K e ith M S cot t an d Coli n F Pa i n
regolith
science
regolith SCIENCE
regolith SCIENCE K E I T H M SCOTT AN D C OL IN F PAIN
© CSIRO 2009 All rights reserved. Except under the conditions described in the Australian Copyright Act 1968 and subsequent amendments, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, duplicating or otherwise, without the prior permission of the copyright owner. Contact CSIRO PUBLISHING for all permission requests. National Library of Australia Cataloguing-in-Publication entry Regolith science/editors, Keith M. Scott, Colin F. Pain. 9780643093966 (hbk.) 9780643097834 (pbk.) Includes index. Bibliography. Regolith. Geomorphology. Scott, Keith M. Pain, C. F. 551.41 First printed in hardback in 2008 Published exclusively in Australia and New Zealand by: CSIRO PUBLISHING 150 Oxford Street (PO Box 1139) Collingwood VIC 3066 Australia Telephone: Local call: Fax: Email: Web site:
+61 3 9662 7666 1300 788 000 (Australia only) +61 3 9662 7555
[email protected] www.publish.csiro.au
Published exclusively throughout the world (excluding Australia and New Zealand) by Springer, with ISBN 978 1 4020 8859 9 Springer Van Godewijckstraat 30 3311 GX Dordrecht The Netherlands Tel: +31 78 657 60 00 Fax: +31 78 657 65 55 Web site: springer.com Front cover: A pseudocoloured 1st vertical derivative magnetic image for the West Wyalong area in NSW, Australia. Data supplied by Geoscience Australia. Back cover (from left): A Ternary image of radiometrics (data supplied by Geoscience Australia); a pseudocoloured total magnetic intensity image (TMI) shaded by a 1st vertical derivative (data supplied by Geoscience Australia); a regolith landform map of Ebagoola, producing in a GIS environment. Set in 10/13 Adobe Minion and ITC Stone Sans Edited by Peter Storer Cover and text design by James Kelly Typeset by Desktop Concepts Pty Ltd, Melbourne Index by Russell Brooks Printed in Australia by Ligare
Contents Preface Acknowledgements Contributors
vii viii ix
9
Regolith geophysics Tim Munday
219
1
Introduction Keith M Scott and Colin F Pain
1
10 Regolith and water Richard G Cresswell and Paul Shand
2
Regolith through time Brad Pillans
7
11 Regolith description and mapping Colin F Pain
281
3
Landscape and regolith Graham Taylor
31
12 Soils and natural resource management Robert W Fitzpatrick
307
Regolith mineralogy Richard A Eggleton
45
4
5
Regolith geochemistry Kenneth G McQueen
6
Rock weathering and structure of the regolith 105 Kenneth G McQueen and Keith M Scott
7
Geomicrobiology of the regolith Frank Reith, Mira Dürr, Susan Welch and Stephen L Rogers
8
73
127
Colour plates
159
Regolith and biota John Field and David Little
175
13 Regolith sampling for geochemical exploration Charles R M Butt, Keith M Scott, Matthias Cornelius and Ian D M Robertson 14 Extraterrestrial regolith Jonathan D A Clarke Appendix 1: Glossary of regolith terms Richard A Eggleton, Colin F Pain and Keith M Scott Appendix 2: Regolith geochemistry of elements Keith M Scott Index
251
341
377
409
433
453
Preface Regolith – the mantle of in situ and transported weathered material that covers landscapes across the world – presents a major challenge to mineral explorers and natural resource managers. The Cooperative Research Centre for Landscape Environments and Mineral Exploration – CRC LEME, (formerly the Cooperative Research Centre for Landscape Evolution and Mineral Exploration) – was established under the Australian Federal Governments Cooperative Research Centres Program to ‘create breakthroughs in mineral exploration and environmental management, through generating and applying new knowledge of the regolith’. Through its 13-year existence in partnership with industry and government end users, regolith geoscience research in CRC LEME has applied the traditionally disparate scientific disciplines of geology, geophysics, geochemistry, geomorphology, soil science, microbiology, molecular biology, biochemistry, hydrogeochemistry, hydrology, plant biology and ecology to:
s s s s s
identifying potential zones of mineralisation determining mineral transport and transformation mechanistic processes in landscapes identifying paleo-landscape features and processes groundwater and salinity mapping identifying ‘natural’ geochemical hazards such as acid sulfate soils, and groundwater acidification.
In addition, the CRC has made significant advances in the dating and understanding of Australian landscape evolution, geochronology and regolith mapping techniques. While the focus of CRC research has been on Australian regolith-dominated landscapes, the regolith geoscience knowledge generated is highly
relevant to other landscapes dominated by cover, such as those in parts of South America, Africa, Indian sub-continent and China. A search of Amazon.com yields six books concerned with regolith on Earth (the remainder are about regolith on extraterrestrial bodies). They are Butt and Zeegers 1992; Kauranne et al. 1992; Cremeens et al. 1994; Ollier and Pain 1996, Eggleton 2001 and Taylor and Eggleton 2001(see references, Chapter 1). Of these, four are authored/co-authored by Australian scientists associated with CRC LEME (Butt, Eggleton, Pain and Taylor) – reflecting the leadership of CRC LEME in regolith geoscience. This book, which has been written by CRC LEME scientists, expands on the 1996 work of Ollier and Pain and the 2001 work of Taylor and Eggleton. The initial idea for this book came from the industry/end-user advisory councils of CRC LEME (Minerals Advisory Council and Land Use Advisory Council). Both end-user groups identified the need for a comprehensive compendium of advances in regolith geosciences aimed at geoscience practitioners with little or no regolith knowledge. The volume is therefore intended for mineral exploration and natural resource management geoscientists, and provides an up to date summary of regolith understanding, and inevitably concentrates on CRC LEME science. However, the science is placed in a broad context to make it a valuable reference book for geoscientists worldwide. Steve Rogers Chief Executive Officer CRC LEME June 2008
Acknowledgements The editors acknowledge the support of CRC LEME (both the Cooperative Research Centre for Landscape Environments and Mineral Exploration and the former Cooperative Research Centre for Landscape Evolution and Mineral Exploration) in encouraging the publication of this book, which largely details aspects of work conducted by CRC LEME during the 13 years of its existence. The CRC was supported by the Australian Government’s CRC Program and specifically supported by CSIRO, The Australian National University, University of Canberra, Geoscience Australia, Curtin University of Technology, Minerals Council of Australia, New South Wales Department of Primary Industries, Primary Industries and Resources South Australia and The University of Adelaide as core parties. The manuscript benefited from discussions with colleagues within CRC LEME and individual chapters were also improved by careful and insightful reviews by:
s
Michael Bird, Mike Thomas and Julie Bell-Lanier (UK); Maite Le Gleuher (France); Eric Tonui, John Dohrewend and Mary Bourke (USA); Gordon Southam (Canada)
s
Brad Pillans, Lisa Worrall, Jonathon Clarke, Mark Raven, David Gray, Michael Whitbread, Paul Wilkes, Patrice de Caritat, Kok Piang Tan, Steve Hill, Ian Robertson, Charles Butt, Vic Gostin, Helen Waldron, Rob Hough and Andy Christie (all from Australia).
A number of illustrations within Chapter 4 are from Taylor and Eggleton (2001) and are used with the permission of John Wiley and Sons. The Visual Resources Unit of CSIRO and CRC LEME (Angelo Vartesi and Travis Naughton) at Kensington WA has (re-) drafted all the figures of this book. Chapter 13 and Appendix 1 are substantially drawn from Butt et al. (2005) and Eggleton (2001), respectively (both published by CRC LEME). The editors have freely drawn upon the work by the chapter contributors to compile Chapter 1. The editors thank their wives (Lyn and Josie) and other family members for their forbearance during this ‘labour of love’. Keith Scott and Colin Pain June 2008
Contributors Charles R M Butt CRC LEME, CSIRO Exploration and Mining, PO Box 1130, Bentley, WA 6102. Jonathan D A Clarke CRC LEME, Geoscience Australia, PO Box 378, Canberra, ACT 2601. Matthias Cornelius Cullen Resources Ltd, 7 Hardy Street, South Perth, WA 6151. Richard G Cresswell CRC LEME, CSIRO Land and Water, 120 Meiers Road, Indooroopilly, QLD 4068. Mira Dürr CSIRO Land and Water, Private Bag 2, Glen Osmond, SA 5064. Richard A Eggleton CRC LEME, Research School of Earth Sciences, Australian National University, Canberra, ACT 0200.
David Little Earth Tech, PO Box 5109, Townsville, Qld 4810. Colin F Pain CRC LEME, Geoscience Australia, PO Box 378, Canberra, ACT 2601. Brad Pillans CRC LEME, Research School of Earth Sciences, Australian National University, Canberra, ACT 0200. Frank Reith CRC LEME, CSIRO Exploration and Mining, PO Box 1130, Bentley, WA 6102. Ian D M Robertson CRC LEME, CSIRO Exploration and Mining, PO Box 1130, Bentley, WA 6102. Stephen L Rogers CRC LEME, CSIRO Exploration and Mining, PO Box 1130, Bentley, WA 6102.
John Field CRC LEME, Fenner School of Environment and Society, Australian National University, Canberra, ACT 0200.
Keith M Scott CRC LEME, CSIRO Exploration and Mining, PO Box 136, North Ryde, NSW 1670, and Research School of Earth Sciences, Australian National University, Canberra, ACT 0200.
Robert W Fitzpatrick CRC LEME, CSIRO Land and Water, Private Bag 2, Glen Osmond, SA 5064.
Paul Shand CRC LEME, CSIRO Land and Water, Private Bag 2, Glen Osmond, SA 5064.
Kenneth G McQueen CRC LEME, Research School of Earth Sciences, Australian National University, Canberra, ACT 0200; and Faculty of Applied Science, University of Canberra, ACT 2600.
Graham Taylor University of Canberra, Canberra, ACT 2601.
Tim Munday CRC LEME, CSIRO Exploration and Mining, PO Box 1130, Bentley, WA 6102.
Susan Welch School of Earth Sciences, Ohio State University, Columbus, Ohio 43210, USA.
1
Introduction Keith M Scott and Colin F Pain
1.1
REGOLITH – WHAT IS IT?
Regolith is ‘the entire unconsolidated or secondarily re-cemented cover that overlies more coherent bedrock’ and which ‘has been formed by weathering, erosion, transport and/or deposition of the older material’. Thus it ‘includes fractured and weathered basement rocks, saprolites, soils, organic accumulations, glacial deposits, colluvium, alluvium, evaporitic sediments, aeolian deposits and ground water’. Or, more simply put, it is ‘everything from fresh rock to fresh air’ (Eggleton 2001: Appendix 1). Merill (1897) applied the term ‘regolith’ to the surface mantle of unconsolidated material. The term languished in obscurity through the first half of the 20th century, but gained currency during the 1970s as space missions revealed the presence of fragmental materials mantling the surfaces of the Moon, Mars, Venus, Mercury, comets and the moons of the outer planets. The term ‘regolith’ was subsequently re-applied with increasing frequency to terrestrial situations. Thus, regolith geology may have been conceived on Earth, but was raised on the surface of the Moon and Mars before returning to its home planet (Clarke 2003) (Chapter 14). Regolith consists of physically broken and, generally, chemically altered rocks. It also contains water, biota and gases. Rocks, when moderately to intensively
chemically altered in situ, form profiles that consist of progressively more altered bedrock towards the surface – an in situ weathering profile. Weathered debris may be moved by surface erosion or moved below the surface – in solution or physically – by groundwater and biota. Such eroded components may be deposited to form transported regolith elsewhere in the landscape. Recognition of the presence of transported material in regolith profiles is becoming increasingly important in exploration, with many recent studies devoted to finding mineralisation beneath such cover. Regolith – both in situ and transported – is almost ubiquitous at the Earth’s surface. In some places it is laterally extensive and more than several hundred metres thick (for example, in Mesozoic-Cainozoic basins; Figure 1.1). In situ deep weathering is particularly common in the tropics, and there is a prevailing idea that ‘tropics’ and ‘deep weathering’ go together. However, deep in situ regolith is found in many places outside the tropics, including northern Europe, the United States, India and southern Brazil. It is common in arid and semi-arid areas. Stierman and Healy (1984) report 70 m of granite weathering in southern California, and it even occurs in Antarctica (for example, Guglielmin et al. 2005).
Regolith Science
Weipa
CANNING BASIN s nd la h g Hi
CARNARVON BASIN EROMANGA BASIN Mesozoic (1200m) PERTH BASIN
Perth
SURAT BASIN
Eas ter n
2
Sydney
EUCLA BASIN MURRAY BASIN Cainozoic (600m)
GIPPSLAND BASIN Lake Tyrell
Figure 1.1: Major Australian Mesozoic–Cainozoic basins (showing maximum thickness of sediments) and other commonly studied areas referred to in the text.
The chemical interaction starts with dissolution of components from the minerals in the rock and oxidation of readily oxidisable ions, such as ferrous iron (Fe2+). How much a particular rock will weather, and what the products of the weathering will be, depends upon the climate, rock type and landscape.
1.2
CLIMATE
Given that deep weathered regolith is found in most parts of the world – and in many different modern climatic zones – it seems clear that many weathering profiles were not formed under the present climate, but under different climates at various times in the past. Early speculation on deep and intense weathering placed the time of weathering in the interglacial periods of the Quaternary. More recently they were ascribed to the ‘Tertiary’. Modern information suggests several periods of deep weathering during the Cenozoic, with some going back as far as the Carboniferous (Chapter 2). Since the composition of the atmosphere is essentially the same everywhere, climate affects only the temperature of weathering, the amount of water available to dissolve the minerals and the seasonality of weathering. Temperature controls the rate of chemical
reactions, and also the rate of biotic processes. It has no other significant effect, and the notion that extensive weathering (such as that which leads to the development of bauxite) is a tropical phenomenon is false. Certainly there are more bauxites in the tropics because the rate of their production is quickest there and so they survive erosion long enough to be recognised. Weathering occurs as long as there is air and water, and the chemistry of weathering is not itself temperature dependent. Water is the agent of mineral dissolution – aided by dissolved CO2, which lowers the pH, and by organic chemicals that may affect pH or change the solubility of minerals by the production of, for example, chelates. The more water that passes over a mineral surface (or through cracks in it), the greater the amount of dissolution. If the water is stagnant, it may become saturated in a particular chemical and this will temporarily prevent further dissolution. The most important contribution climate makes to rock weathering is through the provision (or withholding) of water. Seasonality of rainfall also affects the weathering process. Weathered rocks that are always wet exclude the atmosphere, and hence there is a demarcation between oxidised and reduced parts of the weathered
Introduction
Weathering Uplift
Volcanic rocks
Plutonic rocks Magma generation
Regolith Processes Biosphere
Deposition
Regolith
Diagenisis Atmosphere
Metamorphism
Lithosphere
Sedimentary rock
Figure 1.2: Weathering in the geological cycle (after Wilson 2004).
rock. Regionally, this demarcation is the water table; locally it may be a patch of the regolith that it never dries out as the seasons change. Regions of highly seasonal rainfall, such as the sub-tropics where monsoons are followed by a 6- to 8-month dry season have a zone in the regolith that alternates each year from wet to dry and from oxidising to reducing as the water table rises and falls. This alternation gives rise to a unique regolith profile, which is often dominated by duricrusts – cemented layers at or near the surface – and is commonly referred to as a ‘lateritic profile’. Weathering is an integral part of the geological cycle (Figure 1.2), and is generally regarded as commencing when rocks are exposed to the atmosphere and the physical and/or chemical breakdown of component minerals occurs. Some of the freed components are then removed from their original location by physical or chemical processes and redeposited elsewhere and, when subjected to diagenesis, continue their path in the geological cycle. This book is concerned with the weathering, erosion and deposition processes; that is, regolith processes. Such processes involve the interaction between minerals, air and water, which is enhanced in most cases by the activities of biota (Figure 1.3).
1.3
Hydrosphere
Erosion
ROCK TYPE
Other things being equal, the rock type determines the rate at which the rock weathers and the possible products of weathering. Broadly, igneous rocks
Figure 1.3: The influence of different interactions on regolith (after Taylor and Eggleton 2001).
weather at a rate inversely proportional to their temperature of formation. Basalts weather quickly; granites more slowly. Sediments – being composed of minerals that have survived one episode of weathering – weather much more slowly than igneous rocks, and metamorphics fall in between these extremes. The rock type also dictates possible products. Aluminous rocks, such as arkoses or granites, can, if sufficiently leached, weather to gibbsite and so become bauxite (Section 6.3.1). Basalts have enough Al to do likewise, although rarely in commercial quantities. However, ultramafics have very low Al, but are rich in Mg and Fe so that they weather to a mixture of hematite and opal – with Ni sometimes concentrated to form lateritic Ni deposits (Section 6.3.3). Quartzites barely weather at all, and leave nothing behind when they do eventually dissolve except resistate accessory minerals (Section 6.3.4). Limestone dissolves rapidly, although, because most limestones are relatively free of pore space, water does not have easy ingress, and some limestones survive as karst features surrounded by less soluble, but more permeable, rocks (Section 6.3.5). The products of weathering also depend on the degree to which the rock has been weathered. During the early stages of weathering of granite, the biotite may alter to vermiculite (Section 4.3), cores of plagioclase feldspars may partly alter to smectite or kaolinite, and some quartz and feldspar crystals may be loosened from the rock body through dissolution along grain boundaries or along cracks. With further
3
4
Regolith Science
weathering, only resistant quartz and minor minerals (such as zircon) might remain in a matrix of clays and Fe oxides. Under extreme weathering conditions, even the quartz might dissolve completely and the clays alter to gibbsite (Section 6.3.1). The extent to which the weathering processes continue determines the actual character of the regolith profile.
1.4
LANDSCAPE
All rocks weather. How much of the weathered products remain depends on the balance between the weathering rate and the erosion rate. A steep rock face has barely any regolith. As fast as the minerals weather, they are washed away by rain because there is nothing to hold them in place. Colonisation by lichen may retain a few millimetres of weathered rock, but not much else. By contrast, a flat rock surface, such as on a basalt flow, may retain most of the solid products of mineral weathering, which allows a sequence of weathered products to evolve. Thus, close to the unweathered basalt – whether in a core-stone or at the bottom of the profile – primary minerals are set in a matrix of largely smectite. In more extensively weathered parts, kaolinite and Fe oxides dominate and, ultimately, along fissures where water access is easiest, gibbsite may be present. In pockets or cracks where water periodically accumulates and then dries, calcite may be precipitated. The balance between retention of weathered product minerals, loosened primary minerals and precipitated minerals is a very local phenomenon, and depends critically on the landscape position. It also depends on the colonisation of the locality by organisms – chiefly plants – whose presence slows the rate of erosion of the weathered regolith. Because the extent and character of biotic colonisation is climate-, rocktype- and landscape-controlled, there is a complex interaction between all of the factors that affect the development of a regolith profile.
1.5 REGOLITH AND BIOTA/ GEOMICROBIOLOGY Biota have long been recognised as an important factor in soil, regolith and landscape processes. As early as John Evelyn’s 1679 discourse on tree management, the soil is
described as combining ‘salts and ferments’ and ‘mold’ in addition to ‘sand, gravel, stone, rock or shell’. He goes on to speak of layers, and the uppermost ‘most excellent black mold’ in which to grow vegetation. Thus regolith (particularly soil) is the ‘life support system’ for all biota, and regolith as we know it would be markedly different without the presence of biota (compare weathering on extraterrestrial bodies with Earth: Chapter 14). At a micro scale, root exudates effect the weathering of alumino-silicate minerals (Section 8.2.3) and, at the macro scale, tree fall disturbs large volumes of material (Section 8.3.2). The recent coining of the term ‘critical zone’ emphasises the interdependence of regolith processes and life on Earth (Brantley et al. 2007). Today, the many ways that biota interact with the regolith, such as the growing discussion of carbon sequestration in soils and the field of geomedicine are being investigated. When considering biota, the problem has been that representing biological interactions is so complex that simpler inorganic, abiotic examples have always been used (for example, Wilson 2004). For instance, a relatively simple equation can be written for the reaction between an alumino-silicate mineral and carbonic acid, whereas a more realistic equation would need to contain a complex cocktail of organic acids, and a number of different alumino-silicate minerals all present at the same time (and causing complex interchanges in the immediate environment of the mineral crystal that is being weathered) – not to mention organic ligands and chelates and numerous pathways for different parts of the same mineral crystal (Chapters 7 and 8).
1.6
REGOLITH AND WATER
Water plays a critical role in regolith development – even in arid environments – and surface and sub-surface flow are themselves modified by the structural make up of the regolith. For example, groundwater may flow through both in situ and transported regolith at variable depths, as well as out-flowing as springs on the ground surface or taken up by vegetation (Figure 1.4). The supply of water has become a major issue in Australia during the first decade of the 21st century, as a series of drier-than-usual years followed one after the
Introduction
Regolith (in situ) Biota
We
at
ing her
fron
Soil
t
Regolith (eroded)
Infiltration
Regolith (transported and in situ)
W ate rta ble
Groundwater flow
Jointed rock
Figure 1.4: The effect of groundwater in regolith processes (after Taylor and Eggleton 2001).
other. Coram et al. (2000) estimated that 2.5 million hectares were affected by rising groundwater levels and dryland salinity, and that this was likely to increase four-fold in the next three to four decades. It was recognised that this would impose a large cost on land users. Thus, water and salinity became important issues, and funding for solutions was made available from a number of government sources, including the National Action Plan for Water Quality and Dryland Salinity. Similarly, in areas with acid sulfate soils (ASS), the rereduction of the oxidised ASS by controlled water table management, including re-flooding, can remediate these now degraded areas so that they can again be used for agriculture (see Section 7.7.1 and Chapter 12). Thus it is clear that regolith geoscience plays an important role in understanding the problems and developing land management solutions (for example, Dent et al. 1999; Wilford et al. 2001) (Chapters 9–12).
1.7 ECONOMIC DEPOSITS WITHIN THE REGOLITH Some valuable commodities, such as lateritic Ni, bauxite and beach sand deposits, are formed entirely by regolith processes. Thus, Al hydroxides (bauxite)
are several metres thick over areas greater than 1000 km2 at Weipa (Figure 1.1) and can be economically mined. Furthermore, with advances in highpressure leaching technology, many low-grade lateritic Ni-Co enrichments in the regolith – previously regarded as more as a hindrance to exploration for Ni sulfide deposits than as resources themselves – have become economically viable. However, most economic Au and base metal deposits were formed by hydrothermal processes and subsequently affected by regolith processes. Thus, in the Yilgarn Craton of Western Australia (which accounts for 65% of Australia’s current Au production), although the bulk of the production is derived from primary mineralisation, much exploration and mining over the past three decades has centred on the discovery and exploitation of shallow, low-grade deposits in the regolith. The latter are enrichments in lateritic residuum and saprolite, or as dominantly chemical accumulations in sediments in paleochannels, or in the saprolite beneath them. The deposits are commonly small, with mineable reserves of 0.5–1.5 M tonnes at grades of 1.0–5.0 g/t Au. In many deposits, lateritic and saprolitic Au provides an easily exploitable resource that provides an early cash flow prior to development of deep open-cut or underground mining
5
6
Regolith Science
of a primary mineralisation. In others, the effect of weathering has been to upgrade otherwise uneconomic primary mineralisation and mining ceases at the weathering front (Butt and Scott 2001).
1.8
TERMINOLOGY
Regolith terms are defined in Appendix 1 (compiled from the glossary of Eggleton 2001 and expanded to include relevant terms from the biological sphere of the regolith). It should also be noted that the terms ‘Fe oxides’, ‘Mn oxides’, and so on include both oxides and oxyhydroxides of the appropriate element.
1.9 REFERENCES (INCLUDES REFERENCES FROM PREFACE) Brantley SL, Goldhaber MB and Ragnarsdottir KV (2007). Crossing disciplines and scales to understand the Critical Zone. Elements 3, 307–314. Butt CRM and Scott KM (2001). Geochemical exploration for gold and nickel in the Yilgarn Craton, Western Australia – an introduction. Geochemistry: Exploration, Environment, Analysis 1, 179–182. Butt CRM and Zeegers H (Eds) (1992). Handbook of Exploration Geochemistry Volume 4, Regolith Exploration Geochemistry in Tropical and Subtropical Terrains. Elsevier, Amsterdam. Clarke JDA (2003). The nature of regolith: a planetary scale perspective. In Advances in Regolith. Proceedings of the CRC LEME regolith symposium 2003. (Ed. IC Roach) pp. 74–77. CRC LEME, Perth. Coram JE, Dyson PR, Houlder PA and Evans WR (2000). ‘Australian groundwater flow systems contributing to dryland salinity’. Report for the National Land and Water Resources Audit, Bureau of Rural Sciences, Canberra. Cremeens DL, Brown RB and Huddleston JH (Eds) (1994). Whole Regolith Pedology. Special Publica-
tion 34. Soil Science Society of America, Madison, Wisconsin Dent D, Lawrie K and Munday T (1999). Running down the salt in Australia I: A multi-disciplinary approach. The Land 3, 179–198. Eggleton RA (Ed.) 2001. The Regolith Glossary: Surficial Geology, Soils and Landscape. CRC LEME, Canberra and Perth. Evelyn J (1679). Sylva or a Discourse of Forest Trees. Royal Society, London. Guglielmin M, Cannone N, Strini A and Lewkowicz AG (2005). Biotic and abiotic processes on granite weathering landforms in a cryotic environment, Northern Victoria Land, Antarctica. Permafrost and Periglacial Processes 16, 69–85. Kauranne LK, Salminen R and Eriksson K (Eds) (1992). Handbook of Exploration Geochemistry Volume 5, Regolith Exploration Geochemistry in Arctic and Temperate Terrains. Elsevier, Amsterdam. Merrill GP (1897). A Treatise on Rocks, Rock-Weathering and Soils. Macmillan, London. Ollier CD and Pain CF (1996). Regolith, Soils and Landforms. John Wiley and Sons, Chichester, UK. Stierman DJ and Healy JH (1984). A study of the depth of weathering and its relationship to the mechanical properties of near-surface rocks in the Mojave Desert. Pure and Applied Geophysics 122, 425–439. Taylor G and Eggleton RA (2001). Regolith Geology and Geomorphology. John Wiley and Sons, Chichester, UK. Wilford J, Dent D, Braaten R and Dowling T (2001). Running down the salt in Australia 2: Smart interpretation of airborne radiometrics and digital elevation models. The Land 5, 79–100. Wilson MJ (2004). Weathering of the primary rockforming minerals: processes, products and rates. Clay Minerals 39, 233–266.
2
Regolith through time Brad Pillans
2.1 INTRODUCTION Geological evidence suggests that many parts of the Australian continent have experienced sub-aerial exposure (that is, they have been above sea level) over hundreds of millions of years (for example, BMR Palaeogeographic Group 1990; Figure 2.1). Consequently, there has been a long and complex history of weathering and landscape development, some of which occurred under climates quite different from the present. Dating of regolith materials provides the chronological framework for unravelling this complex history, but reliable numerical estimates of regolith age – using isotopic dating techniques – can be difficult to obtain. Typical problems include the lack of suitable minerals and uncertainties regarding assumptions about closed systems. Furthermore, the generally non-fossiliferous nature of much of the Australian regolith has meant that traditional methods of biostratigraphic dating cannot always be employed. Despite the above problems, a number of dating methods have been successfully applied to the Australian regolith (Pillans 1998, 2005). For timescales of less than 105 years, radiocarbon and thermoluminescence have been extensively used to provide a robust chronology for late Pleistocene regolith. For timescales greater than 105 years, paleomagnetism, oxygen isotopes, K/Ar (including 40Ar/39Ar) and stratigraphic
dating are well-established techniques, while, more recently, cosmogenic nuclides, U-series and (U-Th)/ He methods have been used. The age ranges over which various regolith dating techniques can be applied are summarised in Figure 2.2. Ages of geological materials, including regolith, are usually reported in one of two ways – a numerical age (for example, 100 Ma) and/or a stratigraphic age (for example, Early Cretaceous). Numerical ages can be converted into stratigraphic ages, and vice versa, using the International Geological Time Scale (Figure 2.3). Note that all ages are actually age estimates, with an associated uncertainty that should always be quoted (for example, 100±5 Ma).
2.2 SURVIVAL OF ANCIENT REGOLITH AND LANDFORMS IN AUSTRALIA As shown in Figure 2.1, parts of the Australian continent may have been sub-aerially exposed for hundreds of millions of years. Thus, while the present landscape is being shaped by modern erosional, depositional, weathering and tectonic processes, it invariably contains landforms and regolith that have developed under different conditions in the past. The famous American geomorphologist, William Thornbury, stated (1954) that ‘little of the earth’s topography is
8
Regolith Science
Arnhem Plateau (Mesozoic)
MAXIMUM DURATION OF CONTINUOUS SUBAERIAL EXPOSURE
Kimberley High Surface (Precambrian)
Permian paleokarst
Cambrian river terraces
X XX
Mesozoic
Permian glacial tunnel valleys
Paleozoic Precambrian
Arcoona Plateau (Cretaceous)
X
X
Cenozoic
Mesozoic clays (oxygen isotopes)
X X
X X Mesozoic weathering profile under Jurassic basalt
Pre-Cenozoic deep oxidation (paleomagnetism)
Jenolan Caves (Carboniferous)
Permian glacial striated pavements
Mesozoic (?) weathering profiles in granite
Figure 2.1: Maximum duration of sub-aerial exposure (BMR Palaeogeographic Group 1990). Pre-Cenozoic regolith and landforms – including Mesozoic clays dated by oxygen isotopes (Bird and Chivas 1988, 1989) and oxidised saprolite dated by paleomagnetism (Pillans 2005) – are consistent with the long history of sub-aerial exposure in several regions.
Age (yrs)
2
10
3
10
4
10
5
10
6
10
10
7
10
8
14
C (radiocarbon) 40
39
36
26
K/Ar, Ar/ Ar 10
Be, Cl, Al (cosmogenic)
Numerical Age
Luminescence (TL, OSL) U-Series Electron spin resonance Fission track (U-Th)/He, U/Pb
Calibrated Age
Weathering rinds Amino acid racemization Oxygen isotopes
Correlated Age
Fossils (esp. pollen and spores) Paleomagnetism Weathering stage
Relative Age
Geomorphic position Stratigraphy
Figure 2.2: Regolith dating techniques, showing the age ranges over which each method is applied.
Regolith through time
Phan.
Age (Ma)
SYSTEM PERIOD
251
Neogene Paleogene
542
Cretaceous
65.5
23.0
SERIES EPOCH Pleistocene Pliocene
65.5
Paleozoic
146
Neoproterozoic
Holocene 0.01 1.8 5.3
Miocene
Jurassic 200
Proterozoic
1000
Triassic
23.0 251
Mesoproterozoic
Permian 299 1600
Oligocene
Carboniferous
33.9 359
Devonian
Paleoproterozoic
416 444
Silurian Ordovician 2500
Neoarchean
Mesoarchean
488
Cambrian 55.8
Ediacaran
3200
630
Paleocene 65.5
Paleoarchean Eoarchean
3500 3600
“Early Earth”
Archean
Eocene
542 2800
Hadean
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
EONOTHEM ERATHEM ERA EON Cenozoic Mesozoic
“Precambrian”
Geon
3850
4550
Figure 2.3: Major subdivisions of the International Geological Time Scale (after Gradstein et al. 2004).
older than Tertiary and most of it no older than Pleistocene’. However, there is now abundant evidence to the contrary, particularly in Australia. Twidale (1976) discussed the factors that favour the survival of ancient features in the landscape, including such things as resistant rocks, low rainfall and tectonic stability. In essence, any factors which contribute to low erosion rates will allow ancient regolith and landforms to be preserved. Furthermore, as pointed out by Twidale (1976), erosion can be quite localised, so that ancient features will only be preserved in certain favourable parts of the landscape. A good example is the incision of rivers into otherwise low-relief landscapes, resulting in relict upland surfaces away from
the incising rivers. In Central Australia, plateaux and mesas, capped with resistant silcrete or ferricrete – materials that themselves may have been formed on transported regolith in low-lying parts of the landscape – are typical end-products of this process (Figure 2.4). Glacial scouring is often suggested as a very effective means of removing weathered regolith, as occurred across large areas of northern Europe and North America during the Quaternary. Thus, the limited areal extent of Quaternary glaciation in Australia is another reason why ancient regolith and landforms may have been preserved and, in combination with the other factors outlined above, means that Australia is an ideal laboratory in which to study
9
10
Regolith Science
Figure 2.4: Ferricrete-capped mesas north of Marla, northern South Australia. Paleomagnetic dating of the ferricrete yields an age of around 60 Ma by comparison with the Australian Apparent Polar Wander Path (Schmidt and Clark 2000), a similar age to the Morney weathering profile in the Eromanga Basin (Idnurm and Senior 1978).
regolith and landform evolution on timescales of millions of years. The locations of Australian regions that have been studied are shown in Figure 2.5. Finally, ancient regolith and landforms can be preserved through burial by younger sediments. If the sediments are later eroded, the older materials will be re-exposed at the surface. Indeed, such exhumation has occurred in the heavily glaciated terrain of southern Scandinavia, where exhumed landforms of Cretaceous age have been described by Lidmar-Bergström (1989). Burial and exhumation have also played a significant role in the preservation of ancient features in the Australian landscape (see below).
2.3 MEASURING LONG-TERM DENUDATION RATES The term denudation refers to progressive lowering of the landscape by the combined actions of the various erosional processes that include stream incision, slope
wash, mass movement, glacial and wind erosion. That denudation is spatially and temporally variable in landscapes has long been recognised as a fundamental control on landform morphology and evolution. However, long-term rates of denudation are notoriously difficult to measure because of the general slowness of the processes and their effects in the landscape. Four main methods are used: sedimentary accumulations in basins, Cenozoic lava flows, cosmogenic isotopes and apatite fission track and U-Th/He thermochronology. 2.3.1 Sediment accumulation in basins In the same way that modern denudation is calculated from sediment accumulation in dams, past denudation rates can be calculated using sediment volumes in sedimentary basins. Two major uncertainties are the trapping efficiency of the basin and the area from which the sediment was derived. Using this method, Killick (1998) calculated a denudation rate of around 9 m/Ma for the West Australian Shield (Pilbara and Yilgarn
Regolith through time
X
Darwin
X
Arnhem Land
Groote Eylandt
X
KIMBERLEY PLATEAU
PILBARA CRATON
Lancefield YILGARN CRATON PERTH BASIN COLLIE BASIN
Perth
X X
Davenport Hughenden L. Lewis Range Alice Springs
Yandi
L. Amadeus
Marla
LAKE EYRE BASIN
EUCLA BASIN
X X
EROMANGA BASIN
Brisbane
Trinity Well
Notrab L. Lefroy
Phanerozoic Basins
X X
Mt Tabor
MUSGRAVE RANGES OFFICER BASIN
Paleozoic Tasman Fold Belt
Cairns
Mt Isa
HAMERSLEY Meekatharra
Precambrian shields
X
Tennant Ck
Tanami mine CANNING BASIN
Great Barrier Reef X
GAWLER
Northparkes
CRATON
Great Escarpment
Jenolan Caves
Adelaide Kangaroo Is L. Tyrell
Sydney TASMAN
Melbourne
Wilsons Promontory
SEA
Figure 2.5: Location of sites mentioned in the text, in relation to the major geological domains in Australia.
Cratons; Figure 2.5), from the volume of sediment accumulated in adjacent basins, between the early Ordovician and the end of the Cretaceous. Similarly, Bishop (1985) calculated a Tertiary erosion rate of 3 m/ Ma for south-eastern Australia from the volume of sediment accumulated in the Murray Basin (Figure 1.1). 2.3.2 Cenozoic lava flows as dated reference surfaces In eastern Australia, the wide distribution of K/Ar-dated Cenozoic basaltic lavas has enabled estimates of long-term denudation (Young 1983; Bishop 1985; Stephenson and Coventry 1986; Young and McDougall 1993; Nott et al. 1996; Young and Wray 2000), with rates depending on lithology, relief and stream size: 1. vertical stream incision rates in the range 1–50 m/Ma 2. slope retreat/valley widening rates in the range 10–250 m/Ma 3. headward retreat of gorges in the range 1000–5000 m/Ma.
In essence, this method of calculating erosion rates relies on knowing the age of a lava flow, and the amount of material eroded since the lava was extruded. For example, at Porcupine Gorge, near Hughenden in north Queensland (Figure 2.5), Galah Creek has incised some 40 m since basaltic lava (K/Ar dated at 0.89 Ma) flowed down the paleovalley floor of the creek almost 1 million years ago (Figure 2.6; Stephenson and Coventry 1986). Gibson (2007) has compiled all K/Ar ages from such rocks in Eastern Australia. 2.3.3 Cosmogenic isotopes The Earth is being continuously bombarded by high-energy cosmic rays, principally protons originating from supernova explosions in our galaxy. When these primary cosmic rays pass through the Earth’s atmosphere they produce secondary cosmic rays, including neutrons and muons, which can penetrate many metres into rock. Nuclear interactions between these high-energy particles and elements such as Si and O in rocks, produce long-lived radioactive nuclides (such as 10Be, 26Al, 36Cl) and
11
12
Regolith Science
Figure 2.6: At Porcupine Gorge in North Queensland, Galah Creek has incised about 40 m since the Twins Flow basalt (arrowed, 0.89 Ma) was erupted (Stephenson and Coventry 1986), at a mean vertical incision rate of 44 m/Ma, compared with a mean erosion rate of about 1 m/Ma on the adjacent plateau surface based on cosmogenic nuclide measurements (Pillans unpublished data).
rare stable nuclides (such as 21Ne) in minerals – collectively referred to as cosmogenic nuclides. The concentration of cosmogenic nuclides in the upper layers of long-exposed rock surfaces depends on the balance between nuclide production rate, radioactive decay and erosional loss, expressed in the equation E=
1 m ^ P / N - lh
(Eqn 2.1)
where E = surface erosion rate (cm/year), P = production rate (atoms/g/yr), N = nuclide concentration (atoms/g), l = radioactive decay constant, µ = attenuation length of cosmic rays in rock (Fifield 1999). Erosion rates, calculated from cosmogenic isotope measurements for bare rock surfaces in Australia, vary with lithology, topography and rainfall. The lowest rates (less than 0.5 m/Ma) occur on silcrete and
quartzite plateaux, and the crests of low granite inselbergs in regions that currently receive less than 400 mm/year rainfall (for example, Bierman and Caffee 2002; Belton et al. 2004; Fujioka et al. 2005). Rates are typically one to two orders of magnitude greater on sloping surfaces, under soil cover, along streams, on less-resistant lithologies and/or where rainfall is higher (for example, Weissel and Seidl 1998; Heimsath et al. 2000, 2001; Wilkinson et al. 2005). 2.3.4 Apatite fission track and (U-Th/He) thermochronology Regional-scale patterns of denudation on timescales of the order of ten million years or more can be estimated using apatite fission track data and (U-Th/He) ages, which are primarily controlled by thermal history. Fission tracks in apatite undergo partial annealing in the temperature range 60–120°C, while
Regolith through time
the partial retention zone for (U-Th/He) is 40–80°C. Thus, apatite fission track and (U-Th)/He ages record the time since a sample was subjected to temperatures in those ranges. A measure of long-term denudation is made by converting temperature history into an equivalent depth history, using estimates of past geothermal gradient (compare Kohn et al. 2002; Persano et al. 2005). While there may be uncertainties about past geothermal gradients, these methods indicate that over the past 250 Ma, up to 4 km of material have been eroded from large areas of the Australian continent, at rates that are generally in the range 1–40 m/ Ma (Kohn et al. 2002).
2.4 PRECAMBRIAN REGOLITH AND LANDFORMS As indicated in Figure 2.1, several regions of Australia may have been continuously sub-aerially exposed
since the Precambrian (that is, for more than 540 Ma). Given this astonishingly long history of sub-aerial exposure, it is worth recalling that Precambrian landscapes would have developed under conditions that differed substantially from the present, including an absence of land plants and animals, and a lower atmospheric O2 content. The Earth is unique among planets of the solar system in that it has an atmosphere containing abundant O2 which sustains, and is also sustained by, plant and animal life (see Chapter 14). Atmospheric O2 also plays a key role in rock-weathering processes. However, the composition of the atmosphere, particularly its O2 content, has changed dramatically through the Earth’s history (Figure 2.7), with major implications for the composition of regolith materials. For example, the presence of detrital grains of pyrite, uraninite and siderite (minerals that are not stable in oxidising environments of today) in Archean fluvial sediments from
Ocean chemistry -
+ 4
NO3
NH
2-
Fe2+
SO4 (surface) 2-
S (deep)? 2-
SO 4 (deep)
Oxygen level, log10(pO2) (bar)
0
0
-1
Animals Charcoal
-2
-2
Beggiatoa Detrital siderite
-3
-1
-3
-4
-4 Sulfur MIF (model+data)
-5
-5
-6
-6
-13
-13 Prebiotic (model)
-14 4.4
-14 3.2
3.0 2.8 2.6
2.4
2.2 2.0
1.8
1.1 0.8 0.6
0.4 0.2
Time before present (Ga) Ozone layer Prokaryotes Cyanobacteria BPf006-07
Eukaryotes Animals
Figure 2.7: The history of atmospheric oxygen (after Catling and Claire 2005).
0.0
13
14
Regolith Science
the Pilbara region of Western Australia (Rasmussen and Buick 1999) indicates that the Archean atmosphere was essentially anoxic (that is, it contained only trace amounts of O2). Another indicator of less-oxidising conditions in the Archean, comes from Fedepleted paleosols in which soluble Fe2+ was leached by percolating anoxic rainwater (Rye and Holland 1998). Further evidence for anoxic conditions in the Archean comes from the widespread deposition of banded iron formations (BIFs) – laminated marine sediments composed of alternating magnetite (Fe-rich) and cherty (Fe-poor) layers. A major increase in atmospheric O2 occurred in the early Proterozoic, around 2300–2400 Ma (Figure 2.7), largely driven by the increase in oxygenic photosynthesis by cyanobacteria (blue-green algae) in surface ocean waters (Catling and Claire 2005). After this time, continental redbeds (aeolian and fluvial sediments containing Fe-oxide coated grains) make their appearance (Eriksson and Cheney 1992), and BIFs decline in abundance to finally disappear circa 1800 Ma, though the level of atmospheric O2 was still probably only some 1–3% of what it is today (Catling and Claire 2005). Another major increase in atmospheric O2 occurred at the end of the Proterozoic – circa 600–800 Ma – during which O2 rose to more than 5–18% of present levels (Canfield and Teske 1996). The cause, or causes, of this rise remain unclear, but it was accompanied by a dramatic increase in animal life and also coincided with the widespread occurrence of low-latitude glacial deposits – the so-called ‘Snowball Earth’. Glacial deposits of this age occur in the Kimberley region of Western Australia where they overlie remnants of a dissected high erosion surface (the High Kimberley Surface) that is cut across folded Proterozoic rocks (Ollier et al. 1988). The Neoproterozoic (circa 700 Ma) glacial deposits are in valleys cut into softer rocks between the ridges on which the High Kimberley Surface is developed. Glacially striated pavements, but little sediment, occur on the ridges and Ollier et al. (1988) concluded that the High Kimberley Surface had persisted at, or near, the surface since its inception in the Proterozoic. In contrast, apatite fission track thermochronology indicates kilometre-scale denudation in the Kimberley region since the Late Paleozoic leading Kohn et al. (2002) to conclude that the Kimberley High Surface is an exhumed feature.
2.5 CONTINUOUS EXPOSURE OR BURIAL AND EXHUMATION? Of the all the apparently long-exposed regions shown in Figure 2.1, the Western Australian Shield has long been identified as one of the most ancient landscapes on Earth (for example, Jutson 1914). The shield is made up of Archean and Proterozoic rocks, some of which are more than 4 billion years old (Froude et al. 1983; Wilde et al. 2001), making them among the oldest known rocks on Earth. However, there is little in the way of younger sedimentary cover except for infaulted Permian sediments in the Collie Basin (Figure 2.5) and isolated Permo-Carboniferous glacial deposits along the eastern margin (Eyles and de Broekert 2001). King (1950) suggested that the Yilgarn Craton may have been a great plain since the Late Paleozoic, while Finkl and Fairbridge (1979) concluded that the present land surface was only a few metres below the near horizontal, sub-Proterozoic uncomformity, but in neither case did the authors support their claims with evidence from regolith dating. More recently, direct evidence of long sub-aerial exposure has come from open-pit gold mines at Meekatharra, where Late Carboniferous and Late Cretaceous weathering imprints have been dated by paleomagnetism at Meekatharra in the north of the Yilgarn Craton (Figure 2.5; Pillans 2005, 2007). On the other hand, there is a body of evidence that supports kilometre-scale denudation since the Paleozoic, including apatite fission track thermochronology (Kohn et al. 2002; Weber et al. 2005) and sediment budget calculations for adjacent sedimentary basins (van de Graaff 1981; Killick 1998). A thermal model that can reconcile all of the above observations was suggested by Weber et al. (2005): 1. Surface exposure and weathering in the Late Carboniferous. 2. Rapid burial by about 3 km of Permian sediments. 3. Slow erosion of the Permian cover until re-exposure (exhumation) of the sub-Permian weathering profile in the Late Cretaceous. The thick Permian cover would also explain the lack of Archean-age detrital zircons in late Paleozoic and younger sediments of the adjacent Perth Basin (Sircombe and Freeman 1999; Cawood and Nemchin 2000).
Regolith through time
Another area for which a long continuous sub-aerial exposure has been postulated is the Davenport Range area of the Northern Territory (Figure 2.5). There, the highest land surface – the Ashburton Surface – is cut across folded Proterozoic rocks and was considered to be Cretaceous or older based on regional stratigraphic correlations (Hays 1967). Subsequently, Stewart et al. (1986) reported fluvial sediments of Cambrian age in paleovalleys that are cut into the Ashburton Surface in the Davenport Range, suggesting that the Ashburton Surface is Cambrian or older. Stewart et al. (1986) concluded that the Cambrian river terraces and adjacent Ashburton Surface represented the oldest known persisting landforms in the world, and attributed their survival to ‘marked tectonic stability’ in the region. However, more recent work by Belton et al. (2004), who used apatite fission track thermochronology, indicates that the region underwent kilometre-scale burial and exhumation before, and during, the Mesozoic, and that the Cambrian terraces are exhumed features. Belton et al. (2004) also calculated long-term (10 million year timescale) erosion rates, based on cosmogenic nuclides, of about 0.3 m/Ma for quartzites on the Ashburton Surface, and 2–4 m/Ma for the valley-fill terraces, and concluded that this order of magnitude difference in erosion rates between the ridge tops and the Cambrian terraces was consistent with exhumation of a paleovalley-fill. In summary, it seems extremely improbable that Cambrian and Precambrian regolith and landforms will survive continuous exposure at or near the Earth’s surface. However, burial and exhumation allow such features to be preserved in the modern landscape.
2.6 PERMO-CARBONIFEROUS INHERITANCE IN THE AUSTRALIAN LANDSCAPE During the late Carboniferous and Early Permian (circa 320–280 Ma), Australia was part of the Gondwana supercontinent, which included Antarctica, India, Africa, New Zealand and South America. Gondwana was situated at mid to high latitudes in the Southern Hemisphere, and was extensively covered by large continental ice sheets. Evidence for glaciation is
widespread and includes glacial tills and striated pavements in all states of Australia (Crowell and Frakes 1971), as well as the other Gondwana continents. A particularly well known example of a Permian striated pavement is at Hallett Cove, near Adelaide in South Australia, where large Permian glacial erratic boulders also lie on the present beach (Figure 2.8). Although the timing, character and distribution of glacial events and deposits is debated (for example, Jones and Fielding 2004), it is possible that a large ice sheet – possibly several kilometres thick (like the modern Antarctic ice sheet) – was centred over the Yilgarn Craton in Western Australia (Crowell and Frakes 1971; BMR Palaeogeographic Group 1990). Glacial till and tunnel valleys, dating from this time, are preserved on the eastern margin of the craton (for example, at Lancefield), where glacial melt water drained into the Officer Basin (Eyles and de Broekert 2001). Relict Early Permian landforms, including icescoured channels, U-shaped valleys, rock drumlins and striated pavements, are also preserved along the northeastern margin of the Pilbara Craton (Playford 2001). In contrast, the evidence in eastern Australia (Jones and Fielding 2004) indicates discrete, short-lived episodes of localised mountain glaciation, with substantial non-glacial intervals in between. Indeed, by the late Early Permian (circa 280 Ma), there was extensive development of coal measures (Greta Coal Measures) in the Sydney Basin, including bauxitic weathering profiles (Dickins 1996). Bauxitic weathering is often associated with tropical climates, but Taylor et al. (1992) described Early Tertiary lateritic and bauxitic weathering profiles from southern Australia that formed under a wet cool to cold climate when the region lay at about 60°S latitude – these may be an analogue for the Permo-Carboniferous bauxites. Further evidence that significant areas of Australia must have been ice free – at least for long periods (millions of years) during the major interval (280–320 Ma) of Gondwana glaciation – comes from paleomagnetic dating of thick (greater than 60 m) weathering profiles in the Tanami region, Yilgarn Craton and Northparkes mine (New South Wales) (Figure 2.5; O’Sullivan et al. 2000; Pillans 2005), indicating widespread deep oxidation of the regolith. O’Sullivan et al. (2000) used a combination of apatite fission track thermochronology and
15
16
Regolith Science
Figure 2.8: Exhumed Permian glacial erratic boulders on the beach at Hallett Cove, near Adelaide.
paleomagnetic results at Northparkes to show that the preservation of 320 Ma regolith is the result of burial and exhumation, not continuous sub-aerial exposure. Similarly, striated glacial pavements, such as at Hallett Cove, must be exhumed features because the striations would have been long eroded if they had been continuously exposed since Permian times. K-Ar dating of illitic clays in weathered volcaniclastics within Jenolan Caves – some 200 km ESE of Northparkes (Figure 2.5) – yields ages from 394 Ma (Early Devonian) to 258 Ma (Late Permian), with seven out of 18 ages in the range 342–335 Ma (Early Carboniferous) (Osborne et al. 2006). A zircon fission track age of 345 Ma on one sample is also consistent with the K-Ar ages (Osborne et al. 2006), which would rank Jenolan Caves among the oldest currently open cave systems in the world. The entry of the volcaniclastic sediments into the caves and cave morphology indicate that they were relatively close to the surface in the Early Carboniferous. Subsequently, the caves were buried by Permian and Triassic sediments of the Sydney Basin, to be exhumed in more recent times – a scenario very similar to that proposed for the Northparkes region by O’Sullivan et al. (2000). Exhumed paleokarst features of Paleozoic age are also reported from Wombeyan Caves in New South Wales (Osborne 1993) and the northern Canning Basin in Western Australia (Playford 2001).
In summary, Permo-Carboniferous glacial landforms, weathering profiles and caves, at or near the present land surface in diverse parts of the Australian continent, indicate a significant Late Paleozoic inheritance in the modern landscape. Interestingly, the cooccurrence of deeply oxidised weathering profiles and extensive ice sheets at high latitudes during the interval 320–280 Ma may have no modern analogue because oxidation may have been enhanced by atmospheric oxygen levels up to 50% higher than present (Berner et al. 2003) – Figure 2.7. There is also ample evidence, both in Australia (Dickins 1996) and elsewhere in the world (for example, Montanez et al. 2007), for significant glacial–interglacial climate fluctuations during this time, which means that the development of large continental ice sheets and extensive deep oxidation were probably never exactly coeval.
2.7 MESOZOIC CONTINENTAL BREAKUP, ATMOSPHERIC CO2 AND LANDSCAPE HISTORY In the Late Mesozoic and Cenozoic, progressive continental breakup of Gondwana had profound and lasting consequences for regolith/landform evolution in Australia. Rifted margins were created along the west, south and east coasts – thus establishing the broad continental outline that persists to this day.
Regolith through time
On the east coast, the spreading of the sea floor in the Tasman Sea and the separation of Australia and New Zealand between 85 and 100 Ma (Weissel and Hayes 1977) created a rifted margin consisting of an upland plateau surface that was separated from a lowland coastal plain by a steep escarpment (Ollier 1982). Much debate has centred on the geomorphic evolution of this margin, with two broad schools of thought: 1. Post-rift subsidence along the eastern margin of the pre-existing highlands resulted in westward migration of a continental drainage divide that previously lay east of the present day coast (for example, Ollier and Pain 1994). 2. Erosion of an initially high-standing rift shoulder (for example, Persano et al. 2002). Mesozoic weathering profiles (Bird and Chivas 1989) and volcanics (Nott and Purvis 1995) on the coastal plain of New South Wales could represent remnants of a downwarped surface, or indicate rapid postbreakup denudation along the coast; that is, these would be consistent with either model. However, apatite fission-track and (U-Th)/He ages from the coast are in the range 80 to 112 Ma, which is consistent with erosion of 3–4 km of crust at the time of rifting (Dumitru et al. 1991; Persano et al. 2002) and does not support the post-rift subsidence model. Stability of the continental drainage divide since 180 Ma in north Queensland (Nott and Horton 2000) and since at least 100 Ma in central New South Wales (Persano et al. 2006) also puts the subsidence model in doubt. Deeply weathered saprolite – probably of Mesozoic age – occurs throughout eastern Australia, particularly on granitic rocks (for example, Dixon and Young 1981: Hill et al. 1995; Hill 1999). In Victoria, at Wilsons Promontory (Figure 2.5), Hill et al. (1995) estimated that the Mesozoic weathering profile on granite was at least 300 m thick. Subsequent stripping of this material is recorded in the adjacent Gippsland Basin, where late Cretaceous to Oligocene sediments are dominated by kaolinite and quartz (Hill et al. 1995). The identification of Mesozoic weathering profiles is aided by oxygen isotope analyses of kaolinitic clays as follows: The Australian continent moved north across a marked latitudinal temperature gradient as a
consequence of the continental breakup of Gondwana. Mean annual air temperature (which varies broadly with latitude) is one of the major factors controlling the isotopic composition of meteoric waters, and hence the regolith minerals formed in equilibrium with them. As a result, the isotopic composition of regolith minerals in Australia has become increasingly enriched in 18O (Bird and Chivas 1988, 1989, 1993). By analysing samples from profiles independently dated by other techniques Bird and Chivas (1988, 1989, 1993) calibrated the change in isotopic composition over time, and distinguished four broad age groups of residual kaolinitic clays: post-Mid Tertiary, pre-Mid Tertiary, pre-Late Mesozoic and Permian (Figure 2.9). Their study demonstrated that regolith profiles containing clays with low d18O values (less than +15%) are widespread in Australia (Figure 2.1), and they concluded that a much greater part of the modern landscape than previously recognised may have developed in the Early and Mid Mesozoic. Furthermore, their results suggested that much of the Australian regolith formed in comparatively cold conditions, in contrast to some traditional interpretations that lateritisation and deep weathering largely occurred in tropical and sub-tropical climates (see also Taylor et al. 1992). Indeed, while Cretaceous climates were globally significantly warmer than present, there is evidence of
Age known
Age unknown
N. Queensland samples
SEDIMENTARY CLAYS (all ages) POST-MID TERTIARY RESIDUAL CLAYS
PRE-MID TERTIARY RESIDUAL CLAYS
PRE-LATE MESOZOIC RESIDUAL CLAYS
PERMIAN RESIDUAL CLAYS +7 BPf008-07
+9
+11
+13
+15
+17
+19
+21
+23
18
d Osmow (‰)
Figure 2.9: Oxygen isotope variations in the Australian regolith since the Permian (from Bird and Chivas 1989).
17
18
Regolith Science
cool to cold Cretaceous climates in southern Australia. For example, near Trinity Well in South Australia (Figure 2.5) a 2 m thick diamictite is the only known Cretaceous tillite (Alley and Frakes 2003). From a variety of lines of evidence, it is known that the CO2 concentration in the Earth’s atmosphere was significantly higher than present levels through much of the Mesozoic (up to 10 times higher) and Paleozoic (up to 16× higher) – see Figure 2.10. From a simulation model of granite weathering, Schmitt (1999) concluded that high atmospheric CO2 greatly accelerates the formation of deep kaolinitic profiles on granite, without needing climates as warm and humid as in present conditions. Thus, the widespread development of deeply weathered terrain in the Mesozoic, not only in Australia but in Europe (for example, Migon and Lidmar-Bergström 2001, 2002) and elsewhere, may have been the direct result of high atmospheric CO2. In some cases, weathering profiles show evidence of multi-stage weathering – beginning in the Mesozoic, with further modification in the Tertiary. For example, on Kangaroo Island in South Australia (Figure 2.5), a thick kaolinised weathering profile on Permian sediments is preserved beneath Jurassic basalt (Daily et al. 1974). The oxygen isotopic composition of kaolinite from the profile indicates a pre-Late CO2 x TODAY 18
Best-fit curve Error range Isotopic range
16 14 12 10 8 6 4 2 C 600
O S 500
D 400
C
P Tr
300
200
J
K 100
Age Ma Figure 2.10: Variation in atmospheric CO2 during the Phanerozoic (after Schmitt 1999).
T 0
Mesozoic age (Bird and Chivas 1993), consistent with the K/Ar age of 165–175 Ma of the overlying basalt (McDougall and Wellman 1976). However, paleomagnetic samples from the profile yield a Late Cenozoic age (Schmidt et al. 1976), and alunite bands yield K/ Ar ages of 6.2 and 12.0 Ma (Bird et al. 1990). The combined dating results therefore indicate that, although the weathering profile was originally formed in preJurassic times, ongoing weathering processes have resulted in significant subsequent modification. In the Mid Cretaceous – between 120 and 100 Ma – much of Central Australia was flooded during a series of major marine transgressions that affected some 40% of the present-day continental landmass (Frakes et al. 1987). The sediments, which have undergone little subsequent deformation, have been subaerially eroded and deeply weathered to produce a characteristic landscape dominated by extensive plains and low plateaux that therefore must be younger than 100 Ma (Figure 2.4). Thus, much of the Central Australian landscape has developed through the Cenozoic and is discussed in the next section.
2.8 FROM FOREST TO SALT IN THE CENOZOIC Global climatic changes, including the buildup of the Antarctic ice sheet since the Oligocene and northwards movement of the Australian continent, dramatically affected regolith/landform evolution during the Cenozoic. In the early Cenozoic, Australia was at mid to high latitudes, and warm, humid climates promoted widespread forest vegetation (Martin 2006). These conditions apparently favoured the mobilisation of Fe, and subsequent deep oxidation of the regolith led to precipitation of groundwater ferricretes. Originally precipitated in low lying parts of the landscape, these ferricretes now occur as resistant cap rocks on mesas throughout inland Australia (Figure 2.4). Paleomagnetic dating of ferricretes in the Eromanga Basin in south-western Queensland (Idnurm and Senior 1978) and near Marla (Figures 2.4 and 2.5) in northern South Australia (Pillans 2005) yields ages in the range 55–70 Ma (Maastrichtian to Paleocene). Deep oxidation of regolith is also recorded at many sites across Australia at this time (Pillans 2002, 2005).
Regolith through time
2.8.1 Ages of ferruginous weathering products on the Western Australian Shield Ferruginous weathering products – including nodular and pisolitic ferricretes, surface lag gravels, ferruginous mottles and oxidised saprolite – are common throughout the Western Australian Shield area (Anand and Paine 2002). However, until recently, their ages were not well constrained. Three examples are discussed in Boxes 2.1–2.3. 2.8.2 40Ar/39Ar dating of Mn oxides and alunite supergroup minerals The application of K/Ar and 40Ar/39Ar dating methods to weathered materials has been thoroughly reviewed by Vasconcelos (1999a). Two main groups of potassium-bearing secondary minerals have been successfully dated: Alunite supergroup minerals (alunite and
Box 2.1 During a paleomagnetic investigation of Permian to Cretaceous rocks in the Perth Basin, Schmidt and Embleton (1976) noted the presence of a ‘blanket remagnetisation’ of Tertiary age that they attributed to a period of regional lateritisation. At the time (1976), the age of the weathering-induced remagnetisation was estimated to be Late Oligocene to Early Miocene by comparison with the Australian Apparent Polar Wander Path (AAPWP), but more recent revisions of the AAPWP (Idnurm 1985, 1994) indicate a Late Miocene to Pliocene age (6±4 Ma). Mottled saprolite beneath bauxitic ferricrete at Jarrahdale, south of Perth, also yields a similar paleomagnetic weathering age (Pillans 2005). In a wider study, Pillans (in Anand and Paine 2002, Table 16) reported Tertiary, and pre-Tertiary paleomagnetic ages for oxidised saprolite at Bronzewing, Lawlers, Mt Percy and Kanowna Belle gold mines in the eastern Yilgarn Craton. Deeply oxidised saprolite at Meekatharra also yields Permo-Carboniferous weathering ages (Pillans 2005). Thus, the history of weathering can be traced back at least 300 million years – consistent with the long history of sub-aerial exposure (though not necessarily continuous exposure – see above).
Box 2.2 The well-known nodular and pisolitic ferricretes (including bauxites) in the Darling Ranges near Perth have long been considered to be relict features, which were formed under climatic conditions rather different from the present day – probably during the Tertiary. Pidgeon et al. (2004) determined four (U-Th)/He ages, in the range 7.5 to 10 Ma (late Miocene), for hematite/ maghemite separates from pisolitic nodules from the Morangup Hill area, about 50 km north-east of Perth.
jarosite) and cryptomelane–coronadite–hollanditegroup Mn oxides (see Sections 4.5.1 and 4.4.5 for more compositional details). A probability plot of all published K/Ar and 40Ar/39Ar ages from weathering profiles around the world (Vasconcelos 1999b) indicates that the majority of alunite–jarosite ages are younger than 20 Ma, whereas the Mn-oxide ages have a broader distribution extending back into the late Mesozoic. Formation of alunite is generally favoured by conditions of weak leaching and strong evaporation found in arid and semi-arid environments (Bird et al. 1990), whereas formation of Mn-oxides is favoured by intense leaching in humid environments (Dammer et al. 1999). Thus, the differing age distributions for the two mineral groups might reflect a global shift to greater aridity in the last 20 Ma. While such an arid shift may likely be true for Australia, it is unclear whether this was a global event. Rather, the age distribution may simply reflect the fact that alunite-group minerals are unstable in a humid climatic regime, while Mnoxides, once formed, are stable in both humid and arid climates. Long-term regional fluctuations between humid and arid climatic regimes would therefore favour longer preservation of Mn-oxides compared to alunite-group minerals. From the evidence of a number of studies (for example, Dammer et al. 1996, 1999; Feng and Vasconcelos 2001; Li and Vasconcelos 2002), the formation of Mn-oxides in Australian weathering profiles is episodic. Furthermore, at two sites in Western Australia, K/Ar ages of Mn-oxides increase with depth (Dammer
19
20
Regolith Science
Box 2.3 The Hamersley/Pilbara region of north-western Australia (Figure 2.5) contains huge deposits of Fe-ore that are largely derived from oxidation of banded iron formations (BIFs). Three main types of ore are recognised (Morris and Ramanaidu 2007): 1. Bedded iron deposits (BID), which are BIF-hosted and likely formed by a combination of hydrothermal, metamorphic and weathering processes beginning in the Proterozoic (Webb et al. 2003). 2. Detrital iron deposits (DID), which are minor hematite-goethite, colluvial/alluvial deposits largely derived from BID by erosion. 3. Channel iron deposits (CID), which are highgrade pisolitic, goethite-hematite Tertiary Fe-ore deposits up to 100 m thick occupy meandering paleochannels that are typically less than 1 km wide, but which may be up to several kilometres wide (Ramanaidou et al. 2003). Mining of the two longest paleochannel deposits, the Robe CID (150 km long) and the Yandi CID (80 km long), accounts for nearly 50% of the total iron ore production from the Hamersley Province. At
et al. 1999), suggesting that the formation of these weathering profiles did not occur during a single episode of a downwards-moving weathering front. Li and Vasconcelos (2002) established the duration of weathering in a single 8 cm wide specimen from Mt Tabor in central Queensland (Figure 2.5) – the inner band yielded well-defined 40Ar/39Ar ages of 24.7– 26.0 Ma, while the outer bands gave ages as young as 14.9 Ma. This indicates that precipitation of the Mnoxide spanned a period of around 11 Ma, at an average rate of 0.007 mm/ka. Ages of Mn-oxides from Groote Eylandt (Figure 2.5) – one of the world’s largest supergene Mn oxide deposits – appear to indicate three major episodes of weathering: a pre-Late Eocene episode (prior to 43.7±1.2 Ma), an Oligocene episode (around 30 Ma), and a Miocene episode (6–18 Ma), which are interpreted as representing three episodes of intense chemical weathering under humid climatic conditions, during which times the original sedimentary Mn minerals (of Cretaceous age) were replaced with tetrava-
Yandi, the paleochannel deposits are named the Marillana Formation, which crops out as a series of low mesas along the modern-day Marillana Creek. It is divided into three members (Ramanaidou et al. 2003; Macphail and Stone 2004): the basal Munjina Member (pebble conglomerate with clay lenses, partly carbonaceous); the middle Barimunya Member, which is the main ore horizon (clast-supported conglomerate comprising sub-angular to rounded pisolites of goethite-hematite, maghemite and fossil wood, cemented in a goethitic matrix); and the upper Iowa Eastern member (thin clay and CID units). Pollen from the basal Munjina Member indicates an Early Oligocene (around 30 Ma) age (Macphail and Stone 2004). (U-Th)/He dating of late-stage authigenic goethite in the Barimunya Member yields ages ranging from 18 Ma near the surface to around 5 Ma at depth, consistent with goethite precipitation at progressively lower levels as the water table dropped in response to increasingly arid conditions in northwestern Australia during the Late Tertiary (Heim et al. 2006).
lent Mn-oxides (Dammer et al. 1996). Results from a wider study of supergene Mn deposits in Australia (Dammer et al. 1999) appear to support the model of episodic accumulation associated with Tertiary climatic fluctuations. The prevalence of older ages for Mn-oxides (36–20 Ma) in the central part of the Yilgarn Craton – compared with those from coastal areas in the region (as young as 1.4 Ma) – may be indicative of the time when climate became too dry for Mn-oxides to form in the regolith of the inland areas. Some estimates of long-term geomorphic process rates have been made for mesa-dominated terrains by Vasconcelos and Conroy (2003) who carried out 40Ar/39Ar dating of Mn oxides and alunite supergroup minerals from weathering profiles in the Mt Isa region of Queensland (Figure 2.5). There, the highest profiles (255–275 m elevation on a mesa top) yielded ages in the range 12–16 Ma. Samples from an intermediate elevation site (225–230 m), at the base of the mesa scarp, yielded ages in the range 4–6 Ma, while those from lower elevation (200–220 m) sites yielded ages in
Regolith through time
the range 0.8 to 2.2 Ma. From these data, Vasconcelos and Conroy (2003) concluded that the relationship between age and elevation was consistent with a progressive downward migration of a relatively flat weathering front controlled by the water table, at a mean rate of 3.8 m/Ma over the past circa 15 Ma. They also concluded that the stepped topography in the Mt Isa region resulted from differential erosion of variably weathered bedrock, with an average erosion rate of about 3.3 m/Ma over the past 15 Ma, a rate that is consistent with erosion rates obtained by cosmogenic nuclide measurements in the region (for example, Stone and Vasconcelos 2000). 2.8.3 The ages and origins of silcrete in Australia In many parts of inland Australia, mesas are capped not with ferricrete, but with silcrete. Thiry and Milnes (1991) recognised two main types of silcrete: 1. Groundwater or phreatic silcrete, which preserves sedimentary structures, and generally occurs in local, topographically lower settings in the landscape. 2. Pedogenic or vadose silcrete, which displays vertical differences in structure and mineralogy that are related to infiltration and downward percolation of water. These are often seen as large botryoidal, puddingstone, ropy and lava-like masses that are laterally extensive and formed high in the landscape (Alley 1998). Silcrete also occurs in humid parts of eastern Australia, with a wide distribution from Tasmania to north Queensland (Young 1985; Webb and Golding 1998). These eastern silcretes are morphologically similar to inland groundwater silcretes, but have a lower TiO2 content (Young 1985). They are often called ‘sub-basaltic silcretes’ because they typically occur either beneath, or in close association with, Cenozoic basalts (Taylor and Smith 1975). Much debate has centred on the age and stratigraphic significance of silcretes in Australia, with postulated ages ranging from Jurassic through to Pleistocene. Early workers (for example, Woolnough 1927) generally regarded the silcretes as representing the remnants of an extensive, deeply weathered
peneplain, but it is now clear from more recent work in South Australia that silcrete formation occurred during several phases (Wopfner 1978; Alley 1998). For example, from stratigraphic evidence in the Lake Eyre Basin, Alley et al. (1999) recognised two significant phases of pedogenic silcrete development: the first between Late Eocene and Middle Miocene times, and the second in Late Miocene to Pleistocene times. They also noted that silicification was already in progress prior to the Late Paleocene, but that this may not have been widespread. In Victoria, silcretes appear to be associated only with basalts older than 2 Ma, which led Webb and Golding (1998) to conclude that silcrete formation may have been enhanced by higher groundwater silica levels under a wetter, more humid Tertiary climate. At some locations, silcretes contain fossil leaf impressions, which can be used to give an age estimate. The fossil leaves also provide valuable paleoenvironmental information (for example, White 1994). 2.8.4 The Eucla Basin The Eucla Basin, which contains up to 300 m of Tertiary marine and near-coastal fluvial sediments, represents the largest area (some 400 000 km2) of marine inundation of the Australian continent during the Cenozoic (Figure 2.5). The basin is also characterised by a number of paleovalleys that extend landward into the Precambrian Yilgarn, Musgrave and Gawler Cratons (Hou et al. 2003). In the eastern Eucla Basin, two prominent coastal sand-barrier systems – the Ooldea and Barton Ranges – were formed during Eocene marine transgressions around 39 and 37 Ma, respectively, and contain economically significant deposits of heavy mineral sands that are rich in zircon, ilmenite and rutile (Hou and Warland 2005). The sediments within the basin, including the shorelines and paleovalleys, are dated by a combination of palynology and marine microfossils (particularly foraminifera – Hou et al. 2006). The sources of heavy minerals in the Eucla Basin have been investigated by Reid and Hou (2006) by measuring U/Pb ages of zircons at the Notrab prospect in the eastern part of the basin. The zircons have ages ranging from Neoproterozoic (700 Ma) to Archean (3200), with a major peak (31 of 52 zircon
21
22
Regolith Science
grains dated) at 1189±11 Ma, and a minor peak at 1069±19 Ma (five grains). Most of the remaining grains do not cluster into well-defined age groups, but show a range of Paleoproterozoic to Mesoproterozoic ages from circa 1810–1390 Ma (Reid and Hou 2006). Comparison of this age distribution with zircon age distributions in prospective source regions (Yilgarn Craton, Musgrave Province, Gawler Craton and Albany-Fraser Orogen), indicates the Musgrave Province as the most likely source region (Figure 2.11), which is far to the west of the Notrab heavy mineral deposit. Thus, there may be the potential for future heavy mineral discoveries to the west of Notrab, although – as pointed out by Reid and Hou (2006), – their age data relates only to the zircon fraction of the Notrab deposit, and multiple source regions for the
Musgrave Province
Albany-Fraser Orogen
Reworked Archean detritus from Y ilgarn Craton
Yilgarn Craton
Gawler Craton
Notrab Prospect
1000 BPf010-07
2000
3000
Age Ma
Figure 2.11: Age distribution of zircons from Notrab Prospect, Eucla Basin, compared with predicted detrital zircon signatures of potential source regions (Reid and Hou 2006).
other heavy minerals (such as monazite, titanite, ilmenite and rutile) must be considered. 2.8.5 A brief history of aridity In the mid to late Cenozoic, the progressive development of aridity culminated with the development of the linear dunes, stony deserts (gibbers) and saline lake systems that characterise much of Central Australia today. Pollen evidence suggests that seasonal aridity may have been present in the Alice Springs area (Kemp 1976) and the Lake Eyre Basin (Alley 1998) as early as the middle Eocene, but desert-like conditions were probably not fully established until the Late Pliocene. For example, using cosmogenic 21Ne and 10Be, Fujioka et al. (2005) showed that stony deserts in northern South Australia formed 2–4 Ma. Chen and Barton (1991) showed that a change from fluvio-lacustrine to saline playa sedimentation (interpreted to represent a major arid shift in climate) in Lake Amadeus, south-west of Alice Springs, occurred around 1 Ma. However, in other lake basins (such as Lakes Tyrell and Lefroy; Figure 2.5), the change from freshwater clays to gypsum-dominated sediments appears to be significantly younger than at Lake Amadeus – probably in the range 500 to 700 ka (An et al. 1986; Zheng et al. 1998; English et al. 2001). The earlier onset of saline playa sedimentation at Lake Amadeus is attributed to it being one of the least responsive hydrologic systems in Australia because of its arid climate and low catchment/lake area (Bowler 1981). Thus, differing hydrologic thresholds at Lake Amadeus produced a differing environmental history from other lake basins. The same arid shift is also probably represented in coastal sedimentary sequences near Adelaide in which the oxide-mottled Ochre Cove Formation is unconformably overlain by calcareous Ngaltinga Clay (Pillans and Bourman 1996; Pillans 2003). The change from oxide-dominated to carbonate-dominated weathering is consistent with a significant decrease in rainfall and the onset of the modern climatic regime in that area. In each of these studies, the arid shift is dated by magnetostratigraphy, particularly the identification of the Matuyama/Brunhes paleomagnetic reversal (0.78 Ma) in the sedimentary deposits (Pillans 2003). Regional differences in the timing of the onset of aridity serve as a clear reminder
Regolith through time
that the regolith expression of paleoenvironmental changes may be time-transgressive across the continental landscape. As far as is known, from paleomagnetic and luminescence dating (for example, Hesse 2004) the longitudinal dune systems in Central Australia are entirely of Quaternary age. The same is true of aeolian dust deposits (loess) in Australia (for example, Hesse and McTainsh 2003; Hesse et al. 2003).
et al. 1996), Europe (for example, Migon and LidmarBergström 2002; Théveniault et al. 2007; Ricordel et al. 2007) and Africa (for example, Colin et al. 2005). As in Australia, the same preservation factors (such as tectonic stability, low rainfall, resistant rocks, burial by younger deposits) are important. However, it is probably true to say that favourable preservation factors have combined to allow the survival of ancient regolith and landforms over a much greater area in Australia than elsewhere.
2.9 APPLICATIONS The survival of ancient landforms and weathering profiles in Australia is commonly explained as being the result of prolonged tectonic stability, coupled with postulated low rates of weathering and erosion (for example, Twidale 1976; 2000; Gale 1992). However, while measured rates of long-term (105 –108 yr timescales) weathering (for example, Pillans 1997; Heimsath et al. 2000) and erosion (for example, Bierman and Caffee 2002; Kohn et al. 2002; Belton et al. 2004) in Australia may indeed be low by world standards, they are not low enough to explain the continuous sub-aerial survival of pre-Cenozoic landforms and weathering profiles. Even at a low mean erosion rate of, say, 1 m/Ma, more than 100 m could be lost from surfaces that formed prior to 100 Ma. Burial and exhumation must therefore be significant contributing factors in the preservation of such ancient features in the Australian landscape. In contrast, it does not seem necessary to invoke burial and exhumation to explain the widespread occurrence of relict Cenozoic landforms and regolith in Australia. Rather, their preservation can be explained as a likely response to increased aridity and low rates of tectonic deformation that characterised the late Cenozoic in many parts of Australia. Preservation would also be enhanced by the armouring effect of highly indurated silcretes and ferricretes, which are typically more resistant to erosion than surrounding rocks. While the evidence described in this chapter clearly demonstrates the widespread occurrence of ancient regolith and landforms in Australia, such features are also found on other continents, including South America (for example, Vasconcelos et al. 1994; Ruffet
2.10 REFERENCES Alley NF (1998). Cainozoic stratigraphy, palaeoenvironments and geological evolution of the Lake Eyre Basin. Palaeogeography, Palaeoclimatology, Palaeoecology 144, 239–263. Alley NF and Frakes LA (2003). First known Cretaceous glaciation: Livingston Tillite Member of the Cadna-owie Formation, South Australia. Australian Journal of Earth Sciences 50, 139–144. Alley NF, Clarke JDA, Macphail M and Truswell EM (1999). Sedimentary infillings and development of major Tertiary palaeodrainage systems of southcentral Australia. International Association of Sedimentologists Special Publication 27, 337–366. An ZS, Bowler JM, Opdyke ND, Macumber PG and Firman JB (1986). Palaeomagnetic stratigraphy of Lake Bungunnnia: Plio-Pleistocene precursor of aridity in the Murray Basin, southeastern Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 54, 219–239. Anand R and Paine M (2002). Regolith geology of the Yilgarn Craton, Western Australia: implications for exploration. Australian Journal of Earth Sciences 49, 3–162. Belton DX, Brown RW, Kohn BP, Fink D and Farley KA (2004). Quantitative resolution of the debate over antiquity of the central Australian landscape: implications for the tectonic and geomorphic stability of cratonic interiors. Earth and Planetary Science Letters 219, 21–34. Berner RA, Beerling DJ, Dudley R, Robinson JM and Wildman Jr. RA (2003). Phanerozoic atmospheric oxygen. Annual Reviews of Earth and Planetary Sciences 31, 105–134.
23
24
Regolith Science
Bierman PR and Caffee M (2002). Cosmogenic exposure and erosion history of Australian bedrock landforms. Geological Society of America Bulletin 114, 787–803. Bird MI and Chivas A (1988). Oxygen isotope dating of the Australian regolith. Nature 331, 513–516. Bird MI and Chivas AR (1989). Stable-isotope geochronology of the Australian regolith. Geochimica et Cosmochimica Acta 53, 3239–3256. Bird MI and Chivas AR (1993). Geomorphic and palaeoclimatic implications of an oxygen-isotope chronology for Australian deeply weathered profiles. Australian Journal of Earth Sciences 40, 345–358. Bird MI, Chivas AR and McDougall I (1990). An isotopic study of surficial alunite in Australia. 2. Potassium-argon geochronology. Chemical Geology 80, 133–145. Bishop P (1985). Southeast Australian late Mesozoic and Cenozoic denudation rates: a test for late Tertiary increases in continental denudation. Geology 13, 479–482. BMR Palaeogeographic Group (1990). Australia: Evolution of a Continent. Bureau of Mineral Resources, Canberra. Bowler JM (1981). Australian salt lakes: a palaeohydrological approach. Hydrobiologia 82, 431–444. Canfield DE and Teske A (1996). Late-Proterozoic rise in atmospheric oxygen concentration inferred from phylogenetic and sulfur isotope studies. Nature 382, 127–132. Catling DC and Claire MW (2005). How Earth’s atmosphere evolved to an oxic state: a status report. Earth and Planetary Science Letters 237, 1–20. Cawood PA and Nemchin AA (2000). Provenance record of a rift basin: U/Pb ages of detrital zircons from the Perth Basin, Western Australia. Sedimentary Geology 134, 209–234. Chen XY and Barton CE (1991). Onset of aridity and dune-building in Central Australia: sedimentological and magnetostratigraphic evidence from Lake Amadeus. Palaeogeography, Palaeoclimatology, Palaeoecology 84, 55–71. Colin F, Beauvais A, Ruffet G and Henocque O (2005). First 40Ar/39Ar geochronology of lateritic manganiferous pisolites: implications for the Paleogene
history of a West African landscape. Earth and Planetary Science Letters 238, 172–188. Crowell JC and Frakes LA (1971). Late Paleozoic glaciation of Australia. Journal of the Geological Society of Australia 17, 115–155. Daily B, Twidale CR and Milnes AR (1974). The age of the lateritised summit surface on Kangaroo Island and adjacent areas of South Australia. Journal of the Geological Society of Australia 21, 387–392. Dammer D, McDougall I and Chivas AR (1996). Isotopic dating of supergene manganese oxides from the Groote Eylandt deposit. Economic Geology 91, 386–401. Dammer D, McDougall I and Chivas AR (1999). Timing of weathering-induced alteration of manganese deposits in Western Australia: evidence from K/Ar and 40Ar/39Ar dating. Economic Geology 94, 87–108. Dickins JM (1996). Problems of a Late Palaeozoic glaciation in Australia and subsequent climate in the Permian. Palaeogeography, Palaeoclimatology, Palaeoecology 125, 185–197. Dixon JC and Young RW (1981). Character and origin of deep arenaceous weathering mantles on the Bega Batholith, southeastern Australia. Catena 8, 97–109. Dumitru TA, Hill KC, Coyle DA, Duddy IR, Foster DA, Gleadow AJW, Green PF, Kohn BP, Laslett GM and O’Sullivan AJ (1991). Fission track thermochronology: application to continental rifting of south-eastern Australia. APEA Journal 31, 131–142. English PM, Spooner NA, Chappell J, Questiaux DG and Hill NG (2001). Lake Lewis basin, central Australia: environmental evolution and OSL chronology. Quaternary International 83-85, 81–101. Eriksson PG and Cheney ES (1992). Evidence for the transition to an oxygen-rich atmosphere during the evolution of red beds in the lower Proterozoic sequences of southern Africa. Precambrian Research 54, 257–269. Eyles N and de Broekert P (2001). Glacial tunnel valleys in the Eastern Goldfields of Western Australia cut below the Late Paleozoic Pilbara ice sheet. Palaeogeography, Palaeoclimatology, Palaeoecology 171, 29–40. Feng Y and Vasconcelos P (2001). Quaternary continental weathering geochronology by laser-heating
Regolith through time
40Ar/39Ar
analysis of supergene cryptomelane. Geology 29, 635–638. Fifield LK (1999). Accelerator mass spectrometry and its applications. Reports on Progress in Physics 62, 1223–1274. Finkl CWJ and Fairbridge RW (1979). Paleogeographic evolution of a rifted cratonic margin: S.W. Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 26, 221–252. Frakes LA, Burger D, Apthorpe M, Wiseman J, Dettmann M, Alley N, Flint R, Gravestock D, Ludbrook N, Backhouse J, Skwarko S, Scheibnerova V, McMinn A, Moore PS, Bolton BR, Douglas JG, Christ R, Wade M, Molnar RE, McGowran B, Balme BE and Day RA (1987). Australian Cretaceous shorelines, stage by stage. Palaeogeography, Palaeoclimatology, Palaeoecology 59, 31–48. Froude DO, Ireland TR, Kinny PD, Williams IS, Compston W, Williams IR and Myers JS (1983). Ion microprobe identification of 4100-4200 M yr-old terrestrial zircons. Nature 304, 616–618. Fujioka T, Chappell J, Honda M, Yatsevich I, Fifield K and Fabel D (2005). Global cooling initiated stony deserts in central Australia 2-4 Ma, dated by cosmogenic 21Ne-10Be. Geology 33, 993–996. Gale SJ (1992). Long-term landscape evolution in Australia. Earth Surface Processes and Landforms 17, 323–343. Gibson DL (2007). ‘Potassium-argon ages of Late Mesozoic and Cainozoic igneous rocks of Eastern Australia’. Open File 193 Report. CRC LEME, Perth. Gradstein FM, Ogg JG, Smith AG, Agterberg FP, Bleeker W, Cooper RA, Davydov V, Gibbard P, Hinnov LA, House MR, Lourens L, Luterbacher H-P, McArthur J, Melchin MJ, Robb LJ, Shergold J, Villeneuve M, Wardlaw BR, Ali J, Brinkhuis H, Hilgen FJ, Hooker J, Howarth RJ, Knoll AH, Laskar J, Monechi S, Powell J, Plumb KA, Raffi I, Rohl U, Sanfilippoo A, Schmitz B, Shackleton NJ, Shields GA, Strauss H, Van Dam J, Veizer J, van Kolfschoten T and Wilson D (2004). A Geological Time Scale 2004. Cambridge University Press, Cambridge. Hays J (1967). Land surfaces and laterites in the north of the Northern Territory. In Landform Studies from Australia and New Guinea. (Eds JN Jennings
and JA Mabbutt) pp. 182–210. Australian National University Press, Canberra. Heim JA, Vasconcelos PM, Shuster DL, Farley KA and Broadbent G (2006). Dating paleochannel iron ore by (U-Th)/He analysis of supergene goethite, Hamersley province, Australia. Geology 34, 173–176. Heimsath AM, Chappell J, Dietrich WE, Nishiizumi K and Finkel RC (2000). Soil production on a retreating escarpment in southeastern Australia. Geology 28, 787–790. Heimsath AM, Chappell J, Dietrich WE, Nishiizumi K and Finkel RC (2001). Late Quaternary erosion in southeastern Australia: a field example using cosmogenic isotopes. Quaternary International 83-85, 169–185. Hesse PP (2004). Late Quaternary climates of the Australian arid zone: a review. Quaternary International 118–119, 87–102. Hesse PP and McTainsh GH (2003). Australian dust deposits: modern processes and the Quaternary record. Quaternary Science Reviews 22, 2007–2035. Hesse PP, Humphreys GS, Smith BL, Campbell J and Peterson EK (2003). Age of loess deposits in the Central Tablelands of New South Wales. Australian Journal of Soil Research 41, 1115–1131. Hill SM (1999). Mesozoic regolith and palaeolandscape features in southeastern Australia: significance for interpretations of denudation and highland evolution. Australian Journal of Earth Sciences 46, 217–232. Hill SM, Ollier CD and Joyce EB (1995). Mesozoic deep weathering and erosion: an example from Wilsons Promontory, Australia. Zeitschrift fur Geomorphologie 39, 331–339. Hou B and Warland I (2005). Heavy mineral sands potential of the Eucla Basin in South Australia. MESA Journal 37, 4–12. Hou B, Frakes LA, Alley NF and Clarke JDA (2003). Characteristics and evolution of the Tertiary palaeovalleys in the northwest Gawler Craton, South Australia. Australian Journal of Earth Sciences 50, 215–230. Hou B, Alley NF, Frakes LA, Stoian L and Cowley WM (2006). Eocene stratigraphic succession in the Eucla Basin of South Australia and correlation to
25
26
Regolith Science
major regional sea-level events. Sedimentary Geology 183, 297–319. Idnurm M (1985). Late Mesozoic and Cenozoic palaeomagnetism of Australia - I. A redefined apparent polar wander path. Geophysical Journal of the Royal Astronomical Society 83, 399–418. Idnurm M (1994). New Late Eocene pole for Australia, time-averaging of remanence directions, and palaeogeographic reference systems. Geophysical Journal International 117, 827–833. Idnurm M and Senior BR (1978). Palaeomagnetic ages of Late Cretaceous and Tertiary weathered profiles in the Eromanga Basin, Queensland. Palaeogeography, Palaeoclimatology, Palaeoecology 24, 263–277. Jones AT and Fielding CR (2004). Sedimentological record of the late Paleozoic glaciation in Queensland, Australia. Geology 32, 153–156. Jutson JT (1914). Physiographical Geology (Physiography) of Western Australia. Geological Survey of Western Australia Bulletin 61. Government Printer, Perth. Kemp EM (1976). Early Tertiary pollen from Napperby, central Australia. BMR Journal of Australian Geology and Geophysics 1, 109–114. Killick M (1998). Phanerozoic denudation of the Western Shield of Western Australia. Geological Society of Australia Abstracts 49, 248. King LC (1950). The cyclic land-surfaces of Australia. Journal of the Royal Society of Victoria 62, 79–95. Kohn BP, Gleadow AJW, Brown RW, Gallagher K, O’Sullivan PB and Foster DA (2002). Shaping the Australian crust over the last 300 million years: insights from fission track thermotectonic imaging and denudation studies of key terranes. Australian Journal of Earth Sciences 49, 697–717. Li J-W and Vasconcelos P (2002). Cenozoic continental weathering and its implications for the palaeoclimate: evidence from 40Ar/39Ar geochronology of supergene K-Mn oxides in Mt Tabor, central Queensland, Australia. Earth and Planetary Science Letters 200, 223–239. Lidmar-Bergström K (1989). Exhumed Cretaceous landforms in south Sweden. Zeitschrift fur Geomorphologie 72, 21–40. Macphail MK and Stone MS (2004). Age and palaeoenvironmental constraints on the genesis of the
Yandi channel iron deposits, Marillana Formation, Pilbara, northwestern Australia. Australian Journal of Earth Sciences 51, 497–520. Martin HA (2006). Cenozoic climatic change and the development of arid vegetation in Australia. Journal of Arid Environments 66, 533–563. McDougall I and Wellman P (1976). Potassium-argon ages of some Australian Mesozoic igneous rocks. Journal of the Geological Society of Australia 24, 87–106. Migon P and Lidmar-Bergström K (2001). Weathering mantles and their significance for geomorphological evolution of central and northern Europe since the Mesozoic. Earth Science Reviews 56, 285–324. Migon P and Lidmar-Bergström K (2002). Deep weathering through time in central and northwestern Europe: problems of dating and interpretation of geological record. Catena 49, 25–40. Montanez IP, Tabor NJ, Niemeier D, DiMichele WA, Frank TD, Fielding CR, Isbell JL, Birgenheier LP and Rygel M.C (2007). CO2-forced climate and vegetation instability during Late Paleozoic deglaciation. Science 315, 87–91. Morris RC and Ramanaidou ER (2007). Genesis of the channel iron deposits (CID) of the Pilbara region, Western Australia. Australian Journal of Earth Sciences 54, 733–756. Nott J and Horton S (2000). 180 Ma continental drainage divide in northeastern Australia: role of passive margin tectonics. Geology 28, 763–766. Nott J and Purvis AC (1995). Geomorphic and tectonic significance of Early Cretaceous lavas on the coastal plain, southern New South Wales. Australian Journal of Earth Sciences 42, 145–149. Nott J, Young R and McDougall I (1996). Wearing down, wearing back, and gorge extension in the long-term denudation of a highland mass: quantitative evidence from the Shoalhaven catchment, southeast Australia. Journal of Geology 104, 224–232. Ollier CD (1982). The Great Escarpment of eastern Australia: tectonic and geomorphic significance. Journal of the Geological Society of Australia 29, 13–23. Ollier CD and Pain CF (1994). Landscape evolution and tectonics in southeastern Australia. AGSO
Regolith through time
Journal of Australian Geology and Geophysics 15, 335–345. Ollier CD, Gaunt GFM and Jurkowski I (1988). The Kimberley Plateau, Western Australia. A Precambrian erosion surface. Zeitschrift fur Geomorphologie 32, 239–246. O’Sullivan PB, Gibson DL, Kohn BP, Pillans B and Pain CF (2000). Long-term landscape evolution of the Northparkes region of the Lachlan Fold Belt, Australia: constraints from fission track and paleomagnetic data. Journal of Geology 108, 1–16. Osborne RAL (1993). The history of karstification at Wombeyan Caves, New South Wales, as revealed by palaeokarst deposits. Cave Science 20, 1–8. Osborne RAL, Zwingmann H, Pogson RE and Colchester DM (2006). Carboniferous clay deposits from Jenolan Caves, New South Wales: implications for timing of speleogenesis and regional geology. Australian Journal of Earth Sciences 53, 377–405. Persano C, Stuart FM, Bishop P and Barfod DN (2002). Apatite (U-Th)/He age constraints on the development of the Great Escarpment on the southeastern Australian passive margin. Earth and Planetary Science Letters 200, 79–90. Persano C, Stuart FM, Bishop P and Dempster TJ (2005). Deciphering continental breakup in eastern Australia using low-temperature thermochronometers. Journal of Geophysical Research 110, B12405. doi:10.1029/2004JB003325. Persano C, Bishop P and Stuart FM (2006). Apatite (U-Th)/He age constraints on the Mesozoic and Cenozoic evolution of the Bathurst region, New South Wales: evidence for antiquity of the continental drainage divide along a passive margin. Australian Journal of Earth Sciences 53, 1041–1050. Pidgeon RT, Brander T and Lippolt HJ (2004). Late Miocene (U+Th)-4He ages of ferruginous nodules from lateritic duricrust, Darling Range, Western Australia. Australian Journal of Earth Sciences 51, 901–909. Pillans B (1997). Soil development at a snail’s pace: evidence from a 6 Ma soil chronosequence on basalt in north Queensland, Australia. Geoderma 117–128 Pillans B (1998). Regolith Dating Methods. A Guide to Numerical Dating Techniques. CRC LEME, Perth.
