above-ground biomass and reproductive allocation. Similarly, the number of fruits produced per germinating plant represents the survival fraction of seedlings multiplied by the number of fruits produced per surviving plant. Yet, because each of these variables has a different meaning and different implications (e.g. fruit production per germinating plant is the most important variable from the population dynamics point of view, but the interpretation of its patterns of variation can be improved by distinguishing between the survival and reproduction components), both ‘basic’ variables and ‘combined’ variables were subjected to the statistical analysis.
28.3 28.3.1
Results Seedling Survival
Seedling survival ranged from 18 to 100%, depending on habitat conditions, neighbour treatment and species (Fig. 28.1). In the interdune corridor, removal of neighbouring plants significantly increased the survival of seedlings of both species (Fig. 28.1,
406
R. Kadmon
Seedling survival(%)
Neurada procumbens 100 80
Interdune corridor Dune slope
60 40 20 0
With neighbours
Seedling survival (%)
120
Neighbours removed
Bromus fasciculatus
100 80 60 40 20 0
With neighbours
Neighbours removed
Fig. 28.1 Effects of habitat type and neighbour removal on seedling survival of Neurada procumbens and Bromus fasciculatus
Table 28.1, part a). On the slope of the dune, the effect of neighbour removal was not statistically significant for both species (Table 28.1, part a). Experimentally isolated seedlings of N. procumbens had significantly higher survival in the interdune corridor than on the slope of the dune but a reverse pattern, with significantly higher survival on the slope, was obtained for conspecific seedlings growing under natural density conditions (Fig. 28.1, Table 28.1, part b). Seedlings of B. fasciculatus exhibited higher survival in the interdune corridor when their neighbours were removed but there was no difference between the two habitats when growing under natural density conditions (Fig. 28.1, Table 28.1, part b). Differences in seedling survival between the two species were statistically significant or nearly significant in all cases, with B. fasciculatus showing higher survival than did N. procumbens (Fig. 28.1, Table 28.1, part c).
28.3.2
Above-Ground Biomass
Neighbour removal had a significant positive effect on above-ground biomass but the magnitude of this effect varied between the two species as well as between the two habitats (Fig. 28.2). Individuals growing in the interdune corridor exhibited
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
407
Table 28.1 Results of Chi-square tests for differences in seedling survival between neighbour removal treatments (a), habitat types (b) and species (c). Sample size equals 36 seedlings in each combination of habitat type, species and neighbour treatment Test Habitat or treatment Habitat or control Habitat or treatment Habitat or control a. Neighbour removal, within species, within habitat types Neurada procumbens Bromus fasciculatus Dune slope Interdune corridor Dune slope
Interdune corridor
c2 P
12.98 0.001
0.23 0.629
40.53 0.001
5.44 0.196
b. Habitat type, within species, within treatments Neurada procumbens Neighbour Control removal 21.39 6.92 c2 P 0.001 0.009
Bromus fasciculatus Neighbour removal 18.95 0.001
c. Species, within habitat types, within treatments Dune slope Neighbour Control removal 3.56 13.33 c2 P 0.059 0.001
Interdune corridor Neighbour removal 4.23 0.039
Neurada procumbens Above-ground biomass (g)
Above-ground biomass (g)
0.1 0.01
Reproductive allocation (%)
40 30 20 10
10
22.86 0.001
Interdune corridor Dune slope
1 0.1 0.01
40 30 20 10
0
0
10
10
Reproductive biomass (g)
Reproductive allocation (%)
Control
50
50
Reproductive biomass (g)
1.92 0.165
Bromus fasciculatus
10 1
Control
1 0.1 0.01 0.001
1 0.1 0.01 0.001
With neighbours
Neighbours removed
With neighbours
Neighbours removed
Fig. 28.2 Effects of habitat type and neighbour removal on above-ground biomass, reproductive allocation and reproductive biomass of individual plants of Neurada procumbens and Bromus fasciculatus
408
R. Kadmon
stronger positive responses to removal of their neighbours than did conspecific individuals growing on the slope of the dune and, independently from habitat conditions, individuals of N. procumbens were more sensitive to neighbour removal than were individuals of B. fasciculatus (Fig. 28.2). The effect of habitat type on above-ground biomass of both species was influenced by neighbour conditions. Experimentally isolated plants had higher biomass in the interdune corridor, whereas plants growing under natural density conditions showed a higher biomass on the slope of the dune (Fig. 28.2; also see below and Fig. 28.3). As can be expected from these results, all of the two-way interaction terms (removal x habitat, removal x species and habitat x species) were statistically significant (Table 28.2). Cover by sand significantly reduced the above-ground biomass of N. procumbens (t=2.52, P=0.021) but had no effect (P > 0.1) on that of B. fasciculatus (Fig. 28.4). The interaction between the two main effects (species identity and cover by sand) was statistically significant (Table 28.3), indicating that the two species responded differentially to cover by sand, with seedlings of N. procumbens being more sensitive to sand cover than was the case for seedlings of B. fasciculatus.
Neurada procumbens
Bromus fasciculatus
Interdune corridor Dune slope
20 15 10 5
30
No. of fruits per surviving plant
25
20 15 10 5 0
30
30
25 20 15 10 5 0
100 10 1 0.1
25 20 15 10 5 0
1000
Average fruit weight (mg)
Average fruit weight (mg)
25
0
No. of fruits per seedling
No. of fruits per seedling
No. of fruits per surviving plant
30
With neighbours
Neighbours removed
1000 100 10 1 0.1
With Neighbours neighbours removed
Fig. 28.3 Effects of habitat type and neighbour removal on fruit number per surviving plant, fruit number per germinating plant and average fruit weight in Neurada procumbens and Bromus fasciculatus
Effect (df) Neighbour removal (1) Habitat (1) Species (1) Removal x habitat (1) Removal x species (1) Habitat x species (1) Removal x habitat x species (1) Error (169) Total (176)
MS 9.57 1.24 45.76 1.92 0.77 1.11 0.34 0.20 0.62
F
MS ***
48.0 6.2* 229.6*** 9.6** 3.9* 5.6* 1.7
5.20 0.29 7.61 0.18 6.70 0.06 0.19 0.03 0.15
F
MS ***
159.4 8.99** 232.9*** 5.4* 205.1*** 1.79 5.79*
28.88 2.74 16.06 3.26 12.03 1.68 1.04 0.29 0.79
F
MS ***
98.5 9.3** 54.8*** 11.1*** 41.0*** 5.74* 3.5
5,097 73 9,010 1,938 433 72 108 171 254
F
MS ***
26.7 0.42 52.5*** 11.3*** 2.5 0.4 0.6
4908 1373 10645 5347 779 2 1085 126 207
F ***
39.1 10.9*** 84.7*** 42.6*** 6.1* < 0.1 8.6**
MS
F
0.035 0.0111 0.191 0.002 0.065 0.006 0.004
50.6*** 15.5*** 279.1*** 2.4 95.2*** 8.57** 5.75*
0.001 0.003
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
Table 28.2 Analysis of variance for the effects of neighbour removal, habitat type (interdune corridor vs. dune slope) and species (Neurada procumbens vs. Bromus fasciculatus) on above-ground biomass, reproductive allocation, reproductive biomass, fruit number per surviving plant, fruit number per seedling and fruit weight (* P<0.05, ** P<0.01, *** P<0.001) Above-ground Reproductive Reproductive No. of fruits per No. of fruits per biomass allocation biomass surviving plant germinating plant Fruit weight
409
410
R. Kadmon
Above-ground biomass (g)
10 Control Sand cover
1 0.1 0.01
Reproductive allocation (%)
50 40 30 20 10
Average fruit weight (mg)
Number of fruits per surviving plant
Reproductive biomass (g)
0 1
0.1
0.01
20 15 10 5 0
1000 100 10 1
0.1
Neurada
Bromus
Fig. 28.4 Demographic responses of Neurada procumbens and Bromus fasciculatus to seedling cover by sand
28.3.3
Reproductive Allocation
In general, individuals of B. fasciculatus allocated a higher proportion of their above-ground biomass to fruit production than did individuals of N. procumbens (Fig. 28.2). However, the magnitude of this difference was strongly influenced by
Effect (df)
MS
F
MS
F
MS
F
MS
F
MS
F
Species (1) Cover by sand (1) Species x cover (1) Error (9) Total (39)
17.02 0.26 0.50 0.06 0.52
237*** 3.3 8.7*
< 0.01 < 0.01 < 0.01 < 0.01 < 0.01
1.2 1.3 3.3
16.50 0.36 0.648 0.06 0.52
173*** 3.9 11.6**
7.23 0.27 0.46 0.04 0.26
129*** 3.7 12.4**
0.12 < 0.01 < 0.01 < 0.01 < 0.01
122*** 1.8 1.8
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
Table 28.3 Analysis of variance for differences between treatments (cover by sand vs. control) and species (Neurada procumbens vs. Bromus fasciculatus) in above-ground biomass, reproductive allocation, reproductive biomass, fruit number per surviving plant and fruit weight ( *P < 0.05, **P < 0.01, *** P < 0.001) Above-ground Reproductive Reproductive No. of fruits per biomass allocation biomass surviving plant Fruit weight
411
412
R. Kadmon
neighbour conditions (Fig. 28.2, Table 28.2). Moreover, the degree to which neighbour conditions were important in determining the differences in reproductive allocation between the two species was influenced by habitat conditions, as indicated by the significant three-way interaction term (Table 28.2). Neighbour removal caused an increase in the proportion of biomass allocated to reproduction in both species (Fig. 28.2). Experimentally isolated individuals growing in the interdune corridor allocated a higher proportion of their above-ground biomass to reproduction than did isolated individuals growing on the slope of the dune. Individuals growing under natural density conditions exhibited a reverse pattern, with plants growing on the slope allocating a higher amount of their biomass to fruit production than did plants growing in the interdune corridor (Fig. 28.2). This qualitative difference between experimentally isolated and control plants was found for both species and was statistically significant, as indicated by the significant habitat x removal treatment interaction (Table 28.2). The results of the sand cover experiment indicated that the percentage of aboveground biomass allocated to fruit production was relatively uniform (36–40%) across both treatments and species (Fig. 28.4). Neither the two main effects, nor their interaction were statistically significant (Table 28.3).
28.3.4
Reproductive Biomass
Patterns of differences in reproductive biomass between species, habitats and treatments were relatively similar to those obtained for total above-ground biomass (Fig. 28.2). The main difference was in the magnitude of the interaction between the effects of neighbour removal and species. This interaction was relatively weak (although statistically significant) in the case of the total above-ground biomass, but much stronger and more significant in the case of the reproductive biomass (F ratio = 3.9 vs. 41.0 respectively, Table 28.2). This difference reflects the fact that neighbour removal had a significant positive effect on the proportion of biomass allocated to reproduction, and this effect was much stronger in the case of N. procumbens (Fig. 28.2). Cover by sand had a significant negative effect on reproductive biomass of N. procumbens (t = 2.53, P < 0.021) but no effect on reproductive biomass of B. fasciculatus (Fig. 28.4). The interaction between the effects of sand cover and species was statistically significant (Table 28.3).
28.3.5
Fecundity
The qualitative patterns obtained for the two measures of fecundity (fruit number per surviving plant and fruit number per germinating plant) were similar: (1) B. fasciculatus produced more fruits than did N. procumbens in both habitats and under
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
413
both neighbour treatments, (2) neighbour removal increased fruit production under most habitat and neighbour conditions, and (3) differences in fruit production between habitats were influenced by neighbour conditions, experimentally isolated plants producing more fruits in the interdune corridor, whereas plants growing under natural neighbour conditions produced more fruits on the slope of the dune (Fig. 28.3). The main difference between the patterns obtained for germinating and surviving plants was in the strength and statistical significance of the various effects; all of the main effects and interactions were stronger and more significant when fruit production was calculated on a per-seedling basis (Table 28.2). It can also be seen that the effects of habitat type, the two-way interaction between neighbour removal and species, and the three-way interaction term were not statistically significant in the analysis of the surviving plants but became significant when fruit production was analyzed on a per-seedling basis (Table 28.2). These differences resulted from the fact that seedling survival and fruit production per surviving plant exhibited similar patterns of variation between species, habitats and neighbour treatments (Figs. 28.2 and 28.3). The effect of cover by sand on fruit number was statistically significant in the case of N. procumbens (t=2.75, P<0.013) but not significant in the case of B. fasciculatus (Fig. 28.4). As with most other variables, the interaction between the effects of cover by sand and species was statistically significant (Table 28.3).
28.3.6
Fruit Weight
N. procumbens showed a positive effect of neighbour removal on fruit weight but the magnitude of this effect was influenced by habitat type, with plants growing in the interdune corridor showing larger differences between the two neighbour treatments than did plants growing on the slope (Fig. 28.3). Fruit weight of B. fasciculatus was insensitive to differences in habitat and neighbour conditions (Fig. 28.3). As can be expected from such results, the three-way interaction between neighbour removal, habitat type and species was statistically significant (Table 28.2). The results of the sand cover experiment indicate that the fruit weight of plants subject to cover by sand was not significantly different from that of control plants in both species (Fig. 28.4, Table 28.3).
28.4
Discussion
Patterns of spatial variation in the vegetation of desert dunes have usually been related to changes in abiotic factors and in particular, to local-scale gradients in the mobility of the underlying sand (Grenot 1974; Seely and Louw 1980; Bowers 1982; Walter and Box 1983; Moreno-Casasola and Espejel 1986; Yeaton 1988; Danin 1991; Seely 1991; Abbas et al. 1991). However, in most cases this conclusion was based on observational studies showing that species composition or the relative
414
R. Kadmon
abundance of different plant species vary in a systematic manner along dune gradients. Very few studies have actually measured variations in the stability of desert dunes at scales relevant to the demography of plant populations (Moreno-Casasola 1986; Kadmon and Leschner 1995), and even fewer studies have measured demographic responses of plant populations to changes in habitat conditions along desert dune gradients (Inouye 1991; Tielbörger 1997; Prasse 1999). Moreover, even if correlations between demographic parameters and sand stability are established, the interpretation of such results is equivocal because local changes in the stability of the sand may be correlated with changes in other factors influencing the performance of plants, e.g. water availability or nutrient concentration (Bowers 1982). Biotic interactions such as seed predation and competition may also vary in their intensity along dune gradients (Polis 1991), a fact which further complicates the interpretation of observed patterns of variation in the structure of desert dune communities. Thus, while observational studies of plant distribution are important for describing existing patterns of variation in the vegetation of desert dunes, they provide limited information concerning the mechanisms responsible for the observed patterns, and must be supported by demographic measurements and manipulative experiments to allow interpretation of their results. In the study reported here, manipulation experiments were conducted to investigate the combined effects of topographic position, sand instability and neighbour competition on the demography of two common annual species in the Nizzana sandfield. The results of the study confirmed the use of this experimental demographic approach, and pointed to the existence of complex interactions between the effects of topographic position, sand instability and neighbour competition. For example, in the absence of competitive effects, plants of both species suffered more mortality and produced less fruits on the slope of the dune than in the adjacent interdune corridor. In contrast, plants growing under natural density conditions exhibited a reverse pattern, with both survival and reproduction being higher on the slope of the dune. These results indicate that neighbour competition may have profound effects on patterns of local-scale variation in the demography of desert dune populations. The findings further demonstrate that demographic responses to both neighbour removal and sand manipulation were strongly species-specific, indicating that local gradients in the stability of the sand may be important in providing opportunities for coexistence of competing species. Below, I discuss some of the major interactions which were found in this study, their consequences for the demography of the studied species, and their implications to the understanding of distribution patterns of desert dune populations in particular and natural plant populations in general.
28.4.1
Species-Specific Responses to Habitat Heterogeneity
If two species differ from each other in their demographic responses to changes in habitat conditions, one would expect that the species x habitat interaction would be statistically significant. In the study reported here, significant interactions between
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
415
habitat type and species were recorded for three variables: above-ground biomass, reproductive biomass and fruit weight. The corresponding results obtained for reproductive allocation and the two measures of fruit number were not statistically significant. However, the results of these analyses reflect the combined responses of experimentally isolated plants and plants growing under natural density conditions and, thus, ‘mix’ changes in habitat conditions and corresponding changes in neighbour effects. A more rigorous test for species-specific responses to changes in habitat conditions should be based on the performance of plants growing in the absence of neighbour effects. Such an analysis was performed in a previous work (Kadmon 1994) and revealed statistically significant species x habitat interactions for all the variables examined. The two species differed from each other also in the relative sensitivity of various demographic parameters to changes in habitat conditions. For example, in B. fasciculatus, seedling survival was more sensitive to differences in habitat conditions than was fruit number per surviving plant, whereas a reverse pattern was obtained for N. procumbens (Kadmon 1994). These findings indicate that the demographic processes by which changes in habitat conditions are ‘translated’ into changes in population density may vary considerably among coexisting species.
28.4.2
Species-Specific Responses to Cover by Sand
By artificially covering seedlings of the two species with sand, I attempted to test the hypothesis that individuals of B. fasciculatus are more tolerant to sand instability than are individuals of N. procumbens. The experimental results were consistent with this hypothesis and revealed statistically significant treatment x species interactions for three variables: above-ground biomass, reproductive biomass and number of fruits per plant. As was expected, all of these responses were stronger in the case of N. procumbens. Moreover, there was some indication that B. fasciculatus responded in a positive manner to cover by sand (Fig. 28.4) but these responses were not statistically significant. The overall findings of this experiment support the hypothesis that differential tolerance to sand instability is important in determining the distribution patterns of N. procumbens and B. fasciculatus in the study area. Several previous studies (e.g. Pemadasa et al. 1974; Lee and Ignaciuk 1985; Maun and Lapierre 1986; Pakeman and Lee 1991) have reported species-specific responses of seeds to burial by sand. For example, Maun and Lapierre (1986) have investigated germination and emergence responses of four species of dune annuals to artificial seed burial and found that the sowing depth from which 50% of the seeds emerged, as well as the maximum depth from which a seedling emerged, varied in a consistent manner among the species studied. Differences among annual species in germination responses to cover by sand were reported also by Lee and Ignaciuk (1985). Such findings have indicated that local events of sand accretion may have profound effects on patterns of seedling emergence and, consequently, on the relative abundance of different species in dune communities. The results
416
R. Kadmon
obtained in the study reported here demonstrate that differential reproductive responses to sand accretion may also be important in structuring communities of dune annuals.