Pillans B (2002). Climate-driven weathering episodes during the last 200 Ma in Southern Australia. In Geoscience 2002: Expanding Horizons. Abstracts of the 16th Australian Geological Convention. July 1–5, Adelaide. (Ed. VP Preiss) p. 428. Geological Society of Australia, Sydney. Pillans B (2003). Subdividing the Pleistocene using the Matuyama-Brunhes boundary (MBB): an Australasian perspective. Quaternary Science Reviews 22, 1569–1577. Pillans B (2005). Geochronology of the Australian regolith. In Regolith Landscape Evolution Across Australia. (Eds RR Anand and P de Broekert) pp. 41–61. CRC LEME, Perth. Pillans B (2007). Pre-Quaternary landscape inheritance in Australia. Journal of Quaternary Science 22, 439–447. Pillans B and Bourman R (1996). The Brunhes/ Matuyama polarity transition (0.78 Ma) as a chronostratigraphic marker in Australian regolith studies. AGSO Journal of Australian Geology and Geophysics 16, 289–294. Playford PE (2001). The Permo-Carboniferous glaciation of Gondwana: its legacy in Western Australia. Geological Survey of Western Australia Record 2001/5, 15–16. Ramanaidou ER, Morris RC and Horwitz RC (2003). Channel iron deposits of the Hamersley Province, Western Australia. Australian Journal of Earth Sciences 50, 669–690. Rasmussen B and Buick R (1999). Redox state of the Archean atmosphere: evidence from detrital heavy minerals in ca. 3250–2750 Ma sandstones from the Pilbara Craton, Australia. Geology 27, 115–118. Reid AJ and Hou B (2006). Source of heavy minerals in the Eucla Basin palaeobeach placer province, South Australia: age data from detrital zircons. MESA Journal 42, 10–14. Ricordel C, Parcerisa D, Thiry M, Moreau MG and Gomez-Gras D (2007). Triassic magnetic overprints related to albitization in granites from the Morvan Massif (France). Palaeogeography, Palaeoclimatology, Palaeoecology 251, 268–282. Ruffet G, Innocent C, Michard A, Beauvais A, Nahon D and Hamelin B (1996). A geochronological 40Ar/39Ar and 87Rb/ 86Sr study of K-Mn oxides from
27
28
Regolith Science
the weathering sequence of Azul, Brazil. Geochimica et Cosmochimica Acta 60, 2219–2232. Rye R and Holland HD (1998). Paleosols and the evolution of atmospheric oxygen: a critical review. American Journal of Science 298, 621–672. Schmidt PW and Clark DA (2000). Paleomagnetism, apparent polar wander path and paleolatitude. In Billion-year earth history of Australia and neighbours in Gondwanaland. (Ed. JJ Veevers) pp. 12–17. Gemoc Press, Sydney. Schmidt PW and Embleton BJJ (1976). Palaeomagnetic results from sediments of the Perth Basin, Western Australia, and their bearing on the timing of regional lateritisation. Palaeogeography, Palaeoclimatology, Palaeoecology 19, 257–273. Schmidt PW, Currey DT and Ollier CD (1976). Subbasaltic weathering, damsites, palaeomagnetism, and the age of lateritization. Journal of the Geological Society of Australia 23, 367–370. Schmitt J-M (1999). Weathering, rainwater and atmosphere chemistry: example and modelling of granite weathering in present conditions, in a CO2rich, and in an anoxic palaeoatmosphere. International Association of Sedimentologists Special Publication 27, 21–41. Sircombe KN and Freeman MJ (1999). Provenance of detrital zircons on the Western Australia coastline – implications for the geological history of the Perth Basin and denudation of the Yilgarn Craton. Geology 27, 879–882. Stephenson PJ and Coventry RJ (1986). Stream incision and inferred Late Cainozoic tectonism in the Flinders River headwaters, North Queensland. Search 17, 220–223. Stewart AJ, Blake DH and Ollier CD (1986). Cambrian river terraces and ridgetops in Central Australia: oldest persisting landforms? Science 233, 758–761. Stone JO and Vasconcelos P (2000). Studies of geomorphic rates and processes with cosmogenic isotopes – examples from Australia. Journal of Conference Abstracts 5(2), 961. Taylor G and Smith IE (1975). The genesis of subbasaltic silcrete from Monaro, N.S.W. Journal of the Geological Society of Australia 22, 377–385. Taylor G, Eggleton RA, Holzhauer CC, Maconachie LA, Gordon M, Brown MC and McQueen KG
(1992). Cool climate lateritic and bauxitic weathering. Journal of Geology 100, 669–677. Théveniaut H, Quesnel F, Wyns R and Hugues G (2007). Palaeomagnetic dating of the Borne de Fer (NE France): Lower Cretaceous continental weathering. Palaeogeography, Palaeoclimatology, Palaeoecology 253, 271–279. Thiry M and Milnes AR (1991). Pedogenic and groundwater silcretes at Stuart Creek opal field, South Australia. Journal of Sedimentary Petrology 61, 111–127. Thornbury WD (1954). Principles of Geomorphology. Wiley, New York. Twidale CR (1976). On the survival of paleoforms. American Journal of Science 276, 77–95. Twidale CR (2000). Early Mesozoic (?Triassic) landscapes in Australia: evidence, argument, and implications. Journal of Geology 108, 537–552. Van de Graaff WJE (1981). Paleogeographic evolution of a rifted cratonic margin: S.W. Australia – Discussion. Palaeogeography, Palaeoclimatology, Palaeoecology 34, 163–172. Vasconcelos PM (1999a). K-Ar and 40Ar/39Ar geochronology of weathering processes. Annual Reviews of Earth and Planetary Sciences 27, 183–229. Vasconcelos PM (1999b). 40Ar/39Ar geochronology of supergene processes in ore systems. Reviews in Economic Geology 12, 73–113. Vasconcelos PM and Conroy M (2003). Geochronology of weathering and landscape evolution, Dugald River valley, NW Queensland, Australia. Geochimica et Cosmochimica Acta 67, 2913–2930. Vasconcelos PM, Renne PR, Brimhall GH and Becker TA (1994). Direct dating of weathering phenomena by 40Ar/39Ar and K-Ar analysis of supergene K-Mn oxides. Geochimica et Cosmochimica Acta 58, 1635–1665. Webb AD, Dickens G.R and Oliver NHS (2003). From banded iron-formation to iron ore: geochemical and mineralogical constraints from across the Hamersley Province, Western Australia. Chemical Geology 197, 215–251. Webb JA and Golding SD (1998). Geochemical massbalance and oxygen-isotope constraints on silcrete formation and its paleoclimatic implications in southern Australia. Journal of Sedimentary Research 68, 981–993.
Regolith through time
Weber UD, Kohn BP, Gleadow AJW and Nelson DR (2005). Low temperature Phanerozoic history of the Northern Yilgarn Craton, Western Australia. Tectonophysics 400, 127–151. Weissel JK and Hayes DE (1977). Evolution of the Tasman Sea re-appraised. Earth and Planetary Science Letters 36, 77–84. Weissel JK and Seidl MA (1998). Inland propagation of erosional escarpments and river profile evolution across the southeast Australian passive continental margin. American Geophysical Union Geophysical Monograph 107, 189–206. White ME (1994). After the Greening: the Browning of Australia. Kangaroo Press, Sydney. Wilde SA, Valley JW, Peck WH and Graham CM (2001). Evidence from detrital zircons for the existence of continental crust and oceans on the Earth 4.4 Gyr ago. Nature 409, 175–178. Wilkinson MT, Chappell J, Humphreys GS, Fifield K, Smith B and Hesse P (2005). Soil production in heath and forest, Blue Mountains, Australia: influence of lithology and palaeoclimate. Earth Surface Processes and Landforms 30, 923–934.
Woolnough WG (1927). Presidential address. Journal and Proceedings of the Royal Society of New South Wales 61, 1–53. Wopfner H (1978). Silcretes of northern South Australia and adjacent regions. In Silcrete in Australia. (Ed. T Langford-Smith) pp. 93–141, New England University Press, Armidale. Young RW (1983). The tempo of geomorphological change: evidence from southeastern Australia. Journal of Geology 91, 221–230. Young RW (1985). Silcrete distribution in eastern Australia. Zeitschrift fur Geomorphologie 29, 21–36. Young R and McDougall I (1993). Long-term landscape evolution: Early Miocene and modern rivers in southern New South Wales, Australia. Journal of Geology 101, 35–49. Young RW and Wray RAL (2000). Contribution to the theory of scarpland development from observations in central Queensland, Australia. Journal of Geology 108, 705–719. Zheng H, Wyrwoll K-H, Li Z and Powell C.M (1998). Onset of aridity in southern Western Australia – a preliminary palaeomagnetic appraisal. Global and Planetary Change 18, 175–187.
29
3
Landscape and regolith Graham Taylor
3.1
INTRODUCTION
As weathering occurs, some material is eroded and transported away from its original site, whereas other material may be altered, but still remain at its parental material’s site; that is, as a landscape develops, both transported and residual regolith form. Regolith materials, such as rock fragments, move over the landscape, but water-borne elements released by weathering of rock-forming minerals may also move below the land surface. Organisms may also operate on and below the surface (Chapter 1), so it follows that the processes that lead to regolith formation in the landscape are complex and that they will vary from place to place. In this chapter, two examples of typical landscapes are presented from which to draw some general principles of regolith landscape evolution, while trying to develop an understanding of how regolith processes may operate. The landscapes considered are imaginary, but relate closely to examples of landscape and regolith studied and visited by the author and are not atypical.
3.2
LANDSCAPE 1
This landscape is typical of many in hilly to gently rolling terrains across the globe. It is typical of regions of undulating topography in eastern Australia, where
Paleozoic rocks have weathered in a climate that has remained relatively humid throughout much of the Cenozoic (although more arid periods than at present have occurred). Landscape 1 consists of rounded concavo-convex hills separated by small creeks in valleys that converge to form mid- to large-sized streams with extensive floodplains in their middle to lower reaches. Landscape 1 is underlain by a variety of bedrock lithologies, including granite, folded shales and folded interbedded shales (slates), and sandstones (quartzites). The first, and perhaps most obvious, relationship between the rock type and landscape is that the hills form on the harder (less-weatherable) rocks. The major valley follows the strike of the folded shales simply because they are more easily weathered and hence more easily eroded. The regolith in this landscape, like most, contains both in situ and transported material – even in many of the higher parts of the landscape (Figure 3.1). The nature of the regolith is broadly shown in Figure 3.1, and Figure 3.2 illustrates its nature in more detail at the sites shown on Figure 3.1. Overall, the transported regolith is thickest on the lower hill slopes and in the valley bottom – with changes in its facies depending on landscape position and the processes acting there. The granite hill on the right (Figure 3.1) has a weathering profile typical of those developed on
32
Regolith Science
Sandstone strike ridges Tors
e d a
GT
f00 1-0 7
b
c
Granitoid bedrock
Base of regolith
Folded shales
Joints
Folded sandstone and shales Colluvium
Spring
Ferruginised colluvium
Channel deposits and buried soils
Figure 3.1: A hypothetical landscape formed on mixed geology, illustrating the operation of processes that influence the formation of landscapes and their regolith cover (see text and Figure 3. 2 for details). This is based on landscapes and regolith familiar to the author in the Eastern Highlands and western slopes of Australia.
jointed and coarsely crystalline rocks (Figure 3.2a). From the base, it has a highly irregular weathering front with the bedrock where weathering has penetrated deeply along joints, overlain by saprock in which the minerals of the granite have begun to weather – the biotite has lost Fe2+ and the feldspars have begun to alter, but the saprock is hard to break. Weathering along joint planes is more intense than in the massive rock. This grades upwards into saprolitecontaining corestones of relatively fresh granite. The saprolite is composed of quartz, kaolinite, perhaps a little bleached mica and accessory minerals. It retains the original granite fabric. It may be white or buff coloured or may contain Fe oxide mottles – particularly in the upper parts of the profile above the permanent water table. Above this, the saprolite loses so much material – and is sufficiently close to the surface – that it collapses and begins to move down-slope. This zone, in which all original rock fabrics are lost, is called by some the
mobile zone or the collapsed saprolite zone by others (see Eggleton 2001). Above the mobile zone is the soil. The soil is a welldefined layer, or layers, of material that occurs in the uppermost part of the regolith (for Australian soils see McKenzie et al. 2004). Sitting above the soil on the upper slopes of this hill are granite tors, which were formed as the saprolite and mobile zone are stripped to leave fresh granite corestones. On steepest slopes around the brow of the granite hill, the regolith thins significantly because the surface erosion is at its most effective here. The in situ portion of the regolith is eroded and, because of the relatively fast erosion, little transported regolith occurs either. Weathering occurs from the surface downwards – eating into the bedrock as it continues. Generally, the most weathered materials occur high in the in situ weathered profile and the least weathered materials are found deeper, except where physical weath-
Landscape and regolith
a
b 0
c 0
Soil - red
0
Soil - red to yellow
Collapsed saprolite
Ferruginisation from ground water
1
1
Soil
1
Aplite vein
Saprolite with corestones and joints preserved
4
5
6
Saprock and corestones Irregular weathering front
Fresh granite bedrock
3
Depth (m)
3
Colluvium from granitoid source (up hill) and from saprolite below
2
Depth (m)
Depth (m)
2
4
Saprolite
5
2
Transported regolith
3
Alluvial sequence with paleosols
4
5
unconformity 6
6
Saprock Weathering front
7
Saprolite 7
Fresh rock 8
d
Bedrock
e 0
0
Soil
Soil Transported regolith
Transported regolith 1
1
Collapsed saprolite
3
Depth (m)
Depth (m)
2
Gradational weathering front
Saprolite
2
3
Saprock 4
4
Gradational weathering front
Saprolite and saprock over slate 5
6
Bedrock
5
Weathering front over slate
GT002-07
Bedrock Figure 3.2: Five depth profiles through the regolith at various parts of Landscape 1, illustrating the variation in regolith facies that can occur in such a landscape (see text for details and Figure 3.1 for the position in the landscape).
ering is very much more intense than chemical weathering and the broken relatively fresh rocks are removed by erosion before they can be significantly chemically altered.
Around the lower slopes of the granite hill, regolith begins to thicken as erosional forces give way to deposition. The transported regolith thickens markedly and the in situ regolith may be relatively thick (Figure 3.2b)
33
34
Regolith Science
or, in other circumstances, thinner depending on the erosional/depositional history and weathering history of this part of the landscape. At Profile b in Figure 3.1 the transported regolith is thicker than the in situ regolith (Figure 3.2b). Profile b overlies folded interbedded shales and sandstones with an irregular weathering front, which reflects the great difference in weatherability of the two rocks. The sandstone (or quartzite) beds persist almost unaltered into the saprolite, while the shales are increasingly weathered from the saprock to the top of the in situ regolith. This is unconformably overlain by transported colluvial and sheet-wash debris derived mainly from the granite upslope, with some quartzite and shale (slate) fragments from the saprolite below. This grades upwards into the soil. Springs at the boundary between the granite and folded rocks transport solutes derived from chemical weathering upslope, including Fe2+. Where the springs occur, oxidising conditions result in the oxidation of the Fe2+ to Fe3+, which precipitates and cements the upper parts of the transported regolith forming ferricrete. The central valley is an alluvial plain underlain by predominantly transported regolith overlying in situ weathered shales. The shales weather to predominantly kaolinite, with illite, vermiculite or smectite deeper in the profile. Figure 3.2c shows the regolith facies present. The alluvial regolith lies unconformably on the in situ regolith and is made up of channel sands and gravels, and flood plain muds interspersed with abandoned channels filled with sand and a dark-coloured clay that infilled the oxbow-lakes on the floodplain. Laterally, discontinuous paleosols are common within the muddy parts of the sequence. A detailed explanation of the nature of flood plain paleosols and regolith can be found in Aspandiar et al. 1997 and Rivers et al. 1995. The alluvial sequence grades laterally into the hillslope colluvial mantle at the base of the hills either side of the main valley as does the underlying in situ clayey regolith on the shales. In Figure 3.1 on the left, the hills are underlain by interbedded shales (slates) and sandstones (quartzites), which yield a complex regolith of greatly varying thickness. Some of the less-weatherable quartzites form strike ridge outcrops, while the intervening slates
are more deeply weathered and form comparatively deep regolith. The quartzite ridges produced from differential weathering act as barriers to water and sediment movement, which prevents large volumes of transported regolith moving down-slope, but nonetheless some does move to form a blanket of transported regolith over in situ weathered slate profiles (Figure 3.2d and e). These slate profiles consist mainly of physically weathered fragments of slate mixed with more chemically altered material predominantly consisting of kaolinite – with or without Fe oxides. Again, on this hill the overall regolith is deeper on the hilltop than on the cusp, where erosion predominates, and then progressively thickens again around the lower slopes. The transported regolith on the lower slopes consists mainly of physically weathered blocks of quartzite and small slabs of slate in a matrix of sandy clays. The transported regolith is overlain by soil except where rocky quartzite outcrop occurs. The major processes operating to form this regolith/landscape (Figure 3.1) association are:
s s s s s
weathering (physical and chemical) surface water flow, with attendant erosion and deposition lateral movement of the regolith down-slope under the influence of gravity groundwater movement, both laterally and vertically the activities plants, animals and regolith fauna.
It is not the purpose of this chapter to discuss the details of weathering (see Chapters 5, 6 and 13), but it is clear from the discussion of Landscape 1 that weathering releases fragments of rock that can be transported over, or within, the regolith. Most of these fragments either move over the surface under the influence of gravity or by sheet-wash. Many also move within the regolith as it creeps down-slope – eventually ending up near a creek where it may be eroded and continue its down-slope movement in another direction. However, significant amounts of material are also moved by biological activity and by water infiltrating through the upper parts of the regolith. Biological activity occurs in most landscapes. Perhaps the biggest transporters of regolith materials
Landscape and regolith
are insects (such as ants, cicadas, termites; Chapter 8) that burrow within the regolith. These organisms cause regolith material to move both up and down through the profile. Some termites, for example, move large volumes of material from deep within the regolith (up to 50 m, Thiry et al. 2006) to the surface. Their galleries also allow regolith material to physically descend into their workings after they abandon their nests, which causes significant turnover of material within the regolith over a period of hundreds to thousands of years. The growth and fall of trees accounts for the turnover of the top metre to 1.5 m of regolith over time, and the penetration of roots opens holes within the regolith, which, when the tree dies, fill with regolith from higher in the profile – again causing turnover of much of the regolith. Perhaps the largest biomass on the planet is within the regolith – comprising micro- and meso-organisms (Chapters 7 and 8). These organisms are vital in such processes as mineral weathering (for example, via the acids they secrete), in the storage of elements within the regolith (such as C, P and N) and in the distribution of regolith materials (such as Fe oxides and CaCO3). Many of the micro-organisms may accumulate particular elements as they metabolise and, in this way, move elements from one part of the regolith to another (see Chapters 7 and 10; for example, Hill and Hill 2003). Infiltrating water moves elements from the upper regolith to lower parts or even out of the system as solutes. These elements may also concentrate in the groundwater and precipitate as the Eh and pH of groundwater changes while it moves through the landscape. Perhaps the best example of this is the oxidation of Fe2+ as the groundwater becomes oxygenated – either by sub-surface mixing or by coming into contact with air. Thiry et al. (2006) discuss the precipitation of SiO2 in saline waters during periods of increasing aridity and in bleached parts of the weathering profile during wetter phases of climate. Water entering the profile from the surface, or moving laterally through it, may move fine-grained particulate matter eroded from regolith. These materials (mostly clay minerals) are transported and then deposited as water evaporates – leaving the particles layered on the void surfaces through which water moved. These features are generally known as cutans
(for example, Nahon 1991) and they may make up to 5% of the regolith in weathering profiles. They are common features in regolith and testify to the movement of significant amounts of particulate matter at a microscopic scale.
3.3
LANDSCAPE 2
This landscape is a composite based on examples from the presently arid and semi-arid parts of Australia. It consists of three mesas separated by the main valley with several tributary valleys (Figure 3.3). The main valley has extensive floodplains, while the mesas are bounded by pronounced, but small, cliffs. The region is underlain by granites on the left-hand side and ancient basaltic rocks to the right, in which the granite has caused a significant metamorphic aureole. The major longitudinal valley follows the ancient basic volcanic rocks, with the river now flowing close to the low ridge following the metamorphic aureole. The low mesas on either side of the valley are both underlain by deep regolith, but capped by ferruginous lag and partly by ferricrete. The mesa on the right of Figure 3.3 is a relatively uniform land surface that slopes gently to the right. Its surface is covered by a lag of ferruginous nodules that are red to black in colour, with some minor quartz sand between them. They overlie a desert loam soil consisting of minor quartz sand mixed with significant silt-sized quartz and kaolinite and small amounts of illite and smectite. At the base of the soil is a ferruginous hardpan, or ferricrete, where the top of the clayrich mobile zone is cemented by Fe oxides. This forms a lenticular sheet below the soil, and crops out in the small cliffs surrounding the mesa (Figure 3.4a). Below the ferricrete, the profile simply grades from a kaolinite-rich, massive Fe-oxide mottled zone through basaltic saprolite down to basaltic saprock, which contains corestones of fresh basalt, and finally into fresh bedrock. The mottled zone has large irregular mottles towards the top, many of which have hard centres, and grades downward through progressively smaller mottles until they disappear altogether. From the mobile zone to fresh bedrock, the ratio of kaolinite to smectite decreases until the only clay minerals near the weathering front are smectitic. The upper
35
36
Regolith Science
Weathered metabasalts Calcrete Alluvial gravel
m
Metamorphic aureole in volcanic bedrock
Mottled saprolite v
Ferricrete Silcrete
X
Volcanic bedrock
v
+
vv
Granitic bedrock
Granitic saprock and saprolite Volcanic saprock and saprolite
Alluvium Ferruginous lag Elevation above datum (m) 140
c
120 100 80
X
+
GT
+
f00 3-0 7
a
v
X
X
+ +
v v
X
X
60 40
v
X
X m
+
X X
v
m
m
m m m mm v m m
v v
b
v
v v
v
v
v
v v
v
v
v v
Figure 3.3: Landscape 2 devised from several sites known to the author from the Yilgarn Craton WA and the Mt Isa region, Queensland.
a
Ferruginous lag Soil Ferricrete Mottled mobile zone
b
Soil Calcrete-cemented alluvial transported regolith
c
Ferruginous lag Soil Ferricrete Mottled alluvial transported regolith Silcrete in coarse alluvial regolith
Fe-mottles in saprolite Alluvial transported regolith
Alluvial regolith
Brown saprolite
x x Saprock with corestones
v
v
v
Saprolite with corestones
x
v
v
v v
v v
v
Bedrock
v
v v
+
+ x Bedrock +
+ x
+
+ +
Saprock with corestones
x + +
+
+ + + +
Bedrock
Figure 3.4: Three depth profiles through the regolith at various parts of Landscape 2 illustrating the variation in regolith facies that can occur in such a landscape (see text for details and Figure 3. 3 for the position in the landscape).
Landscape and regolith
part of the mobile zone contains minor amounts of quartz silt and sand that only persist for about 1.5 m. Anatase (TiO2) is common through the in situ weathering profile, but its content slightly increases in the mobile zone. The alluviated valley between the mesas is filled with transported regolith derived from the main valley and its tributaries and from hillslope colluvium (Figure 3.3). In this case, the erosion of the valley removed much of the in situ regolith (weathered basalts at least) – leaving only small remnants below the erosion surface. The alluvium consists of relatively coarse-grained detrital sediments up to about 10–15 m thick. The upper parts of the alluvium have been cemented by carbonates (mostly calcite and high Mgcalcite) to form calcrete lenses. Moving to the left from the alluvium is a small ridge formed by metamorphic rocks next to the granite (Figure 3.3). This ridge is present because these rocks are more difficult to erode than those either side of it, so it remains as a low topographic rise between the transported alluvial and transported colluvial regolith either side. This rise is cut by a small tributary creek leading to the main channel of the major valley. The mesas of the left hand side of Landscape 2 are some 30 m higher than the mesa on the right. They are bounded by cliffs up to 5 m high that give way to steeply inclined colluvial slopes towards the main valley. The cliffs exhibit thin outcrops of ferricrete cementing quartzose gravely sands. They are underlain by alluvial regolith in which large irregular ferruginous mottles occur. These slopes are interrupted by small 1–4 m high cliff-like outcrops of pebbly grey-coloured silcrete that trace around the slope with lens-shaped forms. The surfaces of these mesas are covered by ferruginous nodules in a sandy silty soil similar to the surface of the mesa on the right-hand side of the landscape, although some quartzose and other highly siliceous pebbles also occur on the surface of the left-hand side mesas. The mottled alluvial regolith is underlain by a 20–30 m thick alluvial sequence that is mottled with smaller mottles towards the top, but bleached for most of the section. This passes downward across an erosional surface into granitic saprock with abundant
fresh granite corestones and, eventually, through a very irregular weathering front into fresh granite. Overall, the Landscape 2 regolith relationships are apparently simple, and could be interpreted as a series of mesas representing a former land-surface incised by the contemporary stream system. However, the presence of duricrusts (ferricrete, silcrete and calcrete) in this landscape suggests other interpretations are possible. The complex regolith of the left-hand side mesas encapsulates a complex history and places constraints on how we interpret this history. Several points need to be made here:
s s s s s
to precipitate iron oxides that cement the ferricrete, Fe must be mobilised as Fe2+ Fe2+ must be moved in water to the site of its precipitation the water must encounter oxidising conditions to cause the precipitation of Fe3+ the water carrying the Fe2+ must move downhill but, it may be precipitated as Fe3+ in the capillary fringe of the water table therefore there must be an uphill source for soluble Fe2+, above the site of its eventual precipitation.
Few of these requirements are met in the left-hand mesas of Landscape 2 – thus the landscape must have been different at the time of the formation of the ferricrete. Figure 3.5a shows a scenario that may account for the conditions necessary for form the ferricrete. Bourman (1993, 1995) discusses the origins of ferricrete in Australia and provides further insights to their formation and interpretation (see also Chapter 13). These left-hand mesas also contain silcretecementing alluvial regolith just below the surface. Much the same requirements that applied to the ferricrete formation also apply to the formation of silcrete, except of course we need a source of silica in solution and a mechanism for its precipitation to cement the alluvium. Additionally, the presence of alluvial regolith on the summit of mesas indicates this region must at one time have been a valley bottom where alluviation could occur. This process of preservation of former valley sediments on hill tops is knows as relief inversion (for example, Pain and Ollier 1995).
37
38
Regolith Science
T1
N
T2
F1
F1
T3
Wet season watertable Dry season watertable F1
F1
Bedrock/regolith boundary F2
G
7 -0 05 T0
Contemporary ferruginisation Ferricrete
Figure 3 5: A landscape evolving with the development of ferricrete (see text for details). F1 is ferricrete with an alluvial matrix composed of local and distant materials (note that F1 west is topographically lower than F1 east). F2 is ferricrete composed of alluvial and locally derived F1 materials. All ferricretes are at different levels.
Thiry et al. (2006) suggest that massive groundwater silcrete forms when Si-saturated saline groundwater comes into contact with ‘sweet’ infiltrating water from rain. This results in the precipitation on silica in the saturated sub-surface. As the majority of groundwater moves through the more-permeable old-channel deposits, it is mainly here that silcrete forms. So, in Landscape 2 we have weathering basalt contribut-
ing Si-rich waters to the local groundwater system and rain falling on the floodplain – ultimately resulting in the formation of slabs and lenses of silcrete. Continued weathering and erosion leads to the situation shown in Figure 3.5b. Here the silcrete protected plains erode less than the surrounding regolith, while weathering continues in the alluvium – developing weathering profiles and bleaching and mottling
Landscape and regolith
the alluvium. The basalt hills continue to weather and erode to lower levels. Alluviation of the central valley continues. At times of high groundwater, Fe2+ moves closer to the surface where, in the capillary fringe, it can be oxidised and precipitated cementing the weathered alluvium and causing red-coloured ferruginous mottles in the upper profile (Figure 3.5c). From this point, it is easy to see how Landscape 2 develops: simply by continued weathering and erosion. It is probable that the silcrete and ferricrete ceased forming as a result of climatic drying and falling water tables. Several other points are worth noting:
s
s s
it is possible to form two ferricretes at different levels in the same landscape at the same, or very similar, times, although they cement very different regolith materials the presence of calcrete (calcite-cemented alluvium) testifies to the drying – allowing calcite to be stable in the environment the presence of quartz sand and silt on the righthand mesa surface testifies to the aeolian addition of material to this landscape. Although it is possible to identify it on the basalt-derived regolith (for example, Dickson and Scott 1998), it is not so easily identified in the alluvial regolith, even though these areas must also have received similar quantities of aeolian accession (for example, Tate et al. 2007).
3.4 SOME PRINCIPLES OF REGOLITH GEOLOGY 3.4.1 Landscapes Basically there are two types of landscapes: hills and plains. Hills owe their presence generally to the fact that hard rocks/regolith are less easily eroded than softer materials; valleys, on the other hand, form where softer materials occur in, or under, the landscape. The resultant landscapes depend on the relative weatherability of the rock/landscape and the efficiency of erosion to carry the weathered products away. The only real exception to this is tectonism, which may uplift or downwarp parts of the landscape. Summaries of the role of tectonism in the development of landscapes and regolith can
be found in Ollier and Pain (1996) and Taylor and Eggleton (2001). Plains are most commonly depositional landscapes (such as much of the Murray Basin of SE Australia, Figure 1.1; or the Mississippi Valley in the USA) or some are marine regression plains, such as those left during the Cretaceous as the sea retreated from much of the Eromanga Basin in northern and central Australia. Erosional plains are most commonly associated with continental glaciation, such as that found across much of Europe, northern Asia and North America. However, these are generally covered by a veneer of glacial and fluvioglacial sediments so they are strictly depositional plains over glacially planed bedrock surfaces. Other erosional planes do occur (such as the Tanami region of north-western Australia, Figure 2.5; Pillans 2007) but their origin is enigmatic. Smaller scale erosional plains can be seen in many landscapes where folded rocks of diverse lithology are planed off leaving surfaces where the relief is much less than the scale of variation in lithology or the tectonic structures. 3.4.2 Weathering Weathering is a process that occurs at, or near, the Earth’s surface. It can occur to hundreds of metres but more normally it is restricted to 50 m or so from the surface. It begins at the surface and progressively works downward into the bedrocks – altering them both physically and chemically. Therefore it may be expected that the most weathered products of the process are at the surface and the least weathered at some depth (see especially Chapter 6). However, in structurally and lithologically complex rocks, the age of weathering may not uniformly increase with depth. Rather, weathering preferentially occurs along joint planes and in more weatherable lithologies and, with varying climatic boundary conditions, weathering products will not show a simple correlation between age and depth. As we have seen above, chemical weathering often results in the formation of a typical profile examples of which are shown in Figures 3.2 and 3.4 (and discussed in greater detail in Chapter 6). On relatively uniform bedrock, this weathering profile typically has various facies that are rough zones parallel to the surface; however, it is not appropriate to consider these layers or zones in a stratigraphic sense as
39
40
Regolith Science
they form from alteration, not deposition. A corollary to this is that it is inappropriate to correlate weathering facies from one place to another – as the factors forming the profile will also vary over the region. Another major consideration in weathering is that once materials are near the surface they do not stop weathering, unless the environmental conditions are such that their constituent minerals will no longer chemically alter (see Chapters 4 and 6). So, weathering will continue to alter already altered minerals whether they are in situ or transported. Related to this is that the fact that the rate of weathering depends essentially on the availability of water, the amount of biota and, perhaps, temperature. The surface temperature controls the rate of weathering there, so weathering occurs faster in tropical climates than in arctic climates – except for the weathering of carbonate rocks which occurs faster in cooler climates. But, as chemical weathering also occurs at some depth – say 40 m – then the seasonal surface (climatically controlled) temperature is less important than the temperature controlled by long-term climatic trends and the geothermal gradient. Much physical weathering – such as frost weathering or stress relief jointing – occurs closer to the surface, allowing access to water that causes chemical alteration of the rocks. The availability of water controls how much soluble material can be removed from the site of weathering because weathering will cease once the water becomes saturated with solutes. So, a continued flow-through of water is essential for chemical weathering to continue. The availability of surface water also controls how much material is removed from the surface via erosion, so surfaces in wetter climates are likely to experience more erosion than those in drier climates unless vegetations growth aids in retention of the regolith. One obvious consequence of this is that iron is seasonally flushed from below the dry season water table (where conditions remain reducing) and precipitated toward the top of the wet season water table – leading to a white or pallid zone below a ferruginous or mottled zone. In aseasonal climates, the water table remains more or less at one level and the most soluble products of chemical weathering will be continually flushed and be moved from their point of production.
One last factor to consider relating to weathering is the rate at which changes occur. Generally, weathering reactions are thought to be slow. Laboratory studies indicate dissolution rates for the common rock-forming minerals (quartz and feldspars) to be of the order of 1–103 mm/Ma (Taylor and Eggleton 2001). A rock only needs one of its major minerals (say biotite) to be weathered for the whole fabric to start to collapse – thus the erosion of granite is not dependent on the much slower dissolution of quartz and feldspar. Pillans (1997) shows that basalts in north Queensland produce new soils at the rate of 0.3 m/ Ma, which is much slower than most geologists would have thought. Estimates of erosion rates for the Australian Eastern Highlands of 10 mm/ka (Persano et al. 2002) are common. Thus weathering that modifies rock is not readily seen on a human time-scale. However, recent work at Weipa, Cape York Peninsula (Figure 1.1; Eggleton and Taylor 2006) shows that changes such as the alteration of the hydration state of various regolith minerals can happen on a seasonal basis (such as around tree roots and seasonal water tables). It has also shown that bushfires can significantly alter surface mineralogy of the regolith, so it is possible to observe changes at a time scale <1 year. This suggests that the regolith, like most other Earth systems, is dynamic and continually shifting in response to environmental change. 3.4.3 Correlation and landscape evolution It is common practice to compare regolith facies (such as ferricrete) over a kilometre or larger scale and, where they are the same, to correlate them using the principles of lithostratigraphy, then, by inference, consider them to be of the same age. This has significant implications for the way landscape history is interpreted (Figure 3.6) (see also Section 2.8.1 and Chapter 14). This scenario is not necessarily correct, although the work of Idnurm and Senior (1978) shows that ferricretes in western Queensland can be chronostratigraphically correlated. Such correlations can not be made unless there are definitive age data on which to base the temporal correlations. The reasons for this are shown in Figure 3.6. Landscape 2 also illustrates how two quite different ferricretes can be formed at
Landscape and regolith
OBSERVED LANDSCAPE
1k
m
Ferricrete in alluviual regolith Correlation lines
INFERENCE
INFERRED ORIGINAL LANDSCAPE
07
6-
GT
00
Figure 3.6: The top landscape shows the observed landscape and the correlation of ferricrete across a valley the lower landscape is that commonly inferred from the observations.
the same time, so that the idea of using lithostratigraphic correlation as a surrogate for chronostratigraphic correlation is flawed. 3.4.4 The role of biota in regolith development Biological agents are important in mineral weathering (see Chapters 7 and 8). Common biological agents include tree roots that take nutrients (such as P, K, N,
Ca and Mg) and some non-nutrient elements (such as Si, Ni and Au) from the soil. In the process they aid in mineral breakdown. Similarly, organic debris, left on, and in, the regolith, decomposes – allowing these elements back into the weathering system, as well as producing acids that, in general, aid in chemical weathering. The abundance of decomposing organic matter in tropical regions is one reason weathering generally proceeds faster in the tropics.
41
42
Regolith Science
Microbes also commonly alter their environment by their metabolism, which may produce organic (such as citric or oxalic) and inorganic (such as nitric or, sulfuric) acids that then are available to dissolve minerals. Similarly, some of their metabolic products, such as polysaccharides, are important in holding regolith particles together – altering the physical nature of the materials. Non-photosynthetic microbes must oxidise their substrate to get electrons to provide energy for growth. Consequently, microbial growth will produce oxidation and the complimentary reduction within the regolith. This is particularly important with regard to Fe and Mn within the regolith and their stability (see Chapter 7). Meso-organisms, such as insects, are also important components of the regolith (see Chapter 8). Ants, termites, beetle pupa, earthworms and other insects are abundant in the regolith. All do an enormous amount of work in reorganising regolith materials and introducing organic debris, which may decompose into the regolith. Perhaps the most important insect in the Australian context in regolith are termites. Many, but by no means all, of these organisms build above-ground nests, but nearly all require water to maintain nest humidity. Most termites gather water by digging galleries down to the water table. There are records of termite galleries as deep as 50 m below the surface, and 15 m is common. This means that termites carry regolith materials from significant depths to the surface, or near surface, and so bring lessweathered regolith materials to the surface. Over time, this process is extremely significant in regolith formation. Similarly, vacated termite burrows allow material from the surface to penetrate deeply into the regolith: taking more weathered surface material into the deep regolith. Trees cycle elements through the regolith, but trees also physically bioturbate the regolith in two ways. Firstly, when trees fall, they often bring the top metre or so of regolith to the surface attached to their root bowl. This causes significant turnover of the upper regolith. Secondly, when trees die, their roots rot, producing organic acids to assist in weathering and the void left by the dead root tubule provides a conduit for surface regolith to penetrate deeper (see also Chapter 8).
3.5
CONCLUSIONS
When dealing with regolith in a landscape context, it is essential to think of weathering as an overprint on prior materials and to think of duricrusts in terms of cementing of prior regolith materials. In landscapes generally, the oldest features are the topographically highest – and the regolith associated with them may also be old – but estimating the age of regolith without reliable age determinations can be dangerously misleading (see Chapter 2).
3.6
REFERENCES
Aspandiar MF, Eggleton RA, Orr T, van Eck M and Taylor G (1997). An understanding of regolith and landscape evolution as an aid to mineral exploration - the Charters Towers experience. In Resourcing the 21st Century. Australasian Institute of Mining and Metallurgy 1997 Annual Conference. pp.125–129. AIMM, Ballarat. Bourman RP (1993). Perennial problems in the study of laterite: a review. Australian Journal of Earth Sciences 40, 387–401. Bourman RP (1995). A review of laterite studies in southern South Australia. Transactions of the Royal Society of South Australia 119, 1–28. Dickson BL and Scott KM (1998). Recognition of aeolian soils of the Blayney district, NSW: implications for mineral exploration. Journal of Geochemical Exploration 63, 237–251. Idnurm M and Senior BR (1978). Palaeomagnetic ages of Late Cretaceous and Tertiary weathered profiles in the Eromanga Basin, Queensland. Palaeogeography, Palaeoclimatology, Palaeoecology 24, 263–277. Eggleton RA (Ed.) (2001). The Regolith Glossary. CRC LEME, Perth. Eggleton RA and Taylor G (2006). Pisoliths – formed by accretion or internal reorganization. In Regolith 2006. Consolidation and Dispersion of Ideas. CRC LEME Regolith Symposium, November 2006, Hahndorf (Eds RW Fitzpatrick and P Shand) pp. 67–70. CRC LEME, Perth. Hill SM and Hill LJ (2003). Some important plant characteristics and assay overviews for biogeochemical surveys in western New South Wales. In
Landscape and regolith
Advances in Regolith Proceedings of CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp. 187–192. CRC LEME, Perth. Idnurm M and Senior BR (1978). Palaeomagnetic ages of Late Cretaceous and Tertiary weathered profiles in the Eromanga Basin, Queensland. Palaeogeography, Palaeoclimatology, Palaeoecology 24, 263–277. McKenzie N, Jacquier, D, Isbell R and Brown K (2004) Australian Soils and Landscapes: An Illustrated Compendium. CSIRO Publishing, Melbourne. Nahon DB (1991). Introduction to the Petrology of Soils and Chemical Weathering. John Wiley and Sons, Chichester, UK. Ollier CD and Pain CF (1996) Regolith, Soils and Landforms. John Wiley and Sons, Chichester UK. Pain CF and Ollier CD (1995) Relief inversion: a component of landscape evolution. Geomorphology 12, 151–165. Persano C, Stuart FM, Bishop P and Barfod DN (2002). Apatite (U-Th)/He age constraints on the development of the Great Escarpment on the southeastern Australian passive margin. Earth and Planetary Science Letters 200, 79–90.
Pillans B (1997). Soil development at snail’s pace: evidence from a 6 Ma soil chronosequence on basalt in north Queensland, Australia. Geoderma 80, 117–128. Pillans B (2007). Pre-Quaternary landscape inheritance in Australia. Journal of Quaternary Science 22, 439–447. Rivers CJ, Eggleton RA and Beams SD (1995). Ferricretes and deep weathering profiles of the Puzzler Walls, Charters Towers, North Queensland. AGSO Journal of Australian Geology and Geophysics 16, 203–211. Tate SE, Greene RSB, Scott KM and McQueen KG (2007). Recognition and characterisation of the aeolian component in soils in the Girilambone Region, north western New South Wales, Australia. Catena 69, 122–133. Taylor G and Eggleton RA (2001). Regolith Geology and Geomorphology. John Wiley and Sons, Chichester, UK. Thiry M, Milnes AR, Rayot V and Simon-Concion R (2006). Interpretation of palaeoweathering features and successive silicifications in the Tertiary regolith of inland Australia. Journal of the Geological Society, London 163, 723–732.
43
4
Regolith mineralogy Richard A Eggleton
4.1 INTRODUCTION The minerals of the regolith result either from the weathering of rock-forming minerals or are unweathered fragments of primary minerals. Igneous and metamorphic rocks form at high temperatures, under reducing conditions and at low water availability. By contrast, regolith minerals form at low temperatures under either reducing or oxidising conditions, and in the presence of much water. Their mode of formation may be by hydration and the transformation of a primary mineral or by simple precipitation from solution. Where a primary mineral weathers by hydration, the process is essentially the replacement of a cation by hydrogen. Where oxidation is involved, the main element to be affected is Fe, which is oxidised from Fe2+ to Fe3+. The minerals of the regolith can be divided into three groups:
s s
s
hydrated aluminosilicates formed by weathering primary silicates – the clay minerals precipitated minerals composed of the elements leached from primary minerals during the first process – minerals such as gypsum or halite or the Fe oxides (goethite and hematite) primary unweathered minerals, such as quartz or zircon.
There are so many encyclopaedic data sources available on mineralogy today that it is unnecessary to describe the fundamental properties of the minerals of the regolith here. Detailed, constantly updated species-by-species data can be obtained at internet databases such as http://www.mindat.org and http:// www.webmineral.com. It should be borne in mind that regolith minerals depart somewhat from ideal crystal structures. In fact, most of the clay minerals, the Fe oxides, and the silica minerals depart so far from ideality that, in some instances, they become difficult to identify unambiguously. The current International Mineralogical Association definition of a ‘mineral species’ has been carefully worded to include such complex materials (see Nickel 1995): ‘a mineral is an element or chemical compound that is normally crystalline and that has been formed as a result of geological processes’, and this definition is followed by considerable discussion of ‘crystallinity’ and exceptions! Rock-forming minerals are mostly identifiable quickly by visual examination of hand specimens or by thin-section optical microscopy. Reflectance and emission spectrometry – especially using the VNIR (visible near infra-red) and SWIR (shortwave infra-red) wavelengths and field portable instruments – is now well established as a rapid routine technique to determine
46
Regolith Science
mineralogy, as well as variations in mineral compositions, especially for Fe oxides, hydrous and carbonate minerals (Cudahy and Ramanaidou 1997; Yang et al. 1998). Electron microprobe analysis (EMPA) provides chemical confirmation of previously determined identifications, and X-ray diffraction (XRD) is commonly used for further characterisation or to identify unusual species. But, because most regolith-formed minerals occur as grains smaller than about 2 µm, neither optical microscopy nor EMPA are adequate for identification unless the minerals are present as relatively large (>10 µm) pure aggregates. Identification of the finegrained minerals of regolith almost always has to be made by using a technique requiring some interpretation of its results. Most commonly clays and Fe oxides are identified from their XRD pattern, or from their hyperspectral (SWIR and VNIR) responses – commonly using PIMA® or ASD® instruments. Scanning electron microscopy (SEM) and EMPA provide additional – often more detailed, but generally more expensive – approaches to identification. Identification by XRD or hyperspectral techniques requires a data bank of previously characterised minerals. Typically these are measured on beautifully crystallised museum or synthetic specimens of high purity. Unfortunately, most ‘real’ regolith minerals are less crystallographically and chemically ‘perfect’ than the standards, so their response to XRD or hyperspectral analysis does not always conform to the textbook response. In this chapter, the general properties of regolith minerals will be described, but the emphasis will be on how they differ from the ideal and what effect this has on identification.
4.2
ROCK-FORMING SILICATES
The silicate minerals have crystal structures in which every cation is surrounded by O anions in such a way that the cations are prevented from close approach. The number of O anions around a cation is the coordination number for that cation – and it depends on the cation size. All silicates (except a few high-pressure phases formed in the mantle or by meteorite impact, such as stishovite: the rutile-structure polymorph of quartz) have Si in four-fold coordination to O so that the O
anions are at the corners of a tetrahedron. Magnesium and Fe are dominantly in six-fold or octahedral coordination to Si, and Al – being a little larger than Si but smaller than Mg – may be found in either tetrahedral or octahedral coordination (Figure 4.1). Ca and Na may be squashed into 6-fold coordination, but are most commonly in 8-fold. K – the largest common ion – is in 12-fold coordination in silicates (Figure 4.1). The silicate structures are generally classified on the basis of the polymerisation of the silica tetrahedra (Figure 4.2). For weathering studies, however, the extent of polymerisation of octahedra becomes significant, because some silicates pass whole structural elements onto their weathered products (such as mica sheets to kaolinite sheets). It is therefore useful to consider silicates from the perspective of their octahedral cation linkages. In the pyroxenes, amphiboles and micas, the structures involve two tetrahedra on either side of an octahedron, forming a basic structural unit (I-beams of Thompson, 1978) now referred to as a TOT (Figure 4.3a). As, for example, Eggleton and Boland (1982) and Veblen and Ferry (1983) showed, the TOT units of a primary pyroxene or amphibole may link laterally to alteration product micas or clay minerals because of their TOT structural similarity (Figure 4.3b). 4.2.1 Framework silicates: (quartz, feldspars, feldspathoids and zeolites) None of the framework silicates have cations in octahedral coordination, Al occurs in tetrahedral sites and the other cations – mostly Ca, Na or K – occur in eight- to 12-fold coordination to oxygen. 4.2.2 Orthosilicates (olivine, garnet, alumino-silicates) In orthosilicates, Mg, Fe and Al occur in octahedral coordination (Mg, Fe in garnet in 8-fold); tetrahedral cations are almost exclusively Si, with little tetrahedral Al. Oxygens are densely packed (hexagonal close-packing in olivine; cubic close-packing in kyanite). 4.2.3 Amphiboles and pyroxenes The amphiboles and pyroxenes have strips of octahedra containing Mg, Fe and Al coordinated above and
Regolith mineralogy
6
12
Figure 4.1: Six- and 12- fold coordination represented as packing models, ball and spoke models and for 6-coordination – a polyhedral model.
below by strips of tetrahedra, with some Al substituting for Si in amphiboles. 4.2.4 Layer silicates: micas, chlorite, kaolinite and halloysite, smectites. The basic crystal structure of the common layer silicates was elucidated in the 1930s (Pauling 1930), and it is now recognised that all the minerals of the group have closely related structures. All have two structural units: an octahedral sheet and a tetrahedral sheet. The octahedral sheet comprises a plane of cations in octahedral coordination with planes of anions on either side. The tetrahedral sheet is formed of one plane of anions from the octahedral sheet, a plane of Si and Al cations and a plane of O anions completing the tetrahedra (Figure 4.4). In all the layer silicates, the small
Si cations occur in tetrahedral coordination to O anions, the tetrahedra being linked laterally at three of their corners to other tetrahedra in the form of a continuous hexagonal sheet (Figure 4.2f). The sheetlinking O anions are referred to as the basal oxygens. These two planes of O anions and cations are completed as polyhedra by basal oxygens of the adjacent octahedral sheet, which provide the fourth, or apical, O of the tetrahedra (Figure 4.4). A terminology for describing layer silicates has arisen from the work of the Clay Mineral Nomenclature Committee (Bailey et al. 1971). All components of the structure are planar. Many texts on mineralogy report crystallographic dimensions in Ångström units (104 Å = 1 µm) and clay minerals are commonly referred to in terms of the basal (001) X-ray diffraction
(a)
(b)
(c)
(d)
(e)
(f)
(g)
Figure 4.2: Classification of silicates according to tetrahedral polymerisation. (a) Silica tetrahedron viewed as four oxygens coordinated to a central silicon. (b) Tetrahedron viewed as a coordination tetrahedron with the oxygens at the apices. (c) Silica tetrahedron simplified. (d) Single chain polymer. (e) Double chain. (f) Tetrahedral sheet. (g) Tetrahedral framework.
47
48
Regolith Science
(a) Si tetrahedron
Al, Mg, or Fe octahedron
Si tetrahedron TOT z
(b)
y x 2:1 layer silicate
Pyroxene
Figure 4.3: (a) Tetrahedron-Octahedron-Tetrahedron Unit (TOT) as in pyroxenes, amphiboles and micas (b). TOTs linked as in pyroxenes and 2:1 layer silicates such as biotite.
Tetrahedral sheet 1:1 layer 7.2 Å thick Octahedral sheet
Basal oxygen plane Silicon plane Apical oxygen plane Octahedral cation plane Octahedral anion plane Figure 4.4: Structure of kaolinite – a 1:1 dioctahedral layer silicate
Regolith mineralogy
(a)
(b)
Figure 4.5: (a) Trioctahedral sheet. (b) Dioctahedral sheet. Octahedra are represented as in Figure 4.1.
spacing measured in Ångström units. In this chapter, Ångström units are used when referring to clay minerals and XRD data. Other conventions are:
s s
atoms are referred to as lying in planes two planes of anions with a plane of cations coordinated between them to form linked polyhedra are referred to as sheets sheets linked by common anion planes are referred to as layers.
s s
one octahedral sheet with two flanking tetrahedral sheets (2:1 layer silicates), Figure 4.6a; and, 2:1 layers with octahedral sheets between (2:2 layer silicates), Figure 4.6b.
Layer silicates are classified on two criteria. The first identifies the occupancy of the octahedral sheet. An isolated octahedral sheet, such as in the mineral brucite, [Mg3 (OH) 6], has trigonal symmetry, and a unit cell containing three Mg3 (OH) 6 octahedra. By contrast, the mineral gibbsite, [Al 2 (OH) 6] – despite also having trigonal symmetry and three octahedra in its unit cell – has only two octahedra occupied. Octahedral sheets having all three octahedra occupied are called trioctahedral; those with only two occupied are dioctahedral (Figure 4.5). The second classification criterion refers to the sequence of octahedral sheets and their flanking sheets of [SiAl] tetrahedra. There are only three known configurations for octahedral and tetrahedral sheet sequences:
A single plane of O atoms has a thickness of about 2.6 Å. An overlying anion plane fits into hollows in the first, so that the effective thickness of each plane reduces to approximately 2.3 Å. 1:1 layer silicates have three anion planes, and so are about 7 Å thick; 2:1 layer silicates with four anion planes are about 9.4 Å thick (talc) or 10 Å if there is an alkali cation in the interlayer (micas). 2:2 layer silicates are 14.4 Å to 15.4 Å thick. These measurements – and other names applied from time to time – have led to multiple terminologies for clays: 1:1 layer silicate 7-Å layer silicate kaolinite, 7-Å halloysite 2:1 layer silicate 10-Å layer silicate pyrophyllite, talc, mica, illite 2:2 layer silicate 14-Å layer silicate chlorite, smectite (formerly montmorillonite group), vermiculite Variations available to each layer type are:
s
s
s
one octahedral sheet with one flanking tetrahedral sheet (1:1 layer silicates) (Figure 4.4);
the nature of the octahedral cation – Al or Fe3+ dominate in dioctahedral sheets, and Mg or Fe2+
49
50
Regolith Science
(a)
2:1 layer 9.4 Å thick
(b)
2:2 layer 14.4 Å thick
Figure 4.6: (a) 2:1 layer silicate. (b) 2:2 layer silicate. Symbols are as in Figure 4.4.
s
in trioctahedral sheets. Mn2+, Zn2+, Mn3+, Cr3+, Ti4+ are common minor components in the octahedral sheet substitution in the tetrahedral site of Si4+ by Al3+, leading to a positive charge deficiency, which is compensated by either a large low-charged cation between 2:1 layers (micas), or by positive charge generated in the octahedral sheet by substitution of a trivalent cation for a divalent, such as Al3+ ¢ Mg2+.
In the igneous and metamorphic layer silicates, Al occurs in both tetrahedral and octahedral sites, but rarely exceeds 25% of the tetrahedral sites.
tion of Fe2+ and a change in the coordination of Al 3+ from largely tetrahedral to octahedral (see Wilson 2004 for a detailed summary of mineral weathering). Two examples illustrate these processes. When feldspar weathers, a thin amorphous layer is created through the replacement of K+, Na+ or Ca2+ by H+. The amorphous layer reorganises to a smectite, which has a 2:1 layer with an exchangeable ion in the interlayer region. This also involves a change in Al-coordination from tetrahedral in the parent feldspar to mostly octahedral in the resulting smectite. Chemically, this step in the weathering of an alkali feldspar to smectite might be expressed as: 2 ^ K, Na, Cah7 Al 1.15 Si 2.85 A O 8 + 5H 2 O =
4.3 CLAY LAYER SILICATES The weathering processes that change primary silicates into clays involve progressive hydration, oxida-
Ca 0.15 Al 2 7 Al 0.3 Si 3.7 AO 10 (OH) 2 .4H 2 O + 2.2SiO 2 + (K, Na) 2 O (Eqn 4.1)
Regolith mineralogy
When biotite weathers, the first step involves the loss of K+ and the oxidation of Fe2+. The leaching of the K+ opens the interlayer region, and Mg – probably leached from more weathered parts of the same or from an adjacent biotite – enters along with weakly attached water molecules and expands the interlayer to form vermiculite. K (Mg 2.3 Al 0.2 Fe 20.4+) 7 Si 3 Al A O 10 (OH) 2 + 4H 2 O + 0.3Mg 2 + = Mg 0.3 (Mg 2.3 Al 0.2 Fe 30.4+) 7 Si 3 Al AO 10 (OH) 2 .4H 2 O + K+ + 0.4e(Eqn 4.2) In the second step of feldspar weathering, kaolinite is commonly formed from the smectite by hydration and silica loss: Ca 0.15 Al 2 [Al 0.3 Si 3.7] O 10 (OH) 2 .4H 2 O + 0.2H 2 O = 1.15Al 2 Si 2 O 5 (OH) 4 + 1.4Si (OH) 4 + 0.15Ca (OH) 2 (Eqn 4.3) In the second step of biotite weathering, Mg and Si are leached from the vermiculite and the hydrous minerals kaolinite and goethite are formed: 5Mg 0.3 (Mg 2.3 Al 0.2 Fe 30.4+) [Si 3 Al] O 10 (OH) 2 .4H 2 O + 13H 2 O = 3Al 2 Si 2 O 5 (OH) 4 + 2FeO (OH) + 13Mg (OH) 2 + 9Si(OH) 4 (Eqn 4.4) In both of these examples, the first weathering product mineral has a 2:1 layer with a hydrated interlayer. Millot (1970) called the process ‘bisiallitisation’,
because the new clay mineral has two silica sheets. The second silicate weathering product is kaolinite, which has a single silica sheet. Millot called this step ‘monosiallitisation’. The regolith is generally oxidising, and Fe2+ is rare. Among the clays, only early formed vermiculite may carry Fe2+, and Mg varieties are restricted to weathered mafic rocks. The clay silicates of the regolith can be classified simply according to whether Al, Fe3+ or Mg is in the octahedral sheet, and by layer type (Table 4.1). Because the interlayer region is important in the properties of smectite, this family of layer silicates is here included in the 2:2 group (Table 4.1) even though the interlayer cations are transitory. Many clay minerals when first formed in the regolith are not very well organised in terms of their atomic structure, hence their crystalline character is difficult to characterise using traditional concepts. Because of this, terms have arisen to describe departure from ideality, such as ‘disordered’, ‘poorly ordered’, ‘of low crystallinity’, ‘poorly crystalline’, ‘having short-range order’ and so on. Any, or all, of these terms may be applied to clays and Fe oxides, but it is rarely clear just what is meant when the words are used. ‘Disordered’ may properly be used to refer to the random distribution of atoms in a specific structural site within a crystal, such as Al-Si disorder over the tetrahedral sites in high sanidine, or the irregular stacking sequence of carbon layers in some graphite crystals. ‘Poorly ordered’ suggests some degree of order, perhaps such as the Al-Si distribution in low sanidine or orthoclase, but it may also refer to
Table 4.1: Classification of regolith layer silicates. Layer type
Al dioctahedral
Fe3+ dioctahedral
Mg trioctahedral
1:1
Kaolinite Al2Si2O5 (OH) 4 Halloysite Al2Si2O5 (OH) 4 .2H2O
Hisingerite Fe2Si2O5 (OH) 4
Serpentine Mg3Si2O5 (OH) 4
2:1
Illite K0.9Al2 [Si3.1Al0.9]O10 (OH)2
–
Talc Mg3Si4O10 (OH)2
2:2
Montmorillonite Ca0.3Al1.8 [Si4]O10 (OH)2 .2H2O Beidellite Ca0.2 Al2 [Si3.6Al0.4]O10 (OH)2 .2H2O
Nontronite Ca0.2Fe2 [Si3.6Al0.4] O10 (OH)2 .2H2O
Saponite Ca0.2Mg3 [Si3.6Al0.4] O10 (OH)2 .2H2O Vermiculite Mg0.3 (Mg2.4Al0.2Fe0.4) [Si2.8Al1.2]O10 (OH)2 .nH2O
51
52
Regolith Science
sequences of regular layer stacking in a generally randomly stacked layer crystal. ‘Of low crystallinity’ is more difficult to interpret. It may be used to mean ‘such small crystals that the X-ray diffraction peaks are broadened’, or ‘made up of large crystals but with numerous defects’. ‘Poorly crystalline’ commonly means much the same as ‘of low crystallinity’, but just how the structure departs from perfect crystallinity is unclear. In the case of a clay silicate, the layer sequence might be irregular, the layer type may vary within a packet of layers, or the distribution of atoms or vacant sites in any sheet might be random. A perfect crystal has its atoms repeating regularly in three dimensions, with no breaks in the pattern and no irregularities. Absolute perfection is never observed in nature, but most real crystals, such as quartz or salt, conform very closely to this ideal. While all have defects, they are generally sufficiently few in number and so widely spaced that for most methods of examination (optical, X-ray diffraction or hyperpectral) their structure is indistinguishable from the ideal. It is only when the departure from ideality is sufficiently marked that it affects the observations that terms such as ‘poorly crystalline’ are used. The crystal structure of a clay mineral can be thought of in two parts. Firstly there is the layer type: essentially the gibbsite-type layer or the interlayer, the 1:1 layer (as in kaolinite) or the 2:1 layer (as in mica). How the layers stack upon each other in the z-direction of a classical crystal lattice determines one aspect of the degree of order. If the layers stack regularly – that is, if the arrangement does conform to that of an ideal crystal – then the X-ray diffraction pattern, like that of any well-organised crystal, will show sharp, discrete diffraction peaks (Figure 4.7a). If the layers are not regularly stacked, strictly speaking the arrangement is not a crystal, because there is no regularity of arrangement in the z-direction. Hence there is no direction that can be chosen as the z -axis, and hence there is no c-repeat. The arrangement would be referred to as ‘layer disordered’. Such disorder leads to a merging and a successive diminution in intensity of the X-ray diffraction peaks for reflections, such as hk0, hk1, hk2 and so on, which gives rise to a saw-tooth pattern (Figure 4.7b).
Structures such as this are commonly referred to as one-layer disordered. If the parent structure has, for example, monoclinic symmetry it may be labeled 1Md (1 layer, monoclinic, disordered) though strictly, because monoclinic refers to a three-dimensional ordered arrangement, the word is inappropriate for a layer disordered structure. Similarly the word ‘crystal’ may be inappropriate for such a layer disordered mineral, and the word ‘tactoid’ is sometimes used to refer to an individual layer silicate particle. All smectites have layer disorder – as do many regolith kaolinites and some illites. Secondly there is the layer itself. The tetrahedral sheet of a clay layer is, as far as has been determined, disordered in the disposition of Al and Si in tetrahedral sites. In illite, the Al-Si distribution is assumed to be the same as in muscovite; that is, disordered. Kaolinite has no tetrahedral Al, so the issue does not arise. Smectites yield insufficiently good XRD patterns for Si-Al order to be assessed. The location of the vacant site in the octahedral sheet of dioctahedral micas is known to be wellordered, as is the location of octahedral Al 3+ in trioctahedral micas (Brigatti et al. 2000). Conversely, Mg-Fe ordering between octahedral sites is at most a subtle effect, which may be steered in either direction by other factors (see Holland and Powell, 2006). Vacant site ordering is also well developed in kaolinites and, indeed, it is through the regularity of repeat of the octahedral vacant site in the z-direction that layer-ordered kaolinites are defined. So as far as has been determined, each individual layer of a clay mineral maintains two-dimensional regularity within itself, and for a given simple species (as opposed to interstratifications of more than one layer type), all the layers have this regularity. In short, the vast majority of disorder in clay minerals arises from random displacements between layers. Although the sharpness of the hkl reflections from clays may be affected by layer disorder, it is the thickness of the layer packets or ‘tactoids’ (that is, the number of layers accurately positioned on top of each other) that affects the breadth of the 001 reflections. Clay packets may be as thin as one layer – in which case no 001 reflections will be seen at all by XRD, or
53
Regolith mineralogy
(a)
(b) 002
002
020
110
110
-
--
111
020
111 --
111
-
021 111
021
18
20
22 24 26 2 Cu-K radiation
111
28
30 18
20
22 24 26 2 Cu-K radiation
28
30
Figure 4.7: (a) XRD pattern of the 02l, 11l region of a well-ordered kaolinite. (b) XRD pattern of the 02, 11 region of a poorly ordered kaolinite (Cu Ka radiation). Patterns like this are typical of many soil and transported kaolinites.
be several µm thick yielding narrow, sharp 001 peaks. In between lie most of the regolith clays, which have crystals that are less than 0.1 µm in thickness, which introduces broadening of the 001 reflections (Figure 4.8). It is possible to estimate the mean tactoid thick450 400 350 300 250 2
200 150 100 50 0
20
21
2
22
23
Figure 4.8: X-ray diffraction peak broadened by small crystal size (about 0.025 µm thick). In this case the full width at half maximum (FWHM) (62q) is approximately 0.5° 2q, at least 10× broader than a well-crystallised 2 µm crystal would yield.
ness from the width of a clay mineral peak (for example, Moore and Reynolds 1989). 4.3.1 Smectite Smectites were in the older literature referred to as ‘montmorillonite’, but that name is now restricted to a particular species in the smectite group. The detailed structure of smectites is not well known, because they yield broad and indistinct X-ray diffraction peaks. There is general agreement that they have normal 2:1 layers, with hydrated cations in the interlayer region (Figure 4.9). It is assumed that the layers are regular in their atomic arrangement, just like those of micas. SEM and TEM show that Na-smectite layers have no apparent rigidity, indicating little strength in the inter-particle forces linking one layer to the next. Indeed when placed in water, Na-smectites are thought to completely separate into individual 2:1 layers (Norrish 1954; Foster et al. 1955). This characteristic has important implications for the stability of sodic soils – that is, soils containing appreciable amounts of exchangeable Na (see Section 12.7). The complete separation of Na-smectites into individual 2:1 layers is known as dispersion, and can cause problems of soil crusting when dry (Rengasamy and Olsson 1991). The dispersed clay particles can
54
Regolith Science
Mica
Illite
Smectite
Plan view of tetrahedral sheet Si
Tetrahedral Al
Si Si
Al K
12.5 Å
10 Å 15.5 Å TEf009-07
Layers viewed side on (diagrammatic) Octahedral Al
K
Ca
Water
Si-tetrahedron
Al-tetrahedron
Figure 4.9: Diagrammatic plan (upper) and side (lower) views of the structures of muscovite, illite and smectite. When the interlayer charge is less than abut 0.5 per 4SiAl, hydrated cations enter the interlayer causing the layers to separate to accommodate the water.
also block pores and hence decrease soil permeability (Turner et al. 2008). The singly charged ions K+ and Na+ attract water relatively weakly, whereas Ca2+ attracts water more strongly. Na+ smectites absorb one water layer at humidities between 5% and about 50%, and a second above 50% relative humidity. Ca2+ smectites absorb one layer unless completely dried, and add a second water layer above about 20% humidity. The thickness of a smectite unit is 9.4 Å without water, expanding to 12.5 Å with one water layer, and to 15.5 Å with two. Immersed totally in water, some smectites absorb a third layer, while others expand indefinitely (that is, the layers separate completely). The water in the interlayer can be replaced by molecules that are more strongly attracted to the interlayer cation, such as alcohol, glycerol, urea and many other organic molecules. Natural and synthetic organics, such as pesticides and herbicides, may also enter the interlayer region of smectites. Most smectites as collected are
well hydrated, and so yield a 15 Å basal XRD peak, which collapses to 10 Å when fully dehydrated. Care should be taken when collecting and preparing samples to maintain the hydration state. Smectites typically form the finest particles in a soil. They may be no more than two or three layers thick (30–45 Å), and 1000 Å across. They have very high cation exchange capacity (80–150 cmol+/kg), which is largely derived from exchange sites in the interlayer (see also Chapter 5; Table 5.3). Their small size also gives them a high edge-exchange capacity. A small amount of smectite in a soil therefore has a considerable effect on its properties. Of the several varieties of smectite listed in Table 4.1, the most common compositions reported in the literature fall between those of montmorillonite and beidellite, and is generally closer to montmorillonite (Güven, 1988). Such smectites are the components of the rock bentonite, and are commonly formed by the weathering of volcanic ash. In beidellite – as in most
Regolith mineralogy
smectites – the interlayer charge arises from the Al/Si substitution in the tetrahedral sheet. Montmorillonite is an aluminous smectite, with its net charge arising from in the octahedral sheet by the substitution of Mg2+ for Al3+. The Fe3+ smectite, nontronite, is found in weathered ultramafic rocks, such as the nickel laterites of Western Australia (Murrin Murrin; Gaudin et al. 2004) and east-central Queensland (Marlborough, Foster and Eggleton 2002). Saponite (Mg-smectite) forms in weathered basalts and in regolith over high-Mg rocks such as talc schists and ultramafics (for example, at Marlborough, Queensland). Detailed studies of rock weathering have shown that individual parent minerals can give rise to different smectites. On bulk sampling of weathered rocks or soils, all these phases would be lumped together. Caillaud et al. (2006) found saponite and two different Fesmectites as alteration products of serpentine, depending on the micro-site examined, as well as nontronite and a dioctahedral smectite as an alteration product of adjacent chlorite. 4.3.2 Chlorite Most chlorite is trioctahedral – having talc-like 2:1 layers with brucite-like layers between. The layers have a spacing of about 14.2 Å, and the X-ray pattern is usually quite sharp and clear. Chlorite does not swell with hydration or organics, and is little affected by heating. Chlorite in the regolith is generally residual from bedrock; however, aluminous chlorites occur in some soils. Wilson (2004) summarises the weathering of chlorite in which the first step is the formation of vermiculite by a sequence in which individual brucite-like layers lose Mg 2+ and gain H+ until the interlayer is occupied by hydrated Mg2+ rather than a [Mg(OH)2] octahedral sheet. In some instances, the alteration leads to a regular alternation (interstratification) of chlorite and vermiculite (that is, the species corrensite; Figure 4.10). With further weathering, random-vermiculite chlorite precedes the full conversion to vermiculite. 4.3.3 Vermiculite Vermiculite is structurally mid-way between biotite and chlorite and is similar to smectite. It has a trioctahedral 2:1 layer, and an interlayer of [Mg2+.nH2O] or [Al3+.nH2O]. Vermiculite has a higher layer charge
than smectite, but a similar high cation exchange capacity (100–150 cmoles +/kg) (Table 5.3). Both Mg2+ and Al3+ are able to hold two layers of water molecules in the interlayer. The layers are more strongly held together than are those of smectite, so the basal spacing is smaller (14 Å compared with 15 Å for smectite). On heating, the water can be expelled – collapsing the structure to 12.5 Å and then to 10 Å. If a large vermiculite crystal is rapidly heated, the steam literally blows some of the layers apart, and accordion-like ‘worms’ are formed, from whence its name is derived. Expanded vermiculite is used as an absorbent for potting soils, kitty litter, industrial clean-up and as packing material. Despite the physical expansion of the crystal, its structural basal spacing collapses to 10 Å as a result of being heated. Vermiculite is formed during the weathering of biotite in a manner similar to its formation during chlorite weathering. K+ from the biotite interlayer is leached and replaced by hydrated Mg2+ derived from elsewhere in the weathering crystal or further away in the profile (see Wilson 2004 for a detailed summary). Vermiculite is also produced by the weathering of pyroxenes and amphiboles. Figure 4.3b shows diagrammatically the way in which TOT chains of pyroxene may coalesce to produce a 2:1 layer, which in turn may evolve to vermiculite. 4.3.4 Muscovite and Illite Muscovite has K between the 2:1 layers, held there by electrostatic attraction to charge-unsatisfied O of the Si-O network: these oxygens are bonded to one Si and one Al (Figure 4.11). Muscovite [K2Al4 (AlSi3)2 O20 (OH)4] has one K, two octahedral Al and one tetrahedral Al for every three Si atoms. Substitution of divalent ions (usually Fe and Mg) for trivalent octahedral Al is balanced by increased Si in the tetrahedral sites [Tschermak substitution: (Fe,Mg) 2+oct+ Si4+tet ¢ Al3+ oct+ Al3+ tet ] and results in the formation of phengite [K 2Al3 (Fe,Mg)(AlSi7) O20 (OH)4]. The extent of such substitution has commonly been used as a vector toward mineralisation, especially because it may be so readily measured by hyperspectometers (for example, Hermann et al. 2001; see also Section 4.8 below). Illite has fewer Al replacing Si in the tetrahedral sheet, and correspondingly fewer K, with a formula
55
56
Regolith Science
TOT layer brucite-type interlayer
chlorite TEf010-07
Mg.H2O interlayer
random chloritevermiculite
regular chlorite vermiculite (corrensite)
random vermiculitechlorite
vermiculite
Figure 4.10: Diagrammatic representation of the conversion of chlorite to vermiculite. The ‘cotton reels’ represent a single unit cell (yz section) of the TOT unit of talc-like layers, the shaded rectangles represent brucite-like interlayers, and the outline rectangles represent hydrated-Mg interlayers.
suggested to be K1.8Al4[Si3.1Al0.9]2O20 (OH)4. The illite structure is thought to incorporate in some way 90% muscovite-like domains and 10% pyrophyllitelike domains and, although the octahedral sheet composition need not be pure Al, illites are essentially dioctahedral. The K in illite is still sufficient to hold the 2:1 layers firmly together, so illite is a 10 Å layer silicate. Illite has a CEC of no more than 5 cmoles+/kg (Table 5.3). What has made the understanding of illite difficult is that random interstratification of smectite with illite can occur with little change in the position of the 10 Å XRD peak. The smectite component introduces high CEC and many illites have been reported with CECs of the order of 40 cmoles +/kg (Table 5.3).
Si 4+
O
K+
Al 3+
Figure 4.11: Part of the tetrahedral sheet of mica showing the bonding of K to basal oxygens that are charge unsatisfied because of Al3+ ¢ Si4+ substitution.
Regolith mineralogy
According to Meunier and Velde (2004), these all have interstratified smectite. Deconvolution of the 10 Å XRD peak generally shows it to be composed of a wellcrystallised illite plus poorly crystallised illite ± interstratified smectite. 4.3.5 Kaolin Kaolin is a general term that includes kaolinite, dickite, nacrite and halloysite. It is an aluminosilicate, which is formed during weathering of all the alumino-silicate primary minerals such as feldspars, muscovite, feldspathoids and zeolites, as well as illite, smectite and vermiculite. Kaolinite, dickite and nacrite are platy – commonly forming hexagonal crystals 0.1–2 µm across and about one-tenth as thick. Halloysite has the same composition as kaolinite, but with water between the layers. Its layers are not flat, but are curled or rolled – generally occurring either in tubes or spheres – although platy halloysite with curled edges has been described (Figure 4.12). Kaolin has a very low cation exchange capacity – of the order of 3 cmoles+/kg (Table 5.3). The composition of kaolin is simple and constant (Al2Si2O5(OH)4). There is no structural exchange site – cation exchange in kaolin derives from surface and edge exchange sites. Kaolinite sensu stricto is a triclinic mineral – having regular stacking of the 1:1 layers, yielding a 1-layer structure, 7.2 Å thick (001 spacing). Dickite is a poly-
Figure 4.12: Scanning electron micrograph of tubular halloysite surrounding kaolinite crystals, the products of granite weathering, Hong Kong. (Photo R.A. Eggleton).
typic variant, in which the layer sequence alternates through alternate positioning of the vacant site in the octahedral layer (Newnham and Brindley 1956). This yields a two-layer unit cell with d(001) = 14.4 Å; the individual layers are still 7.2 Å thick. The first X-ray peak from dickite has a 7.2 Å spacing as does kaolinite. However, because of the doubled unit cell, this reflection in dickite is indexed as 002. Because both polytypes are defined on the basis of their regular stacking, the terms are only applicable to regularly stacked crystals, and XRD or electron diffraction is needed to establish this. Identifying which kaolin mineral is present requires a well-ordered crystal, yielding a good sequence of hkl X-ray peaks. Although there are differences in the infra-red patterns of kaolinite and dickite (Section 4.8), distinguishing the two polytypes hyperspectrally may be difficult if disordered kaolinite or any other clays are present. Regularly stacked kaolin is best developed in hydrothermal deposits and in in situ weathering profiles. Kaolinite is particularly common in weathering profiles on granites and other aluminous igneous rocks, and on shales and arkoses among sedimentary rocks (see Chapter 6). Transported or sedimentary kaolin is most commonly disordered: the layer sequence is random, yielding an XRD pattern like that shown in Figure 4.7b, with its hyperspectral response also different to that in residual regolith (Section 4.8). Dickite, which is not common, is known from hydrothermally altered rocks (for example, Choo and Kim 2004) and, more rarely, as an authigenic mineral in shales (Veniale et al. 2002) and in sandstones (Bayliss et al. 1965). Nacrite is a third variant, which is even rarer than dickite. It is generally regarded as a hydrothermal mineral. It also has a two-layer structure, but the interlayer shifts are in the direction perpendicular to those of dickite and kaolinite (Zheng and Bailey 1994). 4.3.6 Interstratified clays In plan, all the clay silicates have the same structure: a hexagonal silica-oxygen tetrahedral sheet (or sheets) and a hexagonal octahedral sheet. They therefore have little difficulty stacking different layer types on top of one another. Illite and smectite layers may alternate: building up random sequences (ISISSIIISISSIIS) or
57
58
Regolith Science
regular sequences (ISISISIS, or IISIISIISIIS). Kaolinite and smectite may alternate in soil clays, and both biotite and chlorite weather to vermiculite through an intermediate random (or sometimes ordered) interstratification, as described above. There is a vast literature on interstratified clays (also incorrectly called mixed-layer clays) because they are formed during diagenesis of shales and are common in sedimentary basins (for example, Weaver et al. 1971; Meunier 2005). 4.3.7 Sepiolite–palygorskite Two fibrous Mg-rich chain-silicates that are found in the regolith are sepiolite and palygorskite. They both have moderate cation exchange capacities (20–30 cmoles+/kg) and a high surface area (around 900 m2 /g). They adsorb metal cations very effectively, and also can absorb more than twice their own weight of water. Sepiolite and palygorskite form in saline evaporitic environments – both marine and terrestrial – such as under arid conditions, in closed basins where high salinity groundwaters accumulate. A reaction between detrital clays – largely smectite – and high-Mg alkaline water derived from mafic rock weathering is generally envisaged (Weaver and Beck 1977). Callen (1984) summarised the environments of sepiolite–palygorskite deposits and notes a marked relation between latitude (30°–40° N and S) and the occurrence of these minerals – interpreting this as a control by aridity on their continental formation. Palygorskite deposits are widespread in palaeochannels on the Eyre Peninsula, South Australia (Keeling and Self, 1996), and are associated with dolomite and illite–smectite clays.
4.4
OXIDES AND HYDROXIDES
4.4.1 Silica minerals Quartz is far and away the most abundant and ubiquitous mineral in the regolith. Most quartz is residual from bedrock – whether locally derived or transported. There are several reasons for its abundance in the regolith beyond its abundance in crustal rocks (note that feldspar is twice as abundant in rocks, yet is quite rare in regolith). Quartz is the hardest of the common rock-forming minerals, and it has very poor cleavage. Under mechanical stress it does not fracture
readily – thus maintaining relatively large grains – and it is resistant to abrasion. Much igneous quartz has incipient fracture surfaces spaced at about 20 µm (Moss et al. 1973). When quartz does fracture, the smallest particles produced are in the silt-size range so that, unlike clay minerals, they do not wash away easily. It is also the least soluble of the common rockforming minerals in normal groundwater (pH around 6). It is 10 to 105× less soluble than feldspar, and its dissolution kinetics are slow. Using laboratory measured rates of dissolution (White and Brantley 1995), a 200 µm quartz grain would take of the order of two million years to dissolve. As outlined at the start of this chapter, the chemical weathering of silicates releases silica into solution. Particularly in arid climates, the dissolved silica is liable to be precipitated – generally as micro-crystalline aggregates. Quartz may precipitate, particularly coating pre-existing quartz, but more common are the silica varieties chalcedony, moganite and opal. Chalcedony and moganite have structures based on multiple twinning of quartz (Miehe and Graetsch 1992), whereas the opal structure may have cristobalite (Opal C) or tridymite-like units (Opal T), both (Opal CT), or be X-ray amorphous (Opal A). Opal yields very broad XRD peaks – typically having full width at half maximum values of ~0.5°2q. Silica precipitation in the regolith leads to hardpans, silica veins and crack fillings and to silcretes. Opaline silica is quite abundant in regolith over ultramafic rocks, as the low Al content of the parent rock does not provide enough Al for kaolinite to form as a host for Si (Chapter 6). Thiry et al (2006) conclude that extensive regions of silica deposition in South Australia in the form of opal, silcrete and red-brown hardpans were formed by a complex interplay of water movement, pH change and climate variation from arid to humid. 4.4.2 Al-oxides and hydroxides Continued weathering of alumino-silicates, particularly of kaolin, eventually removes all the silica and leaves alumina minerals. Gibbsite [Al(OH)3] is the most common alumina mineral, with boehmite [g-AlO(OH)] quite common, whereas diaspore [a-AlO(OH)], corundum and other
Regolith mineralogy
polymorphs of Al2O3 are rare. Gibbsite and boehmite are the main components of bauxite and so are important as ores of aluminium; they are referred to in the mining industry as tri-and mono-hydrate, respectively. Gibbsite is formed by dissolving the silica out of kaolinite. Al 2 Si 2 O 5 (OH) 4 + 5H 2 O = 2Al (OH) 3 + 2Si(OH) 4
(Eqn 8.5) On the basis of crystal size, gibbsite and boehmite can be classed as clay minerals – as in most weathered rocks they occur as submicron-sized crystals – however, they generally yield clear, sharp XRD patterns. Heating, either through metamorphism or forest fire, can convert gibbsite and boehmite to corundum (a-alumina), or the less stable polymorphs g-, c- or e-alumina (see Section 4.5.4 below). 4.4.3 Fe-oxides and hydroxides As described in Section 4.3, an important step in the weathering of primary rock-forming minerals is the oxidation of Fe2+ to Fe3+. Below the water table, where the conditions are likely to be reducing, dissolution releases Fe2+ to the groundwater. As soon as Fe2+ reaches an oxidising environment – for example, above the water table – Fe3+ precipitates within the weathering solution as Fe (OH)3, which then evolves to the mineral ferrihydrite (Schwertmann 1988; see also Chapters 5 and 10; Table 12.1). Ferrihydrite has an approximate composition 5Fe2O3.9H2O. Ferrihydrite is the brown rusty scum visible at springs, where water seeps from cracks in rocks, or as an ‘oil slick’ on some swamp water. Ferrihydrite crystals range from about 20 Å in diameter to 75 Å. The degree of organisation of these particles is low, and the X-ray pattern is very simple and weak, with broad lines. Much ferrihydrite in regolith is missed because it does not yield a marked diffraction pattern. The difficulty of characterising this nanophase material is well illustrated by Michel et al. (2007). The surface area of ferrihydrite crystals ranges from 200 to 800 m2 /g. Ferrihydrite is a strong adsorber of phosphate, silica, organic molecules, and heavy metals. In the laboratory, ferrihydrite transforms to a more stable Fe oxides (usually goethite) over a period
of a few years. In the soil it probably passes in and out of solution with the seasons. Ferrihydrite is of the order of 100× more soluble in normal groundwater than the other Fe oxides. Most ferrihydrite is associated with bacteria (Gallionella and Lepthotrix), which gain their energy from the oxidation reaction, Fe2+ = Fe3+ + e- (see Section 7.4.1). Ferrihydrite also precipitates from Fe3+ solutions as pH increases. Ferric Fe is soluble at pH 2 (very acid), becoming less so with increasing pH. At pH 4 the solubility is negligible (about 1 in 10 million). Very acid waters (mine waters and lakes such as Lake Tyrell in western Victoria (Figure 1.3; Macumber 1992) can hold appreciable Fe3+ in solution and this precipitates as ferrihydrite on dilution (because the pH increases) or on input of alkaline water. Iron, which is derived from pyrite in rocks or coastal muds, commonly rises to the surface and precipitates as ferrihydrite in acid sulfate soils (see also Chapter 12). Cyclic dissolution and precipitation of Fe by reduction/oxidation alternation or pH change moves Fe away from reducing areas toward oxidising areas and is responsible for most of the brown/yellow colour banding of soils and weathered rocks. Precipitation at the top of the water table may yield a ferruginous hardpan (see also Chapters 2, 3 and 13). Goethite [a-FeO(OH)] is the most common of the soil Fe minerals; goethite is the first conversion product from ferrihydrite. It is a yellow-brown mineral, forming as needle-shaped crystals about 1 µm long in synthetic preparations, but typically more equant in soils. Together with ferrihydrite, goethite imparts most of the brown colour to soils. The surface area of soil goethite ranges from 6 to 200 m2 /g, which gives goethite considerable adsorptive ability. Heavy metals, such as Cu, Pb, and Zn, are adsorbed to the extent of about 1 µmol/m2 (20 µmol/g) (see also Chapter 5). Goethite is also an effective anion adsorber – notably of phosphate. At normal regolith pH, phosphate values of about 2–3 µmol/m 2 have been measured both in laboratory and the field. Much of the superphosphate, Ca(H2PO4)2, ploughed into fields becomes unavailable to plants in quite a short time because it is sequestered by goethite. Aluminium occurs in goethite substituting for Fe, up to 32 mole% (Fitzpatrick and Schwertmann, 1982).
59
60
Regolith Science
In the regolith, goethite formed in hydromorphic environments, such as mottles, concretions and ferricretes, tends to have lower Al substitution (0–15 mole%), whereas in freely drained regolith, such as saprolites and bauxites, goethite has Al substitution ranging from 15 to 32 mole% (Fitzpatrick 1988). Fitzpatrick and Schwertmann (1982) explain the difference as resulting from lower pH, and therefore higher Al activity, in the more freely drained regolith. Al substitution is readily estimated from the X-ray diffraction pattern of goethite; substitution of Al reduces the unit cell dimensions (Schulze 1984), as well as reducing the mean crystalline dimension. Because many regolith goethites are extremely finegrained, they give rather broad XRD peaks, and this can lead to misleading estimates of the cell dimension and, in turn, of the Al substitution. Schulze (1984) explains this problem and suggests ways to overcome it. Using full-profile XRD analysis, the cell dimensions of goethite can be reasonably well determined. Lepidocrocite is the [g-FeO(OH)] polymorph with the oxygens in approximate cubic close packing. Recognisable by its orange colour, lepidocrocite is a relatively uncommon mineral, forming in preference to goethite as a direct oxidation product of ferrous Fe and in preference to ferrihydrite if oxidation is slow. It therefore indicates reductomorphic soils. Lepidocrocite – or the very rare polymorph, akaganéite – also seems to be precipitated instead of goethite in the presence of Cl– (such as at Lake Tyrell, Victoria; Dickson and Herczeg 1992). Hematite [a-Fe2O3] is very common in warm or arid regolith, and is red when fine-grained. Its intense colour may mask the presence of goethite. The surface area for soil hematite is about 100m2/g – much the same as goethite. The hematite–goethite ratio in soils increases with soil temperature (decreasing latitude) and decreases with soil moisture content. Locally, hill tops are richer in hematite, and valleys in goethite; globally, the arid regions have hematite rather than goethite. (The proportion of hematite to goethite can be readily determined hyperspectrally; Section 4.8; Cudahy and Ramanaidou 1997.) Hematite has similar adsorption properties to goethite and can also be responsible for the fixation of phosphate. Hematite can accept up to about 15 mole% Al in the structure (Fitzpatrick and Schwertmann 1982); such high levels
indicate hematite crystallisation from solutions saturated with Al (Schwertmann and Kämpf 1985; Fitzpatrick 1988). Determination of the extent of Al substitution is very difficult. As with goethite, high Al tends to reduce the crystal size of the hematite making EMP analysis impossible. Stanjek and Schwertmann (1992) have shown that cell dimensions are at best equivocal as estimators of Al substitution in hematite. Magnetite [g-Fe3O4] is a member of the spinel group of minerals. It is not fully oxidised – having one Fe2+ and two Fe3+ ions within its structure. It is not as common in regolith as other Fe3+ minerals. Although some may be produced directly by bacteria (see Chapter 7), most is residual magnetite from parent rocks. Maghemite [g-Fe2O3] is also a spinel, although its formula is the same as that of hematite. Written as a spinel, maghemite is [Fe8O12], compared with magnetite, [Fe9O12]. Maghemite can form by the oxidation of magnetite, and some soil maghemite may result from the oxidation of 0.1 µm crystals of magnetite formed by bacteria. Most maghemite probably forms in soils by the dehydration of goethite or lepidocrocite on the land surface, often during fires (Fitzpatrick 1988). Lepidocrocite can transform easily to maghemite, as both have cubic close-packed oxygen substructures. Goethite has a structure based on hexagonal close packing, and normally dehydrates to (hexagonal) hematite. In the presence of organic matter, it is thought that maghemite is the common dehydration product. Early in a bushfire, plant fragments in the soil burn providing hot, reducing conditions capable of converting ferrihydrite or goethite to Fe3O4 or possibly FeO. As the fire passes and the carbon is consumed, the reduced oxides change to maghemite. Maghemite is strongly magnetic, and a hand magnet is the quickest means of identification (assuming no magnetite is present). Maghemite is dense (5 g/cm3) and stable in the weathering environment. Maghemite-rich ferruginous nodules and grains accumulate in stream channels and paleochannels in which they are particularly obvious in magnetic surveys (see Section 9.3.1). 4.4.4 Anatase The most abundant polymorph of TiO2 in the regolith is anatase. It commonly is found as very small (0.1 µm) crystals (Figure 4.13), and is a major constituent of
Regolith mineralogy
Parc et al. 1989). Thus they may contain significant amounts of potential pathfinder elements (see Section 5.4.3 and Appendix 2) K-bearing Mn oxides have been used to date regolith by K-Ar and Ar-Ar methods (see Chapter 2).
4.5 OTHER MINERALS
Figure 4.13: Crystals of anatase about 0.1 mm across with smectite in altered sphene (titanite) (Tilley and Eggleton 2005).
the fine-grained alteration assemblage known as leucoxene. Anatase has a cream-coloured appearance when it is concentrated, but mostly it is dispersed uniformly through silicate weathering products. In the lateritic and bauxitic parts of regolith profiles and in silcretes, anatase content becomes residually concentrated (Section 5.4.3) and commonly reaches 2 to 3%, and as high as 40% in some silcretes (Thiry and Simon-Coinçon 1996). Anatase has one very prominent XRD peak at 3.5 Å, which is generally sharp and well resolved from the nearby 3.6 Å kaolin 002 peak. As little as 0.5% anatase can be detected from this peak, although as it is commonly the only anatase peak distinguishable in the XRD pattern, it is wise to cross check a conclusion about anatase with chemical evidence for TiO2. 4.4.5 Mn oxides and hydroxides In the weathering environment, Mn becomes oxidised to the tetravalent state (Chapter 5). The mineralogy of Mn oxides and hydroxides is complex: the more common regolith species are the layer structures having cations other than Mn between MnO6 octahedral sheets: vernadite [d-MnO2], incorporating Ba or K, lithiophorite [(Al,Li)MnO2 (OH)2], birnessite [(Na,K)4Mn14O27.9H2O], and the cryptomelane– coronadite–hollandite group [(K,Pb,Ba)2-1Mn8O16], which have large cations in tunnels which are bounded by columns of MnO6 octahedra. Other Mn-oxides and oxyhydroxides include pyrolusite and nsutite [MnO2], romancheite (containing Ba) todorokite (containing Ca, Na, and K), chalcophanite [ZnMn3O7.3H2O] and asbolane (Ostwald 1992;
4.5.1 Sulfates Gypsum [CaSO4.2H2O] is a common evaporite mineral in arid environments. It occurs both in lake deposits and in the regolith over sulfides. Bassanite [CaSO4.0.5H2O] is formed by the dehydration of gypsum under arid conditions (Akpokodje 1984). Jarosite–natrojarosite [(K,Na)Fe3(SO4)2(OH) 6] precipitate from a reaction between sulfuric acid formed by pyrite oxidation and surrounding silicates. These minerals are common in regolith where pyrite is weathering, and are particularly so in acid sulfate soils (Chapter 12) and mine dumps. Brown (1971) showed that jarosite is only stable in the presence of goethite at pH below 3. Such extreme pH levels are reached during sulfide weathering and also in acid saline lakes such as Lake Tyrrell, western Victoria (Macumber 1992). The persistence of jarosite into environments of higher pH is attributed by Brown to the slowness of its conversion to goethite. Jarosite can take a wide range of metals into its structure, substituting either for the large alkali cation (12-fold coordination) or for the octahedrally coordinated Fe3+ (Scott 1987). Metals released by acid sulfate weathering of sulfide ores, such as Ag, Pb, Tl, Cu or Hg, may reside in jarosite (Becker and Gashrova 2001) or jarositic minerals (Table 4.2; see also Appendix 2), to be released to the environment when the acidity is neutralised and jarosite dissolves. Alunite [KAl3 (SO4)2 (OH) 6] is isostructural with jarosite, but is found both in hydrothermal (advanced argillic) alteration, and weathering environments. In the latter, it may be associated with acid lake and groundwaters, in which it crystallises by reaction between clays and sulfuric acid from pyrite weathering. Alunite, associated with gypsum, kaolin and opal, is also widespread in arid southern Australia, where it has been suggested that it forms from sulfate-rich groundwaters (Bird et al. 1989) under low pH conditions associated with the oxidation of Fe2+
61
Goslarite ZnSO4.7H2O
Argentojarosite AgFe3 (SO4) 2 (HO) 6
Alunite KA13 (SO4) 2 (OH) 6 Barite BaSO4 Gypsum CaSO4.2H2O
Zinc
Silver Native
Others
Chalcophanite ZnMn3O7.3H2O
Coronadite PbMn8O16
Anglesite Pb SO4 Plumbojarosite Pb0.5Fe3 (SO4) 2 (OH) 6
Lead
Mimetite Pb5 (AsO4)3Cl Beudantite PbFe3 (AsO4) (SO4)(OH) 6
Cuprite Cu2O Tenorite CuO
Chalcanthite CuSO4.5H2O Bronchantite Cu4SO4 (OH) 6
Copper Native
Goethite FeOOH Hematite Fe2O3
Scorodite Fe AsO4.2H2O
Hollandite BaMn8O16 Coronadite PbMn8O16 Cryptomelane KMn8O16
Jarosite K Fe3 (SO4) 2 (OH) 6 Melanterite FeSO4.7H2O
Iron
Oxide
Arsenate
Manganese
Sulfate
Smithsonite ZnCO3 Rosasite (Cu,Zn) 2CO3 (OH)2
Cerussite PbCO3
Malachite Cu2CO3 (OH)2
Manganosiderite (Mg,Fe)CO3
Siderite FeCO3 Ankerite Ca(Mg,Fe)(CO3) 2
Carbonate
Commonly occurring minerals in gossans (after Blain and Andrew 1977).
Metal
Table 4.2:
Hemimorphite Zn4Si2O7(OH)2.H2O H2 O
Chrysocolla CuSiO3.2H2O
Silicate
Halite NaCl
Chloargyrite AgCl Embolite Ag(Cl,Br) Iodargyrite Agl
Phosgenite Pb2CO3Cl2 Cottunite PbCl2
Atacamite CuCl2.3Cu(OH) 2
Halide
Pyromorphite Pb5 (PO4)3Cl Plumbogummite PbAl3H(PO4) 2 (OH) 6
Pseudomalachite Cu5 (PO4)2 (OH) 4.H2O
Phosphate
62 Regolith Science
Regolith mineralogy
(ferrolysis) rather than from sulfide weathering (Thiry et al. 2006). 4.5.2 Carbonates The major carbonate mineral of the regolith is calcite [CaCO3]. The polymorphs aragonite and vaterite may be formed within shells (see also Section 7.3.1). Of the other rhombohedral carbonates, dolomite [CaMg(CO3)2] and magnesite [MgCO3], are found quite commonly, but siderite [FeCO3] and ankerite [Ca(Mg,Fe)(CO3)2] rather rarely. Calcite develops in the regolith in many environments and, where Ca is abundant in the bedrock, particularly in semi-arid climates, large regions of regolith cemented by calcite are found, which are generally termed calcrete. Uncemented aggregates of carbonate minerals are also generally termed ‘calcrete’ (see Chapters 5, 6 and 13); these may be calcic, dolomitic or magnesitic. Magnesite may be abundant in high Mg terrains: for example, as regolith carbonate accumulations following the weathering of ultramafics such as those at Kunwarrara in central east Queensland (Wilcock 1998) or on the Yilgarn Craton of Western Australia (see Chapter 6) (Wells 2005; Gaudin et al. 2005). Dolomite and high Mg-calcite occur in some Australian inland regolith – notably in South Australia (Milnes and Hutton 1983) and western New South Wales (McQueen et al. 1999). These authors also report an increase in the Mg content of Ca-Mg carbonates with depth. See also Chapter 13 for discussion of calcrete as a sampling medium. Metal carbonates, such as those of Cu (malachite and azurite), Pb (cerussite), Zn (smithsonite) and Ni (gaspéite), are well known from the supergene region of weathered ore bodies (Table 4.2; Chapter 5). 4.5.3 Phosphates The significance of phosphatic members of the alunite supergroup in the regolith was established by Norrish (1975) and Norrish and Rosser (1983). The group includes crandallite [CaAl3H(PO4)2 (OH) 6], gorceixite [BaAl3H(PO4)2 (OH) 6], and florencite [CeAl3 (PO4)2 (OH) 6], and this isomorphous series can host divalent cations (Ca, Ba, Sr, Pb) and trivalent ions (Y and the REE) in the large (12-fold coordination) sites and tetrahedrally coordinated groups such
as (PO4)3–, (AsO4)3–, (SO4)2–. Of these, the phosphates form a highly insoluble family of minerals that are quite stable in the weathering environment. Banfield and Eggleton (1989) and Taunton et al. (2000a, b) have shown the importance of REE phosphates in controlling the P availability in a weathered granite profile. 4.5.4 Halides, nitrates and borates Evaporation in arid climates commonly leads to the crystallisation of minerals such as halite [NaCl], nitre [KNO3] and borax [Na2B4O5 (OH)4.8H2O]. Halite is by far the most abundant evaporite mineral – occurring across wide regions of arid Australia, particularly in the large salt lakes of South Australia such as Lake Eyre, and the extensive drainage channels of the Yilgarn Craton of Western Australia. Silver halides may be present in gossans (Table 4.2). Nitrates and borates are only found in commercial quantities in evaporative basins of the Andes and Cordillera of America. 4.5.5 Poorly crystalline minerals Amorphous minerals were originally so-called because they lacked a crystal shape. The term has become extended to minerals that are not detectable by methods based on crystallinity such as XRD. Techniques such as scanning and transmission electron microscopy have shown that many ‘amorphous’ minerals do have well-defined morphology. They may be composed of very small, or rather imperfect crystals, as for ferrihydrite, or they may have curved morphology, such as allophane. Both kinds of minerals yield XRD patterns with broad, indistinct maxima and are better termed ‘poorly diffracting’. Their presence has been long known to soil scientists (Gieseking 1975), and routinely estimated by chemical extractions. Examination of regolith minerals by transmission electron microscopy and differential XRD has allowed mineralogical characterisation of these materials, and their importance in regolith mineralogy and geochemistry is gradually being recognised (Tilley and Eggleton 1995, 1996; Singh and Gilkes 1995). Allophane is a hydrated alumino-silicate, which is formed as spheres about 50 Å across. It does not have a single composition, but ranges from [Al2O3.SiO2]to
63
64
Regolith Science
[Al2O3.2SiO2]. It is difficult to recognise because it gives very poor XRD peaks (broad bands centred at about 15 Å, 3.4 Å and 2.5 Å). It is most common in soils derived from volcanic ash, and so is particularly abundant in Japan and New Zealand. Imogolite is a thread-like mineral of composition about [Al2O3.SiO2.2.5H2O]. It may be abundant in volcanic-derived soils. The threads are bundles of 20 Å diameter tubes. At pH 7, both allophane and imogolite have CECs of the order of 20–30 cmol+/kg. Hisingerite – a rare amorphous alteration product of Fe sulfides, carbonates and silicates – has been shown to be a ferric form of spherical halloysite (Eggleton and Tilley 1998). Many specimens of hisingerite have come from mines at depths below the level normally regarded as within the regolith, though the mineral itself is the product of oxidation and hydration. Hisingerite has a formula close to [Fe2Si2O5 (OH)4], and other than the substitution of Mn, Mg and a small amount of Al for Fe, nothing is known about its chemistry. Its fabric of concentric 1:1 layers forming spheres about 140 Å in diameter gives it a high surface area and a high adsorption potential. Aluminium-Fe oxyhydroxides in pisolitic bauxites and laterites commonly yield very weak X-ray diffraction patterns. Tilley and Eggleton (1996) and Singh and Gilkes (1995) have shown that these near-surface regolith materials may contain a high percentage of ultra-fine-grained minerals occurring as crystals with diameter less than 10 Å, including c- and e-alumina, maghemite, akdaleite [5Al2O3.H2O] and very fine goethite. These minerals have extremely high surface areas (around 500 m2 /g). Thus, they may provide important sinks for adsorbed trace metals (see Sections 5.4.3 and 10.7.5 and Chapter 13), but no work has been done on their geochemistry. 4.5.6 Resistate minerals Important components of the residual weathering products are the mineral group commonly called ‘resistates’ or ‘resistant minerals’. These are minerals not significantly affected by the weathering process. Quartz has been considered in some detail earlier, and this is probably the most common resistate mineral under most weathering conditions, but it will not be discussed again here.
Other common resistates include minerals that occur as trace components in parent rock, but which are concentrated by depletion during the process of weathering. The most common are zircon, rutile, ilmenite, magnetite (and other spinels), garnet, tourmaline and monazite. Much rarer resistates include cassiterite, corundum, gold and diamond. Resistates – particularly zircon and Ti-resistates – have been used as indicators of the degree or extent of weathering (see Taylor and Eggleton 2001 and Chapter 6). Additionally, most of these resistate minerals are denser than quartz and the majority of regolith minerals, and are collectively known as ‘heavy minerals’. Because they are denser, they tend to concentrate as lenses and beds as they are eroded, transported, winnowed and deposited. They form a significant ore in many unconsolidated sedimentary sequences.
4.6
MINERAL WEATHERING
How a primary mineral responds to the weathering environment depends significantly on its Fe2+ content and its solubility at pH around 6: the pH of most soil and regolith waters (Chapter 5; Figures 5.7 and 10.18). The Fe2+ in Fe-bearing minerals is quickly oxidised as soon as they reach the oxidising weathering front. In almost every instance, oxidation of the Fe strongly affects the mineral structure – creating nanometric channels and cavities into which water penetrates, and so accelerating dissolution. More specific aspects of mineral weathering are covered elsewhere in this book, and Wilson (2004) and Taylor and Eggleton (2001) give extensive descriptions of mineral weathering. Iron-free minerals dissolve in water very approximately in order of their content of the more soluble elements Ca and Mg. Thus olivines and pyroxenes weather faster than feldspars, and anorthite weathers more readily than albite. Table 4.3 lists the common minerals in order of their solubility on a logarithmic scale with quartz set at 10 and glass at 1. The order is based on experimental results and on field observations, and is only presented as a guide. In different situations and – depending on the mineral’s actual composition and internal integrity (determined by the density of twin planes, dislocations, inclusions
Regolith mineralogy
Table 4.3: Relative mineral solubilities (base-10 logarithmic scale, large numbers = least soluble). Order established from experimental results in the literature (White and Brantley 1995) – modified by field observations. Gibbsite
11
Beidellite
10
Quartz
10
Kaolinite
10
Microcline
9
Muscovite
9
Sanidine
7
Albite
7
Biotite
7
Augite
7
Oligoclase
6
Andesine
5
Hornblende
4
Diopside
4
Bytownite
4
Anorthite
3
Olivine
2
Glass
1
and fractures) – the susceptibility to weathering may be different. The list includes the Fe-bearing minerals biotite, augite, hornblende and olivine, and their position at the more soluble end of the list is at least partly because of the oxidation of Fe2+.
4.7 STRATEGY FOR QUANTIFYING CLAY MINERALS IN A REGOLITH SAMPLE Regolith materials are the most difficult geological samples to quantify by XRD – and in some cases it is impossible to quantify them using XRD alone. A combination of quantitative phase separation followed by electron microscopy and micro-analysis, X-ray diffraction, sub-sample chemical analysis, infra-red analysis, and so on will yield a good answer, but cost a great deal of both time and money. The main difficulty is that XRD interpretational software generally assumes that the clays diffract as 3-dimensional crystals, whereas many of the alumino-
silicate clays – which is most of those in the regolith – do not. Mica does have a regular 3D structure, illite almost does, some kaolinites do, others ( halloysite and smectite) do not. The XRD patterns from all these clays overlap in the region around 4.5 Å (the 02, 11 band), at about 2.5 Å (the 20, 13 band), and at 1.5 Å (the 06, 33 band), making it difficult to separate the contributions when two or more clays are present. The two-index notation for bands is used for materials so thin or so disordered that they diffract as if they have no third dimension. The basal (001, 002 and so on) reflections, as is well known, do allow each clay to be recognised – provided these reflections are present. In a normal procedure, the clay fraction ( less than 2 µm grain size) is extracted from the sample: this is an essential part of a regolith orientation study. This fraction does not represent the silicate clay minerals quantitatively. Much mica and illite, and some kaolin, remains coarser than 2 µm, or stuck to coarser particles, even after determined dispersal. The clay fraction can be used to identify the clays present, but it will not quantify them in the original sample, even if the clay size fraction has been determined quantitatively. Because of its inherently 2D character, smectite in a bulk regolith sample occurs as irregular sheets – some only one layer thick and others not much thicker – draped over the coarser particles. When these are X-rayed, the basal 15 Å spacing is almost invisible because too few consecutive sheets are superimposed and Bragg diffraction does not occur, even from a sample whose clay fraction clearly shows smectite (Figure 4.14). When the clay fraction is prepared by sedimentation on to a substrate, several smectite sheets layer one on the other. As few as six layers yield a very clear signal. Also, there is likely to be a variety of interlayer cations in a bulk sample, giving different hydration states to the smectite and hence producing a spread of the basal spacing. The clay fraction is generally Mg2+ or Ca2+ saturated before analysis, giving uniformly two water layers in the interlayer and a better defined basal peak. This means that smectite may be seriously underestimated by examination of a bulk regolith sample. Although smectite’s 02,11 band may show appreciable intensity, lacking a significant smectite 001 peak, automated analytical software (such as the Rietveld
65
Regolith Science
quartz kaolinite 001
2000
(a) BULK SAMPLE 1500
Intensity (cpi)
66
kaolinite 002
clay 02,11 qtz
1000
(b) ORIENTED CLAY 500
smectite 001
mica 002 mica 001
0 5
10
15 20 2 (Co K)
25
30
35
Figure 4.14: XRD patterns from (a) a bulk sample and (b) the oriented less than 2 µm extract (Co Ka radiation). Note: Smectite does not appear at all in the bulk sample scan, but the ratio of the 02, 11 band to the kaolinite 001 peak warns of a potential problem because in a random kaolin sample, I001>I02,11.
Full Profile) may attribute this part of the spectrum to another clay, such as kaolinite. The greater-thanexpected intensity now attributed to kaolin’s 02, 11 band is allowed for by the software’s orientation parameter, so that what may in fact be a random sample appears to have kaolinite ‘anti-oriented’. That is, the orientation parameter is adjusted by the software as though the kaolinite flakes were sitting on their edges, not randomly. By this adjustment to the orientation parameter, the relative intensities of the kaolinite 001 and 02, 11 bands can be made to fit the pattern, but the result will be wrong. Another incorrect interpretation might be that the sample contains halloysite (which yields an XRD scan with a prominent 02, 11 band). Although such an estimate of the clays proportions may be seriously in error, the estimate of total clay is usually quite good. This is because all the clay diffraction energy has been accounted for. Without extra information, it is really impossible to know what has actually happened during the computer
analysis. To circumvent this problem, the following strategy is adopted. From an orientation study, include the clays known to be present and, of course, all the other minerals. This will allow the software to estimate total clay, and correctly estimate the non-clay minerals. Determine the bulk sample’s cation exchange capacity (CEC). Smectite and vermiculite have CECs *100 cmoles+/kg. If organic matter is absent, these two minerals contribute almost all the CEC in a normal well-weathered regolith sample. If a sample shows no 15 Å or 14 Å smectite or vermiculite peak, but has a CEC greater than about 10 cmoles+/kg, smectite is present either as interstratified illite-smectite, or just as smectite. Another approach is to measure weight loss versus temperature on heating (thermo-gravimetric analysis). Kaolinite and boehmite both lose about 14% structural (OH) over a temperature range of 400–500°C, and the weight loss in this interval can give a fair estimate of (kaolinite + boehmite) %. Bulk chemical analysis is an invaluable aid to the mineral
Regolith mineralogy
analysis. Particularly useful is the K 2O wt %, because, if K-bearing phases such as K-feldspar or alunite are absent, or have been well quantified in the XRD analysis, the remaining K 2O can generally be attributed to mica (illite, muscovite or biotite). From this extra information – of which total CEC is probably the most valuable – the automated analytical software interpretation can be forced to fit more plausibly. If the bulk sample XRD pattern shows a weak or no 001 peak, a possible approach is to set the smectite orientation factor to an artificially ‘anti-oriented’ value; that is, as though the 3D smectite crystals were all on edge in the XRD mount. This may allow the ratio of smectite to other clays to agree with the chemical data. Setting an ‘anti-orientation’ parameter reduces the calculated intensity for 001 and enhances those of the hk bands. Under this strategy, the orientation parameter for the other clays should be held at the value for a random aggregate (this assumes that a serious effort to achieve a truly random sample has been made; having quartz and other granular minerals in a bulk sample helps this enormously). Vermiculite and halloysite add their own complications. Vermiculite has a high CEC like smectite (Table 5.3), but it generally gives sharper and more intense 001 peaks, and Rietveld Full Profile software is able to discriminate fairly well. However, it becomes very difficult to quantify both vermiculite and smectite in the sample unless one has several other types of data (such as CEC, chemistry and water loss on heating). Halloysite has almost the same pattern as poorly ordered kaolinite, though it is possible to quantify both clays using automated analytical software methods using ‘observed’ files for the minerals if they are the only clays in the sample. Some regolith samples, and some volcanic rocks, contain material that yields no diffraction pattern – or only a very diffuse pattern. Among these are volcanic glass, ferrihydrite (the mineralogical equivalent of rust), opal and some alumina species found in lateritic regolith. Such material is commonly referred to as ‘amorphous’, but, as it may be nanocrystalline, the term ‘poorly diffracting material’ (PDM) is preferable. An internal standard such as corundum or ZnO is essential if PDM is suspected (but it is important to ensure that the standard itself does not contain PDM:
chemical purity does not imply crystallographic purity). The first clue to the presence of PDM is a high background in the XRD pattern and low maximum peak intensities. A broad background hump between about 4 Å and 2.7 Å is another indicator of PDM. Rietveld Full Profile programs include a routine for estimating PDM content (generally called ‘amorphous content analysis’), based on the inclusion of a known amount of a standard.
4.8 HYPERSPECTRAL TECHNIQUES IN REGOLITH STUDIES As indicated in Section 4.1, reflectance spectrometry – especially using the 400–2500 nm wavelength interval (1 µm =103 nm) recorded by field portable instruments (such as PIMA® or ASD® instruments) – is now well established as a rapid routine technique to determine mineralogy, as well as variations in mineral compositions for Fe oxides, layer silicates and carbonate minerals. That technology has made possible the rapid collection of data from a large number of samples to help outline mineralogical boundaries in a two- or three- dimensional sense. Most recently automated hyperspectral logging systems have been developed to allow continuous logging of drill core or percussion chips (Huntington et al. 2004). This allows the vertical extent of significant minerals to be routinely recorded. The features most useful for regolith studies are the nature of the neo-formed clay minerals and Fe oxides, with the abundances and compositions of residual micas also useful in some cases. Kaolinite has characteristic spectral responses around 1400, 2160 and 2210 nm (Figure 4.15). Dickite has its adsorption maxima at 2180, rather than 2160 nm (Figure 4.16), and so the ratio of the depths of the 2180 and 2160 nm wavelengths can reflect the degree of incorporation of ‘dickite domains’ (that is, disorder) in kaolinite. Because the hyperspectral response is related to vibration of the Al-OH bonds, such disorder is not directly related to disorder as measured by XRD (Section 4.3). Nevertheless, this parameter has found widespread usage as a discriminant between the ‘disordered kaolinite’, which is commonly found in transported regolith, and the ‘residual kaolinite’ of residual saprolite. The software used to
67
68
Regolith Science
Kaolinite
Reflectance (offset for clarity)
Dickite
Halloysite
Muscovite Phengite Montmorillonite
1200
1400
1600
1800
2000
2200
2400
Wavelength in nm Figure 4.15: Reflectance spectra for Al(OH)-bearing minerals (from CSIRO Exploration and Mining Reference Library).
interpret such regolith material sometimes identifies nacrite and dickite in regolith samples but, as seen in Section 4.3.5, such identifications are unlikely. Halloysite, [Al2Si2O5 (OH)4.0-2H2O] – that is, hydrated kaolinite (Section 4.3.5) – has an extra adsorption feature at 1900 nm relative to kaolinite (Figure 4.15) and is commonly identified especially toward the top of regolith profiles. Montmorillonite, being more hydrated than kaolinite, has a well developed broad 1900 nm water adsorption features plus a broad 1400 nm feature and subdued 2200 nm feature (Figure 4.15). Bound interlayer water in regolith minerals is reflected by an absorption feature at 1900–2000 nm
and the greater depth of this feature in reflectance spectra from transported relative to residual regolith can sometimes be used to differentiate transported and residual regolith (for example, Chan et al. 2001) The nature of the Fe oxides can be determined by the response in the 400–1000 nm spectral region. Hematite shows an adsorption minimum at 860–870 nm, whereas goethite’s minimum is at 910–930 nm (Figure 4.17; Cudahy and Ramanaidou 1997). The spectral response of white mica in the 2200 nm region is dependent upon the amount of phengitic substitution in the octahedral Al sites (Scott and Yang 100
Go A
% reflectance
Kaolinite
2209
2163 2179
Reflectance (offset for clarity)
80
Dickite
Hm A 40
0 450 2100
Go B
20
Halloysite
2000
Go C
60
2200
2300
2400
2500
Wavelength in nm
Figure 4.16: Reflectance spectra for kaolin minerals (from CSIRO Exploration and Mining Reference Library).
Hm B 860
1270
1680
2090
Wavelength (nm) Figure 4.17: Reflectance spectra for Fe oxide minerals (Go=goethite, Hm=hematite; after Cudahy and Ramanaidou 1997).
2500
Reflectance (offset for clarity)
Regolith mineralogy
Muscovite
Phengite
2000
2100
2200
2300
2400
2500
Wavelength in nm Figure 4.18: Reflectance spectra for white mica minerals (from CSIRO Exploration and Mining Reference Library).
1997; see also Section 4.3.4). The adsorption feature for phengites occurs at longer wavelength (Figure 4.18). Because white micas are generally resistant to weathering (Table 4.3), their compositions are retained and sometimes may be used as indications of mineralisation in regolith samples (for example, Scott 1996).
4.9
REFERENCES
Akpokodje EG (1984). The occurrence of bassanite in some Australian arid-zone soils. Chemical Geology 47, 361–364. Bailey SW, Brindley GW, Johns WD, Martin RT and Ross M (1971). Summary of national and international recommendations on clay mineral nomenclature. Clays and Clay Minerals 19, 129–132. Banfield JF and Eggleton RA (1989). Apatite replacement and rare earth mobilization and fixation during weathering. Clays and Clay Minerals 37, 113–127. Bayliss P, Loughnan FC and Standard JC (1965). Dickite in the Hawkesbury Sandstone of the Sydney Basin, Australia. American Mineralogist 50, 418–426. Becker U and Gashrova B (2001). AFM observations and simulations of jarosite growth at the molecular
scale: probing the basis for the incorporation of foreign ions into jarosite as a storage mineral. Physics and Chemistry of Minerals 28, 545–556. Bird MI, Chivas AR, Lock DE and Andrew AS (1989). An isotopic study of surficial alunite in Australia: 1. Hydrogen and sulfur isotopes. Geochimica et Cosmochimica Acta 53, 3223–3237. Blain CF and Andrew RL (1977). Sulfide weathering and the evaluation of gossans in mineral exploration. Minerals Science Engineering 9, 119–150. Brigatti MF, Frigieri P, Ghezzo C, and Poppi L (2000). Crystal chemistry of Al-rich biotites coexisting with muscovites in peraluminous granites. American Mineralogist 79, 1068–1083. Brown JB (1971). Jarosite-Goethite stabilities at 25°C, 1 ATM. Mineralium Deposita 6, 245–252. Caillaud J, Proust D and Righi D (2006). Weathering sequences in rock-forming minerals in a serpentine: influence of microsystems on clay mineralogy. Clays and Clay Minerals 54, 87–100. Callen RA (1984). Clays of the palygorskite-sepiolite group: depositional environment, age and distribution. In Palygorskite-Sepiolite. Occurrences, Genesis and Uses. Developments in Sedimentology 37. (Ed. A Singer and E Galan) pp. 1–38. Elsevier. Chan RA, Greene RSB, de Souza Kovacs N, Maly BER, McQueen KG and Scott KM (2001). ‘Regolith,
69
70
Regolith Science
geomorphology, geochemistry and mineralisation of the Sussex-Coolabah area in the Girilambone region, north-western Lachlan Fold Belt, NSW’. CRC LEME Report 166. (Reissued as Open File Report 148). CRC LEME, Perth. Choo CO and Kim SJ (2004). Dickite and other kaolin polymorphs from an Al-rich clay deposit formed in volcanic tuff, Southeastern Korea. Clays and Clay Minerals 52, 749–759. Cudahy TJ and Ramanaidou ER (1997). Measurement of the hematite-goethite ratio using visible and near infrared reflectance spectrometry in channel iron deposits, Western Australia. Australian Journal of Earth Sciences 44, 411–420. Dickson BL and Herczeg AL (1992). Deposition of trace elements and radionuclides in the spring zone, Lake Tyrell, Victoria, Australia. Chemical Geology 96, 151–166. Eggleton RA and Boland JN (1982). The weathering of enstatite to talc through a series of transitional phases. Clays and Clay Minerals , 173–178. Eggleton RA and Tilley DB (1998). Hisingerite: ferric kaolin mineral with curved morphology. Clays and Clay Minerals 46, 400–413. Fitzpatrick RW (1988). Iron compounds as indicators of pedogenic processes: Examples from the southern hemisphere. In Iron in Soils and Clay Minerals. (Eds JW Stucki, BA Goodman and U Schwertmann) pp. 351–396. D. Reidel Publishing Co., Dordrecht, Netherlands. Fitzpatrick RW and Schwertmann U (1982). Al-substituted geothite – an indicator of pedogenic and other weathering environments in South Africa. Geoderma 27, 335–347. Foster L and Eggleton RA (2002). The Marlborough nickel laterite deposits In Regolith and Landscapes in Eastern Australia. (Ed. IC Roach) pp. 33–36. CRC LEME, Perth. Foster WR, Savins JG and Waite JM (1955). Lattice expansion and rheological behaviour in watermontmorillonite systems. Clays and Clay Minerals 3, 296–316. Gaudin A, Grauby O, Noack Y, Decarreau A and Petit S (2004). Accurate crystal chemistry of ferric smectites from the lateritic nickel ore of Murrin Murrin (Western Australia). I. XRD and multi-scale chemical approaches. Clay Minerals 39, 301–315.
Gaudin A, Decarreau A, Noack Y and Grauby O (2005). Clay mineralogy of the nickel laterite ore developed from serpentinised peridotites at Murrin Murrin, Western Australia. Australian Journal of Earth Sciences 52, 231–241. Gieseking JE (Ed.)(1975). Soil components Volume 2. Inorganic Components. Springer-Verlag, Berlin. Güven N (1988). Smectites. Reviews in Mineralogy 19, 497–560. Hermann W, Blake M, Doyle M, Huston D, Kamprad J, Merry N and Pontual S (2001). Short wavelength infrared (SWIR) spectral analysis of hydrothermal alteration zones associated with base metal sulfide deposits at Rosebery and Western Tharsis, Tasmania, and Highway-Reward, Queeensland. Economic Geology 96, 939–955. Holland TJB and Powell R (2006). Mineral activitycomposition relations and petrological calculations involving cation equipartition in multisite minerals: a logical inconsistency. Journal of Metamorphic Geology 24, 851–861. Huntington J, Mauger A, Skirrow R, Bastrakov E, Connor P, Mason P, Keeling, J, Coward D, Berman M, Phillips R, Whitbourn L and Heithersay P (2004). Automated mineralogical logging of core from the Emmie Bluff, iron oxide copper-gold prospect, South Australia. In Pacrim 2004 Congress, Proceedings, pp. 223–230. The Australasian Institute of Mining and Metallurgy, Melbourne. Keeling JL and Self PG (1996). Garford Paleochannel palygorskite. MESA Journal 1, 20–23. McQueen KG, Hill SM and Foster KA (1999). The nature and distribution of regolith carbonate accumulations in southeastern Australia and their potential as a sampling medium in geochemical exploration. Journal of Geochemical Exploration 67, 67–82. Macumber PG (1992). Hydrological processes in the Tyrell Basin, southeastern Australia. Chemical Geology 96, 1–18. Meunier A and Velde B (2004). Illite. Springer Verlag, New York. Meunier A (2005). Clays. Springer Verlag, New York. Michel FM, Ehm L, Liu G, Han WQ, Antao SM, Chupas PJ, Lee PL, Knorr K, Eulert H, Kim J, Grey CP, Celestian AJ, Gillow J, Schoonen MAA, Strongin DR, Parise JB (2007). Similarities in
Regolith mineralogy
2- and 6-line ferrihydrite based on apir distribution function analysis of X-ray total scattering. Chemistry of Materials 19, 1489–1496. Miehe G and Graetsch H (1992). Crystal structure of moganite: a new structure type for silica. European Journal of Mineralogy 4, 693–706. Millot G (1970). Geology of clays. Springer Verlag, New York. Milnes AR and Hutton JT (1983). Calcretes in Australia. In Soils: An Australian Viewpoint. pp.119–162. CSIRO, Melbourne. Moore DM and Reynolds RC Jr. (1989). X-ray Diffraction and the Identification and Analysis of Clay Minerals. Oxford University Press, New York. Moss AJ, Walker PH and Hutka J (1973) Fragmentation of granitic quartz in water. Sedimentology 20, 489–511. Newnham RE and Brindley GW (1956). The crystal structure of dickite. Acta Crystallographica 9, 759–764. Nickel EH (1995). The definition of a mineral. Canadian Mineralogist 33, 689–690. Norrish K (1954). The swelling of montmorillonite. Transactions of the Faraday Society 18, 120–134. Norrish K (1975). The geochemistry and mineralogy of trace elements. In Trace elements in soil-plantanimal systems. (Eds DJD Nicholas and AR Egan) p. 55. Academic Press, New York. Norrish K and Rosser H (1983). Mineral phosphate. In Soils: An Australian Viewpoint. pp. 335–361. CSIRO, Melbourne. Ostwald J (1992). Genesis and paragenesis of the tetravalent manganese oxides of the Australian Continent. Economic Geology 87, 1237–1252. Parc S, Nahon D, Tardy Y, and Viellard P (1989). Estimated solubility products and fields of stability for cryptomelane, nsutite, birnessite and lithiophorite based on natural lateritic weathering sequences. American Mineralogist 74, 466–475. Pauling L (1930). The structure of micas and related minerals. Proceedings of the National Academy of Sciences 16, 123–129. Rengasamy P and Olsson KA (1991). Sodicity and soil structure. Australian Journal of Soil Research 29, 935–952 Schulze DG (1984). The influence of aluminium on iron oxides: VIII. Unit cell dimensions of Al-sub-
stituted goethites and estimation of Al from them. Clay Minerals 32, 36–44. Schwertmann U (1988). The occurrence and formation of iron oxides in various pedoenvironments. In Iron in soils and clay minerals. (Eds JW Stucki, BA Goodman and U Schwertmann) pp. 267–308. D Reidel Publishing Company, Dordrecht, Netherlands. Schwertmann U and Kämpf N (1985). Properties of goethite and kaolinite in kaolinitic soils of southern and central Brazil. Soil Science 139, 344–350. Scott KM (1987). Solid solution in, and classification of, gossan-derived members of the alunite-jarosite family, northwest Queensland, Australia. American Mineralogist 72, 178–187. Scott K (1996). Composition of white mica in weathered rocks: indicators of rock type and proximity to gold mineralisation, Western Australia. Explore 93, 3–5. Scott KM and Yang K (1997). ‘Spectral reflectance studies of white micas’. Report 439R. CSIRO Exploration and Mining, North Ryde, NSW. Singh B and Gilkes RJ (1995). The natural occurrence of c-alumina in lateritic pisoliths. Clay Minerals 30, 39–44. Stanjek H and Schwertmann U (1992). The influence of aluminum on iron oxides. Part XVI: Hydroxyl and aluminum substitution in synthetic hematites. Clays and Clay Minerals 40, 347–354. Taunton AE, Welch SA and Banfield JF (2000a). Microbial controls on phosphate and lanthanide distributions during granite weathering and soil formation. Chemical Geology 169, 371–382. Taunton AE, Welch SA and Banfield JF (2000b). Geomicrobiological controls on light rare earth element, Y and Ba distributions during granite weathering and soil formation. Journal of Alloys and Compounds 303–304, 30–36. Taylor G and Eggleton RA (2001) Regolith Geology and Geomorphology. John Wiley and Sons, Chichester, UK. Thiry M and Simon-Coinçon R (1996). Tertiary paleoweathering and silcretes in the southern Paris Basin. Catena 26, 1–26. Thiry N, Milnes AR, Royot V and Simon-Coinçon R (2006). Interpretation of palaeoweathering features and successive silicifications in the Tertiary regolith
71
72
Regolith Science
of inland Australia. Journal of the Geological Society of London 163, 723–736. Thompson JB Jr (1978). Biopyriboles and polysomatic series. American Mineralogist 63, 239–249. Tilley DB and Eggleton RA (1995) Tohdite (5Al2O3. H2O) in bauxites from Northern Australia. Clays and Clay Minerals 42, 485–488. Tilley DB and Eggleton RA (1996). The natural occurrence of eta-alumina (h-Al2O3) in bauxite. Clays and Clay Minerals 44, 658–664. Tilley DB and Eggleton RA (2005). Titanite low-temperature alteration and Ti- mobility. Clays and Clay Minerals 53, 100–107. Turner M, Greene RSB, Knackstedt M, Senden TJ, Sakellariou A and White I (2008). Use of gamma emission computed tomography to study the effect of electrolyte concentration on regions of preferred flow and hydraulic conductivity in deep regolith materials. Australian Journal of Soil Research 101–111. Veblen DR and Ferry JM (1983). A TEM study of the biotite-chlorite reaction and comparison with petrologic observations. American Mineralogist 68, 1160–1168. Veniale F, Delgado A, Marinoni L and Setti M (2002). Dickite genesis in the ‘varicoloured’ clay-shale formation of the Italian Apennines: an isotopic approach. Clay Minerals 37, 255–266. Weaver CE and Beck KC (1977). Miocene of the S.E. United States: A Model for Chemical Sedimentation
in a Peri-Marine Environment. Developments in Sedimentology 22. Elsevier. Weaver CE, Beck KC and Pollard CO (1971). Clay water diagenesis during burial: how mud becomes gneiss. Geological Society of America Special Paper 134. Wells M (2005). Murrin Murrin nickel laterite deposit, WA. In Regolith expression of Australian ore systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius M) pp. 115–117. CRC LEME, Perth. White AF and Brantley SL (Eds) (1995). Chemical Weathering Rates of Silicate Minerals. Reviews in Mineralogy 31. Mineralogical Society of America. Washington DC. Wilcock S (1998). Sediment-hosted magnesite deposits. AGSO Journal of Australian Geology and Geophysics , 247–251. Wilson MJ (2004). Weathering of primary rock-forming minerals: processes, products and rates. Clay Minerals 39, 233–266. Yang K, Huntington JF and Scott KM (1998). Spectral characterisation of hydrothermal alteration at Hishihari, Japan. In Water – Rock Interaction. Proceedings of the 9th International Symposium on Water-Rock Interaction, Taupo, New Zealand. (Eds GB Arehart and JR Hulston) pp. 587–590. AA Balkema, Rotterdam, Netherlands. Zheng H and Bailey SW (1994). Refinement of the nacrite structure. Clays and Clays Minerals 42, 46–52.
5
Regolith geochemistry Kenneth G McQueen
5.1 INTRODUCTION The regolith is composed of minerals, water, dissolved compounds, colloids, biota and gases. All of these components can be described and investigated in terms of their elemental constituents. This is the basis of regolith geochemistry, which applies chemical principles to understanding the nature, origin and behaviour of the regolith. Mineral explorers commonly determine the abundances of target and pathfinder elements in regolith materials during exploration programs, but regolith geochemistry can be much more widely applied. For example, geochemical data from the regolith can be used to:
s s s s
distinguish different regolith materials estimate the degree of weathering and chemical leaching determine the parent rock type locate specific chemical environments, such as evaporative zones, ferruginous zones and boundaries between relatively reduced and oxidised materials within the regolith.
Regolith geochemistry can be used to help understand mineral alteration and formation processes, determine and predict element dispersion, fixation
and fractionation in regolith materials and establish the origin and evolution of different regolith materials. Understanding geochemical properties and processes in the regolith has wide application in environmental studies, soil science, land and water management and medical geology. Regolith geochemists are also involved in developing better sampling and analytical methods for regolith materials. Regolith geochemistry is best understood in the context of regolith geology and, because minerals are such a major component of the regolith, regolith mineralogy. Biota are also significant in the regolith (Chapter 1), so this chapter should be referred to in parallel with Chapters 4 and 6–8.
5.2 FUNDAMENTAL CONTROLS ON ELEMENT BEHAVIOUR The basic chemical properties of elements are a good starting point for understanding their behaviour during geological and weathering processes. Different elements are defined by the number of protons in the atomic nucleus (the atomic number), and each atom has an equal number of positively charged protons and negatively charged electrons. The chemical properties of elements can be largely explained by the way
74
Regolith Science
I
Periodic Table
II
III
IV
V
VI
VII
VIII
1
2
H
He
Atomic number 3
Li 11
Transition Elements
B
12
13
Na Mg 19
K 37
Rb
38
Sr
55
56
Ba
Cs 87
Fr
20
Ca
88
6
5
4
Be
Al 21
Sc
22
Ti
39
40
Zr
Y 57
La*
23
V 41
42
25
26
Mn Fe 43
44
Nb Mo (Tc) Ru
72
Hf
24
Cr
73
Ta
74
W
75
Re
76
Os
27
Co 45
Rh 77
Ir
28
Ni 46
Pd 78
Pt
29
Cu 47
Ag 79
Au
30
Zn 48
Cd 80
Hg
31
Ga 49
In 81
Tl
C 14
Si 32
Ge 50
Sn 82
Pb
7
N 15
P 33
As 51
Sb 83
Bi
9
8
O
F
16
S
84
Po
Ar
35
Br
36
Kr
53
52
Te
18
17
Cl
34
Se
10
Ne
I
54
Xe 86
85
At
Rn
89
Ra Ac** 57
*Lanthanides (REE)
La
**Actinides
Ac
58
Ce
89
90
Th
72
Pr 91
Pa
73
74
75
Nd (Pm) Sm 92
U
93
76
Eu
77
Gd
78
Tb
79
Dy
80
Ho
81
Er
82
83
Tm Yb
84
Lu
94
Np Pu
Chalcophile and lithophile in the crust Chalcophile in the crust
Goldschmidt’s Classification Siderophile
Chalcophile
Lithophile
Atmophile
Biophile
Figure 5.1: Periodic Table with superimposed Goldschmidt Classification.
that their outer electron shells interact with those of other elements. Atoms may gain or lose electrons to form negatively charged ions (anions) or positively charged ions (cations). Some elements may form several types of ions. For example, iron can occur (rarely) in elemental form (Fe 0) as ferrous iron (Fe2+) or ferric iron (Fe3+). Similarly gold may occur as elemental (native) gold (Au0), as aurous gold (Au1+) or auric gold (Au2+). Sulfur can exist in a large number of forms including commonly (S2–, S1–, S0 and S6+). These different forms are referred to as different valence or oxidation states. Ordering the elements in ascending atomic number reveals a periodic character for many properties, including melting point, energy of formation, atomic radius, and first ionisation energy. This discovery by D.I. Mendeleev in 1871 led to the Periodic Table – a fundamental tool for understanding the properties and chemical behaviours of the elements (Figure 5.1). The grouping and position of elements in the Periodic Table reflects the structure of their electron orbitals,
which explains the systematic and periodic pattern to their chemical behaviour. Thus, elements that appear in columns have the same number of electrons in their outer shells and are likely to behave similarly. Goldschmidt (1954) also grouped the elements to reflect their chemical properties and behaviour. The Goldschmidt classification – based primarily on the energy of formation of oxides and sulfides – can be incorporated onto the periodic table (Figure 5.1) and provides a useful overview of elemental associations. Lithophile elements (such as Na, K, Si, Al, Ti, Mg and Ca) generally concentrate in the rock-forming silicate and oxide minerals of the crust and mantle. Siderophile elements (such as Fe, Co, Ni and PGE) have an affinity for iron and are concentrated in the Earth’s ironrich core. Chalcophile elements (such as Cu, Ag, Zn, Pb and S) readily form sulfides. Atmophile elements (O, N, H and the inert gases) are the main components of the atmosphere. Biophile elements (C, N, O, H, P and S) make up the main part of the biosphere. Certain elements in each group tend to be more
Regolith geochemistry
volatile (for example, in the lithophile group K is more volatile than either Mg or Ti). More refractory elements such as Mg and Cr tend to concentrate in solid residues. Some elements are distributed in more than one group according to their different behaviour under conditions of high O2, S or H2O activity and different temperatures. For example, all the siderophile elements have some chalcophile tendency. Because the regolith represents the interface between the lithos-
2.0
phere, hydrosphere, biosphere and atmosphere, most of the groupings in Goldschmidt’s classification can be found within the regolith. Another convenient way to understand and predict element behaviour – particularly in terms of cation distribution in crystalline structures such as minerals – is to arrange the cations in terms of their ionic radii and ionic charge (Figure 5.2; Hall 1987). These properties define the electrical potential in the
Cs
Rb K
LARGE ION LITHOPHILE ELEMENTS Ba Pb Sr Eu
1.6
INCOMPATIBLE ELEMENTS
Crystal radius (A)
Na
La Nd Sm Eu Y Yb
Ca 1.2
Li 0.8
8 fold Th U Ce Pb Zr Hf
Mn Fe Co, Zn Mg Cu Ni
HIGH FIELD STRENGTH ELEMENTS
8 fold U
Sc Lu Au V Cr
COMPATIBLE ELEMENTS
12 fold
Nb Ta
Ti PGEs
6 fold 4 fold
Al Be
0.4
0
+1
+2
Si
+3
6 fold
+4
P
+5
+6
Ionic charge Figure 5.2: Ionic crystal radii in crystal structures versus charge for some major, minor and trace elements (after Hall 1987).
75
76
Regolith Science
neighbourhood of the ion. The positions of cations in terms of these parameters indicate which cations are more likely to enter particular sites in mineral structures or substitute for each other according to Goldschmidt’s rules. These rules state that: 1. Ions of similar radii and same charge will enter a crystal in amounts proportional to their concentration in a surrounding liquid. 2. An ion of smaller radius, but with the same charge as another ion, will be incorporated preferentially into a growing crystal. 3. An ion of the same radius, but with a higher charge than another, will be preferentially incorporated into a growing crystal. Variations in ionic radii of up to 15% should not be regarded as significant. The rules do not take account of the type of chemical bond (that is, the degree of ionic versus covalent bonding) or crystal field effects, and so there are many exceptions. However, they provide a first-order approach to explaining cation distributions in minerals. Using this approach, the elements can be broadly divided into groups that show particular characteristics during melt to mineral partitioning and substitution, such as compatible (easily substituted in the mineral) or incompatible (not easily substituted in the mineral) behaviour. Large ions with weak charges (such as K, Rb, Cs and Ba) are accommodated with difficulty in many minerals and tend to concentrate together, particularly in K-feldspars. These are referred to as the large ion lithophile elements (LILE). Small ions with strong charges (such as Zr, Nb, Th and U) develop intense electrostatic fields and do not readily substitute for the major elements in common minerals. These are referred to as the high field strength elements (HFSE). Factoring in a measure of bond type can refine the predictions of cation behaviour in minerals and melts. This has been done using ionisation potential or electronegativity data for the elements to estimate bond type. Trace elements (elements generally present in abundances of ppm or less) are accommodated in primary minerals, in three main ways: 1. substituted isomorphously for major elements in the mineral crystal structure (that is, as a direct
replacement without substantial change to the structure) 2. randomly incorporated in the mineral structures in a non-isomorphous fashion 3. concentrated as essential structural constituents or non essential constituents in particular accessory minerals. Table 5.1 lists the minor and trace elements that have been observed to occur in some of the rockforming and primary accessory minerals. As minerals weather in the regolith, minor and trace elements in the least altered, or more resistate, minerals are retained. However, many are released or re-incorporated in various ways into regolith materials (see also Appendix 2). There are two possible approaches to determining the partitioning and distribution behaviour of elements during primary mineral formation and subsequent weathering: theoretical and empirical. The complexity of chemical weathering and the wide range of possible conditions mean that, at this stage, most knowledge of element distribution in the regolith is based on empirical observation.
5.3 THE CHEMISTRY OF WEATHERING AND ELEMENT DISPERSION/RETENTION 5.3.1 Basic processes Weathering involves the physical, chemical and biological modification of rocks at, or near, the Earth’s surface. Most minerals formed under igneous and metamorphic conditions are unstable at low temperature and under near-surface hydrous conditions, and eventually react to form dissolved components and new mineral precipitates. The relative stability of the major rock-forming silicate minerals during weathering approximates the crystallisation sequence of Bowen’s reaction series (that is, the least stable minerals are those that crystallise first from magmas at higher temperature, Goldich 1938, Figure 5.3). Another indication of a mineral’s susceptibility to weathering is the ratio of silica to other cations in its structure. The higher the proportion of other cations that can be replaced by hydrogen ions, the more weatherable the mineral. Thus mafic minerals
Regolith geochemistry
Table 5.1: Observed substitutions of trace elements for major elements in some primary rock-forming and accessory minerals.
Feldspars
Olivine
Clinopyroxenes
Micas
Apatite
Zircon
Major elements
Coordination
Trace elements
Ca, Na, K
6–9
Ba, Eu, Pb, Rb, Sr
Al, Si
4
Ge
Mg, Fe
6
Co, Cr, Mn, Ni
Si
4
Ge
Ca, Na
8
Ce, La, Mn
Mg, Fe
6
Co, Cr, Ni, Sc, V
Si
4
Ge
K
12
Ba, Cs, Rb
Al, Mg, Fe
6
Co, Cr, In, Li, Mn, Sc, V, Zn
Si, Al
4
Ge
Ca
7–9
Ce, La, Mn, Sr, Th, U, Y
P
4
As, S, V
Zr
8
Ce, Hf, La, Lu, Th, Y, Yb
Si
4
P
Other minerals
Rock-forming minerals and rock types DISCONTINUOUS SERIES
Ultramafic Olivine
Pyroxene
CONTINUOUS SERIES Mg Fe Ca Na Si Al, Ni, Cr, Mn, V Co, REE Sc
Mafic
Ca Na Si Al Sr
Intermediate
Hornblende
Biotite
Mg Fe K Si Al, Sn, Mn, Zn Sc
Ca-rich plagioclase
elements available
Na Cl Br
Gypsum
Ca S Sr
Sulfides
Chalcophile metals S As Se Te
Calcite Dolomite
Ca, Mg, Fe, Mn, Sr, Ba
Na-rich plagioclase
Felsic Alkali Feldspar + Quartz + Muscovite
Halite
K Al Ba Si Rb, Pb, Li F
Decreasing temperature of crystaliisation
Increasing stability during weathering
Kaolinite Gibbsite Goethite Hematite
Resistate accessory minerals
Figure 5.3: Bowen’s reaction series and mineral stability during weathering.
Si Al Fe Cr P e.g. Rutile, Zircon, Chromite
77
78
Regolith Science
ATMOSPHERE (H22 2N2)
CONTROLLING FACTORS
*(+ 2(" (-*-&(" * $-+-0.'(" T$"2-,("
WEATHERING $"' ,(" * (-*-&(" * 9+("0-!( * "2(4(27 9%0-125$#&(,& 92'$0+ *#(1 &&0$& 2(-, 9.* ,2 9 ,(+ * 9-0& ,-"-+.*$6(,&
Chemical 9#(11-*32(-,.0$"(.(2 2(-, 9'7#0-*71(1 9'7#0 2(-,#$'7#0 2(-, 20 ,1%-0+ 2(-, 90$#-6
BIOSPHERE
'$+(" *%0 "2(-, 2(-, LITHOSPHERE + )-0+(,-0 20 "$$*$+$,21
BEDROCK 0(+ 07+(,$0 *1
-+.-1(2(-, 0 (,1(8$ ,#1' .$
HYDROSPHERE (H2#(11-*4$#1.$"($1
KMQf01 1-07
CHEMICAL COMPONENTS
REGOLITH
$1(#3 *
In situ
$"-+!(,$#
0-1(-,#$.-1(2(-,
(1.$01$#
T0 ,1.-02$#
$1.-,1$2-#(1$/3(*(!0(3+
0-3,# ,# 130% "$5 2$01
S 0-3,#5 2$0"'$+(1207 7#0-*-&("%*-5 $0+$ !(*(27
$.-1(2(-,
0-1(-,
Figure 5.4: The principal controls on weathering and the chemical composition of the regolith.
(olivine, pyroxenes and amphiboles) generally weather much more readily than the felsic minerals (plagioclase, K-feldspars, micas and quartz). Elements hosted by mafic minerals therefore tend to be the first released during weathering, followed by elements in plagioclase, alkali feldspars and micas (Figure 5.3). The initial weathering products of the primary rockforming minerals may undergo further weathering with further release of some of their contained elements. The common carbonate minerals are relatively unstable under acidic conditions and sulfides are particularly susceptible to oxidation – being among the least stable minerals under most near surface conditions (Figure 5.3). The main controls on rock and mineral weathering, element dispersion/retention and the chemical composition of the resulting regolith are summarised in Figure 5.4. The nature and rate of weathering are affected by climatic and biological conditions, as well as by geomorphic and tectonic factors that control surface relief. Hydrologic processes at, and below, the surface exert a strong control on the extent of element dispersion and material transport.
The principle chemical processes of mineral weathering involve replacement of more soluble ions by protons (hydrogen ions) and oxidation of some elements. This is promoted by the presence of water (or more commonly aqueous solutions with dissolved components), gases and biological activity. Cations with the highest solubilities in surface and groundwaters are Mg2+, Ca2+, Na+ and K+. The oxides of these cations make up between 13–20% of crustal rocks – depending on rock type. A major control on solubility is the charge on such cations. Generally +1 and +2 charges tend to be soluble for smaller cations, but +3 and +4 charged ions are insoluble. Dissolution/ precipitation
Reduction/ oxidation
Hydrolysis
Hydration/ dehydration and transformation
Figure 5.5: Types of chemical reactions involved in weathering.
Regolith geochemistry
Chemical weathering reactions fall into four main groups: dissolution/precipitation reactions; hydrolysis reactions; hydration/dehydration and transformation processes; reduction-oxidation (redox) reactions (Figure 5.5), with combinations also possible. Dissolution/precipitation reactions
Some compounds dissolve in water by dissociation of ions that go into solution. This is the case for halite: NaCl = Na+ + Cl-
(Eqn 5.1)
The water itself is not affected and the mineral is entirely or congruently dissolved. Other minerals react with water and dissolve congruently as, for example, in the reaction of quartz with water: SiO 2 + 2H 2 O = H 4 SiO 4
(Eqn 5.2)
More commonly, minerals react with water or other dissolved components to form new more stable compounds and species in solution. This is referred to as incongruent dissolution. The alteration of K-feldspar to kaolinite in CO2-rich groundwater, with potassium and some silica going into solution, is an example: 2KAlSi 3 O 8 + 2CO 2 + 11H 2 O = Al 2 Si 2 O 5 (OH) 4 + 4H 4 SiO 4 + 2K+ + 2HCO3 (Eqn 5.3) A range of naturally occurring acids – particularly carbonic acid (H2CO3), sulfuric acid (H2SO4) and various humic acids – can promote dissolution. Examples of acid attack reactions include: the alteration of olivine by carbonic acid: Mg 2 SiO 4 + 4H 2 CO 3 = 2Mg 2 + + 4HCO3 + H 4 SiO 4
(Eqn 5.4)
the dissolution of calcite by carbonic acid: CaCO 3 + H 2 CO 3 = Ca2 + + 2HCO3 (Eqn 5.5) the reaction of calcite with sulfuric acid: CaCO 3 + H 2 SO 4 + 2H 2 O = CaSO 4 .2H 2 O + H 2 CO 3
(Eqn 5.6)
albite dissolution by acid and formation of kaolinite:
2NaAlSi 3 O 8 + 2H + + H 2 O = Al 2 Si 2 O 5 (OH) 4 + 4SiO 2 + 2Na+
(Eqn 5.7)
Under very acid conditions (pH <4), Si is moderately soluble (>5 mg/L) and Al and Fe also become more soluble. Such acid environments can occur around weathering sulfide deposits or in the upper part of organic-rich weathering profiles. The formation of organo-metal complexes under acid conditions can further promote the leaching of Fe and Al – for example, in the breakdown of aluminosilicates (such as anorthite) to release Al as a soluble organic complex and precipitate quartz (Trescases 1992): CaAl 2 Si 2 O 8 + 8H + + (organic anion) = Ca2 + + 2 (Al 3 + + organic anion) + 4H 2 O + 2SiO 2 (Eqn 5.8) Under very alkaline conditions (pH >9.6) Al and silica become highly soluble as Al(OH)4 – and H3SiO4 –, respectively. These conditions can develop in arid environments where there is strong evaporative concentration of dissolved components (particularly carbonate) in surface and groundwaters. Alumino–silicates breakdown to release Al and silica and precipitate components less soluble under the alkaline conditions (such as Ca as calcite; Trescases 1992): CaAl 2 Si 2 O 8 + 4OH - + 5H 2 O + CO 2 = 2Al (OH) 4 + 2H 3 SiO 4 + CaCO 3
(Eqn 5.9) For all these various reactions, dissolved components may re-precipitate under equilibrium conditions or where the reaction is reversed due to changes in the chemical environment. Hydrolysis reactions
Reactions of minerals with water in which the hydroxyl group of the water molecule remains intact are termed hydrolysis reactions. These types of reactions are particularly important in the weathering of many silicate minerals. The alkali felspars are considered to undergo a series of step reactions that ultimately convert them to kaolinite and dissolved K+, Na+ and silica:
79
80
Regolith Science
2 (K, Na) AlSi 3 O 8 + 11H 2 O = Al 2 Si 2 O 5 (OH) 4 + 4H 4 SiO 4
Ca2 + + SO 24 - + 2H 2 O = Ca2 SO 4 .2H 2 O
+ K+ + Na+ + 2OH -
(Eqn 5.16)
(Eqn 5.10) Hydrolysis of olivine releases Fe2+ and Mg2+ and silica: (Fe, Mg) 2 SiO 4 + 4H 2 O = H 4 SiO 4 + Fe 2 + + Mg 2 + + 4OH (Eqn 5.11) Hydrolysis of dolomite releases Ca 2+ and Mg2+ and HCO3 – ions: CaMg (CO 3) 2 + 2H 2 O = Ca2 + + Mg 2 + + 2OH - + 2HCO3 (Eqn 5.12) Many of these reactions involve H+ or OH– and are therefore pH dependent.
Ba 2 + + SO 24 - = BaSO 4
(Eqn 5.17)
Transformation processes involving exchange of ions are particularly important in the phyllosilicates (sheet silicates), where the arrangement of oxygen and silica atoms is retained, but other ions are replaced (such as the transformation of tri-octahedral micas into vermiculite by the replacement of K by hydrated cations: Eqn 4.2; Chapter 4). The transformation of serpentine to garnierite by the partial replacement of Mg in the structure by Ni is another example (Trescases 1979, 1992: Mg 3 Si 2 O 5 (OH) 4 + Ni 2 + = (Mg 2 Ni) Si 2 O 5 (OH) 4 + Mg 2 +
(Eqn 5.18)
Hydration/dehydration and transformation processes
Hydration and dehydration of the regolith can precipitate or modify minerals. Other transformation processes involve the exchange of ions within the crystal structure of particular minerals. Dehydration processes are important in the formation of different Fe compounds in the regolith: for example, formation of goethite from hydrated Fe3+ hydroxide (or other more complex hydrated hydroxides): Fe(OH) 3 .nH 2 O = FeO (OH) + (n + 1) H 2 O (Eqn 5.13) and dehydration of goethite to hematite: 2FeO (OH) = Fe 2 O 3 + H 2 O (Eqn 5.14) Evaporative removal of water increases the concentration of dissolved components until they reach saturation level. Compounds or new minerals can then precipitate. Examples include the formation of secondary carbonates, such as calcite, from Ca 2+ and CO32– in solution: Ca2 + + CO 23 - = CaCO 3
(Eqn 5.15)
and the formation of gypsum and barite by combination of Ca2+ and Ba2+ (respectively) with SO42– from solution:
Reduction–oxidation (redox) reactions
Redox reactions involve the transfer of electrons in the presence of an oxidising or reducing agent (Figure 5.6). For example in the reaction: 4FeO + O 2 = 2Fe 2 O 3
(Eqn 5.19)
Fe2+ is oxidised (loses an electron) to form Fe3+ and the molecular O2 is reduced (gains electrons) to O2–. This is more obvious if the ionic charges are included: 4Fe 2 + O 2 - + O 02 = 2Fe 32 + O 23 - (Eqn 5.20) Manganese is oxidised in a similar way, with Mn 2+ transforming to Mn3+ or Mn4+. For example: 2Mn 2 + CO 23 - + O 2 + 2H 2 O = 2Mn 4 + O 22 - + 2H 2 CO 3
(Eqn 5.21)
Redox reactions are particularly important in the weathering of metal sulfides – the most important of these being pyrite, which is illustrated by: 4FeS 2 + 15O 2 + 14H 2 O = 4Fe(OH) 3 + 16H + + 8SO 24 -
(Eqn 5.22)
Regolith geochemistry
Oxidising Agent
Reduced Species
O
10FeS + 2SO 24 - + 8H 2 O = 6FeS 2 + 4Fe(OH) 3 + 4OH (note this reaction increases pH) (Eqn 5.27)
O Reduction (electron gained)
or
e-
6FeS 2 + 18NO3 + 12H 2 O =
Oxidation 2+
(electron lost)
3+
Fe
Fe KMQf012-07
Reducing Agent
Oxidised Species
Figure 5.6: An example of the reduction–oxidation process.
This reaction of pyrite to Fe3+ hydroxide involves the oxidation of both Fe2+ to Fe3+ and S – to S6+ and proceeds by a series of steps, for example: 4FeS 2 + 14O 2 + 4H 2 O = 4Fe 2 + + 8SO 24 - + 8H +
(Eqn 5.23)
4Fe 2 + + O 2 + 4H + = 4Fe 3 + + 2H 2 O
(Eqn 5.24)
4Fe 3 + + 12H 2 O = 4Fe(OH) 3 + 12H +
Combined reactions
Many chemical weathering reactions involve a combination of reaction types, as already illustrated by some of the examples above. The alteration of fayalite (Fe- olivine) to Fe oxide involves both oxidation and hydrolysis: 2Fe 2 SiO 4 + O 2 + 4H 2 O = 2Fe 2 O 3 + 2H 4 SiO 4
(Eqn 5.29)
Alteration of biotite by dilute carbonic acid to kaolinite and goethite involves, oxidation and carbonation: 4KMg 2 FeAlSi 3 O 10 (OH) 2 + O 2 + 20H 2 CO 3 = 4KHCO 3 + 8Mg (HCO 3) 2 + 4FeO ( OH) + 2Al 2 Si 4 O 10 (OH) 2 + 4SiO 2 + 10H 2 O (Eqn 5.30)
(Eqn 5.25)
The reactions produce a large number of H+ ions (1 mole of pyrite produces 4 moles of H+) so that the weathering solution may become very acid. This acid can assist in the dissolution of other minerals including silicates, carbonates and other sulfides. Ferric iron (Fe3+) can also act as an oxidising agent to breakdown pyrite and other sulfides by the reaction: FeS 2 + 14Fe 3 + + 8H 2 O = 15Fe 2 + + 16H + + 2SO 24 -
9N 2 + 6Fe(OH) 3 + 12SO 24 - + 6H + (Eqn 5.28)
(Eqn 5.26)
This reaction produces yet more H+ ions. Ferric hydroxide is transformed to stable goethite and hematite by dehydration (Eqns 5.12 and 5.13 above): In natural groundwaters, dissolved sulfate and nitrate may also act as oxidising agents for alteration of sulfides such as pyrrhotite and pyrite (see also Chapter 10). For example:
5.3.2 Biogenic processes and bacterial action Vegetation extracts water, nutrients and other dissolved components from the regolith. Root systems can produce and exude organic compounds and acids into the adjacent regolith (rhizosphere) that help to break down minerals and organic material (Chapter 8). Micro-organisms are abundant in the regolith and many play an important role in biochemical weathering. Their activities can increase the rate at which some reactions occur – as they make use of the energy released – or they can change the environmental conditions. Some bacteria oxidise or reduce certain elements as part of their metabolism. These particularly include Fe, Mn and S, but probably any element that can exist in multiple oxidation states. For example, Ferrobacillus bacteria oxidise ferrous iron, Thiobacillus oxidise sulfide. Other bacteria, such as Desulfovibrio, reduce compounds such as sulfate. Micro-organisms can
81
Regolith Science
5.3.4 Eh-pH Chemical interactions between the elements are largely controlled by what happens in the electron shells of atoms – including exchange, loss and addition of electrons. The activity of protons, measured as pH (-ve log of H+ ion concentration), and the activity of electrons, measured as Eh (redox potential or electrode potential, ep in volts), are therefore important controls on chemical reactions and the stability of minerals and ions in solution. Typically, cations are more soluble at low pH and anions more soluble at high pH. Elements with multiple oxidation states are particularly affected by changes in Eh. Under reducing conditions the dissolved ions can exist in the lower oxidation state (such as As3+, Sb3+, Mo3+, Fe2+, Mn2+, Au+ and Cu+), but if the system becomes more oxidising they will tend to lose electrons and be present in
+0.8
+0.6
+0.4
+0.2
Water oxidised r li
mi
to
fw
ate
rs
tab
ilit
y
ALKALINE LAKES
Fe in solution
pe
OCEANS
Up
3+
RIVERS
+1.0
ACID MINE WATER
5.3.3 Trace element behaviour during weathering As the rock-forming minerals weather and chemically breakdown, the trace elements they contain as isomorphous/non-isomorphous substitutions are released to go into solution or form new phases. Trace elements hosted in accessory mineral inclusions may also go into solution depending on the stability of their host mineral. The behaviour of the trace elements will be controlled by the chemical conditions of the weathering environment – particularly water content, availability of complexing and oxidising agents (such as O2, CO2, SO42–, NO3 –, halides and organic compounds), redox potential and H+ ion activity (Eh-pH) and, to a small degree, temperature. The major weathering reactions help to produce or modify these conditions. The main external drivers of these chemical conditions are climate, biological activity, parent rock composition, topography and time. New regolith minerals will take up particular trace elements by incorporation or adsorption. Trace elements that remain in solution over prolonged periods, or that form volatile components, can be significantly transported.
the higher oxidation state (such as As5+, Sb5+, Mo6+, Fe3+, Mn4+, Au2+ and Cu2+). This change in oxidation state may involve a marked change in solubility, with the smaller, more positively charged ions of the more oxidised state generally being less soluble (for example, compounds with oxidised Fe3+ and Mn4+ are usually much less soluble than Fe2+ and Mn2+). Some elements that readily oxidise to 5+ or 6+ states (such as Mo, V, Cr and W) are so highly charged they readily react with oxygen and hydroxide to form large negatively charged oxy-anions (such as MoO42–, VO43–, CrO42– and WO42–) that are more soluble. Under strongly oxidising and acidic conditions, U6+ forms the soluble uranyl cation (UO22+). Eh-pH diagrams are a useful tool to examine and predict the composition of natural aqueous solutions, the oxidation state of ions in solution and the nature of
RAIN
change the pH and also produce specialised compounds that react with minerals, combine with metals or cause significant changes in solubilities (Chapter 7).
Redox potential (Eh)
82
ZO N GR E O OU F FLU ND Goethite CT WA UA TE TIO R FeO(OH) N
2+
Fe in solution
BOGS
0
WATERLOGGED SOILS
-0.2
-0.4
-0.6 KMQf004-07
Lo MARINE we r li EUXINIC mi to fw ate Magnetite rs Water reduced tab Fe3O4 ilit y
2
4
6 pH
8
10
Figure 5.7: The range of Eh-pH conditions for near surface and weathering environments, with the upper and lower stability limits of water. Also shown are the predominance fields for Fe species under surface conditions (25°C and 101.3 kPa).
Regolith geochemistry
Table 5.2:
Examples of inorganic target and pathfinder element complexes that can be important in the regolith.
Element
Species and complexes in the regolith
Antimony
Sb(OH) 2+, Sb(OH)3o, Sb(OH) 5o, Sb(OH) 6 –
Arsenic
HAsO42–, H2 AsO4 –, H3AsO4o
Bismuth
Bi3+, BiOH +, Bi(OH)2+, Bi(OH)3o
Cadmium
Cd2+, CdSO4o, CdCO3o
Chromium
CrOH2+, CrOH3, CrO42–
Copper
Cu2+, CuSO4o, CuCO3o, Cu(CO3) 22–, CuHCO3 +, CuCl +, CuClo
Gold
Au(S2O3) 23–, AuCl4 –, Au(CN)2–
Iron
Fe2+, FeSO4o, FeOH2+, Fe(OH)2+, Fe(OH)3o
Lead
PbCO3o, Pb(CO3) 22–, PbHCO3 +, PbSO4, Pb(SO4) 22–, Pb(OH)2o, PbCl +, PbCl2o
Manganese
Mn2+, MnSO4o, MnHCO3 +, MnCO3o
Mercury
HgOH +, Hg(OH) 2o, HgCl +, HgCl2o, HgCl3 –, HgCl42–
Molybdenum
MoO42–, HMoO4 –, H2MoO4o
Nickel
Ni2+, NiOH +, Ni(OH)2o, NiSO4o, NiCl +, NiCO3o
Palladium
Pd(OH)2, Pd(OH)3 –, Pd(OH) 42–
Selenium
SeO32–, SeO42–, HSeO4 –, H2SeO4
Silver
Ag +, AgOHo, AgClo, Ag(S2O3) 23–,
Uranium
U2O42+, (UO2)3 (OH) 5 +, UO2 (CO3)34–,
Vanadium
VO43–,
Zinc
Zn2+, ZnSO4o, ZnCO3o, ZnHCO3 +, ZnOH +, Zn(OH)2o
Compiled from Thornber (1992); Leverett et al. (2004); Plant et al. (2005); Callender (2005); Williams (1990); Baes and Mesmer (1976); Cameron and Hattori (2003).
solid phases in equilibrium with the solution. Figure 5.7 shows the typical range of Eh-pH conditions found in near-surface environments and the predominance fields for Fe species. The predominance fields for various dissolved species and solid mineral phases will also depend on the abundances of other constituents in solution (such as CO2, Cl–, O2, SO42– and S2–). If these are known – or can be estimated – it is possible to broadly predict how minerals and dissolved ions will behave as Eh or pH conditions vary (such as within a weathering profile or around a weathering sulfide deposit). Non-equilibrium conditions and the great chemical complexity of many weathering environments complicate detailed predictions. 5.3.5 Metal complexing Many elements have low solubilities in ionic form – particularly at the low temperatures and pressures found in the regolith. However, they may be signifi-
cantly more soluble when combined with other elements or radicals as metal complexes. Complex formation is common for the transition elements (Figure 5.1). Inorganic complexes thought to be important in the regolith are with oxygen (O2–) and hydrogen (H+), chloride (Cl–), sulfate (SO42–), thiosulfate (S2O32–), bicarbonate (HCO3 –) carbonate (CO32–), nitrate (NO3 –) and phosphate (PO43–) (see Table 5.2). The particular complex species that may be present, and their coordination with the relevant cation, is strongly dependent on the chemical conditions – particularly the total concentration of the complexing agent in solution. Where decaying organic material is abundant, organic complexing can be important with oxalate (C2O42–), acetate (CH3COO –), fulvic and humic acids as potential complexing agents (Thornber 1992; see also Chapter 8). The dissociation of these organic anions is affected by pH so that complex formation is pH dependent.
83
84
Regolith Science
Rainfall
Lag
Su
rfa
f ce
low
a
w nd
ind
Biogenic dispersion
Gossan Erosion Electrochemical dispersion
Cations
Mechanical dispersion
A Eh
B pH
Gaseous dispersion
Electrons
Anions Conductive sulfide deposit Buried ore deposit
Hydromorphic dispersion
Biogenic dispersion
Hydromorphic dispersion Hydromorphic dispersion Seismic pumping
7 -0
Eh-pH gradient
Sulfide ore deposit with primary dispersion halo
Gravity
Bedrock 4 01 Qf
Water table Weathering front
Transported regolith
KM
Figure 5.8: Types of element dispersion, with examples for weathering ore deposits.
5.3.6 Element dispersion Elements can migrate, or be transported, from their original mineral sites by a number of dispersion mechanisms (Figure 5.8). Known dispersion mechanisms can be grouped as: 1. 2. 3. 4. 5.
hydromorphic (aqueous) electrochemical biogenic gaseous mechanical.
Hydromorphic dispersion
Hydromorphic dispersion of elements or their complexes in solution can occur by diffusion and groundwater flow. Diffusion is the movement of ions or complexes through a medium without the movement of the medium. Groundwater flow is a mass movement or advective process driven through intergranular space by hydrostatic pressure. The degree to which the elements migrate before they become relatively fixed in the regolith by adsorption or by precipitation of insoluble minerals is generally referred to as their mobility. Mobility can be complex and multi-stage,
with Eh-pH changes, element/complex solubility and activity of the complexing agents (such as O2, CO2, S, Cl and P) as major (but not the only) chemical controls. Ionic mobility in the surficial environment can be predicted to a large degree by ionic radii and oxidation state – reflecting Eh-pH conditions (Figure 5.9). Chemical gradients within the regolith, such as changes in Eh or pH (such as around a weathering sulfide ore body), can produce zonal dispersion patterns of different elements. Some elements may also be transported by groundwater in colloids (very fine, <0.1 µm, particles intermediate between those in true solution and those in suspension). Iron and Mn oxides commonly occur as colloids in the regolith and are probably important in dispersing attached transition elements (Southam and Saunders 2005). Aqueous transport is both vertical (mostly downward) and horizontal, with lateral dispersion depending on groundwater flow direction and hydraulic gradient. Variations in temperature and density within the groundwater column can result in buoyancy gradients and cause groundwater convection (Domenico and Schwartz 1998, Simmons et al. 2001). Capillary action, combined with evaporation and seismic pumping (where groundwater
Regolith geochemistry
2.4 Te
2.0
I
Se
Br
S
Cl
OXIDES 1.6
Ionic radius (Å)
Hg O
Au
(OH) F
Ag PGE
Cu
1.2
K Au
Ba Pb
Hg Ag
Bi Hg
Na
Cd Pd
0.4
ANIONS 0.0 -3 KMQf051-07
-2
METALS -1
0
REE
Te Th
U Au Pt V Sb Zn Zr Cu Sn Mn Mg, NiCr W Mo Fe Co Mn Ti V As Mn Fe Se Al Si S
Cu
0.8
Ca
CATIONS +1
+2
Sb
U Te
V
Mo
As P
Cr Se
W
S
OXYANIONS +3
+4
+5
+6
Oxidation state
Figure 5.9: Ionic radius for octahedral coordination versus charge and typical form of element occurrences in the regolith (after Thornber 1992). Elements present as anions, cations and oxyanions tend to be more mobile than metal and oxide forms.
is expelled along compressed fractures and faults during earthquakes), has been documented as a mechanism for upward vertical movement of groundwater and contained elements (Cameron et al. 2002; Mann et al. 2005). Except for seismic pumping, hydromorphic dispersion is largely limited by the height of the water table and overlying capillary fringe – typically extending up to several metres above. It is also strongly influenced by surface and sub-surface flow vectors within the regolith. Where water tables are deep, dispersed elements may reach the surface following erosion and surface lowering. The typical relative aqueous mobilities of elements (as ions and some complexes) in the regolith in terms of broad Eh-pH conditions are shown in Figure 5.10. These are largely based on empirical observations and somewhat generalised.
Electrochemical dispersion
Electrochemical dispersion occurs where a natural redox (Eh) gradient is developed over a weathering conductor, such as a sulfide deposit (Govett and Atherden 1987). Electrons conducted from lower Eh at depth to higher Eh near the surface are balanced by the movement of dissolved ionic species in pore or groundwater surrounding the conductor. There is a net movement of positive ions (such as H+ and metal cations) to the top or cathode of the conductor and a net movement of negative ions towards the bottom or anode (Figure 5.11). The greatest concentration of migrating ions is predicted to be where the current densities in the conductor are greatest at the margins, with a relative deficiency over the negative centre at the top of the conductor. This gives rise to double peak or ‘rabbit-ear’ anomalies across the conductor
85
Regolith Science
Oxidising Acid
Oxidising Neutral to Alkaline
Very high mobility B Br Cl I S He Rn
B Br Cl I S He Rn
Very high mobility High mobility F Na Mg F Sr Pd Se U V
Moderate mobility Al Si Ag Au Hg Sb REE
Moderate mobility Ca Na Mg Ag As Cd Cu Mo Ni Co Zn REE
Low mobility K Li P Rb Si As Ba Be Bi Ge Mo Pb Se Cs Tl
Low mobility K Li Mn P Si Rb Ba Be Bi Cs Pb Ge Hg Sb Tl
Very low mobility Fe Mn Ti Cr Ga Nb Pt Sn Ta W Zr +0.4
Approximate Eh range (volts)
+0.4 to +1.0
High mobility Ca F Li Mg Na Cd Co Cu Ni U V Se Y Zn
Very low mobility Al Fe Ti Cr Te W Nb Ta Pt Sn Th Zr
Reducing Acid
Reducing Neutral to Alkaline
Very high mobility Br Cl He I Rn
Very high mobility Br Cl He I Rn
-0.2 to +0.4
86
High to Moderate mobility Ca F Fe K Mg Mn Na Sr Cd Cu Ni Pb Zn
High to Moderate mobility Ba Ca F Li Mg Mn Na Sr
Low to very low mobility K Fe P Rb Si Ag As Ba Co Hg Li
Low to very low mobility K Fe P Rb Si S Ag As Au Ba Bi Cd Co Cu Cr Hg Li Mo Ni Pb Pt REE Sb U V Zn
KMQf054-07
<4
5 Approximate pH range
6-8
Figure 5.10: The typical relative aqueous mobilities of elements (as ions and complexes) in the regolith in terms of broad Eh-pH conditions (from Perel’man 1977 cited in Levinson, 1980; Gray 2001; Årström and Deng, 2003; Cameron and Hattori, 2003). The mobilities are generalised and do not take into account groundwater flux or biological processes. The presence of Fe and Mn oxides commonly affects the mobility of many transition element cations.
(Govett 1976). Cations concentrated at the upper edges of the sulfide body may be further dispersed by O2 O2 O2 More noble sulfide Cathode e-
Reduction Cations M+ Anions X
Anode More reactive sulfide
Oxidation Electrolyte
Figure 5.11: Electrochemical processes during sulfide weathering (after Thornber and Taylor 1992).
other mechanisms. Dispersion of highly mobile ions – particularly H+ – may result in surface anomalies. The diffusion of H+ to the surface can modify pH conditions in the regolith and in turn mobilise other elements. Biogenic dispersion
Organisms play an important role in transferring elements to different parts of the regolith. This includes vegetation cycling – or phyto dispersion – in which plants take up dissolved components in their roots and transport them to other tissues (Chapter 8). As these decay, or are shed onto the surface, the contained elements are transferred to the upper part of the regolith. Significant concentrations of trace elements have been shown to develop in vegetation litter
Regolith geochemistry
over buried mineralisation (for example, Anand et al. 2007), and presumably over time this may produce anomalous concentrations in the soil. The activities of burrowing organisms – including termites and other invertebrates, worms and some mammals – can transfer and mix material in the regolith. This is referred to as bioturbation and generally affects the upper 2–5 metres of the regolith (the biomantle: Figure 8.17), although some species of termites can bring material to the surface from depths of 70 m (Lock 1985). Micro-organisms colonise mineral surfaces as biofilms and complex associations of cells. Their activities, which involve extraction of elements from minerals and precipitation of various metal ions on their surfaces to produce fine-grained minerals, are an important control on element dispersion in the regolith (for example, Southam et al. 1995; see also Chapter 7). Microbial metabolism also affects the kinetics of many hydrochemical processes – particularly sulfide oxidation and other electron transfer reactions (Nealson and Stahl 1997; Chapter 7). Gaseous dispersion
Gases generated by chemical weathering, element release and biological activity can be highly mobile through the regolith, depending on barometric pressure, rainfall and temperature variations (Hale 2000). Gas species that are thought to be important (Klusman 1993) include:
s s s s s s s
CO2 O2 He Rn possibly volatile elements, such as As, Hg, I, Se sulfur compounds, such as SO2 COS, CS2 and H 2S light hydrocarbons.
Gas transport mechanisms include diffusion through pore spaces in response to concentration gradients, advection due to pressure gradients and bubble migration through groundwater from buoyancy. Microscopic gas bubbles formed by phase separation in groundwater have the potential to transport ions and particles attached by electrostatic or hydrophobic forces (Goldenberg et al. 1989; Wan et al. 2001).
Mechanical dispersion
Mechanical dispersion involves physical transport of rock clasts, mineral particles and precipitates containing the elements of interest. It is largely a near-surface process related to erosion or mass movement, and is typically the predominant dispersion mechanism under arid (limited water for reaction/dissolution), arctic (frozen water and glacial transport) and high relief (mass movement) conditions. It is also important where there is high water flow across the land surface or in channels. Mechanical dispersion is the main dispersal mechanism for elements concentrated in resistate or insoluble secondary minerals (for example, Sn in cassiterite, Zr in zircon, Ti in rutile, Pb in cerussite, Ba in barite). 5.3.7 Adsorption, precipitation and ionic exchange Available ions (and compounds) can be attracted and attached by their electrostatic charge to the surfaces of minerals, colloids and organic materials in the regolith. This is referred to as adsorption. The release of these ions – generally under different conditions – is referred to as desorption. Ions on the surface can also be replaced by ions of like charge in solution and this process is called ionic exchange. The charge on a particle surface can be inbuilt during mineral formation (that is, structural) or the result of adsorption or desorption of particular ions onto the surface. In the case of the inbuilt charge, the charge is ‘permanent’ in the sense that a change in solution pH does not affect the surface charge. In the case of the adsorbed or desorbed ions, if the chargeproducing ion is H+ bonded to surface hydroxyls, then a change in pH will alter the distribution of protons (H+) on the surface, and hence the surface charge. This means that an increase in surface acidity (low pH) promotes greater surface hydroxyl protonation, and therefore greater surface positive charge, whereas an increase in solution alkalinity (high pH) will result in dissociation of protons from the surface hydroxyls, leading to increased surface negative charge. At some pH between the extremes of acidity and alkalinity there can be equal numbers of protonated and de-protonated surface sites and the overall charge will be zero. This is referred to as the point of zero charge
87
88
Regolith Science
(pzc) and can be used to give an indication of whether a mineral surface is likely to be positively or negatively charged at any particular pH (Table 5.3). The adsorption of a particular anion or cation is also influenced by the concentration of competing ions or charged complexes.
Cation exchange capacity
By far the most common ionic exchange is cation exchange, where positively charged cations are bound to negatively charged particle surfaces. The most abundant exchangeable cations are Ca2+, Mg 2+, Na+, K+, Al3+, H+ and NH4+ – although trace element
Table 5.3: Typical cation exchange capacities (CEC) for minerals and soils and point of zero charge (pzc) values for minerals. Typical CEC values Minerals
CEC
(cmoles +/kg)
pH=7
Point of zero charge pH values at pzc
Clay minerals Kaolinite
3–12*
3.3–6
Illite
10–40#
3.3–6
Smectites
80–150
2.5–6
Halloysite
10–50
Palygorskite
16–21
Nontronite
140
Chlorite
10–40
Vermiculite
100–150 9.5
Gibbsite Fe oxides Hematite
<100
6.5–8.6
Goethite
<100
7.6–8.4
Ferrihydrite
<500
6.9
240
2.2
Mn oxides Birnessite MnO2
2–4.5
Quartz
1–2.5
Corundum
8.5–9.1 Typical CEC for different soil conditions
Soils
CEC (cmoles +/kg)
CEC (cmoles +/kg)
Soil type
High pH, high organic matter
Low pH, low organic matter
Krasnozem
10–20
2–6
Chocolate
30–40
3–7
Podzolic
3–10
Alluvial
light and sandy
Heavy clay
10–20
20–30
Dune sand
0–5
*CEC increases slightly with decreasing degree of kaolinite crystallinity. # Pure illite has CEC<5 cmole +/kg; increasing interstatified smectite increases CEC. Source of data: Stumm and Morgan (1996); Salomons and Förster (1984); Jaynes and Bigham (1986); Davis and Kent (1990); Thornber (1992 and references therein); Velde (1995); Post (1999); NSW Department of Primary Industries, Agriculture (2007).
Regolith geochemistry
equivalent to the previously used millequivalents/100 g. Figure 5.12 shows that within an elemental group, the greater the ionic radius the greater the CEC.