28.4.3
Species-Specific Responses to Neighbour Removal
Statistically significant species x neighbour removal interactions were obtained for five of the six variables examined, namely above-ground biomass, reproductive allocation, reproductive biomass, fruit number per germinating plant and fruit weight. These results indicate that the study species differed from each other not only in their responses to natural and experimental changes in the stability of the sand but also in their responses to the removal of neighbouring plants. Two types of hypotheses may explain these differential responses. One possibility is that the two species experienced different neighbour conditions resulting from micro-scale differences in their patterns of within-habitat distribution. An alternative possibility is that the two species differed from each other in their tolerance to competitive effects. The data obtained in this experiment do not allow one to directly distinguish between these two hypotheses. To control for differences in neighbour conditions, one has to apply more complex experimental designs where densities or neighbour conditions of target plants are held constant at different levels (Goldberg and Fleetwood 1987; Pantastico-Caldas and Venable 1993). An alternative approach is to transplant individuals of target species into randomly chosen sites in order to reduce the chance of biased neighbour conditions (Wilson and Keddy 1986a; Goldberg and Fleetwood 1987; Ryser 1993; Reader et al. 1994). The main purpose of the removal experiment conducted in this study was to test whether and how observed patterns of demographic variation are influenced by competitive interactions with neighbouring plants and, therefore, the control treatment was based on plants growing under natural neighbour conditions, rather than on randomly placed transplants. The results obtained from this experimental design allowed me to demonstrate that competitive interactions play an important role in determining patterns of spatial variation in the demography of the study populations (see next section). In reviewing evidence of competition, Gause (1934) concluded that the intensity of competitive interactions tends to decrease from relatively favourable habitats towards habitats characterized by strong abiotic stress. This idea was further developed into a theory of adaptive strategies by Grime (1977), who considered competitive ability and stress tolerance as two extremes of a continuum of adaptive characteristics. Based on this idea, and assuming that a trade-off exists between the adaptations required to cope with competitive effects and those required to tolerate the abiotic stress caused by sand movement, it could be expected that N. procumbens (which was more sensitive to sand instability) would be less influenced by neighbour removal than is the case for B. fasciculatus. The experimental results, however, revealed a reverse pattern, with individuals of N. procumbens showing significantly stronger responses to neighbour removal than did individuals of
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
417
B. fasciculatus. Although it is not possible to directly determine whether this outcome reflects differences in the neighbour conditions experienced by individuals of the two species or differences in their tolerance to competitive effects, some conclusions can be obtained by comparing the performance of experimentally isolated plants with that of plants growing in the control quadrats of the sand manipulation experiment. Such a comparison may indicate differences in tolerance to competition because the quadrats used for the sand manipulation experiments were thinned into constant densities of both species (see Methods). The results indicate that differences between the performance of plants subjected to the two density treatments (complete neighbour removal vs. thinning into constant densities) were always greater in N. procumbens, implying that this species was less tolerant to neighbour competition than was B. fasciculatus. This result contradicts another prediction, namely that species which attain relatively large size when growing with no competition should be more tolerant to competitive effects than are smaller species (Wilson and Keddy 1986b; Gaudet and Keddy 1988; Keddy 1989b). The results indicate that the above-ground biomass of N. procumbens was much larger than that of B. fasciculatus. Further observations of root systems indicated that both root biomass and the depth of root penetration into the sand were much higher in N. procumbens than in B. fasciculatus, implying that the observed differences in competitive responses could not be attributed to differences in below-ground biomass. These results suggest that information on plant size may not be sufficient to predict competitive responses in plant communities.
28.4.4
Interactions Between Neighbour Competition and Habitat Heterogeneity
Statistically significant interactions between the effects of neighbour removal and habitat type were obtained for above-ground biomass, reproductive allocation, reproductive biomass, fruit number per surviving plant and fruit number per germinating plant (Table 28.2). These results indicate that demographic responses to neighbour removal were strongly dependent on habitat conditions. Habitat-specific responses of plants to neighbour removal have been documented in several previous studies (Friedman and Orshan 1974; Keddy 1989b; Reader and Best 1989; Reader et al. 1994; Kadmon 1995; Tielbörger and Kadmon 2000), although only few of these studies involved measurements of plant reproductive success (Friedman and Orshan 1974; Kadmon 1995; Tielbörger and Kadmon 2000). Other studies have applied controlled density manipulation experiments to investigate demographic responses of plants to local gradients in habitat conditions, and found that the shape of the density-dependence curve may vary between habitats (Keddy 1981; Pantastico-Caldas and Venable 1993). These results, together with further results from studies in which habitat characteristics such as resource concentration (Klikoff 1966; Inouye et al. 1980; Wilson and Tilman 1991, 1993; Turkington et al. 1993; Kadmon 1995) or standing crop (Turkington et al. 1993; Reader et al.
418
R. Kadmon
1994) were manipulated experimentally, are consistent with the hypothesis that spatial heterogeneity in habitat conditions may interact with competition for limiting resources in structuring natural plant populations. The experimental results demonstrate that plants growing on the slope of the dune experienced weaker competitive effects than did conspecific plants growing in the interdune corridor. Such a pattern was obtained for both N. procumbens and B. fasciculatus, despite observed differences between the two species in the magnitude of their within-habitat responses to neighbour effects (Figs. 28.1–28.3). These data are consistent with the hypothesis that the intensity of competitive interactions tends to decrease with increasing abiotic stress (Grime 1973, 1977; Stadt et al. 1994). Results from a previous study (Kadmon and Leschner 1995) have shown that the main difference in abiotic conditions between the interdune area and the slope of the dune was in the mobility of the sand, which was tenfold greater on the slopes. It has also been found that both total biomass and total density of annual plants increased from the top of the dunes towards the interdune corridors (Kadmon and Leschner 1995). These findings suggest that differences in densities and standing crop caused by variation in sand mobility were the main causes of the observed pattern of variation in the intensity of competitive effects. The results of the neighbour removal experiments further indicate that competitive effects were important in determining the observed patterns of between-habitat variation in the demographic performance of the species studied. Actually, the experimental results demonstrate that plants growing in the absence of competitive effects performed better in the interdune corridor, whereas conspecific control plants growing under natural density conditions performed better on the slope of the dune (Figs. 28.1–28.3). Thus, competitive effects reversed the patterns of between-habitat variation expected on the basis of differences in abiotic conditions between the two habitats. This result points to the importance of taking into account the potential effects of competitive interactions in interpreting observed demographic responses of desert dune populations to local gradients in sand stability.
28.5
Conclusions and Summary
A variety of abiotic and biotic factors are involved in structuring ecological communities. Identifying such factors, evaluating their relative importance, and determining the demographic mechanisms by which they affect the dynamics and structure of different species in the community is a major goal of ecology. Yet, this is not an easy task because different factors may be correlated with each other or interact with each other in determining the values of demographic parameters. In this study, I focused on two factors which were considered to be important in structuring desert dune communities: spatial heterogeneity in the stability of the sand and neighbour competition (also see Chap. 27, this volume). The study was based on a set of manipulative experiments designed to test whether and how coexisting species of annual plants respond to the combined effects of sand instability and
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
419
neighbour competition. The results demonstrated that the two species studied responded differentially to spatial heterogeneity in sand stability, and that the observed patterns of demographic variation could be attributed, at least partially, to differences in the tolerance of seedlings to cover by sand. Mortality and reproductive responses of individual plants to both natural and experimental changes in the stability of the sand were consistent with observations of distribution patterns made at the community level, indicating that both types of responses were important in structuring the study populations. The results further indicated that competitive interactions with neighbouring plants were important in determining the survival and reproductive success of both species, and that demographic responses of individual plants to neighbour competition were strongly species-specific as well as habitat-dependent. Moreover, as a result of stronger competitive effects in the relatively stable interdune habitat, plants germinating in this habitat experienced higher mortality and lower reproductive rates than did conspecific plants germinating in the abiotically less favourable slope habitat. This finding indicates that competitive interactions may reverse patterns of demographic variation caused by local gradients in sand stability. The overall results indicate that spatial heterogeneity in habitat conditions may interact with competitive effects, as well as with differences among species in tolerance to sand instability and competition, in structuring communities of desert dune annuals. Acknowledgements I thank H. Leschner for her considerable help in the field. The study was carried out at the Nizzana experimental station of the Hebrew University-Minerva Arid Ecosystem Research Center, and was supported by the Israel Science Foundation administrated by the Israel Academy of Sciences and Humanities.
References Abbas JA, Mohammed A, Saleh MA (1991) Edaphic factors and plant species distribution in a protected area in the desert of Bahrain island. Vegetatio 95:87–93 Abrahmson WG (1975) Reproductive strategies in dewberries. Ecology 56:721–726 Aerts R, de Caluwe H (1994) Nitrogen use efficiency of Carex species in relation to nitrogen supply. Ecology 75(8):2362–2372 Bowers JE (1982) The plant ecology of inland dunes in western north America. J Arid Environ 5:199–220 Burger JC, Louda SM (1994) Indirect versus direct effects of grasses on growth of a cactus (Opuntia fragilis): insect herbivory versus competition. Oecologia 99:79–87 Chapin FS, Autumn K, Pugnaire F (1993) Evolution of suites of traits in response to environmental stress. Am Naturalist 142:s78–s92 Condit R, Hubbell SP, Foster RB (1994) Density dependence in two understory tree species in a neotropical forest. Ecology 75(3):671–680 Connell JH (1972) Community interactions on marine rocky intertidal shores. Annu Rev Ecol Systematics 3:169–192 Danin A (1991) Plant adaptations in desert dunes. J Arid Environ 21:193–212 Fowler NL, Antonovics J (1981) Small-scale variability in the demography of transplants of two herbaceous species. Ecology 62(6):1450–1457 Friedman J, Orshan G (1974) Allopatric distribution of two varieties of Medicago laciniata (L.) mill. in the Negev desert. J Ecol 62:107–114
420
R. Kadmon
Gaudet CL, Keddy PA (1988) Predicting competitive ability from plant traits. Nature 334:242–243 Gause GF (1934) The struggle for existence. Hafner, New York Goldberg DE, Fleetwood L (1987) Competitive effect and response in four annual plants. J Ecol 75:1131–1143 Grace JB (1990) On the relationship between plant traits and competitive ability. In: Grace JB, Tilman D (eds) Perspectives on plant competition. Academic Press, San Diego, CA, pp 51–66 Grace SL, Platt WJ (1995) Neighbourhood effects on juveniles in an old-growth stand of longleaf pine, Pinus palustris. Oikos 72:99–105 Grenot CJ (1974) Physical and vegetational aspects of the Sahara desert. In: Brown GWJ (ed) Desert biology, vol 2. Academic Press, New York, pp 103–164 Grime JP (1973) Competitive exclusion in herbaceous vegetation. Nature 242:344–347 Grime JP (1977) Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. Am Naturalist 11:1169–1194 Grime JP (1979) Plant strategies and vegetation processes. Wiley, Chichester Inouye RS (1991) Population biology of desert annual plants. In: Polis GA (ed) The ecology of desert communities. University of Arizona Press, Tucson, AZ, pp 27–54 Inouye RS, Byers GS, Brown JH (1980) Effects of predation and competition on survivorship, fecundity, and community structure of desert annuals. Ecology 61(6):1344–1351 Kadmon R (1994) Ecology of linear dunes. II. Differential demographic responses of annual plants to local scale variation in sand stability. Israel J Plant Sci 42:275–284 Kadmon R (1995) Plant competition along soil moisture gradients: a field experiment with the desert annual Stipa capensis. J Ecol 83:253–262 Kadmon R, Leschner H (1995) Ecology of linear dunes: effect of surface stability on the distribution and abundance of annual plants. Adv GeoEcol 28:125–143 Keddy PA (1981) Experimental demography of the sand-dune annual, Cakile edentula, growing along an environmental gradient in Nova Scotia. J Ecol 69:615–630 Keddy PA (1982) Population ecology on an environmental gradient: Cakile edentula on a sand dune. Oecologia 52:348–355 Keddy PA (1989a) Competition. Chapman and Hall, New York Keddy PA (1989b) Effects of competition from shrubs on herbaceous wetland plants: a 4-year field experiment. Can J Bot 67:708–716 Klikoff LG (1966) Competitive response to moisture stress of a winter annual of the Sonoran desert. Am Midl Naturalist 75:383–391 Lee JA, Ignaciuk R (1985) The physiological ecology of strandline plants. Vegetatio 62:319–326 Lubchenco J (1980) Algal zonation in the New England rocky intertidal community: an experimental analysis. Ecology 61:333–344 MacArthur RH, Wilson EO (1967) The theory of island biogeography. Princeton University Press, Princeton, NJ Mack RN, Pyke DA (1983) The demography of Bromus tectorum: variation in time and space. J Ecol 71:69–93 Maun MA, Lapierre J (1986) Effects of burial by sand on seed germination and seedling emergence of four dune species. Am J Bot 73(3):450–455 Moreno-Casasola P (1986) Sand movement as a factor in the distribution of plant communities in a coastal dune system. Vegetatio 65:67–76 Moreno-Casasola P, Espejel I (1986) Classification and ordination of coastal sand dune communities along the gulf and Caribbean Sea of Mexico. Vegetatio 66:147–182 Pakeman RJ, Lee LA (1991) The ecology of the strandline annuals Cakile maritima and Salsola kali. I: environmental factors affecting plant performance. J Ecol 79:143–153 Pantastico-Caldas M, Venable DL (1993) Competition in two species of desert annuals along a topographic gradient. Ecology 74(8):2192–2203 Pemadasa MA, Greig-Smith P, Lovell PH (1974) A quantitative description of the distribution of annuals in the dune system at Aberffraw, Anglesey. J Ecol 62:379–402
28 Demography of Annual Plants: The Role of Habitat Heterogeneity and Competition
421
Polis GA (1991) The ecology of desert communities. University of Arizona Press, Tucson, AZ Prasse R (1999) Experimentelle Untersuchungen an Gefäßpflanzenpopulationen auf verschiedenen Geländeoberflächen in einem Sandwüstengebiet. Universitätsverlag Rasch, Osnabrück Reader RJ, Best BJ (1989) Variation in competition along an environmental gradient: Hieracium floribundum in an abandoned pasture. J Ecol 77:673–684 Reader RJ, Wilson SD, Belcher JW, Wisheu I, Keddy PA, Tilman D, Morris EC, Grace JB, McGraw JB, Olff H, Turkington R, Klein E, Leung Y, Shipley B, Van Hulst R, Johansson ME, Nilsson C, Gurevitch J, Grigulis K, Beisner BE (1994) Plant competition in relation to neighbor biomass: an intercontinental study with Poa pratensis. Ecology 75(6):1753–1760 Ryser P (1993) Influences of neighbouring plants on seedling establishment in limestone grassland. J Veget Sci 4:195–202 Seely MK (1991) Sand dune communities. In: Polis GA (ed) The ecology of desert communities. University of Arizona Press, Tucson, AZ, pp 348–382 Seely MK, Louw GN (1980) First approximation of the effects of rainfall on the ecology and energetics of a Namib desert dune ecosystem. J Arid Environ 3:25–54 Shabel AB, Peart DR (1994) Effects of competition, herbivory and substrate disturbance on growth and size structure in pin cherry (Prunus pennsylvanica L.) seedlings. Oecologia 98:150–158 Silvertown J, Dale P (1991) Competitive hierarchies and the structure of herbaceous plant communities. Oikos 61:441–444 Sokal RR, Rohlf PJ (1981) Biometry, 2nd edn. W.H. Freeman, New York Solbrig OT, Simpson BB (1977) A garden experiment on competition between biotypes of the common dandelion (Taraxacum officinale). J Ecol 65:427–430 Southwood TRE, May RM, Hassell MP, Conway GR (1974) Ecological strategies and population parameters. Am Naturalist 108:791–804 Stadt KJ, Taylor GJ, Dale RT (1994) Measuring the effect of abiotic stress on competition. Oecologia 100:221–228 Swank SE, Oechel WC (1991) Interactions among the effects of herbivory, competition, and resource limitation on chaparral herbs. Ecology 72(1):104–115 Tielbörger K (1997) The vegetation of linear desert dunes in the north-western Negev, Israel. Flora 192:261–278 Tielbörger K, Kadmon R (1995) The effect of shrubs on emergence, survival and fecundity of four coexisting annual species in a sandy desert ecosystem. Ecoscience 2:141–147 Tielbörger K, Kadmon R (2000) Indirect effects in a desert plant community: is competition among annuals more intense under shrub canopies? Plant Ecol 150:53–63 Tremmel DC, Bazzaz FA (1993) How neighbour canopy architecture affect target plant performance. Ecology 74(7):2114–2124 Turkington R, Klein E, Chanway CP (1993) Interactive effects of nutrients and disturbance: an experimental test of plant strategy theory. Ecology 74:863–878 van Tienderen PH (1992) Variation in a population of Plantago lanceolata along a topographical gradient. Oikos 64:560–572 Walter H, Box EO (1983) The Karakum desert, an example of a well studied biome. In: West N (ed) Ecosystems of the World, vol 5. Temperate deserts and semideserts. Elsevier, Amsterdam, pp 105–159 Weiner J (1988) The influence of competition on plant reproduction. In: Doust JL, Doust LL (eds) Plant reproductive ecology: pattern and strategies. Oxford University Press, New York, pp 228–245 Wilson SD, Keddy PA (1986a) Measuring diffuse competition along an environmental gradient: results from a shoreline plant community. Am Naturalist 127:862–869 Wilson SD, Keddy PA (1986b) Species competitive ability and position along a natural stress/disturbance gradient. Ecology 67:1236–1242 Wilson SD, Tilman D (1991) Components of plant competition along an experimental gradient of nitrogen availability. Ecology 72(3):1050–1065
422
R. Kadmon
Wilson SD, Tilman D (1993) Plant competition and resource availability in response to disturbance and fertilization. Ecology 74(2):599–611 Yeaton RI (1988) Structure of the Namib dune gradients: characteristics of the environmental gradients and species distributions. J Arid Environ 76:744–758
Part D
Research Perspectives / Synthesis and General Conclusions
Chapter 29
Sensitivity of a Sandy Area to Climate Change Along a Rainfall Gradient at a Desert Fringe A. Yair, M. Veste, R. Almog, and S.-W. Breckle
29.1
Introduction
Global climate change has become a strongly and frequently addressed issue in the last decades. The aspect is crucial in dry-land areas, which cover approximately one third of the globe’s total land area. The relationship between average annual rainfall and environmental variables has attracted the attention of many scientists. Climatologists use aridity indices to express relationships between climatic and environmental variables (Köppen 1931; Budyko 1974; Wallen 1967; Bailey 1979). These indices, based on purely climatic variables such as annual precipitation, temperature, evaporation and radiation, tend to imply that the acuteness of aridity is inversely related to annual precipitation. Although aware that soil water content depends on local soil type and precipitation regime, Walter (1939, 1960) asserted that at a larger, global scale, standing biomass is positively correlated to average annual rainfall. This approach is still followed by many researchers who assume a positive relationship between average annual rainfall and environmental variables such as water availability for plants, vegetation cover, productivity, species diversity, soil properties, human activity, and erosion rates for sub-humid to arid areas (Issar and Bruins 1983; Shmida 1985; Seely 1991; Lavee et al. 1991; Kutiel et al. 2000; Meron et al. 2004). This approach is certainly correct at the global scale, as well as for non-irrigated annual crops in dry-land areas. It is, however, questionable for arid and semi-arid areas, usually regarded as highly sensitive to climate change, especially for perennial plants. With decreasing annual rainfall, the number of rainstorms and storm rain amounts decrease. Under such conditions, water availability for plants may be highly dependant on the relationships between rainfall and surface properties which greatly influence the degree to which water will percolate or will be transformed into runoff, thereby significantly affecting the spatial redistribution of water resources. For example, it is well known that rocky hill slopes devoid of extensive soil and vegetation cover are characterized by extremely low infiltration rates, and quickly develop surface runoff. Due to the short duration of most individual rain showers, flow distances are short, resulting in water concentration and deep water percolation at nearby down-slope positions (Yair and Danin 1980; Yair 1983, 1994,
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
425
426
A. Yair et al.
1999). Under given rainfall conditions, the degree of water concentration is expected to be more limited in sandy soils devoid of topsoil crusts, where infiltration depth is quite high due to the low water-holding capacity, and the high porosity of unconsolidated sand. However, due to the latter properties of sand, water preservation in sandy areas is generally good. The poorest water regime is to be expected in finegrained soils (loess or silty-clayey soils) characterized by their high water-absorption capacity which limits the depth of water percolation, and enhances evaporation losses during the hot time intervals between consecutive rainstorms. Under such conditions, we contend that, in dry-land areas, surface and rainfall characteristics which encourage water concentration and water preservation can be considered as the main component in water availability, rather than the absolute average annual rain amount. We therefore postulate that semi-arid and arid ecosystem characteristics will not necessarily correlate with the actual amount of rainwater, but rather with the degree of water concentration and water preservation. In other words, a given climatic change in dry-land areas would be expected to have differential effects in a rocky area, in a loess-covered area, and in a sandy area. Furthermore, a nonuniform effect should be expected within each of the above physiographic units, due to spatial differences (over short distances) in percolation and water redistribution by surface runoff. The common assumption regarding the positive relationship between annual precipitation and water resources disregards the fact that a climatic change in semiarid areas, especially at the desert fringe, is not limited to purely climatic variables such as precipitation, temperature or wind regime. It is almost always accompanied by a parallel change in surface properties, such as associated with sand deposition during a transition to a dry climatic phase, and loess deposition during a wetter phase. The new surface properties can be expected to exercise a strong effect on infiltration, surface runoff, and water availability for plants. This raises an interesting question: does the alteration of surface properties enhance the expected positive effect of rainfall increase, or reduce it or even eliminate it? Earlier studies, conducted in the northern Negev desert, had shown that loess deposition on top of rocky surfaces resulted in a substantial decrease in runoff generation, coupled with rainwater absorption at a shallow depth by the loess mantle (Yair 1983, 1994). Furthermore, a laboratory experiment (Yair and Bryan 2000) showed that slight changes in surface properties may have a dramatic effect on the hydrological regime associated with sand or loess deposition. A sand layer 1–2 mm thick deposited on a fine-grained substratum is sufficient to eliminate surface runoff generation, whereas a fine-grained layer 1–2 mm thick deposited on sand has a reverse effect. Such results lead to the idea that, following a transition to a dry climatic phase, sand deposition may improve water availability because of deep infiltration of all rainwater, and because of good water preservation resulting from limited evaporation losses. A reverse situation may develop where loess is deposited when passing from a dry to a wet climatic phase. The loess mantle, characterized by a high water-absorption capacity, would be expected to limit the depth of rainwater penetration. Infiltrated waters would be completely lost via evaporation, leading to salt accumulation at a shallow depth and, with time, to soil salinization (Yair and Shachak 1987; Yair 1994).