Li
10
Relative hydrated ionic radius
9
Group 2 Elements Co-precipitation
8
Na
Mg
7
Ca Sr Ba
6
K Rb Cs Group 1 Elements
5
60
90 100 70 80 Exchange Capacity (cmoles+/kg)
110
Figure 5.12: Cation exchange capacities versus ionic radius for Group 1 and Group 2 elements.
cations may be locally important around weathering ore deposits. The amount of negative charge on a particular material is referred to as the cation exchange capacity (CEC) and this determines the total amount of exchangeable cations that normally balance this charge. The CEC is a function of the particle surface area and the surface charge density. As mentioned above, the negative surface charge can be permanent – as in the case of the clay mineral montmorillonite – or it can vary with the solution pH – as for other clay minerals, goethite, hematite and Al oxides (that is, changing the pH can have an important effect on the number of cations adsorbed because of the changing number of competing H+ ions). Materials with high CEC can be important hosts for trace elements dispersed in aqueous solution. Common regolith minerals with high CEC include smectite and nontronite clays. Other clay minerals such as kaolinite and illite have low CEC. Cation exchange capacities can by measured experimentally – generally at a fixed pH of 7, since pH affects CEC. Typical values for some minerals and regolith materials are given in Table 5.3. Units of charge are usually expressed as the number of centimoles of charge per kilogram (cmoles – /kg and cmoles +/kg respectively), which is numerically
During precipitation of an insoluble compound, a soluble element may be co-precipitated. In this way, certain trace elements can be incorporated into a mineral structure during solid solution formation or recrystallisation of minerals. The extent of co-precipitation of an element correlates with its solubility and adsorption properties – indicating that similar bonding mechanisms are involved. Co-precipitation of cations is greatest at high cation concentrations and high pH values, whereas co-precipitation of anions is favoured by low pH (Thornber 1992). Many trace element cations co-precipitate with Fe and Mn oxides (for example, Cr3+, Mn3+, V3+ into hematite, Fe2O3) or with carbonates (for example, Mn2+, Cd2+, Fe2+ into calcite, CaCO3). 5.3.8 Sulfide weathering and associated element dispersion Due to the high oxidation potential of S and Fe (and some other base metal cations), sulfides are among the least stable minerals during near-surface weathering. Sulfide weathering is of particular interest because many of the ore deposits that are the targets of mineral exploration are accumulations of metal sulfides. Iron sulfides (particularly pyrite, but also pyrrhotite and marcasite) are the most common and abundant sulfide minerals. The main weathering reactions of pyrite have already been outlined. In the regolith, Fe sulfide weathering generally involves microbial activity (see also Chapter 7). Sulfide weathering ultimately converts metal sulfides to residual, stable metal oxides, sulfates, carbonates, phosphates and some halides and native elements. The process involves dissolution and variable dispersion of elements, oxidation of S and other elements (such as Fe, Cu, Au, As and Se) – as well as precipitation, recrystallisation and dehydration of minerals stable in different parts of the weathering profile. For most near-surface sulfide ore deposits the final product is a gossan composed largely of Fe oxides
89
90
Regolith Science
with minor amounts of other stable weathering products and anomalous to trace contents of elements present in the primary sulfides. The oxidation–reduction reactions of sulfide weathering generate electrons at the sites of oxidation and consume electrons at sites of reduction. Electrons may flow through the semi-conducting sulfides where these are sufficiently concentrated (such as massive sulfides), producing an electrochemical or galvanic cell that promotes the redox reactions and weathering process (Thornber and Taylor 1992). The marked change in oxidation state of sulfur commonly produces eight electrons per atom (that is S2– to S6+). Reduction of O2 dissolved in water consumes the electrons generated by the oxidation of the S in sulfides: MS + 4H 2 O = M 2 + + SO 24 - + 8H + + 8e(Eqn 5.31) 2O 2 + 4H 2 O + 8e- = 8 (OH) (Eqn 5.32) The overall reaction is: MS + 2O 2 = M 2 + + SO 24 where M can be a range of metals (Eqn 5.33) Due to the fact that oxidation and reduction take place at different sites (anode and cathode sites respectively) the oxygen in the sulfate is dominantly derived from the water. Galvanic activity can also occur between different sulfide minerals (such as abundant pyrite and other base metal sulfides). Reduction of O2 on the surface of one mineral (such as pyrite) can be combined with electron loss from the other mineral (such as galena) to release cations (such as Pb2+). Cathodic reduction processes cause the associated solutions to become alkaline, whereas anodic oxidation processes result in the production of acids. Acid is also generated when metals that dissolve as cations at the anode sites hydrolyse, for example: M 2 + + H 2 O = M (OH) + + H + (Eqn 5.34) M (OH) + + H 2 O = M (OH) 2 + H + (Eqn 5.35)
Oxidation of Fe2+ to Fe3+ followed by hydrolysis produces additional acid (as previously described in Eqns 5.24 and 5.25 above). The main controls on the rate of sulfide weathering under these conditions are: 1. the availability of O2 or other oxidising agents 2. the rate at which O2 can be reduced at the cathode–solution interface 3. the conductivity of the weathering sulfides and the surrounding electrolyte solution 4. the energy barriers at the anode due to the anodic reactions and the build up of solution products near the interface (Thornber and Taylor 1992). Sulfate and Fe3+ are the major ionic species generated by sulfidic weathering. Release of these and other ions and complexes into solution initiates the dispersion of elements into the surrounding regolith. At low pH, Fe3+ and most other cations are more soluble than they are at neutral to alkaline pH (Figure 5.7; Thornber 1992). The highly acidic conditions in the immediate vicinity of the weathering sulfides allow many metal ions to remain in solution until they encounter higher pH conditions. Reaction of the acidic solutions with minerals in the surrounding regolith, or mixing with more alkaline groundwater, typically leads to an increase in pH. This increase in pH also promotes the precipitation of Fe3+ hydroxide and generation of other Fe and Mn oxides. Elements that are in solution under the mildly acid conditions are then strongly adsorbed onto these minerals. Thus as the pH increases away from a weathering sulfide body many of the dispersed cations (Ag+, Au+, Ba2+ Co2+, Cu2+, Ni2+, Pb2+, V2+and Zn2+) are fixed by adsorption to form a geochemical anomaly (see Chapter 13). Conversely, elements that form oxy-anions, such as As, In, Mo, Sb, U and V, are fixed at low pH due to their negative charge. The extent of the halo for any particular element will thus vary depending on the form in which it is present and its mobility under the prevailing environmental conditions. Elements released during sulfide weathering may also be precipitated as, or taken up, by other secondary minerals that form from the weathering sulfides, such as secondary sulfides, sulfates, arsenates, carbonates, phosphates, chlorides and oxides (see Table
Regolith geochemistry
Element
Affinity for sulfur
Solubility of sulfide
High
Low
Pd Hg Ag* Cu* Bi Cd Sb Sn Pb Zn Hi* Co Fe As Mn
Nature of secondary oxysalt For: Hg - highly insoluble Ag - soluble Cu - soluble
Pb - insoluble Zn - soluble Fe 2+ - soluble Low
High
*elements which show significant supergene enrichment Figure 5.13: Schürmann’s series linking sulfide solubility and metal substitution to supergene enrichment in weathering ore deposits (after Guilbert and Park 1986).
4.2). In most weathered sulfide deposits, these minerals are commonly developed in particular parts of the profile. Elements may be enriched at different levels either by relative residual concentration (that is, their proportion is increased by other elements being removed) or by addition, reaction and substitution to form new minerals enriched in the element. This is referred to as supergene enrichment. Gold is commonly enriched by residual concentration relative to other elements in the upper strongly oxidised and leached parts of weathered Au-bearing sulfide deposits. Significant Au may still be dissolved and physically or chemically mobilised, but this loss is typically less than that of the other ore and rock components (for example, Colin et al. 1993; Butt 2001). Some metals released from relatively soluble compounds in the upper oxidised parts of the weathering profile may react with sulfides in the lower reduced parts of the profile and displace other elements. If the new secondary sulfides have low solubility, this leads to supergene sulfide enrichment. Metals can be arranged in the order of their ability to displace other metals in sulfide minerals (Figure 5.13). Copper and Ag have particularly favourable properties to form supergene enriched zones (Scott et al. 2001). Other elements that can show significant supergene enrichment under particular conditions include Ni and Zn.
5.4 ELEMENT DISTRIBUTIONS AND WEATHERING RELATED FRACTIONATION IN THE REGOLITH 5.4.1 Physical processes affecting regolith geochemistry Physical processes may selectively redistribute different regolith materials, which, in turn, affect the chemical structure of the regolith. Examples of these processes include:
s s s s s s s
down-slope movement and profile collapse under gravity eluviation of particles in suspension sorting and differential movement of materials with different physical characteristics (for example, during surface deflation) residual concentration of resistant regolith components bioturbation of the upper parts of the regolith heave and swell in soils mixing during erosional transport.
Mixing processes act to homogenise the regolith composition, whereas sorting and selective transport lead to fractionation of different materials and their contained chemical components. 5.4.2 Major element variations during weathering As rocks weather, and their mineral constituents change to new, more stable assemblages, the contained elements are preserved in resistate minerals, partly redistributed to new minerals or taken into solution – in some cases to be incorporated in other parts of the weathering profile. Large amounts of material can be removed in solution and, in this context, weathering can be considered a type of chemical erosion. The end result is a weathering-controlled fractionation of the major, minor and trace elements. The major primary, rock-forming minerals are feldspars, quartz, micas, mafic minerals (Chapter 4) and, in some cases, carbonates and volcanic glass (not strictly a mineral). These are ultimately transformed to quartz, kaolinite and Fe and Al oxides (Figure 5.3). The accompanying chemical changes involve progressive loss of Na+, K+, Ca2+, Mg2+ (some Si4+) and retention of Si4+, Al3+ and Fe3+ (Figure 5.14a).
91
92
Regolith Science
quartz + feldspar + H 2O + mica + mafics
quartz + kaolinite (gibbsite) kaolinite illite (smectite) kaolinite + goethite chlorite (hematite)
Si Al Na K Ca Mg Fe
Si Al Fe mobilised Na+ K Ca Mg (Si)
Trace Elements Retained in primary resistate minerals. Dispersed-sequestered to specific hosts: Clays adsorbed Iron oxides incorporated Manganese oxides substituted Stable weathering accessories (e.g. sulfates, carbonates, phosphates, halides)
(b)
Chemical Fractionation
(a) Major Elements Distribution controlled largely by mineral transformations:
Che
mic
al an dM e Frac t i on c h a n i c a atio l n
Solution Leaching and Dispersion
Detrital dispersion LAG
Upper oxidised zone
Lower oxidised zone Primary zone ORE PROFILE
KMQf055-07
Figure 5.14: (a) Summary of major controls on element fractionation during weathering, (b) fractionation processes during element dispersion from a weathering and eroding ore deposit.
Aluminium is usually one of the least mobile of the major elements during weathering because of the very low solubility of Al2O3 between pH 5 and 8 (the typical range for most weathering conditions). If Al loss during in situ weathering of silicate rocks is considered negligible, comparison of major element oxide abundances in fresh rocks with their weathered equivalents – using Al2O3 as a constant component – suggests the following changes under most weathering conditions:
s s s s s s s s
SiO2 – some decrease Fe2O3 – significant increase FeO – almost total conversion to Fe2O3 MgO – significant decrease CaO – almost complete removal (except in environments with high carbonate or sulfate activity) Na2O – almost complete removal K 2O – decrease H2O – major increase.
Materials exposed and mechanically worked at the surface (such as residual lag) undergo a range of transformations with time and changing environmental conditions. These include variable transport and physical degradation (abrasion and rounding), chemical leaching and precipitation of some elements. The most chemically and physically resistant components, such as quartz and hematite (and their contained elements), tend to be residually enriched. This results in
element fractionation by combined mechanical and chemical processes (Figure 5.14b; McQueen and Munro 2003; McQueen et al. 2004). The degree of fractionation will vary depending on the age or life cycle of the material, and there is often surface intermixing of materials with differing maturity. The geochemical and mineralogical features of weathering in common rock types are discussed in Chapter 6. 5.4.3 Regolith-related element associations Primary geological processes (including primary ore formation) result in the association of particular elements in specific minerals or groups of minerals. During weathering, these associations may persist: the elements may become separated or new associations may be developed (for example, Ni strongly associated with Mg in primary olivine may become strongly associated with Mn in lithiophorite within the regolith). Depending on weathering conditions and the types of regolith formed, particular elements may undergo fractionation during weathering or concentrate in particular regolith components to form regolith-related element associations. To properly interpret geochemical patterns and anomalies in the regolith, it is important to recognise and understand these associations. For example, element associations related to particular ore deposit types are very useful in multi-element geochemical exploration, but they can be complicated or obscured by
Regolith geochemistry
Table 5.4:
Some common element associations in the regolith.
Regolith component
Possible associated elements (see also Appendix 2)
Fe oxides
Fe-As, Bi, Co, Cr, Cu, Mo, Ni, P, Pb, Sb, Th, Ti, V, Zn.
Mn oxides
Mn-Ba, Pb, Co, Cu, Mo, Ni, REE, Sn, U, V, Zn.
Evaporitic carbonates/sulfates
Ca-Mg-Au, Ba, Ni, Sr, U, V, Zn.
Alunite supergroup minerals
Fe-S-Ag, As, Pb, Cu, Zn, REE.
Secondary phosphates
P- Ag, Ba, REE, Sr.
Residual kaolinite–gibbsite
Al-Si-B.
Smectites
Al-Mg-Co, Cu, Mn, Ni, Pb, Ti, Zn.
Resistate minerals Anatase
Ti- Fe, Cr, V, Nb.
Cassiterite
Sn-W, Ta, Nb.
Chromite
Cr-V, Zn.
Gold (including in quartz)
Si- Au, Ag.
Magnetite
Fe-Mn, Cr, Ti, V, Zn.
Rutile-ilmenite
Ti-Cr, Nb, Sb, Ta,V, W.
Zircon
Si-Zr, Hf, HREE, U,
Sources: Lelong et al. (1976); Roquin et al. (1990); Butt and Smith (1992); Pracejus and Bolton (1992); Cornell and Schwertmann (1996 and references therein); Taylor and Eggleton (2001); McQueen and McRae (2004).
different associations developed in the regolith. Some common element associations in the regolith and their controls are described below and in Table 5.4. Associations with Fe oxides
Iron is the fourth most abundant crustal element, and Fe oxides (principally goethite and hematite) are major end products of near-surface chemical weathering (see also Section 4.4.3; Table 12.1). Their ability to take up a wide range of cations means that these minerals exert a significant control over the fate of many trace elements in the regolith, including target and pathfinder elements used in geochemical exploration. Both hematite [a-Fe2O3] and goethite [a-FeOOH] are very stable under ambient conditions. Ferrihydrite [approximate formula 5Fe2O3.9H2O] is a widespread, but unstable, compound in surface environments – generally transforming into more stable Fe oxides. Although not as abundant as goethite and hematite, lepidocrocite [g-FeOOH] is common in non-calcareous soils and other ferruginous weathering products (Cornell and Schwertmann 1996). Maghemite [g-Fe2O3] is commonly present at and near the surface, particularly in ferruginous lag, where
it is thought to form from the other Fe oxides by heat transformation, oxidation of magnetite and, possibly, reaction under aqueous conditions (Anand and Gilkes 1987; Cornell and Schwertmann 1996). Minor and trace elements can be incorporated into Fe oxides by the following mechanisms:
s s s s
adsorption/desorption, followed by structural substitution or neo-formation of trace-element specific minerals isomorphous substitution from a precursor phase (such as ferrihydrite to goethite) co-precipitation/introduction of new trace-element specific minerals overgrowth and inclusion of other resistate minerals.
Under aqueous conditions, metal cations (and complexes) adsorb onto the surface of Fe oxides by interaction with deprotonated, surface hydroxyl groups to form mono- and binuclear inner sphere complexes (Cornell and Schwertmann 1996; Figure 5. 15). The extent of this adsorption is strongly pH dependent – increasing with increasing pH over a narrow range. Increased temperature and addition of
93
94
Regolith Science
surface hydroxyl group coordinated to underlying Fe (derived from the dissociation of H2O or structural OH)
H O Fe
Mz+ deprotonated surface of iron oxide
M
H+
cation adsorbed
lesser extent, Bi, Pb and Sb (for example, Scott 1986; McQueen and Munro 2003; Appendix 2)
O Fe
‡FeOH + Mz+ = ‡FeOMz-1 + H+ mononuclear (FeOH)2 + Mz+ = ‡(FeO)2Mz-1 + 2H+ binuclear
Figure 5.15: Summary of heavy-metal cation adsorption onto the surface of Fe oxides under aqueous conditions (after Cornell and Schwertmann, 1996).
anions (such as thiosulfate or chloride) can also enhance cation adsorption – in the latter case by formation of mixed metal/ligand surface complexes (for example, Schindler 1990; Gunton 2004). The available mineral surface area is a major control on the level of adsorption. Adsorption characteristics are different for the different iron minerals (for example, in typical near-surface environments and soils, the relative strengths of metal adsorption on goethite are Cu>Pb>Zn>Co>Ni>Mn and for hematite Pb>Cu>Zn>Co>Ni>Mn; McKenzie 1980). This gives rise to different associations of these elements with Fe in ferruginous regolith – depending on the types and relative abundances of goethite, hematite and maghemite. Under dry conditions, surface Fe atoms in the Fe oxides may be coordinatively unsaturated with respect to hydroxyl groups. This affects the adsorption capacity of the Fe oxide. Thus periodic wetting and drying may have an effect on the adsorption capabilities and behaviour of the minerals. During deposition of Fe oxides, adsorption is probably a preliminary step in the substitution of cations for Fe3+. A wide range of cations can isomorphously substitute for Fe3+ in Fe oxides including: Al, Cd, Co, Cr, Ga, Ge, Mn, Ni, Pb, Sc, Zn in goethite and Al, Cu, Cr, Ga, Ge, In, Mn, Rh, Si, Sn, Ti in hematite (Cornell and Schwertmann 1996). The main controls on substitution possibilities are ionic radius and charge characteristics of the substituting elements. In ferruginous regolith, the suite of elements associated with Fe can vary with the relative proportions of hematite and goethite. Elements commonly associated with hematite include As, Bi, Cr, Pb, Sb, Th, V and, to some extent, Cu and Zn. Elements commonly associated with goethite include As, Cd, Co, Cu, Ni, Zn and, to a
Associations with secondary Mn minerals
Secondary Mn minerals are a minor, but geochemically important, component of parts of the regolith. Manganese is the tenth most abundant element in the crust, and it generally associates with Fe, Mg, Ni and Co. It is readily oxidised under near-surface conditions and there are more than 30 Mn oxide minerals (Post 1999). Manganese oxides incorporate small metal ions as substitutions for Mn4+ and larger ions within the tunnels and interlayer regions present in their various crystal structures. This gives rise to an association of particular elements with Mn – depending on which manganese minerals are developed. Many of these minerals also have high cation exchange capacities (CEC) – further promoted by large surface areas – due to their typically fine-grained and poorly crystalline nature. Secondary Mn minerals can thus fix and concentrate dispersed cations in a similar fashion to the Fe oxides. Lithiophorite [(Al,Li)MnO2(OH)2], is a widespread Mn mineral in the regolith. Lithiophorite commonly contains high concentrations of Co, Ni, Cu and Zn. The cryptomelane–coronadite–hollandite group [(K,Pb,Ba)2-1Mn8O16] can host Pb, Ba, Cu, Co, Ni, V and Zn (Scott 1987b; Pracejus and Bolton 1992). Associations in evaporative zones
Under arid and semi-arid climatic conditions, groundwaters commonly become saturated with sulfates and carbonates as a result of water loss by evapo-transpiration. This can cause precipitation of common carbonates such as calcite, dolomite and magnesite to form calcrete, as well as sulfates including gypsum [CaSO4.2H2O], barite [BaSO4] and, rarely, celestite [SrSO4], in the upper part of the regolith. Other elements in solution (particularly gold) may co-precipitate with these evaporative minerals. This gives rise to a commonly observed association of Ca-Mg-Ba-Sr-Au (McQueen and McRae 2004). Where Mn and Ni are abundant in the regolith, these elements substitute for Ca and Mg in the carbonate structure. Carnotite [K 2 (UO2)2 (VO4)2.3H2O] can also be associated with regolith carbonates – particularly in valley calcrete deposits – giving rise to a Ca-U-V association (Deutscher et al. 1980).
Regolith geochemistry
Associations with alunite supergroup minerals
Alunite supergroup minerals develop in the regolith where sulfate precipitation occurs – commonly in the presence of Fe – for example, around weathering Febearing sulfides. These minerals have the general formula AB3 (XO4)2 (OH) 6 where: A is a large ion (such as K, Na, Ca, Pb or rare earth elements (REE)) in 12-fold coordination; B is usually Fe or Al; and the XO4 anions are usually SO4, PO4 or AsO4 (Scott 1987a). Plumbojarosite, argentojarosite and beudantite incorporate significant amounts of Pb, Ag and As, respectively. The resulting Fe–S–As–Pb–Ag association can be a feature of the strongly oxidised regolith over base metal sulfide mineralisation, particularly in arid and semi-arid regions. Associations with phosphate minerals
Phosphorous released during weathering of primary minerals – particularly apatite – commonly forms relatively insoluble phosphate minerals in the regolith. These secondary phosphates can host Pb, Ag, Ba, Sr and REE in the A sites of alunite supergroup minerals, and U in autunite/torbenite [Ca(UO2)2(PO4)2.10H2O – Cu(UO2)2(PO4)2.10H2O]. Associations with clay minerals
Particular trace elements can substitute into some clay mineral structures, such as Ni in nontronite. Furthermore, clay minerals with high cation exchange capacity (such as the smectite group; Table 5.3) may adsorb various trace elements – resulting in an association of these elements with Al. However, other clay minerals, such as kaolinite and illite (typically the most abundant clay minerals in regolith), have low cationexchange capacities (Table 5.3) and very limited ability to adsorb trace elements. These therefore do not give rise to important element associations. Associations with resistate minerals
Primary minerals that survive chemical weathering can host particular associations of minor and trace elements. Common resistate minerals include:
s s s s
zircon, which contains Zr, Hf, REE monazite, with P, Ce, La, Th, REE chromite, with Cr, Zn, V rutile and ilmenite, with Ti, V, Cr, W, Ta, Nb, Sb.
Although not a primary resistate mineral, secondary anatase is a common residual Ti-bearing mineral in the regolith. Once formed from the alteration of Ti-bearing silicates, or alteration of ilmenite and rutile, it persists through extreme weathering and may contain Nb, Zr, Cr and V (Scott and Radford 2007; Appendix 2). Native gold is a resistate mineral under some weathering conditions, and generally contains variable amounts of silver and some other elements. Supergene or secondary gold formed during weathering typically has very low contents of other elements. Cassiterite can be an important resistate mineral in weathered Sn-bearing ore deposits. Residual quartz – as well as secondary quartz formed during weathering – can contain and preserve occluded minerals and fluid inclusions. Separation and analysis of quartz from weathering profiles has been suggested as a method for detecting trace elements generally leached from other minerals in the weathering profile (Aung Pwa et al. 1999). These various resistate minerals may form residual concentrations in the regolith – including in surface lag deposits, transported sediments and various cretes, such as silcrete and ferricrete, and be used as sample media (Chapter 13). Combined element associations occur where their controlling host minerals have developed together in the regolith (such as alunite supergroup minerals developed with hematite or goethite).
5.5
GEOCHEMICAL ANOMALIES
Geochemical anomalies are geochemical features different from what is considered normal. They can be the result of: 1. unusual or uncommon processes concentrating particular elements (such as an ore-forming process, weathering and element dispersion from an unusual element concentration such as an ore body) 2. element accumulation or concentration by common processes acting over long periods (such as scavenging and concentration of certain elements by ironstones, ferrruginous regolith or Mn oxides); 3. artificial contamination of sites or samples
95
96
Regolith Science
4. analytical noise or error (such as poor precision of the analytical method, particularly for element concentrations close to the detection limit). Geochemical exploration for ore deposits is based on the search for geochemical anomalies, as are many environmental investigations. Generally, the focus has been on anomalies in element abundances rather than anomalies in element origins. Recognising such anomalies requires some method for knowing or defining the range of normal variation (commonly referred to as the background). Anomalies in element abundances are not defined by absolute values, but by comparative values. The use of particular regolith sample media is discussed in Chapter 13. 5.5.1 Identifying geochemical anomalies Traditionally, geochemical anomalies have been identified by setting threshold values, which mark the upper and lower limits of normal variation for a particular population of data. Values within the threshold values are referred to as background values, and those above or below as anomalies. In mineral exploration, interest is generally in positive anomalies – on the assumption that ore deposits and their weathering have increased element abundances above normal crustal levels. However, negative anomalies can also be important: for example, where they reflect depletion in some elements during host rock alteration accompanying ore formation. Statistical methods have been widely applied to interpret geochemical data sets and define anomalies. Such methods need to be used cautiously because of the particular characteristics of geochemical data. Geochemical data sets seldom represent a single population or distribution – the data are typically spatially dependent and at each sample site a range of different processes have influenced the element abundances measured. The data are also imprecise due to unavoidable variability in sampling methods and media and the level of analytical precision. As a result, no single universally applicable statistical test has been developed for identifying anomalies. Statistical investigation should use a range of techniques to explore the nature of geochemical data before selecting anomalous values (for example, Reimann et al. 2005).
5.5.2 Univariate statistical methods for investigating geochemical data Univariate statistical methods (that is, involving observations with only one variable) can be used to organise and extract information from a data set of values for a single element (such as Au analyses for a group of samples). A first step is to examine the frequency distribution (spread of values) of the data set using frequency histograms, frequency plots or cumulative-frequency plots. This can help to identify the type of distribution of the data, presence of multiple populations and outliers in the distribution. Box and whisker plots are another convenient way of examining the frequency distribution of a data set and for comparing the frequency distributions of multiple data sets. This type of plot shows:
s s
s s
the median (middle value or 50th percentile) a box with upper and lower hinges (or limits) defined by the 75th and 25th percentile values respectively (that is, the upper and lower quartile values) an inner fence (whisker) defined as 1.5× the length of each interquartile range towards maximum and minimum values points beyond the whiskers extending to the maximum and minimum values (Tukey 1977).
The central box contains 50% of the data (Figure 5.16). Values beyond the whiskers are considered outliers and values more than three times the interquartile range from the box hinges are referred to as far outliers. In the past, a simple way of statistically defining an anomaly in a single population of normally distributed data has been to consider values outside two standard deviations from the mean (statistical average) as anomalous (Hawkes and Webb 1962). In other words, the threshold or limit of normal variation was set at two standard deviations from the mean and the anomalous values taken as the top 2.5% of the population (positive anomalies) and the bottom 2.5% of the population (negative anomalies) (Figure 5.17). This is somewhat arbitrary – and rarely do geochemical data fit a normal distribution pattern (they are typically positively skewed towards higher values). There is often more than one population of data
Regolith geochemistry
Box and Whisker Plot
Frequency
25th
50th
Mode
Normal distribution
75th percentiles
34.1%
Frequency Histogram
34.1%
13.6%
13.6%
2.1%
-3 S.D.
Values
0.1 0.6
2.1%
-2 S.D.
2
-1 S.D.
7
16
0 Mean Median 31
50
+1 S.D.
69
84
+2 S.D.
93
+3 S.D.
Percentile 98 99.4 99.9 Rank
Values
Figure 5.16: Example of a box and whisker plot with corresponding frequency distribution. Mode
Probability
Median
Skewed distribution
Mean
Values
Figure 5.17: Examples of frequency plots for normal and skewed distributions.
tion programs, orientation surveys or case studies that compare typical background materials and sites with materials from areas of known mineralisation can be used to establish thresholds. This approach assumes that all the natural variability is covered in the orientation survey. It may miss very subtle anomalies, or anomalies and element association patterns associated with a different or unknown style of mineralisation. A common deficiency with case studies of geochemical
Frequency
present in a geochemical data set (Figure 5.18). For example, samples collected of different media, or of the same media derived from chemically different host rocks, will contain multiple populations, each with their own threshold value. Further, anomalies of interest are defined by outliers that are not part of the background population. Another common approach has been to transform geochemical data to a normal distribution pattern and then apply normal parametric tests. It has been argued that many natural data – including trace element abundances – approximate a lognormal (log10) distribution, so a simple log transformation has typically been applied (see Box 5.1). Again, this is only valid for a single population of data. Also, it has been shown that most geochemical data sets lie between normal and lognormal distributions (Reimann and Fizmoser 2000). An alternative method for handling skewed data is to set the threshold at two median absolute deviations (MAD) from the median (middle or 50th percentile value). The median will lie away from the mean and the skewness of the data (typically to a lower value than the mean) and extreme values (outliers) will have less influence. The medium absolute deviation is defined as the median value of the absolute deviations from the median of all the data (Tukey 1977). The MAD approach is best applied when the data contain less than 10% outliers. Background and anomalous values are commonly established empirically. During geochemical explora-
Multiple populations in a skewed distribution data set
Values Figure 5.18: Example of frequency histogram and frequency curve for a data set with multiple populations.
97
98
Regolith Science
patterns around known mineralisation is insufficient coverage of the surrounding background. 5.5.3 Multivariate statistical analysis of geochemical data Geochemical data sets are inherently multivariate (that is, they generally have more than one variable for each sample, such as a large number of contained and associated elements). Geochemical anomalies are commonly expressed in more than one element. This is because the source or process that has generated the anomaly commonly has an association of elements. Different ore deposit types typically have specific element associations of target and pathfinder elements. A target element is the commodity being sought and a pathfinder element is one that accompanies this element, but may be more widely dispersed or easier to detect. Element associations can be used to advantage by taking a multi-element approach to anomaly detection. Multi-element analysis can also identify other non ore-related associations, such as those generated by normal regolith processes or the result of anthropogenic contamination (see above). A range of multivariate statistical methods can be used to assess the relationships within multi-element data sets. These methods commonly include: 1. ‘scatter plots’ (bivariate plots comparing pairs of elements) 2. correlation matrices (using linear regression to test the correlation between pairs of elements); 3. cluster analysis (hierarchical grouping of elements in a data set with differing degrees of correlation of their abundance) 4. principal component factor analysis (useful for grouping elements into associations) 5. discriminant analysis (a method of optimising the distinction between two or more populations of samples). The main difficulty in assessing multi-element data is multi-dimensional visualisation. A number of techniques can be used to help – including the use of multielement ‘spider diagrams’, which plot values for a range of elements (connected by lines) in each sample. Typically these plots involve normalising the data to a reference sample. They have been widely used for example
in comparing REE data from different samples (see Rollinson 1993). Linked scatter plots, in which particular samples can be identified across a number of bivariate element plots, are another convenient way for visually identifying samples with unusual multi-element characteristics. Triangular diagrams and computer generated rotatable 3D plots can be used to visually examine data sets for three elements at a time. Two-dimensional dendograms produced from cluster analysis are a simple way of assessing multi-element associations. A number of software programs are available for univariate and multivariate statistical analysis of geochemical data and visualisation of the results. The details of these statistical techniques and their application are beyond the scope of this chapter. Where element associations are well known for particular geological materials – including ore deposits – suites of these elements can be statistically combined to detect anomalies. Simple methods that have been used include addition or multiplication of different element concentration or weighted values (for example, Beus and Grigorian 1975; Smith and Perdrix 1983). The combined anomalies can be more robust or indicative of a particular type of source than single element anomalies. For example, anomalies of associated platinum group elements can be used to discriminate between nickel anomalies generated from komatiite-hosted nickel sulfide deposits and anomalies related to weathering of nickel-bearing ultramafic rocks. A number of empirical chalcophile (CHI) indices for ranking gossans and detecting anomalies and regional chalcophile corridors using samples of ferruginous lateritic residuum have been developed (for example, Smith and Perdrix 1983; Smith et al. 1989). These use various combinations of chalcophile and related pathfinder elements and simply weight these in an additive index. For example, CHI-3 = As + 3Sb + 10Bi + 10Cd + 10In + 3Mo + 30Ag + 30Sn was found useful for locating anomalies over massive Cu–Zn sulfide deposits at Gossan Hill and Scuddles in Western Australia. 5.5.4 Exploratory geochemical data analysis In all studies of geochemical abundance data, it is valuable to make some initial assessment of the nature of the distribution of values, presence of outliers, and
Regolith geochemistry
Normal probability plots
X i = a value in the data set µ = the mean of the data set d = the standard deviation. Some examples of normal probability plots are shown in Figure 5.19.
Concentration ppm
100
(a)
80
(b)
60 40 20 -3
-2
-1
0
+1
+2
+3
-3
-2
2 - score 100
Concentration ppm
Box 5.1:
The normal-probability plot adjusts the distribution of values in a data set such that if the data have a normal distribution the raw measured values versus probability plot will be a near straight line. It is a simple test to check if a data set is normally distributed or if there are multiple populations present. For the probability axis, the raw data are transformed to standard normal values: that is, where the mean is 0 and the standard deviation is 1. These values or z-scores are calculated as Zi = (Xi –µ)/d where:
0
+1
+2
+3
+2
+3
2 - score
(c)
80
-1
(d)
60 40 20 -3
-2
-1
0
+1
2 - score
+2
+3
-3
-2
-1
0
+1
2 - score
Figure 5.19: (a) near normal distribution (b) has longer tails than would be expected for normally distributed data (c) lognormally distributed data and (d) two normally distributed populations with overlap.
element correlations. This is referred to as exploratory data analysis (EDA) and commonly uses frequency plots, correlations matrices, bivariate scatter plots and, in some cases cluster analysis or multivariate analysis to examine the data. Normal-probability plots (that is, cumulative frequency plotted on a probability scale) are particularly useful for quick and simple first pass assessment of single data sets (see Box 5.1). EDA can indicate very obvious anomalies, the presence of multiple populations of data and likely element associations. Multiple populations may be indicated by distinct groupings in the frequency distribution of a data set and in some cases these can be highlighted by careful assessment of transformed values (for example, log transformed data versus probability is commonly used; Box 5.1). The population of highest values may represent anomalies, but there may also be anomalies present in the upper values of other individual populations. EDA can also indicate deficiencies in the quality of the geochemical data. Using EDA, natural breaks in data distributions can be highlighted and examined. Threshold values can be varied to see what affect this
has on the allocation of anomalous values. The outliers defined by EDA can be investigated in more detail in terms of their other attributes. In the exploratory analysis of multivariate data, a useful approach is to take into account not only the distance of observations from the centre or centroid of the multi-dimensional data distribution, but also the shape of the data distribution (Filzmoser et al. 2005). The shape and size of multivariate data can be quantified by the covariance matrix using a measure known as the Mahalanobis distance (see Box 5.2). The region for a particular Mahalanobis distance around the mean will form an ellipse when there are only two variables and an ellipsoid when there are more than two variables. The data can then be observed in terms of their Mahalanobis distance from the centroid of the distribution and the outliers can be defined. Samples that are not anomalous in terms of univariate distributions may be revealed as anomalous associations when multivariate relationships are taken into account. Identified outliers can be investigated further using other methods and parameters. For large data
99
100
Regolith Science
Box 5.2: Detecting multivariate outliers using Mahalanobis distances Mahalanobis distances (MD) are based on the means, variances and covariances of all the variables (such as elements) in a multivariate data set (Mahalanobis 1936; Figure 5.20). For a p-dimensional multivariate sample c1….cn MDi = ((ci – t)TC–1(ci – t))½ for i = 1…..n (number of variables)
as values having a large (squared) Mahalanobis distance from the centroid (for example, outside the elipsoid for the 98% quantile, cp0.982). X2 3
where: t is the estimated multivariate location (usually the multivariate arithmetic mean; that is,. the centroid) C is the sample covariance matrix. For multivariate normally distributed data, the values of Mahalanobis distance squared (MDi2) are approximately chi-square distributed with p degrees of freedom (ci2). By setting the (squared) Mahalanobis Distance (MD2) equal to a certain constant (that is, to a certain quantile of ci2) it is possible to define ellipsoids having the same Mahalanobis distance from the centroid. Multivariate outliers can then be identified
sets, this is a very useful approach to reducing the number of samples requiring detailed investigation in terms of anomaly-forming processes. 5.5.5 Space, time and source aspects of geochemical anomalies Groups of geochemical values that are statistically defined should always be examined in terms of their spatial relationships (3D geographic and depth positions) and linked to their regolith and geological context. This can often clear up uncertainties about multiple populations and appropriate thresholds for defining anomalies. Ideally, geochemical data should also be linked to their mineralogical controls. Spatial variance in anomalies and patterns of element associations are the basic method of vectoring to the source of an anomaly. A number of methods are widely used for visualising the 2D spatial distribution of element values and derived data including: contouring, pixel plots and bubble plots. Contour maps are easy to visualise, but may not be the best method of presentation as commonly the geochemical data do not possess the
2
98
_ X2
tile tile an qu 5 7
an
0.
1
qu
0.
0 -1 Anomalies? -2
Ellipse of constant Mahalanobis distance for the 0.98 quantile of the chi-squared distribution
-3 -3
-2
-1
_0 X1
+1
+2
+3 X 1
Figure 5.20: Mahalanobis distances
characteristics required for proper contouring. The data may lack linear dependence – even over short distances – and the contouring may also misrepresent the precision of the sampling and analytical methods. Pixel plots, which are colour coded for different values, or bubble plots, with bubble diameters representative of element concentrations, may be better options. Element dispersion in the fourth dimension (time) is an important aspect of anomaly formation. Environmental conditions may have changed significantly through time at a particular site – affecting the type and degree of element dispersion. If the relative expression of a geochemical anomaly is markedly different under different weathering regimes, setting the appropriate anomaly threshold and recognising the appropriate multi-element association will depend on knowing when the anomaly formed and under what conditions. Protracted weathering or weathering through a range of contrasting chemical regimes commonly results in strong chemical leaching and marked depletion of most elements, so that any geochemical anomalies are very subtle. A multi-element approach
Regolith geochemistry
may improve detection of such anomalies. Analytical or sampling techniques that improve anomaly-tobackground contrast or reduce background variation (noise) may also be required to detect more subtle anomalies. A geochemical anomaly may relate to an anomalous source for an element, or suite of elements, but lie within the level of background variation in terms of element abundance. These types of anomalies are difficult to detect, and many have probably not yet been found. Methods for determining element sources include isotopic analysis, combined geochemical and mineralogical analysis to target particular host minerals and multi-element analysis to detect associations of elements related to a particular source. Isotopic analysis of low-level lead in ironstones/gossans, rock chips and soils has been used as a method to detect anomalies related to ore deposit types with particular Pb isotopic ratios (Gulson 1986). Isotopic analysis of groundwater – particularly using S, Sr and Pb isotopes – is a promising technique for locating ore deposits and other regolith element concentrations that have interacted with groundwater (for example, Andrew et al. 1998; de Caritat et al. 2005).
5.6 REFERENCES Anand RR and Gilkes RJ (1987). The association of maghemite and corundum in Darling Ranges laterites, western Australia. Australian Journal of Soil Research 35, 303–311. Anand RR, Cornelius M and Phang C (2007). Use of vegetation and soil in mineral exploration in areas of transported overburden, Yilgarn Craton, Western Australia: a contribution towards understanding metal transportation processes. Geochemistry, Exploration, Environment, Analysis 7, 267–288. Andrew AS, Carr GR, Giblin AM and Whitford DJ (1998). Isotope hydrogeochemistry in exploration for buried and blind mineralisation. Geological Society of Australia Special Publication 20, 222–225. Appelo CAJ and Postma D (2005). Geochemistry, Groundwater and Pollution, 2nd edn. Balkema, Leiden. Årström M and Deng H (2003). Assessment of the mobility of trace elements in acidic soils using soil
and stream geochemical data. Geochemistry, Exploration, Environment, Analysis 3, 197–203. Aung Pwa, McQueen KG, Scott KM and van Moort JC (1999). Regolith geochemical exploration using acid insoluble residues as a saple medium for gold and base metal deposits in the Cobar region, N.S.W., Australia. Journal of Geochemical Exploration 67, 15–31. Baes CF and Mesmer RE (1976). Hydrolysis of Cations. John Wiley and Sons, New York. Beus AA and Grigorian SV (1975). Geochemical Exploration Methods for Mineral Deposits. Applied Publishing Limited, Wilmette, Illinois. Butt CRM (2001). Dispersion of gold and associated elements in the lateritic regolith, Mystery Zone, Mt Percy, Kalgoorlie, Western Australia. Geochemistry, Exploration, Environment, Analysis 1, 291–306. Butt CRM and Smith RE (1992). Characterisation of the weathering profile. In Regolith Exploration Geochemistry in Tropical and Subtropical Terrains. (Eds CRM Butt and H Zeegers) pp. 299–304. Elsevier, Amsterdam. Callender E (2005). Heavy metals in the environment – historic trends. In Environmental Geochemistry. Volume 9. (Ed. B Sherwood Lollar) pp. 67–105. Treatise on Geochemistry Series (Eds HD Holland and KK Turekian). Elsevier-Pergamon, Oxford. Cameron EM and Hattori KH (2003). Mobility of palladium in the surface environment: data from a regional lake sediment survey in northwestern Ontario. Geochemistry, Exploration, Environment, Analysis 3, 299–311. Cameron EM, Leybourne MI and Kelley DL (2002). Exploring for deeply-covered mineral deposits; formation of geochemical anomalies in northern Chile by earthquake induced surface flooding of mineralised groundwaters. Geology 30, 1007–1010. Colin F, Vieillard P and Abrosi JP (1993). Quantitative approach to physical and chemical gold mobility in equatorial rainforest lateritic environment. Earth and Planetary Science Letters 114, 269–285. Cornell RM and Schwertmann U (1996). The Iron Oxides. VCH Publishers, New York. Davis JA and Kent DB (1990). Surface complexation modelling in aqueous geochemistry In Mineral Water Interface Geochemistry. Reviews in Mineralogy
101
102
Regolith Science
23. (Eds MF Hochella and AF White), pp. 177–260. Mineralogical Society of America, Washington DC. de Caritat P, Kirste D, Carr G and McCulloch M (2005). Groundwater in the Broken Hill region, Australia: recognising interaction with bedrock and mineralisation using S, Sr and Pb isotopes. Applied Geochemistry 20, 767–787. Deutscher RL, Mann AW and Butt CRM (1980). Model for calcrete uranium mineralization. In Conceptual Models in Exploration Geochemistry. (Eds CRM Butt and RE Smith). Journal of Geochemical Exploration 12, 158–161. Domenico PA and Schwartz FW (1998). Physical and Chemical Hydrogeology. John Wiley and Sons, New York. Filzmoser P, Garrett RG and Reimann C (2005). Multivariate outlier detection in exploration geochemistry. Computers and Geosciences 31, 579–587. Goldenberg LC, Hutcheon I and Wardlaw N (1989). Experiments on transport of hydrophobic particles and gas bubbles in porous media. Transport in Porous Media 4, 129–145. Goldich SS (1938). A study in rock weathering. Journal of Geology 46, 17–58. Goldschmidt VM (1954). Geochemistry. Oxford University Press, Oxford. Govett GJS (1976). Detection of deeply-buried and blind deposits by measurement of H+ and conductivity of closely spaced surface soil samples. Journal of Geochemical Exploration 6, 359–382. Govett GJS and Atherden PR (1987). Electrogeochemical patterns in surface soils – detection of blind mineralisation beneath exotic cover, Thalanga, Queensland, Australia. Journal of Geochemical Exploration 28, 201–218. Gray DJ (2001). Hydrogeochemistry in the Yilgarn Craton. Geochemistry, Exploration, Environment, Analysis 1, 253–264. Guilbert JM and Park CF (1986). The Geology of Ore Deposits. Freeman, New York. Gulson BL (1986). Lead Isotopes in Mineral Exploration. Elsevier, Amsterdam. Gunton C (2004). Adsorption: Should we care? In Regolith 2004. (Ed. IC Roach) pp. 112–116. CRC LEME, Perth.
Hale M (2000). Genetic models of remote dispersion patterns. In Geochemical Remote Sensing of the Subsurface. (Ed. M Hale) pp. 3–16. Elsevier, Amsterdam. Hall A (1987). Igneous Petrology. Longman, Harlow, Essex. Hawkes HE and Webb JS (1962). Geochemistry in Mineral Exploration. Harper and Row, New York. Jaynes WF and Bigham JM (1986). Multiple cationexchange capacity measurements on standard clays using a commercial mechanical extractor. Clays and Clay Minerals 34, 93–98. Klusman RW (1993). Soil Gas and Related Methods for Natural Resource Exploration. John Wiley and Sons, Chichester. Lelong F, Tardy Y, Grandin G, Trescases JJ and Boulange B (1976). Pedogenesis, chemical weathering and processes of formation of some supergene ore deposits. In Supergene and Surficial Ore Deposits: Textures and Fabrics. Handbook of Stratabound and Stratiform Ore Deposits 3. (Ed. KH Wolf) pp. 93–173. Elsevier, Amsterdam. Leverett P, McKinnon AR and Williams PA (2004). A supergene exploration model for Cobar style deposits. In Proceedings Exploration Field Workshop Cobar Region 2004. (Eds K G McQueen and K M Scott) pp. 46–50. CRC LEME, Perth. Levinson AA (1980). Introduction to Exploration Geochemistry. 2nd Edn. Applied Publishing Ltd, Wilmette. Lock NP (1985). Kimberlite exploration in the Kalahari region of southern Botswana with emphasis on the Jwaneng kimberlite province. In Prospecting in Areas of Desert Terrain. pp. 183–190. Institute of Mining and Metallurgy, London. Mahalanobis PC (1936). On the generalised distance in statistics. Proceedings of the National Institute of Science of India 12, 49–55. Mann AW, Birrell RD, Fedikow MAF and de Souza HAF (2005). Vertical ionic migration: mechanisms, soil anomalies, and sampling depth for mineral exploration. Geochemistry, Exploration, Environment, Analysis 5, 201–210. McKenzie RM (1980). The adsorption of lead and other heavy metals on oxides of manganese and iron. Australian Journal of Soil Research 18, 61–73.
Regolith geochemistry
McQueen KG and McRae A (2004). New ways to explore through the regolith in western New South Wales. In Pacrim 2004 Congress, Proceedings, pp. 232–238. The Australasian Institute of Mining and Metallurgy, Melbourne. McQueen KG and Munro DC (2003). Weatheringcontrolled fractionation of ore and pathfinder elements at Cobar, NSW. In Advances in Regolith. (Ed. IC Roach) pp. 296–300. CRC LEME, Perth. McQueen KG, Munro DC, Gray D and Le Gleuher M (2004). Weathering-controlled fractionation of ore and pathfinder elements Part II: The lag story unfolds. In Regolith 2004. (Ed. IC Roach) pp. 241– 246. CRC LEME, Perth. Nealson KH and Stahl DA (1997). Microorganisms and biogeochemical cycles: What can we learn from layered microbial communities? In Geomicrobiology: Interactions between Microbes and Minerals. Reviews in Mineralogy 35. (Eds JF Banfield and KH Nealson) pp. 5–34. Mineralogical Society of America, Washington DC. NSW Department of Primary Industries, Agriculture (2007). Web site.
. Perel’man AI (1977). Geochemistry of Elements in the Supergene Zone. Israel Program for Scientific Translations, Keter Publishing, Jerusalem, John Wiley and Sons, New York. Plant JA, Kinniburge DG, Smedley PL, Fordyce FM and Klinck BA (2005). In Environmental Geochemistry. Volume 9. (Ed. B Sherwood Lollar) pp. 17–66. Treatise on Geochemistry Series (Eds HD Holland and KK Turekian). Elsevier-Pergamon, Oxford. Post JE (1999). Manganese oxide minerals: crystal structures and economic and environmental significance. Proceedings of the National Academy of Science USA 96, 3447–3454. Pracejus B and Bolton BR (1992). Geochemistry of supergene manganese oxide deposits, Groote Eylandt, Australia. Economic Geology 87, 1310–1335. Reimann C and Fizmoser P (2000). Normal and lognormal data distribution in geochemistry: death of a myth. Consequences for the statistical treatment of geochemical and environmental data. Environmental Geology 39, 1001–1014.
Reimann C, Filzmoser P and Garrett RG (2005). Background and threshold: critical comparison of methods of determination. Science of the Total Environment 346, 1–16. Rollinson H (1993). Using Geochemical Data: Evaluation, Presentation, Interpretation. Longman, Essex. Salomons W and Förstner U (1984). Metals in the Hydrocycle. Springer Verlag, Berlin. Schindler PW (1990). Co-adsorption of metal ions and organic ligands: Formation of ternary surface complexes. In Mineral-Water Interface Geochemistry. Reviews in Mineralogy 23. (Eds MF Hochella and AE White) pp. 281–309. Mineralogical Society of America, Washington DC. Scott KM (1986). Elemental partitioning into Mnand Fe-oxides derived from dolomitic shale-hosted Pb-Zn deposits, northwest Queensland, Australia. Chemical Geology 57, 395–414. Scott KM (1987a). Solid solution in, and classification of, gossan-derived members of the alunite-jarosite family, northwest Queensland, Australia. American Mineralogist 72, 178–187. Scott KM (1987b). Significance of a lithiophorite interface between cryptomelane and florencite. American Mineralogist 72, 429–432. Scott KM and Radford NW (2007). Rutile compositions at the Big Bell Au deposit as a guide for exploration. Geochemistry, Exploration, Environment, Analysis 7, 353–361. Scott KM, Ashley PM and Lawrie DC (2001). The geochemistry, mineralogy and maturity of gossans derived from volcanogenic Zn-Pb-Cu deposits of the eastern Lachlan Fold Belt, NSW, Australia. Journal of Geochemical Exploration 72, 169–191. Simmons CT, Fenstemaker TR and Sharp JM (2001). Variable density groundwater flow and solute transport in heterogeneous porous media: approaches, resolutions and future challenges. Journal of Contaminant Hydrology 52, 245–275. Smith RE and Perdrix JL (1983). Pisolitic laterite geochemistry in the Golden Grove massive sulphide district, Western Australia. Journal of Geochemical Exploration 18, 131–164. Smith RE, Birrell RD and Brigden JF (1989). The implications to exploration of chalcophile corridors in the Archaean Yilgarn Block, Western
103
104
Regolith Science
Australia, as revealed by laterite geochemistry. Journal of Geochemical Exploration 32, 169–184. Southam G and Saunders JA (2005). The geomicrobiology of ore deposits. Economic Geology 100, 1067–1084. Southam G, Ferris FG and Beveridge TJ (1995). Mineralized bacterial biofilms in sulfide tailings and in acid mine drainage systems. In Microbial Biofilms. (Eds HM Lapping-Scott and HW Costerton) pp. 148–170. Cambridge University Press, Cambridge. Stumm W and Morgan JJ (1996). Aquatic Chemistry: an Introduction Emphasising Chemical Equilibria in Natural Waters, 3rd edn. Wiley Interscience, New York. Taylor G and Eggleton RA (2001). Regolith Geology and Geomorphology. John Wiley and Sons, Chichester, UK. Thornber MR (1992). The geochemical mobility and transport of elements in the weathering environment. In Regolith Exploration Geochemistry in Tropical and Subtropical Terrains. (Eds CRM Butt and H Zeegers) pp. 79–96. Elsevier, Amsterdam.
Thornber MR and Taylor GF (1992). In Regolith Exploration Geochemistry in Tropical and Subtropical Terrains. (Eds CRM Butt and H Zeegers) pp. 119–138. Elsevier, Amsterdam. Trescases JJ (1979). Remplacement progressif des silicates par less hydroxydes de fer et de nickel dans les profils d’altération tropicale des roches ultrabasiques; accumulation résiduelle et épigénie. Sciences Geologiques Bulletin 32, 181–188. Trescases JJ (1992). Chemical weathering. In Regolith Exploration Geochemistry in Tropical and Subtropical Terrains. (Eds CRM Butt and H Zeegers) pp. 25–40. Elsevier, Amsterdam. Tukey JW (1977). Exploratory Data Analysis. AddisonWesley, Reading. Velde B (1995). Composition and mineralogy of clay minerals. In Origin and Mineralogy of Clays. (Ed. B Velde) pp. 8–42. Springer-Verlag, New York. Wan J, Veerapaneni S, Gadelle F and Tokunaga TK (2001). Generation of stable microbubbles and their transport through porous media. Water Resources Research 37, 1173–1182. Williams PA (1990). Oxide Zone Geochemistry. Ellis Horwood, Chichester, UK.
6
Rock weathering and structure of the regolith Kenneth G McQueen and Keith M Scott
6.1 INTRODUCTION The previous two chapters have considered the mineralogical and chemical changes that occur during weathering. As rocks weather chemically, and their mineral constituents change to new, more stable assemblages, the contained elements are preserved in resistate minerals, partly redistributed into new minerals or taken into solution – in some cases to be incorporated in other parts of the weathering profile. Such elemental changes within the weathering profile may be large and, in some cases, result in economic mineral deposits (for example, the concentration of Al in bauxite deposits; Chapter 1). This chapter considers the mineralogical and geochemical changes that accompany progressive rock weathering and discusses the typical weathering profiles for common rock compositions.
6.2 STRUCTURE OF THE REGOLITH The structure of the regolith at any particular site depends on the extent to which chemical weathering has transformed the bedrock composition, as well as the degree of physical and chemical addition and removal of materials. Well-developed profiles show a
vertical zonation, which may include from depth to surface:
s s s s s
a zone of partially weathered bedrock that retains the primary rock fabric a clay-rich or sandy plasmic/arenose zone in which the primary rock fabric has been destroyed a ferruginous mottled zone; a ferruginous, bauxitic or siliceous duricrust/residuum a soil layer a surface lag of chemically and physically resistant materials (Figure 6.1).
The zone in which the primary rock fabric is preserved is referred to as the saprolith. The zone in which the parent fabric has been destroyed, new fabrics formed or soil developed is termed the pedolith. Weathering occurs throughout the profile down to the weathering front, which is defined as the boundary between fresh rock and saprolith (that is, rock that shows some sign of chemical weathering). Depending on bedrock type and landscape setting, various parts of this mature zonation may be absent, eroded or buried. Across a landscape (or paleolandscape), there is generally significant lateral variation in the regolith and its chemical structure, and these variations may be down to a
106
Lateritic residuum or ferricrete
Regolith Science
Lag Soil Loose
Lateritic gravels
Cemented
Lateritic duricrust Mottled zone Cementation front
Regolith
Pedolith
Plasmic (clay) or arenose (sandy) zone
Pedoplasmation front
Saprolite
Saprolith
Saprock Weathering front Bedrock
Figure 6.1: Idealised regolith profile (after Eggleton 2001).
scale of metres or tens of metres. Typically weathering will penetrate preferentially and deeper along fractured or more permeable zones. Less-weathered blocks (lithorelicts) may be preserved well above the weathering front and even to the surface (for example, granite tors; Figure 3.1). Importantly, the most recently weathered material is that closest to fresh rock. Generally, this means that in a profile the youngest weathering zone is at the base and the oldest at the top. Within the saprolith, the lowermost saprock layer is the least chemically altered zone, with less than 20% of weatherable minerals altered. It will generally retain evidence of the most recent weathering. Under humid, oxidising and acid conditions, most sulfides and
carbonates are the least stable minerals and hence the first to breakdown. Consequently S will be depleted, or present as sulfate either in solution or as precipitates, and elements hosted by the sulfides (As, Cd, Co, Cu, Mo, Ni, Zn) pass into solution or are incorporated into neo-formed (secondary) minerals (Section 5.3.8; Appendix 2) . Elements present within carbonates (Ca, Fe, Mg, Mn and Sr) will also be released and may be depleted. Initial alteration of ferromagnesian minerals will release some Fe and Mg and contained trace elements (Figure 5.3). Changes in chemical conditions at the weathering front may also cause some dissolved elements – derived from more extensive weathering higher in the profile – to be precipitated or substituted into existing minerals (for example, supergene Au enrichment; Figure 6.2; Figure 5.14) Chemical modification is more extensive in the saprolite, where more than 20% of weatherable minerals are altered, although the primary bedrock fabric is maintained. Progressive destruction of ferromagnesian minerals and feldspars results in depletion of Mg, Ca, Na, K, some depletion in Si and retention of Al and Si within the main weathering products of kaolinite, other clay minerals and secondary silica. Hosted trace elements that are released include Co, Cr, Cu, Mn, Ni, Ba, Cs, Rb and Sr. Muscovite will generally persist through the saprolite and retain some of the K, Rb and Cs of the parent rock composition. Under reducing conditions (such as below the water table), the Fe2+ and Mn2+ released from weathered ferromagnesian minerals are mobile, but will precipitate as insoluble Fe and Mn oxides where conditions become more oxidising. Trace elements (particularly As, Co, Cr, Cu, Ni and Zn) that are adsorbed or incorporated in these stable oxides will be partially retained in the profile, or even relatively enriched. More extensively altered parts of the saprolite (typically the leached upper saprolite) are marked by alteration of all but the most chemically resistant primary minerals – as well as progressive destruction of the less-stable secondary minerals (such as smectites). This leads to further release and potential leaching of Mg, Fe, Co, Cu, Cr, Ni and Zn and relative enrichment in Al and Si. The REE, which have generally been considered immobile, can show significant depletion and enrichment within the saprolite depending on their primary host (such as
Rock weathering and structure of the regolith
Qtz Goe Goe Kaol Qtz
Goe Hem Qtz (Qtz)
Goe Kaol
Goe Hem Goe
Smec
Smec Serp
Au-depleted Zone
Qtz Kaol
Smec Serp Talc Goe
Smec Chl Kaol (Qtz)
Qtz Feld Mica
PORPHYRY
Talc
Amph Pyrox Plag
Chl (Qtz)
BASALT
Serp Amph Talc Pyrox
PERIDOTITE
Pedogenic calcrete
Saprolite
Soil
Saprock
Lateritic ferruginous zone
Fresh rock
Clays with ferruginous mottles
Zone of gold enrichment
Figure 6.2: Supergene enrichment in Au caused by concentration of Au derived from higher in the weathering profile (after Butt 1989).
highly weatherable pyroxenes and apatite versus resistate zircon) and the weathering conditions. The REE – particularly Ce – can be mobile and variably fractionated under intense weathering conditions (see Section 6.8; Duddy 1980). Major primary mineral alteration, chemical leaching and secondary mineral growth can eventually destroy the primary rock fabric to produce the pedolith (Figure 6.1). This typically contains a clay- or quartz-rich plasmic/arenose zone (strongly enriched in Al and Si relative to the parent bedrock) and a mottled zone (in which darker Fe oxides are segregated from the more pallid clay minerals). Under some conditions, weathering in the plasmic zone can result in the alteration of kaolinite to gibbsite and leaching of Si. Most primary quartz and resistate accessory minerals are retained – resulting in residual enrichment in Si and elements such as B, Cr, Hf, Nb,
Rb, REE, Th, Ti, V, W and Zr. The mottled zone in the upper part of the pedolith with Fe- and clay-rich zones is enriched in Fe relative to the zone below. Marked accumulation of the Fe oxides – particularly over mafic and ultramafic rocks – can produce a very ferruginous zone, which is typically composed of ferricrete (if cemented by Fe oxides) or ferruginous residuum (if less consolidated). This very ferruginous zone is enriched in elements associated with hematite and goethite, as well as those present in highly resistate accessory minerals. Accumulation of Al oxides, such as gibbsite and boehmite, can form a bauxitic zone and precipitation of dissolved silica can produce an almost pure SiO2 accumulation of silcrete. These chemically stable and, in most cases, physically resistant materials can harden to form a duricrust and protect this zone of the regolith from erosion and further major chemical alteration.
107
Regolith Science
Organic
character from thin, coarse-grained lithosols – that have only one or two poorly differentiated horizons – in some deserts, to thick, organic-, silt- and/or clayrich soils differentiated into several horizons, in more humid regions. Common soil horizons (Figure 6.3) include:
O1 Litter O2 Peaty horizon
Solum (pedogenic processes)
A Humic horizon E Pale eluvial horizon
s s
B Illuvial horizon
Unconsolidated parent material
108
s s s
C Mineral horizon
s Saprolite
Bedrock
D Weathered rock
R Fresh rock KSf010-07
s
Figure 6.3: An idealised soil profile (after Butt et al. 2005).
A soil profile will generally develop close to the surface in the zone of greatest biological activity and accumulation of organic detritus. Decay of organic matter and the activity of microorganisms and vegetation result in concentration of humic acids, organic complexes, carbonates and nitrates. The soil chemistry will reflect climate and water content as well as the composition of the underlying regolith/ bedrock and any colluvial, alluvial and aeolian additions. See Section 13.2.2 for more detail on soils as a sampling medium. A mature soil in equilibrium with the local environment is typically differentiated into a number of horizons. These are soil layers that are approximately parallel to the land surface and differ from adjacent, genetically related, layers in their physical, chemical and/or biological properties, or in characteristics such as colour, pH, structure, fabric, texture, consistency, and types and number of organisms. Soils may vary in
s
Litter (or O1): organic matter on the ground surface O (or O2): fibrous (peaty) or massive organic matter A: near-surface mineral horizon containing humified organic matter E (or A2): pale, commonly sandy, eluvial horizon with little organic matter. Iron and Mn oxides and clays leached or translocated to lower horizons B: illuvial horizons enriched in clay, and/or Fe and Mn oxides and/or organic matter derived from overlying horizons C: mineral horizon from which the overlying horizons are presumed to have been derived. It is only slightly affected by pedogenic processes so that remnant geological structures or fabric may be retained D: layers below the C horizon unaffected by the pedogenic processes that formed overlying horizons, such as previously formed saprolite or transported overburden R: continuous fresh rock.
The A, E and B horizons are referred to as the solum. This terminology is ideally suited to well-differentiated soils, such as podzols, but less readily applied to arid-zone soils that have weak horizon differentiation. In soils in semi-arid Australia, the B horizon may be differentiated from the A horizon by a contrast in texture and/or being more sodic or alkaline, rather than the characteristics noted above. For more detail on soil types see CSIRO (1983) for Australia and FAOUNESCO (1988) for a more global view. A surface lag may form above the soil if the finer or less dense material is preferentially removed by sheetwash or wind erosion. As quartz and hematite are the most chemically and physically stable minerals under most surface conditions, this lag is relatively enriched in Fe and Si. The presence of stable accessory minerals
Rock weathering and structure of the regolith
in the lag or underlying duricrust can also result in relative enrichment in Ba, Cr, Mn, Nb, Ti, W and Zr. Other minor and trace elements can be accumulated and concentrated in the Fe and Mn oxides that persist in the lag (such as As, Bi, Co, Cr, Ga, Ni, Pb, Sc, Th and V). The water table is an important chemical and physical boundary within the weathering profile – marking the interface between the zone of water saturation and the overlying zone of partial water content. In weathering profiles that are in a steady state, the water table will generally occur in the saprock or lower saprolite zone. However, its position can change seasonally, or over longer time periods, with climate change and landscape evolution. Under wetter conditions the water table can rise up through the weathering profile. As conditions become drier, the water table may progressively fall. The depth to the water table will also depend on the geomorphic setting and local hydrologic factors, including vegetation pumping. Below the water table, conditions are typically reducing (low Eh), and water movement is generally slow through the water-filled fractures and pore spaces (Figure 6.4). Permeability and hydraulic gradient control water movement. Above the water table, gases (including oxygen from the atmosphere) have greater access to the regolith via voids and fractures and conditions are typically more oxidising (higher Eh). Water generally moves rapidly through this zone transporting dissolved and suspended components and leading to strong leaching with further void formation. As well as moving downwards and laterally, water and dissolved components can also move upwards through this zone by capillary action – particularly where evaporation exceeds rainfall. The change in Eh near the water table is commonly sufficient to produce a redox zone marked by the precipitation of insoluble Fe and Mn oxides. The pH may also be reduced in this zone due to the release of H+ when divalent cations dissolve: M 2 + + 2H 2 O = M (OH) 2 + 2H +
(Eqn 6.1)
Acid production is even greater if the cation is Fe2+, which may be oxidised and hydrolysed (sometimes referred to as ferrolysis):
O2
O2 Soil
Voids abundant Water moves laterally and vertically (capillary)
Oxidising Eh higher
Saprolite
pH lowered by Fe 2+ oxidation/hydrolysis Redox zone
Reducing Eh lower
Few voids Slow water movement
Water table Fe/Mn oxides
Saprock
Weathering front
Fresh rock
KSf001-08
Figure 6.4: Eh and pH conditions about the water table.
Fe 2 + = Fe 3 + + e-
(Eqn 6.2)
and Fe 3 + + 3H 2 O = Fe(OH) 3 + 3H +
(Eqn 6.3)
Thus acidity is generated during chemical weathering even in the absence of sulfides (see Chapter 5.3.1 Eqns 5.22–5.25 and Chapter 10). Groundwater compositions can also have a significant affect on the chemical conditions above and below the water table, including on a regional scale. For example, in the southern Yilgarn Craton of Western Australia, conditions in the upper 5–30 m of the regolith are commonly acid and oxidising, whereas in the northern Yilgarn Craton they are commonly neutral and weakly to moderately oxidising (Gray 2001).
6.3 FIELD EXAMPLES OF WEATHERING PROFILES ON COMMON ROCK TYPES Particular zones within a regolith profile are commonly recognised by their colour, induration, textural
109
110
Regolith Science
features and mineralogy (such as recognition of mottling). Some of these features can be difficult to determine in pulverised and homogenised drill chips – a common means of observing and sampling weathering profiles during mineral exploration. Field descriptions of zones within profiles based on information obtainable from drill chips are therefore simplified from those in Figure 6.1. Thus, in the descriptions of weathering in common rock types (below), the weathered materials are divided into lower saprolite (which also includes saprock) and leached upper saprolite (which may include some pedolith material), respectively. Examples are for complete, well-developed and largely intact profiles. Such profiles are more commonly preserved in areas with a long history of weathering, tectonic stability and low relief or where they are protected from erosion by well-developed duricrusts or other deposits (such as lava flows). As well as climatic conditions and geomorphic setting, the degree and depth of weathering are influenced by rock type, structure and the presence of sulfide mineralisation. Studies in the Yilgarn Craton of Western Australia indicate that the extent and depth of weathering commonly follow the succession: granite and mafic rocks
amounts of other clays, such as halloysite and smectite, may also form. Further weathering – particularly under humid climatic conditions – can lead to breakdown of kaolinite and removal of silica in solution to produce Al hydroxides (gibbsite and boehmite): that is, a bauxite zone above the kaolinite-rich saprolite. Some elements that are retained in resistate minerals – such as Zr and Ti, in zircon and rutile – are residually concentrated in the upper part of the profile (Butt 1985; Figure 6.5). Soluble Na, K and Ca are strongly leached from the upper saprolite. The K present in muscovite is, however, retained. In the example of granitic weathering from the Darling Ranges, WA (Figure 6.5), the formation of a surficial siliceous capping (duricrust) has diluted all other components, but this horizon, and the thin soil formed above it, contain the same minerals as in the leached upper saprolite – though not in the same proportions. Weathering profiles in felsic volcanic rocks show a similar pattern of mineral development to weathering in granitoids, but with stronger ferruginous mottling where they have a higher original Fe content (Anand and Paine 2002). 6.3.2 Weathering in mafic rocks Most mafic rocks are igneous or metamorphosed igneous rocks and generally contain dominant plagioclase, ferromagnesian minerals (such as olivine, pyroxenes, amphiboles, biotite and chlorite) and variable amounts of other minerals, including quartz. Muscovite, carbonates and sulfides (mainly pyrite or pyrrhotite) may be present in alteration and mineralised zones. With the commencement of weathering, the carbonate minerals (dolomite and calcite) rapidly dissolve, pyrite and pyrrhotite weather to form goethite, and the ferromagnesian minerals alter to smectite and kaolinite. This results in loss of Ca, but retention of Al, Fe and Mg (the latter two may even be slightly residually enriched) at the base of the lower saprolite. With more intense/longer weathering, plagioclase and chlorite also break down to form smectites and more kaolinite and goethite, with Mg being lost from higher in the lower saprolite. Further weathering results in the complete destruction of plagioclase and chlorite and the formation of more clay minerals (kaolinite and smectites) and goethite/hematite, with
<0.1 <0.1 Na2O % 3.9 4.3
2.6
0.9
0.1
10 Fe2O3 % 12 2 3.0 2.9 3.0 3.1
0.4
0.4
0.9 TiO2 % 0.7 .7 0.6
20 Al2O3 % 28 8 35
2.9
Grey
0.3
Fresh Rock
0.3
Grey
16
Grey
Lower Saprolite
15
230
5
Grey
17
300 300
White
150
Leached Upper Saprolite
17
Brown
400 Zr ppm
Duricrust
260
Yellow-brown
160
Magh
Hem
Gibbsite
Kaol
Goe
Musc
Biotite
Plag
Kspar
Soil
5
0.5
Colour
2.0
Qtz
Thickness (m)
Rock weathering and structure of the regolith
KSf014-07
Key to Figures 6.5 - 6.9 Calcite Calc Chlorite/vermiculite Chl/V Dolomite Dol Goethite Goe Gyp/Ba Gypsum/barite
Hem Kaol Kspar Magh Musc
Hematite Kaolinite K-feldspar Maghemite Muscovite
Parag Plag Qtz Smec
Paragonite Plagioclase Quartz Smectite
Figure 6.5: A generalised profile through weathered granite, Darling Ranges, WA (after Gilkes et al. 1973; Anand 1984).
soluble elements – particularly K and Mg – being lost from the leached upper saprolite. Aluminium and Fe (plus insoluble trace elements, such as Cr) become enriched in this zone. The formation of soil continues the process of enrichment in clay minerals and Fe oxides, although the formation of near surface calcrete/sulfates in the soil may dilute other components.
Quartz and muscovite are retained through the weathering cycle. Figure 6.6 depicts a generalised weathering profile on mafic rocks from Mt Magnet (WA), showing the mineralogical variation. Weathering of Cenozoic basalts (containing olivine, pyroxene and glass) under more temperate conditions in south-eastern Australia, has produced
111
Fe2O3% 16 18 16 10
19
MgO % 1 0.3 6 5
1.5
K2O % 0.5 0.2 0.2
Grey
0.5
0.4
Al2O3 % 11 15
Brown to brown-grey
13
14
Cr ppm 600 450
Gyp/Ba Pyrite
Smec
Kaol
Chl/V
Brown
13
Fresh Rock
Buff to red-brown
350
Lower Saprolite
Pink-brown
250
Leached Upper Saprolite
Parag
Musc
Goe
Hem
Dol
Calc
Plag
Qtz Soil and calcrete
Colour
300
10
Thickapprox 3.0 ness (m)
Regolith Science
25
112
KSf015-07
Figure 6.6: A generalised profile through weathered mafic rocks, Mt Magnet, WA (after Scott and Martinez 1990).
similar suites of minerals and elemental depletions/ enrichments in saprolite and soil (Loughnan 1969; Moore 1996). The importance of water in weathering is illustrated by comparison with the weathering of basalts on the Moon where only physical, rather than chemical, weathering occurs (Section 14.2). 6.3.3 Weathering in ultramafic rocks Ultramafic rocks consist mainly of olivine and pyroxenes, but they are commonly serpentinised or altered by metamorphism to amphibole–chlorite or talc-bearing assemblages. They are rich in Mg and poor in Al relative to other rock types. Chemical weathering of olivine- and serpentine-rich rocks produces smectitic clays, such as nontronite and saponite,
in the saprolith. These are altered to kaolinite in the upper saprolite, with loss of Mg. Some of this Mg may be precipitated as magnesite nodules in the lower saprolite. Iron – originally present in olivine and disseminated magnetite – forms goethite in the saprolite, which is dehydrated to hematite near the top of the weathering profile. Excess Si released during the weathering commonly forms deposits of opaline silica within the profile. Under free-draining conditions in high-relief terrains, Ni released from the weathering of olivine and serpentine can substitute for Mg in serpentine close to the weathering front to produce garnierite. In low-relief and poor-draining environments Ni may be taken up in smectitic clays or goethite in the lower saprolite.
1.6 Fe2O3% 27 12
13
1.1 MgO % 3.8 13 16
0.5 K2O % <0.1 0.4
7.8 Al2O3 %
<0.1
18
11
<0.1
18
14
<0.1
20
12
Grey
6.9
Fresh Rock
9.1
Khaki to brown
7.5
Khaki
2200
Saprolite
2800
60-70
Yellow to khaki
<0.1
Yellow to light brown
8.6
Brownyellow
10
Ferruginous duricrust
8.4
Brown
2500 1800 1100 Cr ppm
Soil
2300
Talc Kaol
Chl/V
Musc
Hem Goe Magh
Dol
Calc
Qtz
Colour
2800
2
0.5 Thickness (m) Calcrete
Rock weathering and structure of the regolith
Figure 6.7: A generalised profile through weathered ultramafic rocks, Panglo, WA (after Scott 1990).
Significant concentration of Ni by these mechanisms can produce lateritic nickel deposits. Unweathered serpentinised ultramafic rocks at Panglo (Yilgarn Craton of WA) are dominated by chlorite and talc, with only trace quartz (Figure 6.7). With the commencement of weathering, chlorite readily breaks down to form goethite, kaolinite, vermiculite and smectite in the lower saprolite. Thus, although the Mg in talc is retained, Mg from chlorite is gradually depleted up the profile. Aluminium and Fe from chlorite are retained in the neo-formed clays and Fe oxides. Ferruginous duricrust above the upper saprolite is particularly Fe-rich, indurated by calcrete, and commonly incorporates a transported component (indicated by the presence muscovite) so that Mg
and Cr contents are diluted. The soil formed above ferruginous duricrust shows further dilution of the diagnostic ultramafic components (Mg and Cr), although these are still present in significant amounts. Chromium present in spinels is retained during weathering, but that in chlorite (up to 3500 ppm) is transferred to goethite (8700 ppm) formed from the chlorite breakdown (Scott 1990). 6.3.4 Weathering in detrital sedimentary rocks Detrital sedimentary rocks, and their low-grade metamorphosed equivalents, contain mainly quartz, muscovite, chlorite and, in some cases, plagioclase, carbonates, epidote and minor sulfides, together
113
Cr ppm
Al2O3%
K2O %
Fe2O3%
320
10
0.9
7.6
Leached Upper Saprolite
White to pale brown
230
16
2.6
1.6
White to pale brown
210
19
2.3
0.8
White to grey
230
18
2.6
1.7
Yellow
170
20
1.6
1.5
Black to grey
160
17
1.8
1.2
Goe Magh Musc Parag Chl/V Talc Kaol Smec Pyrite
Brown
Plag Calc Dol Hem
<1.0
Soil and calcrete
Qtz
Colour
approx 25
Thickness (m)
Regolith Science
>30
114
Lower Saprolite
Fresh Rock KSf017-07
Figure 6.8: A generalised profile through weathered shale, Panglo, WA (after Scott and Dotter 1990).
with trace amounts of accessory minerals (zircon, rutile, ilmenite and apatite) and carbonaceous matter. As many of the accessory minerals are the weathering products of other rocks, they are less affected by further chemical weathering under a new weathering regime. In profiles on weathered sedimentary rocks (especially sandstones), it can be difficult to distinguish the weathered and unweathered zones. Physical processes generally play a more important role in the development of weathering profiles on these rocks types – including increased fracturing and destruction of the rock fabric by stress relief, drying and wetting, freeze/thawing and veining by secondary minerals (such as Fe oxides).
Mottling and strong bleaching are other clues to identifying the weathered zone. In unweathered Archean shale (interbedded with mafic volcanic rocks) in the Panglo area (Yilgarn Craton of WA), the major minerals are quartz, plagioclase, mica (muscovite + paragonite), carbonates, chlorite with or without pyrite. Weathering commences along cleavage/bedding planes in the shale, with the carbonates, pyrite and feldspar breaking down to form goethite and kaolinite. After this initial mineralogical change there is little further change throughout the lower saprolite (Figure 6.8). In the upper leached saprolite, smectite and hematite are developed, but chemical compositions are not
Rock weathering and structure of the regolith
markedly different from those in the lower saprolite, except for residual concentration of Cr in neo-formed goethite and resistate rutile (Scott and Dotter 1990). The soil – with higher Fe but lower K and Al than in the leached upper saprolite – probably has a transported component at Panglo (Figure 6.8). Non-Archean shales tend to have higher K contents than their Archean equivalents. The K may be present as K-feldspar and/or biotite, both of which weather out leaving lower K through the saprolite (for example, Dickson and Scott 1997). Chromium contents in shales in less mafic environments may be lower than at Panglo. Commonly, up to 1% organic matter may be present in black carbonaceous postProterozoic shales, with higher abundances in mineralised shales (Appendix 2). This organic matter (kerogen) has variable H/C and O/C ratios (Saxby 1976) which affects its weathering susceptibility, but ultimately it is broken down during weathering (Section 7.3.1) to produce a pallid saprolite: that is, black shales weather to white saprolite. In arenaceous rocks, the feldspars and phyllosilicates (micas, chlorite) weather to kaolinite and goethite/hematite, but the abundant quartz component and accessory resistates remain largely unaffected. Quartz-rich sandstones break down mechanically, but do not change chemically during weathering – apart from the formation of minor Fe oxide coatings in fractures/voids and, in some cases of extreme weathering, precipitation of additional secondary silica. There are commonly silty interbeds, which weather as described above. 6.3.5 Weathering in carbonate-rich rocks Limestones contain abundant calcite, with only minor silica/quartz and clays. Weathering involves dissolution of the calcite – with the loss of the Ca – and accumulation of the insoluble components to form a thin soil. (Dolomite tends to dissolve less rapidly than calcite, but, when it does, it releases Ca and Mg.) If there are significant clay and silt impurities or interbeds in the carbonate rock, these can be weathered and residually concentrated to form thicker (up to 30 m) accumulations of smectite, kaolinite or bauxite (gibbsite/boehmite). Minor Fe and Mn released from the carbonates during dissolution are generally oxi-
dised to form insoluble goethite, hematite and Mn oxides. The Fe oxides produce deep red-coloured residual soils known as terra rossa. Profiles on carbonate-rich rocks are very distinctive: with sharp contacts between the unweathered (undissolved) rock and the weathering residue (Robertson et al. 2006). Unless there is a significant, noncarbonate component, there is generally no saprolith. Percolation of water along fractures and accompanying dissolution of the carbonate may lead to spectacular karst topography, with jagged and pinnacled rock surfaces forming the weathering front (Figure 6.9). Quartz, residual clays, insoluble Fe and Mn oxides and wind-blown dust commonly accumulate in hollows and cavities in the unweathered limestone and may form irregular and thick deposits. Zirconium, Ti and Si – present as minor components in the original rock – are residually concentrated, although much of the quartz and clay is derived from the more silty interbeds. Other primary constituents may be residually concentrated (such as at Laowanchang in Guizhou Province, China, gold has been concentrated in this way). Further concentration of resistate minerals/elements occurs during the formation of the soil. Chlorite in the soil but not in underlying saprolite or rock (Figure 6.9) reflects the presence of an introduced component in the soil at Laowanchang.
6.4 EFFECTS OF EROSION AND DEPOSITION ON PROFILE DEVELOPMENT The weathering profiles described above are mature profiles for which there has been sufficient time and/ or intensity of weathering for the zones to become well developed. However, in the case of recent exposure, the profiles may be immature, with only some features of the saprolite present and little or no surficial ferruginisation and soil development. Mature weathering profiles may be partly eroded and a new weathering profile developed across the unconformity. In inset valleys (paleovalleys), clayrich sediments commonly develop ferruginous megamottles (more than 100 mm across) as a result of later weathering (Ollier et al. 1988; Figure 6. 10). These mega-mottles are formed by the mobilisation and concentration of Fe oxides commonly around roots or
115
Zr ppm
Al2O3 %
TiO2 %
Fe2O3%
SiO2%
CaO%
Soil
Red
860
15
3.6
19
50
<0.1
Saprolite
Red-brown
420
17
3.8
13
56
0.2
White to grey
0.7
0.1
<0.1
2.7
55
Calc Kaol Chl Goe Hem Anatase
Colour
-
Regolith Science
Qtz
116
Chert Fresh rock
Figure 6.9: A generalised profile through weathering limestone, Laowanchang Au deposit area, Guizhou Province, China (data from J Mao pers. comm. 2007).
around shrinkage cracks (Anand 2001; Johnson and McQueen 2001). After a period of erosion – especially where a channel has developed – renewed deposition may commence, with accumulation of vegetable matter, which, in some cases, is pure and thick enough to form lignite, which may be enriched in elements such as Mo and V. Such a reducing horizon can cause precipitation of secondary sulfides and U mineralisation such as at Mulga Rock, WA (Figure 6.11; Douglas et al. 2005). The addition of transported material to weathering profiles can particularly affect the chemical composition of the soil layer. Most soils are affected by some colluvial movement of adjacent material (metre scale movement), but, unless this movement introduces material derived from a different rock type or from material with a different degree of weathering, it is unlikely to be recognised. Aeolian additions to profiles are obvious where they are present as sand
dunes, but they may not be so obvious when the introduced material is finer grained silt and clay. This extraneous component can also be worked into the profile by soil churning or bioturbation (Sections 3.4.4 and 8.3.2). This material can be recognised in the soil by detailed analysis of the particle size distribution, particle morphology and chemical and mineralogical characteristics (Scott 2005; Dickson and Scott 1998; Tate et al. 2007).
6.5 WEATHERING REGIMES Chemical and physical conditions in the regolith can change significantly through time particularly in response to erosional, climatic, hydrologic and biological changes. At any one location, this can be the result of crustal plate movement, global or local climate change, as well as local tectonic and geomorphic factors. Regolith that has formed and been preserved over a long time span may therefore have
Rock weathering and structure of the regolith
Hill belt
Plains DPB Plains
Modern drainage
Tertiary inset-valleys
5
Breakaway Dissected paleolandsurface on basement 1
6 7
4
2 4
3
8
BIF
1
2
Quaternary colluvium and alluvium Tertiary sediments
Ferricrete (inverted ferruginised Tertiary sediments) Saprolite
Lateritic residuum
Basement
4
3
Lag
Soil
Lateritic residuum
Silty hardpanised colluvium
Ferricrete
Collapsed mottled saprolite
Gravelly hardpanised colluvium Gravels
Unconformity
Lateritic residuum Collapsed mottled saprolite
Mottled saprolite
Mottled saprolite
Fe saprolite
Saprolite
5
6 Soil Silty hardpanised colluvium
7
8 Soil
Soil
Soil
Hardpanised colluvium Red clay
Nodularslabby ferricrete
Mega mottled clay
Gravel
Mottled clay
Gravel Megamottled clay
Gravel
Unconformity Collapsed mottled saprolite
Dolomite Pisolitic clay
Grey clay
Lateritic res iduum
Lateritic residuum
Gravel & sand Unconformity
Figure 6.10: Formation of mega-mottles as the result of multiple weathering events (after Anand 2001).
117
118
Regolith Science
U (%) 0
Fe (%) 1
0
Cu (%) 3
0
1
Zn (%) 0
Ce (%)
10 0
0.5 0
Se (%) 0.1 0
Al (%) 15 0
Ti (%)
Zr (ppm)
1.5 0
200
35
Silty Clay
Depth (m)
36
Clay rich 37 Lignite
38
Lignite 39
40
Figure 6.11: Development of a reducing environment leading to sulfide and U precipitation in clay-rich lignite, Mulga Rock, WA (after Douglas et al. 2005).
experienced a number of different weathering regimes. It is important to remember that not only is weathering a continuous process, but also the rate and type of weathering can vary depending on the stability of the minerals in the regolith and underlying bedrock under the conditions of the prevailing weathering regime. Changes in weathering conditions can have important implications for the dispersion of target and pathfinder elements used in geochemical exploration. Significant changes result in complex, multistage geochemical dispersion effects that may not be fully explained by the current weathering conditions. Patterns from different weathering regimes may be preserved in different parts of the landscape or they may be superimposed – depending on the history of landscape development, relative uplift, erosion and deposition. Figure 6.12 shows how the superimposition of a second weathering regime may largely obscure the effects of initial weathering. 6.5.1 Effects of long-term climate change Climate is a major control on weathering regimes and, for many weathering profiles, climatic conditions have changed significantly during their formation (see
Chapter 2). The simplest case is for two markedly different climates, but, for very old weathering profiles – particularly in some Proterozoic and Archean terrains – the climatic conditions and weathering processes have been much more complex. Unravelling this complexity requires detailed studies using techniques such as paleomagnetic or isotopic dating (McQueen et al. 2007; Anand et al. 1997). For example, over much of southern Australia, major climatic variations through the Late Cretaceous and Cenozoic generally resulted in early deep chemical weathering under predominantly warm humid conditions (with high availability of organic material) onto which has been superimposed drier chemical weathering under increasingly arid conditions since the Late Cenozoic. The detailed picture is more complex, with fluctuations to at least two cooler-dry episodes before the Oligocene (McGowran and Li 1998). Paleomagnetic dating of ferruginous mottling in weathering profiles across the region indicates a number of periods of major hematite fixation – reflecting episodes of major oxidation and profile drying following intensive chemical weathering (Pillans 2006; McQueen et al. 2007; Smith et al. in press). These dated periods occurred during the
Rock weathering and structure of the regolith
1c 1b 1a Regime 1
erosional stripping surface
Overprint Regime 2
surface
surface
uplift 3 2a
Regime 1
2 etc Regime 2 in cover
Overprint
stable
exposure level
burial
Figure 6.12: Some possible outcomes of superimposition of different weathering regimes.
Paleocene (circa 60 Ma), Miocene (circa 16 and 12 Ma) and at some sites in the Late Cenozoic (circa 5 Ma). In some profiles, there is also evidence of earlier deep oxidation episodes, including in the Jurassic and Carboniferous (Pillans 2006). See Chapter 2 for further discussion of dating regolith materials. 6.5.2 Geochemical dispersion under different weathering regimes Where weathering has occurred under a wide range of contrasting conditions, some minerals formed under an early weathering regime have a stability range that covers the conditions of later weathering. Thus kaolinite, Fe oxides, quartz and secondary silica will persist over a very wide range of weathering conditions and may retain minor and trace elements that are incorporated into their structures or concentrated in occluded accessory and secondary minerals. However, weakly adsorbed elements can be further dispersed under changed chemical conditions (particularly changed water activity, Eh and pH). Less-stable weathering products – such as smectitic clays, carbonates, sulfates, halides that are stable under one weathering regime – may be completely broken down under different conditions, releasing their contained major, minor and trace elements. The availability of water under different weathering regimes is the major control on the resulting regolith geochemistry, element dispersion and fixation. Weathering regimes can therefore be considered as part of a spectrum from high water availability (wet,
mainly tropical and temperate, typically with annual rainfall greater than 500 mm) to low water availability (dry, mainly arid to semi-arid with annual rainfall less than 500 mm, or frozen). Seasonal fluctuations from wet to dry are also a common type of weathering regime in the subtropics. Wet conditions – characterised by high water tables and a typically abundant and active biota – tend to favour hydration/hydrolysis reactions and mobility of reduced species, such as Fe2+, and Mn2+. Groundwater pH conditions are generally neutral to acid – the latter particularly in zones with abundant decaying organic material and around oxidising sulfides. High organic content promotes organo-complexing of many elements (Section 7.5). Most major elements are strongly leached – resulting in relative enrichment in Al, Fe and Ti. Trace elements are also strongly leached – particularly Cu, Zn, Cd and Ag – but some, including As, Bi, Cr, Mo, Pb and V, are retained or relatively concentrated in Fe-rich zones where these develop (Lecomte and Zeegers 1992). Dry conditions typically favour oxidation reactions and a change to more complex, groundwater compositions, particularly with higher salinity, increased activity of carbonate and sulfate and regional neutral to alkaline pH. Some major elements (such as Ca, Mg and Na) may be retained in smectitic clays or enriched by evaporative concentration and re-precipitation in the regolith. More alkaline conditions favour mobility of trace elements such as As, Mo and U. Chloride and thiosulfate complexing can increase the solubility and mobility of some elements, such as Au, whereas other elements – especially Pb, Ag, Ba and Hg – become relatively fixed as insoluble chlorides and sulfates. Marked pH gradients around sulfide deposits that continue to weather generally results in dispersion of elements such as Cu and Zn to form distinct secondary haloes. Weathering conditions in the Yilgarn Craton of Western Australia have spanned the range from warm humid conditions in the Late Mesozoic and Early Cenozoic to arid conditions in the Late Cenozoic (Chapter 2). There early-formed lateritic profiles – which are characterised by a modest amount of strong element leaching in the upper saprolite – have been modified under the more arid conditions, with the leached saprolite becoming more extensive as the water
119
120
Regolith Science
a
Water table
Time
Soil Ferruginous zone Mottled zone
Saprolite
Fresh rock
b
Ferruginous zone Mottled zone Leached saprolite
Saprolite
Fresh rock
Continued slow weathering
Water table KSf012-07
Figure 6.13: Effect of aridity on lateritic profiles of the Yilgarn Craton (a) initial profile formation (b) modification (after Butt 1989).
table fell (see Figure 6.13). Under these conditions, the groundwaters became quite saline (Gray 2001).
6.6 DISCRIMINATING PARENT ROCK TYPES In deeply weathered terrains, it is generally difficult to establish the parent rock type for regolith materials because of the profound mineralogical and chemical changes related to weathering and, in some cases, to secondary cementation and induration. In saprock and saprolite, the original rock fabric may be preserved to give clues to the original parent. In many cases, however, these features are destroyed – particularly in the upper part of the weathering profile. A number of studies have investigated the use of geochemical criteria – based on the least mobile elements – to discriminate parent rock types for in situ regolith.
The most widely used technique employs Ti/Zr ratios to discriminate between regolith derived from igneous rock types over the range from mafic to felsic (Figure 6.14; Hallberg 1984). These two trace elements are typically concentrated in rutile and zircon – minerals that are highly resistant to weathering – and are thus considered to be relatively immobile. (Some Ti may also be present in ilmenite, sphene, micas, amphiboles and pyroxenes and, as these minerals weather, Ti is released, but re-precipitated as stable anatase [TiO2], which is effectively immobile (Butt 1985).) Furthermore, because Hf is more concentrated in the zircon of more fractioned igneous rocks, the Zr/Hf ratio in zircon can also be used to gain information about the parentage of highly weathered material (Figure 6.15). Minor and trace element contents of resistate minerals are also useful as guides to mineralisation (for example, Scott and Radford 2007).
Rock weathering and structure of the regolith
Rhyolite
600
Dacite
Zr ppm
500 400
15 Al2O3
ABS - Average bulk sediment CAS - Calc-alkaline suite SPG - Strongly peraluminous granite
Andesite
300
Note: Al2O3, TiO2 - wt % Zr- ppm SPG
700
200 100 0
sh ale s
0
Basalt
ABS
2000 4000 6000 8000 10000 12000 Ti ppm
ne
sto
Figure 6.14: Ti/Zr plot for discriminating weathered volcanic rocks (after Hallberg 1984).
d an
Of the major elements, Al is generally the least mobile and plots of Al2O3-TiO2-Zr can also be used to discriminate regolith from different parent rocks. Including Al allows characterisation of regolith for a wider range of rock types, including sedimentary and metamorphic rocks (for example, Garcia et al. 1994; Figure 6.16). Interpretation of compositional variation for these three elements is based on the premise that sedimentation involves weathering, transport, mixing from different sources and sorting. In the first three processes, the contents of the less-soluble elements, such as Al, Ti and Zr, may vary in response to the degree of leaching of the soluble elements. However, their relative proportions are transferred from the source area into the bulk sediment or regolith without, or with little, modification. This material is then sorted according to the hydraulic properties 60
Gabbro 50
Carbonatites Granodiorite Granite
Zr/Hf
40 30
Pegmatite 20 10
Pegmatite 0
1.0
1.5
2.0
CAS
s
27.0
Hf(%) Figure 6.15: Zr/Hf for zircons in igneous rocks (after ÿerný et al. 1985; Wang et al. 1996).
Zr
300 TiO2
Figure 6.16: The 15Al2O3 –300TiO2–Zr weathered rock type discriminant diagram (after Garcia et al. 1994).
of its mineral components to produce a chemical fractionation between complementary shales and sandstone (Figure 6.16). Other rock types, such as felsic and mafic igneous rocks or immature volcanicderived sediments, will also plot in specific fields.
6.7 CHARACTERISING AND IDENTIFYING REGOLITH MATERIALS Major chemical changes that occur during weathering – particularly progressive loss of Na+, K+, Ca2+, Mg2+ (some Si4+) and retention of Si4+, Al3+ and Fe3+ (Section 5.4.2), can be used to help to characterise regolith materials that are at different stages of mineralogical/chemical evolution. The K/Al and Mg/Al ratios can commonly be used in this way by comparing two mobilised elements with a relatively immobile major element. Figure 6.17 shows an example of the compositional distribution in terms of this index for three different types of regolith found in both the Kalgoorlie region of Western Australia and the Cobar region of eastern Australia (Johnson and McQueen 2001; McQueen 2006). The three different regolith materials in these examples are in situ saprolite/ saprock, lacustrine clays (difficult to distinguish visually from weathered saprolite) and younger ferruginous alluvium/colluvium. These materials have different regolith histories that are reflected in their
121
122
Regolith Science
0.35
Smectite/chlorite/dolomite
Ferruginous alluvium
0.25
Muscovite Kaolinite
0.20 0.15 0.10 0.05
Saprolite on mafic rocks Transported clays
Saprolite on siltstones/sandstones
0 0
0.1
0.2
0.3
0.4
0.5
0.6
K/Al (wt%) Figure 6.17: Compositional variability of regolith in terms of Mg/Al versus K/Al (after McQueen 2006).
mineral and chemical composition. The lacustrine clays were deposited in the Late Mesozoic to Early Cenozoic and appear to have been derived from a deeply weathered landscape. They were well sorted during erosion and transport to produce a kaolinite– smectite–quartz dominant sediment. The saprolite/ saprock contains significant muscovite and illite, which have been variably altered to kaolinite – depending on degree of in situ weathering and depth in the profile. The younger alluvium/colluvium was deposited in the Late Cenozoic when climatic conditions were significantly drier and chemical weathering less intense. Erosion of less altered profiles, more limited sorting and low levels of post-depositional weathering produced material that retained significant amounts of weakly altered phyllosilicates. These sediments can also contain local concentrations of dolomitic calcrete, which will be apparent in their Mg/Al ratio. There is some compositional overlap between the three materials, and the clearest distinction is between saprolite and transported clays. This approach to chemically distinguishing regolith types can be applied within particular regions where there are regolith components characterised by different parent rock compositions, different degrees of weathering or different histories of sorting or remixing/ homogenisation during transport. The differences can be established with an orientation survey. Trace element characteristics – particularly for those elements hosted by resistate minerals – can be useful for identifying different types of regolith. For
6.8 DEGREE OF WEATHERING AND WEATHERING HISTORY The Chemical Index of Alteration (CIA) has been widely used to quantify degree of rock weathering (Nesbitt and Young 1982). This index, CIA = 100Al2O3/ (Al2O3+CaO+Na2O+K2O), reflects the breakdown of feldspars and mica to kaolinite, but it has a major drawback in that it estimates the total history of chemical weathering from the primary source rock: that is, including that already present in sedimentary rocks subjected to further weathering. It is thus difficult to apply this index as a direct measure of the in situ weathering of a particular regolith sample. However, it can be useful for comparing samples within profiles developed on a variably weathered common rock type (Figure 6.18). Weathering trends within a profile, or 0 10 20
Depth (m)
Mg/Al (wt%)
0.30
example, regolith derived from mafic or ultramafic rocks typically has higher contents of Ti, V and Cr – hosted in rutile/anatase, ilmenite, magnetite and chromite – than regolith derived from felsic or meta-sedimentary rocks. Aeolian material in the regolith may have high Zr (and Hf) contents relative to in situ material due to the greater abundance of zircon grains in the introduced material (Tate et al. 2007).
30 40 50 60 70 60
KSf028-07
70
80
90
100
CIA (Chemical Index of Alteration)
Figure 6.18: Chemical Index of Alteration (CIA) variation with depth in weathered metasediments, Cobar region (after McQueen 2006).
123
Rock weathering and structure of the regolith
A (Al2O3)
10
100
Profile CBAC 215, Cobar region Down profile (less weathered)
10 90
80
Muscovite
Most weathered saprolite
1
20
19m 25m 31m 37m 43m 49m
30 70
0.1 40
Saprock
60
Feldspar
Weathering trend for sedimentary rocks
50
50
0.01
40
Igneous rock trend
FM (Tot Fe2O3+MgO)
CNK (CaO+Na2O+K2O)
Figure 6.19: A-CNK-FM plot for a weathering profile through metasediments, Cobar region (after McQueen 2006).
group of profiles, on common rock types can also be distinguished using the relative mole proportions of Al2O3, (CaO+Na2O+K2O) and (total Fe2O3+MgO) (Nesbitt and Young 1989). Thus an A-CNK-FM plot for samples from a single weathering profile within metasediments in the Cobar region (Figure 6.19) reflects the relative increase in Al and loss of Fe, Mg and alkalis as weathering proceeds. This approach is complicated by the inclusion of Fe, which can be accumulated at redox boundaries or rapidly mobilised from Fe-bearing carbonates in weathering sedimentary rocks. Formation of secondary dolomite can also complicate Mg contents of strongly weathered regolith.
20
Depth (m)
Depth (m)
Profile CBAC 217
10
30 40 50 60 70
The REE are commonly used as petrogenetic indicators because of their similar chemical properties, typically low solubilities and assumed resistance to fractionation in supracrustal environments. However, a number of studies have shown that under some weathering conditions REE are significantly mobilised and fractionated (for example, Nesbitt 1979; Duddy 1980; Sharma and Rajamani 2000; Gray 2001; McQueen 2006). Far from being immobile, REE can be significantly redistributed during weathering to the point where their depletions in one part of the profile, and subsequent enrichments in other parts, can provide an index for the intensity and style of chemical weathering. Analysis of REE distributions through deeply weathered profiles in the Cobar region indicates significant leaching in the upper
Approximate initial ratio
20
Approximate initial ratio
30 40 50 60
0
2
4
6
Ce/Nd
8
10
12
Y
Figure 6.20: REE distribution in a weathering profile through metasediments, Cobar region (after McQueen 2006).
Profile CBAC 215
10
La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu
Note: REE(Y) normalised to North American Shale Composite (Gromet et al. 1984)
60
70
0
1
2
3
4
5
6
Ce/Nd
Figure 6.21: Ce/Nd plots for weathering profiles through metasediments, Cobar region (after McQueen 2006).
7
124
Regolith Science
parts, and enrichment in the lower zone close to the weathering front (Figure 6.20). The light rare earth elements (LREE) – particularly Ce – appear to have been the most mobile under the weathering conditions that pertained in this region. This pattern is less marked in the least-weathered profiles and there is also some indication that profiles in different settings have different patterns of REE mobility and enrichment (McQueen 2006). The REE may therefore provide a basis for evaluating the extent of chemical weathering and leaching. Comparison of Ce (the most mobilised LREE) with Nd (another LREE with similar properties and initial distribution) is one possibility for a REE mobility and weathering index. Examples of Ce/Nd plots for different weathered profiles in the Cobar region are shown in Figure 6.21. Departure of the Ce/Nd ratio from the initial ratio (typically around 1.5 to 2.5 for sedimentary rocks) indicates the relative depletion or enrichment of Ce and provides an indication of chemical leaching of the rocks.
6.9 REFERENCES Anand RR (1984). Mineral weathering in lateritic saprolite. PhD Thesis (unpublished), University of Western Australia. Anand RR (2001). Evolution, classification and use of ferruginous regolith materials in gold exploration, Yilgarn Craton, Western Australia. Geochemistry, Exploration, Environment, Analysis 1, 221–236. Anand RR and Paine M (2002). Regolith geology of the Yilgarn Craton, Western Australia: Implications for exploration. Australian Journal of Earth Sciences 49, 3–162. Anand RR, Fraser SJ, Jones MR, Shu L, Munday TJ, Phang C, Robertson IDM, Scott KM, Vasconcelos P, Wildman JE and Wilford J (1997). ‘Geochemical exploration in regolith-dominated terrain, North Queensland’. CRC LEME Restricted Report 63R/ CSIRO Exploration and Mining Report 447R. (Reissued as Open File Report 120, 2002. CRC LEME, Perth.) Aspandiar MF (1998). Regolith and landscape evolution of the Charters Towers area, North Queens-
land. PhD Thesis. Australian National University, Canberra. Butt CRM (1985). Granite weathering and silcrete formation on the Yilgarn Block, Western Australia. Australian Journal of Earth Science 32, 415–432. Butt CRM (1989). Genesis of supergene gold deposits in the lateritic regolith of the Yilgarn Block, Western Australia. In The Geology of Gold Deposits: the Perspective in 1988. (Eds RR Keays, WRH Ramsay and DI Groves) pp. 460–470. Economic Geology Monograph 8. Proceedings of Bicentennial Gold ’88. The Economic Geology Publishing Company, El Paso, Texas. Butt CRM, Scott KM, Cornelius M and Robertson IDM (2005). Sample media. In Regolith expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 53–79. CRC LEME, Perth. Cerný P, Meintzer RE and Anderson AJ (1985). Extreme fractionation in rare-earth granitic pegmatites: selected examples of data and mechanisms. Canadian Mineralogist 23, 381–421. CSIRO (1983). Soils: An Australian Viewpoint. Division of Soils, CSIRO, Melbourne/Academic Press, London. Dickson BL and Scott KM (1997). Interpretation of aerial gamma ray surveys – adding the geochemical factors. AGSO Journal of Australia Geology and Geophysics 17, 187–200. Dickson BL and Scott KM (1998). Recognition of aeolian soils of the Blayney district, NSW: implications for exploration. Journal of Geochemical Exploration 63, 237–251. Douglas GB, Butt CRM and Gray DJ (2005). Mulga Rock uranium and multielement deposits, Officer Basin, WA. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 415–417. CRC LEME, Perth Duddy IR (1980). Redistribution and fractionation of rare-earth and other elements in a weathering profile. Chemical Geology 30, 363–381. Eggleton RA (Ed.) (2001). The Regolith Glossary: Surficial Geology, Soils and Landscapes. CRC LEME, Perth.
Rock weathering and structure of the regolith
FAO-UNESCO (1988). ‘Soil map of the world, revised legend’. World Soil Resources Report 60, United Nations Food and Agriculture Organization, Rome. Garcia D, Fonteilles M and Moutte J (1994). Sedimentary fractionation between Al, Ti and Zr and the genesis of strongly peraluminous granites. Journal of Geology 102, 411–422. Gilkes RJ , Scholz G and Dimmock GM (1973). Lateritic deep weathering of granite. Journal of Soil Science 24, 523–536. Gray DJ (2001). Hydrogeochemistry in the Yilgarn Craton. Geochemistry: Exploration, Environment, Analysis 1, 253–264. Hallberg JA (1984). A geochemical aid to igneous rock identification in deeply weathered terrain. Journal of Geochemical Exploration 20, 1–8. Johnson CB and McQueen KG (2001). The nature of gold-bearing palaeochannel sediments in the Gidji area north of Kalgoorlie, Western Australia. Quaternary International 82, 51–62. Lecomte P and Zeegers H (1992). Humid tropical terrains (rainforests). In Regolith Exploration Geochemistry in Tropical and Subtropical Terrains. (Eds CRM Butt and H Zeegers) pp. 241–294. Elsevier, Amsterdam. Loughnan FC (1969). Chemical Weathering of the Silicate Minerals. Elsevier, New York. McGowran B and Li Q (1998). Cainozoic climate change and its implications for understanding the Australian regolith. Geological Society of Australia Special Publication 20, 86–103. McQueen KG (2006). Unravelling the regolith with geochemistry. In Regolith 2006 – Consolidation and Dispersion of Ideas. (Eds RW Fitzpatrick and P Shand) pp. 230–235. CRC LEME, Perth. McQueen KG, Gonzalez OR, Roach IC, Pillans BJ, Dunlap WJ and Smith ML (2007). Landscape and regolith features related to Miocene leucitite lava flows, El Capitan northeast of Cobar, NSW, Australia. Australian Journal of Earth Sciences 54, 1–17. Moore CL (1996). Evaluation of regolith development and element mobility during weathering using the isocon technique. Geological Society of Australia Special Publication 20, 141–147.
Nesbitt HW (1979). Mobility and fractionation of rare earth elements during weathering of a granodiorite. Nature 279, 206–210. Nesbitt HW and Young GM (1982). Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature 299, 715–717. Nesbitt HW and Young GM (1989). Formation and diagenesis of weathering profiles. Journal of Geology 97, 129–147. Ollier CD, Chan RA, Craig MA and Gibson DL (1988). Aspects of landscape history and regolith in the Kalgoorlie region,Western Australia. BMR Journal of Australian Geology and Geophysics 10, 309–321. Pillans B (2006). Highlights from the LEME geochronology project. In Regolith 2006 – Consolidation and Dispersion of Ideas. (Eds RW Fitzpatrick and P Shand) pp. 279–283. CRC LEME, Perth. Robertson IDM, Craig MA and Anand RR (2006). ‘Atlas of regolith materials of the Northern Territory’. Open File Report 196, CRC LEME, Perth. Saxby JD (1976). The significance of organic matter in ore genesis. In Handbook of Stratabound and Stratiform Ore Deposits.1. Principles and General Studies, Volume 2. Geochemical Studies. (Ed. KH Wolf) pp. 111–133. Elsevier, Amsterdam. Scott KM (1990). ‘The mineralogical and geochemical effects of weathering on volcanics from the Panglo Deposit, Eastern Goldfields, WA’. CSIRO Division of Exploration Geoscience Restricted Report 143R. (Reissued as Open File Report 24, 1998. CRC LEME, Perth.) Scott KM (2005). Blayney-Orange district, New South Wales. In Regolith Landscape Evolution Across Australia. (Eds RR Anand and P De Broekert) pp. 76–79. CRC LEME, Perth. Scott KM and Dotter LE (1990). ‘The mineralogical and geochemical effects of weathering on shales at the Panglo Deposit, Eastern Goldfields, WA’. CSIRO Division of Exploration Geoscience, Restricted Report 171R. (Reissued as Open File Report 48, 1998. CRC LEME, Perth.) Scott KM and Martinez A (1990). ‘The mineralogical and geochemical effects of weathering in mafic and
125
126
Regolith Science
ultramafic profiles, Mt Magnet, WA’. CSIRO Division of Exploration Geoscience Restricted Report 178R. (Reissued as Open File Report 30, 1998. CRC LEME, Perth.) Scott KM and Radford NW (2007). Rutile compositions at the Big Bell Au deposit as a guide for exploration. Geochemistry, Exploration, Environment, Analysis 7, 353–361. Sharma A and Rajamani V (2000). Major element, REE and other trace element behaviour in amphibolite weathering under semiarid conditions in southern India. The Journal of Geology 108, 487–496.
Smith ML, Pillans BJ and McQueen KG (in press). Palaeomagnetic evidence for periods of intense oxidative weathering, McKinnons mine, Cobar NSW. Australian Journal of Earth Sciences. Tate SE, Greene RSB, Scott KM and McQueen KG (2007). Recognition and characterisation of the Aeolian component in soils in the Girilambone region, north western New South Wales, Australia. Catena 69, 122–133. Wang RC, Fontan F, Xu SJ, Chen XM and Monchoux P (1996). Hafnium zircon from the apical part of the Suzhou Granite, China. Canadian Mineralogist 34, 1001–1010.
7
Geomicrobiology of the regolith Frank Reith, Mira Dürr, Susan Welch and Stephen L Rogers
7.1 INTRODUCTION
7.2
Microorganisms are the dominant life forms on Earth, with bacteria alone constituting 50% of the living biomass. Soils and surface waters contain up 1012 cells/g, and sediments and shallow regolith materials up to 106 cells/g, whereas in the deep sub-surface (between 50 m and 10 000 m) 104 to 105 cells/g have been detected (Juniper and Tebo 1995; Paul and Clarke 1996). They play a critical role in sustaining the fertility of soils, producing foods and beverages, developing new types of medicines and improving the processing of mineral resources. Microorganisms have been known to the scientific community for more than 300 years – and their metabolic, ecologic and genetic capabilities have been explored during the last 100 years – but their relevance to geological and geochemical processes, such as dissolution and diagenesis of minerals and associated metals, and their influence on the formation of mineralisation, has only recently been recognised (Ehrlich 1996, 2002). This chapter outlines the broad concepts of microbial ecology, microbial biochemistry and molecular microbial (genetic) techniques relevant to the study of regolith geomicrobiology.
The term ‘microorganism’ was coined for a large number of genetically, metabolically and ecologically diverse organisms, which have one characteristic in common: they are small (that is, less than one to tens of microns), mostly single-celled organisms. The universal phylogenetic tree of life (Figure 7.1), which is based on 16S rRNA analysis, contains three domains: bacteria, archaea and eucarya (Pace 1997). The domains bacteria and archaea comprise the prokaryotes, which are entirely made up of single-celled organisms. Similarly, the domain eucarya, consists largely of microorganisms (principally fungi and algae), but also contains higher organisms such as mammals. Prokaryotes in particular have developed a wide array of physiological processes to generate metabolic energy required for cell function (Table 7.1; Figure 7.2), whereas eukaryotes are metabolically more limited and generate energy mainly by photoautotrophy (such as algae and higher plants) and chemoorganotrophy (such as fungi and higher animals; Madigan and Martinko 2006). Eukaryotic cells are bigger, structurally more complex, contain defined cell organelles and have a membrane enclosed nucleus
MICROORGANISMS
128
Regolith Science
Bacteria
Domains
Mitochondrion Proteobacteria
Gram positive
Cyanobacteria Flavobacteria Thermotoge Thermodesulfobacterium
Archaea Green non-sulfur bacteria
Eucarya Animals Entamoebae Slime moulds
Fungi
Methanosarcina Methanobacterium Halophiles ThermoMethanoproteus coccus ta eo ThermoPyroa h coccales rc dictium ya r C r u a rc a n E ha e ota
Aquifex
Green and purple sulfur bacteria
Prokaryotes
Plants and algae Ciliates Flagellates Trichomonads
Diplomonads
Microsporidia FRf007-08
Eukaryotes
Figure 7.1: The domains of life, as defined by the universal phylogenetic tree based on 16S ribosomal RNA sequences.
containing the DNA; prokaryotes are much smaller, have no defined organelles and their DNA is located in the cytoplasm. Fossil records demonstrate that prokaryotes have inhabited Earth for more than 3.5 Ga (Barns and Nierzwicki-Bauer 1997). During their long evolution, prokaryotes have developed numerous genetic, physiological and ecological capabilities that allow them to survive wherever water is available and temperatures range between –50 and 121°C (Stetter et al. 1990; Kashefi and Lovley 2003). Thus, prokaryotes are found in environments ranging from deep marine sediments (Parkes et al. 1994), deep-sea hydrothermal vents (Juniper and Tebo 1995), hydrothermal geysers, deep rock fractures (Pedersen 1993), deserts (Adams et al. 1992), polar regions (Vincent and James 1996) to acidic, heavy metal-polluted mine wastes (Ledin and Pedersen 1996). Depending on their ability to reduce O2 (that is, use O2 as an electron acceptor; see Figure 5.6) during metabolism, microorganisms can be divided into obligate aerobes (require O2), obligate anaerobes (grow in absence of O2, which is toxic to them) and facultative anaerobes (able to change from aerobic to anaerobic metabolism depending on O2 availability) (Figure 7.2; Ehrlich 2002). Facultative anaerobes may be well
suited to many regolith environments, such as soils or sediments (Paul and Clark 1996). Another division divides microbes into heterotrophic and autotrophic organisms (Figure 7.2; Madigan and Martinko 2006). Heterotrophs need organic compounds as carbon sources for cell growth, whereas autotrophs are able to fix CO2 from their environment. Depending on the energy sources used, microbes can be divided into phototrophs and chemotrophs (Figure 7.2). Phototrophs use sunlight to create metabolic energy, whereas chemotrophs gain energy through the reduction and oxidation of chemical compounds. Among chemotrophs, organotrophs use reduced carbon sources (such as glucose or acetate) that they oxidise to CO2, whereas lithotrophs use reduced inorganic compounds such as H2, Fe2+ or H2S, to create metabolic energy. However, many microorganisms, particularly prokaryotes, change their metabolic mechanisms according to the ambient environmental conditions (Madigan and Martinko 2006).
7.3 BIOGEOCHEMICAL ELEMENT CYCLES The Earth is a closed system that contains a limited quantity of each element. Approximately 20 elements
Geomicrobiology of the regolith
Aerobic (requires oxygen)
ion rat spi Re
En erg yf or bio syn the sis
Phototrophs (sunlight)
Microorganisms Anaerobic (does not require oxygen)
Chemotrophs (oxidation/reduction of compounds)
Source of carbon for cell growth Autotrophs (from CO2)
Heterotrophs (from organic compounds)
KSf005-08
Figure 7.2: Criteria used to group microorganisms. Note: These are broad groupings and are not mutually exclusive.
are used by organisms as cell components (that is, C, H, O, N, P, S, Mg, Na, Ca, Fe, B, Cr, Co, Cu, Fe, Mn, Mo, Ni, Se, W, V and Zn; Chapter 8; Madigan and Martinko 2006). Several other elements (that is, species of As, Se, Te, Sb, Hg, and U) are directly oxidised or reduced by prokaryotes in a variety of catabolic reactions in order to gain energy, and a number of other elements are indirectly influenced by microbial activity, such as through microbial influence on redox conditions, pH, formation complexing ligands and stability of complexes; among these are many heavy and precious metals (such as Pb, Cd, Ag and Au; Southam and Saunders 2005; Reith et al. 2007). In order to use an element in their metabolism, organisms need to convert it from one oxidation state to another, and thus they recycle them in the system. In particular, microorganisms play an essential role in the recycling of elements, because of their large total biomass, their metabolic diversity and their persistence in all habitats that support life (Ehrlich 2002). Microorganisms also serve as intercycle couplers by linking the cycles of individual elements, because one
metabolic reaction often influences more than one element cycle; for example, the oxidation of glucose (a reduced C compound) may be coupled with the reduction of nitrate (or sulfate) in the processes of denitrification (or sulfate-reduction)(Figure 7.3). 7.3.1 Carbon Cycle Carbon is recycled through all Earth’s major reservoirs (that is, the atmosphere, biosphere, hydrosphere, pedosphere, deeper regolith materials and lithosphere) by two processes: the geological C cycle, which operates over large time scales (millions of years); and the biogeochemical C cycle, which operates at shorter time scales (days to thousands of years; Des Marais 1997; Janzen 2004). Microorganisms control large parts of the biogeochemical C cycle and also a part of the geological C cycle, because they are able to:
s s s
fix CO2 from the atmosphere produce CH4 and CO2 during decomposition of organic matter precipitate carbonate minerals
129
Chemolithotrophic prokaryotes Fermentative bacteria and fungie plus synthrophic bacteria
Fe-oxidation, Mn-oxidation
Fermentationc, d
Denitrification
Redox boundary (fluctuating oxic– anoxic conditions)
Anoxic zone denitrifying zone
Denitrifying Bacteria
Sulfur-oxidising prokaryotes
Sulfur /sulfide oxidation
NO3 – /0.5N2
H2-bacteria
H2 oxidationb
+0.7
Nitrifying bacteria
Nitrification
Cyanobacteria, algae, autotrophic prokaryotes
Methano- and methylotrophic bacteria
0.5O2 /H2O
Associated microorganisms
Cl-oxidation
+0.8
Dominant and potential redox couples
Aerobic chemoorganoheterotrophic bacteria and fungi
Photosynthesis and autotrophya
Redox potential E’0 (at pH 7) [V]
Organic carbon oxidation and respirationb
Oxic zone
Geochemical zone
Associated microbial process Overall biogeochemical effect
Increaseg,k
Decreasel
Consumption of H + and LMWOAs
Decreasej
Increasei
Decreasef
Increaseh
Increasef
Increaseg
Decreasef
Excretion of LMWOAs, i.e. lixiviants for metal complexation, lowering pH
Oxidation of metals, precipitation of metal oxides, adsorption of trace metals
Production of H2SO4, oxidation of metals and metal ions, lowering pH, release of metals
Consumption of H +, fixation of CO2, biomass production, raising pH
Production of HNO3, lowering pH
Production of CO2 , lowering pH
Decomposition of organic material, production of CO2, lowering pH
Fixation of CO2, biomass produc tion, raising pH
Effect on metal mobility in regolith
Table 7.1: Linking important microbial metabolic capabilities and populations with geochemical zonation and trace metal element mobility in regolith.
130 Regolith Science
-0.3
Methanognesisd
Methanogenic zone
h
g
f
e
d
c
b
Corresponding process under anoxic conditions. Corresponding process under anoxic conditions with terminal electron acceptors other than O2. Also present below the redox boundary in the anoxic zone. Often associated with syntrophic H + reduction by syntrophic bacteria Often facultative anaerobic microorganisms. Madigan and Martinko 2006. Barker et al. 1997. Lebedeva et al. 1979
-0.3
Acetegenesisd
Acetogenic zone
a
-0.2 -0.3
Sulfate/sulfur reductiond
Sulfate/sulfurreducing zone
+0.7 +0.4 +0.2 +0.2 +0.1 +0.1 0.0
Metal reduction
Redox potential E’0 (at pH 7) [V]
Metal/metalloidreducing zone (e.g. Ag, As, Fe Co, Cr, Mn, Se U)
Geochemical zone
Associated microbial process
o
n
m
l
k
j
i
Methanogenic archaea plus synthrophic bacteria
Nordstrom and Southam 1997. Nealson and Stahl 2007. Stone 1997. Dassonville and Renault 2002. Saunders et al. 1997. Lloyd 2003. Southam and Saunders 2005.
CO2/CH4
Acetogenic plus synthrophic bacteria
Sulfate- and sulfurreducing prokaryotes plus synthrophic bacteria
SO42– /H2S S/H2S
CO2/Acetate
Metal-reducing prokaryotes
Associated microorganisms
Mn(IV)/Mn (II) Cr(VI)/Cr(III) Fe(III)/Fe(II) Ag(I)/Ag(0) As(V)/As(III) Se(VI)/Se(IV) U(VI)/U(IV)
Dominant and potential redox couples
Decreaseo
Consumption of H +, precipitation of metal sulfides
Utilisation of H +, H2, acetate (ligands), (raising pH)
Decreaseo
Decreasel
Decreasen
Transformation of trace metal ions from mobile oxidized to less mobile reduced forms
Consumption of H +, H2, CO2, LMWOAs
Increasem
Dissolution of Mn- and Fe-oxides
Overall biogeochemical effect
Effect on metal mobility in regolith
Geomicrobiology of the regolith 131
132
Regolith Science
(a)
Decomposition (aerobic and anaerobic)
CARBON
CO2 CYCLE (CH 2)n
(b) Denitrification (anoxic)
Carbon fixation
(c) Sulfate reduction
S
N2 fixation NH3 NITROGEN NO3CYCLE
Intercycle
H2S SULFUR SO 42CYCLE
S RE GO BA LIT CT H ER IU M
Nitrification (oxic)
(d) Metal oxidation
Sulfide, sulfur oxidation
(e) Phosphorus re-mineralisation
Coupling Metal (ox)
FRf008-08
METAL CYCLE
Metal (red)
Metal reduction
Phosphate PHOSPHORUS Phosphate (inorganic) (organic) CYCLE
Phosphorus assimilation
Figure 7.3: Schematic diagram illustrating the key-role of microorganisms in the biogeochemical (a) carbon (b) nitrogen (c) sulphur (d) metal and (e) phosphorus cycles – as well as their ability to connect these cycles (adapted from Madigan and Martinko, 2006).
s
catalyse the transformation of plant polymers into soil humus (Janzen 2004; Castanier et al. 1999 a, b).
The starting point of the biogeochemical C cycle is the fixation of CO2 and the formation of complex organic C compounds (Figure 7.4) Carbon dioxide fixation is a mostly driven by photosynthesis and, to a smaller extent, chemosynthesis (geochemical processes) (Madigan and Martinko 2006). In terrestrial ecosystems, higher plants dominate CO2 fixation, whereas in surficial marine systems complex microbial communities of cyanobacteria and unicellular green algal plants are the primary photosynthetic pro-
ducers of biomass (Janzen 2004). In some environments, such as the vicinity of black smokers at spreading centres or in deep sub-surface environments, chemosynthesis is the dominant process of CO2 fixation (Fyfe 1996). The main function of microorganisms in soils and deeper regolith materials is the decomposition of organic matter to CO2. Under oxic conditions, most organic compounds are directly oxidised to CO2 by aerobic hetero-organotrophic microorganisms (Paul and Clarke 1996) but, under anoxic conditions, decomposition occurs via a number of intermediate metabolic reactions and involves a succession of highly
Geomicrobiology of the regolith
Carbon is taken out of the rapid biogeochemical cycling and transferred into the slow geological cycle by the burial of refractory organic material, such as oil and coal, and via the precipitation of carbonate minerals, such as calcite, vaterite, aragonite or dolomite (DesMarais 1997). Biotic or biomediated carbonate precipitation is very common and occurs as part or consequence of the following processes:
Carbon cycle Atmospheric co2
Photosynthesis (plants)
Organic C
Photosynthesis (cyanobacteria)
Mineralisation (aerobic bacteria + fungi) Mineralisation (anaerobic bacteria)
CO2 CO CH4
1. production of skeletal carbonate by eukaryotes 2. autotrophic variation of the CO2 partial pressure (photosynthesis, methanogenesis) 3. heterotrophic mediation by bacteria and fungi.
SRf001-08
Mass balance estimates by Castanier et al. (1999 a, b) suggest that heterotrophic mediation is the most significant for the formation of these carbonates, rather than abiotic processes.
Figure 7.4: Biogeochemical carbon cycle.
specialised groups of microorganisms (Segers 1998). In the first step, complex polymers (such as cellulose or proteins) are hydrolysed to monomers by cellulolytic and other hydrolytic microorganisms (Millar 1974). The resulting monomers, (such as sugars or amino acids) are converted to low molecular weight organic acids (LMWOAs) such as lactic, citric, succinic, formic or butyric acids, H2 and CO2 by fermentative microorganisms. These products are then converted to acetate by hetero- and homoacetogenic bacteria and the acetate is then used by methanogens that produce CO2 and CH4 (Drake 1994; Segers 1998). When CH4 migrates to oxic environments methanotrophic prokaryotes convert it to CO2 (Segers 1998).
7.3.2 Nitrogen cycle Nitrogen is cycled through all of the main Earth’s reservoirs in four main steps: N fixation, N assimilation, nitrification and denitrification (Herbert 1999; Gutknecht et al. 2006). Nitrogen is a major component of cells and is particularly important for the formation of nucleic acids, amino acids and amino sugars. A major source for the formation of these compounds is the fixation of N2 from the atmosphere by anaerobic processes catalysed by prokaryotes that reduces N2 to NH3 (Figure 7.5). The NH3 is then oxidised (through a number of intermediate steps) to
Nitrogen cycle Atmospheric N2 Symbiotic N2 Fixation (Rhizobium spp., Bradyrhizobium spp., Frankia spp.)
Non-symbiotic N2 Fixation (free living soil bacteria, marine cyanobacteria)
Organic N
Gaseous NO, N2O
Plants
Denitrification (anaerobic bacteria)
Organic N Organic N
Decomposition (heterotrophic bacteria + fungi)
– NO 3
+ NH4
SRf002-08
Figure 7.5 :Biogeochemical nitrogen cycle
Ammonia oxidation (autotrophic ammonia oxidisers)
– NO 2
Nitrification (autotrophic nitrifying bacteria)
133
134
Regolith Science
Sulfur cycle Assimilatory SO4
2-
reduction
2SO 4
Organic S
Decomposition (heterotrophic bacteria)
Dissimilatory sulfate reduction - anaerobic Desulfovibrio spp.
S
Oxidation Acidothiobacillus spp. (i.e. FeS2)
S
2-
Oxidation Thiothrix spp. Dissimilatory S reduction Desulfuromonas spp.
Figure 7.6: Biogeochemical sulfur cycle.
nitrate, which can be assimilated by plants. Within the plants, nitrate is reduced in a process called assimilatory nitrate reduction, and incorporated into complex organic molecules such as amino acids, peptides and proteins (Madigan and Martinko 2006). The most important microorganisms that catalyse N2 fixation in terrestrial environments are symbiotic, N2-fixing bacteria and actinomycetes (that is, Rhizobium spp., Bradyrhizobium spp. and Frankia spp.), that are associated with root nodules of legumes (Jones et al. 2003; Section 8.2.3). A number of free-living soil and regolith bacteria, such as Clostridium spp., Bacillus spp. and Azotobacter spp., are also able to fix atmospheric N2 (Gutknecht et al. 2006). In aquatic systems, free-living (discrete) cyanobacteria carry out most of the N2 fixation (Herbert 1999). During the next stage of the N cycle (nitrification), NH3 is converted to nitrate under oxic conditions (Herbert 1999; Gutknecht et al. 2006). Two groups of bacteria are important for this process: Nitrosomonas spp. and Nitrobacter spp. In a first step, Nitrosomonas spp. convert NH3 to nitrite and, in a second step, Nitrobacter spp. converts nitrite to nitrate, which is assimilated by plants (Herbert 1999; Gutknecht et al. 2006). The final aspect of the N cycle is denitrification, which is performed by a variety of bacteria under anoxic conditions (Herbert 1999; Gutknecht et al. 2006). During denitrification, nitrate is reduced via a number of intermediate products, such as nitrous oxide and NH3, back to N2, which is released into the atmosphere – thus closing the N cycle.
7.3.3 Sulfur cycle Most of the Earth’s S is tied up in rocks, sediments and regolith materials as metal sulfide (particularly pyrite) and sulfate minerals (such as gypsum, jarosite, and barite), some of which are likely to be the product of microbial transformations in the earlier periods of the Earth’s history, for which acid sulfate soils appear to be one contemporary analogue (Nordstrom and Southam 1997; Southam and Saunders 2005; see also Chapter 12). The largest reservoir of S for the biosphere is sulfate dissolved in the Earth’s oceans. Sulfur is also present in the atmosphere, which it enters mostly as H2S and SO2 from volcanic eruptions, microbial processes (such as dissimilative sulfate reduction or decomposition of organic matter) and anthropogenic sources, (such as industrial and transport discharges). Biogeochemical S transformations are more complex than N transformations, because of the variety of oxidation states and organic compounds in which S can exists, and the interrelationship between abiotic and biochemical reactions in the S cycle. The main steps in the biogeochemical S cycle are dissimilatory and assimilatory S and sulfate reduction, and sulfide and elemental S oxidation (Figure 7.6). Dissimilatory sulfate and S reduction are anaerobic processes, in which SO42– or S0 is reduced to H2S. Hydrogen sulfide is toxic because it inhibits the function of Fe-containing cytochromes in organisms (Truong et al. 2006). The speciation of the sulfide in the environment depends on the ambient pH – H2S is dominant below pH 7; HS – and S2– are dominant
Geomicrobiology of the regolith
above pH 7 (Morse et al. 1987). Reduction of sulfate and S is widespread in terrestrial and aquatic environments and mediated by a diverse group of bacteria, such as Desulfovibrio spp., Desulfomonas spp. and Desulfobacter spp. (Nealson and Stahl 1997). These bacteria play a direct role in forming sulfide minerals because H2S is detoxified biologically by precipitation
with metal ions (such as Fe2+) during the formation of metal sulfides (such as pyrite; Figure 7.7a). The process of assimilative sulfate reduction turns SO42– to reduced to sulfhydryl groups (R–SH) – which are components of many organic molecules, such as the amino acids methionine and cysteine – is a common mechanism in many plants, fungi and prokaryotes (Madigan and b
2 µm
1 µm
a c
d
20 µm FRf005-07
2 µm
Figure 7.7: Examples of biomineralisation: (a) Framboidal pyrite, euhedral pyrite, and nano-crystalline ZnS produced by sulfate reducing bacteria in a biofilm on wood (b) Nano-crystalline ZnS formed by sulfate reducing microbes in a biofilm growing on wood (c) Iron-oxide coated twisted filaments and hollow tubes that are characteristic of the neutrophilic Fe-oxidising bacteria Galionella spp. and Leptothrix spp. respectively (d) Microbial colony on the surface of a ferruginous pisolith.
135
136
Regolith Science
Martinko 2006). The decomposition of organic S-compounds – known as desulfurylation – also leads to the production of H2S, and so does S disproportionation, in which thiosulfate (S2O32–) is converted to H2S and SO42– (Madigan and Martinko 2006). Under oxidising conditions, H2S, HS – and S2– are rapidly chemically oxidised, but S-oxidising bacteria, such as Acidothiobacillus spp., Leptospirillum spp. and Beggiatoa spp., are also able to carry out this reaction (Nordstrom and Southam 1997). As a result of the rapid chemical oxidation of H2S, these bacteria often inhabit environments at the oxic/anoxic interface or where stable metal sulfides are present in oxic environments. These include mineralised zones in the regolith, S-rich waste rock piles in mines and acid sulfate soils (Chapter 12; Southam and Saunders 2005). Here the metal sulfides are oxidised – leading to a sharp decrease of ambient pH and the release of associated heavy metals (Ledin and Petersen 1996). Several hundred Fe- and S- oxidising species have now been discovered and many are used to promote the dissolution of various metal sulfides such as sphalerite (Figure 7.7b; Nordstrom and Southam 1997). Iron- and S- oxidising bacteria are today commonly used in industrial bioleaching processes to extract metals from sulfide ores (Krebs et al. 1997). The microbial breakdown of sulfide minerals leads to the release of the metals and complexing ligands, such as thiosulfate, into the environment (Southam and Saunders 2005). Potentially microbial S-cycling could be responsible for the formation of geochemical anomalies found in the regolith above sulfide mineralisation. Furthermore, microbiological reactions involving S have the potential to lead to the numerous environmental problems in acid sulfate soils (ASS) (see Chapter 12), as well as surface soils and sediments containing elevated concentrations of Fe sulfide minerals (Morse et al. 1987).
7.4 MICROORGANISMS, MINERAL WEATHERING AND TRACE METAL SOLUBILISATION IN THE REGOLITH Geomicrobial weathering of minerals and dispersion of trace elements in the regolith depend on:
s s s
the physical and chemical properties of minerals (such as crystal structure, crystal defects, particle size, morphology, surface area) the solution variables (such as element concentrations, speciation, available ligands, redox potential and pH) the composition (population structure) and biochemical activity of indigenous microbiota (that is, prokaryotes, algal plants and fungi; Banfield and Hamers 1997; Barker et al. 1997; Little et al. 1997; Stone 1997).
A number of studies have shown that microbiota accelerate the weathering (see Section 7.5) of metal sulfides, silicates, phosphates, carbonates and clay minerals (Bennet et al. 1996; Barker et al. 1997; Nordstrom and Southam 1997; Garcia-Vallès et al. 2000). These studies have also shown that microbial processes leading to mineral dissolution and trace element dispersion are mediated by free-living microorganisms, as well as organisms living in biofilms on mineral surfaces (Barker et al, 1997; Little et al. 1997). A biofilm is a complex aggregation of microorganisms characterised by mineral surface attachment through a protective and adhesive matrix of exopolymeric substances (mucilages and polysaccharide gels), as well as structural heterogeneity, genetic diversity and complex interactions of the microbial community (Little et al. 1997). Microbial biofilms are hotspots for geomicrobial weathering and have been shown to accelerate the weathering of a wide range of minerals – as well as building materials such as concrete, steel or other metals – promoting the degradation of these surfaces (Little et al. 1997; Beech 2004). Estimates of the magnitude of microbial impact on mineral weathering rates in the regolith range from less than 2 times (Drever and Vance 1994) to orders of magnitude higher compared with abiotic processes (Schwartzman and Volk 1989; Section 7.4.3). Minerals in contact with organic-rich waters in the regolith are particularly liable to extensive microbially mediated dissolution (Hiebert and Bennett 1992; Bennett et al. 1996). Because minerals generally contain ppm to % levels of sorbed or structurally incorporated trace metals (Chapter 5; Appendix 2), accelerated microbial mineral weathering also accelerates the dispersion of
137
Geomicrobiology of the regolith
trace elements (Table 7.1). The microbially mediated transformation of minerals via oxidative attack – coupled with shifts in pH and redox chemistry, and the excretion of organic and inorganic metabolites that serve as complexing ligands – have a profound impact on geochemical behaviour of many trace elements (Figure 7.8; Huang and Schnitzer 1986). Microbial processes influence the solubilisation, distribution, speciation and precipitation of As, Se, Mo, Sn, Sb, Te, Hg, W, Cd, Hg, Pb, U, Ag, Cu and Au species (see Section 7.5) in a range of interlinked redox cycles under a wide range of environmental conditions common in the Australian regolith (Southam and Saunders 2005; Reith et al. 2007). This is of interest to both the mineral explorers and environmental scientists because patterns of mineral distribution and re-distribution resulting from microbial processes are potentially different from abiotic weathering and solubilisation reactions. 7.4.1 Microbially mediated changes of pHand redox conditions Mineral dissolution rates are relatively constant at near neutral pH (Figure 7.8). These rates increase as pH either increases or decreases as a function of the point of zero charge (pzc; see Section 5.3.7) and the degree of protonation/deprotonation of ions at the mineral surface. The most common mechanism by which microorganisms control pH – and thus mineral weathering in the regolith – is by producing carbonic acid as a result of respiration (Barker et al. 1997). 3
(a)
Degradation of organic C in soils and groundwater leads to CO2 abundances that are several orders of magnitude higher than in the atmosphere (Keller and Wood 1993) – and water in equilibrium with these elevated pCO2 levels has a minimum pH around 4.5 (Drever 1994). Although this level of acidity may not affect mineral framework ions (such as Si, Al, and Fe), abundances of more easily leachable ions (such as K, Na and Ca) correlate with pCO2, indicating that elevated pCO2 in soil and water can result in cation exchange with mineral surfaces (Drever 1994). Microorganisms also generate a number of stronger inorganic acids as a result of metabolic processes (see also Section 7.4.3). Chemolithotrophic nitrifying bacteria produce nitric acid from reduced N compounds – leading to substantial changes the geochemistry of basaltic rock surfaces (Lebedeva et al. 1979). Sulfuroxidising bacteria produce sulfuric acid (from the oxidation of sulfide minerals: see Section 7.4.2), which can dramatically accelerate mineral weathering in areas impacted by acid mine drainage or acid sulfate soil (ASS) formation (Chapter 12). In addition, prokaryotes also contribute to indirect mechanisms for producing or consuming acidity: for example, Fe-oxidation reactions mediated by Fe-oxidising microorganisms such as Galionella spp. and Leptothrix spp. initially consume protons; however, subsequent precipitation of Fe oxides generates acidity (ferrolysis; Section 6.2). By influencing the pH, microorganisms not only increase mineral dissolution rates and trace element 3
(b)
Bytownite
-8
Bytownite Labradorite
Al/Si(sol)/Al/Si(min)
Al/Si(sol)/Al/Si(min)
Labradorite 1 mM oxalate
log rate
(c)
2
-9
1
-10
2
1
Inorganic -11
0 2
4
6
8
pH
10
12
0 2
4
6
pH
8
10
2
4
6
pH
8
10 FRf004-07
Figure 7.8: (a) Feldspar dissolution rate mol Si/cm2/s versus pH in 1 mM oxalate and in inorganic solutions. Al/Si ratios of product solutions normalised to Al/Si ratios in plagioclase feldspars in (b) inorganic solutions and (c) complexing organic acids (oxalate, citrate, succinate) (after Welch and Ullman 1993).
138
Regolith Science
dispersion, they also affect mechanisms and stoichiometries of weathering reactions, which vary with changing pH. For example, biotite dissolution at near neutral pH is characterised by leaching of the interlayer K+ ion and expansion of the mineral lattice to form a hydrobiotite or vermiculite structure (Section 4.3). However, under more acidic conditions, biotite dissolution is characterised by a leaching of K+ from the interlayer along with leaching of Fe2+ and Mg2+ from the octahedral sites (Kalinowski and Schweda 1996; Malstrom and Banwart 1997). Feldspar dissolution at neutral pH is characterised by a preferential leaching of cations Na, Ca, and K – as well as Si – leaving a smectite enriched in Al (Section 4.3). However, under mildly acidic conditions (pH less than around 4), cations and Al3+ are preferentially leached – leaving a residual material enriched in Si (Figure 7.8; Welch and Ullman 1993, 1996, 2000). The difference in dissolution stoichiometry can dramatically affect the nature of secondary mineral phases formed in the regolith as a result of mineral alteration. Besides their inf luence on ambient pH, microbiota also control the oxidation–reduction (redox) potential of micro- and macro-sites within the regolith (Southam and Saunders 2005). Thus, they mediate the formation of biogeochemical zones with differing mineral dissolution/formation and element dispersion/re-concentration characteristics; these biogeochemical zones are often linked to the location of the material within the regolith profile. In order to survive, all organisms need to generate metabolic energy via coupled oxidation/ reduction reactions and, to create metabolic energy, organisms transfer electrons along an electron transport chain located in the cytoplasm and the cytoplasmic cell membrane to a dominant external electron acceptor (oxidising agent), such as O2 , nitrate, Fe 3+ or sulfate, which are reduced. The highest net gain of energy is achieved by oxidising reduced C sources (such as glucose) and transferring the electrons to O2 (Madigan and Martinko 2006). Once the O2 is depleted, nitrate, then Mnand Fe-oxides, sulfates and, finally, CO2 are reduced, resulting in progressively lower redox potentials in these zones (Table 7.1; Southam and Saunders 2005). Thus, these coupled oxidation/
reduction reactions exert a strong inf luence over the surrounding geochemical conditions and metal mobility (Table 7.1). In regolith zones close to the surface, soil and ground waters are in contact with the atmosphere and aerobic bacteria dominate (Table 7.1) and are known to increase mineral weathering and trace element dispersion (Southam and Saunders 2005). However, even in these dominantly oxic zones, anoxic microzones with differing geochemical characteristics exist – commonly within the core of soil aggregates that are physically protected from O2 infiltration. For instance, in column experiments with soils and sediments that were considered to be oxic, the O2 saturation in numerous microzones decreased from 100% to zero due to the activity of the resident microbiota and lead to differences in speciation and thus geochemical behaviour of trace elements such as As (van der Lee et al. 1999). Underlying the oxic zone, a progression of anoxic zones with decreasing redox potentials is commonly found (Table 7.1; Lovley and Chapelle 1995). Under moderately reducing conditions (Eh = 0.2 to 0.7), Mn- and Fe-reducing bacteria couple the oxidation of organic matter to the reductive dissolution of high-surface area Mn- and Fe-oxides (Southam and Saunders 2005). In this process, the microorganisms strip electrons from organic C and transfer them to the Fe- and Mn- minerals – thereby reducing Mn4+ to Mn2+ and Fe3+ to Fe2+. Thus, Fe and Mn, plus sorbed trace elements, such as Co, Ni, V, Ba, REE, Au and As (see Chapters 5 and 10; Appendix 2), are released into regolith waters (Saunders et al. 1997). As minerals containing Mn4+ and Fe3+ become depleted in the system at Eh <0.2, microbial activity may lead to the formation of metallic sulfides (Table 7.1; Section 7.6; Southam and Saunders 2005). 7.4.2 Oxidative breakdown of sulfide minerals The direct microbial oxidation of a number of sulfide minerals by Fe- and S-oxidising prokaryotes drastically increases the mobility and reactivity of associated metals such as Fe, Mn, Co, Ni, Zn, Pb and Hg (Nealson and Stahl 1997; Lovley and Chapelle 1995). The most relevant sulfide oxidation process in the regolith is the oxidation of Fe and S from pyrite, which
Geomicrobiology of the regolith
leads to the production of acidity (see Equation 7.3 below) and the mobilisation of trace metals in the resulting acid sulfate soils (ASS; see Chapter 12) or acid mine drainage (Nordstrom and Southam 1997). The main pathway for pyrite dissolution features the oxidation of Fe2+ by O2 according the following reaction (see also Section 5.3): 14Fe 2 + + 3.5O 2 + 14H + = 14Fe 3 + + 7H 2 O (Eqn 7.1) This is followed by reduction of Fe2+ by sulfide: FeS 2 + 14Fe 3 + + 8H 2 O = 15Fe 2 + + 2SO 24 - + 16H + (Eqn 7.2) resulting in an overall reaction of pyrite dissolution: FeS 2 + 3.5O 2 + H 2 O = Fe 2 + + 2SO 24 - + 2H + (Eqn 7.3) The rate-limiting step (Equation 7.2) is catalysed and accelerated by up to 1000 times by Fe- and S-oxidising bacteria (Nordstrom and Southam 1997). However, the exact processes involved have only recently been described by Sand et al. (2001) and Rawlings (2002). Bacteria, such as Acidithiobacillus ferrooxidans or Leptospirillum ferrooxidans, have a strong affinity for mineral surfaces, such as pyrite, to which they rapidly attach by forming a biofilm. A layer of extracellular polymeric substances (EPS) that serves as a reaction space is produced by these bacteria when they attach to a mineral and form biofilms (Sand et al. 2001). Biofilms on pyrite contain high concentrations of protons and dissolved and chelated Fe-species, which are derived from the bacterial oxidation of Fe2+ to Fe3+ (Sand et al. 2001; Rawlings 2002). The high concentration of Fe3+ and protons mounts the attack on the valence bonds of the pyrite, which is then degraded via intermediate thiosulfate (Sand et al. 2001). Fe3+ is reduced to Fe2+ in the process and then re-oxidised to Fe3+ by Fe-oxidising bacteria. Oxidation of other metal sulfides, such as galena or sphalerite, proceeds via different intermediates, such as polysulfides and elemental S (Sand et al. 2001). The general functions of Fe- and S-oxidising bacteria in the solubilisation of metal sulfides are to:
1. provide sulfuric acid for a proton hydrolysis attack 2. provide the reactive environment for the chemical reactions to take place 3. keep Fe or other metals in the oxidised state. 7.4.3 Effect of organic ligands on mineral dissolution and trace element mobilisation All microorganisms produce and excrete organic or inorganic (see Section 7.4.1.) substances that affect mineral weathering reactions due to their acidity, and may also act as lixiviants for trace metals due to their complex-forming capabilities. Thus, most microbially mediated rock weathering (other than the weathering of sulfides) involves the excretion of metabolites that corrode minerals through chemical interaction (Barker et al. 1997; Stone 1997). Microbes produce and excrete low molecular weight organic acids (LMWOAs), such as fermentation products (lactic, formic, oxalic, acetic and succinic acids) and Krebs cycle compounds (citrate, succinate, a-ketoglutarate, oxalacetate and pyruvate) that may have profound effects on mineral weathering reactions (Barker et al. 1997; Stone 1997). Laboratory studies with minerals, rocks or soils have shown that LMWOAs excreted by microorganisms increase mineral dissolution and trace metal mobilisation rates by 2 to more than 100-fold (Huang and Kiang 1972; Figure 7.8). Vandevivere et al. (1994) measured a 200-fold increase in Si release rate from a Ca-rich plagioclase in an organic-acid-producing bacterial culture compared with an abiotic control (see also Section 4.3). Similarly, Berthelin (1971) measured a 200-fold increase in Al concentration from dissolving granitic sand in an acid-producing microbial culture (pH around 3) compared with an abiotic control at the same pH. Little (2007) showed that microbially produced organic acids increased dissolution of major elements from bulk and rhizosphere soils 2 to 10 times and trace elements 10 to 1000 times higher than control experiments run under the same conditions (see also Chapter 8). Thus, microbial generation of organic ligands can increase mineral dissolution rates – even in the absence of significant pH changes in the bulk solution (Stone 1997). In a series of studies, Welch and Ullman (1993, 1996,
139
140
Regolith Science
1999 and 2000) have shown that LMWOAs catalyse the reaction by forming complexes with ions from minerals – either at the mineral surfaces, which weakens metal-oxygen bonds, or in solution, which lowers solution saturation state. In general, difunctional organic ligands that form bidentate complexes with Al and Fe have a much larger effect than monofunctional ligands such as acetate or formate (Stone 1997). The relative impact of organic acids is greatest at pH around 4 to 6, where the inorganic mineral dissolution rates are lowest and there are deprotonated carboxylic groups to interact with ions from minerals. High-molecular-weight organic compounds (such as microbial extracellular slime moulds) can also affect mineral weathering by complexing with ions from minerals very much like the low-molecular-weight organic compounds. In addition to organic acids and slime moulds, microorganisms also excrete a number of elementspecific soluble organic-complexing agents in response to nutrient stress. Of particular interest in regolith environments is the release of highly Fe3+ -specific bidentate ligands (siderophores) by bacteria and fungi (Kraemer et al. 2005). The major siderophoric molecules (that is, hydroxamates, phenolates and catecholates) form highly stabile complexes with Fe3+ compared with low-molecular-weight chelators (Kalinowski et al. 2000). When the microbes take up siderophore- Fe3+ complexes, Fe3+ is reduced to Fe2+ and released into the cell. Recent experiments have shown a significant enhancement of siderophore-promoted dissolution of insoluble Fe- and Al-oxides when compared with LMWOAs (Watteau and Berthelin, 1994), while other batch-reaction studies suggest that the two-ligand system – siderophores and oxalic acid – is an effective mediator for microbial induced Fe3+ -oxide reduction (Cheah et al. 2003). 7.4.4 Other microbial transformation processes In addition to direct influences on mineral weathering and trace element dispersion, microorganisms also contribute to mineral weathering reactions indirectly by 1. affecting hydration at the mineral–cell interface
2. taking up ions (P, Mg, K, Fe) released from minerals thereby changing solution saturation state, 3. serving as nucleation sites for the formation of secondary mineral phases 4. stabilising regolith materials and thereby increasing reaction time for ‘abiotic alteration’. Another microbial process that may increase the dispersion of trace elements in regolith is the biomethylation of metals and metalloids that leads to their volatilisation (Aspandiar et al. 2006). In the process of biomethylation – which is mediated by a number of prokaryote and fungal species – a metal-carbon bond is established by combining a methyl group (-CH3) with a metal or metalloid (Craig et al. 2003). Since identifying biomethylated mercury (CH3-Hg; Choi et al. 1994), biomethylated volatile compounds of several other trace elements (that is, Co, As, Se, Sb, Te, I and Pb) have been documented under laboratory conditions, and in natural and anthropogenic environments that contain high concentrations of these elements: for example, geothermal environments, swamps, sewage plants and landfill sites. Biomethylation of several orerelated metals, such as As, Hg, Se, Sb and I, occurs in the regolith, but the total concentrations and kinetics of formation, persistence and retention – and thus the influence of biomethylation on the element dispersion in these uncontaminated environments and around weathering ore bodies – are little understood (Aspandiar et al. 2006). Volatile biomethylated compounds in the regolith may be rapidly transferred back to their solid inorganic state, and thus immobilised, giving rise to only limited element dispersion or, depending on the element, the opposite may occur.
7.5 MECHANISMS OF MICROBIAL MINERAL FORMATION AND TRACE ELEMENT PRECIPITATION Microorganisms are capable of directly and indirectly catalysing the formation of secondary minerals by using a number of different active and passive mechanisms that are generally referred to as biomineralisation (Frankel and Bazylinski 2003; Bazylinski and Frankel 2003). Microbes been shown to mediate the formation of:
Geomicrobiology of the regolith
s s s
s s s
carbonates, such as calcite, aragonite, siderite and vaterite (Castanier et al. 1999a, b; Rivadeneyra et al. 2006) sulfides, such as pyrite, greigite, sphalerite and chalcopyrite (Figure 7.7a, b; Lichtner and Biino 1992; Druschel et al. 2002) Fe- and Mn-oxides, such as goethite, magnetite and todorokite (Figure 7.7c, d; Frankel and Bazylinski 2003; Bazylinski and Frankel 2003; Châtellier et al. 2004) amorphous silicates phosphates, such as vivianite and amorphous Ca phosphate (Weiner and Dove 2003) clay minerals (Tuck et al. 2006).
These biominerals often display mineralogy, chemistry, isotope composition or morphology that is different to the products of abiotic mineral formation (Weiner and Dove 2003). Microbially mediated biomineralisation has a profound impact on environmental geochemistry of major and trace elements and may also be important in the formation of a number of ore deposit systems (Southam and Saunders 2005). The ability of microorganisms to form bio-minerals – combined with their high metabolic rates, microscopic sizes (and thus high surface-area-to-volume ratios), and negative surface charges on cell walls and exopolymeric substances, – enables them to rapidly adsorb and accumulate a variety of metals (Frankel and Bazylinski 2003; Bazylinski and Frankel 2003). It has been demonstrated that microorganisms can precipitate free or complexed As, Se, Mo, Sn, Sb, Te, Hg, W, Cd, Hg, Pb, U, Ag, Cu and Au ions from surrounding solutions (Beveridge 1989; Southam and Saunders 2005). Thus, they are likely to contribute to the precipitation of trace metals in the regolith as geochemical anomalies or even secondary mineralisation. 7.5.1 Indirect mechanisms of biomineralisation Indirect mechanisms of biomineralisation are characterised by the ability of microorganisms to modify their local micro-environment by creating physicochemical conditions (such as pH and redox potential) suitable for the purely chemical precipitation of min-
erals to occur. One of the most relevant examples for the regolith geosciences is the biologically induced sulfate reduction followed by the chemical precipitation of metal sulfides (see Section 7.7.1). Although sulfate reduction in the presence of organic matter is thermodynamically favourable at low temperature, it is kinetically inhibited under abiotic conditions, and thus is the rate limiting step for metal sulfide formation. In the presence of dissimilatory sulfate-reducing bacteria (SRBs), sulfate is readily reduced to sulfide, which is then free to react with metal cations to form metal sulfide minerals (Figure 7.7; Jong and Parry 2003). Another example of an indirect microbial mechanism that leads to the formation of important regolith minerals may be the decomposition of urea followed by an increase in ambient pH and precipitation of carbonate minerals such as calcite and vaterite (Figure 7.9; Section 7.7.2). 7.5.2 General principles of direct mineral precipitation Direct mineral precipitation depends not only on the activity of microorganisms to maintain physiochemical condition conducive for mineral precipitation, but also on the direct interaction of microbial cell membranes with free or complexed ions. Direct microbial precipitation of minerals and metals in the regolith can be divided into two categories: intracellular, microbially controlled mineralisation, such as magnetite and Au precipitation by magnetotactic bacteria (Frankel and Bazylinski 2003, Bazylinski and Frankel 2003), and extracellular, bacterially induced mineralisation (BIM), which occurs outside microbial cells by the interaction of metals and/or minerals with cell surfaces (Beveridge 1989). This latter process may be more important for the formation of a wide range of major secondary minerals, such as Fe- and Mn-oxides, sulfides and carbonates. This may be because extracellular biomineralisation it is not limited by space constraints inside the cell and it can be accelerated by microbial metabolism (Frankel and Bazylinski 2003). Extracellular biomineralisation can be divided into passive and active processes (Fortin and Beveridge 2000; Southam 2000). Passive extracellular biomineralisation refers to simple non-specific binding of
141
142
Regolith Science
cations and recruitment of solution anions – resulting in surface nucleation and growth of minerals – and depends on the ionic charges present on the surface of the organism, which is largely determined by the structure and composition of the microbial cell wall (Frankel and Bazylinski 2003). Negative charges on most cell and exopolymer surfaces can result in binding of cations by non-specific electrostatic interactions – effectively contributing to local supersaturation. Binding also helps to stabilise surfaces of nascent mineral particles – decreasing the free energy barrier for critical, crystal-nucleus formation. Active mineralisation occurs by the direct and element selective redox transformation of surface-bound metal ions, or by the formation of cationic or anionic by-products of metabolic activities that form minerals on the bacterial surfaces (Frankel and Bazylinski 2003). Extracellular biomineralisation often leads to the formation of microfossils or pseudomorphs as discussed by Southam and Saunders (2005; Figure 7.10): Mineral precipitation covers and reduces the available cell surface area for nutrient uptake. As a consequence, covered cells are starved and die, and their proton motive force dissipates. Thus, mineralisation proceeds to completion and a fossil is formed. Once metals are immobilised on bacterial cell envelopes, organic compounds in cells are less likely to be degraded – thereby preserving the cell structure. These metal precipitates are generally hydrous, amorphous aggregates, but they can become more crystalline with time (that is, by dissolution–re-precipitation or solid-state transformation: Banfield et al. 2000). In many cases, precipitation of minerals on bacteria represents the earliest stage in authigenic mineral formation and has been observed by several authors (for example, Watterson 1992; Reith et al. 2006). 7.5.3 Microbially induced precipitation of secondary Au in the regolith Microbially controlled mineralisation, as well as bacterially induced mineralisation (BIM), may play an important role in the mineralisation of Au in the regolith. Recent research suggests that biomineralisation of Au may have led to the formation of secondary octahedral Au as well as secondary Au grains in the regolith (see Section 7.7.2; Figure 7.10), and may have
contributed to the formation of the Witwatersrand paleo-placer Au deposit in South Africa (Mossman et al. 1999; Reith et al. 2007). Secondary Au in deposits within intensely weathered Archean rock from Western Australia is commonly present as octahedral Au particles (Wilson 1984; Freyssinet and Butt 1988). The geochemical, abiotic formation of these particles through mobilisation of the original detrital, dispersed Au requires temperatures between 240°C and 300°C and pressures of 2–3 × 105 kPa for more than 1000 years (Frimmel et al. 1993). However, in a number of studies with Bacillus subtilis, Southam and Beveridge (1994, 1996) have shown that secondary octahedral Au – similar in morphology to the Western Australian samples – may be the product of rapid biomineralisation at temperatures between 60°C to 90°C. In these studies, B. subtilis deposited more than 100 µg/g (dry weight bacteria) Au as fine-grained intracellular colloids (5–50 nm: 0.005–0.05 µm). Eventually, autolysis of cells was initialised, proteins were released and pseudo-crystalline Au was formed, which was transformed into crystalline octahedral Au. In recent research by Lengke et al. (2006a, b, c) focussed on the mechanisms of Au3+ -chloride and Au+ -thiosulfate precipitation by other environmentally relevant bacteria. They show that the reduction mechanism of Au3+ -chloride to elemental gold by the cyanobacterium Plectonema boryanum involved the formation of an intermediate Au+ -species, Au+ sulfide – with S originating from cyanobacteria proteins (probably cysteine or methionine). In another study, Lengke and Southam (2005) assessed the precipitation of Au thiosulfate complexes (Au(S2O3)23–) by an A. ferrooxidans isolated from deep Au mines. They found that the (Au(S2O3)23–) complex was stable in the bacterial systems until the oxidation of free thiosulfate was complete, after which the bacteria oxidised thiosulfate bound in Au(S2O3)23– complexes – leading to the precipitation of native Au. The Au was precipitated inside the bacterial cells as fine-grained colloids, ranging between 5 and 10 nm (0.005–0.01 µm) in diameter, and in the bulk fluid phase as crystalline µm-scale octahedral Au. The presence of Au nanoparticles within, and immediately surrounding, the bacterial cell envelope highlights the presence of localised
Geomicrobiology of the regolith
reducing conditions produced by the bacterial electron transport chain via energy generating reactions within the cell. Several other studies have also shown that prokaryotes are able to actively and passively precipitate Au. The accumulation of complexed and colloidal Au was investigated by Karamushka et al. (1987a, b) and Ulberg et al. (1992). They found that the accumulation of Au by a strain of B. cereus was dependent on the chemical structure of the cell envelopes, and involved functional groups of proteins and carbohydrates. In addition, they discovered that the accumulation of Au was directly dependent on the metabolic activity of the cells and especially on metabolic reactions on the plasma membrane – in particular the hydrolysis of ATP (adenosine triphosphate) by the enzyme ATPase. In a study with 30 different microorganisms, Nakajima (2003) found that the bacteria were more efficient in removing Au3+ - and Au+ -complexes from solution than other groups of microorganisms. These results were confirmed by Tsuruta (2004), who studied Au3+ chloride bioaccumulation in 75 different species, and found that Gram-negative bacteria – such as Acinetobacter calcoaceticus and P. aeruginosa – had the highest ability to accumulate Au. Energy-dependent uptake of Au has also been shown for other bacteria, such as Spirulina platensis (Savvaidis et al. 1998). Addition of an Au3+chloride solution to cell suspensions of Geobacter metallireducens oxidised c-type cytochromes, which are thought to be involved in electron transfer to metals (Kashefi et al. 2001). Further experiments conducted with mesothermophilic and hyperthermophilic dissimilatory Fe3+ -reducing bacteria and archaea demonstrated that some of these organisms – such as Pyrobaculum islandicum, Shewanella algae, Desulfuromonas vulgaris and Geovibrio ferrireducens – are capable of precipitating Au by reducing Au3+ -complexes to Au0 (Kashefi et al. 2001). The origin of coarse secondary Au grains in the regolith has long been the subject of discussion among geologists. Three models have been established to explain their formation: detrital origin, chemical accretion and a combination of both (Boyle 1979). When Watterson (1992) first reported structures resembling gold encrusted microfossils on placer gold specimens from Lillian Creek in Alaska, a biological
mechanism for their formation was considered. Goldencrusted cell-like microfossils, or bacterial pseudomorphs, detected on Lillian Creek Au grains consisted of branching and anastomosing, axial and lateral direct budding cells, which were oval, spheroidal or kidney-shaped, ranged in size from 0.9 to 1.5 µm, and morphologically resembled Pedomicrobium manganicum cells (Figure 7.10a, b). After studying 18 000 Au grains from different sites in Alaska, and observing similar lacelike networks of µm-size filiform Au on the majority of these grains, he concluded that these structures were of bacterial origin. Subsequently, numerous specimens from different locations in Australia – such as Watts Gully in South Australia (Keeling 1993) and the Palmer River Goldfields in northern Queensland (Bischoff 1997; Reith et al. 2006) – have been studied and Pedomicrobium-like pseudomorphs were detected in most crevices (>70%) of secondary Au grains that showed little or no signs of transport. However, after producing analogous structures abiotically with natural and artificial Au amalgams, using hot HNO3 dissolution, Watterson (1994) concluded that the observed morphologies alone could not be considered adequate evidence of microbial origin of these Au grains. Nevertheless, a recent CRC LEME study with the metallophilic bacterium Cupriavidus (formerly Ralstonia) metallidurans has shown that the bacterium is associated with secondary Au grains has the ability to precipitate Au (Figure 7.10b; Section 7.7.2.) and may thus provide a critical link the microbial Au precipitation observed in the laboratory and bacterial pseudomorphs observed on natural secondary Au grains present in the regolith (Reith et al. 2006).
7.6
ISOTOPIC BIO-SIGNATURES
Although it is difficult to constrain the direct impact of microbial metabolic processes on mineral weathering and formation in the regolith, biological processes often produce characteristic ‘bio-signatures’ by forming mineral phases that have a distinctive geochemistry, mineralogy or isotopic composition (Weiner and Dove 2003). Biologically mediated reactions tend to favour the lighter isotopes, so that ‘light’ isotopic ratios for elements (such as H, C, O, N, S,
143
144
Regolith Science
and even heavier elements such as Fe, Zn and Cu) reflect biologically mediated reactions (Beard et al. 2003; Zhu et al. 2002). For example, the S isotopic composition of sulfides reflects fractionation due to selective uptake of isotopically light S in sulfate into cells for use as an electron acceptor during sulfate reduction (Southam and Saunders 2005). Therefore, the low temperature biologically mediated formation of metal sulfides produce mineral phases with a distinctive isotopic composition. Although there is little S fractionation associated with oxidation of sulfide minerals, microbially mediated pyrite oxidation – via oxidation of Fe2+ to Fe3+ and reaction of Fe3+ with pyrite surfaces – will produce SO42– that has an oxygen isotopic signature that is distinct from that formed by pyrite oxidation by oxygen (Canfield 1998, 2001). Another example of how isotopic analyses can be used to elucidate mechanisms of mineral formation in the regolith is a study on carbon isotopic fractionation during the formation of regolith carbonates. Regolith carbonates (calcretes) are commonly used during Au exploration in arid parts of Australia and biotic or biomediated carbonate precipitation in the environment is very common (see Section 7.7.2 and Chapter 13). However, it is not known to what extent, regolith carbonates are formed as a result of metabolic or purely chemical processes (Figure 7.9; Lintern et al. 2006; Schmidt Mumm and Reith 2007). If carbonates are the product of biological activity isotopic bio-signatures may reveal their origin: A large fractionation Amino acids Organic matter
Gold-AA complexes
Metabolic breakdown of amino acids
Destabilisation of Au-AA
Release of urea Urease degradation of urea Release of NH3 and CO2 NH4+ and HCO3-
Ca 2+
[Au]
Co-precipitation
Ca2+ + HCO3 -+ OH - = CaCO3 + H2O
GOLD
KSf006-08
Figure 7.9: Model for the microbially mediated formation of Au-in-calcrete anomalies (from Schmidt Mumm and Reith 2007).
of C isotopes in the regolith is associated with uptake of CO2 and assimilation of C into plant biomass (Madigan and Martinko 2006). The d13C of plant biomass ranges from around –10 to –30‰, which reflects the variation associated with C3 and C4 plants (Appendix 1; Chapter 8; Madigan and Martinko 2006). Degradation of plant organic matter, with the production of organic acids and, ultimately CO2, from respiration of plant biomass, will produce carbonate ions that reflect their biogenic origin. In contrast, little fractionation of C will associated with abiotic transformation of CO2 species (purely chemical formation and dissolution of carbonates).
7.7
CASE STUDIES
7.7.1 Sulfur biogeochemistry in acid sulfate soils Previous sections have shown the importance of regolith microbiota in mediating biogeochemical cycles and mineral transformations. Quantification of microbially mediated reactions is especially important in environments where the rates of biochemical transformations surpass purely chemical reaction rates. Acid sulfate soil environments are formed by natural accumulation of bacterially formed Fe sulfides, predominantly pyrite, in estuarine environments (Berner, 1984; Morse et al. 1987). Acid sulfate soils (ASS) are abundant in lowland coastal areas around the world, where sea-level regression, and subsequent landscape change, culminates in burial of high sulfide-bearing estuarine sediments. Australia is characterised by extensive development of ASS (1.2 × 106 ha) along much of the coastline, as well as inland (see Chapter 12). These developments become problematic when changes in soil water regimes, resulting from increasing utilisation and development of coastal soils, lead to widespread oxidation of the metal sulfides (White et al. 1997; Chapter 12). The products of the oxidation process, such as sulfuric acid and mobilised metals, instigate a host of environmental and economic problems that include loss of aquatic populations and habitats, decreased soil productivity and loss of infrastructure (Sammut and White 1996). The role of bacteria in oxidising sulfide minerals. and subsequent production of sulfuric acid, in the
Geomicrobiology of the regolith
oxidation of pyrite – as well as in reducing sulfate and producing sulfide minerals – are well studied under laboratory and field conditions, such as in acid mine drainage situations (Evangelou 1995; Tyson et al. 2004). In contrast, the diversity and role of S-oxidising and -reducing prokaryotes in ASS environments (Chapter 12) has remained largely unstudied – partly because quantification of microbial diversity and function in environmental samples was elusive due to lack of suitable tools. However, the latest developments in molecular biosciences have fundamentally altered the science of environmental microbiology (see Section 7.8; Barns and Nierzwicki-Bauer 1997). In particular, advances in the use of 16S rDNA phylogenetic techniques have facilitated detection and characterisation of multiple species in complex environmental samples – and the increasing availability of 16S rDNA sequence data on international databases is allowing for the rapid increase in the sensitivity of this technique to describe previously unknown organisms (Pace 1997). Recently Dürr et al. (2006) studied microbial diversity and function in ASS within the Tweed River catchment in northern New South Wales. Results show that the diverse bacterial microflora present at sites constitute the main control on the S and Fe cycling. The ASS examined exhibit typical coastal ASS profile morphology, with oxidised actual ASS (pH <3.5) overlying reduced potential ASS (pH around 7). At the study sites, the oxidation of reduced S compounds – such as sulfide (S2–), hydrogen sulfide (H2S) and thiosulfate (S2O32–) – is carried out by a range of phylogenetically diverse bacteria, including chemolithotrophic (such as Acidithiobacillus spp.), phototrophic (Allochromatium spp., Chlorobium spp., Rhodobacter spp. and Rhodovulum spp.), and heterotrophic (Rhodovulum spp., Rhodocyclus spp., Rhodopseudomonas spp.) bacteria. The majority of these S-oxidising bacteria are chemoautotrophs, but some organisms also use reduced organic C sources, which supports heteroorganotrophic growth (Madigan and Martinko 2006). A number of S-oxidising bacteria are metabolically so flexible that they conduct S oxidation in oxic conditions, but operate as sulfate reducers in anoxic conditions. The bacterial oxidation of inorganic S in the oxidised zone occurs through the S oxidase pathway
(Sox). The Sox pathway is the most widespread and best characterised of the bacterial S oxidation pathways, and bacterial genera – including representatives of the chemolithotrophic, chemoautotrophic and phototrophic groups – use this pathway (Friedrich et al. 2005). The Sox pathway is responsible for the oxidation of hydrogen sulfide, sulfide, elemental S, sulfite and thiosulfate ( Friedrich et al. 2005). The high concentration of the Sox genes within the top layers of the ASS profiles suggests an important role in the oxidation of secondary oxidation products – including jarosite (see Section 4.5 and Chapter 12) – at the sites. Jarosite is considered a store of acidity: slowly releasing sulfuric acid into the soil profile. The activity of S-oxidising bacteria in the top layers of the ASS profile may also contribute to: 1. the ‘second pulse effect’ observed proceeding rainfall events, where oxidation products are mobilised up the soil profile by a rise in the water table and flushed out of the system by a second rainfall event 2. the increase in acidity through further oxidation of S minerals and thus mobilisation of metals and toxicity of the leachate. The reduction of sulfate (SO42–) and elemental S (S0) in anoxic zones of the profiles is carried out by sulfate-reducing bacteria (SRB). SRB are a large and phylogenetically diverse group that include approximately 20 genera – among them are d-Proteobacteria, Gram-positive bacteria and thermophilic bacteria (Madigan and Martinko 2006). SRB are found in the reducing part of the acid sulfate soil profiles and play a key role in the cycling of organic matter in this environment. SRB use the dissimilatory sulfite reductase (Dsr) pathway for the reduction of various S compounds. The Dsr pathway is endemic to the dissimilative sulfate-reducers – present in all known SRB and has been used to identify dissimilative sulfate-reducing organisms in a number of studies (Chang et al. 2001; Joulian et al. 2001; Perez-Jimenez et al. 2001). The Dsr gene was also detected in the oxidised zone in resident phototrophic green and purple S bacteria (Figure 7.11). The presence of S-oxidation and -reduction pathways indicates an abbreviated S cycle within the oxidised zone, which is possibly associated with
145
146
Regolith Science
anoxic microzones within the oxic soils. This discovery may have a major influence on land- management practices in the area, because it suggests the potential for re-reduction of the oxidised ASS by controlled water table management including re-flooding. Thus these now degraded areas (previously used for agriculture) may – with the use of sophisticated remediation procedures – be used again (see Chapter 12). Measurement of the Dsr gene thus provides a potential tool for assessing the remediation potential of highly degraded ASS. 7.7.2 Geomicrobial cycling of Au and formation of Au- anomalous calcrete CRC LEME research (Reith et al. 2005, 2006, 2007; Reith and McPhail 2006, 2007; Schmidt Mumm and Reith 2007) has shown that microbiota resident in auriferous Australian soils and other regolith materials may be involved in the biogeochemical cycle of Au – from its solubilisation to the formation of secondary Au grains and Au- anomalous carbonate (calcrete) (Figures 7.9, 7.10 and 7.12) The indigenous microbiota in biologically active microcosms within auriferous soil are capable of solubilising up to 80 wt.% of the Au contained in these materials during the first 50 days of incubation, after which the solubilised Au was readsorbed by mineral- and organic soil fractions. In contrast, no Au was solubilised in sterilised microcosms incubated under otherwise identical conditions. Molecular (PCR-DGGE, cloning and sequencing of 16s rDNA) and physiological profiling (CLPP) of bacterial communities during the incubation of the microcosms, combined with amino acids analyses, indicated that changes in the structure of the bacterial community from carbohydrate- to amino acid-utilising populations occurred concurrently with, and appear to be linked to, the observed solubilisation and re-precipitation of Au (Reith and McPhail 2006, 2007). These results suggest the following model of Au solubilisation and re-precipitation in the soil microcosms:
s
s
The bacterial community in the early stages of incubation apparently produced surplus amino acids, which are known to directly solubilise native Au and stabilise it in solution (Korobushkina et al. 1983) Some of the detected amino acids, such as glycine, are also precursors for microbially produced cya-
s
nide that is known to form stable Au complexes (Korobushkina et al. 1983). The bacterial community in the later stages of incubation appears to have used these ligands – as a result, the Au complexes were destabilised and Au was re-precipitated in the soil.
Molecular profiling allowed the differentiation of bacterial communities from auriferous and adjacent non-auriferous soils at the Tomakin Park Au Mine (Reith and Rogers 2008). These results – in combination with results of Bacillus cereus spore counts, which were up to 1000 times higher in soils that displayed Au concentrations of 150 to 1000 ppb compared with soils with background Au concentrations, and microcosms amended with dissolved AuCl4 – – suggest that the presence of highly mobile Au in the regolith, as observed at many Australian sites, may influence the composition of the resident microbiota (Reith et al. 2005). The results suggest that a geomicrobial exploration technique, in which B. cereus spore counts are measured and used as a pre-screening method to target areas for further sampling and more complete geochemical analysis, could be developed (Reith et al. 2005; Reith and Rogers 2007). Scanning electron microscopy revealed bacterial pseudomorphs on untreated secondary Au grains from two field sites used in this study (Figure 7.10a, b). The presence of active bacterial biofilms on the surface of Au grains was confirmed using confocal stereo laser microscopy combined with nucleic acid staining (Reith et al. 2006). Molecular profiling showed that unique, site-specific bacterial communities are associated with these Au grains, which differed from those dominating the surrounding soils. 16S ribosomal DNA clones belonging to the genus Ralstonia – and bearing 99% similarity to C. metallidurans – were present on all 16S rDNA-positive Au grains from both locations, but were not detected in the surrounding soils. The ability of C. metallidurans to actively accumulate Au from solution was successfully tested, suggesting that C. metallidurans may contribute to the formation of secondary Au grains in the regolith (Figure 7.10c, d; Reith et al. 2006). An explanation for the presence of active precipitation of Au3+/Au+ -complexes in C. metallidurans and other bacteria may lie in the toxicity of these complexes to the organisms (Karthikeyan and Beveridge 2002).
Geomicrobiology of the regolith
a
B b
c
d
Figure 7.10: Bacterioform Au and ‘golden’ bacteria: (a) Bacterioform Au on secondary Au grain from the Hit or Miss Gold Mine at the Palmer River, Qld, Australia (b) Secondary electron micrograph of metallic Au associated with a C. metallidurans cell (centre) (c) Detailed view of branching network of rounded and oval budding cell-like structure with apparently preserved cell wall structures (d) Morphology (left) and distribution of Au (right) in a cell cluster resembling the bacterioform structures detected on secondary Au grains.
Even at low concentrations, Au3+/Au+ -complexes can cleave peptide and protein disulphide bonds, disrupt cell walls and membranes, and cause cell death (Witkiewicz and Shaw 1981). Thus, for microorganisms, reducing the toxic complexes and accumulating the native metal forms an effective mechanism to avoid Au toxicity and detoxify their immediate environment (Karthikeyan and Beveridge 2002). Regolith carbonate (calcrete) is an important sampling medium for geochemical exploration in Australia (Lintern and Butt 1993; Chapter 13). Most current models for calcrete formation in semi-arid and arid environments in Australia do not include a microbial component (Lintern et al. 2006). The genesis of pedogenic carbonates may be ascribed to abiotic and biologically mediated precipitation – and
Topsoil
Sox B Gene
Actual acid sulfate soil Oxidised zone Transition zone Potential acid sulfate soil Reduced zone DSR Gene Figure 7.11: Sulfur cycling in acid sulfate soils (ASS), showing the distribution of the Sox and Dsr (S-oxidising and -reducing genes) in the different zones
147
148
Regolith Science
Secondary Gold Grains in Regolith
Microbial Gold Precipitation
Decomposition of Plant Tissue and Re-mobilisation of Gold
Incorporation into Plant Tissue
Stabilisation and Transport of Gold in the Liquid Phase
Microbial Gold Solublisation
Submicroscopic Gold in Primary Mineralisation or Adsorbed to Regolith Phases
Figure 7.12: Geomicrobiological cycle of Au in the Australian regolith
both mechanisms have been observed in the natural environments. Mass balance estimates by Castanier et al. (1999a, b) suggest that the heterotrophic microbially mediated carbonatogenesis accounts for the most of the formation of these carbonates, rather than abiotic processes. One of the most common heterotrophic processes leading to the formation of carbonate is the degradation of urea (ureolysis). Urea is a product of the microbial breakdown of purines and amino acids by microorganisms (Vogels and Drift 1976; Cunin et al. 1986). Recent CRC LEME research (Schmidt Mumm and Reith 2007) demonstrates that microbial ureolytic communities, which are resident in the carbonaceous sands and calcrete, are capable of mediating the formation of gold anomalous carbonates. That study investigated the formation of Au anomalies in calcrete (2.5 to 50 ppb) and microbial degradation of urea in calcrete in aeolian sand dunes (2–4 m thick) overlying Au mineralisation at the Barns Au-in-calcrete anomaly in South Australia. The results of their microbial study suggested that the genesis of calcrete may be biomediated through the microbially mediated urea decomposition, which occurred in all samples in a depth profile and pro-
vided a pH and pCO2 environment conducive to carbonate precipitation. Schmidt Mumm and Reith (2007) proposed a coupled model of biomediated and inorganic mechanisms that control Au and calcrete precipitation as shown in Figure 7.9. If the amino acid breakdown can be applied to those involved in Au -complexation, the process would have a destabilising effect on Au-amino acid complexes (Reith and McPhail 2006). The urea released during the break down of these amino acids is transformed by urease to CO2 and NH3 – establishing pH and pCO2 (or aCO32–) conditions conducive to biologically mediated carbonate precipitation in the calcrete-bearing environments. Together with the destabilised Au- amino acid complexes, this process potentially provides a tight link between carbonate formation and related Au enrichment in calcrete. Additional experiments have been conducted that support this model. Microbial enrichment cultures obtained from the calcareous materials from three depths (10, 64 and 210 cm) overlying the mineralisation at Barns were incubated in a urea- and CaCl 2containing microbial growth media (pH 8) amended with 100 ppb of a gold as an Au–aspartic acid complex. Within 96 to 240 h, the urea was changed to NH4 +, the pH in solution rose by approximately 1 unit to pH 9 and Ca 2+ aq was precipitated as Ca carbonate crystals. Gold was uniformly dispersed in these carbonates and enriched up to 800 times compared with the microbial growth medium. The results of the microbial fingerprinting analyses showed that natural microbial communities at the three depths, as well as communities in the enrichment cultures, are dominated by alkaliphylic, halotolerant Bacillus spp.: 82 of 86 clones sequenced from these samples at three depths belonged to the genus Bacillus. Bacillus spp. are known for their metabolic versatility, which includes the utilisation of many organic substances derived from plant or animal sources, ureolysis, N2-fixation and facultative lithotrophy (Claus and Berkeley 1986). Bacillus spp. are also capable of forming endospores, which are dormant, non-reproductive structures produced when a bacterium detects that environmental conditions are becoming unfavourable (Claus and Berkeley 1986). The primary function of endospores is to
Geomicrobiology of the regolith
ensure the survival of a bacterium through periods of environmental stress (Madigan and Martinko 2006). The ability to endure environmental stress makes alkalophilic Bacillus spp. ideally suited for the environmental conditions at the Barns area – that is, extremely low nutrient availability, a prolonged period of extremely dry, high salt contents and pH values between 8 and 10 – that are also characteristic of arid regions of Australia. The results demonstrate that Bacillus spp. resident in the calcareous materials at the Barns site are capable of forming Au- anomalous Ca carbonates via the ureolysis pathway.