29 Sensitivity of a Sandy Area to Climate Change Along a Rainfall Gradient
427
Finally, the term resilience is often used to describe the degree to which an ecosystem can be disturbed or affected by climatic change, and yet revert to its initial composition and structure. A system disturbed beyond its level of resilience will develop into a new ecosystem. Such a dramatic change can be triggered by human activity. For example, the introduction of heavy grazing into a grassland area is often responsible for the replacement of grass cover by shrubland, with severe bush encroachment (Jeltsch et al. 1997). The same has been predicted for semi-arid areas, due to increasingly warmer and drier climatic conditions (Schlesinger et al. 1990). However, semi-arid and particularly arid ecosystems are regarded by some ecologists as being resistant to drought (Thiery 1982; Holling 1983; Wiens 1985). Perennial plants in these ecosystems are adapted to extreme variability in climatic conditions from year to year, and over a timescale of decades. Under such conditions, a rather extreme climatic change, mainly rainfall, would be required in order to seriously affect natural semi-arid and arid ecosystems. The sandy area along the Israeli-Egyptian border offers quite unique conditions for the analysis of the possible effects of climatic change on a sandy arid ecosystem, characterized by semi-stable to stable dunes (Chap. 2, this volume). The whole area is composed of uniform quartzitic sand. However, the rainfall gradient is very sharp, passing from approx. 170 mm average annual rainfall in the north to approx. 90 mm in the south, over a distance of 35 km. The northern area is classified as arid and the southern area as hyper-arid (Chap. 6, this volume). In view of the specific sand properties introduced above (limited water absorption by sand grains, and high porosity), one would expect deeper water penetration and water preservation with increasing annual rainfall. Such a view would be in accordance with the prevailing idea regarding the positive relationship between average annual rainfall and environmental variables. This view may be valid in the case of unconsolidated sand but it is questionable for sandy areas stabilized by a thin topsoil crust rich in fine-grained particles and biological elements. Earlier studies conducted in the southern part of the area, within the Nizzana research site (Yair 1990, 2001; Kidron and Yair 1997) had already shown that the topsoil biological crust plays an important role in the local water regime, strongly affecting rainwater infiltration, runoff generation, and the spatial redistribution of water resources (Chaps. 6, 16 and 17, this volume). Field observations show a differential development of the topsoil crust from north to south along the rainfall gradient. The topsoil crust is thicker and darker in the northern than in the southern part of the sand field, pointing to differences in the composition and properties of the crust which may affect its hydrological behaviour and, consequently, the water regime along this rainfall gradient (Almog and Yair 2007). In addition, the area covered by the topsoil crust increases from south to north.
29.2
Aim of Study
Contrasting views exist on how topsoil biological crusts affect infiltration, runoff generation, and soil moisture regime. Some authors (Booth 1941; Loope and Gifford 1972; Perez 1997; Eldridge and Tozer 1997) contend that the cohesive
428
A. Yair et al.
and flexible biological elements of the crust absorb raindrop energy and prevent the development of a rain crust conducive to surface sealing and runoff generation. Such soil crusts would be expected to increase infiltration and soil moisture content. Other authors (Bond 1964; Avnimelech and Nevo 1964; Roberts and Carson 1971; Yair 1990; Dekker and Jungerius 1990) adopt a reverse view. They claim that water absorption, and the swelling of microorganisms and fine-grained elements forming the crust limit infiltration rate and, under suitable rainfall conditions, develop surface runoff, resulting in spatial differences in the soil water regime. The major aim of this study is to check the validity of existing models regarding the positive correlation between annual rainfall and environmental variables in a sandy area with an extensive cover of a topsoil biological crust. The following aspects will be dealt with along the rainfall gradient described above: 1. Spatial distribution pattern of rain amounts at the annual and individual rainstorm scales. 2. Vegetation cover and plant species diversity in different habitat types along the gradient. 3. Spatial variability of topsoil crust cover and properties (Chaps. 10 and 11, this volume). 4. Spatial distribution of dead and living perennial shrubs. 5. Effects of topsoil crust properties on infiltration, runoff generation, and soil moisture regime (Chaps. 17, 18 and 20, this volume).
29.3
Methodology
The data collected cover the rainfall years 1998–2000 and 2001–2003. The study is based on five monitoring sites (Fig. 29.1). Sites N1, N3 and N5 were equipped with rain recorders, and sites N2 and N4 with rain gauges. Investigations of vegetation cover and species diversity were limited to sites N1, N3 and N5. Plots of 5–5 m were located along three parallel lines at sites covering the different ecotopes (Chap. 2, this volume) along a transect incorporating the following: the dune crest, north-facing dune slope, dune slope, dune base, interdune area and south-facing dune slope. In addition, a field survey of the cover of dead and living perennial shrubs was conducted along three transects (100×1 m) at sites 1, 3 and 5, following the approaches of Bauer (1943), Barour et al. (1987) and Kent and Coker (1992). The vegetation transects were located at the dune base where, for topographical and hydrological reasons, the density of the vegetation is highest in the area (Chap. 18, this volume). It is assumed that the survival and mortality of the perennial vegetation is a good indicator of the effect of water availability at the decadal timescale. Runoff plots, located at the base of north-facing slopes and equipped with stage recorders and runoff and sediment collectors, were established at the five sites. Runoff plots had a uniform area of 8 m2. Soil water content was measured after each rainstorm down to a depth of 80 cm. At each site, the topsoil crust (Chap.10, this
29 Sensitivity of a Sandy Area to Climate Change Along a Rainfall Gradient
429
Fig. 29.1 Research area and location of monitoring sites
volume) was analyzed for its thickness, particle size distribution, organic matter content, and water-absorption capacity. Scanning electron microscope photographs provided an insight into the appearance of the biological crust elements.
29.4 29.4.1
Results Rainfall
Annual rain amounts recorded during the monitoring period are presented in Table 29.1. During the 4 years of this study, annual rain amount was below the long-term average. However, it is interesting and important to note that the longterm trend of rainfall decrease from north (site N5) to south (site N1) is not evident in all years. Rain amounts followed the long-term trend in 1998–2000 and 2001–2002, but deviated from the long-term trend the year after. In 2002– 2003, the southern site (N1) showed a rain amount slightly higher than the long-term average, while the northern site (N5) showed only half of the long-term average. This is due to the synoptic patterns prevailing in the eastern Mediterranean region (Chap. 4, this volume). Rainfall can penetrate the area from the northwest, northeast,
430
A. Yair et al. Table 29.1 Annual rain amounts along the climatic gradient Annual rainfall (mm) Site
1998–1999
1999–2000
2001–2002
2002–2003
N5 N3 N1
49.8 30.3 28.8
55.3 53.1 27
146.5 76.2 57.7
82.7 95.1 99.3
Table 29.2 Spatial distribution of selected rain events Storm rain amount (mm) N1 N3 N5 30 Nov. 2001 28–30 Jan. 2001 9–12 Dec. 2002 17–22 Dec. 2002
2 6.1 14.2 24.4
11.5 8.7 11.1 8.8
20.9 8.7 4.1 0
southwest and southeast. Furthermore, during many storms the rain-bearing clouds are limited in their dimensions, do not cover the whole north-western Negev dune field, and often split into localized, patchy and spotty rain cells. Table 29.2 displays some examples of the spatial rainfall pattern described above. Needless to say, the non-uniform and non-systematic spatial distribution of rain amounts at the annual and rainstorm timescales must affect spatial vegetation patterns, without even considering the possible effects of the biological topsoil crust and other factors on spatial differences in water resources.
29.4.2
Vegetation Changes Along the Rainfall Gradient
Vegetation cover and plant species diversity were investigated in different habitat types. In total, 90 plant species (see list in Veste et al. 2005) were identified along the gradient within the study plots: 64% (57 species) are annuals and 36% (33 species) are perennials (Fig. 29.2; site N1, Chap. 7, this volume). No significant difference in the number of perennial species is observed in the area investigated (Fig. 29.2A). Plants occurring at all sites are Artemisia monosperma, Asthenaterum forsskalii, Atractylis cuneata, Erodium crassifolium, Moltkiopsis ciliata, Retama raetam and Thymelaea hirsuta (Veste et al. 2005). Stipagrostis scoparia and Cornulaca monocantha have their most extensive distribution at the most southern site, N1, where the area with mobile sand is relatively extensive. When annual plants, highly sensitive to annual rainfall amount and timing of rainfall, are considered, the highest number of species is found in the central site N3, the lowest at the wetter northern site N5 (Fig. 29.2A). The lowest mean vegetation cover of perennials is found at the southern site N1. A sharp increase is observed at site N3, with no difference between latter site and site N5 (Fig. 29.2B). When annual
29 Sensitivity of a Sandy Area to Climate Change Along a Rainfall Gradient
431
plants are considered, there is no difference between sites N1 and N3, but the mean cover of annuals is slightly higher at site N5. Figure 29.3 displays the percentage cover of perennial shrubs in the four major ecotopes, namely the crests, interdunes, and the north- and south-facing slopes. Apart from the crest, where the percentage cover increases from site N1 to site N5 with increasing annual rainfall, no clear systematic trends can be detected along the rainfall gradient for any of the other ecotopes. However, increasing surface stability from south to north
no of species
60 50
perennials
A
40
annuals
30 20 10
mean cover [%]
0 25
B
20 15 10 5 0 1
3
5
site Fig. 29.2 Species diversity (A) and mean vegetation cover (B) along the rainfall gradient
crest 40
B
A A
interdune A
A,B
30
30 A
20
10 0
10
N-slope
40 A
B
A
20
30 20
A
10 0
0 40
A,B
A
30
S-slope
10 1
3
5
1
site Fig. 29.3 Vegetation cover at the major geo-ecological units
3
site
5
0
cover [%]
cover [%]
20
40
432
A. Yair et al.
enables more perennial plants to establish on the dune crests. The highest cover (27–32%) is found in the interdune corridors, with no significant differences along the gradient. Of special interest is the fact that, at site N5, no difference exists between the crest and the interdune area. At this site, almost the whole dune from top to bottom is stabilized and covered by the topsoil crust. Site N3 has a similar, quite high cover on both the north- and south-facing slopes. The percent cover of north-facing slopes is slightly higher at the southern site N1 than at the northern site N5. This trend is, however, reversed for south-facing slopes at sites N1 and N5. Figure 29.4 displays the percent cover of the topsoil crust along the gradient. An increase in percent cover is observed from south to north for the crests of the dunes (Chap. 10, this volume). Crust cover in the interdune corridors is quite uniform throughout the area, with a very slight increase from site N1 to site N5 (from 93 to 97%). Crust cover increase from south to north is not gradual. It is approximately 50% at site N1, 90% at site N3 and 93% at site N5. Figure 29.5 presents the percentage of living and dead perennial shrubs at the dune base at sites N1, N3 and N5. The data obtained clearly show a linear increase in the percentage of dead shrubs with increasing average annual rainfall. A reverse trend is observed for the living shrubs. Such a pattern is indicative of the negative long-term trend in water availability with increasing annual rainfall (Chap. 19, this volume).
29.4.3
Hydrological Aspects
Table 29.3 presents the data on rainfall–runoff relationships for sites N1, N3 and N5 during the years 2001–2003. As could be expected, due to the fluctuations in rainfall, the trends in the spatial distribution of annual rainfall are not similar for the 2 years. However, a clear trend appears in both years for runoff generation. The runoff volumes collected decrease with increasing annual rainfall. In 2001–2002, rainfall
100
mean crust cover [%]
crest 80
north-facing slope interdune
60 40 20 0 1
3
5
Fig. 29.4 Mean topsoil crust cover along the rainfall gradient
Perennial vegetation cover (%)
29 Sensitivity of a Sandy Area to Climate Change Along a Rainfall Gradient
433
Plot
35 N1
30
N2
N3
25
N4
R2 = 0.85
20
N5
Dead
Alive
15 10 5
R2 = 0.95
0 80
100
120
140
160
180
Average Annual Rainfall (mm) Fig. 29.5 Dead and living perennial vegetation along the rainfall gradient (March 2002)
Table 29.3 Rainfall–runoff relationships (2001–2003) Annual Annual Rainfall year Plot rainfall (mm) runoff (l) 2001–2002
2002–2003
N5 N3 N1 N5 N3 N1
146.5 76.2 57.7 82.7 95.1 99.3
4.4 7.5 14.4 1 8.5 28.8
at the northern site N5 was three times higher that at the southern site N1. However, the runoff volumes collected at site N1 were 3.7 times higher than those at site N5. The year after, annual rain amounts were quite uniform all over the area, yet the ratio of runoff to rainfall for plots N1/N5 is almost 30 times higher. The runoff data obtained are supported by data collected during sprinkling experiments conducted in the area. Under wet surface conditions, the rain threshold required for runoff generation was 10–12 mm h−1 at plot N1, 19–22 mm h−1 at plot N3, and above 40 mm h−1 at plot N5. These findings may be indicative of two opposing processes: (1) an increase in infiltration and deep rainwater penetration with increasing rainfall, or (2) high water absorption by the topsoil crust with increasing annual rainfall, which limits infiltration depth. In the case of high infiltration rate with increasing annual rainfall, one would have expected an increase in infiltration depth, particularly for the wet, first rainy season. However, the data presented in Fig. 29.6 show a reverse trend for both years. In the first rainy season, despite the differences in rainfall and runoff, depth of water percolation is negatively correlated with annual rainfall. Water percolation reached 50 cm at site N1, 30 cm at site N3 and only 15 cm at site N5. Soil moisture
434
A. Yair et al. April 5 2002
March 15 2003 Moisture content (% by weight)
Moisture content (% by weight) 0
2
4
6
8
0
10
Depth (cm)
N1
20 40 60
Cumulative rain amount: 56 mm
80 100 2
4
6
8
N1
60
Cumulative rain Amount: 88 mm
80
0
4
8
12
16
20
24
N3
20
60
Cumulative rain amount: 76 mm
Cumulative rain amount: 78 mm
80 100
100
0
2
4
6
8
10
N5
40 80
24
40
80
100
20
40
0
N3
60
60
16
20
10
40
20
12
100
0
20
0
8
0
0
0
4
0
0
4
8
12
16
20
24
N5
20 40
Cumulative rain amount: 146 mm
60 80
Cumulative rain amount: 64 mm
100
Fig. 29.6 Soil moisture profiles along the rainfall gradient (two rain events at three sites, 2002–2003)
graphs show an additional interesting fact. Water absorption by the topsoil crust is always highest at site N5, and decreases southwards along the rainfall gradient. The runoff and moisture data obtained draw attention to the important role that should be attributed to differences in surface properties along the rainfall gradient considered. A gradual change in the properties of the topsoil crust is observed, for all variables, along the rainfall gradient (Table 29.4). Crust thickness (Figs. 29.7 and 29.8), organic matter content, and percent of silt and clay increase with increasing annual rainfall. Such trends explain the significant increase in the field capacity of the crusts from the southern drier to the northern wetter area. An additional factor contributing to the increase in water absorption by the topsoil crust with increasing rainfall is related to the composition of the biological elements of the crust (Fig. 29.8). The biological crust in the climatologically drier area is composed of a thin microbial film, with a dominance of cyanobacteria, and many voids between the sand grains (Chap. 20, this volume). The crust in the northern wetter area is quite different. It is very rich in large mosses known for their high waterabsorption capacity (Kidron 1995).
29 Sensitivity of a Sandy Area to Climate Change Along a Rainfall Gradient
435
Table 29.4 Properties of the topsoil crust along the rainfall gradient (average of 20 samples) Silt and clay content Organic matter content Crust thickness Field capacity Site location (%) (%) (mm) (% by weight) N5 N3 N1
49.2 37.8 27.7
11.2 7.1 5.3
7.6 4.2 2.7
19.5 11.9 6.1
Fig. 29.7 Topsoil crust thickness along the rainfall gradient, sites N1, N3 and N5
Fig. 29.8 ESEM images of the crust. Top Topsoil crust at the southern plot (N1). Thin and compacted sheet of biological elements (left), with many voids between the sand grains (right). Bottom Topsoil crust at the northern plot (N5). High concentration of large mosses acting as a sponge (left) and close-up view of one of the mosses (right)
436
29.5
A. Yair et al.
Discussion
The data presented above clearly demonstrate the complex relationships among average annual rainfall and environmental variables. Despite the fact that the sandy substratum is uniform all over the area, spatial variability along the rainfall gradient is not gradual and not systematic. Some aspects show a positive relationship with annual rainfall, while others show no clear relationship or even a negative relationship. The difference in long-term average annual rainfall from north to south resulted in a differential development of the topsoil crust. The crust is better developed and more extensive in the wetter area where it is thicker, and richer in organic matter, fine-grained particles and mosses than is the case for the southern topsoil crust (Table 29.4 and Fig. 29.8). This may be regarded as a positive environmental effect of rainfall increase. However, the better developed topsoil crust plays a negative role when the higher vegetation is considered. The crust in the wetter area (site N5) is able to absorb all rainfall during most rainstorms, strongly limiting the depth of rainwater penetration, subsurface flow water movement, surface runoff generation and, consequently, water availability for perennial shrubs (Fig. 29.6). At the same time, the thin crust in the drier area (site N1) absorbs less water, allows deeper water penetration and surface runoff generation during some of the rainstorms. Runoff generated over the crust infiltrates at the dune base. The overall result is deeper water penetration and higher water availability for vegetation in the crusted area. Water availability at the dune base, and at local concavities, is further enhanced by the process of subsurface flow. The processes described above explain why the survival of perennial plants is lower in the wetter than in the drier area (Fig. 29.5). Nevertheless, in this case the non-uniform spatial distribution of rainfall at the annual and rainstorm timescales also plays an important role. The quite uniform species diversity at the regional scale may be regarded as indicative of a low sensitivity of the perennial vegetation of the sandy area to climate change along the rainfall gradient considered (90–170 mm). The three aspects discussed above show different trends at the regional scale. Topsoil crust properties change positively with increasing rainfall; water availability, biomass (Chap. 25, this volume) and survival of the perennial vegetation show a negative correlation with average annual rainfall; species diversity and number of species show no correlation with increasing rain. A similar complexity is observed at the smaller, local scale of representative eco-geomorphic units within the dune system (Fig. 29.3). Vegetation cover increases with increasing annual rainfall for the crest of dunes, is quite uniform for the interdune corridors, but irregular and non-systematic for north- and southfacing slopes. Different factors may explain these results. The quite uniform cover in the interdune corridors may be explained by a combination of two factors: (1) these represent the most stable environment in the whole area; (2) the increase in water absorption by the topsoil crust with increasing annual rainfall eliminates the expected positive effect of rainfall increase. Surface runoff and subsurface flow are not regarded as important factors in the soil moisture regime of interdune
29 Sensitivity of a Sandy Area to Climate Change Along a Rainfall Gradient
437
areas. The positive relationship between vegetation cover and annual rainfall for the dune crests may be explained by the surface stability factor. Surface stability enhances the extent of vegetation and topsoil crust covers. However, surface stability is highly dependant on wind speed. A positive relationship exists between the relative elevation of dunes and wind speed, leading to decreased stability of elevated dunes and increased stability of low dunes. As the relative elevation of the sandy ridges is higher in the southern than in the northern part of the area, surface stability increases with increasing rainfall (Fig. 29.4), resulting in a more extensive vegetation cover on dune crests from south to north. The role of surface runoff and subsurface flow is evident for north-facing slopes. Water concentration by these two processes at the dune base is considered responsible for the dense vegetation belt observed at the base of north-facing slopes at site N1, where runoff frequency and magnitude are highest (Table 29.3). This high density explains the higher vegetation cover observed at the southern and drier site N1, contrasting with the northern wetter site N5 where runoff frequency and magnitude are extremely low. A reverse trend is observed at sites N1 and N5 for the south-facing slopes. This result may be explained by the combination of the limited frequency and magnitude of runoff generation at site N1 (Chap. 17, this volume), and the more extensive topsoil crust cover and surface stability at site N5. In view of the arguments presented above, it appears that the central site N3, where many of the characteristics are less extreme than at sites N1 and N5, on the whole enjoys the best water regime prevailing along the rainfall gradient considered (Figs. 29.2, 29.3 and 29.4).