7.8 APPLICATION OF MOLECULAR TOOLS AND TECHNIQUES To study the diversity of complex microbial populations, as well as their functional activity in the regolith, numerous culture-dependent and culture-independent techniques have been developed (Figure 7.13; Barns and Nierzwicki-Bauer 1997). Culture-dependent techniques use microbial growth media to enrich and isolate microorganisms from regolith samples (Barns and Nierzwicki-Bauer 1997). However, all culture-dependent methods share a main bias: they rely on having to culture the organisms in a growth medium in vitro. Depending on the type of regolith sample, only 0.001 to 10% of the total microbial species contained therein can be successfully cultured using existing culturing techniques (Alexander 1977). In recent years, culture-independent methods of characterising microbial population in regolith samples have been developed (Figure 7.13). These methods are based on the analyses of cellular components, such as deoxyribonucleic acid (DNA), ribonucleic acid (RNA) or phospholipid fatty acid (PLFA) (Barns and Nierzwicki-Bauer 1997). These cellular components are extracted directly from regolith materials without prior cultivation. The main advantage of molecular methods is that they do not rely on culturing the organisms and, thus, provide a more accurate insight into the in situ composition and activity of microbial communities in regolith samples. Figure 7.13 shows how they are used in sequence/combination to obtain a detailed representation of the phylogenetic and functional relationships in microbial
communities. Phylogenetic methods aim to identify key organisms, community structures and genetic relationships, and are based on the extraction, amplification, fingerprinting, sequencing and analyses of, typically, ribosomal DNA and RNA (Barns and Nierzwicki-Bauer 1997). Methods assessing the functional aspects of microbial communities target genes that encode proteins/enzymes responsible for key biogeochemical transformations (DiChristina et al. 2005). The identification of bio-indicator organisms or genes present in elevated numbers in regolith samples with anomalous trace metals concentrations may lead to the development of specific nucleic-acid probes for these bio-indicators that can be used directly for bioprospecting (Reith and Rogers 2007). Nucleic-acid probes, which are complementary to signature sequences of functional or ribosomal DNA or RNA of particular species or groups of organisms, are fluorescently or isotopically labelled allowing for their detection at very low concentration (Barns and Nierzwicki-Bauer 1997). Nucleic-acid probes can bind to bulk community DNA or RNA bound in filter membranes, making it possible to quantify the amount and activity of particular bio-indicator organisms (Barns and Nierzwicki-Bauer 1997). Microarrays are one of the exciting new tools in molecular microbiology, which allow researchers to study complex microbial communities. Microarrays are gene-chips on which several hundred thousand nucleic acid probes can be placed and used simultaneously, as shown in Figure 7.13 (Widada et al. 2002). This makes microarrays extremely powerful tools for detecting multiple genes from regolith samples, which can provide vast information relating to the phylogenetic and/or functional structure of microbial ecosystems. Expression microarrays are used to assess gene expression (that is, which genes are active) in environmental samples. This type of microarray provides information relating to the activity of the microbe(s): that is, ‘listening-in’ to what the microbes are saying and key genes associated with environmental attributes can be found. For example, genes that are important when a species is subjected to elevated concentrations of heavy metals can be identified. Using signatures of genes involved in metal cycling will lead to the development of specific
149
150
Regolith Science
Field sterile sampling Whole cell in situ hybridisation
REGOLITH SAMPLE Isolation
Colony/dot blot
CULTURES
Extraction and purification of community DNA/RNA from regolith materials
Dot/slot blot
DNA
RNA
Primer construction For gene sequences PCR R T-PCR (i) polymerase chain reaction (PCA)-DNA (gene presence) (ii) reverse transciptase-PCR-Messanger RNA (gene Expression)
Dot/slot blot
Sequencing of Individual bands
Labeling
GENOMIC LIBRARIES
PCR PRODUCTS
Cloning PCR
Reverse Hybridization
Molecular fingerprinting analysis Sequencing
MICROARRAY
Southern blot
GENETIC FINGERPRINTS
CLONE LIBRARIES
Sequencing of Individual bands
Sequencing of Unique clones
Colony/dot blot
Immobilized Probes Hybridization analyses
SEQUENCE DATABASE Comparative sequence analysis
NUCLEIC ACID PROBES
PHYLOGENETIC TREE
Figure 7.13: Schematic diagram of molecular microbial tools for studying the microbial populations and activities in the regolith.
Geomicrobiology of the regolith
nucleic-acid probes and microarrays for these bioindicators that can be used directly for bio-prospecting. Identification of genes that facilitate heavy metal transformations may then lead to the development of microarrays with specific probes for these genes. These could then be applied as exploration tools after RNA extraction from regolith samples. Heavy-metal-specific bacterial sensors are another promising bio-prospecting tool, which allow measurement of the concentration of mobile heavy metals in soil samples to µg/kg (ppb) levels (Tibazarwa et al. 2001). Antibody-based or immunoassays offer alternative approach for metal ion detection in natural samples. An immunoassay makes use of the binding between an antigen and its homologous antibody in order to quantify the specific antigen or antibody in a sample (Blake et al. 1998; Johnson 2003). A number of biosensor and antibodies for heavy metal complexes have been developed. Methods to test for Hg, Cd, Co, Pb and U, have been successfully applied in ecotoxicology studies, and may be transferable to exploration purposes (Blake et al. 1998; Johnson 2003). Biosensor and immunoassays have several advantages for mineral exploration because they are quick, inexpensive, simple to perform, portable and highly sensitive and selective and thus allow rapid assessment of mineralisation in the field. However, for this to become reality, microbial bio-indicators and biosensors, as well as immunoassay methods, must first demonstrate their full potential and cost competitiveness and benefit over existing methods.
7.9
CONCLUSIONS AND APPLICATIONS
The understanding of microbially mediated processes in the regolith that is gained from geomicrobiological studies, like the ones presented in this chapter, are applicable in a number of basic and applied fields such as biogeochemistry and astrobiology, as well as geochemical exploration, mineral processing and bioremediation. By identifying biogeochemical processes that lead to the dispersion and accumulation of trace elements in regolith – and quantifying the reaction kinetics of these processes in different materials – geomicrobiology will help
mineral explorers to develop predictive biogeochemical modelling tools. Ultimately, it will be possible to incorporate appropriate microbiological data into numerical geochemical models to predict dispersion, transport and concentration of trace elements in, and around, mineralised zones. Identification of the microbiota that control the solubilisation, transport and precipitation of trace metals in the regolith may also lead to the discovery of indicator organisms or genes, and the may foster the development of microbial biosensors for the underlying mineralisation. Furthermore, identifying the biochemical and physiological pathways used by microorganisms during the turnover of metals, may lead to the development of bio-processing capacities for metal-containing ores and biomediation techniques for metal-contaminated sites. By linking basic scientific endeavour with industrial applications, geomicrobiological research has already established itself as one of the key areas of interest for regolith geoscience – and much of our future understanding of regolith processes will come from understanding the regolith microorganisms. With these prospects, and the ever-increasing capabilities of molecular microbial methods, an exciting decade of regolith geomicrobial research beckons.
7.10
REFERENCES
Adams JB, Palmer F and Staley JT (1992). Rock weathering in deserts: mobilisation and concentration of ferric iron by microorganisms. Geomicrobiology Journal 10, 99–114. Alexander M (1977). Introduction to Soil Microbiology. John Wiley and Sons, New York. Aspandiar MF, Anand R and Gray D (2006). ‘Mechanisms of metal dispersion through transported cover: a review.’ Report 230. CRC LEME, Perth. Banfield JF and Hamers RJ (1997). Processes at minerals surfaces with relevance to microorganisms and prebiotic systems. Reviews in Mineralogy 35, 81–117. Banfield JF, Welch SA, Zhang H, Ebert T and Penn RL (2000). The role of aggregation in crystal growth and transformations in nanophase FeOOH biomineralization products. Science 289, 751–754.
151
152
Regolith Science
Barker WW, Welch SA and Banfield JF (1997). Geomicrobiology of silicate mineral weathering. Reviews in Mineralogy 35, 391–428. Barns SM and Nierzwicki-Bauer SA (1997). Microbial diversity in ocean, surface, and subsurface environments. Reviews in Mineralogy 35, 35–79. Bazylinski DA and Frankel RB (2003). Biologically controlled mineralization in prokaryotes. Reviews in Mineralogy and Geochemistry 54, 217–247. Beard BL, Johnson CM, Skulan JL, Nealson KH, Cox L and Sun H (2003). Application of Fe isotopes to tracing the geochemicaland biological cycling of Fe. Chemical Geology 195, 87–117. Beech IB (2004) Corrosion of technical materials in the presence of biofilms – current understanding and state-of-the-art methods of study. International Biodeterioration and Biodegradion 53, 177–183. Bennett PC, Hiebert FK and Choi WJ (1996) Microbial colonization and weathering of silicates in a petroleum-contaminated groundwater. Chemical Geology 132, 45–53 Berner RA (1984). Sedimentary pyrite formation: An update. Geochimica et Cosmochimica Acta 48, 605–615. Berthelin J (1971). Alteration microbienne d’une arene granitique: Note preliminaire. Science du Sol 1, 11–29. Beveridge TJ (1989) Role of cellular design in bacterial metal accumulation and mineralization. Annual Review of Microbiology 43, 147–171. Bischoff GCO (1997). The biological origin of bacterioform gold from Australia. Neues Jahrbuch für Geologie und Paläontologie Abhandlungen H6, 329–338. Blake DA, Blake RC, Khosraviani M and Pavlov AR (1998). Immunoassays for metal ions. Analytica Chimica Acta 376, 13–19. Boyle RW (1979). The geochemistry of gold and its deposits. Geological Survey of Canada Bulletin 280. Canfield DE (1998). Isotope fractionation and sulfur metabolism by pure and enrichment cultures of elemental sulfur-disproportionating bacteria. Limnology and Oceanography 43, 253–264. Canfield DE (2001). Isotope fractionation by natural populations of sulfate-reducing bacteria. Geochimica et Cosmochimica Acta 65, 1117–1124.
Castanier S, Le Me’tayer-Levrel G and Perthuisot JP (1999a). Ca-carbonates precipitation and limestone genesis – the microbiogeologist point of view. Sedimentary Geology 126, 9–23. Castanier S, Le Me´tayer-Levrel G and Perthuisot JP (1999b). Bacterial roles in the precipitation of carbonate minerals. In Microbial Sediments. (Eds R Riding and S Awramik) pp. 32–39. Springer Verlag, Berlin. Chang YJ, Peacock AD, Long PE, Stephen JR, McKinley JP, MacNaughton SJ, Hussain AKMA, Saxton AM and White DC (2001). Diversity and characterization of sulfate-reducing bacteria in groundwater at a uranium mill tailings site. Applied and Environmental Microbiology 67, 3149–3160. Châtellier X, West MM, Rose J, Fortin D, Leppard GG and Ferris FG (2004). Characterization of ironoxides formed by oxidation of ferrous ions in the presence of various bacterial species and inorganic ligands. Geomicrobiology Journal 21, 99–112. Cheah S, Kraemer SM, Cervini-Silva J and Sposito G (2003). Steady-state dissolution kinetics of goethite in the presence of desferrioxiamine B and oxalate ligands: implications for the microbial acquisition of iron. Chemical Geology 198, 63–75. Choi SC, Chase T and Bartha R (1994). Metabolic pathways leading to mercury methylation in Desulfovibrio desulfuricans LS. Applied and Environmental Microbiology 60, 4072–4077. Claus D and Berkeley RCW (1986) Genus Bacillus. In Bergey’s Manual of Systematic Bacteriology Volume 2, pp. 1105–1139, Springer, New York. Craig PJ, Eng G and Jenkins RO (2003). Occurrence and pathways of organometallic compounds in the environment – General considerations. In Organometallic compounds in the environment. (Ed. PJ Craig) pp. 1–56. John Wiley and Sons, Chichester, UK. Cunin R, Glansdorff N, Pierard A and Stalon V (1986). Biosynthesis and metabolism of arginine in bacteria. Microbiological Reviews 50, 314–352. Dassonville F and Renault P (2002). Interactions between microbial processes and geochemical transformations under anaerobic conditions: a review. Agronomie 22, 51–68. Des Marais DJ (1997). Long-term evolution of the biogeochemical carbon cycle. Geomicrobiology: Inter-
Geomicrobiology of the regolith
actions between Microbes and Minerals. Reviews in Mineralogy 35, 429–445. DiChristina TJ, Fredrickson JK and Zachara JM (2005). Enzymology of electron transport: energy generation with geochemical consequences. Reviews in Mineralogy and Geochemistry 59, 27–52. Drake HL (1994). Acetogensis and acetogenic bacteria, and the acetyl-CoA ‘Wood/Ljungdahl’ pathway: past and current perspectives. In Acetogenesis. (Ed. HL Drake) pp. 3–60. Chapman and Hall, New York. Drever JI (1994). The effect of land plants on the weathering rate of silicate minerals. Geochimica et Cosmochimica Acta 58, 2325–2332. Drever JI and Vance GF (1994). Role of soil organic acids in mineral weathering processes. In Organic Acids in Geological Processes. (Eds ED Pittman and MD Lewan) pp. 138–161. Springer Verlag, Germany. Druschel GK, Labrenz M, Thomsen-Ebert T, Fowle DA and Banfield JF (2002). Geochemical modelling of ZnS in biofilms: an example of ore depositional processes. Economic Geology 97, 1319–1329. Dürr M, Wakelin SA, Rogers SL, White I, Macdonald BCT and Welch S (2006). Archeal diversity of an acid sulfate soil in coastal northern NSW. In Regolith 2006 – Consolidation and Dispersion of Ideas. (Eds RW Fitzpatrick and P Shand) pp. 63–67. CRC LEME, Perth. Ehrlich HL (1996). How microbes influence mineral growth and dissolution. Chemical Geology 132, 1–3. Ehrlich HL (2002). Geomicrobiology. Marcel Dekker, Inc., New York. Evangelou B (1995) Pyrite oxidation and its control. CRC Press, Boca Raton, Florida. Fortin D and Beveridge TJ (2000). Mechanistic routes to biomineral surface development. In Biomineralization: From Biology to Biotechnology and Medical Application. (Ed. E Bäuerlein) pp. 7–24. WileyVCH, Weinheim, Germany. Frankel RB and Bazylinski DA (2003). Biologically induced mineralization by bacteria. Reviews in Mineralogy and Geochemistry. 54, 95–114. Freyssinet P and CRM Butt (1988). ‘Morphology and geochemistry of gold in a laterite profile, Bardoc
Mine, Western Australia.’ CSIRO Exploration Geoscience Restricted Report MG 59R. 24 pp. (Reissued as Open File Report 5, 1998). CRC LEME, Perth. Friedrich CG, Bardischewsky F, Rother D, Quentmeier A and Fischer J (2005). Prokaryotic sulfur oxidation. Current Opinion in Microbiology 8, 1–7. Frimmel HE, LeRoex AP, Knight J, and Minter WE (1993). A case study of the postdepositional alteration of Witwatersrand Basal Reef Gold Placer. Economic Geology 88, 249–265. Fyfe WS (1996). The biosphere is going deep. Science 273, 448. Garcia-Vallès M, Urzì C, De Leo F, Salamone P and Vendrell-Saz M (2000). Biological weathering and mineral deposits of the Belevi marble quarry (Ephesus, Turkey). International Biodeterioration and Biodegradation 46, 221–227. Gutknecht JLM, Goodman RM and Balser TC (2006). Linking soil process and microbial ecology in freshwater wetland ecosystems. Plant and Soil 289, 17–34. Herbert RA (1999). Nitrogen cycling in coastal marine ecosystems. FEMS Microbiology Reviews 23, 563–590. Hiebert FK and Bennett PC (1992). Microbial control of silicate weathering in organic-rich groundwater. Science 258, 278–281 Huang WH and Kiang WC (1972). Laboratory dissolution of plagioclase in water and organic acids at room temperature. American Mineralogist 57, 1849–1859. Huang PM and Schnitzer M (1986). Interactions of Soil Microbes with Natural Organics and Microbes. Soil Science Society of America Special Publication 17. Madison, Wisconsin Janzen HH (2004). Carbon cycling in earth systems – a soil science perspective. Agriculture Ecosystems and Environment 104, 399–417. Johnson DK (2003). Fluorescence polarization immunoassays for metal ions. Combinatorial Chemistry and High Throughput Screening 6, 245–255. Jones DL, Farrar J and Giller KE (2003). Associative nitrogen fixation and root exudation - What is theoretically possible in the rhizosphere? Symbiosis 35, 19–38.
153
154
Regolith Science
Jong T and Parry DL (2003). Removal of sulfate and heavy metals by sulfate reducing bacteria in shortterm bench scale upflow anaerobic packed bed reactor runs. Water Research 37, 3379–3389. Joulian C, Ramsing NB and Ingvorsen K (2001). Congruent phylogenies of most common small-subunit rRNA and dissimilatory sulfite reductase gene sequences retrieved from estuarine sediments. Applied and Environmental Microbiology 67, 3314–3318. Juniper SK and Tebo BM (1995). Microbe-metal interactions and mineral deposition at hydrothermal vents. In The Microbiology of Deep-Sea Hydrothermal Vents. (Ed. DM Karl) pp. 219–253. CRC Press, Boca Raton, Florida. Kalinowski BE and Schweda P (1996). Kinetics of muscovite, phlogopite, and biotite dissolution and alteration at pH 1-4, room temperature. Geochimica et Cosmochimica Acta 60, 367–385. Kalinowski BE, Liermann LJ, Givens S and Brantley SL (2000). Rates of bacteria-promoted solubilization of Fe from minerals: a review of problems and approaches. Chemical Geology, 169, 357–370. Karamushka VI, Gruzina TG, Podolska VI and Ulberg ZR (1987a). Interaction of glycoprotein of Bacillus pumilis cell wall with liposomes. Ukrainskii Biokhimicheskii Zhurnal 59, 70–75. Karamushka VI, Ulberg ZR, Gruzina TG, Podolska VI and Pertsov NV (1987b). Study of the role of surface structural components of microorganisms in heterocoagulation with colloidal gold particles. Prikladnaia Biokhimiia Mikrobiologiia. 23, 697–702. Karthikeyan S and Beveridge TJ (2002). Pseudomonas aeruginosa react with and precipitate toxic soluble gold. Environmental. Microbiology 4, 667–675. Kashefi K and Lovley DR (2003). Extending the upper temperature limit for life. Science 301, 934. Kashefi K, Tor JM, Nevin KP and Lovley D (2001). Reductive precipitation of gold by dissimilatory Fe(III)-reducing bacteria and archaea. Applied and Environmental Microbiology 67, 3275–3279. Keeling JR (1993). Microbial influence in the growth of alluvial gold from Watts Gully, South Australia. South Australia Geological Survey Quarterly Geology Notes 126, 12–19.
Keller JD and Wood W (1993). Possibility of chemical weathering before the advent of vascular plants. Nature 364, 223–225. Korobushkina ED, Karavaiko G and Korobushkin IM (1983) Biochemistry of gold. In Environmental Biogeochemistry. (Ed. R Hallberg) pp. 325–333. Ecological Bulletins 35, Stockholm, Sweden. Kraemer SM, Butler A, Borer P and Cervini-Silva J (2005). Siderophores and the dissolution of ironbearing minerals in marine systems. Reviews in Mineralogy and Geochemistry 59, 53–84. Krebs W, Brombacher C, Bosshard PP, Bachofen R and Brandl H (1997). Microbial recovery of metals from solids. FEMS Microbiology Reviews 20, 605–617. Lebedeva EV, Lyalikova NN and Bugel’skii YY (1979). Participation of nitrifying bacteria in the weathering of serpentized ultrabasic rock. Mikrobiologia 47, 898–904. Ledin M and Pedersen K (1996). The environmental impact of mine wastes-Roles of microorganisms and their significance in treatment of mine wastes. Earth Science Reviews 41, 67–108. Lengke MF and Southam G (2005). The effect of thiosulfate-oxidizing bacteria on the stability of the gold-thiosulfate complex. Geochimica et Cosmochimica Acta 69, 3759–3772. Lengke MF, Fleet M and Southam G (2006a). Morphology of gold nanoparticles synthesized by filamentous cyanobacteria from gold(I)-thiosulfate and gold(III)-chloride complexes. Langmuir 22, 2780–2787. Lengke MF, Fleet ME and Southam G (2006b). Bioaccumulation of gold by filamentous cyanobacteria between 25 and 200°C. Geomicrobiology Journal 23, 591–597. Lengke MF, Ravel B, Fleet ME, Wanger G, Gordon RA and Southam G (2006c). Mechanisms of gold bioaccumulation by filamentous cyanobacteria from gold(III)-chloride complex. Environmental Science and Technology 40, 6304–6309. Lichtner PC and Biino GG (1992). A first principles approach to supergene enrichment of a porphyry copper protore: I. Cu-Fe-S subsystem. Geochimica et Cosmochimica Acta 56, 3987–4013. Lintern MJ and Butt CRM (1993). Pedogenic carbonate: an important sampling medium for
Geomicrobiology of the regolith
gold exploration in semi-arid areas. Exproration Research News 7, 7–11. Lintern MJ, Sheard MJ and Chivas AR (2006). The source of pedogenic carbonate associated with goldcalcrete anomalies in the western Gawler Craton, South Australia. Chemical Geology 235, 299–324. Little D (2007). The rhizosphere: organic acid microbe mineral interactions. A multidisciplinary investigation of forest soil biogeochemistry in temperate Australian landscapes. PhD thesis, Australian National University, Canberra. Little BJ, Wagner PA and Lewandrowski Z (1997). Spatial relationships between bacteria and mineral surfaces. Reviews in Mineralogy 35, 123–155. Lloyd JR (2003). Microbial reduction of metals and radionuclides. FEMS Microbiology Reviews 27, 411–425. Lovley DR and Chapelle FH (1995). Deep subsurface microbial processes. Reviews of Geophysics 33, 365–381. Madigan MT and Martinko JM (2006). Brock Biology of Microorganisms. 11th edn. Prentice Hall, New York. Malstrom M and Banwart S (1997). Biotite dissolution at 25°C: The pH dependence of dissolution rate and stoichiometry. Geochimica et Cosmochimica Acta 61, 2779–2799. Millar CS (1974). Decomposition of coniferous leaf litter. In Biology of Plant Litter Decomposition. (Eds CH Dickinson and GJF Pugh) pp. 105–128. Academic Press, London. Morse JW, Millero FJ, Cornwell JC and Rickard D (1987). The chemistry of the hydrogen sulfide and iron sulfide systems in natural waters. Earth Science Reviews 24, 1–24. Mossman DJ, Reimer T and Durstling H (1999). Microbial processes in gold migration and deposition: modern analogues to ancient deposits. Geoscience Canada. 26, 131–140. Nakajima A (2003). Accumulation of gold by microorganisms. World Journal of Microbiology and Biotechology 19, 369–374. Nealson KH and Stahl DA (1997). Microorganisms and biogeochemical cycles: what can we learn from layered microbial communities? Reviews in Mineralogy 35, 5–34.
Nordstrom DK and Southam G (1997). Geomicrobiology of sulfide metal oxidation. Reviews in Mineralogy 35, 361–390. Pace NR (1997). A molecular view of microbial diversity and the biosphere. Science 276, 734-740. Parkes RJ, Crag BA, Bale SJ, Getliff JM, Goodman K, Rochelle PA, Fry JC, Weightman AJ and Harvey SM (1994). Deep bacterial biosphere in Pacific Ocean sediments. Nature 371, 410–413. Paul EA and Clark FE (1996). Soil Microbiology and Biochemistry. 2nd edn. Academic Press, San Diego, California. Pedersen K (1993). The deep subterranean biosphere. Earth Science Reviews 34, 243–260. Perez-Jimenez JR, Young LY and Kerkhof LJ (2001). Molecular characterisation of sulfate-reducing bacteria in anaerobic hydrocarbon -degrading consortia and pure cultures using the dissimilatory sulfite reductase (dsr AB) genes. FEMS Microbiology Ecology 35, 145–150. Rawlings DE (2002). Heavy metal mining using microbes. Annual Review of Microbiology. 56, 65–91. Reith F and McPhail DC (2006). Effect of resident microbiota on the solubilization of gold in soils from the Tomakin Park Gold Mine, New South Wales, Australia, Geochimica et Cosmochimica Acta 70, 1421–1438. Reith F and McPhail DC (2007). Microbial influences on solubilisation and mobility of gold and arsenic in regolith samples from two gold mines in semiarid and tropical Australia. Geochimica et Cosmochimica Acta 71, 1183–1196. Reith F and Rogers SL (2008). Exploration geomicrobiology – the new frontier. In Biogeochemistry in mineral exploration. (Ed. CE Dunn) pp. 393–413. Elsevier, Amsterdam. Reith F and Rogers SL (2008). Assessment of bacterial communities in auriferous and non-auriferous soils using genetic and functional fingerprinting analyses. Geomicrobiology Journal 25, 203–215. Reith F, Lengke MF, Falconer D, Craw D and Southam G (2007). The geomicrobiology of gold. The International Society for Microbial Ecology Journal 1, 567–584 Reith F, McPhail DC and Christy AG (2005). Bacillus cereus, gold and associated elements in soil and
155
156
Regolith Science
regolith samples from Tomakin Park Gold Mine in south-eastern New South Wales. Journal of Geochemical Exploration 85, 81–89. Reith F, Rogers SL, McPhail DC and Webb D (2006). Biomineralization of gold: biofilms on bacterioform gold. Science 313, 333–336. Rivadeneyra MA, Martin-Algarra A, Sanchez-Navas A and Martin-Ramos D (2006). Carbonate and phosphate precipitation by Chromohalobacter marismortui. Geomicrobiology Journal 23, 89–101. Sammut JI and White I (1996). Acidification of an esturine triburary in eastern Australia due to drainage of acid sulphate soils. Marine and Freshwater Research 47, 669–684. Sand W, Gehrke T, Jozsa PG and Schippers A (2001). (Bio)chemistry of bacterial leaching - direct vs. indirect bioleaching. Hydrometallurgy 59, 159–175. Saunders JA, Duke L and Roden EE (1997). Effects of anaerobic bacterial respiration on the geochemistry of silica: evidence from the Holocene diagenetic alteration of logs. Annual Meeting of the Geological Society of America Abstracts with Programs 29, 295. Savvaidis I, Karamushka VI, Lee H and Trevors JT (1998). Micro-organism-gold interactions. BioMetals 11, 69–78. Schmidt Mumm A and Reith F (2007). Biomediation of calcrete at the gold anomaly of the Barns prospect, Gawler Craton, South Australia. Journal of Geochemical Exploration , 13–33. Schwartzman DW and Volk T (1989). Biotic enhancement of weathering and the habitability of Earth. Nature 340, 457–459. Segers R (1998). Methane production and methane consumption: a review of processes underlying wetland methane fluxes. Biogeochemistry 41, 23–51. Southam G (2000). Bacterial surface-mediated mineral formation. In Environmental Microbe-Mineral Interactions. (Ed. DR Lovely) pp. 257–276. ASM Press, Washington DC. Southam G and Beveridge TJ (1994). The in vitro formation of placer gold by bacteria. Geochimica et Cosmochimica Acta 58, 4227–4230. Southam G and Beveridge TJ (1996). The occurrence of sulfur and phosphorus within bacterially derived crystalline and pseudeocrystalline gold formed in
vitro. Geochimica et Cosmochimica Acta 60, 4369–4376. Southam G and Saunders JA (2005). The geomicrobiology of ore deposits. Economic Geology 100, 1067–1084. Stetter KO, Fiala G, Huber G, Huber R and Segerer A (1990). Hyperthermophilic microorganisms. FEMS Microbiology Reviews 75, 117–124. Stone A (1997). Reactions of extracellular organic ligands with dissolved metal ions and mineral surfaces. Reviews in Mineralogy 35, 309–341. Tibazarwa C, Corbisier P and Mench M (2001). A microbial biosensor to predict bioavailable nickel in soil and its transfer to plants. Environmental Pollution. 113, 19–26. Truong DH, Eghbal MA, Hindmarsh W, Roth SH and O’Brien PJ (2006). Molecular mechanisms of hydrogen sulfide toxicity. Drug Metabolism Reviews 38, 733–744. Tsuruta T (2004) Biosorption and recycling of gold using various microorganisms. Journal of General and Applied Microbiology 50, 221–228. Tuck VA, Edyvean RGJ and West JM (2006). Biologically induced clay formation in subsurface granitic environments. Journal of Geochemical Exploration 90, 123–133. Tyson GW, Chapman J, Hugenholtz P, Allen EE, Ram RJ, Richardson PM, Solovyev VV, Rubin EM, Rokhsar DS and Banfield JF (2004). Community structure and metabolism through reconstruction of microbial genomes from the environment. Nature 428, 37–43. Ulberg ZR, Karamushka VI and Vidybida AK (1992). Interaction of energized bacteria calls with particles of colloidal gold: pecularities and kinetic model of the process. Biochimica et Biophysica Acta 1134, 89–95. van der Lee GEM., de Winder B, Bouten W and Tietema A (1999). Anoxic microsites in Douglas fir litter. Soil Biology and Biochemistry 31, 1295–1301. Vandevivere P, Welch SA, Ullman WJ and Kirchman DL (1994). Enhanced dissolution of silicate minerals by bacteria at near-neutral pH. Microbial Ecology 27, 241–251. Vincent WF and James MR (1996). Biodiversity in extreme aquatic environments-lakes, ponds and
Geomicrobiology of the regolith
streams of Ross Sea sector, Antarctica. Biodiversity and Conservation 5, 1451–1471. Vogels GD and Drift CVD (1976). Degradation of purines and pyrimidines by migroorganisms. Bacteriological Reviews 40, 403–468. Watteau F and Bertlein J (1994). Microbial dissolution of iron and aluminium from soil minerals: efficiency and specificity of hydroxamate siderophores compared to aliphatic acids. European Journal of Soil Biology 30, 1–9. Watterson JR (1992). Preliminary evidence for the involvement of budding bacteria in the origin of Alaskan placer gold. Geology 20, 1147–1151. Watterson JR (1994) Artifacts resembling budding bacteria produced in placer-gold amalgams by nitric acid leaching. Geology 22, 1144–1146. Weiner S and Dove PM (2003). An overview of biomineralization processes and the problem of the vital effect. Reviews in Mineralogy and Geochemistry 54, 1–29. Welch SA and Ullman WJ (1993). The effect of organic acids on plagioclase dissolution rates and stoichiometry. Geochimica et Cosmochimica Acta 57, 2725–2736. Welch SA and Ullman WJ (1996). Feldspar dissolution in acidic and organic solutions: Compositional and pH dependence of dissolution rate. Geochimica et Cosmochimica Acta 60, 2939–2948 Welch SA and Ullman WJ (1999). The effect of microbial glucose metabolism on bytownite feldspar dis-
solution rates between 5 and 35°C. Geochimica et Cosmochimica Acta 63, 3247–3259. Welch SA and Ullman WJ (2000). The temperature dependence of Bytownite feldspar dissolution in neutral aqueous solutions of inorganic and organic ligands at low temperature (5–35oC). Chemical Geology 167, 337–354. White I, Melville MD, Wilson BP and Sammut J (1997). Reducing acidic discharges from coastal wetlands in eastern Australia. Wetlands Ecology Management 5, 52–72. Widada J, Norjiri Y and Omori T (2002). Recent developments in molecular techniques for identification and monitoring of xenobiotic-degrading bacteria and their catabloc genes in bioremediation. Applied Microbiology and Biotechnology 60, 45–59. Wilson AF (1984). Origin of quartz-free gold nuggets and supergene gold found in laterites and soils - a review and some new observations. Australian Journal of Earth Sciences 31, 303–316. Witkiewicz PL and Shaw CF (1981). Oxidative cleavage of peptide and protein disulphide bonds by gold(III): A mechanism for gold toxicity. Journal of the Chemical Society, Chemical Communications 21, 1111–1114. Zhu XK, Guo Y, Williams RJP, O’Nions RK, Matthews A, Belshaw NS, Canters GW, de Waal EC, Weser U, Burgess BK and Salvato B (2002). Mass fractionation processes of transition metal isotopes. Earth and Planetary Science Letters 200, 47–62.
157
Regolith Science
Calcium
Aluminium
Iron
Figure 8.5: Sub-millimetre variation and metal zonation in soil chemical composition around a tree root hair. Ca (purple pink) in the root cortex, Fe (blue green) in the soil matrix and distinct Al (yellow green) accumulation at the root surface. Scanning Electron Microscope back scattered electron image (Field of View = 1 mm).
(a) 318000 m
319000 m
320000 m
321000 m
322000 m
323000 m
(b)318000 m
319000 m
320000 m
321000 m
322000 m
323000 m
6629000 m
6629000 m
6628000 m
6628000 m
6627000 m
6627000 m
6626000 m
6626000 m
6625000 m
6625000 m
6624000 m
6624000 m
6623000 m
6623000 m 32/A/39
32/A/40
TMf003-08 -45
-38
-36
-27 mgal >10
>20
5 >30
40
80 >40
120 metres >50
>60
Figure 9.2: (a) Pseudocolour image of the Bouguer gravity (with NE illumination) for an area of variable regolith thickness in the Grants Patch area of Eastern Goldfields, Western Australia (see Figure 9.3 for location). Interpreted lithological boundaries and structures are shown with white lines. These data were collected on a 200 m (NS) by 50 m (EW) grid using Scintrex CG3 Autograv meters (Bell et al. 2001). (b) Results from a 3D gravity inversion (the VPmg potential field inversion program) predicting regolith thickness for the same area. Regolith thickness interpreted from available drill holes is indicated by the circles (Bell et al. 2001). Images courtesy of Geoscience Australia.
159
Regolith Science
o
121 00’ 0
WA
Northing (m) 1100N 1000N
1200N
Depth 100 (m)
Bardoc
Meta-sediments
Ora Banda
Grants Patch Gravity Survey Figure 9.2
150
Broad Arrow
Grants Patch
200
Paddington
Line 10
Flag
Bla ck
121o 00’
Base of paleochannel Saprolite Basement
Wollubar Sstn
50
121o00’
900N
Perkolilli shale
Clay
aN od o rlg da Yin
BP9 Black Flag
.
Kanowna Kunanaling
Lake Yindarlgooda
g
llin
Kalgoorlie S.
K
a an un
lg
oo
da
Boulder
nd
ar
le ie Va Bonn
Yi
160
nan
Han
Coolgardie
o
121 00’ HW64
Line 18
Wollubar
ar
ub oll W
Location of SAM survey Figure 9.7
87.85
100
87.50 Gravity (mgal) 87.15
0 -100
0
86.80 23 Depth Saprolite (m) 47 2.30 g/cm 200
Channel sediments 2.00 g/cm3
400 600 Northing (m)
20 km
Paleochannel and drilling line
Overburden 2.40 g/cm3
TM028-07
Lake Lefroy
Kambalda
G and WS pipeline Granodiorite 2.70 g/cm3 800
Main highway Railway BP9
Production Bore
(Source: Kern and Commander 1993)
Figure 9.3: Paleochannel investigations using gravity and shallow seismic reflection techniques. Line 18, covering the Wollubar Paleochannel, was traversed with a gravimeter and the modelled (solid red line) versus observed gravity at stations (circles) is shown in the lower left inset figure. Results from a shallow seismic reflection survey across Line 10 are shown in the upper right. Both techniques resolve variability relating to the sediments within the paleochannels (after Cooper 1994).
Regolith Science
(a)
(b)
metres metres
(c)
metres
Figure 9.4: Imaged airborne radiometric and magnetic data sets for the West Wyalong area in NSW, Australia. (a) The Ternary image of radiometrics shows a variable response relating to the nature and origin of regolith materials present (see section 9.4.1). (b) A pseudocoloured total magnetic intensity image (TMI) shaded by a 1st vertical derivative. (c) The pseudocoloured 1st vertical derivative magnetic image showing high-frequency variations relating to accumulations of maghemite gravels and paleochannels with these materials. The result of their combined interpretation is presented in Figure 9.5. Data supplied by Geoscience Australia.
161
162
Regolith Science
Interpreted Base of Transported cover
Depth below surface (m)
0m
Low Resisitivty Ohm.m
500m
High
1000m
25 50
75
100 Figure 9.6: ERI traverse (~1km long) across a paleochannel in Western Australia. The channel has eroded into the underlying basement rock to a maximum depth of 60 m. Resistivity has been imaged to 100 m. The image depicts a resistivity crosssection, with the reds, orange and yellow colours indicative of resistive zones. Blues and greens are indicative of relatively conductive zones. Drill hole data indicate the presence of transported sands and gravels (relatively resistive) overlying transported silts and clays (relatively conductive). The thickness of the channel sediments (indicated by the white bar on the drillholes) correlates well with an interpreted depth defined from the ERI data: the base of the channel is indicated by a marked change in resistivity (data supplied by GeoForce Pty Ltd).
6534000N
Sirius open pit
6532000N
Brittannia open pit
N 0
500 m
384000E
386000E
Figure 9.7: Pseudocoloured first vertical derivative image of SAM EQMMR (equivalent MMR) data, calculated from TFMMR data collected over an area of deep weathering in the Eastern Goldfields, Western Australia. The image is illuminated by a north-east sun angle. White lines denote the axis of strong linear anomalies, including features correlating with the known mineralised shears (Sirius ‘A’ and Britannia ‘B’) (Image courtesy of E. Stolz, Geoscience Australia and Goldfields).
Regolith Science
Elevation (m)
420 400
3.0
380
log10 2.5 conductivity mS/m
360 340
Elevation (m)
2.0
320
420
(b) 3.0
400
log10 2.5 conductivity mS/m 2.0
380 360 340 320
420 Elevation (m)
(a)
(c)
400
Sheetwash
380
Alluvium
360
Saprolite
340
Saprock/ Fresh rock
320 318000
320000
322000
Figure 9.9: Conductivity–depth sections for PROTEM ground TEM (section a) and TEMPEST airborne EM data (section b) along a transect located in the Eastern Goldfields, Western Australia. The upper and lower bounds of a ‘conductive unit’ with a minimum threshold of 100 mS/m are shown over the TEMPEST section. The regolith profile derived from drilling (section c) is also shown (Worrall et al. 2001).
163
Regolith Science
(a)
(d)
Flight line map
Interval conductivity map
Surface 10-15 m
(b)
1e+02
Depth
164
(c) Transformation/Inversion
Response
(CDI/LEI) 1e+01 1e-05
1e-04 Time (s)
1e-03
1 0
Conductivity 10 100
10 Interval Depth (m) 20
1000
Conductivity Conductivity depth profile
30
Figure 9.10: Schematic representation of fixed wing Time domain EM data acquisition and interpretation. (a) Data are acquired along parallel flight lines. (b) The receiver towed beneath and behind the aircraft measures the electromagnetic response of the ground as it decays over time. (c) The measured response is used to determine the conductivity–depth function by transformation or inversion. (d) The conductivity–depth functions may be represented as layers that are combined to produce an interpreted conductivity–depth section and/or interval conductivities that map the spatial distribution of conductivity as it varies with depth (adapted from Fitterman and Deszcz-Pan 2001).
Regolith Science
Caltowie
Caltowie
Jamestown
Jamestown
CDI 00 - 05m
CDI 05 - 10m
Caltowie
Caltowie
Jamestown
Jamestown
CDI 10 - 15m
CDI 15 - 20m
Caltowie
Caltowie
Jamestown
Jamestown
CDI 25 - 30m
CDI 20 - 25m
Caltowie
Caltowie
Jamestown
Jamestown
CDI 30 - 40m
CDI 40 - 60m Conductivity (mS/m)
0
6
12 km 0
40
80
120
160
200
Figure 9.11: Interval conductivity images for 5 and 10 m depth slices below the ground surface. These images have been shaded by a DEM. Conductive areas are largely confined to the top 30 m of valley fill sediment (after Wilford 2004).
165
166
Regolith Science
Jamestown Caltowie
0
3
6 km
Figure 9.13: Three-dimensional perspective of 3-band gamma-ray spectrometry image for the Jamestown region of South Australia draped over a digital elevation model for the area. The white line delineates slopes above (erosional landscapes) and below 1.5 degrees (depositional landscapes) (after Wilford 2004).
(a)
(b)
Figure 9.14: Pseudocoloured total magnetic intensity images modulated by a 1st vertical derivative image from two airborne magnetic surveys. Data for image (a) was acquired at 200 m line spacing and 80 m flying height, with the image (b) generated from data acquired with 50 m line spacings and 60 m flying height. The fine detail in image (b) allows accumulations of maghemite gravels in paleochannels to be readily resolved. Images are approximately 10 km across (images courtesy of Geoscience Australia).
Regolith Science
Salmon
Wights 862
852
280
842
843
833
251
351
824
1251
1351
814
E 1451
W
260
m AHD 240
220
Sand / silt
Lower saprolite
Lateritic duricrust and gravel
Felsic gneiss
Upper saprolite
Mafic gneiss
Fault Water table (2002) Piezometer screened section
Figure 9.15: Conductivity–depth sections for a transect across a regolith-dominated terrain in south-western Australia. From the top they are derived from EMFlow, Zohdy’s method, an Occam’s inversion, and a ‘blocky’ three layer inversion (after Munday et al. 2006).
167
168
Regolith Science
Residual Regime - 10% of landsurface Erosional Regime - 40% of landsurface Depositional Regime - 50% of landsurface
m - mafic bedrock f - felsic bedrock Erosional Regime mELA4 fESS6
mELA4
mENL2
mELA4 mENL2
mRLA2
Residual Regime mRLA1
Depositional Regime DCT1
DAT1
DAT5
Infusion of Fe into saprolite
Deposition Gossan
1 km Coarse lag of quartz and iron segregation on red clayey soil Ferruginous saprolite lag on red clayey soil Iron segregation, pods and stringers Red clayey soil, or sand on felsic lithologies
Primary sulfide Sandy hardpanised colluvium and alluvium Gravelly clayey hardpanised colluvium and alluvium Transported saprolitic clays Lateritic debris
Authigenic clays and manganese nodules
Saprolite
Lateritic residuum
Granite
Mottled zone
Basalt
Ferruginous saprolite
Komatiite
Figure 11.9: Schematic cross section for the Mount McClure district, Yilgarn Craton, showing the regolith distribution and landforms with the RED mapping units (Anand et al. 1993).
Regolith Science
SVep1
SMel4
857
SVep1 SMel2 CHer4
AOap
CHel3 CHer2
SV SMel2 SMel2 1 SVep1 SMel2 CHfs1 MCAULEY 484 CHfs10 CHer4 SVer1 CHpd1 SMel2 CH CHer4 SVer1
0
0
5 km
10 km
Figure 11.10: Extract – with magnified tile – from the Leonora Thematic Image map showing saprolites as colour-filled polygons and transported units as transparent polygons draped on an east-west gradient TMI image. Diamond symbols – generally along magnetic lineaments – indicate known gold occurrences from the MINLOC database. Regolith polygons are identified by four letter mapping symbols, in some cases with additional digits for greater distinction (that is, Chfs20 =sheet flow sediments forming colluvial fan). The symbols may be read from the map legend, which also indicates if polygons are associated with greenstone, granite saprolite or transported materials. Together, regolith and geology provide much more understanding of the exploration context (from Craig 2001).
169
170
Regolith Science
(a)
N
(b)
Figure 11.11: ‘Thumbnails’ of two regolith landform maps of Ebagoola (Pain et al. 1994). (a) Map compiled from aerial photographs and field checking. (b) Map producing in a GIS environment using different data layers and decision rules derived from field observations.
Regolith Science
0 N
ey all
Metres below modern surface
di
pa
n
rra
Di
n ba
v leo
240
0
25 km
Areas with no alluvial cover Figure 11.12: The geometry of the sub-alluvial land surface in the Lower Balonne area, southern Queensland. The Dirranbandi paleovalley is the main buried paleovalley in the area, and has a major paleo-tributary joining it from the east.
171
172
Regolith Science
a
Duplex soils prone to waterlogging and salinity. Duplex soils prone to waterlogging. Soils not prone to waterlogging and salinity.
b
Saline and sulfate metasediments. 500 mm rainfall isohyet GULF ST VINCENT
ADELAIDE
Regional Scale 0
10
20km
0 SOUTH AUSTRALIA
Catchment boundary Topographic contour interval (5m) Creek lines Dams
ADELAIDE
200m
Toposequence Dipwells Piezometers and Eh electrodes installed at P11-P15
x
Y 1 2
SOIL LAYERS Top Layers
3
Brown Brown with grey mottles and red stains Grey with red stains
30m
Bottom Layers Red* or red with yellow mottles
4
Yellow with red mottles Grey with yellow or red mottles Yellow and grey with weathered grey rock
Saline
Fresh water flow
Groundwater flow
7
Salt crystals on soil surface
Sodic
*Could also be yellow
C 400m
Figure 12.12: Descriptive soil–regolith model showing toposequences with three selected profiles, soil features (such as relict purple mottles and current very poorly drained saline soils with grey and red stains) and direction of perched fresh water flow and groundwater flow (after Fritsch and Fitzpatrick 1994; Fitzpatrick et al. 1996).
Regolith Science
3
5
1
2
1 0
1 4
6
2 2m
3 5
0
4
12
m
30
m
50m
0 0
6
6 2m
2m
125m
1500m
A and E horizons Dark reddish brown (sandy clay) Dark reddish brown (loamy) Dark brown (loamy) Black (loamy) Reddish yellow (loamy)
Nodules Calcite Black brown magnetic and non magnetic
Geology Siltstone Tillite/quartzite Interbeds of quartzite, sandstone, mudstone, siltstone, shale partly carbonaceous
B horizons Grey Dark reddish brown to reddish brown (clayey) Dark brown (loamy)
Structure Prismatic/columnar Slickensides Rock gravels and fragments Soil layers Sodic Exchangeable Na % >15 Saline ECse > 20dS/m Saline ECse 2-4dS/m
C and D horizons Grey brown (loamy with sand and carbonate gravel) Yellowish red/pale brown (clayey) Reddish brown/yellowish (clayey) (Alluvium) Yellowish red/pale yellow (clayey) (Colluvium)
Landscape soil units 1 to 6 Water flow Saline groundwater flow Freshwater flow
Figure 12.16: Whole-of-landscape 3D process model for the Jamestown case study area (South Australia) showing: (i) EM-38 map partly draped over the 3D aerial photograph drape of study area with boundaries of landscape–soil units (LSU), (ii) photographs of representative soil profiles for each LSU, (iii) geology, (iv) cross-section of typical toposequence showing the main morphological, saline and sodic soil–regolith features/layers and (v) groundwater and fresh surface water flow paths. The EM-38 map designates high conductivity values in red (subsoil expressed dry saline land), medium values in yellow– turquoise and low values in dark blue (after Fitzpatrick et al. 2003d)
173
174
Regolith Science
5
SOIL TYPE No. SUMMARY OF MAJOR MANAGEMENT OPTIONS
SOIL & WATER PROBLEMS
Strongly waterlogged saline and sodic
Strongly waterlogged and saline sulfidic
Poorly drained grey saline and sodic soil
Very poorly drained black saline sulfidic soil
SOIL DESCRIPTION
SOIL FEATURES Top Layers
1
Black with red stains
Bottom Layers Red Yellow Red with yellow mottles Yellow with grey mottles Greyish-blue with yellow mottles cracks ss
shiny clay surfaces (slickensides)
k
carbonate accumulation
c
ironstone gravel
n
saline zone sodic Fresh water flow Ground water flow Estimated relative importance of water flow directions
Elevation height above sea level (m)
Grey with red stains
Fence off Plant salt tolerant grasses Exclude stock throughout winter Stabilise bare eroded areas with salt tolerant trees, shrubs and native grasses Add lime to areas that are strongly sulfidic to prevent acid conditions Do not drain strongly sulfidic areas
m 0
230
Brown
1 2 3
7
Fence off Plant salt tolerant grasses Exclude stock throughout winter Salt tolerant trees Consider subsurface drains
0.1
2
210
0.2 0.3
3 4 200
5 7
6
LAKE
190
180
700
600
500
400
300
200
100
0
100
Distance from edge of lake (m)
Figure 12.18: Sequence of soils down a slope (two of the seven soils are illustrated) linked to a 3D mechanistic model of soil and water processes with summaries of management options associated with each soil type from Woorndoo, south-western Victoria (Table 12.4; after Fitzpatrick et al. 2003b).
8
Regolith and biota John Field and David Little
8.1
INTRODUCTION
The effects of biota on the Earth’s regolith occur at scales from 10-9 to 102 metres (Figure 8.1). Organisms – ranging from organic molecules and bacteria (Chapter 7), through those visible to the naked eye, to the largest living organisms (the giant trees) – all take part in the formation and distribution of the regolith (Field 2003). Although the importance of the biological component of the regolith was recognised hundreds of years ago (Chapter 1.5), re-emphasis has been given to it by the recognition that life depends on the environment provided by the ‘critical zone’ (Brantley et al. 2007). At the small scale, initial weathering by organisms may be physical, such as when fungal hyphae and root hairs disrupt mineral grains but that physical disruption expedites the chemical breakdown of minerals by releasing CO2 and, in a moist environment, vastly enhances the levels of carbonic acid above the abiotic concentrations at the weathering interface, while exuding low-molecular-weight organic acids (LMWOAs) (see also Chapter 7). At larger scales, material brought to the surface by burrowing animals has even led to the discovery of major mineralisation, such as malachite lumps in material excavated by wombats at Moonta, South Australia (Pryor 1962). Finally, at the regional scale, there is the general
protection and moistening effect of vegetation and litter, which encourage the ecosystem to colonise, exploit and grow, while maintaining moist conditions for mostly organic acids to attack the alumino-silicates and enhance weathering effects. This chapter considers the effects of biota on regolith materials at small to large scales (the effects of microorganisms having already been discussed in Chapter 7).
8.2 BIOTIC EFFECTS AND PROCESSES AT SMALL OR MICRO SCALES 8.2.1 Microorganisms and weathering – geomicrobial reactions A huge variety of microorganisms is involved in the breakdown of rocks and primary minerals: bacteria and archaea (discussed in Chapter 7), algae, fungi, including mycorrhizae (Figures 8.2 and 8.3; Section 8.2.2), protozoans, and symbionts such as lichens (see Section 8.3.4). A few examples will be used here to demonstrate the widespread distribution and critical roles microorganisms play in weathering and regolith formation (see Banfield et al. (1999) for an excellent overall review). The cycling of some elements is controlled by microbes: for example, the cycling of N (see also Section 7.3.2). Nitrogen is not available from
176
Regolith Science
Metres ____
10
____
_____ Atoms
-9
_____
Molecules
-8
_____
Macromolecules
10 (nm)
____
10
____
____
Particles
-10
10
_____
-7
Clay minerals _____
10-6 (µm)
____
____
____
____
____
10
10
Colloids Sesquioxides
Silt _____
-5
_____
-4
Sand
_____
10-3 (mm) 10-2 (cm) 10
-1
_____
Gravel Pebbles
_____
Cobbles ____
100
_____
Boulders
Minerals
S E C O N D A R Y M I N E R A L S P R I M A R Y M I N E R A L S
Biota
Pores and Function
Organic molecules
MICROPORES
Polysaccharides Humic and fulvic acids
MICRO-ORGANISMS Viruses Algae
Absorbed and intercrystalline water
MESOPORES
Bacteria Fungal hyphae Plant-available water
MICRO FAUNA MICRO FLORA Protozoa Nematoda Root hairs Microfungi, fungal hyphae Algal bodies, fungi, fruiting bodies Fine roots Mites, springtails Termites, ants MESO FAUNA Millipedes, centipedes, earthworms Rodents, fossorials COARSE ROOTS FLORA - herbs, forbs, grasses FAUNA - wombats, pigs, dogs
MACROPORES Aeration Fast drainage Root holes, burrows
Pipes and Tunneling Gully head collapse
MACRO FAUNA - buffaloes, elephants ____
____
1
10
102
Underground flows, caves
_____
MACRO FLORA - trees
Aquifers
_____ JFf001-08
Figure 8.1: Scales relevant to biota in the regolith: mineral and biological constituents of soils and the size and function of pores.
alumino-silicate weathering, but is fixed from atmospheric sources by microbes both in a symbiotic relationship with plants and as free standing organisms. The mineralisation and nitrification of N in the regolith are processes carried out almost exclusively by microorganisms, and the movement into plants that are not N fixers
is managed by the soil fungi. Nitrogen is moved out again during decomposition of plant remains by a suite of micro (and macro) organisms (such as the gut flora of all the invertebrates from woodlice to earthworms). Algae have been shown to be involved in active weathering in local topographic lows – even in arid
Regolith and biota
Figure 8.2: Root hair and mycorrhizal fungal hyphae (scale bar = 5 µm).
Figure 8.3: Fungal hyphae physically etch mineral grains (scale bar = 50 µm).
environments (Smith et al. 2000), where endolithic and epilithic algae are responsible for algal boring and plucking and etching of limestones (see also Section 8.3.5). Soil fungi, and especially soil mycorrhizae, are an extremely important component of the regolith and have been studied extensively because they are critical to plant growth, and thereby all forms of agriculture and forestry. Fungi are implicated in all the myriad processes in the rhizosphere and are generally studied in conjunction with a host plant. However, they can be shown to take part in weathering reactions even when isolated from all plants. For example, Yuan et al. (2004) demonstrated that several isolates of mycorrhizal fungi were capable of replacing interlayer K with protons and effused oxalate – leading to further weathering of both phlogopite and vermiculite. The minerals (soil and primary), organic matter (living and dead) and microorganisms are three critical components of regolith that control weathering (Huang 2000) and none of these can be considered in isolation. The regolith ecosystem has to be considered as a whole and understood at the atomic, molecular and microscopic level to make sense of weathering changes. The most important location for these changes is the rhizosphere, where the kinds and combinations of biomolecules are at a maximum (enhanced biological activity) and are also often different to the bulk soil (Theng and Orchard 1995).
8.2.2 The rhizosphere, rhizoplane, rhizodeposition and root exudates The rhizosphere – the narrow zone of soils surrounding plant roots and directly impacted on by their activity (Allaby 1998; Eggleton 2001) – and rhizoplane (the surface of the root or root hair in contact with the organisms, organic matter and alumino-silicates in the rhizosphere) are of particular interest because this is one of the most biologically active zones in soils and regolith (for example, Bolan et al. 1997; Jones et al. 2003) (Figure 8.4). Root exudates and the rhizosphere microbial community are likely to prove particularly important in soil weathering because the exudates are likely to increase mineral weathering through their acidity and complexing ability, and also provide nutrition for microbial communities that mine metal nutrients from soil minerals – thereby indirectly increasing chemical weathering rates (for example, Chorover and Amistadi 2001; Martino et al. 2003). However, there is still considerable controversy regarding specific pedogenic implications of many rhizosphere processes (Drever and Stillings 1997; Jones et al. 2003). 8.2.3 Soil mineral–organic acid– microorganism interactions There is a diverse set of organic acids involved in the biochemical reactions taking place in the regolith. These include ascorbic, aspartic, citric, fulvic and humic (individual acids and acid groups), hydroxybenzoic, malic, oxalic and tannic acids. These are
177
178
Regolith Science
Water + nutrients transported from tree into and out of roots
- Amino acids - Sugars - Carboxylic acids - Protons - Electrons
Soil matrix
Root uptake - Amino acids - Sugars -Carbohydrates
Metal chelates: Fe, Mn, Ca, Mg
Rhizosphere - Primary alumino-silicate minerals e.g. Na/K feldspars - Soil organic matter
- Soil bacteria - Soil fungi + organic acids + siderophores - Clays - Fe /Al oxide precipitates - Quartz
Soil matrix
Figure 8.4: The rhizosphere
commonly grouped as the carboxylic acids, or the LMWOAs (the term used in this chapter; see also Appendix 1). In general, organic acids form (and are often called) ligands by accepting an electron, and they can react with ions (particularly metal and metalloid cations and anions) to form (organic) complexes. This group of biochemical reactions are at the core of alumino-silicate mineral weathering – with rates markedly increasing in the presence of virtually all organic acids (Chapter 7.4.3). There are multiple sources of these organic acids, but one is from root exudates. Little (2007) found that Al was accumulated as either clay minerals or Al oxides in the rhizoplane, and for up to 100 µm beyond, with Fe and Ca showing slight accumulation in the outer parts of the root (Figure 8.5, page 159). Little et al. (2005a) showed that concentrations of elements such as Al, and Fe were up to 10 times higher, while Si was 2–5 times greater in solutions containing the dicarboxylic LMWOAs (oxalic, malic and citric acids) when compared with control solutions such as NaCl.
However, these acids cannot be considered in isolation because each particular acid, and its concentration and distribution, is, in turn, controlled by the microorganisms, the location in the regolith as related to the surrounding rhizosphere(s) and the aluminosilicate mineral context. 8.2.4 The rhizosphere as an ecosystem and the effects on regolith The soil microbial community is heterogenous – varying between vegetation and soil type, soil horizons, between rhizosphere and non-rhizosphere soil compartments, and at even finer scales (Marilley and Aragno 1999; Marschner et al. 2005). What is clear, however, is the extent of influence. Root length and distribution – as well as fungal hyphae and other organisms associated with roots – are so extensive that virtually all the active soil (A and B horizons) is involved. Thus, up to 1.6 m of soil is involved for grasses and much greater depths for large trees (Figures 8.6 and 8.7). There are a number of ‘guilds’ of
Regolith and biota
0
0.33
Depth (m)
0.66
1.00
1.33
1.66 Figure 8.6: The root system of rye (a grass) grown in dry sandy soil showing only primary and secondary roots (from Gilkes 1998).
soil microorganisms – bacterial and fungal – responsible for biogeochemical processes of metal elements in soils. Saprotrophic fungi are responsible for much of the litter decomposition in forest soils: liberating elements from dead plant and animal materials back into the surface soil environment (Cairney and Meharg 2002; Smith 1982). Two important components of these recycling processes are dissolved CO2 (a weak acid) and LMWOAs, which are able to change soil solution pH, provide vital nutrition for other microorganisms or directly react with soil minerals (Jones et al. 2003; Marschner and Kalbitz 2003). In an
Figure 8.7: Extensive tree roots (2.5 m of soil in photo) allow greater infiltration of water and provide organic matter throughout the regolith – driving weathering.
ideal forest, the mobilised elements would then be readily available for uptake by tree roots, although this is not necessarily the case due to the inability of the trees to take up all forms of dissolved metal nutrients. It is nevertheless possible to observe mutual relationships between saprotrophic and mycorrhizal fungi, which overcome this barrier – at least in part (Cairney and Meharg 2002; Stone 1997). One aspect of the rhizosphere that has received considerable attention in recent years is the formation of mycorrhizal symbioses, which have evolved as an improved nutrient uptake pathway for most terrestrial plant species (Brundett 2002). Mutualistic or symbiotic relationships are observed between plant roots and microorganisms that are able to extract elements such as N and P from the regolith, transform the elements into forms that can be taken up by, or transfer the element directly to, the plant roots (Curl and Truelove 1986; Hagerberg et al. 2003). These
179
180
Regolith Science
relationships are strongly developed in many Australian ecosystems, where complex nutrient conservation and recycling mechanisms have evolved in nutrientlimiting landscapes (for example, Marschner et al. 2005). Mycorrhizal fungi are especially important agents for uptake of P and K (Fomina et al. 2005). The abundance and diversity of microorganisms is strongly influenced by the roots of many plant species and, in most cases, the populations in the rhizosphere are much larger and more diverse than in adjacent non rhizosphere soils (Gomes et al. 2003; Marschner et al. 2005). In addition, there are many studies that demonstrate the ability of mycorrhizal fungi and symbiotic bacteria to enhance metal nutrient uptake (such as Fe and K) in plants – as well as to accumulate trace metals, such as Cd, Cu, Pb and Zn, thus reducing their toxic effects (Cromack et al. 1979; Hulme and Hill 2003; van Hees et al. 2004; Raapana and Field 2006). There are three major groups of micro symbionts associated with tree roots in Australian ecosystems that show varying degrees of host specificity. The Frankia spp., mycorrhiza and some Rhizobium spp. are particularly important for nutrient uptake in plants of the Casuarinaceae, Myrtaceae and Mimosaceae families, respectively (Reddell et al. 1997; Brundett 2002). These micro symbionts expand the surface area of infected tree roots, make otherwise unavailable nutrients available for uptake and extend the effective rhizosphere into parts of the regolith that tree roots cannot reach (Hoffland et al. 2003; Paris et al. 1995) – thus having a potentially profound influence on chemical weathering in soils. Recent technological advances have improved the investigation of root growth, as well as in situ sampling of rhizosphere soil water (Vogt et al. 1998; Arocena et al. 2004). Rhizotrons fitted with small digital cameras and micro suction cups allow detailed, non-destructive examination of root growth – as well as nutrient and water uptake and exudates from plant roots (Pan et al. 1998; Arocena et al. 2004). These rhizotrons can be installed beneath mature field-grown trees or agricultural lands – or scaled down versions (such as root study containers or rhizoboxes) can be used as controlled environments in nursery-scale microcosm and mesocosm investigations of rhizosphere biogeochemistry (for example, Arocena et al. 2004; Sandnes et al. 2005). Rapidly developing methodologies are also
being applied in order to understand a wide range of molecular-scaled microbe-to-mineral interactions that occur in soils (Vancura and Kunc 1987; Welch and Banfield 2002), and might be particularly useful in helping to distinguish between rhizosphere communities – where root exudates are major sources of microbial nutrition – from adjacent non-rhizosphere communities. It can be argued that deoxy ribonuclease (DNA) extraction followed by polymerase chain reaction (PCR) and Biolog sole C source utilisation experiments – when used in combination – are particularly effective tools for examining microbial community structural and functional diversity in samples from a range environments (Zak et al. 1994; Hernesmaa et al. 2005; Little et al. 2005b). Investigation of organic acid– soil mineral interactions in the rhizosphere is becoming more common as the importance of root and microbial metabolites in soil biogeochemical processes is increasingly recognised (Qin et al. 2004; Little et al. 2005a). In addition, there have been growing numbers of studies examining bioturbation (Holt et al. 1980; Field and Anderson 2003), and the effects of bacterial and fungal communities and organic acid anions on dissolution from pure mineral separates (Barker et al. 1998; Welch and Banfield 2002; Welch et al. 2002). The remainder of this section covers the application of a range of analytical techniques adapted for, and used in, fine-scale investigations of some of the different aspects of rhizosphere biogeochemistry. Since the rhizosphere is such a complex microenvironment, it is difficult to develop analytical techniques that are specifically targeted at the biotic and abiotic components and the interactions between them. In most cases standard techniques for examination of bulk soil samples can be modified for use in the rhizosphere (for example, Kirk et al. 2004; Seguin et al. 2004). These techniques are often very expensive – and can be time consuming – as well as being limited in their application due to small sample sizes, difficulties associated with sampling live roots from field grown plants, and changed conditions when examining the rhizosphere in nursery and/or microcosm experiments (Haag and Matschonat 2001; Singh et al. 2004). It is therefore imperative to understand the biogeochemical processes that are likely to be important in the selected environment in order to target the
Regolith and biota
specific biogeochemical processes under investigation. In addition, it is often difficult to replicate all soil habitat factors without introducing artefacts to the experiment (Haag and Matschonat 2001). Nevertheless recent advances in geochemical sampling techniques – and the application of molecular biology and genomic studies to soil ecosystems – now allow soil ecologists to begin elucidating nano- to micro-scale biogeochemical pathways in soils and regolith. To date, few investigations use truly multidisciplinary approaches to examine the organic acid–microbe– mineral interactions occurring in the rhizosphere (Little et al. 2005a, b), which is surprising given that the complex root–microbe–mineral interactions, and their importance in soil formation, were first eluded to in 1904 (Hiltner 1904), which was nearly 40 years before Jenny restated soil formation as a function of interactions between climate, geology, topography and biology – all acting over time (Jenny 1941). Understanding soil biogeochemistry in the rhizosphere of forest soils is particularly challenging due to the fine scales of investigation and the myriad of root– microbe–soil interactions occurring there. A particularly important consideration is which combination of molecular and culture-based approaches are most suitable for examining the resident soil microbial communities. Culture-based techniques, such as Biolog, give an indication of different microbial characteristics to those measured using restricted length fragment polymerisation (RLFP), fatty acid metal esters (FAME) or polymerase chain reaction denaturing gradient gel electrophoresis (PCR DGGE) techniques (Buyer and Drinkwater 1997; Ramsey et al. 2006). However, while these techniques are all capable of identifying potential shifts in microbial community structure and function, the differences identified by one method may not necessarily be reflected by the results of other techniques. For example, when using FAME profiling in combination with PCR DGGE, Kozdroj and van Elsas (2000) found that a greater response to artificial root exudates were observed in the culturable fraction of the soil microbial community. In another study by (Widmer et al. 2001), it was demonstrated that DNA extraction with RLFP, phospholipid fatty acid and Biolog techniques identified a number of similarities and differences between microbial communities from
different sites – showing also that differences in community structure identified by one technique may not necessarily be reflected by similar differentiation using another, because each takes advantage of different aspects of the microbial community. Thus, the complementary use of two or more techniques can give a more comprehensive view of specific aspects of the potential structural and functional diversity of soil microbial communities. 8.2.5 Biotic paleoforms in regolith Rhizomorphs, root channels and pore casts have been recognised in paleosols by a variety of workers (for example, Anand and Paine 2002). Their interpretation can lead to paleoclimate analysis and recognition of the processes acting at the time that the paleosol formed. Paleosols between basalt flows have been shown to contain ‘fossilised’ root traces among a suite of other characteristics, such as horizonation, ped structure, fossil plants and gastropods, and weathering indices consistent with pedogenesis, such as high K in paleosurface horizons (Sheldon 2003). In the sandstones of McMurdo Dry Valleys, Antarctica, Wierzchos et al. (2006) have demonstrated that endolithic microorganisms have their structures, such as cell walls, mineralised after death – along with allochthonous clay minerals and sulfate-rich salts filling the sandstone pores.
8.3 BIOTIC EFFECTS AND PROCESSES AT LOCAL OR MESO SCALES As seen in Section 7.4, there is a complex interaction between the biotic and abiotic components of the regolith. Organisms are highly dependent on the soils in which they grow and, in turn, the physical structure and chemical and mineralogical composition of the regolith depend on the actions of these same organisms (Corbet 1935; Vancura and Kunc 1987). Locally active soil fauna and flora are important in maintaining soil stability and soil turnover – as well as organic matter addition and turnover (Burges 1958; Nannipieri et al. 2003), which all have important consequences for pedogenic and regolith evolution. At the local scale, individual organisms such as large trees may become ecosystems in their own right.
181
182
Regolith Science
Other organisms such as termites or ants may be organised into groups or colonies. 8.3.1 Biochemical uptake and redistribution of regolith materials by plants Plants include the largest sedentary organisms – trees – and, as a result, their effects can be seen at the local and meso scale and on a timescale allowing direct observation by humans. However, and despite the obvious feedback between organisms and the soils they depend upon, there has been little research on how organisms affect regolith characteristics – particularly the biochemistry and geochemistry – and especially in Australian landscapes (Hamilton 1972; Leonard and Field 2003, 2004). Rather, most studies aim to understand plant physiological processes, such as nutrient uptake, where essential and metal elements are either in deficit or in excess (Gadd 2004; Fomina et al. 2005), or where they are only biologically available (bioavailable) in specific forms. For example, most plants require Fe3+ for their nutritional requirements rather than Fe2+, but Fe3+ is relatively insoluble in soils (Figure 5.7). Some groups of metals can be regarded as both biologically important and derived from weathering (K, P, Ca, Mg, Zn, Cu, Ni, Co, Fe, Mn and Si) and these appear to vary less across organisms and space – particularly when compared with the non-nutrients (such as Al, Na, As, Ag, Au, Pb and U). The important elements range from highly deficient levels, where they are individually or collectively so low as to be limiting factors, to levels so high that they are toxic. In addition, plants of different genera take up a variety of elements at quite different rates and in different ionic and chelated forms (Leonard and Field 2004). For example plants, and plant parts, that are highest in Fe may or may not be the highest in Mn. Plants can be divided into accumulators (and hypo accumulators), passive plants and excluders (Baker 1981). Biogeochemical cycling, geobotany, bioprospecting and biointrusion
Of particular interest to this study is how trees alter soil mineralogy and geochemistry as a result of biogeochemical cycling processes, such as rainfall modification through stem flow and canopy drip,
Rainfall, dust: N, C, O, H, S, F Metals and salts leached from leaf / wood / bark surfaces Stemflow and canopy drip deliver dissolved nutrients to soil surface for uptake
Water uptake plus minor inputs from parent material
Nutrients translocated to areas of active growtht Nutrients added to soil as leaf, wood and bark litter Soil fungi and bacteria decompose litter to give humic acids + metal ions for re-uptake Small leaching losses to groundwater
Figure 8.8: A simplified biogeochemical cycle based on a single tree.
litter fall and its decomposition (Figure 8.8), and the alteration of the underlying soil by uptake of metal nutrients (Figure 8.9) (Field 1983; Gilkes 1998; Gobran et al. 2001). However, from a geochemical-exploration point of view (Section 13.8), very little is known about the processes of uptake, distribution, redistribution and loss from plants in Australia. Plant biochemistry reflects the chemistry of the materials in which plants grow. Hence, geobotany, bioprospecting, biomining and biointrusion are all fields of active research. Geobotany can be defined as the study of the relationships between the geochemistry (and mineralogy) of the substrate lithology and overlying regolith and the biochemistry of the different organs of the plant. Bioprospecting is a process whereby plant organs are sampled within an area to search for signs of concealed mineralisation. Because plants draw on large quantities of regolith, soils and groundwater within the rooting zone, they act as integrators and concentrators. The use of accumulator plants can increase the likelihood of finding geochemical anomalies because the plants actively collect particular elements and concentrate them in specific organs (Section 13.8). Work by Hulme and Hill (2003) has begun to differentiate amalgamators (usually perennial plants that sample a wide area of regolith) from penetrators (plants with deep tap root-type accession in the regolith) and to establish geobotanical associations of widely dispersed plants that can be used in geobotanical mineral exploration. Biomining uses
183
Regolith and biota
A
B Acacia 2. 0
2.0
0
3.0
Depth 0.5 (m)
3.0
1.0 0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Al % 1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
B
C Eucalypt
0 2.0 3.0
3.0
4.0
2.0
2.0
Depth 0.5 (m)
5.0
1.0 5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
Distance (m) Figure 8.9: Distribution of Al (%) in the sidewall of a costean dug from the tree bole of an Acacia falciformis at A to an Eucalyptus mannifera at C. Areas of marked concentration or depletion generally coincided with the presence of larger roots (after Leonard and Field 2003).
plants to concentrate elements – usually metals from otherwise sub-economic ore deposits – into plant organs which are then harvested and ‘mined’ (commonly by ashing and then extraction). Biointrusion is the study of the ways in which plant roots (and other organisms such as vertebrate and invertebrate bioturbators) access materials that have been buried or sealed off to prevent leakage or contamination of the environment by toxic material. The role and efficacy of clay seals on waste dumps and tailing dams can be radically affected by colonising the surface with different plants and their concomitant ecosystems.
whereby nutrients are redistributed within the leaves, twigs, branches, bole and roots in response to growth points, atrophy and abscission, so as to maintain critical elements within the organism for its advantage and ongoing growth (Banks 1989; Little et al. 2003; Leonard and Field 2004; Heinrich and Banks 2006). Additional adaptations include:
s s
Species-specific differentiation and internal redistribution
For some time, the different paths of elements into, within and through the biota have been recognised. These paths vary between even closely related organisms. The ability of many Australian species of plants to internally recycle elements is also well known,
s
very deep roots (for example Eucalyptus camaldulensis down to 10 m vertically: Davies 1953; Hulme and Hill 2003) very widespread rooting systems (out 20 m from the bole and well beyond the generally accepted distribution under the drip zone: Mylius 1992; Dexter 1967) extensive shallow root mats and fungal mat associations (Rao 2005).
However, the variety of elements and the ways in which they are transferred and stored is not well understood (Leonard and Field 2003; Raapana and Field
184
Regolith Science
2006). The plants that accumulate high levels of particular elements that may be toxic in other plants (such as Ni) are being researched because these accumulator plants (amalgamators or penetrators) represent options for industry to mine biologically (biomine) by growing particular plants on sub- economic mineralisation or stock piles and subsequently harvest the plants and process them as ‘ore’ (Baker 1981). The recent research of Leonard and Field (2003; 2004), Moerkerken (2005) and Raapana and Field (2006) demonstrated the differences in uptake and redistribution between species. In particular, whether or not a plant species fixes N2 becomes a major factor in regolith bulk chemistry, as do root architecture, soil organic matter and the depth to which weathering is effective. Some organisms require more or less of each element and concentrations do not always reflect the generally accepted levels for macro or micro nutrients (and non-nutrients) – particularly when concentrations in individual organs are examined (Raapana and Field 2006; Leonard and Field 2004). In terms of redistribution within a tree, Leonard and Field (2004) have shown that the individual tree species is very important (including root architecture, rooting depth, rooting volume and the distribution of roots versus the regolith horizons), but Hulme and Hill (2003) have shown that, in addition, season and proximity to rainfall events in the arid zone – and the regolith in which the tree is growing, including source lithology, groundwater chemistry, regolith geochemistry, and landform setting – can all contribute to plant uptake and distribution within the plant. Roach and Walker (2005) also demonstrated the importance of aspect and season for sampling vegetative organs. All these characteristics also seem to contribute to the processes of redistribution within the plant, and also whether or not particular elements are macro or micro nutrients, whatever their relative abundances in the regolith on which the tree is drawing. Both Leonard and Field (2003) and Hulme and Hill (2003) showed that Zn (an important micro nutrient) does not seem to be differentially redistributed, and that other element concentrations can vary between sites even if only a single plant species is considered. On the other hand, Zn was shown by Reid et al (2005) to have higher concentrations in spinifex over mineralisation.
Hulme and Hill (2003) concluded that many elements were preferentially concentrated in particular organs of the River Red Gums. They also showed that concentrations of elements, such as As, can be used in addition to Au for Au prospecting. Leonard and Field (2004) demonstrated the different ways that trees deal with macro and micro nutrients. Thus, K is not withdrawn from Acacia leaves before abscission to anywhere near the extent that it is in a eucalypt; and the converse is true for Fe in the branch tissue, but not in the bole (Figure 8.10). Although Hulme and Hill (2003) discuss a simple model of uptake that does not include microorganisms, it is clear that the effects on elemental, ionic and complex concentrations, the affinity of these to regolith particles, and bonding and synergistic and antagonistic reactions, are amplified if microorganisms are interposed at the interface between roots and regolith (Little and Field 2003; Little et al 2005b). Of course, individual members of the biota and the regolith interact in a multi-dimensional space, with cause and effect becoming bidirectional; that is, a two-way feedback loop is created. The niche that an organism finds most favourable (has a competitive advantage in) becomes the niche that the organism inhabits and, in so doing, creates feedback mechanisms, whereby that organism enhances the very conditions that make it most competitive in its niche: is this a chicken or egg question? Aluminium, iron, manganese and silica mobility in the presence of biota
Metal oxides are ubiquitous in the regolith and are very significant and reactive molecules, particularly in acidic to neutral environments. However, organic compounds exert very strong controls on the environment – and on the metal oxides, their formation, and surface properties (Kwong et al. 1977). For example, the microbial oxidation of Mn can produce Mn oxide coatings on soil particles up to 5–10 times faster than abiotic oxidation (Tebo et al. 1997). The metal oxides have important reactions with soil nutrients, such as P, and therefore have enormous influence on ecosystem functioning and biomass. The extraction of nutrients such as P from the bulk matrix often leads to the precipitation of metal oxides such as Mn, Fe and Al on the sheath of fungal
Regolith and biota
phytolith is most common – and varies in size, shape and ornamentation, depending on the plant part (stem, leaf and root) and taxa. Generally, phytoliths have a similar distribution in regolith to organic matter (Alexandre et al. 1997), although specific patterns develop in different soil types (Hart and Humphreys 2003). In general, most biogenic Si is in the form of phytoliths, which seem to form two pools – those that are being recycled (about 90%) and those that remain in the soil as phytoliths for much longer periods of time (Alexandre et al. 1997). In systems that are high (or even saturated) in Si phytoliths appear to remain for longer and Si re-precipitation may take place close to the base of the active soil profile. In the case of an equatorial rainforest, the uptake by plants of Si increases the weathering and breakdown of aluminosilicate minerals, but does not increase the denudation rate because the Si is being cycled by the biota (Alexandre et al. 1997). Phytoliths represent a direct impact on regolith biogeochemistry, with the control over alumino-silicate weathering by uptake of Si by plant roots (remove the products and move the reaction forward), the relocation of this Si onto, and then into, the soil horizons, and then its redistribution back down through the regolith.
(a)
Young leaves Old leaves Green twigs Small branch - all Medium branch - heartwood Medium branch - sapwood Medium branch - phloem Medium branch - all Large branch - heartwood Large branch - sapwood Large branch - phloem Large branch - all Bore cole - heartwood Bore cole - sapwood Bole - bark Bore cole - all Medium root >3mm wood Medium root >3mm bark All fine root <3mm 0
4000
K ppm
8000
(b)
Young leaves Old leaves Green twigs Small branch - all Medium branch - heartwood Medium branch - sapwood Medium branch - phloem Medium branch - all Large branch - heartwood Large branch - sapwood Large branch - phloem Large branch - all Bore cole - heartwood Bore cole - sapwood Bole - bark Bore cole - all Medium root >3mm wood Medium root >3mm bark All fine root <3mm 0
200
400 Fe ppm
Acacia
600 Eucalypt
Figure 8.10: Distribution and redistribution of elements within the physiology of trees of two different species – Acacia falciformis and Eucalypt mannifera – for (a) the nutrient element K and (b) the micro nutrient Fe (after Leonard and Field 2004).
hyphae, and on the surface of root hairs – demonstrating the power of the biochemical processes fuelled by the biota (McLean et al. 2002). Recycling of elements, such as Si and Al, between the soil and trees has been documented by Gilkes 1988; Figure 8.11). Phytoliths
Phytoliths (often, and literally, called a ‘plant stone’) are microscopic crystalline formations made up from a variety of elements and minerals (Hart 2001) that occur in various parts of many plants. The Si (or opal)
8.3.2 Redistribution of regolith materials by plants and animals Plants redistribute regolith materials in a number of obvious ways, but, in general, some of the processes responsible for the greatest quantities transferred are far less obvious. The chemical transfer of materials in solution by nutrient cycling, the uptake and storage of regolith materials in solution and storage in organic solids, and the control of biotransfers (even in the absence of vegetation such as following a conflagration) probably amount to the transfer of exponentially greater volumes of material, than the more obvious physical transfer. The latter also has many processes of redistribution, including root growth, that proceed very slowly making them less obvious, although the lifting of paths and general bulging upwards of the soil surface close to the boles of trees are easily measured once identified.
185
186
Regolith Science
Biogeochemical cycle of Si Precipitaion 2 kg Si/ha/yr
Biogeochemical cycle of Al Precipitaion 0.5 kg Al/ha/yr
Litter fall 22-25 kg Si/ha/yr
Stem flow <1 kg Si/ha/yr
litter fall 1.5 kg Al/ha/yr
Stem flow 0.5 kg Al/ha/yr
Vegetation 180-190 kg Si/ha
Vegetation 2 kg Al/ha
Organic matter 100 kg Al/ha
Organic matter 450-1400 kg Si/ha/yr
Soil solution 1 kg Si/ha Weathering and exchange 3-30 kg Si/ha/yr Soil minerals 2,000,000 kg Si/ha
Mineralisation of organic matter + dissolution of phytoliths 3-30 kg Si/ha/yr
Leached to subsoil <1-26 kg Si/ha/yr Neoformation <3-<30 kg Si/ha/yr
Plant uptake 2 kg Al/ha/yr
Top metre of soil
Top metre of soil
Plant uptake 26-28 kg Si/ha/yr
Soil solution 0.1 kg Al/ha Weathering and exchange 66 kg Al/ha/yr Soil minerals 1,000,000 kg Al/ha
Rate
Leached to subsoil 5 kg Al/ha/yr
Mineralisation of organic matter + dissolution of phytoliths 5 kg Al/ha/yr
Neoformation 60 kg Al/ha/yr
Stock
Figure 8.11: An approximate biogeochemical cycle for (a) Si and (b) Al in the solum and vegetation of temperate deciduous forests (after Gilkes 1998).
Root growth
Root growth and expansion occurs as a result of rapidly proliferating cells, and their subsequent expansion, at the tip (apical meristem), immediately behind the root cap and, to a lesser extent, along the sides (lateral meristem) of the roots and root hairs causes the regolith material to be moved upwards and to the sides with considerable force (1.45 MPa axially and 0.91 MPa radially; Bennie 1991) (Figure 8.12). Material usually doesn’t return to its initial location at any time in the future on even slightly sloping land (Figure 8.13 and Section 8.3.5). Tree fall
Tree fall creates a micro topography of pit and mound on slopes, and Norman et al. (1995) showed that the steeper the slope the greater the down-slope (and not back into the pit) movement of regolith material, while pits also became more elongate than round on steeper slopes (greater than 45 degrees). Tree fall breaks up, and brings to the surface, bedrock, saprolite and sub soil (B horizons), increases
Figure 8.12: Wedging apart of rocks in regolith by large roots and tree boles.
Regolith and biota
(a) As biotic action displaces regolith material (root grows, animal burrows) the outward transfer (y) faces less opposing force than any of (x) which would result in compaction;
y x x x
z
(b) Net outward transfer of material resulting from forces in (a) is perpendicular to the surface (z);
(c) Over time the transferred
z
material ‘falls’ back due to gravity (g) giving a net downslope movement (n); and
Figure 8.14: An example of the root plate from a large fallen eucalypt.
1956; Putz 1983), but the process seems to take several thousand years (Gabet et al. 2003). In fact, Osterkamp et al. (2005) have shown that partial rock veneers from tree fall (or root throw) can stabilise slopes in a subalpine setting over periods of several hundred years. Thus vegetation may not only stabilise a landscape during its life, but also once toppled and decomposing. Litter fall and litter dams
w
f
n z
(d) The ‘collapse’ of the biotic expansion leads to further net movement (f) leaving a depression which will slowly fill from upslope (w). JFf002-08
Figure 8.13: Down-slope transfer of regolith material by the growth and death of roots.
the heterogeneity of almost all soil characteristics and soil processes, such as metabolic and respiration rates, and may destroy (or sometimes enhance) soil horizonation. The pit and mound topography may have a micro relief of up to 0.5 m above and below soil level with mounds and pits being round to elliptical in shape and up to 5 m in diameter on average (Figure 8.14). Coverage seems to range widely, with northern hemisphere forests having up to 40% of the landscape covered by this micro topography (Denny and Goodlett 1956), while less than 1% is reported under tropical conditions (Putz 1983). The pits appear to fill more rapidly than the mounds flatten (Denny and Goodlett
Litter fall is an integral part of nutrient cycling, but is a much less obvious part of the effects of vegetation on the formation of regolith (Section 8.4.4). Its impacts include the redistribution of regolith components from depth to the surface (nutrients, non-nutrients, water, alumino-silicates, and so on.) brought up within the vegetation from shallow and deep roots. Litter dams have been described in recent decades by workers such as Mitchell and Humphreys (1987), Eddy et al. (1999) and Howell et al. (2006). Once flow is impeded by litter, there is a pervious dam or weir against which more litter is deposited, along with alumino-silicates and living and dead organisms. As a result a micro- through to meso-patterned ground develops, with varying areas of bare ground or fresh sediment. Litter dams represent an important surface storage for both moisture and organic and aluminosilicate materials and can substantially reduce the movement of all materials to drainage lines. 8.3.3 Bioturbation and biotransfer Bioturbation is the physical rearrangement (moving, sorting, mixing, churning, compacting, ingesting and
187
Regolith Science
B and C horizon materials cemented into mound with organic matter and released after mound abandoned
B and C horizon materials redeposited on soil surface, reincorporated into A horizons by wind and water
Organic matter from soil surface imported to mound
A horizon organic matter + resistates B horizon organic matter + resistates
Tree fall Ants
Animal burrow
Termites C horizon
Figure 8.15: A schematic diagram of common forms of bioturbation and the soil horizons affected.
comparable to the corrected soil formation rate of from 9 to 27 m3/Ma (Heimsath et al. 2000; Wilkinson et al. 2004). Individual and local processes, such as tree fall (700 m3/Ma) and wombat burrowing/mounding (100 m3/Ma), have rates an order of magnitude higher (Figure 8.16). The rate of addition of material to the surface by all these processes far exceeds the rate of erosion by a factor of three or four (for example, Anderson 2001). In addition to bioturbation sensu stricto, many biotic effects enhance or retard other less- or nonbiotic processes. For example, the production of loose regolith materials from burrowing provides easily erodible material for slope-wash or gully erosion.
0.14
Minimum
0.12
Maximum
0.10 0.08 0.06 0.04
Termites
Ants
Echidnas
0
Treefall
0.02
Wombats
digesting) of regolith and soil by the biota – both from within and acting externally. In common usage, the term is applied to animate activity, with the general exception of the inclusion of tree fall. Biota exploit the regolith for nutrition, for protection and shelter and, in the process, create a new bio-fabric. Bioturbation is most obvious when it results from macro (vertebrates) and meso fauna (invertebrates) (Figure 8.15). As a result of bioturbation, regolith particles are dislodged, dislocated, built up (aggregates), broken down, reworked biologically, biochemically, chemically and physically and relocated to become loose, or compacted and/or cemented and more or less resistant to erosion. Some authors (for example, Johnson et al. 2005) use terms such as ‘conveyor belt’ (ants, termites and worms), ‘mix master’ (mole rats and pocket gophers), ‘crater maker’ (wombats and rabbits), and ‘mound builder’ (prairie dogs and viscachas) to describe the activities of organisms. While these actions, and therefore the terms, may sometimes be apt and descriptive, they also somewhat misleading because most bioturbator species in fact carry out combinations of more than one of these processes (not to mention varying levels of biochemical alteration as well) and the terminology is not mutually exclusive, so the additional jargon will not be used here. Compared with the rates of other regolith processes such as soil formation, bioturbation rates are significant: turning over an average of 30 m3/Ma (30 cm of topsoil in <10 000 yrs; Anderson 2001), which is
Rate of mounding (m3/ha/year)
188
Figure 8.16: Rates of soil formation and bioturbation (volumes of material moved) in a dry sclerophyll forest, Southern Tablelands, NSW (Field and Anderson 2003).
Regolith and biota
Organic Matter Dominant
winnowing by wind and water
surface stones
A1 very diffuse stones
A A2
bioturbation unable to move stones upward, so they "sink"
A3 possible stone line
diffuse stones
B1 B2 B B3
bioturbation unable to move stones upward, so they ‘sink’ tree fall moves stones to surface
B4
lighter no texture and vertical eluviation movement of stones by free root growth, death and nutrient uptake heavier texture and illuviation
Biofabric Horizons of organic matter gain and loss of alumino-silicates
Biofabric Horizons of accumulation of organic matter and alumino-silicates
probable stone line
B/C saprolite
C
saprolith
Some rock fabric
More rock fabric
Alumino-silicates Dominant rock
Full rock fabric JFf003-08
Figure 8.17: Regolith, soil and the concept of a biomantle.
Animals trample and compact regolith materials and concentrate flow along tracks, enhancing erosion; fauna add nutrients in faeces, urine and dead bodies altering nutrient cycles; and the combined biotic and abiotic processes may occur at the same time or sequentially (Carr et al. 2004). Changes to the processes of biochemical alteration also occur as a result of bioturbation in all its forms. The exposure of soil that was previously at depth (even very shallow depths) allows erosion by raindrop impact and concomitant leaching and oxidation in a more oxic environment. Material brought to the surface by earthworms as casts has a new microorganism load – substantially altering organic content (weathered, decomposed, digested, gut mucous, altered microorganism load, and so on). The story for termites is similar, with the addition of intestinal flora via their faeces and saliva which they add to the mate-
rial they use to construct the mound. The latter is chemically and physically altered so that it is cemented, will resist erosion and will support the growth of some additional organic crusts, lichens and bryophytes. Mammals and marsupials not only dig and burrow, they defecate and urinate and these processes cause localised, but intense, changes to a whole range of biological populations and biochemical transfers, stores and processes. The result of bioturbation is the formation of a biomantle (Figure 8.17). Nearly all forms of bioturbation lead to net downslope transfer (however low the slope angle; Figure 8.13), or net transfer in the direction of the dominant processes of wind or water movement (Figure 8.17). Bioturbation also leads to the creation of microtopography, and large (and very large) pores and spaces of preferential permeability all lead to changes in the movement of fluids (air, water and solutions)
189
190
Regolith Science
into, and through, the regolith. Air, containing CO2 and O, and soil water and dissolved ions all gain preferential access to larger volumes of regolith – increasing all forms of chemical (biochemical) weathering and other regolith processes. In this discussion, bioturbation will not include strictly physical processes (such as frost heave, frost shattering, wetting and drying, and shrink–swell) although there are many situations where these processes may be retarded (such as insulating organic matter and freeze–thaw processes) or augmented (addition of organic colloids and cations to the shrink–swell process). ‘Bombturbation’ (Hupy and Schaetzl 2006: the effects of armament explosion) and other human induced changes, such as ploughing, ripping, mounding, creating temporary or permanent beds, clearing of vegetation and the effects of domestic livestock, will also not be considered in this chapter. Bioturbation by vertebrate fauna
Practically all vertebrates cause bioturbation. A brief list of the macro fauna in an international context (vertebrates, including warm and cold blooded) implicated (from a brief review of the literature) in bioturbation shows the diversity: aardvarks, anteaters, armadillos, badgers, beavers, canids (coyotes, dingoes, dogs, foxes, jackals, wolves), chinchillas, elephants, felines, fossorial animals such as moles, mole rats, and gophers (pocket and other sorts), ground squirrels, kangaroo rats, lizards, skinks and monitors, marmots, megapodes (mound building – incubator birds), prairie dogs, rodents of various sorts (mice, rats, and so on), snakes, susliks (a type of ground squirrel), tortoises, tuco-tucos (a type of rodent or hamster), viscachas, and viverrids (civets, genets, mongoose). While in Australia an even more eclectic collection includes bandicoots (of various types) bettongs (various), bower birds, brush turkeys, dingoes, echidnas, goannas, lizards and skinks, lyrebirds, macropods (kangaroos, potoroos, wallabies, and so on), mallee fowl, (the possibility of) marsupial moles, numbats, platypus, snakes, stick-nest rats and wombats. In fact, the numbers and varieties of fauna actively taking part in bioturbation are enormous: for example, 15% or so of all the Australian birds (White 2003) bioturbate soils or regolith:
‘... ten species extensively probe the soil during feeding; 11 rake the forest litter; 26 nest in soil burrows; 44 nest in shallow scrapes in the ground, often in colonies; 3 build incubating mounds; and 10 construct mud nests in trees and caves.’ Most recent investigations of bioturbation by meso and macro fauna were undertaken in Africa and the Americas (Ekundayo and Aghastise 1997; Whitford and Kay 1999; Aufreiter et al. 2001) and often outside the geosciences: for example, a substantial review of mammals by zoologists Reichman and Smith (1990). In Australia, studies have demonstrated the contributions of bioturbation to soil formation in the context of deep, old weathering profiles across the Australian continent: for example, Humphreys (1981, 1989, 1994), and Mitchell (1985, 1988) and their co-researchers (Cowan et al. 1985; Humphreys and Field 1998; Wilkinson et al. 2003) and, more recently, Field and co-workers (Field and Anderson 2003; Field 2003, 2004, 2006). Large bioturbators, such as wombats, rabbits, echidna, reptiles, lyrebirds and scrub turkeys all create obvious regolith changes in human timescales and hence they have been the focus of this research. Many vertebrates have a distribution that relates to particular regolith and soil characteristics and/or types: for example, regolith moisture conditions (and landscape position; wombat burrow versus floodplain distribution; Anderson 2001), texture, depth or stoniness, or the distribution of resources on which to feed such as particular invertebrates in well-drained soils. Animals can also create the circumstances that cause other geomorphic processes to change the regolith: for example, beavers construct dams that cause sedimentation, avulsion and the creation of new channels; and various animals burrow into stream banks, leading to bank collapse and stream changes, both within and alongside the drainage lines. Vertebrate bioturbation is broken down into three broad classes: 1. the refugia - burrows, dens, warrens, nests and mounds (Box 8.1) – essentially long-term and fairly stable structures 2. foraging structures, such as tunnels, diggings, scrapes and scratchings (Box 8.2), which are an
Regolith and biota
Box 8.1
Refugia – burrows, dens, warrens, nests and mounds
Animals seeking shelter and protection make use of available hollows, small caves, ready-made depressions, and so on, or can construct or augment places to shelter, to raise young and stay safe from predators. Some animals may have one or more tunnels or nests (for example, male wombats have two or three dens or burrows, with a female inhabiting each) or different types of tunnels (for example, aardvarks dig some tunnels for food and other burrows for temporary or permanent nests; Smithers 1983). There is a continuum through two or more tunnels and/or openings, with animals such as suricates and mongooses, to those animals that either singly or in pairs, or as communities, build much more elaborate multi-tunnel, multi-entrance burrow systems. Material removed from tunnels and burrows becomes mounds. Some of the most elaborate burrow systems are dug by rabbits and rodents, which have the highest bioturbation rates on a body weight basis. These warrens and burrow systems can have multiple entrances, complex interconnections below ground, and several nests and be inhabited by families or small communities of related animals. These complex systems can have a series of characteristics such as slopes, interconnectedness, arrangement of nest sites, shape and orientation of openings that are quite specific to the animals constructing them. They can also be built over time by the same individual or by groups as the family expands (such as prairie dog systems; Longhurst 1944; and evolving suslik burrows in Russia; Ognev and Tomilin 1947). North American kangaroo rats create mounds (30 cm high and up to 4 m in diameter) that take
almost continuous process as animals seek food and other resources throughout their lives 3. the surficial effects of grazing animals (Box 8.3). Each of these forms of bioturbation has the potential to initiate or enhance other geomorphic and regolith processes, such as the interaction of overland flow with litter dams, and rill formation, but larger scale processes, such as piping and gully initiation, often result from tunnels and burrows, and bank collapse along drainage lines.
around 2 years to construct and begin to break down after a year or more of non-habitation of the burrow (Schooley and Wiens 2001). The mounds are spaced at from 2.5 to 10/ha and markedly alter the spatial patterning of soils and regolith and, in so doing, the distribution of vegetation and other fauna. The pattern of distribution of mounds is, in turn, affected by localised topographic conditions (such as regolith and soil texture and drainage). Mounding can reach extremes, such as the Mima mounds of the Western USA (up to 2 m high by 25–50 m in diameter), which are made of soil and other regolith materials by gophers. Bioturbation varies from animals that produce long-lived and often complex networks of burrows and chambers to those that have relatively simple and short-lived excavations for refugia, such as most moles. Slightly more complex, but still only linear, burrows are constructed by some animals such as the Australian platypus, which digs a convex upwards burrow with the entry under water in a pool in a creek or river, and with a dry nest above the tunnel entry height. Many other animals, such as Australian wombats, dig tunnels from the surface sloping down into the regolith and with a simple larger cavity at the deepest point. Wombat burrows can range in length from 1 metre to more than 13 metres, and in diameter from 50 centimetres to almost a metre. The construction of nests and nesting behaviour is usually limited to birds; this can range from simple nest construction on the surface to the mound nests produced by megapodes (1 m high, 2–3 m wide and 4–5 m long).
Bioturbation by invertebrate fauna
Soil and other regolith material provide an environmentally stable, nutrient-rich, moist habitat for invertebrates. As a result, there are anywhere from a few tens to tens of millions of individuals beneath a square metre of surface litter (Viles 1988; Gabet et al. 2003) and 10–15 000 kg/ha (Coleman and Hendrix 2000). The quantities moved or bioturbated by invertebrates are considerable, and frequently compare favourably with rates of regolith and soil production and erosion. The macro invertebrates, such as soil-living bees, beetle
191
192
Regolith Science
Box 8.2 Foraging – tunnels, diggings, scrapes and scratchings Many animals forage for food on and within the surface horizons of the regolith, and others may dig for water adjacent to springs or in drainage channels. Some animals will also create dust or mud baths and coat their skin by rolling or wallowing (such as hippopotamuses and water buffalo), fluffing (poultry and many birds), blasting (elephants using their trunks), or throwing material onto their skin (bison, cattle) to protect themselves from parasites or sunburn. Tunnels and burrows are dug in order to move through the regolith, to find and recover food, and to avoid obstacles. Many animals dig among the organic materials in the litter and A1 horizons, such as ground-dwelling birds, while others, such as dogs and cats, create shallow excavations to bury food, and even their own excreta. Surface and sub-surface bioturbators have multiple effects on regolith processes, including predation of invertebrates, eating of fungi (such as native truffles) and microorganisms, habitat alteration, and damage to living organisms, such as plants, mosses, and lichen, and impacts on parts of the regolith, such as surface crusts. In so doing, bioturbators can have marked effects on soil leaf litter decomposition, the structure of the surface soils and the larger scale processes such as infiltration.
larvae (such as scarabs), cicadas, crayfish, crickets, dung beetles, earthworms, mites, nematodes, spiders (trapdoors and funnel webs in particular), springtails and, particularly, ants and termites move through the regolith, pushing materials aside, creating tunnels, chambers, mounds and nests – some of which are obvious, but these organisms have many effects that are not seen without measurement. These effects include changes to permeability, destroying and creating regolith horizons and layers, and causing biochemical weathering within their gut, as well as the effects of saliva, mucilage excretion and digestive excreta. Invertebrates can bring large volumes of relatively unweathered materials (or subsoil) to the soil surface, which can then be redistributed across the landscape
Box 8.3
Animal tracks and terracettes
The movement of fauna across slopes transfers litter and topsoil, compacts surface soil and leads to the creation of tracks, litter dams and terracettes. The effects of native herbivores and, to a greater extent, domestic livestock, compact surface soil and change bulk density and other soil characteristics, although these changes are less conspicuous than mound building and foraging. Surficial effects on the redistribution of organic materials by the biota provide an interesting example of how biotic processes in regolith formation can interact to give more than a sum of the parts. Kangaroo and wallaby tracks (many other animals use these tracks, but the nomenclature attributes them to kangaroos and wallabies) move litter aside: exposing the surface aluminosilicate/organic matter horizon – the A1 soil horizon. This shallow depression forms a mini bank, bund or long, linear litter dam, and channels water across-slope at a near contour level until it overflows and concentrates the water in a mini rill. The impact of animals trampling and causing erosion is highest on the floodplains, along drainage lines and on stream banks. Active erosion and destabilisation, leading to additional erosion, occurs as animals access water and seek to cross streams. Banks are often steep – resulting from on-going fluvial processes of channel evolution – and animals clambering or climbing the banks cause substantial sediment production. Animal excreta, and even the bodies of dead animals, have an effect on the regolith in and around the stream banks.
via the actions of water and gravity. For example, earthworms, ants and termites may be responsible for from 0.5–10, 0.004–1.8 and 0.013–0.4 m3/m2 /ka, respectively (Mitchell 1988; Whitford 2000). Obviously, conditions do not favour all three organisms at the same time (termites dominate dry-seasonal environments, whereas earthworms dominate under humid conditions) so these rates are not additive (see Boxes 8.4 and 8.5). Many invertebrate animals move within the soil by forcing themselves into interstices and pushing
Regolith and biota
Box 8.4
Nests and mounds: ants and termites
Miklos (1999) estimates that ants and termites together can bring 20–30 cm of soil to the surface each 1000 years. This exposed material is made more susceptible to chemical weathering (for example, Field and Anderson 2003). Ants and termites can concentrate organic matter in tunnels and mounds deep in the soils and regolith at the expense of inorganic minerals that they transport to the mound exterior. This generates complex fine-scale networks of nutrient sinks and sources across the landscape, and transports relatively unweathered regolith materials to the soil surface. Ants and termites are particularly active in tropical, semi-arid and arid ecosystems (for example, Holt et al. 1980; Cammeraat et al. 2002). While the ant or termite mound is maintained, these materials are essentially retained – with soil organic matter and mineral nutrients being released only slowly back in to the landscape. This release becomes more rapid after the mound has been abandoned and begins to disintegrate (Holt et al. 1980; Cammeraat et al. 2002). The role of termites in mixing and fractionating both alumino-silicates and organic materials is extremely important in arid and semi-arid through to seasonally dry environments. Their role in decomposing otherwise inedible organic materials, such as cellulose and lignin, is critical to nutrient cycling and soil organic C in these drier environments. Termites can exponentially accelerate the breakdown of otherwise resistant materials, including resinous, high-
particles and aggregates apart, while others excavate by eating or digging and can move quite large quantities of material from the voids they are producing to form mounds or nests. At a landscape scale, the mixing or slow stirring of the regolith and soil causes displacement outwards towards lower densities and pressures (and may leave a void) – usually perpendicular to the surface. On a slope, the subsequent ‘fall’ back to the surface (or into a void) is likely to be vertical under gravity. The result is a slow creep of material down-slope, which is analogous to particle diffusion (Figure 8.13). Heimsath et al. (2002) showed that grains keep being exposed at the surface, only to be buried again and so a net down-slope trajectory is
wax and high-lignin timbers, such as those of the Australian eucalypts, into rapidly weathering organic materials. Termites are among the few organisms that can decompose eucalypt stumps and provide the basis for further fungal attacks on dead and dying timber because they can also function as a vector by transferring spores. Termite nests have a considerable effect on regolith hydrology: creating deep macropores and many minor interconnecting pores and chambers that allow exponentially higher quantities of water, air and gases, such as CO2, to move through the regolith and soil. Termites can reinforce horizonation when combined with other processes, such as erosion and winnowing, while breaking it down in other cases, such as duplex soils where termites essentially bring the B horizon – and in some regions the saprolite – up to form their mounds, which subsequently erode and wash and winnow out across the topsoil. Because termites can bring subsoil and saprolitic materials to the surface, and may even concentrate particular grainsizes and/or mineralogies, they can be used to sample the regolith in bioprospecting (Petts 2006). As with termites, there are arguments suggesting that ants destroy horizonation, whereas ants may be responsible for promoting, if not causing, texture contrast soils. It seems to depend upon the substrate regolith material and the environmental conditions, as well as species and community organisation.
produced, despite a Monte Carlo pattern of movement in the first place. Some invertebrates (such as earthworms: Box 8.5) actually ingest alumino-silicates while the majority only consume organic matter as they move through the soil (such as termites eating roots), but gut microorganisms affect weathering, both in the gut and, subsequently, in the soil – particularly in faecal material and castings. Dung beetles produce a nest filled with acquired dung formed into neat balls, which also contain dung beetle eggs (or an egg). The nest may be connected to the surface by a cylindrical tunnel back filled with dung by the adult on departure. Isolated egg-shaped cocoons containing dung beetle eggs
193
194
Regolith Science
Box 8.5 Since the early work of Darwin (1837, 1881), it has been clear that earthworms can be said to create topsoil both biomechanically and biochemically. Earthworms can transfer considerable quantities of material from within the topsoil to the surface and even bury A and litter horizons in the process. In humid environments, earthworms may dominate the invertebrate fauna because of their relatively large size and their abundance in soils with relatively high organic matter and high levels of activity in temperate and warmer environments. Earthworms have a number of effects. First they create voids: excavating tunnels and ingesting soil materials in the process. The tunnels can be up to 5 mm in diameter for northern hemisphere species – and more for some large Australian native earthworms. These tunnels become macropores – allowing the preferential movement of water and gases – and so increasing aeration, infiltration and drainage. Secondly, earthworms exude mucilaginous material that may help to stabilise the tunnel walls and support a range of microorganisms (Edwards and Shipitalo 1998), Thirdly, they process the ingested material through the gut – adding a vast array or microorganisms. In the process of consuming soil, the gut contents are subjected to both biochemical, biological and physical breakdown during digestion. The process also removes some nutrients, mixes plant and fungal fibres from the ingested soil along with secretions from the gut wall and com-
and/or larvae may also be located contiguously. Dung beetles are the best example of the incorporation of relatively fresh, inoculated organic materials into topsoils. Spiders produce burrows that are essentially vertical, widening upwards, which may, or may not, have a capping. Solitary ground-living bees construct flaskshaped chambers with a smooth internal wall, which are connected to the surface. Less-obvious processes include use of saliva and gastrolith (Box 8.6). 8.3.4 Consequences of bioturbation within the regolith Invertebrates that burrow and mound soil tend to produce a biomantle of topsoil, subsoil and overlying
Earthworms pacts the material – and additional microorganisms are added. Casts or faeces also contain intestinal mucus, which provides an additional substrate for all sorts of microorganisms while binding the material into very different, and more stable aggregates: all changes that will affect the biochemical reactions from then onwards. The resultant casts are bound together by all the secretions, derivatives and subsequent processes of microorganisms. A multitude of chemical reactions are promoted within and between organic materials, alumino-silicate minerals and living and dead microorganisms and can create organo-mineral bonds. Some authors argue that earthworms restrict ongoing soil forming processes in the actual casts (Edwards and Shipitalo 1998), while others, such as Sharpley et al. (1979), have shown that casting can increase erosion. Pastures with good levels of earthworm activity had sediment run-off loads almost an order of magnitude higher than pastures without earthworms (Sharpley et al. 1979). On the other hand, the latter results also showed that higher earthworm numbers encouraged infiltration and reduced nutrient loss (specifically N and P) from paddocks. Earthworms are generally described as being less important in Australia than Europe and North America due to lower latitudes, lower precipitation and general dryness of the continent – and very low levels of organic matter favouring bioturbation by other invertebrates such as termites and ants.
saprolite and, as such, are primary agents producing horizonization. On the other hand, uprooted trees break up bedrock, transport soil up and onto the surface and increase the heterogeneity of soil, and disrupt soil horizonation. Biomantles form as single layers in fine-fraction parent materials, and two layers when formed on mixed fine-and-coarser materials (Figure 8.17). Within the solum or biomantle, there is a complex interacting set of fast and slow, often countermanding translocations, taking place, as material is loosened and then transferred. The result can be several different forms of biostratification or horizon formation, or the destruction and homogenisation of layers. Bioturbation can, depending on the balance of organisms,
Regolith and biota
Box 8.6
Saliva, corpses, gastroliths, excreta and gut flora
There are many less-obvious interactions between biota and the regolith. For example, some birds construct nests out of mud and dung with or without saliva. The geochemical interactions between these chemically very active organic substances (enzymes, organic acids, chelates, inoculants containing complex and diverse microorganisms, and so on) and regolith are only postulated currently and require further research. The death of any organism and subsequent incorporation onto, and into, the regolith provides another example of incredibly rich organic material (both as organic substrates for more biota, plus inoculants, chemically aggressive and active substances, and so on) being added to the ongoing weathering of alumino-silicates. Gastroliths are stones swallowed by animals (usually alumino-silicates, but they can be formed in the stomach by precipitation – rather like human kidney stones), which were much more common in earlier geological periods – particularly in dinosaurs (aquatic dinosaurs may have used them for ballast) and large reptiles – but, even today, gastroliths may make up as much as 1% of the weight of crocodiles and alligators. They are also common today in seals and sea lions and almost all herbivorous birds: effectively providing the grinding medium (sand, grit, gravel, even cobbles, and shell grit, and so on) in the very muscular gizzard for physically reducing food material during digestion. Gastroliths may be highly polished (most of those recognised from the fossil record are) or may have little or no polish and be quite angular in modern birds. Gastroliths from dinosaurs can be several kilograms in weight, and modern ostriches can have stones up to 10 centimetres in length (Wings 2003). Gastroliths may or may not be found among the skeletons of dead animals, but are not often released from the gizzard of healthy, living animals.
either create horizonation or homogenise the biomantle (Figure 8.17). Whether biota actually incorporate saprolite and rock into the biomantle depends on the processes involved. Tree fall on shallow regolith brings fresh
and weathering material to the surface, while it is doubtful whether gophers or rabbits actually bring even saprolite to the surface. However, the production of regolith and soil does not require the physical transfer of rock or saprolite upwards and onto the surface. Many authors are mistaken in their belief that biota, such as gophers, do not accelerate the production of regolith or produce soil, because they do not burrow among rocks or dig up the saprolite (for example, Gabet et al. 2003; cf. Heimsath et al. 2000). However, virtually all bioturbation processes increase the infiltration of water and the incorporation of organic material through increased porosity, reduced bulk density and the mixing of litter from both the surface and the root biomass. Also, the rate at which rock weathers nearly always increases with shallow to moderate burial, because moisture and temperature conditions within the regolith are more conducive than exposed conditions on the surface to all biotic chemical weathering processes, as well as even strictly abiotic geochemical weathering – if such a thing exists. The only exceptions to such a generalisation are predominantly abiotic physical processes such as freeze–thaw or frost shattering. There is an argument that invertebrate bioturbators, such as ants, termites and earthworms, can turn finer material in the biomantle layer over to the exclusion of coarser particles (such as gravel and stones), and put in place a stone line or a texture contrast soil (coarser loams or loamy sands over clays or clay loams). A stone line is a non-geological, non-depositional (that is, nonstratigraphic) layer that can be considered to be an horizon in a pedological context. Some larger bioturbators, such as rabbits, wombats and tree fall, can also act to destroy the texture contrast by inverting soil layers or remixing them. However, others – because their activities are limited to only one or two horizons or layers – will enhance horizonation, rather than destroy it (Figure 8.15). There is a landscape zonation of specific bioturbators with particular environments. For example, wombats generally dig into softer floodplain and drainage line soils, rather than shallower, and more resistant soils with stronger texture contrast, and heavier textures in mid and upper slope positions (Field and Anderson 2003). Some authors go further: arguing that either climatic/microclimatic zones or
195
196
Regolith Science
vegetation types can be use to predict levels of bioturbation. Carr et al. (2004) argue that bioturbation reaches a maximum under temperate humid Californian grassland (about four times the volume on average; around 100 cm3/m2) when compared with shrubland or forest. The dominance of burrowing rodents in these sites may explain the result, but the sample size is very small. On the other hand, Anderson (2001) and Field and Anderson (2003) clearly showed that in a similar environment (temperate humid NSW) – but on another continent with very different biota – tree fall moved the most materials, and in grasslands there are no other bioturbators to replace these processes (Figure 8.16). Stone lines can be seen as a residue of bioturbation. As organisms cycle the materials they are capable of moving, some larger, denser particles or aggregates are left behind. If any sort of random walk process (not really random because most movement is at least partly upwards towards lower densities, aeration and/or away from saturation) can move some materials but systematically leave others behind, eventually a horizon, or layer with just that residual material, concentrates near the base of the bioturbation and a stone line is created (Figure 8.17). For example, North American gophers can produce a stone layer, just below the typical depth of burrows, from pebbles and cobbles that are too large for the animals to move during their burrowing and tunnelling activities (Johnson 1989). In addition, if a particle that is too large to be moved by bioturbators lands on the surface (dropping stone, gastrolith, bone, shell, or man-made article, such as shards, points, ceramic fragments), it will slowly, but systematically, be lowered under gravity through the biomantle (almost literally, as material is moved up and piled out on top of it) until it reaches the stone line. If erosion on the surface due to wind or water exceeds the rate at which material is being bioturbated onto the surface, then not only will the artifacts be left at the surface, but the surface is also likely to armoured with larger, resistant objects but it is likely to be lowered in the process. Stone lines (Johnson 2002), or a stone layer (Paton et al. 1995), do not form where there are no resistant larger particles (Figure 8.17). In a similar manner, and sometimes with the assistance of winnowing by wind and water, texture contrast soils are formed.
8.3.5 Biotic processes on regolith surfaces The regolith surface is the main interface between air and water – where a truly diverse biota has a dominant role. On fresh rock surfaces, the endolithic bacteria carry out forms of biogenic weathering – as do lichens. The following all alter the run-off or infiltration, while enhancing or retarding weathering:
s s s s
the surface effects of litter fall – the mulch effect of keeping the surface moist the biochemicals in litter that lead to hydrophobicity and run-off, or to aggregation and infiltration the fungal mats from symbiotic partnerships between the higher plants and microorganisms the cryptograms constituted from algae and fungi or algae alone.
The following manage erosion and sedimentation, while enhancing weathering:
s s s s s
litter dams root holes bioturbation pores lichens liverworts, and other bryophytes.
Biotic surface processes are critical to the functioning of most landscapes. Biogenic weathering, biofilms, biotic crusts, biomineralisation and endolithic microorganisms
Biogenic weathering is where ‘life processes’ dominate the formation and evolution of regolith. A biofilm is a complex aggregation of microorganisms, which is marked by the excretion of a protective and adhesive matrix and characterised by surface attachment, structural heterogeneity, genetic diversity, complex community interactions and an extracellular matrix of polymeric substances. Some, such as desert varnish, impede further weathering when they form; others, such as lichens, lead to enhanced weathering. Biotic crusts result from the growth effects of biota and reorganise the alumino-silicates in, and on, the surface of the regolith – often positively altering its stability, and changing the chemical composition and physical characteristics. Biomineralisation is a simple combination of the biota causing changes to alumino-silicate materials to produce new or secondary minerals.
Regolith and biota
Endolithic means living within, or penetrating deeply into, stony substances. An endolith is a (micro) organism (usually an archaeon, bacterium, or fungus) that lives, at least partly (fungi and lichens), within rock (coral, animal shells, or in the pores between mineral grains) and includes endolithic lichens. The overlap between biogenic weathering and discussion in earlier sections (for example, Section 8.2) is complex. Specific examples will be given here under each named process and then, where relevant, these will be referred back to earlier subsections. Biogenic weathering occurs within one or two years of exposure of a fresh carbonate surface (Hoppert et al. 2004) in humid temperate environments. It also takes place when lithobiontic organisms, both epilithic (external to the alumino-silicate mineral surfaces) and endolithic (or cryptoendolithic) species colonise the surface, induce and accelerate weathering and actively penetrate the rock independently of pre-existing pores, fissure or other inhomogeneities in sandstones and volcanogenic rocks – even in environments as cold as the Antarctic (Johnston and Vestal 1991; Guglielmin et al. 2005). Initial colonisation takes place as algae and ascomycetes penetrate the rock and slowly – over several years – a more complex colonisation pattern by lichens develops. The rates of weathering are very much climatedependent and also vary across lithologies. Biogenic weathering can also be demonstrated to occur in a laboratory setting (Cervini-Silva et al. 2005) using the same biogenic substances that are ubiquitous in soils (ligands such as oxalate, ascorbate, citrate and humic acids, and chelating agents) to weather P-bearing minerals, such as rhabdophane, to form more- or less-soluble organic complexes (ligand plus an ion) with both Ce and phosphate. Surficial root mats, fungal mats, surface hydrophobicity, surface crusts, and cryptogamic crusts
These phenomena are special cases of concepts that have been discussed already in this chapter or covered later. However, they also represent more intense examples, more extreme processes and ever greater influence of biota over regolith evolution. Surficial root mats are generally described for rainforest environments, where high inputs of precipitation and high temperatures lead to very rapid leaching of very
quickly decomposing organic materials on, and in, the surface soils. The vegetation has evolved dense shallow root systems (plus deeper systems, including tap roots) to efficiently absorb the results of the very high turnover nutrient cycle. Rao (2005) demonstrated that fungal mats, and underlying root mats, redistribute low levels of both nutrients and water in the surface soils under dry sclerophyll forests and woodlands on the Southern Tablelands of NSW. The ecological system biases the distribution of both nutrients and water, so that the eucalypts get the lion’s share and competitors are starved of both. Symbiotic fungi enhance the effects of the root mat. The fungal mat is hydrophobic – shaped into what resemble octagonal or hexagonal ‘tiles’ about 10–15 mm thick, lying immediately below the eucalypt litter and above the A1 horizon that contains the eucalypt root mat. The fungal mat appears to be an effective barrier to the germination of any plants – either simply by its aridity, but possibly also through allelopathic chemicals, including biphenols of various sorts. By so doing, this combination of regolith surface phenomena is used by the eucalypt overstorey to control competition for scarce resources. Biological soil crusts are highly specialised communities of living cyanobacteria and tiny mosses and/or lichens that combine with their by-products and alumino-silicate minerals to form a crust millimetres in thickness bound together by organic materials. The crust can reach centimetres in thickness and vary in their makeup to include cyanobacteria plus or minus combinations of liverworts, mosses, lichens and a variety of free-standing fungi. They bind the surface and may cover all areas not growing higher plants and create a considerable micro topography with their filaments of cyanobacteria and green algae. The latter migrate outwards on wetting, creating additional sheath material that binds the soil further, holding it together on drying. The biotic crusts stabilise surfaces, markedly reducing both wind and water erosion, promoting soil moisture retention, fixing N for plants, conserving other nutrients, increasing infiltration and soil water retention and promoting seedling germination and plant growth. The growth of the filaments from the fungi in the symbiosis further adds to soil stability, when they exude polysaccharides – aiding in the cementing together of aggregates of soil (as distinct
197
198
Regolith Science
weathering and regolith formation. They then expand using this explanation: ‘Fixed C released by the photosynthetic symbiont stimulates growth of fungi and other microorganisms. These microorganisms directly or indirectly induce mineral disintegration, hydration, dissolution and secondary mineral formation. … C sources … can suppress or enhance rates of chemical weathering by up to three orders of magnitude, depending on the pH, mineral surface structure and composition, and organic functional groups. Mg, Mn, Fe, Al and Si are redistributed into clays that strongly absorb ions. Microbes contribute to dissolution of insoluble secondary phosphates, possibly by release of organic acids. Below fungi-mineral interfaces, mineral surfaces are exposed to dissolved metabolic by-products. Through this indirect process, microorganisms can accelerate mineral dissolution, leading to enhanced porosity and permeability and colonisation by microbial communities.’
Figure 8.18: Early stage succession, as higher plants colonise weathered material shed by surficial weathering of a granitic face.
from lichens and mosses that bind soil with additional rhizines or rhizoids). Non-filamentous soil crusts can actually reduce surface roughness, reduce soil infiltration and restrict the entry of air, germination of seedlings and seedling emergence. The very different effects of soil biotic crusts are crust specific, and vary across environments as well as surface soil materials. Lichens and bryophytes
Although most bioturbation studies focus on ‘largescale’ visible processes, vegetation establishment and growth also have important physical implications for mineral weathering and pedogenesis. Mosses and lichens – through their rhizines or rhizoids, as well as cryptogamic mats with rhizomes or filaments – are able to exploit minute flaws and cracks in mineral grains and relatively unweathered rock surfaces. As these root-like bodies grow, they expand tiny cracks and allow greater water penetration and add organic matter to the developing weathering rind. Increasingly complex grass and herb communities, followed by shrubs and, eventually, trees, then become established, with roots that grow deeper and expand the cracks in the weathering parent material further (Figure 8.18). The lichen-based weathering model has been developed to a level where a number of authors now see it as a means to understanding biotic or biogenic mineral weathering in general (for example, Banfield et al. 1999). They state that microorganisms are known to increase rates and add additional processes to rock
8.3.6 Biota and regolith physical characteristics The incorporation of organic material changes regolith characteristics such as bulk density, horizonation, texture, structure, moisture content, porosity, permeability, sorptivity, and hydraulic conductivity. Exploitation of the weathered regolith by plant roots influences soil insulation, compaction and aeration – as well as soil water conditions and the microbial community. Organic matter virtually controls soil structure, which, in turn, controls infiltration and run-off. The biota have an overall controlling influence on regolith water content and transfer:
s s s
right from the very minor effects of tiny root holes left after the death of rootlets and root hairs, up to the largest roots and root macro pores from the myriad burrowing invertebrates and the micro though to macropores they leave behind from the effects of soil organic matter on soil aggregation and bulk density to infiltration and bulk soil water storage
Regolith and biota
s
from the minute processes of root cell exudation of water to hydraulic lift by huge trees.
Biotic surface conditions
As discussed above, biotic regolith surface processes can have seemingly opposite results: on the one hand, stabilising surfaces, increasing infiltration and thereby reducing run-off and erosion; and, on the other, reducing infiltration and increasing run-off. Living and dead biota insulate surfaces from heat and reduce evaporation, increase infiltration and enhance weathering. These controls are, in turn, affected by environmental controls (moisture and temperature, even if only at the micro scale) and the nature of the surface materials – producing a wide spectrum of likely scenarios, the complexities of which cannot be dealt with here. Bulk density, soil structure and soil or regolith horizons
Organisms seeking to exploit the regolith (particularly the organic component) for food and shelter rearrange the structure and fabric during the process. In doing so, they create a biofabric and/or a biostratigraphy. Once the organism has passed (or the root dies), a void is created that will stay open, fill with material from the passing organism, or collapse. Whatever happens, a heterogeneity is created that is biomechanically (and often biochemically) different and more likely to be aerated, be more porous and/or permeable to water and solutions, and have different nutrient and biological characteristics from that of the surrounding regolith. The result is a reduced bulk density, enhanced soil structure (Figure 8.19) and horizonation. The types and effectiveness of the variety of bioturbation acting on any site varies by vegetation and/ or ecosystem type, although there is little consensus as to whether particular systems, such as forests, have more or less than grasslands. Permeability, porosity, infiltration and drainage – roots holes, tunnels and burrows
Burrowing animals and insects, and recently deceased plant root systems, create passageways for air and water movement and thus change soil morphology. The pas-
Figure 8.19: Fungal hyphae bind soil aggregates (scale bar = 20 µm).
sageways, or krotovinas, formed by these processes create loose material on the surface and also a vast network of micropores (root hairs, micro fauna), macropores (earthworms, and other invertebrates) and mega-pores (large root death, vertebrate tunnels, and so on) that become back filled once vacated by organisms moving (animals) or dying (roots). These krotovinas are filled with materials that are lower in bulk density, contain more organic matter, and are more porous and permeable. Krotovinas are likely to be narrow and elongated, with varying levels of connectivity. In addition, they were shown by Noguchi et al. (1999) to connect with many other features such as:
s s s s s s
other former root pores faunal burrows of all sizes (infilled or collapsed but still preferentially permeable pores and burrows) dehydration and shrink–swell cracks soil inter and intra aggregate spaces saprolite and rock fissures and preferential flow paths soil horizon boundaries.
Moles in temperate humid environments have been shown to double soil porosity (Mellanby 1971), increase soil moisture and invertebrate numbers, such as earthworms (an interesting interaction between predator and prey and resultant bioturbation), while decreasing bulk density. Rodents, with their extensive burrow
199
200
Regolith Science
systems, create vast interconnected systems of open and back-filled macropores, which have huge impacts on air and water movement and, in turn, slope form and evolution. Earthworm tunnels generate macropores, which can increase soil porosity by 3–10 times (Edwards and Bohlen 1996). These macropores have high levels of connectivity within, and with the surface, and can increase infiltration (and reduce concomitant surface run-off) during rainfall while encouraging drainage and aeration once rain has ceased. Because organic matter reduces bulk density, it increases both the hydraulic conductivity and sorptivity of soil and increases the heterogeneity of soil physical characteristics. In addition, because almost all forms of bioturbation create pores, organic matter increases infiltration rates up to three times (Williams and Vepraskas 1994). Devit and Smith (2002) showed that root canals from shrubs in the desert not only carried water more quickly into the regolith, but also more deeply, while fibrous roots, such as those under native tussock grasses in eastern Australia, can cause a full order of magnitude increase in both hydraulic conductivity and sorptivity (Rath 1993). Hydraulic lift by vegetation
Hydraulic lift (or hydraulic redistribution) is a process whereby plants take water from depth and, using the xylem and capillary rise, bring water up through their roots into shallower layers where it may later exude from the roots and moisten the rhizosphere and beyond. Under Earth’s gravity it is physically impossible to lift water more than 10 metres by suction. However, the very fine capillaries in plant xylem exert much greater localised forces, and plants can lift water an order of magnitude higher. During the day, lifting is driven by the evaporation of water out through the stomata, as well as consumption of the water as the H and O are split: H reacts with C from CO2 to form carbohydrates during photosynthesis and oxygen is released as a by-product. Capillary lift is undertaken by plants that have tap roots extending down into regolith layers for their own water supply and is more highly developed in plants that are adapted to dry environments and/or environments that are seasonally or periodically dry such as drought-prone areas. The xylem in roots is
very effective at preventing leakage, but it is not watertight, particularly in the smaller, younger roots. If there is a hydraulic gradient between roots in a dry soil and others in a moist or wet soil, then water will be moved along this gradient to reduce the potential. There is a need for the hydraulic gradient to be great enough to overcome the forces of gravity, and the resistance of the xylem in order to get a net mass transfer of water. Thus, there is a process of self irrigation, whereby moisture continues to be brought up during non-sunlight hours and leaks into the rhizosphere to be taken up again once photosynthesis begins – and the in-plant store is consumed during the days. Thus, the plant (and its neighbours) has two reservoirs of water to use to drive photosynthesis when the stomata open. The effects are localised around each plant, with water close (< 2.5 m) to the tree in the shallow soil indistinguishable in isotopic signature to the groundwater, while 5 m from the tree in the same shallower soil horizon the isotopic signature returns to that of soil water from recent precipitation. From the point of view of the formation and evolution of the regolith, there are the obvious effects of hydraulic lift that have already been mentioned in transferring water that will promote weathering by other forms of biota and in the non-biotic weathering spheres. There is also the possibility of the transfer of solutes (weathering products) in solution in the water in roots as water is drawn from considerable depth, encouraging chemical and biochemical reactions to go forward and the effects on the hydro-bio-geochemical cycling of elements. The support provided to the rhizosphere by extending the water supply into otherwise drying periods extends the time during which organisms can be active, and increases the metabolic rates throughout the root zone. Increased metabolism and increased time for chemical reactions means increased biotic weathering by all the organisms in the regolith. Biota and slope processes, erosion and mass movement
The growth of roots, the construction of casts and mounds, chambers, galleries, tunnels, burrows, nests, tracks (which funnel water), and feeding scratches, scrapings, tunnelling, digging and sorting all transfer material, which, with the added effect of gravity, lead to
Regolith and biota
slope processes. Parallel changes to porosity and permeability, as already discussed, create preferential flow paths within the regolith, wherein water moves more easily and quickly and, in turn, aids the transfer of materials through the regolith and also enhances mass movement of materials down-slope. In extreme cases, krotovina lead to tunnelling, piping and gully-head collapse. Bioturbation of regolith can be broken down into four major effects on landscapes and slopes: 1. the augmentation of erosion or movement of surface material down-slope, commonly called biotransfer 2. the mixing versus homogenisation and/or horizonation or overturning of material, which leads to effect 3 3. enhanced weathering and regolith formation 4. enhancement of processes, such as infiltration, sub-surface transfer of water, solutes and solids. If the relocation of material by bioturbation is considered in three dimensions on even low-angle slopes (Figure 8.13), it quickly becomes obvious that this transfer process moves material outwards (it can really only move upwards and perpendicular to the surface, unless compaction takes place) and laterally (with some compaction or ‘flow’). When the space created by biota is vacated, it will be filled predominantly from up-slope and vertically above (by gravitational forces and usually to lower bulk density) and so leads to net downhill creep. As a result of root growth, Gabet et al. (2003) calculated the slope-dependent lateral rates of movement for northern hemisphere grass root growth down a slope of 10° at 10 m 3/m/Ma, shrubs at 90 m3/m/Ma, and trees at 100 m3/m/Ma, (compare tree fall, which moved material at 1000 m3/m/Ma, and gophers at 5000 m3/m/Ma: Gabet et al. 2003). The form of hillslopes can be controlled by biota causing, or enhancing, hillslope processes (surface erosion, sediment transport and deposition on hillslopes) by the release of uncompacted materials on to the surface by vertebrates and invertebrates when spoil-type mounds and constructed mounds are considered. Roering et al. (2002) showed that bioturbation under forests, on a loess-mantled hillslope on the South Island of New Zealand, increased sediment
erosion (Holocene conditions) and led to the development of broadly convex slopes, where, under earlier lower sediment flux rates in grassland-dominated landscapes (Pleistocene conditions), flatter locally incised slopes developed. In areas where slopes are inhabited by gophers, it would be difficult not to see the tunnelling by gophers in pursuit of plant roots, and their resultant mounding, as the major source of sediment (Thorn 1978; 1982). Add to this the likely impacts on infiltration into the slope of water along the multitude of current and past open or infilled macropores, and the conclusion is that gopher activity is the dominant sedimentmovement process (Black and Montgomery 1991). Gabet (2000) has developed a slope erosion model based on slope and gopher activity. Burrowing by wombats in the more easily dug, sometimes moist soils along floodplains and in drainage lines – often in less stable and usually sodic soils (Field and Anderson 2003) – can lead to up-slope piping from run-off and, eventually, to extensive gully systems (Field 2004; 2006). In Eastern Australia, these processes are widespread and have taken place in dispersible, sodic valley fill soils and have led to extensive gullying, channel incision, channel widening (lateral pipes and bank collapse) and lengthening. An understanding of the effects of biota in moving materials around and down-slope has been around for a long time. Davis and Snyder (1898) saw soil creep as a result of animals burrowing and the growth and death of plant roots. Some authors (for example, Caine 1986) even attribute the form of hillslopes, hillslope erosion, sediment transport and deposition on hillslopes, and soil production rates primarily to bioturbation types, distribution and intensity (Viles 1988). These processes are now seen as comparable to the abiotic processes (biota still have effects on water content, insulation against temperature changes, clay formation, water table control, and so on) such as freeze– thaw and frost heave, shrink–swell of clays on wetting and drying, and other mass movement processes. Gabet et al. (2003) present a formula that can be used to calculate sediment fluxes resulting from root growth and decay of from around 8 m3/Ma in temperate grassland to around 32 m3/Ma under temperate forest on a slope of 10°. These values are an order of magnitude
201
202
Regolith Science
less than the same authors suggest for tree fall and a further order for bioturbation by gophers. Animals also directly impact on slopes by loading (combined animal weights) and resultant mass movement is also going to affect regolith formation and landscape evolution. Increased slope mass movement, as discussed above, when combined with loading by herds of animals such as reindeer, may be the trigger that initiates landslides.
8.4 BIOTIC EFFECTS AT BROAD SPATIAL AND LONG TIMESCALES: BIOGEOMORPHOLOGY AND ZOOGEOMORPHOLOGY Biogeomorphology – a term coined by Viles (1988) and more recently the term zoogeomorphology, popularised by Butler (1995), are both used to describe the ways in which animals act as geomorphic agents. Clearly faunal bioturbation and biotransfers can be considered as examples of both bio- and zoogeomorphology. Bioturbation can be considered at scales from individual mineral grains, the scale of individual organisms, or collectively at scales of whole ecosystems or catchments and landscapes. These ecosystems can range in size from catchments of tens of thousands of square kilometres to forests of only a square kilometre or even tens of hectares. The most appropriate timescale to consider the formation of soils, regolith and landscapes across the Australian continent is a geological timescale – and that is most appropriately carried out at a continental, regional or broad spatial scale. Obviously, at a continental or regional scale, there is a very strong correlation of climate and biota and this is then reflected in the influence of biota on, for example, erosion. As soon as long timescales are introduced, particularly in Australia, the effects of climatic change must be considered. Climate change and concomitant changes in the biota (plus or minus feedback into climate) lead to cycles of weathering and denudation, which are followed by more stable periods of more- or less-intense weathering. The cycles of weathering and erosion tend to swing between humid tropical conditions and the dominance of groups of plants such as rainforests, and more arid conditions leading to
vegetation retreat and the production of aeolian material. The effect of climate and vegetation on depth and intensity of weathering has been discussed many times in the literature of pedology and geomorphology, and can be summarised in the diagrams of Strahkov (1967) (Figure 8.20). The effects of the blanket of living vegetation and litter on infiltration and evapotranspiration underlie these models for weathering. There are a multitude of effects: surface roughness reducing the speed of movement of both wind and water, reducing evaporation and transpiration, increasing storage and redistribution of water by standing vegetation, litter and wellstructured regolith surficial layers; and then there are effects on other variables, such as temperature, and the mediation and habitat provision for other biota. If the regolith is maintained in a shaded and moist condition following enhanced accession and storage of water, then almost all the chemical reactions within it will be maximised and exceed the likely rates should the system dry out, or cool. 8.4.1 Erosion and weathering controls Continent and larger scale comparisons of erosion and weathering are dominated by discussion of the effects vegetation cover, the evolution of soil mantles and the partitioning of water (for example, Douglas 1977). Vegetation shelters the regolith from sunlight and heating and cooling, from the erosive effects of wind while favouring deposition and from erosion by water, while storing and redistributing water and nutrients to produce more biomass and maintain the ecosystem. After all, vegetation is the reason why arid environments do erode under wind and infrequent run-off at quite high rates while areas with much higher rainfall do not – even under most storm events. Vegetation is also a critical component in distribution of resources, such as water, and thereby erosion and deposition. Take for example banded vegetation in arid and semi-arid regions. Ecologists often discuss the bands in terms of redistribution of very limited resources – usually water (but also importantly, wind) – from across the whole landscape to concentrations within the bands, with bare and scalded areas simply regarded as extended and
203
3000
Tundra
Moist temperate
Temperate grassland
Desert and semidesert
Tropical grassland
Tropical forest Rainfall Equator
Mean annual precipitation (mm)
Regolith and biota
2000
Leaf fall 25
Temperature
20
1000
15 10
Evaporation 0
5 0
Soil armour (laterite, ferricrete) Zone of ochre and alumina Kaolinite (clay) zone Zone of dominant hydrolysis Rock debris, little altered chemically Fresh rock
Figure 8.20: Weathering mantles, climate and biota (after Strahkov 1967).
contributing catchments for the vegetation bands. The effects on regolith formation, evolution and distribution from growing vegetation: the deposition and stabilisation of sediment and solutes in the band which were derived from the whole area, the uptake and cycling of nutrients, the enhanced weathering as a result of the plants and their concomitant soil organisms, the associated fauna and bioturbation, and the inputs and impacts of organic matter on the regolith characteristics are among the most obvious. At more localised scales, vegetation changes are strongly correlated with regolith characteristics such as soil depth (average 66 cm depth under open forest/ woodland and 28 cm under treeless heath and heathlands; Wilkinson et al. 2003, 2004). However, one must inevitably ask the question again – which came first: the vegetation type or the soil characteristic that gave a particular vegetation type a comparative advantage? The most productive way to see this argument is
to see the vegetation association and soil or regolith development as co-evolution (Field 2004, 2006). Nevertheless Wilkinson et al. (2003, 2004) and others argue that weathering or soil production is inversely related to soil depth. However, maintaining a soil cover of any depth requires a stable vegetation cover. Aspect is another example of the interaction between a landscape-scale variable and biota that provide a feedback loop, which, in turn, affects landscape form and thereby process. Selkirk et al. (2001) show that in the Blue Mountains west of Sydney, the valley physiography (working through aspect) affects fire regimes and the vegetation that, in turn, interact with moisture availability, temperature and sunlight. Moisture availability, temperature and incident sunlight, in turn, are components in regolith formation and erosion (soil depth) and the success of vegetation types. In the same environment, Wilkinson et al. (2003) demonstrated the relationship between spur
Regolith Science
cross-sectional shape (that is, convexity) and soil depth and vegetation type. 8.4.2 Hydrobiogeochemical cycles At the largest scale, ecosystems store and move large quantities of chemical elements around in geologic time frames – the hydrobiogeochemical cycle (Figure 8.21). For example, Gilkes (1998) calculates that typical vegetation growing on a substrate in a landscape for a million years at around 5 tonnes of biomass/ha/year – and containing a typical concentration of about 5 ppm Zn – will translocate from the weathering zone around 50 tonnes of Zn/ha up, and into, a cycle between the upper soil horizons, litter and living vegetation. In terms of constituents of biota, Zn is only a micronutrient. The term ‘hydrobiogeochemical cycle’ is used when elements are followed from the underlying hard rock up through the regolith and soil and into standing biota, to seek out the sources, transfers and sinks for these within the overall system. During the last quarter of last century, work began to outline the paths in particular ecosystems. The most comprehensive and widely read of these studies are the Hubbard Brook experiments (for example, Likens et al. 1997) in north-eastern USA, followed by work at Plynlimon in the UK (for example, Kirby et al. 1991) and in Australia in New England, NSW (for example, Zakaria 1977; Lam 1979; Day 1980; Field 1983). All these studies detailed the hydrobiogeochemistry of nutrients and non-nutrients, cation and anions, through vastly different ecosystems, but they all proved the importance of the biota to the removal and cycling of elements out of, and back into, the regolith. The transfers are huge when considered on a regolith-formation timescale. Several studies have used mass balances to calculate inputs from weathering and sought to balance them with outputs in streams (Markewitz and Richter 1998; Kearns et al. 2003; Alexandre et al. 1997). Groups of elements within these cycles have quite different paths: essential nutrients of C, H and O are not supplied from weathering, but are cycled within the regolith as organic matter; macro nutrients are sourced from a variety of sources – from the atmosphere for N, from the weathering of apatite, fluorapa-
F7a precipitaion input water , solutes, dust, gas diffusion
F6 respiration evapotranspiration F3 foliage accretion
F7b evaporation gas diffusion
F4 Litter fall
F2 wood accretion
F14 erosion/ deposition
F1 vegetation uptake Solum = the inhabited soil
204
F11 mineralisation of organic matter, N fixation
F5 Stem flow S3 dead biomass including litter
S4 soil water/solution
F10 neoformation F8 weathering release S5 minerals susceptible to weathering
F9 ion exchange F12b water, gas + dissolved element input from subsoil
F12a water, gas + dissolved element output to subsoil
S
= Stock
S2 below ground living biomass
F13 particulate migration
F = Flux
Stocks S1 S2 S3 S4 S5
Above ground biomass: leaves, wood, animals, fungi etc. Below ground living biomass: roots, animals, fungi etc. Dead biomass including litter Soil water and dissolved elements Minerals susceptible to weathering
F1 F2 F3 F4 F5 F6
Uptake by vegetation via root system Accretion by wood Accretion by foliage Litter fall including leaves, wood, animals etc. Stemflow of water and dissolved ions Respiration (O2, CO2), evapotranspiration (H2O), fire (N, S, C) Precipitation includes dust and dissolved elements, evaporation (H2O), gas diffusion (e.g. CO2) inputs/outputs Weathering of minerals lon exchange from clay, sesquioxide and organic matter surfaces Neoformation of minerals Mineralisation of organic matter; N-fixation by legumes Movement into/out of solum of water, gas and dissolved elements Eluviation/illuviation of colloidal particles Erosion/deposition of soil materiaIs
Fluxes
F7 F8 F9 F10 F11 F12 F13 F14
Figure 8.21: A schematic hydrobiogeochemical cycle (after Gilkes 1998).
Regolith and biota
tite and other accessory minerals for P, and the K-spars and muscovite for K. Each element is then cycled through the regolith and is systematically transferred between organic components, labile compartments and living organisms (and also partitioned within the organisms), such as plants (Figure 8.22). The micro nutrients nearly all come from the weathering of minor and accessory minerals, and are subsequently cycled. The biota have a critical role in the release of nutrients from alumino-silicate weathering – and in the subsequent uptake storage and release again from organic materials, whether they are nutrients or not (for example, Si cycling; Alexandre et al. 1997). In the case of P, the role of organic ligands produced by as-yet-unidentified microorganisms is fundamental to the breakdown of apatite, strengite and variscite and fluoroapatite with subsequent sequestering in secondary Fe, Al, and lanthanide minerals (Kearns et al. 2003). The lanthanide minerals can be very insoluble, but the role of soil fungi then becomes important in their cycling. The fluxes and stores vary according to species, genera and plant type (C3, C4, N fixer, phenotypes, and so on) as well as temporal and spatial dimensions. On a single site, for example, a succession of vegetation types will, over time, have an evolving set of fluxes and stores – as will periods of
Photosynthesis fixation
drought or flood. Even in a period of a few days to weeks, stores and fluxes may change with soil moisture conditions, as a saturated soil first drains and then dries: a good example is the deficiency in B for sensitive species (to B concentrations), such as the Pinus genera. Additionally, across sites with different soils, regolith and/or lithologies, there will be very different hydrobiogeochemical cycles. Other landform characteristics (aspect, slope position and slope steepness) can also change the fluxes and stores across a landscape. Fluxes and stores change for different individual elements and groups of elements across the periodical table. The results can vary over several orders of magnitude in what appear to be as close to possible similar conditions for some elements (such as Fe) and as much as 106 times for others (Au). The non-essential elements can also be toxic – particularly in areas that are highly concentrated, such as over a weathering ore body. One of the problems with research that relies on published values is the wide variation in these aspects of biomass uptake and storage when compared with regolith fluxes and stores. Not only do the values change with time and growth stages, even for only one species, but they also differ markedly across species and varieties.
ORGANIC COMPARTMENT Litter
ATMOSPHERE
Respiration (CO2, etc) Evapotranspiration
Aerosols
Ev
Biomass dead (humus)
Mesofauna Microflora
Vegetation
Ra in
ap
ora tio n
Biological uptake
Alteration SOIL MINERALS Neoformation
Mineralisation
Exudation leaching
LABILE COMPARTMENT Exchange sites
Soil solution
Figure 8.22: Processes, stores and fluxes within the living biota (flora, herbivores and carnivores), litter (detritivores and carnivores) and regolith (soil, saprolite and groundwater) compartments (after Gilkes 1998).
205
206
Regolith Science
8.4.3 Vegetation bands, patterned ground, tiger striping, and resource distribution and concentration When resources are limited to the point where ecosystems cannot function effectively over the whole landscape, but where redistribution of all the resources available to a smaller area can allow functioning, a mosaic or banding often occurs. Ecologists, geomorphologists and pedologists all argue for a parallel and causal evolution of climate and banded vegetation with landforms, regolith and soils. Some authors have argued that climatic deterioration can be a trigger for the evolution of this redistribution (Goodspeed and Winkworth 1978); others suggest grazing pressure may be the accelerant or underlying cause (Valentin et al. 1999), while, more recently, several have shown that simpler explanations, such as partitioning of water, can alone cause banding (Dunkerly 1997; Ursino 2007). Dunkerly and Brown (1995) suggest a system of germination and colonisation of litter dams being selected and scaled up through an unspecified process); some areas gain a competitive advantage at collecting and maintaining resources, to the detriment of surrounding areas – areas of vegetation and bare areas (Ursino 2007). Surface sealing, or crusting, across bare areas is a critical component of the processes of water redistribution (Greene 1992; Valentin et al. 1999). The patterning can be approximately round (or spotted on roughly flat topography, but leading to a localised build-up of material under vegetation) (Dunkerly and Brown 1995; Valentin et al. 1999), through to long linear bands, such as those along contours on gently sloping land (Mabbutt and Fanning 1987). Linear vegetated bands can be expected to be perpendicular to the slope (0–2% slope) or dominant wind direction (Valentin et al. 1999). Valentin et al. found that slope gradient and mean annual rainfall are correlated with the distance (or wavelength) between the vegetated and bare areas. Wakelin-King (1999) ascribes banded vegetation to a widely distributed, but single, geomorphic unit – the sheet-flow plains, but others see it as much more widespread (Mabbutt and Fanning 1987). Patterning, banding, banded mosaic, or (banded) tiger bush (striping or swirling like the pattern on tiger fur) are two or three phase sequences of alternat-
ing bare ground, or source areas, with bands of vegetation or sinks. Essentially resources such as water, nutrients and organic matter are harvested across the entire landscape and collect in the bands and are therefore able to support an enhanced biomass in a limited area. A strong feedback mechanism begins to develop whereby the density of plants (and all the other organisms in the ecosystem – such as termite mounds; Valentin et al. 1999) are strongly related to that systems ability to absorb (infiltrate and store) water, organic matter and nutrients, which, in turn, produces a greater density of plants and other biota. There is also a development of plant physiological mechanisms (either by single plant adaptation or by selection of fitter organisms) within the inter-grove and isolated vegetation (such as mulga and grasses) to be more water and nutrient efficient (Cook and Dawes-Gromadzki 2005), while those within groves seemed to consume greater resources and have enhanced biomass (including, incidentally, invertebrates such as termites, spiders and ants) (Cook and Dawes-Gromadzki 2005). The bottom line is that the banding controls erosion, sedimentation and most other slope processes, and hence the biota, are intimately involved in the shape of the landscape and the formation and evolution of the regolith. 8.4.4 Fire and vegetation effects on regolith The Australian regolith is particularly strongly affected by fire. The age and long weathering history of the Australian landscape (Chapter 2) has led to the very low productivity of Australian ecosystems. Fire may destroy vegetation and lead to very high levels of erosion if the right seasonal weather conditions follow (for example, Blake et al. 2005; Shakesby et al. 2003). Fire creates a mosaic of units within the landscape each bounded by a fire scar. These scars are more than vegetation boundaries because, as we have seen, vegetation affects the regolith from the surface down. Surficial mats affect hydrology and erosion, litter affects organic chemistry in the surface horizons and beyond, and roots mine the regolith for nutrient ions; therefore fire becomes a land-forming agent. More localised effects of vegetation-fuelled fire are the ‘cooking’ of rocks, regolith and soils – not only
Regolith and biota
altering organic matter content, but also changing the overall structure (splitting and spalling of rocks), mineralogy and resistance to weathering and erosion. Any comminution affects the surface-to-volume ratio and therefore the rate of weathering, so spalling and splitting enhance the subsequent breakdown of primary mineral dominated rocks into weathering products. Fire can also ‘create’ new formations from elements that previously made up long-standing vegetation, such as ‘wood-ash stone’ (a calcite agglomeration), which is produced during the slow burning of a large standing tree bole (Humphreys et al. 1987). Heat from fires can cause seemingly opposite effects, with increasing strength or case hardening on the one hand and mechanical failure on the other, where responses from weathered samples mean that post-fire materials are comminuted and more readily break down (Allison and Bristow 1999). Many authors stress the widespread and major effects fire has on the landscape and regolith (for example, Allison and Bristow 1999) when temperatures commonly reach 400 or 500°C (Goudie et al. 1992) and sometimes as high as 1000°C (Allison and Bristow 1999). (The importance of maghemite formation in dating and indicating surface conditions is emphasised in Chapters 2 and 6.) The rapid rise in temperature, and subsequent cooling with nightfall or rain, imposes an extreme temperature gradient promoting rock disintegration and subsequent weathering susceptibility (Adamson et al. 1983). The opposing effect of case hardening appears to be related to rock thermal characteristics, material properties – such as dehydration of particular minerals in both the primary, and then the secondary mineralogy associated with weathering – and environmental constraints, such as diurnal temperature ranges and the prevalence of thermally induced storms (Pye 1982; Goldammer 1990). Allison and Bristow (1999) give the examples of dehydration of both free and locked water in serpentinite and wehrlite (in serpentine) and actinolite (in gabbro, which dehydrates to chlorite and water) at 500°C, which reduces the modulus of elasticity and the physical strength of the rock. They show, on the other hand, that dolerite raised to 500°C has free water dehydration reactions, while locked water is retained within the mineral matrix – leading to a case-hardening effect and reduced weathering (until some other effect, such
as another fire at higher temperatures, causes breakdown). We also know that anthropogenic effects stretching back through millennia have altered vegetation through the use of fire, with concomitant changes to the landscape. Anthropogenic vegetation changes are just the imposition of yet another layer of the effects of biota (and biota changes) on the Australian landscape and regolith – albeit an extremely intense series of changes that are taking place at the moment. Each of the issues raised above illustrates the coevolution of Australian landscapes and regolith with the Australian biota. 8.4.5 Water table control and dryland salinity Biota – through evapotranspiration and root distribution – are a major control on surface and near-surface water tables. Vegetation layers (overstorey, understorey, herb and grasses and litter layers) act as sponges to incoming precipitation (they can even ‘create’ precipitation through fog drip): absorbing it and redirecting it, storing or evapotranspiring it. As a result, much less than 50% of incoming precipitation gets to the mineral soil surface in eucalypt forest and woodland, on average, on an annual cycle (Field 1983). This process of interception is slashed when vegetation is cleared. As a result, huge increases in regolith infiltration and run-off occur. Locally, this can cause rising water tables. Salinity in regolith and groundwater (including regolith and soil water tables) is controlled by a host of factors (for example, Bann and Field 2006a; Chapter 12). Because most Australian regolith and soil contains appreciable levels of Na salts and insufficient buffers of Ca, Mg and K (Stahl and Field 2003), the water becomes saline. Processes such as capillary rise bring soil water to the surface. If evaporation takes place on a bare or sparsely vegetated site, even a low electrical conductivity (a measure of water impurities) in the soil water can lead to the concentration of a range of salts and inorganic and organic complexes on the soil surface. The result is dryland salinity (Bann and Field 2006b). The challenge is to understand how changing salinity levels affect regolith processes. Obviously, EC will have major effects on biochemical and chemical
207
208
Regolith Science
equilibria, on hydration, and on osmosis, when the biota attempt to withdraw nutrients and water. There are also toxicity effects from the Na itself (not to mention Cl and other ions present). The net result is usually loss of some vegetation (but not necessarily other biota: Bann and Field, 2006a) with concomitant effects of changes to the surface hydrology (run-off, infiltration, and so on) through changes to the surface (crusting and sealing) and sub-surface (soil structure, permeability, porosity, sorptivity and hydraulic conductivity) characteristics of the regolith.
8.5 CONCLUSIONS Biota are both a part of the landscape and a factor controlling the functioning of that landscape. The functioning of the biota is also critical to the management of landscapes across all timescales: from short-term human scales to geological timescales. Biota also function across virtually all spatial scales from the microbial to the continental. In fact, biota can be argued to make the Earth different from all the other planets. Individual organisms can dramatically accelerate the rate of weathering in their vicinity – a lichen or a simple vascular plant can accelerate the very beginnings of the weathering of a virtually bare volcanic flow when compared with adjacent fresh rock. This increased rate continues unabated through a succession to forest – and an individual tree has effects on its site occurring during its lifetime (Hamilton 1972; Leonard and Field 2003) – and affect regolith and soil characteristics such as depth, horizonation, pH, organic matter, Eh, texture (and differentiation, duplex profile) and soil-profile formation under a single tree. The soil in which vegetation grows can carry a ‘signature’ from that vegetation for considerable periods of time (Little 2001) and different species have different signatures. Soil along a transect between two species of trees, shows markedly different characteristics near each tree bole (Leonard and Field 2003), in the area affected by stem flow, under the drip zone and in the intervening area (Hamilton 1972). Many of these patterns in soil are strongly related to root growth, leaching and weathering as the result of the selective uptake of elements by the vegetation, the cycling and subsequent deposition in precipitation and litter fall. Trees take part in
more obvious bioturbation when uprooted – rotating the root ball and bringing subsoil up to the surface. Studies in forest science suggest that the quantities of material contained in the biota and turned over by vegetation are also of comparable time and spatial scales. Catastrophic events, such as fires, can then transfer very large quantities of materials within landscapes and become major land-forming events. Meso and macro fauna transfer material at rates comparable to weathering and soil production and rates of erosion and deposition. Wombats, kangaroos, wallabies, lyrebirds and rabbits can move substantial quantities of material at particular places and times. Collectively, they are a major bioturbation factor. Smaller fauna such as earthworms, termites, ants and other insects are also important in bioturbation and, again, the dominance of one or more groups depends on the environment and timescale. Termites are very important in the seasonal tropics right through to the humid temperate, whereas earthworms are quite rare in some humid temperate forests such as the dry sclerophyll, and barely exist in the semi-arid and arid environments (although riparian and terminal drainage systems can locally support high numbers following major rains). The micro biota are also extremely important and, despite their very small size, sheer numerical dominance makes their weathering and bioturbation effects important in most landscapes. There are clear opportunities for research in many aspects of the biota–regolith interface. While there is a large knowledge base relating to biota–regolith interactions, there still remains little understanding of how these interactions influence the regolith in real landscapes, especially in Australia. There are many avenues open to the study of biogeochemical interactions in the Australian soils and regolith, as there have been few pedogenic studies in the rhizosphere for most of the dominant Australian flora. This is surprising given the recognised harshness of Australian environments and the uniqueness of its flora. Although the rhizosphere is only a narrow interface between plant and soil, it is a zone of very high biogeochemical activity. This soil compartment is thus worthy of investigation to better understand how plants modify their soil environment, encourage mineral weathering processes and take up the potential pathfinder and target elements used during bioprospecting.
Regolith and biota
In summary, the biota are a critical component of all soils, regolith and landforms and the biota are also drivers and components of most of the important processes in the formation and ongoing evolution of the landscape.
8.6
REFERENCES
Adamson D, Selkirk PM, Mitchell PB (1983). The role of fire and Lyre birds in the sandstone landscape of the Sydney Basin. In Aspects of Australian Sandstone Landscapes. (Eds RW Young and GC Nanson) pp. 81–93. University of Wollongong, NSW. Alexandre A, Meunier J-D, Colin F and Koud J-M (1997). Plant impact on the biogeochemical cycle of silicon and related weathering processes. Geochemica et Cosmochimica Acta 61, 677–682. Allaby M (Ed.) (1998). A Dictionary of Ecology. Oxford University Press, Oxford. Allison RJ and Bristow GE (1999). The effects of fire on rock weathering: some further considerations of laboratory experimental simulation. Earth Surface Processes and Landforms 24, 707–713. Anand RR and Paine M (2002). Regolith geology of the Yilgarn Craton, Western Australia: Implications for exploration. Australian Journal of Earth Sciences 49, 3–162. Anderson GR (2001). The influence of macro and meso biota on regolith development and evolution. BSc (Hons) thesis. Australian National University, Canberra Arocena JM, Gottlein A and Raidl S (2004). Spatial changes of soil solution and mineral composition in the rhizosphere of Norway-spruce seedlings colonized by Piloderma croceum. Journal of Plant Nutrition and Soil Science 167, 479–486. Aufreiter S, Mahaney WC, Milner MW, Huffmann MA, Hancock RGV, Wink M and Reich M (2001). Mineralogical and chemical interactions of soils eaten by chimpanzees of the Mahale Mountains and Gombe Stream National Parks, Tanzania. Journal of Chemical Ecology 27, 285–311. Baker AJM (1981). Accumulators and excluders – strategies in the responses of plants to heavy metals. Journal of Plant Nutrition 3, 643–654. Banfield JF, Barker WW, Welch SA and Taunton A (1999). Biological impact on mineral dissolution:
application of the lichen model to understanding mineral weathering in the rhizosphere. Proceedings of the National Academy of Science of the United States of America 96, 3404–3411. Colloquium Paper. Banks JCG (1989). A history of forest fire in the Australian Alps. In The Scientific Significance of the Australian Alps. The Proceedings of the 1st Fenner Conference. (Ed. R Good), pp. 265–280. Australian Alps National Parks Liaison Committee, Canberra. Bann GR and Field JB (2006a). Dryland salinity in south-east Australia: which scenario makes more sense? In Proceedings Australian Earth Sciences Convention, Melbourne. <www.earth2006.org.au> Bann GR and Field JB (2006b). Dryland salinity and agronomy in south-east Australia: groundwater processes or soil degradation associated with intensive grazing? In 13th Australian Society of Agronomy Conference, Perth. Ground Breaking Stuff. <www. regional.org.au/au/asa/2006> Barker WW, Welch SA, Chu S and Banfield JF (1998). Experimental observations of the effects of bacteria on aluminosilicate weathering. American Mineralogist 83, 1551–1563. Bennie AT (1991). Growth and mechanical impedance. In Plant Roots: The Hidden Half. (Eds Y Waisel, A Eshel and U Kafkafi) pp. 393–414. Marcek Dekker, New York. Blake WH, Wallbrink PJ, Doerr SH, Shakesby RA and Humphreys GS (2005). Magnetic enhancement in wildfire-affected soil and its potential for sediment-source ascription. Earth Surface Processes and Landforms 31, 249–264. Black TA and Montgomery DR (1991). Sediment transport by burrowing animals, Marin County, California. Earth Surface Processes and Landforms 16, 163–172. Bolan NS, Elliot J, Gregg PEH and Weil S (1997.) Enhanced dissolution of phosphate rocks in the rhizosphere. Biology and Fertility of Soils 24, 169–174. Brantley SL, Goldhaber MB and Ragnarsdottir KV (2007). Crossing disciplines and scales to understand the Critical Zone. Elements 3, 307–314. Brundett MC (2002). Tansley review no. 134: Coevolution of roots and mycorrhizas of land plants. New Phytologist 154, 275–304.
209
210
Regolith Science
Burges NA (1958). Micro-organisms in the Soil. Hutchinson University Library, London. Butler D (1995). Zoogeomorphology. Animals as Geomorphic Agents. Cambridge University Press, Cambridge, UK. Buyer JS and Drinkwater LE (1997). Comparison of substrate utilization assay and fatty acid analysis of soil microbial communities. Journal of Microbiological Methods 30, 3–11. Caine N (1986). Sediment movement and storage on alpine slopes in the Colorado Rocky Mountains. In Hillslope Processes. Binghampton Symposium in Geomorphology: International Series No. 16. (Ed. AD Abrahams) pp. 115–137. Allen and Unwin, Boston, Massachussetts. Cairney JWG and Meharg AA (2002). Interactions between ectomycorrhizal fungi and soil saprotrophs: implications for decomposition of organic matter in soils and degradation of organic pollutants in the rhizosphere. Canadian Journal of Botany 80, 803–809. Cammeratt LH, Willott SS, Compton SG and Incoll LD (2002). The effect of ants’ nests on the physical, chemical and hydrological properties of a rangeland soil in semi-arid Spain. Geoderma 105, 1–20. Carr J, Gould N and Sterling, J (2004). ‘Does bioturbation activity vary with vegetation cover in the Santa Cruz Mountains, California’. Research Report Series No. 4. pp. 27–30. Department of Geography, University of Liverpool, UK. Cervini-Silva J, Fowele DA and Banfield J (2005). Biogenic dissolution of a ceri-phosphate mineral. American Journal of Science 305, 711–726. Chorover J and Amistadi MK (2001). Reaction of forest floor organic matter at goethite, birnessite and smectite surfaces. Geochemica et Cosmochimica Acta 65, 95–109. Coleman DC and Hendrix PF (Eds) (2000). Invertebrates as Webmasters in Ecosystems. CABI Publishing,Wallingford, UK. Cook GD and Dawes-Gromadzki T (2005). Stable isotope signatures and landscape functioning in banded vegetation in arid-central Australia. Landscape Ecology 12, 216–224. Corbet AS (1935). Biological Processes in Tropical Soils, with Special Reference to Malaysia. W. Heffer and Sons, Cambridge, UK.
Cowan JA, Humphreys, GS, Mitchell, PB and Murphy CL (1985). An assessment of pedoturbation by two species of mound-building ants, Camponotus intrepidus (Kirby) and lridomyrmex purpureus (F. Smith). Australian Journal of Soil Research 22, 95–107. Cromack K, Sollins P, Graustein WC, Speidel K, Todd AW, Spycher G, Li CY and Todd RL (1979). Calcium oxalate accumulation and soil weathering in mats of the hypgeous fungus Hysterangium crassum. Soil Biology and Biochemistry 11, 463–468. Curl EA and Truelove B (1986). The Rhizosphere. Springer-Verlag, Berlin, New York. Darwin C (1837). On the formation of mould. Transactions of the Geological Society of London 5, 505. Darwin C (1881). On the formation of vegetable mould through the action of worms. J. Murray, London. Davies N (1953). ‘Investigations on the soil and water relations of the river red gum forests’. Final Report Murray Management Survey. Report no. R 124. Forestry Commission of New South Wales, Resources Branch, Sydney. Davis WM and Snyder WH (1898). Physical Geography. Ginn and Co. Publishers, Boston and London. Day DG (1980). Drainage density changes during rainfall. Earth Surfaces Processes 1, 63–70. Denny C S and Goodlett JC (1956). Microrelief resulting from fallen trees. In Surficial Geology and Geomorphology of Potter County, Pennsylvania (Ed. CS Denny) pp. 59–66. Professional Paper 288. United States Geological Survey, Reston, Virginia. Devit DA and Smith SD (2002). Root channel macropores enhance downward movement of water in a Mojave Desert ecosystem. Journal of Arid Environments , 99–108. Dexter BD (1967). ‘Flooding and regeneration of river red gum, Euclayptus camldulensis (Dehn)’. Bulletin Number 20, Forests Commission of Victoria, Melbourne. Douglas I (1977). Humid Landforms. ANU Press, Canberra. Drever JJ and Stillings LL (1997). The role of organic acids in mineral weathering. Colloids and Surfaces A: Physicochemical and Engineering Aspects 120, 167–181.
Regolith and biota
Dunkerly DL (1997). Banded vegetation: development under uniform rainfall from a simple cellular automaton model. Plant Ecology 29, 103–111. Dunkerly DL and Brown KJ (1995). Runoff and runon areas in a patterned chenopod shrubland, arid western New South Wales, Australia: characteristics and origin. Journal of Arid Environments 30, 41–55. Eddy J, Humphreys GS, Hat DM, Mitchell PB and Fanning PC (1999). Vegetation arcs and litter dams: similarities and differences. Catena 37, 57–73. Edwards CA and Bohlen PJ (1996). Biology and Ecology of Earthworms. Chapman and Hall, London. Edwards WM and Shipitalo MJ (1998). Consequences of earthworms in agricultural soils: aggregation and porosity. In Earthworm Ecology. (Ed. CA Edwards) pp. 147–161. CRC Press, Boca Raton, Florida. Eggleton RA (Ed.) (2001) The Regolith Glossary: Surficial Geology, Soils and Landscapes. CRC LEME, Canberra. Ekundayo EO and Aghastise VO (1997). Soil properties of termite mounds under different land use types in a typic paleudult of midwestern Nigeria. Environmental Monitoring and Assessment 45, 1–7. Field JB (1983). The sources of solutes in, and chemical budgets for three small rural catchments, New England, N.S.W. PhD thesis, University of New England, Armidale, NSW. Field JB (2003). Biota, regolith and landscapes: At the hillslope, profile and lesser scales. In Advances in Regolith. Proceedings of the CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp.115–118. CRC LEME, Perth. Field JB (2004). Geomorphology and the biota. In Proceedings of the 11th Australia New Zealand Geomorphology Group Conference 2004. (Ed. D Fabel) p. 24. Australia New Zealand Geomorphology Group, Melbourne. Field JB (2006.) The effects of biota on the formation of regolith: where we are up to, and whence from here. In Regolith 2006 Consolidation and Dispersion of Ideas. Hahndorf, South Australia (Eds RW Fitzpatrick and P Shand) pp.284–287. CRC LEME, Perth. Field JB and Anderson GR (2003). Biological agents in regolith processes: case study on the Southern
Tablelands, NSW. In Advances in Regolith. Proceedings of the CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp.119–121. CRC LEME, Perth. Fomina M, Alexander IJ, Colpaert JV and Gadd GM (2005). Solubilization of toxic metal minerals and metal tolerance of mycorrhizal fungi. Soil Biology and Biochemistry 37, 851–866. Gabet EJ (2000). Gopher bioturbation; field evidence for nonlinear hillslope diffusion. Earth Surface Processes and Landforms 25, 1419–1428. Gabet EJ, Reichmen OJ and Seabloom EW (2003). The effects of bioturbation on soil processes and hillslope evolution. Annual Review of Earth and Planetary Sciences 31, 249–273. Gadd GM (2004). Microbial influence on metal mobility and application for bioremediation. Geoderma 122, 109–119. Gilkes RJ (1998). Biology and the regolith: an overview. In The State of the Regolith. (Ed. RA Eggleton) pp.110–125. Special Publication No. 20. Geological Society of Australia, Sydney. Gobran GR, Wenzel WW and Lombi E (Eds) (2001). Trace Elements in the Rhizosphere. CRC Press. Boca Raton, Florida. Goldammer JG (Ed.) (1990). Fire in the Tropical Biota. Springer-Verlag, Berlin Gomes NCM, Fagbola O, Costa R, Rumjanek NG, Buchner A, Mendona-Hagler L and Smalla K (2003). Dynamics of fungal communities in bulk and maize rhizosphere soil in the tropics. Applied and Environmental Microbiology 69, 3758–3766. Goodspeed MJ and Winkworth RE (1978). Fate and effect of runoff – with particular reference to arid and semi-arid plain lands. In Studies of the Australian Arid Zone. III. Water in Rangelands. Proceedings of the Symposium, Rangelands Research Unit, Alice Springs Field Centre, 15–18 October. (Ed. HMW Howes). Division of Land Resources Management, CSIRO, Melbourne. Goudie AS, Rogers A and Viles HA (1992). The Students Companion to Geography. Blackwell, Oxford. Greene RSB (1992). Soil physical properties of three geomorphic ones in a semi-arid Mulga Woodland. Australian Journal of Soil Research 30, 55–69. Guglielmin M, Cannone N, Strini A, Lewkowicz A. (2005). Biotic and abiotic processes on granite
211
212
Regolith Science
weathering landforms in a cryotic environment, Northern Victoria Land, Antarctica. Permafrost and Periglacial Processes 16, 69–85. Haag D and Matschonat G (2001). Limitations of controlled experimental systems as models for natural systems: a conceptual assessment of experimental practices in biogeochemistry and soil science. The Science of the Total Environment 277, 199–216. Hagerberg D, Thelin G and Wallander H (2003). The production of ectomycorrhizal mycelium in forests: Relation between forest nutrient status and local mineral sources. Plant and Soil 252, 279–290. Hamilton C (1972). The nature and causes of spatial variation in forest ecosystems. PhD thesis. Australian National University, Canberra. Hart DM (2001). The elemental composition of phytoliths. In Phytoliths:Application in Earth Sciences and Human History. (Eds JD Meunier and F Colin) pp. 313–316. Balkena. Hart DM and Humphreys GS (2003). Phytolith depth functions in surface regolith materials. In Advances in Regolith. Proceedings of the CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp. 159–163. CRC LEME, Perth. Heimsath AM, Chappell J, Dietrich WE, Nishiizumi K and Finkel RC (2000). Soil production on a retreating escarpment in southeastern Australia. Geology 28, 787–790. Heimsath AM, Chappell J, Spooner NA and Questiaux DG (2002). Creeping soil. Geology 30, 111–114. Heinrich I and Banks JCG (2006). Variation of phenology, tree growth and wood anatomy of Toona sinensis and Toona ciliata in relation to different environmental conditions. International Journal of Plant Science 167, 831–841. Hernesmaa A, Njorklof K, Kiikkila O, Fritze H, Haahtela K and Romantschuk M (2005). Structure and function of microbial communities in the rhizosphere of Scots pine after tree-felling. Soil Biology and Biochemistry 37, 777–785. Hiltner L (1904). Uber neuere Erfahrungen und Probleme auf dem Gebiete der Bodenbakteriologie unter besonderer Berucksichtigum der Grundungung und Brache. Arbeiten der Deutschen Landwirtschaftlichen Gesellschaft 98, 59–78.
Hoffland E, Giesler R, Jongmans AG and van Breeman N (2003). Feldspar tunneling by fungi along natural prodcutivity gradients. Ecosystems 6, 739–746. Holt JA, Coventry RJ and Sinclair DF (1980). Some aspects of the biology and pedological significance of mound-building termites in a red and yellow earth landscape near Charters Towers, North Queensland. Australian Journal of Soil Research 18, 97–109. Hoppert M, Flies C, Pohl W, Günzl B and Schneider J (2004). Colonization strategies of lithobiontic microorganisms on carbonate rocks. Environmental Geology 46, 421–428. Howell J, Humphreys GS and Mitchell PB (2006). Changes in soil water repellency and its distribution in relation to surface microtopographic units after a low severity fire in eucalypt woodland, Sydney, Australia. Australian Journal of Soil Research 44, 205–217. Huang PM (2000). Abiotic catalysis. In Handbook of Soil Science. (Ed. ME Sumner). CRC Press, Boca Raton, Florida. Hulme KA and Hill SM (2003). River red gums as a biogeochemical sampling medium in mineral exploration and environmental chemistry programs in the Curnamona Craton and adjacent regions of NSW and SA. In Advances in Regolith. Proceedings of the CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp. 205–210. CRC LEME, Perth. Humphreys GS (1981). The rate of ant mounding and earthworm casting near Sydney, New South Wales. Search 12, 129–131. Humphreys GS (1989). Earthen structures built by nymphs of the cicada, Cyclochila australasiae (Donovan) (Homoptera, Cicadidae). Australian Entomological Magazine 16, 99–108. Humphreys GS (1994). Bioturbation, biofabrics and the biomantle: an example from the Sydney Basin. In Soil Micromorphology: Studies in Management and Genesis. Developments in Soil Science. Vol. 22. (Eds AJ Ringrose-Voace and GS Humphreys) pp. 421–436. Elsevier, Amsterdam. Humphreys GS and Field RJ (1998). Mixing, mounding and other aspects of bioturbation: implications
Regolith and biota
for pedogenesis. In Proceedings of the 16th World Congress of Soil Science, Montpellier, France. 20–26 August. International Union of Soil Sciences, Wageningen, Netherlands. Humphreys GS, Hunt PA and Buchanan R (1987). Wood-ash stone near Sydney, NSW: a carbonate pedologial feature in an acidic soil. Australian Journal of Soil Research 25, 115–124. Hupy JP and Schaetzel RJ (2006). Introducing ‘Bombturbation’, a singular type of soil disturbance and mixing. Soil Science 171, 823–836. Jenny H (1941). Factors of Soil Formation: a System of Quantitative Pedology. McGraw-Hill, New York. Johnson DL (1989). Subsurface stone lines, stone zones, artifact-manuport layers, and biomantles produced by bioturbation via Pocket Gophers (Thomomys Bottae). American Antiquity 54, 370–389. Johnson DL (2002). Darwin would be proud: bioturbation, dynamic denudation, and the power of theory in science. Geoarchaeology: An International Journal 17, 7–40. Johnson DL, Domier JEJ and Johnson DN (2005). Reflections on the nature of soil and its biomantle. Annals of the Association of American Geographers 95, 11–31. Johnston CG and Vestal JR (1991). Photosynthetic carbon incorporation and turnover in Antarctic cryptoendolithic microbial communities: Are they the slowest growing communities on Earth? Applied Environmental Microbiology 57, 2308–2311. Jones DL, Dennis PG, Owen AG and van Hees PAW (2003). Organic acid behaviour in soils – misconceptions and knowledge gaps. Plant and Soil 248, 31–41. Kearns JP, Cervini-Silva J, Fournelle J, Schloss P, Williamson L, Reuss R, Handlesman J and Banfield JF (2003). Mechanisms of Biogeochemical Influence on Phosphorus Bioavailability in Cold Terrestrial Ecosystems. Eos Transactions American Geophysical Union 84(46) Fall Meeting Supplement, Abstract B21C-0720, American Geophysical Union, Washington DC. Kirby C, Neal C, Turner H, and Moorhouse P, (1991). A bibliography of hydrological, geomorphological,
sedimentology, biological and hydrochemical references to the Institute of Hydrology experimental catchment studies in Plynlimon. Hydrology and Earth Systems Science 1, 755–763. Kirk JL, Beaudette LA, Hart M, Moutoglis P, Klioromos JN, Lee H and Trevors JT (2004). Methods of studying soil microbial diversity. Journal of Microbiological Methods 58, 169–188. Kozdroj JJ and van Elsas JD (2000). Response of the bacterial community to root exudates in soil polluted with heavy metals assessed by molecular and cultural approaches. Soil Biology and Biochemistry 32, 1405–1417. Kwong NG, Kee KF and Huang PM (1977). Influence of citric acid on the hydrolytic reactions of aluminium. Soil Science Society of America Journal 41, 692–697. Lam KC (1979). Production of runoff, solutes and sediments in a small catchment on the Northern Tablelands, New South Wales. PhD thesis. University of New England, Armidale, NSW. Leonard JA and Field JB (2003). The effect of two very different trees on soil and regolith characteristics. In Advances in Regolith. Proceedings of the CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp. 263–266. CRC LEME, Perth. Leonard JA and Field JB (2004). Differential distributions of cations in the regolith and vegetation. In Regolith 2004. Proceedings of the CRC LEME Regional Regolith Symposia 2004. (Ed. IC Roach) pp. 215–219. CRC LEME, Perth. Likens GE, Tartowski SL, Berger TW, Richey DG, Driscoll CT, Frank HG and Klein A (1997). Transport and fate of trifluoroacetate in upland forest and wetland ecosystems. Proceedings of the National Academy of Sciences of the United States of America 94, 4499–4503. Little DA (2001). The subalpine inverted treeline: soil patterns and nutrient processes. BSc (Hons) thesis. Australian National University, Canberra. Little DA (2007). The rhizosphere: organic acid microbe mineral interations. A multidisciplinary investigation of forest soil biogeochemistry in temperate Australian landscapes. PhD thesis. Australian National University, Canberra. Little DA and Field JB (2003). The rhyzosphere, biology and the reogolith. In Advances in Regolith.
213
214
Regolith Science
Proceedings of the CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp. 271–274. CRC LEME, Perth. Little DA, Field JB and Banks JCG (2003). Soil and regolith patterns and nutrient processes across an abrupt woodland-grassland ecotone. In Advances in Regolith. Proceedings of the CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp. 275– 277. CRC LEME, Perth. Little DA, Field JB, Welch SA and Rogers SL (2005a). Metal mobilisation from rhizosphere and nonrhizosphere soils using low molecular weight organic acids. In Regolith 2005 – 10 years of CRC LEME. Proceedings of the CRC LEME Regional Regolith Symposia 2005. (Ed. IC Roach) pp. 197–201. CRC LEME, Perth. Little DA, Welch SA, Macdonald LM and Rogers SL (2005b). Microbial community structural and functional diversity in the rhizosphere of cooccurring forest trees. In Regolith 2005 – 10 years of CRC LEME. Proceedings of the CRC LEME Regional Regolith Symposia 2005. (Ed. IC Roach) pp. 202– 206. CRC LEME, Perth. Longhurst W (1944). Observations on the ecology of the Gunnison prairie dog in Colorado. Journal of Mammalogy 25, 24–36. Mabbutt JA and Fanning PC (1987). Vegetation banding in arid Western Australia. Journal of Arid Environments 12, 41–59. Marilley L and Aargno M (1999). Phylogenetic diversity of bacterial communities differing in degree of proximity of Lolium perenne and Trifolium repens roots. Applied Soil Ecology 13, 127–136. Markewitz D and Richter DD (1998). The bio in aluminium and silicon geochemistry. Biochemistry 42, 235–252. Marschner B and Kalbitz K (2003). Controls of bioavailability and biodegradability of dissolved organic matter in soils. Geoderma 113, 211–235. Marschner P, Grierson PF and Rengel Z (2005). Microbial community composition and functioning in the rhizosphere of three Banksia species in native woodland in Western Australia. Applied Soil Ecology 28, 191–201. Martino E, Perotto S, Parsons R and Gadd GM (2003). Solubilization of insoluble inorganic zinc
compounds by ericoid mycorrhizal fungi derived from heavy metal polluted sites. Soil Biology and Biochemistry 35, 133–141. McLean JS, Lee J-U and Beveridge TJ (2002). Interactions of bacteria and environmental metals, finegrained mineral development and bioremediation strategies. In Interactions between soil particles and microorganisms: impact on the terrestrial ecosystems. IUPAC Series on Analytical and Physical Chemistry of Environmental Systems. (Eds PM Huang, J-M Bollag and N Senesi) pp. 227–262. John Wiley and Sons, Chichester, UK. Mellanby K (1971). The Mole. Williams Collins and Sons Ltd, Glasgow. Miklos AAW (1999). Stone-lines and oxic horizons: biogenetic organizations-soil fauna. In Abstracts, Symposium: Slope Processes that Produce Stonelines. Regional conference on geomorphology. 17–22 July, Rio de Janeiro. p. 103. International Association of Geomorphologists (IAG). Mitchell PB (1985). Soil bioturbation. PhD thesis. Macquarie University, Sydney. Mitchell PB (1988). The influences of vegetation, animals and micro-organisms on soil processes. In Biogeomorphology. (Ed. HA Viles) pp. 43–83. Blackwell, Oxford. Mitchell PB and Humphreys GS (1987). Litter dams and microterraces formed on hillslopes subject to rainwash in the Sydney Basin, Australia. Geoderma 39, 331–357. Moerkerken (2005). Distribution of precious metals in two dry sclerophyll forest trees: early steps in research at Tomakin Park Goldmine, New South Wales. In Regolith 2005 – 10 years of CRC LEME. Proceedings of the CRC LEME Regional Regolith Symposia 2005. (Ed. IC Roach) pp. 235–239. CRC LEME, Perth. Mylius SJ (1992). The Effects of Eucalypts on Soils of Southern New South Wales. BSc (Hons) thesis. Australian National University, Canberra. Nannipieri P, Ascher J, Ceccherini MT, Landi L, Pietramellara G and Renella G (2003). Microbial diversity and soil functions. European Journal of Soil Science 54, 655–670. Noguchi S, Tsuboyama Y, Sidle RC and Hosoda I (1999). Morphological characteristics of macropo-
Regolith and biota
res and the distribution of preferential flow pathways in a forested slope segment. Journal of the Soil Science Society of America 63, 1413–1423. Norman SA, Schatzl RJ and Small TW (1995). Effects of slope angle on mass movement by tree uprooting. Geomorphology 14, 19–27. Ognev SI and Tomilin AG (1947). The Mammals of the USSR and Adjacent Countries. Academy of Sciences Press, Moscow and Leningrad (Russian). Osterkamp WR, Toy TJ and Lenart MT (2005). Development of partial rock veneers by root throw in a subalpine setting. Earth Surface Processes and Landforms 31, 1–14. Pan WL, Bolton RP, Lunquist EJ and Hiller LK (1998). Portable rhizotron and color scanner system for monitoring root development. Plant and Soil 200, 107–112. Paris F, Bonnaud P, Ranger J, Robert M and Lapeyrie F (1995). Weathering of ammonium- or calciumsaturated 2:1 phyllosilicates by ectomycorrhizal fungi in vitro. Soil Biology and Biochemistry 22, 1237–1244. Paton TR, Humphreys GS and Mitchell PB (1995). Soils: A New Global View. UCL Press Limited, London. Petts A (2006). From the Tanami to Tanzania: Termite Adventures. In Regolith 2006 Consolidation and Dispersion of Ideas. Hahndorf. (Eds RW Fitzpatrick and P Shand) pp. 270–273. CRC LEME, Perth. Pryor G (1962). Australia’s Little Cornwall. Rigby Ltd, Adelaide. Putz FE (1983). Treefall pits and mounds, buried seeds, and the importance of soil disturbance to pioneer trees on Barro Colorado Island, Panama. Ecology 64, 1069–1074. Pye K (1982). Morphological development of coastal dunes in a humid tropical environment, Cape Bedford and Cape Flattery, North Queensland. Geografiska Annaler 64, 212ñ227. Qin F, Shan X-Q and Wei B (2004). Effects of lowmolecular-weight organic acids and residence time on desorption of Cu, Cd and Pb from soils. Chemosphere 57, 253–263. Raapana JAW and Field JB (2006). Macrozamia communis (L Johnson) as a biochemical biogeohemical
indicator of mineralisation on the South Coast of New South Wales. (Understanding regolith processes). In Regolith 2006 Consolidation and Dispersion of Ideas. Hahndorf. (Eds RW Fitzpatrick and P Shand) pp. 284–287. CRC LEME, Perth. Ramsey PW, Rillig MC, Feris KP, Holben WE and Gannon JE (2006). Choice of methods for soil microbial community analysis: PLFA maximizes power compared to CLPP and PCR-based approaches. Pedobiologia 50, 275–280. Rath H (1993). An investigation of the physical and chemical properties of soils under regrowth dry sclerophyll forest on the Southern Tablelands of NSW. BSc (Hons) thesis. Australian National University, Canberra. Rao L (2005). Spatial distribution of soil hydrophobicity under dry schlerophyll forests. Masters thesis, Fenner School, Australian National University. Reddell P, Yun Y and Shipton WA (1997). Cluster roots and mycorrhizae in Casuarina cunninghamiana: their occurrence and formation in relation to phosphorus supply. Australian Journal of Botany 45, 41–51. Reichman O and Smith SC (1990). Burrows and burrowing behaviour by mammals. Current Mammalogy 2, 197–244. Reid N, Hill SM and Lewis DM (2005). Tanami geobotany and biogeochemistry: towards its characterisation, role in regolith evolution and implications for mineral exploration. In Ten Years of CRC LEME. Proceedings of the CRC LEME Regional Regolith Symposia 2005. (Ed. IC Roach) pp.256–259. CRC LEME, Perth. Roach IC and Walker SD (2005) Biogeochemical expression of gold in native tree leaves at Wyoming and Tomingley, NSW. In Regolith 2005 – Ten years of CRC LEME. (Ed. IC Roach) pp. 277–282. CRC LEME, Perth. Roering JJ, Almond P, Tonkin P and McKean J (2002). Soil transport driven by biological processes over millennial time scales. Geology 30, 1115–1118. Sandnes A, Eldhuset TD and Wollebaek G (2005). Organic acids in root exudates and soil solution of Norway spruce and silver birch. Soil Biology and Biochemistry 37, 259–269.