29.6
Implications for the Sensitivity of the Sandy Area to Changing Climatic Conditions
The results obtained cast doubt on the generality of the prevailing assumption that an increase in average annual rainfall, especially in semi-arid areas usually regarded as highly sensitive to climate change, should always have a positive effect on the ecosystem. For some aspects, the relationship with annual rainfall is positive and, for others, negative or inexistent. Quite often, a single factor has a strong and determinant influence on the outcome. For example, a thick topsoil crust, able to absorb all rainwater during most rainstorms, seriously limits infiltration depth and water availability for the perennial vegetation, and counteracts the expected positive effect of a relatively high annual rain amount. On the other hand, the high frequency and magnitude of surface runoff in the drier part of the area has a positive effect on water resources via the process of water concentration (Chaps. 24, 25, 27 and 28, this volume). The data presented above bear some importance in view of the expected climate change related to global warming. In terms of the response of the sandy area considered to the foreseen increase in aridity, any attempt to project the results of the present study at a larger scale needs to consider changes and effects in both rainfall and wind regimes. The following scenarios may be advanced.
438
A. Yair et al.
- Along the gradient considered, a decrease in the average annual rain amount may have negative effects in the southern part of the area (site N1), due mainly to a decrease in the frequency and magnitude of runoff events responsible for the creation of “wet belts” at the base of the sandy ridges and at local concavities. In addition, it is quite possible that the spatial extent of the topsoil crust will decrease, further limiting runoff frequency and magnitude, while increasing the extent of the area with mobile sand. This scenario may be regarded as a desertification process due to rainfall decrease. Desertification will be greatly enhanced if, parallel to the reduced rainfall, wind velocities increase. Such a combination would be expected to increase the extent of the unstable and mobile sand surfaces. The scenario presented above is consistent with prevailing views on the positive relationship between average annual rainfall and environmental variables. - A reverse situation – improvement of the water regime – may develop in the northern wetter part of the study area (site N5). A decrease in rain amount may lead to a change in the composition of the topsoil crust (Chaps. 10 and 13, this volume). Under drier conditions, the composition of the topsoil crust will change. Mosses, dominant under present-day conditions at site N5, may be replaced by a crust rich in cyanobacteria, fungi and lichens, similar to the crust prevailing, at present, in the southern area. Under such conditions, the frequency and magnitude of runoff and of subsurface flow will increase, leading to a positive effect of water concentration. The overall result may be regarded as a positive effect despite the transition to a drier climatic phase. However, in the case of an increase in the frequency of strong winds, one would expect an increase in surface instability which would, nevertheless, be limited to the dune crests. A positive aspect of this scenario is that, due to the spatial shift in topsoil crust types, and of their effects on the water regime, a transition into a drier phase may not result in a reduction in species diversity for the sandy area as a whole. Species found today in the south would be sustained in the north.
29.7
Conclusions
The scenarios described above lead to the conclusion that any climatic change in the area will not be limited to purely climatic variables such as rainfall or wind regimes. It will be accompanied by a parallel change in surface properties, including changes in the extent of the topsoil biological crust, the composition and properties of the topsoil crust, and the extent of bare sand surfaces. These would be expected to have positive or negative effects in different sectors along the rainfall gradient considered, as well as within each of the sectors. Moreover, this would be expected to change the effects of perennials and shrubs on other life forms (Chaps. 27 and 28, this volume). Finally, the conclusions derived from the present study should not automatically be extrapolated to all other sandy arid areas. For the time being, this should be limited to sandy areas with similar conditions, namely a winter rain climate and an
29 Sensitivity of a Sandy Area to Climate Change Along a Rainfall Gradient
439
extensive occurrence of topsoil biological crusts indicative of stable surface conditions. The results obtained may not be applicable to sandy areas with strong winds or high-intensity summer rainstorms which prevent or limit the development of biological topsoil crusts. Acknowledgements We are grateful to Mrs. H. Leshner of the Hebrew University of Jerusalem for her assistance with the vegetation survey. Special thanks are due to Prof. E. Verrecchia of the University of Neuchâtel for his assistance in the use of the ESEM facilities, and to Mrs. M. Kidron of the Department of Geography for drawing the illustrations.
References Almog R, Yair A (2007) Negative and positive effects of topsoil biological crusts on water availability along a rainfall gradient in a sandy arid area. Catena 70:437–442 Avnimelech Y, Nevo Z (1964) Biological clogging of sands. Soil Sci 98:222–226 Bailey HP (1979) Semi-arid climates: their definition and distribution. In: Hall AE, Cannell GH, Lawton HE (eds) Agriculture in semi-arid environments. Springer, Berlin Heidelberg New York, pp 73–96 Barour MG, Burk JH, Pitts WD (1987) Terrestrial plant ecology. Benjamin Cummings, Menlo Park, CA Bauer HL (1943) The statistical analysis of chaparral and other plant communities by means of transect sampling. Ecology 24:45–60 Bond RD (1964) The influence of microflora on physical properties of sand. Effects associated with filamentous algae and fungi. Austr J Soil Res 2:123–131 Booth WE (1941) Algae as pioneers in plant succession and their importance in erosion control. Ecology 22:38–46 Budyko MI (1974) Climate and life. Academic Press, New York Dekker LW, Jungerius PD (1990) Water repellency in the dunes with special reference to the Netherlands dunes of the European coast. Catena suppl 18:173–183 Eldridge DE, Tozer ME (1997) A practical guide to soil lichens and bryophytes of Australia’s dry country. Department of Land and Water Conservation, Sydney Holling CS (1983) Resilience and stability of ecological systems. Annu Rev Ecol Systems 4:10–23 Issar A, Bruins HJ (1983) Special conditions in the deserts of Sinai and the Negev during the latest Pleistocene. Palaeogeogr Palaeoclimatol Palaeoecol 42:63–72 Jeltsch F, Milton SJ, Dean WRJ, von Rooyen N (1997) Analysing shrub encroachment in the southern Kalahari: a grid-based modeling approach. J Appl Ecol 34:1497–1508 Kent M, Coker P (1992) Vegetation description and analysis – A practical approach. Wiley, New York Kidron GJ (1995) The impact of microbial crusts upon runoff-sediment yield relationships on longitudinal dune slopes, Nizzana, Western Negev, Israel (in Hebrew with English summary). PhD Thesis, The Hebrew University of Jerusalem Kidron GJ, Yair A (1997) Rainfall-runoff relationships over encrusted dune surfaces, Nizzana, Western Negev, Israel. Earth Surface Processes Landforms 2:1169–1184 Köppen W (1931) Grundriss der Klimakunde. Gruyter, Berlin Kutiel P, Kutiel H, Lavee H (2000) Vegetation response to possible scenarios of rainfall variation along a Mediterranean–extreme arid climatic transect. J Arid Environ 44:277–290 Lavee H, Imeson Ac, Pariente P, Benyamini Y (1991) The response of soils to simulated rainfall along a climatological gradient in an arid and semi-arid region. Catena suppl 19:19–37 Loope WI, Gifford GF (1972) Influence of a soil micofloral crust on selected properties of soils under pinyon-juniper in southeastern Utah. J Soil Water Conserv 28:27–52
440
A. Yair et al.
Meron E, Gilad E, von Hardenberg J, Schachak M, Zarmi Y (2004) Vegetation patterns along a rainfall gradient. Chaos Solitons Fractals 19:367–376 Perez FL (1997) Microbiotic crusts in the high equatorial Andes and their influence on Paramo soils. Catena 31:173–198 Roberts FG, Carson BA (1971) Water repellence in sandy soils of southwestern Australia. Austr J Soil Res 10:35–42 Schlesinger WH, Reynolds JF, Cunningham GL, Huenneke LF, Jarrell RA et al. (1990) Biological feedbacks in global desertification. Science 247:1043–1048 Seely MK (1991) Sand dunes communities. In: Polis GA (ed) The ecology of desert communities. University of Arizona Press, Tucson, AR, pp 348–382 Shmida A (1985) Endemic plants of Israel (in Hebrew). Rotem, Bull Israel Plant Centre 3:3–47 Thiery RG (1982) Environmental stability and community diversity. Biol Rev 57:691–710 Veste M, Eggert K, Breckle SW, Littmann T (2005) Vegetation entlang eines geo-ökologischen Gradienten in der Negev. In: Veste M, Wissel C (Hrsg) Beiträge zur Vegetationsökologie der Trockengebiete und Desertifikation. UFZ Bericht 1/2005, Leipzig, pp 65–81 Wallen CC (1967) Aridity definitions and their applicability. Geogr Ann A 49:367–384 Walter H (1939) Grasland, Savanne und Busch der ariden Teile Afrikas in ihrer ökologischen Bedingtheit. Jahrb wiss Bot 87:750–860 Walter H (1960) Grundlagen der Pflanzenverbreitung. I. Standortslehre. Einführung in die Phytologie III/1. Ulmer, Stuttgart Wiens AJ (1985) Vertebrate responses to environmental patchiness in arid and semiarid ecosystems. In: Pickett STA, White PS (eds) The ecology of natural disturbance and patch dynamics. Academic Press, New York Yair A (1983) Hillslope hydrology, water harvesting and areal distribution of some ancient agricultural systems in the northern Negev desert. J Arid Environ 6:283–301 Yair A (1990) Runoff generation in a sandy area; the Nizzana sands, Western Negev, Israel. Earth Surface Processes Landforms 15:597–609 Yair A (1994) The ambiguous impact of climate change at a desert fringe: Northern Negev, Israel. In: Millington AC, Pye K (eds) Environmental change in drylands. Wiley, Chichester, pp 199–226 Yair A (1999) Spatial variability in the runoff generated in small arid watersheds: implications for water harvesting. In: Hoekstra TM, Shachak M (eds) Arid lands management: toward ecological sustainability. University of Illinois Press, Chicago, IL, pp 212–222 Yair A (2001) Effects of biological soil crusts on water redistribution in the Negev Desert, Israel: a case study in longitudinal dunes. In: Belnap J, Lange OL (eds) Biological soil crusts: structure, function, and management. Springer, Berlin Heidelberg New York, pp 303–314 Yair A, Bryan RB (2000) Hydrological response of desert margins to climatic change: the effect of changing surface properties. In: McLaren SJ, Kniveton DR (eds) Linking climate change to land surface change. Kluwer, London, pp 49–63 Yair A, Danin A (1980) Spatial variations in vegetation as related to the soil moisture regime over an arid limestone hillside, northern Negev, Israel. Oecologia 47:83–88 Yair A, Shachak M (1987) Studies in watershed ecology of an arid area. In: Berkofsky L, Wurtele G (eds) Progress in desert research. Rowman and Littlefield, Totowa, NJ, pp 45–93
Chapter 30
General Conclusions – Sand Dune Deserts, Desertification, Rehabilitation and Conservation S.-W. Breckle, A. Yair, and M. Veste
30.1
Sand Deserts and Sand Dunes
Moving sand dunes represent a natural phenomenon in most arid and hyper-arid sand deserts, such as the Sahara, Namib, Taklamakan and Rub’al Khali. The preconditions for large sand dunes or even extensive “sand seas” are, on the one hand, the geological situation with a large source of sand provided by the weathering of parent rocks and, on the other hand, the climate, which is normally very arid and exhibiting typically strong wind systems. These dune systems – e.g. in the Gobi, the Rub-al-Khali and the Namib – are typical sand deserts. The water regime of these sand deserts is rather favourable in comparison with that of adjacent rock, gravel or clay deserts. The biomass resulting from 1 mm of rainfall on sandy soils is 2.5 times higher than that produced on fine-texture soils (Le Houérou 1986). This can always be seen in some specific stands of plants, mostly in the stable dune valleys where eventually sometimes even water can be found. The mobility of these dune systems is controlled by the specific wind regime, which may cause different types of dune morphology and dune types (Bagnold 1941; Besler 1980; Lancaster 1982; Tsoar 1984; Tsoar and Møller 1986; Cooke et al. 1993). There are also less arid deserts. These are found along desert margins or in the form of semi-deserts, where fixed sand dune systems start to become mobile for various reasons (Wang et al. 2006), mainly by overgrazing and trampling, together with firewood collection. In geological timescales, climate change can be a trigger for the reactivation of stable sand dunes (Lancaster 1987; Littmann 1988). The Nizzana dunes are an example of dunes at a desert margin, more or less stable during the last centuries. They are very small in comparison with other sand dune areas but nevertheless a good example of where, on the one hand, the dynamics of desert ecosystems (as shown in previous chapters) and, on the other hand, the vulnerability of these systems and the problem of desertification as well as the necessary measures of rehabilitation can be studied. Thus, some comparisons with other, selected sand dune ecosystems, with their specific dynamics and threats, can be made but also some general conclusions can be drawn.
S.-W. Breckle et al. (eds.) Arid Dune Ecosystems. Ecological Studies 200, © Springer-Verlag Berlin Heidelberg 2008
441
442
30.2
S.-W. Breckle et al.
Desertification – the Degradation of Sandy Desert Ecosystems and Threat to Adjacent Areas
Desertification is a major threat for the world’s drylands. Especially desert margins and semi-arid areas are affected by land degradation as a consequence of inappropriate land use and over-exploration of natural resources. The United Nations Conventions to Combat Desertification (UNCCD) defines desertification as “the degradation of the land in arid, semi-arid and dry sub-humid areas caused by various factors, including climatic changes and human activities”. Dryland ecosystems and especially desert margins are very vulnerable to over-exploitation and inappropriate land-use practices and climatic changes. Various aspects are involved in the initialisation and acceleration of desertification (Mainguet 1999; Breckle et al. 2001; Müller et al. 2006), e.g. overgrazing, deforestation, lumbering, overuse of water resources, salinisation, mining, erosion, sand movement, and climatic drought. Land degradation leads to a dramatic decrease in soil fertility, water availability, net primary production, plant cover and biodiversity. The ongoing climate change and climate fluctuations will accelerate these desertification processes. In many affected countries, the remobilisation of sand dunes and enhanced sand mobility are a major threat for villages, buildings, roads, railways and technical infrastructures. The Sinai-Negev sand field is a perfect example of how grazing and land use has destroyed the vegetation cover to a great extent (see Chap. 6, this volume). Already in the 1970s, the desertification
Fig. 30.1 Sand dunes near Nizzana are converted into agricultural fields for the production of flowers and vegetables in greenhouses
30 General Conclusions – Sand Dune Deserts, Desertification
443
problem of the Sinai-Negev sand dunes was indirectly recognized on satellite images (Otterman et al. 1975). Nowadays, agricultural cultivation along the northern and southern margins of the sand field is increasing, and large areas are covered by greenhouses for the production of flowers and vegetables (Fig. 30.1). The sandy soils of the Negev are very suitable for drip irrigation (Tsoar 1990). The increasing water demand for irrigation is affecting the coastal aquifer, and brackish water is moving into the wells (Tsoar 1990).
30.3
Designing Shelterbelts
Stabilisation and rehabilitation of mobile sand dunes is a major task for combating desertification. A key challenge is the reduction of surface wind speed as the source of sand transportation. Chemical products have been often used to stick sand grains together and to prevent saltation (Veisov et al. 1999). Such methods are unacceptable from an environmental viewpoint. They alter or even damage ecosystem functioning. Mechanical and biological barriers are a useful tool to increase the surface roughness and to decrease wind speed. Biological methods for sand dune stabilisation are increasingly and successfully being used in many deserts (Mainguet 1999; Veisov et al. 1999; Gao et al. 2006). In the first stage of stabilisation, straw checkerboards are often used to reduce surface wind speed (Fig. 30.2). An efficient straw checkerboard of 10 to 20 cm in height and 1×1 m in size decreases sand flow by more than 99.5% up to wind speeds of 6 m s−1 (Qiu et al. 2004). In the second stage, the development of natural or planted vegetation takes place. Only re-vegetation is a permanent solution for sand dune stabilisation. An efficient shelterbelt depends on the regional wind field and on average as well as maximal wind speeds. The spacing depends on wind velocity and wind direction. It is recommended to space windbreaks at a distance of 5–25 times their height (Mainguet 1999). For the establishment of an effective shelterbelt, therefore, detailed information about the local wind field is needed but, unfortunately, there is commonly an important lack of knowledge in this respect. Species used for plantations have to be selected according to their growth, structure and water consumption. Soil water availability limits vegetation density and the biomass of perennials (see Chap. 26, this volume). Therefore, only local plants should be selected for shelterbelts. Indeed, plants from foreign regions have sometimes been introduced for the reclamation of sand dunes and as wind-breaking shelterbelts. Various Populus species are used, sometimes together with fruit trees, such as Morus. In fact, this is a rather old tradition in many parts of Central Asia (Fig. 30.3). In South Africa, Acacia species from Australia were taken to stabilise the sand dunes in the Cape Flats. These alien plants invaded other sensible ecosystems in the country and are nowadays a major threat for the indigenous biodiversity and ecosystem functioning in many regions of South Africa (Bromilow 2001). Also in China, several tree species originating from Europe, North America and Australia
444
S.-W. Breckle et al.
Fig. 30.2 Straw checkerboards decrease surface wind speed and stabilise the surface. The construction is very labour-intensive but is now used in many countries
Fig. 30.3 Populus trees are planted extensively in many parts of Central Asia, serving as shelterbelts and wind-breaking walls, here in Khoreshm (Uzbekistan)
30 General Conclusions – Sand Dune Deserts, Desertification
445
are used in reforestation projects (Gao et al. 2006). Many species were tested for the rehabilitation of the Loess Plateau with its extreme erosion problems (Ichizen et al. 2005). The effects of these introduced plants on biodiversity, ecosystems functioning and genetic diversity are still unknown.
30.4
Stabilisation of Sand Dunes in the Aralkum
The most conspicuous cases of desertification with visible effects in the wider vicinity can be observed within the Aralkum Syndrom (Agakhanjanz and Breckle 1993; Letolle and Mainguet 1996; Breckle 2003). The Aralkum Syndrom is a complex pattern of various kinds of desertification problems caused by increasing lack of water, mainly by overexploitation for excessive irrigation practices, and thus desiccation of lakes. Not only has the Aral Sea almost disappeared but also Lake Chad in Africa, Lobnor in Sinkiang, and the Dead Sea as well as the Great Salt Lake in the USA and the Lago Enriquillo in the Dominican Republic face strong water losses, and are becoming smaller and more saline. New salt flats develop, as well as new sand deserts. From the desiccated seafloor, the Aralkum, salt-dust clouds are transported hundreds of kilometres and are contaminating irrigation fields in neighbouring countries (Uzbekistan: Amudarya river plains, Kazakhstan: Syrdarya river plains, Turkmenistan, etc., many examples; see NASA 2006). The huge stretches of salt deserts with puffy salt crusts (about 65%) need an urgent action for rehabilitation, which is possible only by means of phytomelioration (Fig. 30.4) through extensive halophyte plantings of Haloxylon aphyllum or Halocnemum strobilaceum (Fig. 30.5; Meirman et al. 2001; Breckle 2003; Wucherer et al. 2005a, b). Dust storms are a common feature in arid lands but are increasingly disturbing cultivated lands and settlements (Littmann 2006; Gao et al. 2007). About 25% of the desiccated seafloor of the Aral Sea consists of sand deserts with newly developed aeolian activities. This area is threatening the adjacent regions by dust and sand storms (Youlin et al. 2001). Again, it needs a strong action of phytomelioration to accelerate the natural succession processes of invasion of psammophytes, e.g. Calligonum species, Haloxylon aphyllum, H. persicum, Tamarix species, in order to minimize sand storms (Fig. 30.6). The use of saplings of H. aphyllum in spring on sandy soils yields the best results. Nevertheless, even if saplings are taken and used in fall and on solonchak soils, there is still some success of rooting (Meirman et al. 2001; Wucherer and Breckle 2005). A strict management system for grazing to prevent further degradation of sandy areas around villages and newly established “afforestations” is necessary, in combination with establishing shelterbelts around the villages (Figs. 30.7, 30.8). This also has been demonstrated in the Nizzana area for the sand dunes and their grazing intensity (see Chap. 6, this volume). It has to be stressed that in the Aralkum, the sand desert is by far less dangerous for the adjacent regions than are the salt deserts (Wucherer and Breckle 2005).
446
S.-W. Breckle et al.
Fig. 30.4 One row of a phytomelioration trial with cuttings of Haloxylon aphyllum (Saxaul) on the dry seafloor of the Aral Sea 1 year after planting. In the background: an annual plant cover by Atriplex pratovii
Fig. 30.5 A 1-year-old sapling of a well-growing Haloxylon aphyllum, and creation of a sand dune in the lee
30 General Conclusions – Sand Dune Deserts, Desertification
447
Fig. 30.6 Various Calligonum species (Polygonaceae) occur in the sand deserts of the Aralkum. They are subject to sand covering or to sand deflation. The extensive root system is then visible
Fig. 30.7 Villages around the former coastline of the Aral Sea are in a desolate condition because of frequent dust and sand storms
448
S.-W. Breckle et al.
Fig. 30.8 Creation of shelterbelts by planting resistant trees, e.g. Ulmus pumila around the village of Bugen
Salt-dust particles are small and can be transported for hundreds of kilometres. Sand grains are larger and are subject to saltation and short-distance transport only. Nevertheless, common sand storms (Bisqunaq) in the wider Aral Sea basin are also a threat for the health of local inhabitants and cause damage to equipment not only in the villages (Fig. 30.9). The productivity of the vegetation by natural succession is much stronger on sand dunes than in the salt desert. The relatively good water conditions in a sandy area (Yair 2001, and see Chaps. 18 and 29, this volume) under an arid climate are a good precondition for phytomelioration. Thousands of hectares are already “afforested” in several, mostly sandy parts of the new Aralkum in the south-eastern Uzbek sector (GTZ Project, Novizkiy, unpublished data). Conservation and sustainable use of biological diversity has became a key issue in Kazakhstan (Karibaeva et al. 1998). Barsa Kelmes Island has been a nature reserve since 1939 but today, open access to the east via the dried-out seafloor has had a major impact (TERRA 2007). The totally flat seafloor, with a mixture of salt crust desert and, in the older parts, of sandy desert has enabled the fauna to occupy a much larger area. By enlarging the size (16,795 ha) of the Barsa Kelmes nature reserve considerably (59,884 ha), and by demarcating two additional core areas (Kaskakulan, 109.942 ha; Syrdarya Delta, 11,244 ha), a good step forward in nature
30 General Conclusions – Sand Dune Deserts, Desertification
449
Fig. 30.9 Frequent sand and dust storms cause damages to equipment – here, a stranded ship
Fig. 30.10 The new sand desert in the Aralkum inhibits the establishment of a rich fauna, including reptiles – here, an Agama species
450
S.-W. Breckle et al.
Fig. 30.11 Open Artemisia semi-desert with old Haloxylon aphyllum shrubs and small trees on Barsa Kelmes Island, now declared a core area of nature conservation
conservation with future developmental goals has been accomplished (Ogar and Geldeev 2007). The declaration by the Kazakh Government aims to activate the natural dynamics and to conserve remnant fauna (Fig. 30.10) comprising various reptiles, and also gazelles, Saiga antelopes and wild asses, as well as remnants of the open Saxaul desert forests (mainly H. aphyllum) with a rich biodiversity on the former Barsa-Kelmes Island (Fig. 30.11).
30.5
Stabilisation of Sand Dunes in the Tengger Desert
In the last decades in the northern provinces of China, large shelterbelts have been established to reduce sand erosion (Sun and Fang 2001; Ximing et al. 2001; Li et al. 2004b; Gao et al. 2006; Veste et al. 2006). Protective forest systems (Zhao et al. 2005) have been established in semi-arid regions, such as in the Horqin sandy region in northeast China. Here, an annual rainfall of between 315 and 490 mm can promote the growth of trees on the sand dunes. This is a major difference with the Nizzana area. The southern parts of the Tengger Desert along the Yellow River have only 180 mm of annual precipitation. High sand movement is common and is endangering the roads and railway lines along the river.
30 General Conclusions – Sand Dune Deserts, Desertification
451
Fig. 30.12 Stabilised sand dunes in the Tengger (Shapotou) sandy region of Inner Mongolia, China
Therefore, in 1957 a sand dune protection system was initiated. Since then, the vegetation protective system in Shapatou is 16 km long and 500 m wide on the northern, and 200 m wide on the southern side of the railway (Li et al. 2004b). In the first phase of stabilisation, straw checkerboards are used. In the centre of the checkerboards, mainly seedlings of Artemisia ordosica, Hedysarum scoparium, Caragana korshinskii, Eragrostis poaeoides and Calligonum mongolicum were planted. Other plants, especially annuals, established spontaneously, as did biotic crusts (Li et al. 2002). The degree of vegetation coverage in experimental plots with A. ordosica and C. korshinskii depends on plant density and mixture. In monocultures of A. ordosica, the maximum observed vegetation cover was 18.6% and, for C. korshinskii, 19.6%, whereas in mixed stands of both species the cover was up to 25.8% (Fig. 30.12) (Li and Shi 2003). The reasons for the higher vegetation cover in the mixed stands are the different root systems of the species. In contrast to Artemisia, Caragana can use water from deeper layers. Between the checkerboards, a biological soil crust develops (Fig. 30.13). As we could show for the Nizzana dunes (see Chap. 29, this volume), mosses have a negative impact on deep-rooting shrubs also in sand dunes at the Shapatou research site (Fearnehough et al. 1998; Mitchell et al. 1998; Li et al. 2004a). The number of shrubs declined on stabilised plots where mosses established (Fig. 30.14). Despite the climatic differences between northern China and the Negev, these sand dune ecosystems have some similarities (Fig. 30.15); in both regions, effects of climate change may be a risk (He 2001). Vegetation cover and plant functional types as well as ecosystem processes are comparable.
452
S.-W. Breckle et al.
Fig. 30.13 Biological soil crusts developed between the straw checkerboards on stabilised sand dunes of the Tengger Desert at the Shapotou experimental site (Inner Mongolia, China)
Fig. 30.14 Top Sand dunes in the Haluza sands in the eastern extension of the Sinai-Negev sand field, and bottom stabilised sand dunes of the Tengger Desert (Inner Mongolia, China)
30 General Conclusions – Sand Dune Deserts, Desertification
Vegetation cover (%)
A
453
shrubs grass
16
annuals
12 8 4
Biological crust cover (%)
0
B
cyanobacterial crust
16
moss crust
12 8 4 0 mobile
12 years 29 years Surface age
37 years
Fig. 30.15 Vegetation cover with shrubs, grasses and other annuals on mobile and stabilised sand dunes at the Shapotou experimental site in the Tengger Desert, Inner Mongolia, China (A) and a cover of cyanobacterial crusts and mosses (B), the plots representing different surface ages (0, 12, 29, 37 years in 1996; after Mitchell et al. 1998)
30.6
Restoration of Sand Dunes in Southern Africa
A major challenge for ecological restoration aiming to repair ecosystems with respect to their health, integrity and self-sustainability exists for arid and semiarid areas, especially those of high biodiversity (Aronson et al. 1993). The Succulent Karoo and the Strandveld belong to the hotspots of biodiversity in arid regions, with a large number of endemic genera and species (Milton et al. 1997; Desmet and Cowling 1999a; Veste and Jürgens 2004). Sand dunes (Fig. 30.16) occur along the coastal belt of the Strandveld and up to 10 km inlands. The Strandveld receives 50–150 mm mean annual rainfall, the northern sand dunes between Port Nolloth and Alexanderbay 20–60 mm with rainfall predominantly in the winter months (Cowling et al. 1999). The vegetation consists mainly of succulent and non-succulent dwarf shrubs (Milton et al. 1997). Characteristic for the flora are the high numbers of Asteraceae, Mesembryanthemaceae, Crassulaceae, Eurphorbiaceae and other succulents. The conservation of this unique flora and fauna is a major task in recent South African environmental laws. Intensive livestock grazing leads to serious landscape degradation in the Strandveld. Furthermore, opencast mining for diamonds and other minerals is common in the sand dune areas of the Succulent Karoo and the southern Namib Desert (“Diamanten-Sperrgebiet”).
454
S.-W. Breckle et al.
Fig. 30.16 Succulent dwarf shrubs covering the sand dunes of the Succulent Karoo (photograph southeast of Alexanderbay, South Africa)
These mining activities have damaged vast areas of the sandveld succulent vegetation (Fig. 30.17), creating enormous dumping sites with overburden sands (Carrick and Krüger 2006). Huge-scale sand removal and relocation is characteristic for opencast mining. Mine pits produce unnaturally deep depressions. Strong southerly winds and intensive rain events are the main erosion forces in the Namaqualand lowlands. These extreme wind conditions, in combination with the low rainfall, prevent an autogenic recovery of the damaged sites (Carrick and Krüger 2006). Many of these disturbed sites are invaded by alien species. The consequences for the ecology of this unique biome have to date not been adequately assessed. Since 1998, the Environmental Conservation Act limits environmentally damaging activities and requires that developers include the costs for ecological rehabilitation into their operational budgets. Natural vegetation has to be restored after mine closure. Positive experiences with seed collection and propagation of woody species were recorded in reclamation projects in the semi-arid fynbos. Often, technical approaches (cultivating, fertilizing, reseeding, irrigation) and exotic species were used in strip-mining restoration (Halbich 2003). However, such technical procedures failed in areas with rainfall below 200 mm year−1 (Milton 2001). For the conservation of biodiversity, the goals of restoration have to go beyond the stabilisation of damaged ecosystems, and the reestablishment of ecosystem services such as the control of sand movement and dust. The reestablishment of ecosystem function and self-sustaining vegetation cover in these sandy ecosystems is very important for biodiversity conservation. General principles for restoration
30 General Conclusions – Sand Dune Deserts, Desertification
455
Fig. 30.17 Opencast diamond mining in the sand dunes between Alexanderbay and Port Nolloth damages the fragile desert vegetation of the sandveld
were adapted to the arid winter-rainfall areas (Table 30.1). After the failure with alien plants (cf. above), locally adapted indigenous plant species (e.g. Burke 2003; Anderson et al. 2004) have been successfully used for the restoration of sand dunes of the fynbos, Strandveld and southern Namibia. The re-vegetation of sand dunes can be based on a seed bank and seed dispersal. Therefore, different planting experiences were conducted to evaluate different planting designs for indigenous perennials (Burke 2003; Mahood 2003). Succulents were successfully relocated from pre-mined to post-mined areas. The reconstruction of vegetation structure is also important, and planting in multispecies clumps is recommended to increase survival under the environmental conditions prevailing in the Succulent Karoo (Blignaut and Milton 2005). The approaches used by South African ecologists for the restoration of these ecosystems are focused on the reestablishment of ecosystem processes and functions (nutrient cycling, dispersal, recruitment). Animals also play an important role in supporting various rehabilitation processes. Indeed, they are the major pollinators, and import seeds and organic matter into the ecosystems. Burrowing rodents and insects improve soil fertility and soil permeability, with a positive effect on re-vegetation (Desmet and Cowling 1999b). The South African experience emphasises also the importance of involving local communities in the planning, decision making and execution of successful projects (van Rooyen 1998; Milton 2001). The indigenous knowledge of local inhabitants can be useful to understand processes leading to desertification and also in selecting
456
S.-W. Breckle et al.
Table 30.1 General principles for ecological rehabilitation in the arid and winter rainfall regions of southern Africa (after Milton 2001) General principles 1. 2. 3. 4. 5. 6. 7. 8. 9.
Set clear and ecologically and economically feasible goals for rehabilitation Salvage living components of the ecosystem Use locally adapted, indigenous plant and animal species Retain and capture resources (organic matter, nutrients and water) Mimic the original uneven shape of the landscape Enable plants and animals to assist in the rehabilitation process Do not rely on traditional agricultural and horticultural approaches Keep good records of what works and what does not work Budget sufficient time and money for rehabilitation
appropriate techniques for resource management. New scientific hypotheses can be based on such information. This will also promote public acceptance of rehabilitation measures and changes in land-use management.
30.7
Conclusions
The establishment of sustainable rehabilitation measures in sandy drylands requires the understanding of the main ecological processes governing these ecosystems. The Nizzana case study, as an excellent example, can here provide useful information about these ecological processes and their interrelations with vegetation patterns. In this volume, various chapters on the geology, geo-ecology, geomorphology and land use, climate, flora, vegetation and fauna, as well as the formation of biological crusts give a sound basic knowledge of the Nizzana Sands system. The application of various remote-sensing techniques and many other methods to reveal ecosystem patterns and processes, such as topo-climate and micro-climate, including evaporation, transpiration and dew effects, to reveal aeolian sand transport and atmospheric input of dust, and to understand salt dynamics in soil, nitrogen input pathways, and runoff and erosion processes on dune surfaces and in dunes is a precondition in unravelling successional stages in the recovery processes of dunes on the crust and on the vegetation level, as well as interrelations between vascular plants and the crust and its spatial and temporal dynamics. This can also be seen in assessing plant water status, biomass distribution, and the vegetation patterning of perennials and annuals, as well as their heterogeneity and competition. Dynamic processes have to be seen in various temporal and spatial scales; extrapolation of observations and measurements along environmental gradients is an additional tool for a synthesis, and provides hints for sustainable management measures in such sensitive arid regions under climate change conditions. Thus, sand dune ecology has a strong importance for various applications in drylands.
30 General Conclusions – Sand Dune Deserts, Desertification
457
References Agakhanjanz OE, Breckle S-W (1993) Umweltsituation in der ehemaligen Sowjetunion. Naturwiss Rundsch 47:99–106 Anderson P, Hoffman MT, Holmes PM (2004) The potential of Cephalophyllum inaequale (L. Bolus) for the restoration on degraded arid landscapes in Namaqualand, South Africa. Restoration Ecol 12:343–351 Aronson J, Floret C, Le Floc’h E, Ovalle C, Pontainer R (1993) Restoration and rehabilitation of degraded ecosystems in arid and semi-arid lands. I. A view from the south. Restoration Ecol 1(1):8–17 Bagnold RA (1941) The physics of blown sand and desert dunes. Methuen, London Besler H (1980) Die Dünen-Namib: Entstehung und Dynamik eines Ergs. Stuttgarter Geographische Studien 96 Blignaut A, Milton SJ (2005) Effects of multispecies clumping on survival of three succulent plant species translocated onto mine spoil in the Succulent Karoo Desert, South Africa. Restoration Ecol 13(1):15–19 Breckle S-W (2003) Rehabilitation of the Aral Sea environment, Kazakhstan. In: Proc Int Worksh Combating Desertification – Rehabilitation of Degraded Drylands and Biosphere Reserves, Aleppo, May 2002. UNESCO-MAB Dryland Series no 2, pp 47–57 Breckle SW, Veste M, Wucherer W (eds) (2001) Sustainable land-use in deserts. Springer, Berlin Heidelberg New York Bromilow C (2001) Problem plants of South Africa. Briza, Pretoria Burke A (2003) Practical measures in arid land restoration after mining – a review for the southern Namib. S African J Sci 99:413–41 Carrick PJ, Krüger R (2006) Restoring degraded landscapes in lowland Namaqualand: lessons from the mining experience and from regional ecological dynamic. J Arid Environ DOI 10.1016/j.jaridenv.2006.08.006 Cooke R, Warren A, Goudie A (1993) Desert geomorphology. University College of London Press, London Cowling RM, Esler KJ, Rundel PW (1999) Namaqualand, South Africa – an overview of a unique winter-rainfall desert ecosystem. Plant Ecol 142:3–21 Desmet PG, Cowling RM (1999a) Biodiversity, habitat and range-size aspects of a flora from a winter-rainfall desert in north-western Namaqualand, South Africa. Plant Ecol 142:23–33 Desmet PG, Cowling RM (1999b) Patch creation by fossorial rodents: a key process in the revegetation of phytotoxic arid soils. J Arid Environ 43:35–46 Fearnehough W, Fullen MA, Mitchell DJ, Trueman IC, Zhang J (1998) Aeolian deposition and its effect on soil and vegetation changes on stabilized desert dunes in northern China. Geomorphology 23:171–182 Gao J, Veste M, Sun B, Beyschlag W (eds) (2006) Rehabilitation and stability of ecosystems in arid and semi-arid areas. Science Press, Beijing Gao J, Nishio T, Ichizen N, Veste M, Breckle S-W (2007) Desertification and rehabilitation in China. Japan. J Arid Land Stud 17:101–112 Halbich TFJ (2003) Mine rehabilitation in the arid Succulent Karoo vegetation zone on the South African west coast, Namakwa Sands – case study. In: Heavy Minerals. South African Institute of Mining and Metallurgy, Pretoria, pp 113–118 He Q (2001) Climate change and sandy desertification combating in China. J Beijing Forestry Univ 23:61–65 Ichizen N, Takahashi H, Nishio T, Liu G, Li D, Huang J (2005) Impacts of switchgrass (Panicum virgatum L.) planting on soil erosion in the hills of loess plateau in China. Weed Biol Manage 5:31–34 Karibaeva KN, Baitulin IO, Glukhovtsev IV, Kovshar AF, Kurochkina LY, Nesterenko VA, Safonov GA, Shabanova LV, Aralbaev NK (1998) National report on conservation and balanced use of biological diversity of the Republic of Kazakhstan. Ministry of Ecology and Natural Resources, Almaty, Kazakhstan
458
S.-W. Breckle et al.
Lancaster N (1982) Linear dunes. Progr Phys Geogr 6:445–504 Lancaster N (1987) Formation and reactivation of dunes in the Kalahari: palaeoclimatic implications. Palaeoecol Afr 18:103–110 Le Houérou HN (1986) The desert and arid zones of Northern Africa. In: Evenari M, Noy-Meir I, Goodall DW (eds) Hot deserts and arid shrublands. Elsevier, Amsterdam, pp 101–147 Letolle R, Mainguet M (1996) Der Aralsee – eine ökologische Katastrophe. Springer, Berlin Heidelberg New York Li Z, Shi W (2003) Analysis of water consumption of artificial sand-fixing plants and simulation of ecological moisture fitness. Israel J Plant Sci 51(2):101–108 Li XR, Wang XP, Zhang JG (2002) Microbiotic soil crust and its effect on vegetation and habitat on artificially stabilized desert dunes in Tengger Desert, North China. Bull Fertil Soils 35:147–154 Li XR, Chen YW, Yang LW (2004a) Cryptogam diversity and formation of soil crusts in temperate desert. Ann Arid Zone 43:335–353 Li XR, Xiao H-L, Zhang JG, Wang XP (2004b) Long-term ecosystems effects of sand-binding vegetation in the Tengger desert, Northern China. Restoration Ecol 12(3):376–390 Littmann T (1988) Jungquartäre Ökosystemveränderungen und Klimaschwankungen in den Trockengebieten Amerikas und Afrikas. Bochumer Geographische Arbeiten 49 Littmann T (2006) Dust storms in Asia. Geogr Rundsch Int Edn 2:8–12 Mahood K (2003) Strip mining rehabilitation by translocation in arid coastal Namaqualand, South Africa. MSc Thesis, University of Stellenbosch, Stellenbosch Mainguet M (1999) Aridity. Droughts and human development. Springer, Berlin Heidelberg New York Meirman G, Dimeyeva L, Dzhamantikov K, Wucherer W, Breckle S-W (2001) Phytoreclamation on the dry sea floor of the Aral Sea. In: Breckle SW, Veste M, Wucherer W (eds) Sustainable land-use in deserts. Springer, Berlin Heidelberg New York, pp 318–322 Milton SJ (2001) Rethinking ecological rehabilitation in arid and winter rainfall regions of southern Africa. S African J Sci 97:1–2 Milton SJ, Yeaton RI, Dean WRJ, Vlok JHJ (1997) Succulent Karoo. In: Cowling RM, Richardson DM, Pierce SM (eds) Vegetation of southern Africa. Cambridge University Press, Cambridge, pp 131–161 Mitchell DJ, Fullen MA, Trueman IC, Fearnehough W (1998) Sustainability of reclaimed desertificated land in Ningxia, China. J Arid Environ 39:239–251 Müller JV, Veste M, Wucherer W, Breckle S-W (2006) Desertifikation und ihre Bekämpfung – eine Herausforderung an die Wissenschaft. Naturwiss Rundsch 59(11):585–593 NASA (2006) NASA Earth Observatory. htpp://www.earthobservatory.nasa.gov Ogar NP, Geldeev B (2007) The project Development of Natural, Scientific, Technical and Economical Rationale for Territorial Extension of the State Wildlife Preservation Barsakelmes 2004/05 – GIS Terra, May 2007. Almaty, http://www.gis-terra.kz/eng/zapovednik_e.htm/ Otterman J, Waisel Y, Rosenberg E (1975) Western Negev and Sinai ecosystems: comparative study of vegetation, albedo, and temperatures. Agro-Ecosystems 2:47–59 Qiu GY, Lee IB, Shimizu H, Gao Y, Ding G (2004) Principles of sand dune fixation with straw checkerboard technology and its effects on the environment. J Arid Environ 56:449–464 Sun B, Fang T (2001) Desertification in China and its control. In: Breckle SW, Veste M, Wucherer W (eds) Sustainable land-use in deserts. Springer, Berlin Heidelberg New York, pp 418–426 TERRA (2007) Development of natural, scientific, technical and economical rationale for territorial extension of the State Wildlife Preservation Barsakelmes. http://www.gis-terra.kz/eng/ zapovednik_e.htm (viewed May 2007) Tsoar H (1984) The formation of seif dunes from barchans – a discussion. Zeitschr Geomorphol 28:99–104 Tsoar H (1990) The ecological background, deterioration and reclamation of desert dune sand. Agric Ecosystems Environ 33:147–170 Tsoar H, Møller JT (1986) The role of vegetation in the formation of linear sand dunes. In: Nickling WG (ed) Aeolian geomorphology. Allen and Unwin, Boston, MA, pp 75–95
30 General Conclusions – Sand Dune Deserts, Desertification
459
van Rooyen AF (1998) Combating desertification in the southern Kalahari: connecting science with community action in South Africa. J Arid Environ 39:285–297 Veisov SK, Cherednichenko VP, Svintsov IP (1999) The fixation of drifting sands. In: Babaev AG (ed) Desert problems and desertification in Central Asia – the researches of the Desert Institute. Springer, Berlin Heidelberg New York, pp 143–153 Veste M, Jürgens N (2004) Zonobiom III: die Karoo. In: Walter H, Breckle S-W (eds) Ökologie der Erde. Band 2, Spezielle Ökologie der tropischen und subtropischen Zonen. Elsevier, Amsterdam, pp 514–537 Veste M, Gao J, Sun B, Breckle SW (2006) The Green Great Wall – combating desertification in China. Geogr Rundsch Int Edn 2(3):14–20 Wang X, Eerdun H, Zhou Z, Liu X (2006) Significance of variations in the wind energy environment over the past 50 years with respect to dune activity and desertification in arid and semiarid northern China. Geomorphology DOI 10.1016/j.geomorph.2006.09.003 Wucherer W, Breckle S-W (2005) Desertifikationsbekämpfung und Sanierung der Salzwüsten am Aralsee. Sukzession und Phytomelioration, Naturschutz und nachhaltige Entwicklung. Bielefelder Ökologische Beiträge (BÖB), vol 19, pp 1–94 Wucherer W, Breckle S-W, Kaverin VS, Zhamantikov K, Ogar NP (2005a) Phytomeliorative Eigenschaften von Haloxylon aphyllum und Perspektiven der Anpflanzungen in der Region am Aralsee. In: Veste M, Wucherer W, Homeier J (eds) Ökologische Forschung im globalen Kontext. Festschrift zum 65. Geburtstag von Prof. Dr. S.-W. Breckle. Cuvillier, Göttingen, pp 109–128 Wucherer W, Veste M, Herrera Bonilla O, Breckle S-W (2005b) Halophytes as useful tools for rehabilitation of degraded lands and soil protection. In: Proc 1st Int Forum Ecological Construction of the Western Beijing, Mentougou District Government, Beijing Municipal Science & Technology Commission, Beijing, pp 87–94 (in English), pp 169–175 (in Chinese) Ximing Z, Li X, Zhang H (2001) The control of drift sand on the southern fringe of the Taklamakan desert – an example from the Cele oasis. In: Breckle SW, Veste M, Wucherer W (eds) Sustainable land-use in deserts. Springer, Berlin Heidelberg New York, pp 350–356 Yair A (2001) Sedimentary environments in the desiccated Aral Sea Floor: vegetation recovery and prospects for reclamation. In: Breckle SW, Veste M, Wucherer W (eds) Sustainable landuse in deserts. Springer, Berlin Heidelberg New York, pp 310–317 Youlin Y, Squires V, Qi L (2001) Global alarm: dust and sandstorms from the world’s drylands. UN, Bangkok Zhao H-L, Zhao XY, Zhou R-L, Zhang T-H, Drake S (2005) Desertification processes due to heavy grazing in sandy rangeland, Inner Mongolia. J Arid Environ 62:309–319
Subject Index
A Abiotic stress, 401, 402, 416, 418 Above–ground biomass, 101, 404–412, 415–417 Abundance, 93–98, 101, 102, 105–107, 115, 121–122, 128, 139, 143, 144, 164, 165, 167 Accumulation, 11, 74, 143, 165, 201, 202, 208, 225–229, 233, 236, 271, 286, 289, 307, 309, 353 Active crest, 20, 23, 253, 267 Active dunes, 31–33, 163, 247, 293 crest, 19, 20, 159, 163, 201–209 sands, 168 Adsorption, 227, 236 Advection, 59, 189 Advective effects, 189 Advective mixing, 199 Aeolian activity, 66, 67, 202, 203, 208, 222 dust, 74, 164 particles, 287, 299, 300 sand, 11, 26, 27, 203, 211–223, 231, 233, 237, 456 sand incursion, 26–30 sand transport, 211–223, 456 sediments, 45, 227 transport, 162, 211 Aerial photographs, 35, 65, 84, 161, 164–167, 169, 170, 201, 287, 382 Aerodynamic roughness, 211 Aerodynamic surface, 184, 212 Aerosols, 274, 276, 353 Aggregation, 137, 225, 228 Agricultural activity, 101, 333 Agricultural cultivation, 443 Agur, 25, 26, 28, 30, 31, 37 Air temperature, 53, 185, 193, 196, 197, 306, 310–312, 316
Airborne dust, 235 Airborne salts, 231 Airflow, 31, 50, 51, 59, 211, 212, 215, 216, 233, 313 Albedo, 83–85, 175–177 Algae, 21, 73, 149, 152–154, 166, 228, 248, 252, 271, 285, 286, 299 Alkaline, 3, 68, 75 Alluvial areas, 14 Alluvial belt, 9 Ammonia, 321, 330, 331, 333 Anabasis articulata community, 116 Animals, 125, 126, 130, 137, 144, 211, 213, 236, 455, 456 Annual course, 61, 176, 193, 196, 316 plant populations, 385–398 plants, 6, 79, 106, 107, 116, 126, 144, 232, 286, 339, 386, 387, 394, 395, 397, 401–419 plant species, 100, 101, 106, 107, 118, 387, 389–391, 393 variation, 391, 394, 398 vegetation, 74, 121, 122, 213, 299, 386 Ant lion, 130, 139 Anthropogenic pressure, 37 Anticyclone, 50, 56, 61 Ants, 125, 128, 129, 137, 142–145, 305 Arabian desert, 2, 49 Arable land, 68, 73, 298 Aral sea, 445–448 Aralkum, 3, 354, 445, 447–449 Aralkum Syndrom, 445 Arenosol, 67–71, 73–77, 108, 113–116, 121, 226, 228–234, 237 Arid climates, 33, 49, 61, 75, 188, 227, 353, 382 Arthropod groups, 142 Association, 105, 120, 197, 237 461
462 Atacama, 3, 49, 305, 307 Atmospheric deposition, 225, 271–276, 283, 319, 321, 333 Atmospheric input, 271–283, 329, 456 Atmospheric nitrogen, 271, 319, 320, 326, 329, 331 and deposition, 319, 320, 329–331 Atmospheric processes, 282 Australian deserts, 2 Avdat, 9, 54, 196, 308, 312, 382 Azores High, 50, 56
B Background signal, 166, 170 Barchans, 39 Barotropic waves, 50, 51 Barsa Kelmes, 448, 450 Bedouin, 5, 35, 41, 80, 82–84, 86, 87, 101, 287 Besor, 10, 25, 26 Between–habitat, 121, 391, 393, 418 Between–habitat differences, 121, 391, 393 Between–year differences, 390–392 Bidirectional wind regimes, 37, 38 Biodiversity, 5, 126, 442, 443, 445, 450, 453, 454 Biodiversity conservation, 454 Biogenic crust(s), 87, 157, 175, 178, 182, 287 Biological crust(s), 1, 6, 73, 74, 149–154, 166–168, 228, 229, 233, 235–237, 239, 241, 244, 246 Biological fixation, 319–334 Biological soil crusts, 149, 154, 159, 183, 305–316, 321, 324, 325 Biological topsoil crust(s), 1, 4, 5, 17, 20, 24, 239, 243, 253, 257, 267, 287, 299, 430, 439 Biomass, 62, 101, 159, 177, 182, 246, 247, 333, 358, 377–379 Biomass index, 379–381 Biomass production, 228, 229, 271, 327, 328, 383 Bio–physical border, 25 Biotic interactions, 144, 145, 414 Bioturbation, 144, 229, 236, 237 Bisqunaq, 448 Black sea troughs, 58 Blowouts, 20, 202, 248, 286 Border, 4, 25, 35, 37, 45, 66, 82–87, 101, 114, 117, 132, 157, 158, 165, 170, 171 Borderline, 17, 82, 86 Boundary layer, 60, 62, 186, 188, 312
Subject Index Bowen ratio, 188 Braided forms, 35, 38, pattern, 35, 39, 40, 42 Brownification, 225, 227 Bulk density, 65, 67, 71 Bulk deposition, 272, 274, 277–280, 282 Bulk samplers, 178, 279 Buried channels, 42–45 Buried drainage systems, 42–46 Burrowing, 144, 145, 236, 237, 455 Burrows, 115, 130, 132, 138 Burrows structure, 130 Bush encroachment, 427
C Calcisol(s), 67, 68, 73–77, 114, 116, 228, 229, 232 Camouflage, 128 Capillary movement, 353 Carbonate(s), 3, 65, 68, 73, 74, 114, 163, 164, 225, 226, 230, 232, 235, 272, 274, 276, 277, 280, 282, 283 Carbonate accumulation, 225 Carbonatization, 225 Carrying capacity, 83 Catena, 1, 67–69, 71, 359 Catenary position(s), 106, 122, 159, 170 Cation exchange capacity, 67, 75, 76 Cementation, 228 Chamaephyte(s), 94, 99, 101 Channel(s), 10–12, 15, 17, 19, 22, 23, 43, 45, 165, 262, 267 Checkerboards, 443, 444, 451, 452 Chenopodiaceae, 101, 107, 114, 225, 321, 354, 355, 358, 359, 373, 379 Chihuahua, 2 China, 2, 3, 189, 443, 450–453, Chloride(s), 231, 272, 277, 279, 282, 354–356, 361, 362 Chlorophyll, 21, 22, 150, 160–165, 169, 171, 247, 248, 287, 306, 313, 315, 327 Chlorophyll-a, 21, 22, 150, 161, 164, 171, 247, 248 Chlorophyll fluorescence, 306, 313, 315 Chorotype(s), 93–99, 113 Classification, 53, 56, 67, 93, 114, 120, 121, 151, 160 Clay, 1, 20–23, 27, 30, 33, 46, 68–74, 81, 149, 161, 162, 164, 166–170, 228, 229, 231, 242 Clay layers, 23 Clearing, 291, 296, 298
Subject Index Climate(ic), 5, 6, 27, 31, 33, 38, 42, 49, 50, 53, 56 change, 6, 425–439, 441, 442, 451, 456 fluctuations, 15, 442 gradient, 56, 382, 383, 430 phases, 14, 15, 426, 438 variables, 425, 426, 438 Clogging, 243, 247, 252, 257, 286 Cloud clusters, 58 cover, 53, 186 C/N ratio, 229, 235, 236 CO2 exchange, 315, 368, 369 Coalescence of two dunes, 34 Coastal desert, 307 Coastal plain, 25, 27, 49, 100, 272, 273, 282, 329, 333 Cohesiveness, 33, 80 Colorado Plateau, 3 Community patterns, 402 Compartmentation, 355 Competition, 198, 348, 386, 401–403, 405, 414, 416–419, 456 Competitive effects, 347, 385, 402, 404, 414, 416–419 interactions, 401, 402, 416, 418, 419 Condensation, 184–187, 189–191, 193, 195, 196, 312, 313 Conservation, 102, 441, 448, 450, 453, 454 Constancy, 32, 107, 109, 115–117 Constancy table, 107, 109 Consumption, 144, 443 Continental deserts, 49 Convection, 195 Coppice dunes, 22, 36, 212 Cornulaca monacantha community, 95, 101, 102, 107–109, 112, 114–117, 120, 121, 339, 354, 356, 362, 363, 378, 379, 403 Cover, 1, 5, 6, 13, 14, 17, 20–23, 25, 28, 32, 37, 53, 68, 77, 79, 81, 86 Cracking, 228 Crest, 11, 12, 19, 20, 23, 24, 33, 35–37, 41, 42, 73 Crest line, 33, 36, 37, 41, 203, 206, 208 Cross–border habitat, 83 Crosswinds, 36 Crust activity, 306, 311, 313, 314, 333 cover, 5, 6, 21, 23, 79, 123, 244, 246, 247, 285, 333, 428, 432, 437, 453 effects, 338, 347, 348 formation, 163, 225, 228, 229 recovery, 286, 291, 299, 300
463 regeneration, 292, 300 types, 122, 149–151, 159, 168, 170, 324, 325, 328, 339, 340, 343, 345, 346, 348, 438 Crusted area, 17, 21, 22, 244, 259–261, 286, 288, 349, 436 surface, 164, 221, 222, 261, 293, 297, 349 Cryptogamic crusts, 149, 178, 332 Cumulus clouds, 58, 59 Cyanobacteria(ae/al), 1, 21, 149–151, 153, 154, 161, 164, 166, 168, 169, 228, 235, 294, 299 crusts, 22, 24, 295, 325, 327, 331, 332, 453 sheath, 161, 257 Cyano–lichens, 23 Cycling of nutrients, 1, 5, 285, 455 Cyclogenesis, 50 Cyclogenetic activity, 56 Cyclones, 51, 56, 58 Cyclonic fronts, 49, 61 systems, 56, 61 Cyprus Lows, 50
D Dead shrubs, 432 Decomposition, 229, 235, 336 Deficiency, 75 Deflation, 211, 212, 220, 271, 286, 447 Defunct wadis, 42 Degradation, 158, 165, 171, 213, 442, 445, 453 Demography, 6, 401, 403, 414, 416–419 parameters, 414, 415, 418 processes, 386, 415 responses, 386, 402, 404, 410, 414, 417–419 success, 394 variables, 388 variation, 401, 416, 419 Density(es), 21, 31, 61, 65, 67, 71, 79, 82, 86, 101, 107, 113, 114, 117, 118, 120, 121, 130, 137 Density manipulation, 417 Deposition, 1, 5, 10–12, 20–23, 27, 37, 38, 53, 74, 162, 185, 202 Depositional sequence, 11 Desert annuals, 385, 386, 394, 395 dunes, 80, 251, 402, 413, 414 fringe, 171, 298, 425, 426 margins, 49, 333, 441, 442 plant communities, 105, 121 types, 1, 382 vegetation, 105, 120, 455
464 Desertification, 6, 87, 158, 286, 438, 441–443, 445, 455 Desiccation, 101, 179, 228, 316, 445 Dew fall, 164 formation, 186–188, 305–316 Dew point, 179, 185–187, 191–193, 195, 305, 306, 309–316, 379 Dewpoint temperature, 179, 185, 186, 193, 311–315, 379 Dew–rise, 306 Diachronic analysis, 86 Diamanten-Sperrgebiet, 453 Diamond mining, 455 Diaspores, 338, 346–348 Digging, 126, 133, 136, 139, 143, 144 Digging activity, 144 Digital aerial photographs, 166, 167 Distribution patterns, 106, 107, 117, 121, 122, 233, 386, 414, 415, 419 Disturbances, 6, 40, 56, 131, 159, 189, 190, 222, 274, 286–288 Disturbance gradient, 5 Diversity, 24, 107, 108, 113, 115–117, 122, 125, 126, 128 Domain–specific unmixing, 165, 166, 170 Dominance, 98, 100, 105–108, 113, 120, 121, 168, 434 Downward flux, 185, 193, 314 Drainage basin, 13 network, 11–13, 15, 45 patterns, 42, 43, 46 system, 9, 12, 42–44, 353 Drained area, 259, 260 Drift potential, 31–33, 293 Droughts cycles, 54 period, 56, 368, years, 50, 55, 61, 62, 158, 391, 394 Dry deposition, 233, 235, 237, 274, 280 land, 425, 426 season, 61, 193, 355, 358, 370, 373, 391, 395 Dune activity, 42 base, 20, 21, 106, 114, 120, 122, 150, 151, 154 crest, 12, 19, 20, 23, 37, 73, 79, 100, 159, 163, 167 field, 49, 56–62, 154, 159, 166, 179–182, 248, 252, 271 height, 178, 201, 206–209, 276
Subject Index morphology, 30, 39, 182, 441 ridge, 19, 22, 73, 106, 126, 127, 140, 201, 202, 208 slope, 5, 19–21, 67, 71, 73, 77, 100, 101 stabilization, 33, 81, 122 tops, 5, 77, 101, 106, 108, 114, 121, 122 valleys, 73, 136, 225, 441 volume, 207 Dust deposition, 74, 235, 274, 276, 282 retention, 279 storms, 51, 53, 56, 180, 274, 276, 277, 279, 282, 283, 333, 445, 449
E Echinops philistaeus community, 113, 114, 120 Echiochilon fruticosum–Thymelaea hirsuta community, 115, 213 Ecological rehabilitation, 454, 456 Ecosystem engineering, 125, 130–136, 144, 145 Ecotopes, 182, 337, 338, 341–343, 354, 377, 379, 428, 431 Eddy correlation, 184, 186, 189, 195, 196, 198 Egyptian border, 4, 86, 101, 117, 165, 170, 278, 427 El Nin~o, 55 Electrical conductivity, 67, 72, 226, 262 Element groups, 276 Element(al) analysis, 272 deposition, 273, 280, 283 mass, 272 Emergence probabilities, 396 rates, 393, 395–397 Encrusted area, 259–261, 349 Endemic, 100, 125, 453 Energy balance, 177, 179, 184, 188 budget, 175–177, 187, 188, 193, 306 ENSO, 55, 62 Eolian activity, 246, 248 Erg, 1, 2, 4, 9, 17 Erosion, 11, 37, 38, 73, 80–82, 116, 158, 162, 168, 202 Establishment, 1, 5, 24, 101, 122, 222, 248, 271, 286, 287, 299 Evaluation, 196, 197, 237, 300, 381 Evaporation, 1, 6, 33, 49, 75, 81, 102, 116, 123, 159 Evaporative demand, 367
Subject Index Evapotranspiration, 49, 59, 81, 178, 183–198, 379 Evapotranspiration models, 184–188 Evenness, 107, 108, 113, 122 Exploitation, 87, 442
F Fluvisols, 67, 68, 70, 71, 73, 75–77, 228, 232 Faecal pellets, 130 Fauna(al), 11, 125, 127, 128, 136, 144, 145, 305, 448–450, 453, 456 Fecundity, 340, 343, 345–348, 412 Feedback, 161, 162, 348, 354, 358 Fertile islands, 333, 334, 385 Field capacity, 67, 80, 262, 382, 434, 435 Flora, 79, 81, 82, 93–103, 105, 114, 154, 453, 456 Flow, 11, 13, 21, 22, 26, 31, 37, 50, 51, 59, 74 Fluorescence, 306, 311–315 Fluvial sediments, 11, 19, 46 Fog droplets, 306–308 interception, 307 Food web, 144, 145 Friction velocity, 181, 186 Fruit number, 405, 408, 409, 411–413, 415–417 production, 404, 405, 410, 412, 413 weight, 404, 405, 408–411, 413, 415, 416 Fynbos, 454, 455
G Gas exchange, 197, 367–374 Gaza, 25, 82 Geo–ecological gradient, 57, 316, 320, 329, 380, 382 units, 17, 19, 23, 377, 431 Geological history, 9–13 Geophyte, 94, 98, 99 Germination, 1, 5, 74, 75, 79, 212, 222, 252, 285, 286, 300, 337, 338, 346, 347, 386, 388, 389, 391 Germination rates, 388 Gevulot, 4, 54, 272, 329 Global warming, 437 Gobi, 2, 441 Goethite, 227 Gradients, 184, 186, 189, 193, 195, 199, 367, 382, 385, 394, 404, 413, 414, 417–419, 456
465 Grain–size analysis(es), 160, 164, 168, 169 variability, 170 Graminoid perennials, 108, 113 Grazing, 5, 25, 34, 35, 81–83, 86, 87, 209, 211, 213 Great Basin, 2, 332 Groundwater recharge, 262, 266 Guttation, 306
H Habitat conditions, 385, 394, 401, 405, 408, 412, 414, 415, 417–419 heterogeneity, 401, 414, 417 type(s), 101, 106, 117, 121, 340, 385–389, 391–393, 404–409, 413, 415, 417, 428, 430 Hallamish, 17, 19, 23, 26–28, 31, 32, 34, 35, 38, 39, 41, 42, 46, 80–84, 87, 101, 102, 248, 267 Hallamish–Shunra, 17 Halophytes, 225, 354, 355, 373 Haluza, 25, 28, 39, 40, 44, 45, 73, 150, 151, 154, 306–308, 316, 321, 323–325, 328 Haluza–Agur, 9, 17, 19, 20, 23, 25, 26, 30, 31, 39, 45, 84, 86, 87 Hamada, 1 Har Keren, 25, 26, 30, 45 Hard Crust, 339, 341–343, 346, 354, 356, 357, 361 Haredin dunefield, 45 Harvester ants, 129, 144 Harvesting, 83 Hemicryptophyte, 94, 98–99 Herbivorous animals, 144 Herbivory, 401 High–magnitude events, 219, 220 Hillocks, 108, 113, 115, 116 Holocene, 11, 12, 14, 15, 27 Hovav Plateau, 279, 282 Human activity, 4, 5, 42, 287, 425, 427 disturbance, 40, 299 impact, 100, 116, 171, 228 interference, 81 pressure, 37, 41 Human–induced pressure, 25 Humidity, 60–62, 176, 184–187, 189–193, 195, 197, 199, 306, 307, 310–314 Humus accumulation, 225, 229–231 Hydration, 161, 305, 316
466 Hydraulic conductivity, 251, 373 lift, 373 Hydrographs, 254, 256, 260 Hydrological behavior, 239, 246 regime, 15, 426 Hydrophobicity, 252 Hygroscopy, 185, 314 Hyper–arid areas, 49 climate, 49, 56
I Indigenous knowledge, 455 Induration, 74, 229 Infiltration, 1, 4, 13, 20, 21, 33, 68, 81, 83, 122, 144, 228, 236 Infiltration rate, 1, 4, 20, 21, 122, 239, 240, 243, 248, 252, 257, 259, 260, 266, 286, 338, 360, 379, 382, 425, 428, 433 Inorganic crusts, 229 ions, 357 solutes, 355 Intercepted fog, 307 Interception, 157, 237, 307, 308, 360, 385, 397 Interdune(s), 5, 11, 12, 19, 22, 23, 27, 39, 40, 46, 65, 68, 71, 73–75, 79 area(s), 23, 27, 39, 40, 46, 65, 71, 73–75, 79, 117, 151, 154, 163, 168, 227, 229 corridor(s), 11, 12, 22, 23, 100, 101, 106, 107, 115, 116, 159, 162, 201, 202, 212, 213 Interdune Playa, 68, 168 Inverse texture effect, 81, 82 Invertebrate diversity, 126, 139 morphospecies, 136 survey, 125, 126, 128 Ion accumulation, 225, 355 analysis, 277, 354, 379 contents, 354, 355, 359, 360, 363 patterns, 354, 355, 373 relations, 353, 355, 361 Iranian deserts, 3 Irano–Turanian, 94, 99, 113 Iron oxide(s), 27, 30, 67, 68, 73, 227, 228 Isotopic effects, 320 shift, 322, 323 Israeli–Egypt border, 157, 171
Subject Index K K+/Na+ relation, 357 Kalahari, 2, 33, 202, 211 Karoo, 305, 453–455 Keren, 25, 26, 30, 45 Khamsin, 51, 53, 190, 191, 195, 196, 277, 278 Khamsinic depressions, 51, 190, 195, 277 Khamsinic intrusions, 278 Kyzyl–Kum, 3
L Labilization, 191 Land cover, 451, 453, 454 degradation, 445, 453 management, 185 use, 6, 25, 79–87 Landsat image, 42, 44, 45, 84, 85, 157 satellite, 157 Thematic Mapper (TM), 28, 43, 161, 165 Latent heat, 177, 178, 186–189, 191 Lateral Water Flow, 262, 266 Lavan, 9–14 Leaching, 74, 80, 231–233, 237, 347, 353 Leaf gas exchange, 367–374 Leaf washout, 280, 283 water potential, 357, 367, 370, 371, 373 wetness, 178, 179, 186, 187, 308, 311–313, 315 Lichen activity, 311–313, 316 Lichens, 23, 24, 73, 114, 149–151, 154, 183, 228, 252, 285–287, 294 Life forms, 93, 94, 105, 113, 438 Linear Dunes, 17, 33–35, 39, 41, 42, 73, 79, 81, 178, 201, 226, 231, 232, 234, 237, 379, 381 Linear dune ridges, 232 Litter accumulation, 271 Livestock, 82, 87, 286, 453 Living shrubs, 432 Local–scale gradients, 385, 394, 413 Lodgement, 338, 345, 347 Loess, 11, 26, 81, 282, 308, 426 deposits, 11 plateau, 445
M Manipulation experiments, 414 Mapping, 42, 43, 65, 157, 159, 164, 165 Marine–borne salts, 225 Mass germination, 395
Subject Index Meandering, 38, 45, 46 Mediterranean, 4, 10, 25, 32, 45, 49–51, 53, 56, 58, 59 Mediterranean oscillation, 50, 56, 61 Meso–scale model, 381, 382 Microbial crust(s), 285 Microclimate, 175–182, 197, 309, 333 Microclimatic boundary conditions, 178, 182, 306, 311, 312, 316 characteristics, 182 parameters, 197, 237, 306, 313 Microfauna, 127, 136 Microlysimeter(s), 306, 308–310 Microphytic crust, 79, 213, 222, 262, 285 Micro–scale model, 381, 382 Microsites, 122, 126, 145 Migration, 37, 83, 128, 131, 132, 144, 225, 232, 237 Migratory route, 139 Military tension, 82 Mineral component, 164, 287–289, 293, 299 Mineral crust(s), 74, 159, 161, 163, 164, 169, 171, 267 Mini–catchment, 347 Mobile crest, 240, 244 Mobile sand(s), 100–102, 108, 113, 114, 121, 122, 149, 158, 257, 338, 378, 379, 382, 430, 438, 443 Model validation, 196–198 Modelling, 175, 179, 278, 316, 330, 377–383 Modelling biomass, 379–383 Moisture, 1, 62, 65, 68, 79, 81, 161, 236, 251, 252, 261, 266, 271, 285, 288, 305–310
N Namaqualand, 454 Namib, 3, 202, 305, 306, 355, 441, 453 Namibia, 2, 3, 305, 307, 455 Nanophanerophytes, 113, 115 Natufian site, 27 Natural vegetation, 215, 217, 219, 221, 454 Near–surface flow, 221 Nebkhas, 22, 36, 108, 201, 203, 208, 212 Negative crust effects, 338, 348 Negev climate, 31, 53, 61 desert, 9, 11, 13, 25, 28, 32, 79, 81, 149, 150, 157, 161, 183, 248, 305, 306
467 Neighbour competition, 405, 414, 417–419 effect, 395, 396 removal, 403–409, 412–414, 416–418 Neighbour removal experiment, 403, 404 Neighbour removal treatments, 405, 407 Nematodes, 127, 128, 137, 143, 305 Neutron probe, 261, 262 N–fixing microbes, 229 Nile Delta, 25, 26 Nitrate, 277, 330 Nitrogen deposition, 319, 320, 329, 333 fixation, 1, 267, 271, 285, 320, 327, 328, 332–334 input, 6, 319, 320, 327, 329, 330, 333, 334, 456 15 N abundance, 321–324, 326, 327, 329, 331, 332 15 N abundance method, 320, 322, 327, 331, 334 N deposition, 327, 329, 331, 332, 334 15 N determination, 321 N fixation, 237, 319, 320, 322, 325 N2 fixation, 320, 322, 325–328, 331, 332 δ15N signature, 322, 323 δ15N values, 324, 326, 327, 331 Nizzana channel, 12, 17, 19, 22, 23, 262, 267 Nizzana Research Site, 1, 4, 17, 65, 93–123, 158, 159, 212, 222, 338, 346, 379, 397, 398, 427 Nizzana Research Station, 4, 5, 131, 266, 299 Nizzana watershed, 9 Noaea mucronata–Artemisia monosperma community, 114 Nocturnal dew, 312 Nocturnal Wetting, 311, 313, 314 Nomadic societies, 81 Number, 14, 61, 79, 83, 86, 94, 99, 100 Nurse plants, 385 Nurse–plant effects, 398 Nutrient, 1, 5, 75, 80, 82, 122, 143–145, 154, 158, 161, 164, 228, 267, 271 Nutrient availability, 122, 144, 271, 349 Nutrient Elements, 80, 271
O Oblique winds, 38 Off–road vehicles, 81 Organic matter, 20, 21, 23, 67, 74, 75, 143, 158, 161, 162, 228, 229, 235, 271, 282, 297, 403, 429, 434–436, 455, 456
468 Osmotic adaptation, 357, 361 Osmotic gradient, 358 Overgrazing, 286, 287, 441, 442 Oxygen, 10, 74, 75
P Palatable species, 82 Particle size distribution, 289, 429 Pasture Management, 82 Patagonia, 2 Patchiness, 145, 386 Patterns of distribution, 106, 402, 404 Peak flow rate, 260 Pedogenesis, 65 Penman approach, 188 Percolation, 81, 197, 198, 261, 262, 266, 425, 426, 433 Perennial plants, 24, 81, 102, 106, 107, 113, 115, 116, 168, 300, 339, 386, 387, 394, 402, 425, 427, 432, 436 vegetation, 5, 62, 79, 116, 166, 233, 287, 288, 290, 291, 296, 298–300, 306, 377, 428, 433, 436, 437 Permanent quadrats, 107, 117, 387 Permeability, 80, 81, 455 pH, 65, 66, 68–70, 72, 74, 75, 127 Phanerophyte, 94, 99 Photosynthesis, 6, 164, 314–316, 368, 369 Photosynthetic activity, 161, 306, 314, 316 Phreatophytes, 373 Physiotypes, 355 Phytogenic hillocks, 108, 113, 115, 116 Phytomelioration, 445, 446, 448 Phytosociological studies, 105, 107, 121 Piston hypothesis, 266 Pitfall trap, 127, 145 Plant communities, 105–109, 113, 114, 116–118, 120–122, 144, 401, 417 cover, 1, 77, 79, 100, 107, 108, 113–116 diaspores, 338, 346 distribution, 117, 121, 386, 414 populations, 385, 394, 401, 414, 418 species, 21–23, 74, 75, 93, 94, 99, 100 Plant species diversity, 428, 430 Plant water potential, 368, 370, 373 Plant water status, 305, 367, 368, 456 Playa, 1, 21, 23, 46, 65–68, 71, 73, 77, 81, 100 Playa sediments, 22, 73, 231, 233 Pleistocene sediments, 11 Plinth, 159, 338, 340–346
Subject Index Pliocene shoreline, 9, 10 Ploughing, 83, 116 Political border, 25, 83 Pollinators, 143, 144, 455 Population density, 401, 415 Population dynamics, 394, 405 Populations, 144, 154, 385, 394, 401, 404, 414, 416, 418, 419 Pore clogging, 243, 244, 252, 257 Pore size, 67, 68, 71, 73, 252, 257, 272 Pore size distribution, 67, 68, 71, 257 Pore spaces, 79 Pore volume, 71 Porosity, 20, 68, 73, 212, 251, 252, 262, 279, 426, 427 Potassium, 75, 271, 279, 280, 283, 355, 356, 359, 361, 362 Precipitation, 4, 21, 61, 74–76, 79, 81, 82, 126, 178–179, 229, 231, 235, 236, 242, 243, 305 Predators, 139, 144 Predawn leaf water potential, 371 Predawn water potentials, 112, 118, 137, 357, 368, 370 Pressure, 25, 32, 37, 41, 49–51, 56, 59–61, 75, 76, 83, 176, 179, 181, 184 Prevailing winds, 38, 353 Psammophytes, 100, 101, 355, 445
Q Quantum yield, 315
R Regosols, 67, 68, 71, 75–77, 116 Radar, 42–45 Radiation, 53, 125, 175–179, 183, 186–188, 191, 193, 306, 307, 309–312, 379, 425 Radiocarbon dates, 27 Rain crust, 257, 286, 428 efficiency, 81 intensities, 4, 239, 240, 242, 243, 254, 256, 257, 266 storm, 260 threshold, 13, 433 Rainfall anomaly, 55, 62 distribution, 49, 178 gradient, 4, 17, 23, 24, 57, 59, 65, 80 variability, 55, 61, 62 Rainfall–Runoff Relationships, 254, 255, 432, 433
Subject Index Rainstorm, 240–244, 254, 257, 259, 266, 372, 428, 430, 436 Rainwater penetration, 426, 433, 436 Rainy seasons, 49, 56, 57, 59, 61, 178, 198, 261, 271 Recharge of soil water, 198 Reclamation, 443, 454 Recovery processes, 5, 285, 287, 291, 299, 456 vegetation of, 83, 87 Red Sea trough, 53 Redistribution, 4, 24, 123, 178, 225, 226, 233, 235, 248, 251, 252, 254, 261, 267, 382, 425–427 Redness index, 27, 28, 30 map, 28 Redoximorphism, 227 Reforestation, 445 Reg, 1 Regeneration, 164, 171, 228, 229, 232, 292, 300 Regeneration time, 327 Regional climate, 49 pressure field, 51 Registan, 3 Regosols, 67, 68, 71, 75–77, 116 Regression function, 378 Rehabilitation, 6, 441, 443, 445, 454–456 Relative growth rate, 328 Relative humidity, 176, 184, 185, 310 Relative neighbour effect, 395, 396 Relief energy, 379, 380, 382 Remote Sensing, 157, 159, 161, 164, 171, 316, 456 Removal experiment, 403, 404, 416 Reppelency, 240 Reproduction, 143, 338, 385, 386, 391, 394, 395, 397, 401, 402, 404, 405, 412, 414 Reproductive allocation, 405, 407, 409–412, 415–417 biomass, 405, 407, 409–412, 415–417 success, 346, 348, 387, 388, 391, 392, 397, 404, 417, 419 Resilience, 62, 427 Restoration, 453–455 Re–vegetation, 443, 455 Ripple marks, 20, 108, 202, 208, 293, 294 Roadside vegetation, 117 Roadsides, 101 Rocky belt, 9 Rodents, 125, 455 Root Penetration, 74, 75, 417 Root system, 20, 75, 322, 371–373, 447
469 Root zone, 21, 193, 198, 226, 233, 237, 373 Rooting depth, 372 Roughness, 36, 74, 175, 181, 182, 184, 186, 188, 211–215, 221, 443 Rub al–Khali, 2, 441 Runoff generation, 4, 6, 13, 239, 240, 243, 248, 252, 254, 257, 261, 286, 426–428, 432, 433, 436, 437 plots, 240–242, 428 rates, 20, 24 Rusty mottles, 23, 68 Red species, 102
S Sabkha, 1 Sahara, 25, 49, 51, 73, 80, 441 Saharo–Arabian, 49, 94, 98, 99, 102, 320 Salina, 1 Saline soils, 225, 233 wedges, 266 Salinity, 76, 225, 226, 231, 233, 237, 262, 267, 271, 309, 353–355, 374, 403 Salinization, 225, 231, 283, 353, 426 Salt accumulation, 353–355, 358–360, 426 content(s), 20, 71, 116, 233, 355, 357, 358 crusts, 353, 445 deposition, 225, 277 dynamics, 225–237, 456 flats, 445 Saltation, 213, 215, 217, 219, 221, 222, 443, 448 Salt–dust, 445, 448 Salty islands, 358 Sand barrier, 12–15 crusts, 306, 316 dams, 15 deflation, 211, 212, 220, 286 deposition, 20, 21, 276, 426 deserts, 441, 445 flux, 31, 216, 221 instability, 402–404, 414, 415, 418, 419 mobility, 21, 83, 100, 282, 286, 297, 418, 442 movement, 14, 15, 181, 202, 203, 211, 213, 214, 216, 217, 221, 222, 403, 404, 416, 442, 450, 454 stability, 101, 121, 122, 402–404, 414, 418, 419 storms, 445
470 Sand (cont.) texture, 82 transport, 31, 32, 37, 82, 203, 211–213, 215–217, 219–222 traps, 213, 214, 220 types, 28 water content, 261 Sand cover experiment, 403–405, 412, 413 Sand dune field, 49, 56, 57, 59–61, 180, 271–278, 282, 283, 329, 330, 333, 379, 380 mobility, 33 stabilization, 33 Sand fields, 17, 26, 44, 45, 149–151, 154, 157 Sand red colour, 27 Sand–moving wind, 35, 201 Sandveld, 454 Sandy belt, 9 ridges, 19, 65, 68, 253, 292, 437, 438 nebkhas, 22 Satellite images, 83, 443 Saturation deficit, 59, 60, 189, 192, 193 pressure, 59, 185, 314 vapour pressure, 188 Sea–breeze, 62, 307, 309 Sea–spray, 235, 279 Sea–spray salts, 274, 277, 280, 283 Sebkha, 1 Sediment budget, 206 flux, 222 yield, 242, 244, 246–248 Sedimentation, 228, 231, 237, 246, 248 Sediments, 9–12, 19, 22, 23, 27, 45, 46, 68, 71, 73, 162, 206, 213, 217 Seed bank, 340, 395, 445 burial, 415 dispersal, 339–341, 344–348, 388, 455 germination, 285 production, 387–389, 391, 395, 396 survival, 388 trapping, 347, 348 Seedling densities, 387, 389, 390, 395 emergence, 347, 415 survival, 337, 338, 387, 388, 405–407, 413, 415 Seed–trapping effect, 347 Seif, 36–38, 40–42 Seif dunes, 33, 36–38, 40, 41 Seig, 82
Subject Index Serir, 1 Shading, 315, 316 Shapotou, 451–453 Sharav, 53, 180, 282 Sharav conditions, 276 Shelterbelts, 443–445, 448, 450 Shifting sand, 65, 212 Shrub canopies, 236, 339, 397 cover, 86, 87 gathering, 81, 83 vegetation, 74, 86, 87, 226, 228, 237 Shunra, 9, 10, 13, 14, 17, 25, 26, 28, 31, 39, 44, 45 Siberian anticyclone, 50, 56, 61 Silt, 11, 20, 21, 23, 33, 46, 68–70, 72–74, 79, 81, 149, 161, 162 Silt fraction, 20, 226 Sinai, 4, 9, 17, 25–27, 30, 35, 37, 40, 41, 44, 45, 49 Size index, 377 Skimming flow, 212 Slip faces, 39, 40, 202–206, 208 Slope angle, 23, 175, 237, 247, 248, 251 Sodiophilic plants, 355 Sodium, 30, 355 Soil aggregates, 144 crusts, 61, 149, 150, 154, 158, 159, 162, 164, 171 distribution, 67 fauna, 305 fertility, 283, 442, 455 formation, 227 lichens, 306, 311, 316, 319, 321, 325 map, 65, 74 moisture, 1, 62, 81, 251, 252, 261, 271, 285, 288, 306, 310, 427, 428, 433, 434, 436 patterns, 353–363 processes, 225–237 properties, 65, 106, 145, 286, 309, 367, 425 stability, 286, 349 structures, 130, 382 surface, 73, 75, 100, 121, 122, 154, 158, 159, 166, 211, 229, 236, 280, 294, 305, 337–340, 353, 362 Soil water availability, 357, 367, 368, 383 regime, 285, 428 Solar radiation, 53, 175, 309 Solonchaks, 23, 67–69, 71, 73–77, 116, 228, 231, 445 Soluble salts, 67, 74, 225–229, 231, 233–237
Subject Index Sonora, 3 Southern Africa, 453 Sowing, 83, 388, 415 Sowing experiments, 337, 388, 389, 391 Space photographs, 157 Spatial distribution, 57, 106, 166, 168, 178, 225, 239, 253, 266, 380, 428 diversity of soil crusts, 159 heterogeneity, 145, 367, 401, 403, 418, 419 variability, 23, 24, 136–139, 145, 159, 161, 170, 226, 230, 233, 235, 237, 251, 259, 267 of soil, 236, 237 Species diversity, 151 numbers, 5, 100, 115, 122, 340, 341, 345, 347, 348 Specific humidity, 60, 62, 184–195, 197, 306, 311–313 Spectral endmember spectra, 166 Spectral mixture analysis, 165 Spectral reflectance, 28, 287, 316 Spectral unmixing, 165, 166 Sprinkling experiment, 239, 240, 433 Stabilisation, 443, 445, 450, 451, 454 Stability parameter, 184 Stabilization, 5, 33, 79–87, 121, 122, 232, 252, 285–287, 299 Stabilized sand dunes, 79, 182 Stagnant water, 23, 68 Standing crop, 401, 417, 418 Stipagrostis scoparia-Heliotropium digynum community, 108, 112, 113 Stomatal closure, 368 Stomatal control, 193 Storm synoptics, 276 Strandveld, 453, 455 Stratigraphic sequence, 11 Straw checkerboards, 443, 451 Stress gradient, 396 Subcanopy plants, 385 Subsistence, 83, 87 Subsoil, 74, 75, 226, 235–237 Substrate domain, 168 Subsurface flow, 21, 122, 198, 251–267 lateral flow, 262, 267 migration, 232 water, 261 water flow, 262 Succession, 161, 285–300, 319, 445, 448, 456 Succulence, 355 Succulents, 305, 453, 455 Sudanian, 94, 99
471 Sulphate, 231, 273, 274, 277, 282 Summer storm, 32 weather, 50 winds, 32, 40 Surface albedo, 83, 175, 176 changes, 202 crust, 66–68, 73, 74, 76, 77, 100, 113–116, 121, 122 disturbance, 287, 288, 299, 300 instability, 21, 24, 267, 438 level, 204, 205, 207 properties, 1, 6, 13, 15, 24, 157, 159, 164, 165, 175 roughness, 36, 74, 181, 186, 188, 443 runoff, 4, 20, 21, 74, 158, 161, 162, 164, 183, 225, 240 sealing, 257, 285, 428 stability, 1, 20, 222, 271, 431, 437 stabilization, 1, 5, 287, 299 structure, 73, 106, 122, 340 types, 163, 168, 170 wetting, 316 Surface reflectance spectra, 160 Surface root system, 372 Survival, 102, 212, 285, 337, 338, 373, 385–388, 391, 401 Surviving plant, 404, 405, 408–413, 415, 417 Sverdrup model, 195, 198 Synoptic patterns, 56, 181, 429 Synoptic types, 51 Syntaxonomic system, 105, 120, 121 Syntaxonomy, 120, 121
T Taklamakan, 3, 441 Takyr(s), 1, 22 Takyric, 66–69, 74 Taproot, 101, 372 Temperatures, 5, 17, 50, 53, 54, 59, 60, 125, 126, 175 Tengger desert, 450–453 Thar, 2 Thermal instability, 51 Therophytes, 94, 98, 99, 113 Topoclimate, 175–182 Topsoil crust(s), 5, 6, 17, 20, 23, 24, 73, 239, 240, 252, 296, 299 Trampling, 81, 83, 168, 211, 213, 287, 288, 299, 300, 337, 441 Trans–border access, 82
472 Transect(s), 19, 56–58, 159, 160, 162, 164, 168–170, 178, 181, 208 Translocation of salts, 65, 232 Transpiration, 6, 59, 62, 177, 183–199, 368, 370, 373, 456 Transported mass, 217–222 Transverse dunes, 39 T–root system, 372 Trophic levels, 143, 145
U Underestimated evaporation, 188 Understory species, 394
V Vapour fluctuations, 189, 190 Vapour flux, 59, 61, 62, 184–186, 188–190, 312 Vapour flux models, 184 Vapour pressure, 59, 60, 62, 176, 179, 185, 188, 193, 196, 309, 310 Variation, 27, 59, 68, 71, 77, 82 Vascular plant(s) abundance, 337 response, 338, 339, 341, 343, 345–349, 351 species, 93, 102, 339, 342 VEGDUNE model, 379–381 Vegetated domain, 166 Vegetated linear dunes, 33, 35, 232 Vegetation canopy, 216, 217, 221, 222 cover, 1, 5, 13, 14, 17, 20, 21 density, 82, 198, 211, 213, 216, 220, 443 domain, 168 growth, 79, 81, 267 pattern, 57, 121, 123, 126, 226, 367, 377, 380, 382, 430, 456 Vertical differentiation, 145 Volume balance, 207
Subject Index preservation, 261, 426, 427 redistribution, 123, 178, 248, 251, 252, 261, 267, 382, 426 regime, 5, 17, 21, 22, 24, 266 repellency, 240 source, 82, 83, 305, 372, 373 storage, 59, 62, 198, 373 stress, 79, 367 vapour, 59–62, 176, 179 yield, 261 Water vapour flux, 61 Water–soluble elements, 233, 272, 275, 282 Water–soluble ions, 226, 353 Waveform, 37, 44 Weather types, 50, 51, 56 Weathering, 65, 225, 227, 228, 441 Wet belts, 438 deposition, 233, 237, 274 years, 21, 22, 54, 100, 298, 300, 368, 395, 397, 398 Wind climate, 31 conditions, 179, 222, 454 directions, 31, 32, 38, 181, 189, 213 energy, 17, 31, 33, 179, 211, 213 erosion, 99, 106, 108, 189, 194, 237, 238, 255, 274, 302 field, 51, 59, 178, 181, 283, 443 power, 38, 82 regime, 4, 5, 37, 38, 203, 207, 209 speed, 1, 5, 31, 51–53, 61 Windbreaks, 443 Winter, 2–4, 17, 25, 31 storm, 32, 203, 213, 219, 221 winds, 40, 203, 214, 244 Wood–gathering, 25
X Xero–halophyte, 23, 355, 361, 373 W Wadis, 25, 26, 42, 45, 46, 372 Wake interference flow, 221 Water availability, 5, 24, 76, 122, 123 capacity, 67, 75–77, 229 content, 65, 67, 136, 179, 198, 228 interception, 385 movement, 251, 261, 262, 266, 267, 436 penetration, 4, 22–24, 229, 251, 262, 426, 427, 433, 436 percolation, 197, 261, 262, 425, 426, 433 potential, 198, 357, 367, 368, 370, 371
Y Year–to–year fluctuations, 105 Year–to–year variations, 386 Yermic Arenosols, 67, 68, 79, 226, 228, 229, 233, 237 Yevul, 4, 57, 58, 61, 196, 272, 329
Z Zero plane model, 189–193, 195, 196, 199 Zonal circulation, 50, 56, 62, 283
Taxonomic Index
A Acacia, 443 Acacia tortilis, 372 Acinipe zebratus, 128, 130, 134 Adesmia metallica, 128, 135, 137 Agama, 134, 449 Aizoon hispanicum, 94, 100, 112 Algae, 21, 73, 149, 152–154, 166, 228, 248, 252, 271, 285, 299 Allium papillare, 94, 100, 102 Ammochloa palaestina, 94, 100 Anabasis, 137, 235, 236, 326, 327, 355, 357, 358, 360, 362, 371, 378, 379 Anabasis articulata, 22, 23, 100, 107, 109, 112, 114, 116, 127, 141, 225–229, 328 Anthemis melampodina, 94, 100, 111, 116, 117, 119 Ant lion, 130, 139 Ants, 125, 128, 129, 137, 142–145, 305 Arenarius, 144, 145 Arenivaga, 143 Argyrolobium uniflorum, 94, 100, 109, 115 Artemisia monosperma, 21, 82, 94, 100, 105, 109, 112, 114, 120, 193, 197, 321, 354, 356, 358, 370, 371, 373, 378, 430 Artemisia ordosica, 451 Artemisia tridentata, 373 Arthraerua leubnitziae, 305 Arthropod groups, 142 Astragalus spinosus, 101 Asteraceae, 94, 107, 114, 387, 453 Asthenatherum forsskalii, 95, 100, 109, 115, 430 Astragalus corrugatus, 94, 102 Atractylis carduus, 94, 100, 110, 114, 120 Atractylis cuneata, 430 Atriplex confertifolia, 359 Atriplex dimorphostegia, 94, 100, 102, 111, 116
Atriplex halimus, 94, 101 Atriplex pratovii, 446
B Banksia prionotes, 373 Banksia, 372 Beetles, 125, 128, 135, 142, 144, 305 Bellevalia eigii, 94, 100, 112, 116 Birds, 125, 126, 128, 131, 144 Brachythecium, 21 Brassicaceae, 94 Bromus fasciculatus, 95, 101, 107, 110, 115, 117, 118, 402, 406–411 Bryum, 21, 152, 154
C Calligonum comosum, 95, 102, 109 Calligonum mongolicum, 451 Calligonum, 445 Callipeltis aperta, 95, 102 Calothrix, 152, 319 Cannabis sativa, 95, 101, 117 Carabidae, 135 Caragana korshinskii, 451 Carrichtera annua, 95, 100, 112, 116 Chenopodiaceae, 101, 107, 114, 225, 321, 354, 355, 358, 359, 373, 379 Chroococcidiopsis, 150, 152 Chroococcus, 319 Chrysothamnus nauseosus, 373 Ciconia ciconia, 128, 131 Cistanche salsa, 95, 102 Cleome arabica, 95, 101 Colchicum ritchii, 95, 102, 110 Collema tenax, 23, 152, 314, 321, 324, 325, 328, 334 Collema, 150, 316 473
474 Collembola, 129, 136, 143 Convolvolus lanatus, 403 Cornulaca monacantha, 95, 101, 102, 107–109, 112, 114–117, 120, 121, 339, 354, 356, 362, 363, 373, 378, 379, 403 Cornulaca, 101, 355, 358, 379 Crassula, 305 Galerida cristata, 128, 132 Cutandia memphitica, 95, 100, 101, 114 Cyanobacteria, 1, 21, 149–151, 153, 154, 161, 164–169, 228, 235, 294, 299 Cyperus macrorrhizus, 95, 100, 107, 108, 110, 114
D Daucus litoralis, 95, 100, 111 Dipcadi erythraeum, 95, 102, 110 Diploschistes diacapsis, 150, 152, 154, 315, 321 Diptera, 129, 136, 143
E Echinops philistaeus, 95, 100, 109, 112–114, 120 Echiochilon fruticosum, 22, 95, 100, 112, 115, 213, 339, 403 Emex spinosa, 95, 100, 116 Eragrostis poaeoides, 451 Eremobium aegyptiacum, 95, 108, 110 Erodium crassifolium, 85, 110, 430 Erodium laciniatum, 95, 101, 107, 108, 111, 114, 115, 117, 118, 387 Euphorbia grossheimii, 96, 102, 112
F Fabaceae, 94, 319, 320 Fulgensia fulgens, 23, 150, 152, 154, 314, 321, 324, 325, 328
G Galerida cristata, 132 Glaucium corniculatum, 96, 101, 102 Graminoids, 108, 113 Graphopterus serrator, 134, 135 Grasshoppers, 128, 130, 144 Grus grus, 128 Gymnarrhena micrantha, 96, 100
H Halocnemum strobilaceum, 445 Haloxylon aphyllum, 445, 446, 450
Taxonomic Index Hedysarum scoparium, 451 Helianthemum sessiliflorum, 96, 100, 109, 115, 339 Heliotropium digynum, 20, 96, 100, 107–109, 112–114, 339, 354, 356, 378, 386, 403 Heliotropium rotundifolium, 96, 101 Hormuzakia aggregata, 96, 100, 112 Hypecoum littorale, 96, 102, 110 Hypogymnia physodes, 331 Hypotrichs, 143 Hystrix indica, 133, 144
I Ifloga spicata, 96, 100, 101, 107, 108, 110, 115, 117, 118, 120, 339, 340, 344, 387 Insects, 134, 455 Invertebrates, 125–127, 130, 136, 138, 140, 143–145, 305
L Lanius excubitor, 128 Launaea tenuiloba, 100, 109 Leontice leontopetalum, 96, 100, 112, 116 Leontodon laciniatum, 100, 116 Lepismatidae, 136, 143 Lichens, 24, 73, 114, 149–152, 154, 183, 228, 252, 285–287, 294 Loeflingia hispanica, 96, 102 Lotus halophilus, 97, 100, 108, 111, 119 Lycium shawii, 97, 110, 115
M Macrochloris multinucleata, 152–154 Malva parviflora, 97, 100, 116 Mammals, 125, 126, 128, 133, 138, 144, 145 Mantodea, 128, 134, 136 Messor sp., 129, 144, 145 Messor semirufus, 144 Microcoleus, 151–153, 161, 164, 319 Moltkiopsis ciliata, 21, 97, 100, 101, 107–109, 112, 114, 115, 117, 339, 354, 356, 378, 386, 403, 430 Moricandia nitens, 97, 101, 112, 116 Morus, 443 Mosses, 21, 22, 24, 73, 114, 149–152, 154 Myrmeleonidae, 130
N Nematoda, 143 Neurada procumbens, 17, 100, 111, 117, 119, 402, 406–411
Taxonomic Index Noaea mucronata, 21, 97, 109, 112, 114, 354, 356, 378 Nostoc, 152–154, 319, 321 Nostoc microscopicum, 152, 154
O Oligomeris liniifolia, 100 Orthoptera, 128, 134, 136 Oscillatoria, 152, 154, 321
P Pancratium sickenbergeri, 97, 102, 109 Panicum turgidum, 82, 97, 110 Paronychia palaestina, 97, 100, 102 Peganum harmala, 97, 100, 112, 116 Phagnalon barbeyanum, 97, 100, 102 Phormidium, 150, 152, 154, 321 Picris asplenioides, 97, 100, 102, 111, 115, 117, 118 Pimelia arabica, 137, 144 Plantago coronopus, 97, 100 Poaceae, 94 Polycarpon succulentum, 97, 100, 111, 117, 118 Populus, 443, 445 Prosopis, 333 Protozoa, 127, 143 Pseudevernia furfuracea, 331 Pulchripennis, 128, 130, 134
R Ramalina maciformis, 312 Reptiles, 125, 126, 128, 129, 133, 134, 449, 450 Retama, 235, 236, 271, 279–282, 326, 333, 370, 371 Retama raetam, 21, 97, 102, 109, 115, 127, 141, 193, 328 Rhizobium, 319, 320, 322 Rumex, 390, 392, 393, 396 Rumex pictus, 97, 100, 102, 107, 108, 110, 114, 117, 118, 387, 392 Rumex vesicarius, 97, 102
S Sarcobatus vermiculatus, 373 Scarites, 134, 135 S. cartilaginea, 150, 152, 154, 320, 321, 331, 334
475 Schizothrix, 150, 152 Scotocerca inquieta, 128, 133 Scrophularia hypericifolia, 98, 100, 109 Scytonema, 152–154, 321 Senecio, 117, 390, 392, 393, 396 Senecio glaucus, 98, 101, 107, 108, 111, 114, 115, 117, 118, 340, 345, 387 Silene villosa, 98, 100, 110 Sixalix eremophila, 98, 102, 110, 119 Snails, 20, 136, 144, 305 Squamarina, 23, 314, 324, 325, 327, 331, 332 Squamarina cartilaginea, 150, 152, 154 Squamarina lentigera, 150, 152, 154, 311, 312, 314, 315, 321 Stipa capensis, 98, 100, 111, 116, 117, 119 Stipagrostis ciliata, 98, 112, 354, 356 Stipagrostis plumosa, 82, 98, 109, 115, 120, 339 Stipagrostis scoparia, 20, 81, 82, 98, 100, 101, 105, 108 Suaeda aegyptiaca, 355
T Tamarix, 445 Tenebrionids, 141, 144 Termites, 129, 143 Thymelaea, 370, 371 Thymelaea hirsuta, 22, 82, 98, 109–112, 115, 116, 193, 213 Tmethis pulchripennis, 128, 130, 134 Tmethis pulchripennis asiaticus, 130, 134 Tortula, 154 Trichocoleus, 150, 151, 153 Trichocoleus sociatus, 151, 153 Trigonella arabica, 98, 100 Trigonella stellata, 98, 100, 111, 116, 117, 119 Trisetaria glumacea, 98, 102
U Ulmus pumila, 448 Urginea maritima, 98, 102 Urospermum picroides, 98, 100, 112, 116
V Vertebrates, 125–127, 130, 136, 138, 140, 143 Vulpia brevis, 98, 102 Vulpia pectinella, 98, 100, 102