215
216
Regolith Science
Schooley RL and Wiens JA (2001). Dispersion of kangaroo rat mounds at multiple scales in New Mexico, USA. Landscape Ecology 16, 267–277. Seguin V, Gagnon C and Courchesne F (2004). Changes in water extractable metals, pH and organic carbon concentrations at the soil-root interface of forested soils. Plant and Soil 260, 1–17. Selkirk PM, Adamson DA and Downing AJ (2001). Landform and vegetation change in the Greaves Creek Basin: an asymmetric hanging valley in the Blue Mountains, New South Wales. Australian Geographer 32, 45–75. Shakesby RA, Chafer C, Doerr SH, Blake WH, Wallbrink P, Humphreys GS and Harrington BA (2003). Fire severity, water repellency characteristics and hydrogeomorphological changes following the Christmas 2001 Sydney forest fires. Australian Geographer 34, 147–175. Sharpley AN, Syers JK and Springett JA (1979). Effect of surface-casting earthworms on the transport of phosphorus and nitrogen in surface runoff from pasture. Soil Biology and Biochemistry 11, 459ñ462. Sheldon ND (2003). Pedogenesis and geochemical alteration of the Picture Gorge subgroup, Columbia River basalt, Oregon. Geological Society of America Bulletin 115, 1377–1387. Singh BK, Millard P, Whiteley AS and Murrell JC (2004). Unravelling rhizosphere–microbial interactions: opportunities and limitations. Trends in Microbiology 12, 386–393. Smith OL (1982). Soil Microbiology: a Model of Decomposition and Nutrient Cycling. CRC Press Inc, Boca Raton, Florida. Smith BJ, Warke PA and Moses CA (2000). Limestone weathering in contemporary arid environments: a case study from southern Tunisia. Earth Surface Processes and Landforms 25, 1343–1354. Smithers R (1983). The Mammals of the Southern African Subregion. University of Pretoria. Pretoria, South Africa. Stahl N and Field JB (2003). Stream salinity in a coastal catchment, Bega, NSW: A conceptual model. In Advances in Regolith. Proceedings of the CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp. 391–393. CRC LEME, Perth.
Stone AT (1997). Reactions of extracellular organic ligands with dissolved metal ions and mineral surfaces. Reviews in Mineralogy 35, 309–344. Strakhov NM (1967). Principals of Lithogenesis. Oliver and Boyd, Edinburgh. Tebo BM, Ghiorse WC, van Waasbergen LG, Siering PL and Caspi R (1997). Bacterially mediated mineral formation: insights into manganese (II) oxidation from molecular genetic and biochemical studies. Reviews in Mineralogy 35,.225-266. Theng BKG and Orchard VA (1995). Interactions of clays with microorganisms and bacterial survival in soil: a physicochemical perspective. In Environmental impact of soil component interactions. Vol II. (Eds PM Huang, J Berthelin, J-M Bollag, WB McGill and AL Page) pp. 123–143. Lewis Publishers, CRC Press, Boca Raton, Florida. Thorn CE (1978). A preliminary assessment of the geomorphic role of pocket gophers in the alpine zone of the Colorado front range. Geografiska Anneler 60A, 181–187. Thorn CE (1982). Gopher disturbance: its variability by Braun-Blanquet vegetation units in the Niwot Ridge alpine tundra zone, Colorado Front Range, USA. Arctic and Alpine Research 14, 45–51. Ursino N (2007). Modeling banded vegetation patterns in semiarid regions: interdependence between biomass growth rate and relevant hydrological processes. Water Resources Research 43, WO4412. Valentin C, d’Herbes JM and Poesen J (1999). Soil and water components of banded vegetation patterns. Catena 37, 1–24. van Hees PAW, Jones DL, Jentschke G and Godbold DL (2004). Mobilization of aluminium, iron and silicon by Picea abies and ectomycorrhizas in a forest soil. European Journal of Soil Science 55, 101–111. Vancura V and Kunc F (1987). Interrelationships between microorganisms and plants in soil. In Proceedings of International Symposium. Liblice, Czechoslovakia, 22–27 June. Czechoslovak Academy of Sciences, Prague. (English version: Elsevier, Amsterdam; New York. 1989.) Viles HA (1988). Biogeomorphology. Blackwell, Oxford. Vogt KA, Vogt DJ and Bloomfield J (1998). Analysis of some direct and indirect methods for estimating
Regolith and biota
root biomass and production of forests at an ecosystem level. Plant and Soil 200, 71–89. Wakelin-King GA (1999). Banded mosaic (‘tiger bush’) and sheetflow plains: a regional mapping approach. Australian Journal of Earth Sciences 46, 53–60. Welch SA and Banfield JF (2002). Modification of olivine surface morphology and reactivity by microbial activity during chemical weathering. Geochimica et Cosmochimica Acta 66, 213–221. Welch SA, Taunton AE and Banfield JF (2002). Effects of microorganisms and microbial metabolites on apatite dissolution. Geomicrobiology Journal 19, 343–367. White ME (2003). Earth Alive: From Microbes to a Living Planet. Rosenberg Publishing, Dural, NSW. Whitford WG (2000). Keystone arthropods as webmasters in desert ecosystems. In Invertebrates as Webmasters in Ecosystems. (Eds DC Coleman and PF Hendrix) pp. 25–41. CABI Publishing, New York. Whitford WG and Kay FR (1999). Bioturbation by mammals in deserts: a review. Journal of Arid Environments 41, 203–230. Widmer F, Fleibach A, Laczko E, Schulze-Aurich J and Zeyer J (2001). Assessing soil biological characteristics: a comparison of bulk soil community DNA-, PLFA-, and BiologTM -analyses. Soil Biology and Biochemistry 33, 1029–1036. Wierzchos J, Ascaso C and McKay CP (2006). Endolithic cyanobacteria in halite rocks from the hyperarid core of the Atacama Desert. Astrobiology 6, 415–422.
Wilkinson MT, Humphreys GS, Chappell J, Fifield K and Smith B (2003). Estimates of soil production in the Blue Mountains, Australia, using cosmogenic 10Be. In Advances in Regolith. Proceedings of the CRC LEME Regional Regolith Symposia 2003. (Ed. IC Roach) pp. 441–443. CRC LEME, Perth. Wilkinson MT, Humphreys GS, Chappell J, Fifield K, Smith BL and Hesse P (2004). Soil production, landscape evolution and vegetation dynamics in the Blue Mountains, Australia. In American Geophysical Union, Fall Meeting 2004, abstract #H51C1147. Williams JP and Vepraskas MJ (1994). Solute movement through quartz-diorite saprolite containing quartz veins and biological macropores. Journal of Environmental Quality 23, 810–815. Wings O (2003). Observations on the release of gastroliths from ostrich chick carcasses in terrestrial and aquatic environments. Journal of Taphonomy 1, 97ñ103. Yuan L, Huang J, Li X and Christie P (2004). Biological mobilization of potassium from clay minerals by ectomycorrhizal fungi and eucalypt seedling roots. Plant and Soil 262, 351–361. Zak JC, Willig MR, Moorhead DL and Wildman HG (1994). Functional diversity of microbial communities: a quantitative approach. Soil Biology and Biochemistry 26, 1101–1108. Zakaria AS (1977). Controls upon the Mineral Outputs from three small catchments in New England. PhD thesis. University of New England, Armidale, NSW.
217
9
Regolith geophysics Tim Munday
9.1 INTRODUCTION Geophysical methods represent an important tool for investigating the sub-surface, and a variety of technologies have been developed with this purpose in mind. As such, they are potentially well suited to defining the variability associated with regolith, to understanding how thick it is and how it changes beneath the surface. Geophysics commonly involves measurements made at, or near, the Earth’s surface – with the observed response or measurement determined by the internal distribution of its physical properties. Analysis of these measurements by various means can reveal how the Earth’s physical properties vary spatially and with depth. In this chapter, particular consideration is given to the physico-chemical characteristics of regolith materials and how these affect observed geophysical responses. Those geophysical techniques that have a demonstrated application for mapping variability through the regolith are the focus for this review and, while the emphasis is on technologies that are notable for their effectiveness in the arid and semi-arid Australian conditions, mention is also made of methods that have application in the regolith of the humid tropics in India, Africa and South America. Only limited consideration is given to techniques appropriate for regolith devel-
oped under the more temperate conditions commonly found in North America and Europe, although all of those discussed have value in such environments and, in some instances, techniques such as ground-penetrating radar may be more effective. A full and comprehensive examination of geophysical methods for characterising the regolith is beyond the scope of this review. Rather, basic elements that may aid in determining which technologies might be used in mapping across regolith terrains, why they work and what they might resolve, are presented. References to more detailed exposition of the various technologies discussed are also presented. This is particularly so for borehole geophysical techniques, which, although often very effective for regolith characterisation – including the determination of petrophysical properties – commonly have only limited spatial coverage.
9.2 GEOPHYSICAL TECHNIQUES AND THEIR APPLICATION IN REGOLITHDOMINATED TERRAINS Geophysical techniques can be divided into two main categories: those that measure natural fields of the Earth – specifically gravitational, magnetic, electrical
220
Regolith Science
and electromagnetic – and those methods that use artificial sources to generate local electrical or electromagnetic fields, or seismic waves. Natural field techniques also lend themselves to data acquisition by air, whereas active source methods are commonly confined to ground surveys. One exception is airborne electromagnetics, which employs an active transmitter mounted on an aircraft, or towed beneath a helicopter, to induce a current in the ground, which, in turn, generates secondary fields in conductive material that are then detected and measured by a receiver. In airborne surveys, it is common to use several methods simultaneously: for example, magnetics and radiometrics. As a general rule, active source methods are capable of generating higher resolution data, and therefore a more detailed picture of the sub-surface, compared with natural source methods. Applied geophysical techniques that may be used in the elucidation of regolith landscapes can be divided into the following methods of exploration:
s s s s s s s s
gravitational magnetic seismic electrical, electromagnetic radioactivity radar borehole logging.
The choice of method often depends on the target characteristics, the scale at which information is required and the cost or budgetary constraint. For example, the determination of the regional distribution of buried paleochannel systems might employ airborne electromagnetic (EM) or magnetic surveys, whereas understanding the geometry, and variability of sediments within a paleochannel in order to target a borehole, would be best achieved by ground seismic, time domain EM, gravity or electrical resistivity imaging (ERI) traverses (see Table 9.1 for commonly used geophysical abbreviations). Some geophysical methods provide a direct indication of the presence of a material being sought: for example, magnetics can be used to identify the location and concentration of maghemite gravels in transported cover, whereas in other situations the method employed may only indicate the nature of the regolith
system by defining its geometry and extent (such as a paleochannel system imaged by airborne EM data). It is also perhaps worth noting that geophysical techniques only detect discontinuities in the sub-surface (Telford et al. 1990). Consequently, in regolith settings, variations in the observed geophysical response arise when one part of the regolith differs significantly from another part. The use of geophysics in regolith environments presents significant challenges – particularly as regolith materials have physical properties that have a profound effect on geophysical response (see for example, reviews by Doyle et al. 1981; Butt 1981; Palacky 1981, 1987, 1989; Doyle and Lindeman 1985; Smith and Pridmore 1989; Dentith et al. 1994; Papp 2002). Deeply weathered materials and an alluvial-colluvial cover have a varied influence on the observed response from the underlying bedrock – particularly to those exploring for mineral systems through the regolith – causing problems in investigation and interpretation. For example, Stolz (2000) described a situation where conductive regolith masked the response of deep Nisulfide ore bodies to ground EM systems, with IP (induced polarisation) effects and other geological noise originating in the regolith making recognition of weak bedrock anomalies difficult. However, these characteristics may also be used to advantage when trying to characterise the regolith itself. In these instances, the source of (geological) noise becomes a signal, which, with careful analysis, can elucidate the character of regolith materials at depth. From the perspective of mapping regolith variability in three dimensions, various geophysical techniques have relevance, but, to be effective, it may help to have some knowledge and information about the types of materials that may be encountered at changing depths below the ground surface, if only to assist in the choice of method to be employed. Features of the regolith that both enhance and hinder the application of geophysical methods through the regolith are summarised in Table 9.2.
9.3 PETROPHYSICAL PROPERTIES OF REGOLITH MATERIALS The physical basis for mapping vertical and lateral variations in the regolith can be related to their
Regolith geophysics
Table 9.1: Commonly used geophysical abbreviations (this chapter and Chapter 11). AEM
Airborne ElectroMagnetic
AGG
Airborne Gravity Gradiometer
ASTER
Advanced Spaceborne Thermal Emission and Reflection Radiometer
CST
Constant Separation Traversing
DEM
Digital Elevation Model
EM
ElectroMagnetic
EQMMR
Equivalent MagnetoMetric Resistivity
ERI
Electrical Resistivity Imaging
ERT
Electrical Resistivity Traversing
FD
Frequency Domain
GAP
Ground Acoustic Penetration
GIS
Geographic Information Systems
GPR
Ground Penetrating Radar
HEM
Helicopter ElectroMagnetic
IP
Induced Polarisation
Landsat MSS
Landsat MultiSpectral Scanner
Landsat TM
Landsat Thematic Mapper
MIP
Magnetic-Induced Polarisation
MMR
MagnetoMetric Resistivity
MT
MagneTotelluric
MTM
MicroTremor array Method
MRS
Magnetic Resonance Sounding
NOAA
National Oceanic and Atmospheric Administration (US)
RAP
Resonance Acoustic Profiling
SAM
Sub-Audio Magnetics
SNMR
Surface Nuclear Magnetic Resonance
SP
Self Potential
SPOT
Satellite Pour l’Observation de la Terre
SQUIDs
Superconducting Quantum Interference Devices
TD
Time Domain
TEM
Time domain ElectroMagnetic
TFMMIP
Total Field MagnetoMetric-Induced Polarisation
TFMMR
Total Field MagnetoMetric Resistivity
TMI
Total Magnetic Intensity
UAV
Unmanned Aerospace Vehicle
VDI
Vertical Derivative Image
VES
Vertical Electric Sounding
221
222
Regolith Science
Table 9.2: Features of regolith-dominated terrains that both enhance and hinder the application of geophysical methods in regolith dominated terrains (adapted from Smith and Pridmore 1989; Dickson and Scott 1997; Davis and Annan 1989). Negative features
Positive features
Thick (up to 100 m+) and variable oxidised cover – can result in a complex observed response.
Relatively gentle topography in arid environments such as Australia; aids the acquisition of low-level airborne data and sparse vegetation cover assists easy access.
Commonly highly conductive, but not always. Large and highly variable electrical properties often generate a strong and confusing background response, which can mask bedrock responses.
Highly conductive regolith can assist in the mapping of regolith thickness using electrical, electromagnetic and MMR methods.
Highly conductive, damp regolith attenuates high frequency radar signals.
Dry, sub-horizontal regolith unit boundaries and variations in the near surface lend themselves to detection using GPR.
Transitional, gradational changes in material properties limit the application of seismic methods in the near surface.
Marked velocity contrasts between the regolith and fresh bedrock assist the mapping of this boundary using shallow seismic methods.
Saline groundwater – very conductive making electrical methods difficult when the regolith is saturated with saline pore water.
Differential weathering – promotes contrasting physical properties that can be used to map buried lithostructural boundaries.
Development of maghemite – source of noise in magnetics and super paramagnetic effects in ground EM.
Maghemite may concentrate in paleochannels – assisting in their mapping when using high-resolution airborne techniques.
Low density and marked density contrasts in the regolith can generate anomalies of the same order as those from the underlying bedrock.
Variable density regolith can assist the use of gravity for paleochannel mapping.
Transported cover can obscure variations in radioelement distribution associated with weathered lithologies.
Concentration of Th and U in Fe-rich materials can aid the understanding of landscape development.
petrophysical properties. Regolith materials exhibit a range of densities, elastic moduli, magnetisation and conductivity (resistivity), and contrasts within the regolith between transported cover and underlying materials, and at the weathering front combine to produce a range of geophysical responses that are often difficult to interpret and analyse. These complexities have become more significant with the development of more sensitive geophysical technologies, and the acquisition of data at finer scales, along with improvements in data modelling and inversion. Consequently, as geophysical data analysis becomes more refined, there has been accompanying demand for more accurate information on the (petro)physical characteristics of geological materials, including the regolith, to assist their interpretation. Unfortunately, information on the petrophysical properties of regolith materials remains somewhat fragmented, and studies that have focussed specifically on this issue are, at best, limited.
Exceptions are those reported by Emerson et al. (2000) and Emerson and Macnae (2001). They examined the petrophysical characteristics of different regolith units developed over contrasting lithologies found in the Yilgarn Craton of Western Australia (Table 9.3). The petrophysical measurements reported are not invariably identical to those that would be measured in situ, as sampled materials are likely to be differently hydrated and under different confining pressures, and the sample collection process tends to provide a bias towards the more robust samples. Nonetheless, the measurements constrain the range of possible physical properties of regolith. Other studies which have examined the petrophysics of the Australian regolith include those of Tracey and Direen (2002), Wilford (2002), Rutherford et al. (2001), Bell et al. (2001), Lawrie et al. (2002), Wildman and Compston (2000) and Emerson (1990). Elsewhere, Palacky (1987), Palacky and Kadekaru (1979),
Regolith geophysics
Davis and Annan (1989), Telford et al. (1990) and Robineau et al. (2007) have examined the petrophysical properties of rocks and the regolith from other regions around the world. Results from this body of work indicate that the regolith commonly exhibits decreasing porosities from the surface to the saprock (see Figure 6.4), with very high values (averaging around 30%) in the saprolite, which are largely responsible for the low densities, velocities and resistivities. However, the ion mobility of the constituent clays, moisture content and the presence of conductive solids will influence their electrical properties. A key driver to the observed electrical response of regolith materials is the salinity of the groundwater – with high salinities significantly elevating the observed conductivity response. High salinities and moist profiles tend to attenuate high-frequency signals from radar signals – limiting their depth of penetration. Transported regolith commonly contains maghemite gravels, which are concentrated through geomorphic processes and accumulate in or adjacent to paleochannels (Section 4.4.3). Consequently, these materials exhibit high magnetic susceptibilities. This tends to contrast with in situ regolith where measured magnetic susceptibility decreases as a result of the oxidation of magnetite, on weathering, to hematite or goethite. Occasionally, higher susceptibilities may be observed with local concentrations of magnetite in association with shear zones within the regolith, or in saprolite where magnetite may occur as magnetite rims on chromites (for example, Brand 1997). In another seminal study – underpinning our understanding of how to interpret airborne radiometric data sets – Dickson and Scott (1997) determined the distribution of the radioelements K, U, and Th in different rock types and in overlaying regolith materials and soils as part of a study to examine the changes in radioelement distribution brought about by weathering and pedogenesis. Their study shows a general loss of K with weathering, whereas U and Th are lost from felsic rocks, but enriched in soils derived from intermediate and basic rocks. Overall, further studies on a larger sample group of regolith materials are warranted to explore further the systematics of physical property variation between and within regolith material types.
9.3.1 Regolith characteristics and influence on observed geophysical response The influence of particular petrophysical characteristics of the regolith on the observed geophysical response are discussed below. Magnetic
Resistant magnetite and secondary maghemite may be concentrated by weathering and landscape development (Sections 4.4.3 and 5.4.3). Weathering of magnetite can produce non-magnetic species of Fe oxide in the regolith (Dentith et al. 1992). However, this can vary and, in some cases, the magnetic character may be preserved. When maghemite is developed it can display a variable remanence, susceptibility and super-paramagnetic behaviour (Clark and Emerson 1991). Maghemite gravels may be concentrated via geomorphic process and accumulate in paleochannels. In some landscapes, ground magnetic data are dominated by short-wavelength anomalies with very high amplitudes that are often attributable to the presence of maghemite in gravels, which are present either as lags or as bands in alluvium and colluvium. Gravity
The observed shape of a gravity profile across regolith-dominated terrain is a function of the gravity responses of the regolith and the underlying basement (Braine and Macnae 1999). The relatively low densities (compared with unweathered rock) and marked density contrasts that characterise regolith materials can combine with variations in thickness to generate anomalies of the same order as those relating to bedrock. Studies by McCrea et al. (1990), Kew and Gilkes (2007) and others concerning the densities of regolith materials have indicated that differences in bulk density may be attributable to the amount the quartz present in the parent material. Consequently, separating the influence of underlying lithologies, which can introduce high background levels in gravity data from the response relating to the regolith, can be difficult in some circumstances. Seismic
The complex zonation of a regolith profile, along with transitional boundaries and irregular interfaces
223
2–10 1–3
1–20
3-15
Ferricrete (loose pisoliths, nodules nodular duricrust)
Mottled zone
Ferruginous saprolite
65
18
173 (149)
3 (clay) –12 (sands)
5–10
Channel or lacustrine clays
1000 (dry) 10 (wet)
Resistivity (1m) (1kHz)
2384
10–15
Nominal thickness (metres)
Opaline silica
Alluvium, colluvium –commonly forms a hardpan
Regolith material (from surface to bedrock)
2.26 (0.23)
2.02 (0.17)
2.86 (0.34)
2.25
2.32
2.23 (0.31)
Dry bulk density (g/cm3)
94 (71)
25 (16)
5024
15
4
3603
Magnetic susceptibility (k (si × 10 –5))
24 (10)
22 (5)
19 (6)
15
8
23 (5)
Apparent porosity (%)
2416
3358 (569)
3867
2148 (507)
P-wave velocity (m/s) (500kHz)
Hematite, goethite, kaolinitic clay; ~50 % Fe oxides diffused through clay-rich saprolite
Goethite, hematite + clays
Hematite, goethite, kaolinitic clay, ± maghemite
Frequently kaolinitic, may be gritty or silicified, commonly mottled, grey clays are more smectitic
Clay/Fe saprolitekaolinitic lithic clasts, with ferruginous clasts, sandy clay and clay horizons. Crypto-crystalline silica cement in hardpan, maghemite in finer grain fraction
Mineralogy
Highly weathered rock with preserved texture; discontinuous, containing nodular material as breccia fragments; ferruginous saprolite commonly has a collapsed upper zone without original rock texture
Segregations of Fe within a clay-rich matrix
Irregular sub-horizontal distribution as discontinuous lenses, sheets and pods; duricrust is hard and massive with porphyritic texture; Fe rich over mafic/ultramafics
Curvilinear or basinal, not always present, not part of Archean weathering
Hardpan imparts laminated (sub-) horizontal fabric
Features
Table 9.3: Nominal physical characteristics (mean values (and standard deviations in brackets)) of physical properties of regolith materials characteristic of the Eastern Goldfields in Western Australia (adapted from Emerson et al. 2000, Berkman 1995, Bishop et al. 2001 and Tracey and Direen 2002).
224 Regolith Science
Variable
Saprock
Notes
1.
5–100+
Saprolite
Fresh rock
Nominal thickness (metres)
Regolith material (from surface to bedrock)
2.96 (0.16)
2.66 (0.22)
1.78 (0.28)
Dry bulk density (g/cm3)
96 (59)
743
112
Magnetic susceptibility (k (si × 10 –5))
1.6 (1.6)
8 (6)
35 (10)
Apparent porosity (%)
5955 (540)
4540 (935)
2030 (519)
P-wave velocity (m/s) (500kHz) Mineralogy
as for saprock, but no alteration
As for bedrock, but with minor alteration of quartz, feldspar, amphibole, pyroxene, mica, talc, chlorite, carbonate ± magnetite and/or pyrrhotite/ pyrite
Smectite and kaolinite clays, some zones of re-precipitated silica; various original rock silicate minerals in progressive stages of breakdown
Features
Commonly steeply dipping lithological units (pyroxenites, dolerites, gabbros) sometimes sheared, meta-sediments, granite–gneiss
Slightly weathered Archean bedrock with texture preserved; isovolumetric weathering; <20% of weatherable minerals altered.
Weathered bedrock; fine fabrics are retained; multi-coloured clay rich; >20% of weatherable minerals have been altered; proportion of clay may increase upwards; secondary cementation by Fe oxides, silica and carbonates not uncommon; Quartz veins and Fe segregations may be present.
Stratigraphy is notional and thicknesses nominal. Considerable lateral and vertical variations are common.
> 5000+
131 (68)
7 (5)
Resistivity (1m) (1kHz)
Regolith geophysics 225
226
Regolith Science
in solution) is low. Their significance becomes negligible at high ionic concentrations, particularly for clays with low CECs, such as kaolinite (Table 5.3; Emerson and Yang 1997). Marked conductivity contrasts are observed within a regolith profile, which affects electrical and electromagnetic data. Regolith materials are commonly porous and contain water along with dissolved salts. The presence of conductive clays and saline groundwaters yield conductivities in ranging between 10 and 10 000 mS/m (resistivities of 0.1 to 1001m). Polarisable clays in the regolith can cause induced polarisation (IP) effects in electromagnetic (EM) data. In areas of high conductivity, electromagnetic coupling can result – superimposing signal (noise) on deeper sources. At the base of the weathering profile, the transition between saprolites to essentially unweathered rock is reflected in a marked change in conductivity (Figure 9.1). The base of weathering can be very irregular and a deep, conductive regolith can abut unweathered resistive rocks, which may be close to the surface. Rocks that may be resistive to weathering can remain unaltered high in the weathering profile. Where they exhibit physical properties that cause them to be either resistive or conductive – depending, for example, on the concentration of dissolved salts in contained pore water – then the application of electrical methods and
hinder the application of seismic methods. However, the marked velocity contrasts between the regolith and bedrock may aid the application of shallow seismic techniques in resolving the nature and geometry of the saprolite–saprock/bedrock boundary. Similar velocity contrasts may be usefully resolved higher in the regolith. Electrical methods
The electrical conductivity (the reciprocal of resistivity) of the regolith is dominated by an electrolyte (salt) that occurs in moisture-filled pores within an insulating matrix (McNeill 1980, 1990). The electrical conductivity of the bulk regolith is primarily controlled by: 1. the concentration of dissolved electrolytes – the concentration of ionic conductors in solution 2. the amount and composition of clays – particularly those with a moderate to high cation exchange capacity (CEC)( see Section 5.3.7) 3. the moisture content – the extent to which the pores are filled by water 4. the porosity: shape and size of pores, number, size and shape of interconnecting passages 5. the temperature. Clay content and type become important when the concentration of ionic conductors (for, example, salts
Breakaway Ferruginous saprolite
Land surface
Ferricrete
Colluvium/alluvium Paleochannel
100 m+
Regolith
vp Saprock
1000 1000 dry - 50 wet
0.005
200
0.01
100
0.05
20
1-100
Ore body
Resistivity m 1000+
- 0.02
0.05-0.02
vp v
10-3 10-3
Saprolite
v
Conductivity S/m <10-3
5-20 0.01-1
Shear zone v
vp v
v v Unweathered rock
vp vp
Fault
pv pv
Unweathered rock
Figure 9.1: Schematic regolith section and representative conductivity/resistivity attributes for different regolith materials.
Regolith geophysics
the interpretation of derived data can become very difficult (Butt 1981). Ground-penetrating radar (GPR)
Variations in the electrical properties of regolith materials are primarily controlled by their volumetric water and solute content, which, in turn, give rise to radar reflections. These variations determine the velocity, attenuation and the power of the radar signal that is reflected at boundaries where electrical properties vary in the ground. The dielectric and conductivity properties of the regolith also determine the velocity and attenuation of high-frequency radar waves through the ground. The presence of conductive materials within the regolith can significant affect the range and resolution of GPR. Radiometrics
The radiometric response from weathered materials can vary significantly (Dickson and Scott 1997). Transported overburden will effectively mask the response from underlying bedrock. These materials are often characterised by a relatively high response in Th and U, which can be related to the concentration of resistant host minerals (such as zircon and monazite) or by incorporation into neo-formed Fe oxides (Appendix 2). Geomorphic processes control the distribution and relative concentration of radioelements. Gamma-ray spectrometric surveys may be used to help elucidate the nature and distribution of regolith materials in the landscape and contribute to an understanding of its development (see Section 9.4.1). It should be noted that approximately 90% of the recorded radiometric response comes from the top 30 cm of the regolith. 9.3.2 Implications for mapping the regolith Given these petrophysical characteristics, how can they be used to help map the regolith? Magnetic susceptibility data indicates a moremagnetic transported cover and, hence, suggests the possibility of mapping the presence and variability of these layers. Dauth (1997) and Wildman and Compston (2000), among others, have described the application of ultra-high-resolution airborne magnetics for mapping concentrations of the more magnetic
components (maghemite) in transported regolith cover. Dentith et al. (1992) used an increase in susceptibility with increasing depth in the Southern Cross Greenstone Belt (Yilgarn Craton, Western Australia) to map the regolith base. Although magnetic techniques show promise, further work is required to better characterise the magnetic behaviour of regolith materials other than maghemite in order to realise their full potential. Although variations in the density of the regolith often show no clear pattern, it has long been noted among Australian practitioners of the gravity method, that variability in the overburden in some geological settings is often relatively well defined by the gravity data. Examples include the Yilgarn Craton of Western Australia, where, even though there may be dense units within the regolith, these are generally not significant when compared with its thickness and to the density contrast between the regolith and underlying basement. In that environment, gravity can be an effective tool for mapping the extent of thick paleochannel sedimentary sequences (for example, Smyth and Barret 1994). Available velocity data clearly demonstrates that the depth to fresh bedrock – that is, the base of the saprock – should be obtainable with seismic techniques and this has been demonstrated in several published papers (for example, Dentith et al. 1992; Leslie et al. 2000). Discontinuities within the regolith caused by cementation, or by marked changes in material type (for example, transported gravels over a clay-rich saprolite), may result in changes in velocity that are detectable. Resistivity data for different regolith materials indicates that the saprolites are the most conductive feature of the regolith and that electrical/electromagnetic surveys should be able to define the contact between the saprolite and underlying saprock, which is around 30 times more resistive. Electromagnetic discrimination between saprolite and saprock has been reported in several studies (for example, Worrall et al. 2001; Wilford et al. 2002; Stolz 2000). Where there is a resistivity contrast between transported cover and in situ regolith, then this boundary may also be defined (for example, Munday et al. 1998, 2001; Munday and Sumpton 1999).
227
228
Regolith Science
9.4 GEOPHYSICAL TECHNOLOGIES, PRINCIPLES AND APPLICATIONS As indicated in Table 9.3, particular attributes of regolith materials lend themselves to measurement with a range of geophysical techniques. Some of these techniques are better suited to some settings and conditions than others. Table 9.4 summarises the principal technologies having application in exploring through, and within, the regolith. All the geophysical techniques described below can be deployed in ground surveys. Some have application in airborne surveys, where speed and area of coverage present particular advantages. However, these advantages can be offset by a need for detail and more direct measurements, which can only be forthcoming using ground surveys or borehole methods. In the following section, we briefly examine relevant technologies, their mode of operation and how they may see application in the study of the regolith. We also offer comment on several techniques that might be regarded as not being in the mainstream at present – including passive seismics, seismoelectrics and ground acoustic penetration. These have indicated some potential to inform on regolith characteristics, but they require further investigation and evaluation. 9.4.1 Technology, operating principle and application Gravity
Gravity surveying – a natural source method – commonly involves measurements of subtle changes in vertical gravitational acceleration caused by density contrast in the sub-surface geology. Gravity anomalies occur at boundaries between rocks and regolith materials that have different densities. Measurements are made by a gravimeter or a gradiometer. A gradiometer measures the spatial rate of change in the gravity field: measuring high-frequency signals associated with near-surface lateral density variations more clearly than vertical gravity field instruments. Until more recently, observations were made at the Earth’s surface with gravimeters positioned on the ground over a widely spaced grid either by vehicles or helicopters (for example, Lo et al. 1999). In groundbased gravity surveys, the vertical component of the
acceleration due to gravity is measured. Ground-based techniques have seen extensive application across regolith terrains (for example, Shevchenko et al. 2001; Meyers et al. 2001; Bell et al. 2001; Wilkes et al. 2004, 2006). Applications of gravity methods include the location of paleochannels (Figure 9.3, page 160) and geological mapping through thick regolith cover, including estimating regolith thickness in areas of good constraint (Figure 9.2, page 159). Recent developments in airborne gravimeters and gradiometers have seen a rapid expansion in the use of this method across a variety of regolith dominated environments, including Australia and overseas (for example, Lane 2004; Dransfield 2007). Airborne systems measure differential curvature gravity gradients and these are transformed through processing to recover the vertical gravity and the vertical gravity gradient. With some gradiometers (such as the BHP Billiton FALCON™ airborne gravity gradiometer (AGG)) other components of the full tensor are also recovered (Lee 2001), although, from a mapping perspective, the vertical gravity gradient is the most useful: providing greater resolution and greater sensitivity to small sources by comparison with the vertical gravity component. There is no question that high-precision ground surveys currently provide geological detail beyond the capability of airborne systems, but, with the use of airships and helicopters, we are now realising veryhigh-resolution gravity data from the air, combined with the benefits of rapid data acquisition and coverage (see Dransfield 2007). Magnetics
Magnetic surveys, which are undertaken with a magnetometer on the ground or in the air, measure the strength and/or the direction of the magnetic field in the vicinity of the instrument. There are two basic types of magnetometers: the scalar magnetometer, which measures the total strength of the magnetic field; and the vector magnetometer, which measures the component of the magnetic field in a particular direction. Examples of vector magnetometers include fluxgates and superconducting quantum interference devices (SQUIDs). The benefits arising through the measurement of vector information, relating to both
Regolith geophysics
the components and gradients of the Earth’s magnetic field, include: the better definition of structural features; continuous mapping of the magnetisation direction of magnetic source rocks, of remanence indicators and of magnetic susceptibility distribution; and mapping of geological boundaries (Schmidt and Clark 2000). Observed variations in the Earth’s magnetic field (geomagnetic field) result from the magnetic properties of the regolith, where present, and underlying rocks. All magnetic anomalies caused by the regolith or unweathered rocks are included in the measured geomagnetic field. An understanding of the local magnetic anomalies is necessary to enhance and interpret these features. Although most regolith materials are non-magnetic, anomalies can occur where the most common magnetic mineral – magnetite, is present in significant quantities. As mentioned previously, in the Australian regolith this is commonly as maghemite gravels, which have accumulated in paleovalley sediment fill during landscape development, or as a surface lag across the contemporary land surface. While ground magnetic surveys are conducted for high-resolution definition of structures and other buried features, improvements in airborne magnetic surveys in the past four decades – with advances in both data acquisition, with very low level surveys, and image processing techniques – have seen the widespread use of these data in exploration and mapping through regolith terrains (Brodie 2002). Specific so-called ‘regolith’ filters have been developed to separate the magnetic effects of regolith materials from fresh basement rocks (Dauth 1997; Gunn et al. 1997). The application of the Euler deconvolution method for estimating depth to basement from airborne magnetic data sets has been described by Milligan et al. 2004. Excellent examples of using the magnetic response of regolith materials that have accumulated in paleovalleys to map paleodrainage systems have been presented by Mackey et al. (2000), Wildman and Compston (2000), Lawrie et al. (1999) and Gibson and Wilford (2002). An illustration of their appearance in aeromagnetic data is presented in Figure 9.4 (page 161). How they are used to construct a regolith landform map is illustrated in Figure 9.5 (see also Chapter 11)
Electrical
There are many methods employed in electrical geophysical surveying. Some make use of naturally occurring fields in the Earth, while others rely on the generation of artificial currents in the ground. Electrical methods, which are principally ground based, use direct currents or low-frequency alternating currents to investigate the electrical properties of the subsurface. Among the most common techniques having application for exploration through the regolith are the resistivity method, the induced polarisation (IP) method, and the self-potential (SP) method (Telford et al. 1990). However of these, only the resistivity method has seen wide application in the characterisation of regolith materials. More recently, technical developments have seen a variation on the resistivity and IP methods, which follow the magnetometric resistivity (MMR) and magnetic induced polarisation (MIP) techniques reported by Seigel and Howland-Rose (1990) and involve the measurement of the magnetic field induced by a time-varying current applied to the Earth. Referred to as sub-audio magnetics (SAM) (Cattach et al. 1993), this approach has seen wide employment in exploration across regolith-dominated environments in the past decade. Of particular interest is the ability to use SAM in highly conductive ground conditions, such as salt lakes. Resistivity
This method involves transmitting an electric current into the ground between two electrodes and measuring the resulting potential differences. This directly measures the apparent resistivity of the area beneath the electrodes and includes deeper layers as the electrode spacing is increased. Horizontal and vertical discontinuities in the ground resistivity result from variations in regolith type, bedrock lithology, water content and pore-water chemistry. The spacing of electrodes can be increased about a central point – resulting in a vertical electric sounding (VES), which can then be modelled to create a geoelectric sounding. A variety of electrode configurations have been designed for the resistivity surveys, but the most common are the Wenner and Schlumberger array configurations (Telford et al. 1990). The Wenner array is particularly suited to high-resolution near-surface
229
Theory
Measures electrical resistivity in the sub-surface, which varies due to regolith physical, mineralogical properties and pore fluid chemistry.
Measures changes in the propagation of high-frequency electromagnetic energy in the ground to produce an image of the sub-surface.
Measures subsurface conductivity through low frequency electromagnetic induction. Response a function of regolith physical properties and chemistry of pore fluids.
Bedrock topography
Regolith thickness
Fractures, faults, shears
Lithostructural boundaries
Electrical resistivity
Bounding surfaces
Sedimentary layering structures
Sub-horizontal boundaries and bounding surfaces in the regolith
Voids
Groundpenetrating radar (GPR)
Regolith bounding surfaces
Bedrock topography
Fractures, faults, shears Regolith thickness
Lithostructural boundaries in the basement
Electromagnetic frequency and time domain
Measures seismic velocity of regolith materials and underlying fresh rock. Observed response a function of density and physical properties.
Faults, fractures
Internal bounding surfaces and layering
Bedrock topography
Regolith thickness
Seismic reflection and refraction
Measures gamma-rays emitted from the Earth’s surface. Quantity of K, U, Th and daughters.
Spatial distribution of materials and erosion
Regolith material type at the surface
Radiometrics
A summary of geophysical technologies, applications to regolith studies and limitations.
Applications for mapping the regolith
Table 9.4:
Gravity
Measures Earth’s gravity field and small anomalies caused by density contrasts.
Paleochannels
Faults
Bedrock topography
Overburden thickness
Magnetics
Measures Earth’s magnetic field and anomalies caused by magnetic minerals in the regolith or basement.
Mineralogy
Maghemite gravels
Borehole
Measures resistivity, inductive conductivity, velocity, density, SP, radioactive properties, or hole width – depending on tool employed.
Regolith physical and chemical properties
Lithology
230 Regolith Science
Groundpenetrating radar (GPR) Antenna pulled by hand or vehicle, with continuous measurements possible. Data are recorded digitally. Potentially up to 100 m, but usually in top 20 m. Less in conductive ground. Exploration depth in saline environments very limited.
Electromagnetic frequency and time domain
Stationary, sledge, or manual surveys. Data acquired in stationary mode, or continuously. Data recorded digitally.
Frequency domain ~10 cm to 100 m Time domain ~1–100s m
Noise from power lines, fences, infrastructure, storms.
Methodology
Investigation depth
Limitations
Array length is several times the depth of investigation.
Near surface to >100 m. Depth of investigation increases as electrode array is expanded
Electrical current is injected into the ground though a stationary or moving array of electrodes. Data is recorded digitally.
Electrical resistivity
Requires velocity contrasts between materials. Refraction requires velocity to increase with depth.
Refraction: 0–100s m Reflection: 5–100s m
Energy is induced at the surface by vibration or shots. Arrival of seismic S or P wave is sensed by geophones and recorded digitally.
Seismic reflection and refraction
Near surface to 1000s m
Depth, target size, and density contrast determine detectability.
Limited depth of investigation.
Ground stations, with digital GPS locations and, more recently, in the air.
Gravity
Top 30–45 cm of the Earth’s surface
Gamma-ray data are collected on the ground or in the air using sodium-iodide scintillation detectors.
Radiometrics
Noise from infrastructure, detectability depends on size and depth of target.
Near surface to 1000s m
Data may be acquired on the ground, in vehicle, or in the air using a magnetometer. Data are recorded digitally.
Magnetics
Vary with tools employed. Casing type may limit tools.
Depth of drill hole
Probes are lowered by cable down hole. Data are recorded digitally.
Borehole
Regolith geophysics 231
232
Regolith Science
512000E
516000E
520000E
524000E
147°10'
528000E
147°15' C
A3
G2
Billys Lookout
NORTHERN
6268000N
TERRITORY WESTERN
A3
AUSTRALIA
G2
QUEENSLAND SOUTH AUSTRALIA
G2
Forbes 1:250 000 sheet area
33°45'
G1 G2
6264000N
TASMANIA
Magnetic paleodrainage deposits
G1
Gold bearing quartz veins
G2
A1
SD2
Boundary between depositional and erosional areas
G1
A2
G1
NEW SOUTH WALES VICTORIA
Hard rock gold workings (shafts) 6260000N
G1
G1
SD3
G1
G2
0
5 km
Hiawatha Goldfield
A2
G2 33°50'
6256000N G1
A1
Seismic Line (Fig 9.12) G2 SD1
G1 6252000N
A2
G2 G2
G1
G1 O1
D1
A1
G1
A3 G1
6248000N G1
O1
Wyalong
D2 Goldfield
33°55'
O2 6244000N O1 A3 A1 6241000N
O1
A2 TMf007-08 16/l55/44
A1 A2
Alluvium/colluvium in modern valley floors; sediment derived from low-response weathered granite. Alluvium/colluvium in modern valley floors; sediment derived from high-response Ordovician metasediments, and high-response weathered granite. A3 Thick alluvium in Bland Creek paleochannel; variable response depending on provenance. C Granite-derived colluvium forming distal low-angle colluvial fan; high response. G1 Granite, highly weathered to 60 m-100m lag of magnetic pisoliths, veneer of residual/colluvial sediments; low-response. Erosional plains and rises. G2 Granite, mostly highly weathered; high-response; erosional plains and rises. D1 Diorite, fresh; high response; steep rise. D2 Diorite, weathered, masked by residual/colluvial deposits; low-response; erosional plains and rises. SD1 Paleozoic strata, fresh; high total response; steep ridge. SD2 Paleozoic strata, weathered outcrops on rises; high-response. SD3 Paleozoic strata, weathered, veneer of residual/colluvial sediments; low-response; plains and rises. O1 Ordovician metasediments, slightly weathered; high-response; low ridges. O2 Ordovician volcanics, highly weathered; low-response; erosional plains and rises.
Figure 9.5: Regolith landform map from a combined interpretation of the airborne radiometrics and magnetic data sets shown in Figure 9.4 (page 161). The erosional terrain largely comprises in situ highly weathered granitic plains and rises. G1 has a low-potassium and moderate- to low-Th signature, and is characterised by ferruginous lag derived from mottled granite. G2 has a high-K signature, and appears to be less weathered. Magnetically delineated paleochannels (grey screen), containing highly magnetic detrital ferruginous pisoliths, sand and clay, cross this area, and exit to the north-east. Colluvial and alluvial deposits associated with modern drainage form a veneer across these units with two distinct provenances (A1 and A2). Variably weathered and covered diorite (D1 and D2) and sedimentary rocks (SD1- 3 and O) surround the low-relief granitic landscape. In the north-east, granitic colluvium (C) is associated with steep granite hills in the north-east (Lawrie et al. 1999).
Regolith geophysics
investigations and for its ease of operation (Griffith and Barker 1993). Recent advances in microprocessor control technology, allow for the rapid collection of adjacent multiple soundings along a transect, which are then modelled to create a 2D geoelectric cross-section showing lateral and vertical variations of resistivity with depth (see for example, Zhou et al. 1999). Referred to as constant separation traversing (CST) or electrical resistivity traversing/imaging (ERT/ERI), this approach generates an image that is essentially a smooth or blurred interpretation of the true electrical cross-sectional structure of the sub-surface (see Figure 9.6, page 162). There is abundant literature describing the successful use of resistivity methods for regolith investigations in a variety of settings, including the weathered volcanics in Nicaragua (Mendoza and Dahlin 2008), the granites of sub-tropical West Africa (Acworth 2001; Beauvais et al. 1999, 2003, 2004; Ritz et al. 1999), the lateritic weathering mantles developed from ultramafic rocks of New Caledonia (Savin et al. 2003; Robineau et al. 2007), the deep in situ and transported regolith profiles of eastern and western Australia (Calvert and Acworth 2000; Timms and Acworth 2002; Whitford et al. 2005) and Quaternary sediments in Canada (Smith and Sjogren 2006). With the deployment of electrode arrays, production rates of several kilometres/day can now be attained, making this ground technique an effective tool for mapping regolith variability and the geometry of the interface between transported regolith cover and basement (for example, Baines et al. 2002) in a range of landscapes. Magnetometric resistivity (MMR)
MMR is an electrical surveying method in which the ground is energised with commutated direct current through a pair of widely spaced electrodes and the anomalous conductivity distribution is surveyed by measuring the secondary magnetic field arising from current flow. The magnetic measurement direction is perpendicular to the line between electrodes. This technique is used to explore through a conductive surface layer. Street (1989) described its application in the study of paleochannel systems in the eastern Yilgarn Craton, Western Australia. He demonstrated
that MMR could be used to define lateral changes in regolith thickness. Sub-Audio Magnetics (SAM)
SAM was developed in the early 1990s for simultaneously mapping the magnetic and electrical characteristics of the Earth (Cattach et al. 1993), and has demonstrated potential for regolith mapping in very conductive terrains. Stolz (2005) and Stolz and Roache (2007) describe SAM as a very effective technique for mapping the base of regolith in terrains with very saline groundwater and transported cover. They report that SAM defines sub-surface bedrock structures by detecting their weathered expression at the regolith– bedrock interface (Figure 9.7, page 162), and note that the derived information complements that available from magnetic, gravity and spectral techniques. As mentioned previously, SAM involves the measurement of the magnetic field induced by a timevarying current applied to the Earth. A time-variant electrical signal is generated in the ground by the SAM transmitter using electrodes set on either side of the prospect area (Meyers and Cooper 2004). Electrical current flowing in conductive regolith and geology within the survey grid generates an electromagnetic field at right angles to the current flow, and this is measured at surface by a very sensitive, rapid-sampling magnetometer. The part of the time -varying signal relating to conductivity in the ground is called the total field magnetometric resistivity (TFMMR). Another part of the SAM signal relates to induced polarisation (IP) effects, and is called total field magnetometric induced polarisation (TFMMIP). Useful TFMMIP data cannot always be extracted from SAM field data, but efforts are underway to improve SAM TFMMIP data acquisition and processing. Total magnetic intensity (TMI) data are also recorded during SAM surveying (for example, Meyers et al. 2005). For SAM surveys, transmitter wires and current electrodes are commonly placed along the geological grain or strike, at least 500 m from the edges of the survey grid. Survey transect lines are orientated perpendicular to the strike of the transmitter electrodes. The line spacing between transects can vary from 100 m to less than 10 m, depending on survey
233
234
Regolith Science
requirements. Sample readings are spaced at about 0.5 m to 1 m, depending on the walking speed of the field crew. The primary fields from the transmitter wire and electrodes are removed from the data, and the residual TFMMR, TMI and TFMMIP data are imaged to show spatial patterns in the sub-surface that relate to the regolith and geology. Depth information cannot be directly estimated from the TFMMR data, but the data can be treated like a monopole potential field, and modelled using gravity algorithms to predict subsurface geometry of conductive sources (Meyers and Cooper 2004; Whitford et al. 2005). A SAM TFMMR response detects variations in current channelling and exclusion in the strike direction of the transmitter electrodes, and can penetrate down to 200 m in some areas (Meyers and Cooper 2004). Figure 9.8 shows an idealised cross-section of the regolith over a shear zone and a steeply dipping black shale unit. Current flowing out of the plane of the page becomes channelled into more conductive regions in the sub-surface, such as paleochannels, zones of deeper weathering, shears, conductive bedrock lithologies and mineralisation. The amplitude of the TFMMR
anomalies and their shapes reflect information about the geometry of these sub-surface conductors. Stolz (2005) notes that, in hypersaline environments, SAM is likely to be more useful for mapping regolith features relating to faults and shears compared with airborne TEM, because the linear current flow of the SAM transmitter dipole drives current channelling along the structural trends in the orientation of the transmitter dipole, resulting in clearer images of long narrow structures. Highly chargeable features, such as carbonaceous shales and sulfide mineralisation, may also produce TFMMIP anomalies (for example, Meyers et al. 2005). The reader is referred to the special issue of Exploration Geophysics (Vol. 36 (2), 2005) for a compilation of articles describing the application of SAM surveys in exploring through regolith cover. Although not discussed in any detail, developments in electrical geophysics concerning the use of distributed arrays and digital acquisition systems, are worthy of mention. The collection of networked, 3D electrical geophysical data (resistivity, IP, electromagnetic (EM) and magnetotelluric (MT)) with systems such as MIMDAS (for example, Garner and Webb 2000; Webb and James 2001; Rutley et al. 2001) have
SAM MMR response
SAM IP response Land surface Transported regolith Saprolite V
V
V
V V V
V
V
V
V
Mineralised shear V
V
Black shale
V
Fresh rock
Current channelling
V V
V V
V
V
V
V
Figure 9.8: Schematic section through the regolith and the expected SAM responses caused by current channelling in more conductive regions of the regolith profile (after Meyers and Cooper 2004).
Regolith geophysics
the potential to provide detailed resolution in the near surface as well as obtaining good data from depth in areas of regolith cover. Electromagnetic (EM)
EM surveying techniques involve the measurement of the varying response of the ground due to the propagation of electromagnetic fields. Primary fields are generated by passing a current through a loop or coil positioned on the ground, or in the air. A secondary field is induced in the ground, and these fields are detected by the alternating currents that are induced to flow in a receiver coil: a process known as electromagnetic induction. The primary field travels from the transmitter to the receiver via paths above and below the ground surface. In the presence of a conducting body (such as a conductive saprolite), the magnetic component of the electromagnetic field penetrating the ground induces eddy or alternating currents to flow in the conductor. These eddy currents generate the secondary electromagnetic field, which is measured by the receiver (Peters 2001). The difference between the transmitted (primary) and received (secondary) electromagnetic fields will be determined by the geometry and electrical properties of conductors in the ground. Materials that are highly conductive produce strong secondary electromagnetic fields. Regolith materials, particularly porous saprolite that contains saline pore water, generate such fields. As the induction of current flow results from the magnetic component of the electromagnetic field, there is no need to have physical contact between transmitter or receiver and the ground. Consequently, EM surveys can proceed effectively on the ground or from the air. Electromagnetic measurements of the earth are commonly made using ground-based instruments, and there are numerous ways EM (TEM) transmitters and receivers can be arranged (Peters 1996, 2001). Increasingly time domain EM instrumentation is becoming available for shallow (upper 20 m) investigations that have been made possible by advances in switching and digital acquisition technology, which enable very fast turn-off transmitters and high-speed sampling of the early part of the receiver waveform. Mayes (1992), Barrett et al. (2002) and Hatch et al.
(2002) describe the application of this technology in near-surface investigations of the regolith. Hatch et al. (2006, 2007) describe the use of Zonge Engineering’s NanoTEM system, which increases the resolution of ‘standard’ TEM by using smaller loops and measuring the decay very quickly after the transmitter turn-off and then collecting decay data at a very fast sampling rate. When configured on a sledge, using a three-turn, 3 × 3 m transmitting antenna, and a three-turn 1 × 1 m receiving antenna, the system can be towed behind a slow moving vehicle and data collected at 3–4 second intervals. Although having a limited depth penetration (~10 m) the system allows data collection that is at least 40 times more rapid than the conventional land-based TEM data acquisition systems. As mentioned above, EM systems can also be mounted below helicopters or specially adapted fixed-wing aircraft (Lane 2002) and, as with ground systems, an airborne electromagnetic (AEM) system detects the decay of secondary magnetic fields associated with currents induced in conductive bodies by a pulsed transmitter primary field. AEM systems are capable of rapid systematic coverage of large areas at a relatively low cost, without causing ground disturbance. However, there are usually some trade-offs in spatial resolution, near-surface vertical resolution, and depth of penetration against carefully acquired ground-based data. Nonetheless, AEM systems have seen extensive application in mapping regolith variability, and the more powerful systems are in common use for mineral exploration through conductive regolith cover. Recent developments in AEM technologies that are relevant to those considering their use in regolith studies have been eloquently summarised by Thompson et al. (2007) and Macnae (2007). A variety of AEM systems are currently available for commercial operation, including: 1. Time domain (TD) fixed wing AEM systems, such as TEMPEST (Lane et al. 2000) and GEOTEM (Smith et al. 1996) 2. TD helicopter systems, such as HOISTEM (Boyd 2001); SKYTEM (Sørensen and Auken 2004); VTEM (Witherly et al. 2004); GEOTEM (Sattel 2006) and REPTEM (Lines 2007)
235
236
Regolith Science
3. Frequency domain (FD) helicopter systems, such as DIGHEM (Huang and Fraser 2001); RESOLVE and Hummingbird (Valleau 2000). Helicopter EM (HEM) systems have EM transmitters and receivers suspended below the helicopter. Frequency domain systems, such as RESOLVE, work on the principle of transmitting a relatively low-power, sinusoidally varying current of a fixed frequency, which induces currents in the ground and measures the combined primary and secondary responses with a receiver. Usually frequency domain systems have paired coil sets, which transmit and receive a fixed frequency. These coils may be oriented horizontally or vertically. Some systems have several sets which transmit different frequencies and measure different responses. Time domain HEM systems, such as HOISTEM, VTEM and SKYTEM, transmit a square or half-sine wave pulse from a transmitter loop slung around a frame. They record the response of a decaying signal in the ground at various times after the transmitter pulse has been switched off in receiver coils or loops located within or adjacent to, but offset from, the transmitter loop. Fixed-wing systems, such as TEMPEST and GEOTEM, operate in the time domain and transmit a square or half-sine wave pulse from a transmitter loop slung around the aircraft, and measure the response of an induced, decaying signal in the ground at various times after the transmitter pulse has ceased. These measurements are made in a multicoil receiver, which is housed in a ‘bird’ towed about one hundred metres behind and below the aircraft. The spatial resolution of an AEM system varies with the system type; with sample time or frequency, and with ground conductivity (Spies and Woodgate 2004). It is different in the horizontal and vertical direction. Commonly, resolution is considered in terms of the volume of the ground that contributes most of the response for each measurement. Systems such as TEMPEST return a weighted-average response over lateral distances of several hundred metres. With helicopter systems, this decreases to the order of tens of metres. The smallest lateral features that can be resolved near surface are around 20–40 m for helicopter systems and 50–100 m for
fixed-wing systems, where good conductivity contrasts are present. These figures increase with depth. Normally, the highest resolution is measured along a flight line, with the perpendicular resolution being determined by line spacing. Recent developments in AEM system technologies (Thompson et al. 2007; Macnae 2007) – with the better definition of system geometry and the greater attention to calibration – have contributed to substantial improvements in the definition of conductivity at shallow depths. These trends, coupled with our ability to better process, image and display AEM data, have made these systems a more relevant geophysical technology for the systematic mapping of weathered environments. The processing of AEM data and the presentation of conductivity data as maps or sections is now commonplace. This is particularly effective in regolith and flat to shallow-dipping sedimentary environments (Lane 2000, 2002; Macnae et al. 1998; Macnae and Bishop 2001; Munday et al. 2001; Sattel 1998) where the assumption that each observation can be treated in isolation and that the sub-surface is represented as a series of horizontal layers holds reasonably well, at the scale of the footprint of most AEM systems. Through the application of approximate transforms or layered inversions, conductivity–depth values can be calculated for each observation and then stitched together into sections to provide a representation of the 2D variation of conductivity. This is sometimes referred to as a ‘parasection’ and an example of the resulting product is shown in Figure 9.9 (page 163) (Lane 2002; Worrall et al. 2001). Furthermore, the conductivity depth profiles can be combined into a 3D gridded volume from which arbitrary sections, horizontal depth slices (or interval conductivity images) and isosurfaces can be derived (Lane and Pracilio 2000; Lane et al. 2004). The schematic in Figure 9.10 (page 164) summarises the process of acquiring AEM data, inverting the resulting data, and presenting the results as conductivity images. Figure 9.11 (page 165) shows a set of interval conductivity images depicting spatial changes in conductivity with depth for a regolith setting. The representation of essentially continuous and gradational conductivity distributions as discrete conductive
Regolith geophysics
‘units’ or bounding layers is an effective way summarising information from large AEM surveys. This is particularly so when the application calls for mapping the conductivity, depth to top and thickness of a semicontinuous layer of transported or in situ regolith. Several (semi-) automated schemes exist for picking the layer boundaries, such as a layered or blocky inversion (Sattel 1998). The use of conductive unit parameters (Lane 2000) is illustrated in Lawrie et al. (2000), Worrall et al. (2001), Worrall and Gray (2004) and Munday et al. (1998, 2006) who describe examples comparing calculated layer boundaries in the regolith with boundaries interpreted from borehole data. Ground-Penetrating Radar (GPR)
GPR provides a means of exploring the shallow subsurface with electromagnetic waves (radar), which are usually in the 10 to 1000 MHz band. In principle, it is similar to reflection seismic or sonar techniques. The radar antennae, which are generally in contact with the ground for the strongest signal strength, produce a short pulse of electromagnetic energy that is transmitted into the ground (Davis and Annan 1989; Daniels 2004). The propagation of the pulsed energy depends on the high-frequency electrical properties of the ground, and these are primarily controlled by the water content. When the propagating wave hits a buried boundary with different dielectric constants, part of the transmitted signal is reflected. The receiving antenna records variations in these reflected returns. The two-way travel times of reflected radar waves give the depths where changes in electrical properties occur. A ground-penetrating radar record or image appears similar to a very shallow-penetration seismic section. The depth range of GPR is limited by the electrical conductivity of the ground, and the transmitting frequency. As conductivity increases, the penetration depth also decreases. In moist clay-rich materials with a high soluble salt content (that is, high conductivity) penetration may be limited to a few centimetres. Good penetration (to depths of 10–25 m) may be obtained in dry sandy soils. Higher frequencies do not penetrate as far as lower frequencies, but they do provide a better resolution. Ground-penetrating radar (GPR) has been effectively used to explore alluvial-fan complexes and alluvial systems in Europe and North
America (for example, Davis et al. 1985), but has seen limited application in the more conductive regolith settings such as found in Australia. Nonetheless GPR has seen application in some Australian (for example, Dupuis et al. 2007; Davey et al. 2003; Campbell 2007) and West African regolith settings (Beauvais et al. 2003). Useful information pertaining to GPR methods is found in Olhoeft (2004). Seismic
Seismic reflection and refraction surveys are used extensively as a means of investigating the subsurface. They have considerable potential for defining variability within the regolith – particularly discontinuities and bounding surfaces that may reflect changes in material type or the depth of the weathered zone. Seismic energy, which is generated as shock waves produced by mechanical or explosive means, travels outwards in all directions through the layers below the surface. The time taken for the energy to travel from the source to a series of geophones oriented in a line or profile towards the source is then measured (Telford et al. 1990; Drummond 2002). Knowledge of the time taken and the velocity of the propagating waves are used to determine the path of the energy, which may be reflected or refracted. Travel times for both types of path depend on the physical properties of the sub-surface materials and their geometry. Two-way time can be converted to depths using estimates of seismic velocities generated during the data-processing stage. With regolith investigations, the objective in the application of seismic methods is to use the observed arrival times to deduce information about the attitude of bounding surfaces, such as that between transported and in situ regolith, or the geometry of the weathering front. Seismic reflection surveys are used where the primary information required is the geometry of interfaces in the regolith and the underlying basement (Drummond 2002). Refraction surveys are used when variations in the composition of the regolith or the underlying basement are the targets, and not the detailed geometry of interfaces. In seismic reflection surveys, the information is usually displayed as continuous reflection-time cross-sections, which essentially describe the geometry of discontinuities at depth.
237
Regolith Science
0
Wyalong Line
Interpreted section Transported cover
0
“Hill”
100 3316 m/s
Transported cover
Saprolite ?
?
200
02
+
+
+
300
40 Depth (m)
Two-wat time (ms)
15 Wedge
+ 200
Bedrock
+ 400
+ 600
+
+ +
CDP 800
100
+ +
+
200
Stacked section 1000
1200
1400
1600
Two-wat time (ms)
238
100
200
300 TMf027-07
Figure 9.12: A continuous reflection-time cross-section, showing regolith structure at depth for a line in NSW (see Figure 9.5 for location). It shows discontinuities in a transported cover overlying in situ saprolite. The time for the seismic energy to travel to the reflectors and back (two-way time) is shown as a linear scale on the left, while depths are shown on the right (after Leslie et al. 2000).
Examples of shallow seismic investigations applied to regolith investigations are numerous, and the work of Cooper (1994), Leslie et al. (2000), Jones et al. (2000), Deen et al. (2000), Miller and Xia (1997) and BurVal Working Group (2006) describes some examples. Figure 9.12 illustrates the type of cross-sectional information that can be obtained in shallow seismic surveys: in this instance, showing shallow responses from discontinuities in transported cover overlying saprolite. Radiometrics
Radiometrics is also known as gamma-ray spectrometry. A radiometric survey measures the spatial distribution of three radioactive elements (K, Th and U) in the top 30–45 cm of the Earth’s crust. The concentrations of K, Th and U are measured by detecting the gamma rays produced during the natural radioactive decay of these elements and the daughter products in their decay chains. Gamma rays can be measured on
the ground or from a low flying aircraft. They are detected by interaction with a sodium iodide detector, where the gamma rays lose some or all of their energy. This creates light pulses within the detector, which are detected by photomultiplier tubes and converted to voltage pulses where the voltage is proportional to the energy lost by the individual gamma rays. Typically measurements are made in 256 12 keV energy channels in the range 0–3 MeV and the number of events are recorded each second in each energy channel. Each of the decay series has distinctive spectral peaks and shapes and the data are converted to ground concentrations of K, Th and U in the subsequent dataprocessing phase that follows acquisition. During weathering, radioelements within rocks are released and redistributed. Processes involved in the development of regolith materials exert a significant influence on their distribution and concentration. Commonly, the radioelement characteristics of rego-
Regolith geophysics
lith materials differ markedly from their parent materials. In general terms, intensely weathered regolith materials typically show a depletion of K resulting from leaching, and elevated U and Th values associated with clays and/or Fe oxides. The presence of calcrete and ferricrete is important – as both have their own radiometric signatures. Ferruginous materials tend to accumulate Th and, to a lesser extent, U – with K generally being severely depleted. Other regolith processes that may influence the surface concentration of radioelements include clay eluviation, colluvial and aeolian transport and soil movement (Dickson and Scott 1989, 1997; Wilford et al. 1997; Cook et al. 1996; Bierwirth 1996; Bierwirth et al. 1997). The observed response from a gamma ray survey in a weathered landscape is shown in Figure 9.4a (page 161), and a resulting product from their combined interpretation with airborne magnetics is presented in Figure 9.5. Wilford (1992, 1995, 2004) and Dickson et al. (1996) reported that radiometric images can be used to separate areas of high geomorphic activity with shallow regolith from stable surfaces that are less geomorphically active and that have deeper and more highly weathered regolith (Figure 9.13, page 166). Gamma-ray spectrometry surveys have been widely used in soil and regolith mapping as has been described by Dauth (1997), Taylor et al. (2002) and Praciloio et al. (1998). The ability of gamma rays to pass through vegetation to a detector in an aircraft is an advantage in agricultural regions, where cropping can mask the underlying soils, making it difficult to interpret data from other remote sensing systems such as Landsat TM (Bierwirth et al. 1997). Borehole geophysics
Borehole logging, using a variety of geophysical tools, offers the opportunity for determining the composition, variability and physical properties of the materials around the borehole. In regolith studies, geophysical-well logging can provide unique insights into the composition, structure and variability of the sub-surface, and is also widely used for validating the results obtained from airborne geophysical data acquisition, such as AEM (for example, Lane et al. 2004). For additional information on the specifics of borehole geophysical logging, the reader is referred to
Telford et al. (1990), with more recent reviews given by Killeen (1997), Chopra et al. (2002) and McMonnies and Gerrie (2007). 9.4.2 Emerging technologies There are several geophysical technologies that may provide useful information pertaining to particular characteristics of the regolith, but which are, at present, not in widespread use. They are ground-based techniques and include those described in the following sub-sections. Microtremor array method (MTM)
Recent developments in the area of passive seismic techniques (for example, Asten 2004, 2005; Asten et al. 2005) hold the promise of regolith-mapping applications using the microtremor array method (MTM). This works well in areas where conventional seismic methods are difficult to justify for reasons of cultural noise, environmental restraints and safety. Recent studies have demonstrated the usefulness of the microtremor array method for estimation of thickness and shear-velocity of sands overlying lowervelocity clays in 50 to 100 m of Quaternary cover in the Perth Basin, Western Australia. The measurement of regolith thickness is another potential application of the technique (Asten 2004), and it is likely that further progress in the broader application of this technology will be forthcoming in the near term. Seismoelectric methods
Mechanical wave propagation through porous media can generate minute electromagnetic signals, known as seismoelectric effects, by electrokinetic coupling mechanisms that involve the motion of charge in the electric double layer at the solid–liquid interface (Pride 1994). These can be measured electrodes at the surface (Waring et al. 2002; Thompson et al. 2007). These signals, are of interest for the information they may be able to provide on the type of pore fluid and the properties of the porous medium, such as porosity and permeability (for example, Thompson and Gist 1993; Garambois and Dietrich 2001; Beamish and Peart 1998; Waring et al. 2002; Rosid and Kepic 2003). Recent results (for example, Dupuis et al. 2007) have demonstrated that seismoelectric methods can be
239
240
Regolith Science
used to trace sub-surface interfaces in an unconfined sand aquifer in a manner analogous to multichannel seismic reflection surveying. Surface Nuclear Magnetic Resonance (SNMR)
SNMR or magnetic resonance sounding (MRS) is an active, non-invasive geophysical tool that provides information on the physical properties of water-saturated rocks and it is specifically used for groundwater investigation (Yaramanci and Hertrich 2006). The measured signal is directly related to the volume of groundwater, and its relaxation time is correlated with the size of the water-filled pores (Bernard and Legchenko 2003; Lubczynski and Roy 2004; Legchenko et al. 2004; Kenyon 1997). Inversion of MRS data can reveal the vertical distribution of the water content in regolith materials (for example, Dippell et al. 2003). Porosity and hydraulic conductivity of the rocks may also be inferred. The maximum depth of investigation with MRS is about 100 m. Ground Acoustic Penetration (GAP) or Resonance Acoustic Profiling (RAP)
This technique, developed in Russia, is apparently based upon the novel interpretation of acoustic signals generated within the ground in a manner similar to seismic methods. However, unlike the seismic technique, it uses the acoustic oscillations generated between materials and registers changes in their acoustic density (impedance) (Cornelius 2003). Although the technology was trialled in several field situations in Australia, details on its specifics remain limited. Results from these field trials suggested that the technique had the potential to delineate paleochannels and recognise boundaries between transported and residual materials (Koulmametiev and Matveev 2002). In the absence of other well-documented case studies and examples – particularly against the more conventional shallow seismic methods – it remains on the periphery of tried-and -trusted geophysical technologies for regolith characterisation.
9.5 GEOPHYSICAL SURVEY PLANNING AND DESIGN FOR REGOLITH SETTINGS Critical to the successful application of geophysics in regolith investigations is the choice of suitable tech-
nologies given a particular environment and target. Having a clear objective at the outset can save money and disappointment once the data have been collected and interpreted. Some geophysical methods are clearly unsuited to some settings: for example, GPR would be wholly unsuited to defining the saprolite–fresh rock interface where the regolith is thick, conductive and partially or wholly saturated, whereas it might be very effective at defining significant regolith material boundaries in dry arid settings (for example, Mills and Speece 1997). Poor planning, the incorrect choice of technique and an inappropriate specification – coupled with poor operation, lack of calibration and an inadequate sampling strategy – can all contribute to poor results and a disappointing outcome. In planning a geophysical survey, consideration should be given of the geology underlying the regolith, geological structure and regolith materials likely to be present. Some appreciation of the size and depth of the target should also form part of the planning processes, as these will not only determine the best technique to employ but also the sampling interval or orientation to use. For example, the electrodes used in the sub-audio magnetic (SAM) method should be placed along the strike of the structure or lithologies under investigation, to enhance their relationship with the overlying regolith and for the detection of significant geological features beneath it (for example, Jackson 2005; Stolz 2005). The consequences of varying the orientation of the electrodes and survey lines in a SAM survey in a regolith setting have been illustrated by Meyers et al. (2005), who showed that an equivalent MMR (EQMMR) response changed significantly when transmitter electrodes were placed in different orientations over the same site. In another example, ground magnetic data in areas of transported cover can be dominated by high-frequency high-amplitude responses, which are attributed to the widespread distribution of ferruginous maghemite pisoliths (which have an extremely high magnetic susceptibility) in the colluvium (Butt et al. 2001). Low sensor heights with ground surveys and, in these circumstances, low-level, high-resolution airborne magnetic surveys may be more appropriate to resolve accumulations of maghemite gravels and/or the presence of ferricrete (for example, Wildman and Compston 2000). Figure 9.14 (page 166), illustrates the
Regolith geophysics
consequence of changing line spacing and resolution in an airborne magnetic survey over a regolith setting – emphasising the importance of sampling spacing and survey height in resolving variability in the regolith. Available resources (funding and people) should also be a consideration. For further discussion of some of these issues – particularly as they relate to the use of airborne geophysical surveys in mapping features in the regolith – refer to Mackey et al. (2000) and Lawrie et al. (2003). It should always be appreciated that the application of geophysics may not necessarily produce a unique, unambiguous answer (Telford et al. 1990). The basis of geophysical interpretation requires the deduction of sub-surface structure and physical character through a process of inversion and/or computer modelling. However, geophysical interpretation suffers from inherent ambiguity and non-uniqueness because different geological configurations can produce similar observed responses. This can result in several possible answers for a given data set. An example of this is illustrated in Figure 9.15 (page 167) which shows the conductivity response with depth for the ground for an E–W section through a weathered landscape in south-western Australia. The conductivity depth sections shown were generated from the transformation and inversion of TEMPEST airborne electromagnetic data using different algorithms (see Munday et al. 2006, for details). The regolith geology is shown in the bottom section. All the derived models fit the field data and, as such, are geophysically ‘correct’, but the resulting conductivity depth sections, although quantitatively similar, show some marked differences. The predictions of conductivity are the result of a chain of processes – each with its own effect and assumptions (Lane 2002). Applying different conductivity transformations or inversions to the same measurements commonly yields quantitatively similar, though subtly different, answers. Using different measurement systems – even if based upon the same principles – may also yield a different answer. While this is to be expected, it should be appreciated by users of geophysical technologies. It is interesting to note that the apparent success of the approximate methods such as EMFlow in defining the regolith conductivity structure can be attributed to the regolith behaving essentially as a 1D body to an
AEM system, and that errors arising from that assumption coupled with calibration and altitude/ geometrical uncertainties, are larger than the approximations made to speed up the conversion of data to conductivity and depth (Macnae 2007). The previous example also illustrates that a key aspect to any survey is the consideration of subsequent processing and interpretation phase. Delivery of useful products that can be reliably interpreted for regolith variability requires skilled analysis and interpretation – with an understanding of system noise, resolution and survey parameters. This knowledge has to then be linked to an appreciation of transformation and inversion procedures, and assumptions made in their application. Advice from suitably qualified professionals should always be sought when considering what technology to use and when, and what processing and interpretation strategies might be effectively employed.
9.6 DEVELOPMENTS AND IMPLICATIONS The past few decades have witnessed significant advances in the acquisition, processing and interpretation of geophysical data. There is now a greater appreciation of geophysical technologies that are relevant to the study of regolith variability. In part, this acceptance comes from our ability to sample more quickly and at a higher resolution, making it much easier to map variations relating to regolith materials in the near surface. It has also been helped by the evolving speed of computers and better data storage systems, coupled with improvements in electronics and noise reduction. This has resulted in geophysical systems that are more compact, portable, easier to use and yield higher quality data relatively quickly. Our ability to rapidly and effectively map in the near surface is reflected most particularly with airborne systems, where we can acquire airborne magnetic and radiometric data very close to the surface in fixed wing ‘crop dusters’ and helicopter platforms. In the same way, we are now able to measure the electromagnetic and gravity response at high resolution with developments in helicopter systems and airship platforms. New developments in drones or unmanned aerospace vehicles (UAVs) are likely to provide a
241
242
Regolith Science
different dimension to data acquisition, with significant potential to lower the cost of surveys in the short to medium term (Thompson et al. 2007). Recently, the possibility of acquiring airborne gravity gradiometer, magnetics and EM data simultaneously from a helicopter platform has been demonstrated (Rajagopalan et al. 2007). Our ability to undertake highresolution, integrated geophysical surveys, for near-surface investigations of regolith terrains is now a reality. However, key to the future application of these airborne systems in the characterisation of regolith variability is the care given to the attitude or geometry of airborne platforms, noise reduction and the acquisition of good terrain models for data correction. This will ensure good lateral and vertical resolution in the near surface. Major advances have been made in enhancing the signal-to-noise ratios in both ground and airborne data acquisition systems, with data streaming and full waveform recording. The use of multiple receivers is also becoming common for some ground techniques – permitting faster ground coverage and enhanced mapping of sub-surface variability in 2 and 3D. Hardware developments have been accompanied by significant advances in our ability to filter, model, transform and invert data. Similarly, we now have the ability to image and display derived data in conjunction with other data sets – thereby improving interpretability. Overall, we have never been better placed to see geophysical techniques deployed in the spatial characterisation of regolith landscapes, but the considered choice of system and interpretation procedure(s) to match the application and environment will remain crucial – to avoid disappointing outcomes.
9.7 REFERENCES Acworth RI (2001). The electrical image method compared with resistivity sounding and electromagnetic profiling for investigation in areas of complex geology – a case study from groundwater investigation in a weathered crystalline rock environment. Exploration Geophysics 32, 119–128. Asten MW (2004). Passive seismic methods using the microtremor wave field for engineering and earthquake site zonation. In Proceedings of the SEG 74th
Annual Meeting. Denver. Session NSG-1. (Extended Abstract on CD). Society of Exploration Geophysicists, Tulsa, Oklahoma. Asten MW (2005). Estimating the shear velocity profile of Quarternary silts using microtremor array (SPAC) measurements. Exploration Geophysics 36, 34–40. Asten MW, Stephenson BR and Davenport P (2005). Shear-wave velocity profile for Holocene sediments measured from Microtremor Array Studies, SCPT, and Seismic Refraction. Journal of Environmental and Engineering Geophysics 10, 235–242. Baines D, Smith DG, Froese DG and Bauman P (2002). Electrical resistivity ground imaging (ERGI): a new tool for mapping the lithology and geometry of channel-belts and valley-fills: Sedimentology 49, 441–449. Barrett B, Heinson G, Hatch M and Telfer A (2002). Geophysical methods in saline groundwater studies: locating perched water tables and fresh-water lenses. Exploration Geophysics 33, 115–121. Beamish D and Peart RJ (1998). Electrokinetic geophysics – a review: Terra Nova 10, 48–55. Beauvais A, Ritz M, Parisot J-C, Dukhan M and Bantisimba C (1999). Analysis of poorly stratified lateritic terrains overlying a granitic bedrock in West Africa, using a 2-D electrical resistivity tomography. Earth and Planetary Science Letters 173, 413–424. Beauvais A, Ritz M, Parisot J-C, and Bantisimba C (2003). Testing etching hypothesis for the shaping of granite dome structures beneath lateritic weathering landsurfaces using ERT methods. Earth Surface Processes and Landforms 28, 1071–1080. Beauvais A, Ritz M, Parisot J-C, Bantisimba C and Dukhan M (2004). Combined ERT and GPR methods for investigating two stepped lateritic weathering systems. Geoderma 119, 121–132. Bell B, Fullagar PK, Paine J, Whitaker A and Worrall L (2001). Exploring through cover – the integrated interpretation of high resolution aeromagnetic, airborne electromagnetic and ground gravity data from the Grant’s Patch area, Eastern Goldfields Province, Archaean Yilgarn Craton. Part B: Gravity inversion as a bedrock and regolith mapping tool: Exploration Geophysics 32, 194–197.
Regolith geophysics
Berkman DA (1995). Gravity survey methods and tables. In Field geologists manual. 3rd edn (Ed. DA Berkman) p. 324. Monograph 9. The Australasian Institute of Mining and Metallurgy, Melbourne. Bernard J and Legchenko A (2003). Groundwater exploration with the magnetic resonance sounding method. In Proceedings of the 16th ASEG Geophysical Conference and Exhibition. February, Adelaide. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Bierwirth P (1996). Investigation of airborne gammaray images as a rapid mapping tool for soil and land degradation – Wagga Wagga, NSW. AGSO Record 1996/22. Bierwirth PN, Hardy S, Wilson P, Philip S, Smith D, Heiner I and Grundy M (1997). Radio Ga Ga – Integrating gamma-radiometrics into landscape modelling of soil attributes: results of an ACLEP exchange: Australian Collaborative Land Evaluation Program Newsletter 6, 15–18. Bishop J, Sattel D, Macnae J and Munday TJ (2001). Electrical structure of the regolith in the Lawlers District, Western Australia. Exploration Geophysics 32, 20–28. Boyd G (2001). Normandy heli-borne time domain EM system. In Proceedings of the 15th ASEG Geophysical Conference and Exhibition. August, Brisbane. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Braine MF and Macnae J (1999). Removing the regolith: EM defined structure fails to correct gravity at Elura. Exploration Geophysics 30, 115–122. Brand NW (1997). Chemical and mineralogical characteristics of weathered komatiitic rocks, Yilgarn Craton, Western Australia: Discrimination of nickel sulphide bearing and barren komatiites. PhD thesis. University of Western Australia, Perth. Brodie RC (2002). ‘Airborne and ground magnetics.’ Open File Report 144. CRC LEME, Perth. BurVal Working Group (2006). Groundwater Resources in Buried Valleys - A Challenge for Geosciences. Leibniz Institute for Applied Geosciences, Hannover. < http://www.burval.org/> Butt CRM (1981). The nature and origin of the lateritic weathering mantle, with particular reference to Western Australia. In Geophysical Exploration in
Deeply Weathered Terrains. (Eds HA Doyle, JE Glover and DI Groves) pp. 11–29. Publication 6. Geology Department and University Extension, University of Western Australia, Perth. Butt CRM, Gray DJ, Robertson IDM, Lintern MJ, Anand RR, Britt AF, Bristow APJ, Munday TJ, Phang C, Smith RE and Wildman JE (2001). ‘Geochemical exploration in areas of transported overburden, Yilgarn Craton and environs, Western Australia’. Open File Report 86. CRC LEME, Perth. Calvert HT and Acworth RI (2000). Electrical and electromagnetic surveys of dryland salinity, Yass, NSW, Australia. In Proceedings of the 2000 SEG/ Calgary International Exposition and 70th Annual Meeting of the SEG. Calgary, Canada. (Extended Abstract on CD). Society of Exploration Geophysicists, Tulsa, Oklahoma. Campbell T (2007). Geophysical characterization of variable limestone terrains. In Proceedings of the 19th ASEG International Geophysical Conference and Exhibition. November, Perth. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Cattach MK, Stanley JM, Lee SJ and Goyd GW (1993). Sub-audio magnetics (SAM) – A high resolution geophysical method for simultaneously mapping electrical and magnetic properties. Exploration Geophysics 24, 387–400. Chopra P, Papp É and Gibson D (2002). ‘Geophysical well logging.’ Open File Report 144. CRC LEME, Perth. Clark DA and Emerson DW (1991). Notes on magnetisation characteristics in applied geophysical studies. Exploration Geophysics 22, 547–555. Cook SE, Corner RJ, Groves PR and Grealish GJ (1996). Use of gamma radiometric data for soil mapping. Australian Journal of Soil Research 34, 183–194. Cooper M (1994). High resolution geophysical investigations of the Roe paleodrainage system, Kalgoorlie. BSc (Hons) thesis. Curtin University of Technology, Perth. Cornelius M (Ed.) (2003). ‘Resonance Acoustic Profiling trials in Australia’. CRC LEME Open File Report 146. CRC LEME, Perth. Daniels JJ (Ed.) (2004). Ground Penetrating Radar. 2nd edn. The Institute of Electrical Engineers, London.
243
244
Regolith Science
Dauth C, 1997. Airborne magnetic, radiometric and satellite imagery for regolith mapping in the Yilgarn Craton of Western Australia. Exploration Geophysics 28, 199–203. Davey A, Joseph J and Heinson G (2003). 2D Regolith mapping with ground penetrating radar and time domain EM at a recently discovered Cu-Au orebody: White Dam, Curnamona Craton, SA. In Advances in Regolith. (Ed. IC Roach) pp. 86–90. CRC LEME, Perth. Davis JL and Annan AP (1989). Ground-penetrating radar for high resolution mapping of soil and rock stratigraphy. Geophysical Prospecting 37, 531–551. Davis JL, Annan AP and Vaughan CJ (1985). Placer exploration using radar and seismic methods. Canadian Institute of Mining Bulletin 80, 67–72. Deen T J, Gohl K, Leslie C, Papp É and Wake-Dyster K (2000). Seismic refraction inversion over a palaeochannel in the Lachlan fold belt, central NSW. Exploration Geophysics 31, 389–393. Dentith MC, Evans BJ, Paish KF and Trench A (1992). Mapping the regolith using seismic refraction and magnetic data: results from the Southern Cross greenstone belt, Western Australia. Exploration Geophysics 23, 97–104. Dentith MC, Frankcombe KF, Ho SE, Shepherd JM, Groves DI and Trench A (Eds) (1994). Geophysical Signatures of Western Australian Mineral Deposits. Geology and Geophysics Department (Key Centre) and UWA Extension Publication 26, Publication 7. Australian Society of Exploration Geophysicists, Perth. Dickson BL and Scott KM (1989). ‘Radioelements in weathered shales and mafic volcanics, Panglo gold deposit, Eastern Goldfields, WA’. Restricted Report 41R. CSIRO Division of Exploration Geoscience, Perth. Dickson BL and Scott KM (1997). Interpretation of aerial gamma-ray surveys – adding the geochemical factors. AGSO Journal of Australian Geology and Geophysics 17, 187–200. Dickson BL, Fraser SJ and Kinsey-Henderson A (1996). Interpreting aerial gamma-ray surveys utilising geomorphological and weathering models. Journal of Geochemical Exploration 57, 75–88. Dippell S, Golden H and Jenke G (2003). Nuclear magnetic resonance: applications for groundwater
exploration in Queensland. In Proceedings of the 16th ASEG Geophysical Conference and Exhibition. February, Adelaide. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Doyle HA and Lindeman FW (1985). The effect of deep weathering on geophysical exploration in Australia – a review. Australian Journal of Earth Sciences 32, 125–135. Doyle HA, Glover JE and Groves DI (1981). Geophysical Exploration in Deeply Weathered Terrains. Geology Department and University Extension, The University of Western Australia, Publication 6. Perth. Dransfield M (2007). Airborne gravity gradiometry in the search for mineral deposits In Proceedings of Exploration 07: Fifth Decennial International Conference on Mineral Exploration. 9–12 September, Toronto, Canada. (Ed. B Milkereit) pp. 341–354. Prospectors and Developers Association of Canada, Toronto, Canada. Drummond B (2002). ‘Seismic surveys for imaging the regolith.’ Open File Report 144. CRC LEME, Perth. Dupuis JC, Butler KE and Kepic AW (2007). Seismoelectric imaging of the vadose zone of a sand aquifer. Geophysics 72, A81–A85. Emerson DW (1990). Notes on mass properties of rocks- density, porosity and permeability. Exploration Geophysics 21, 209–216. Emerson DW and Macnae J (2001). Further physical property data from the Achaean Regolith, Western Australia. Preview 92, 33–37. Emerson DW and Yang YP (1997). Effects of water salinity and saturation on the electrical resistivity of clays. Preview 19–24. Emerson DW, Macnae JC and Sattel D (2000). The petrophysics of the Lawlers regolith, Western Australia. Exploration Geophysics 31, 229–235. Fitterman DV and Deszcz-Pan M (2001). Using airborne and ground electromagnetic data to map hydrologic features in Everglades National Park. In Proceedings of the SAGEEP 2001. (Paper on CD). Environmental and Engineering Geophysical Society, Denver, Colorado. Garambois S and Dietrich M (2001) Seismoelectric wave conversions in porous media: field measure-
Regolith geophysics
ments and transfer function analysis. Geophysics 66, 1417–1430. Garner S and Webb D (2000). Broadband MT and IP electrical property mapping with MIMDAS. In SEG 70th Annual Meeting. (Extended Abstract on CD). Society of Exploration Geophysicists, Tulsa, Oklahoma. Gibson DL and Wilford J (2002). Aspects of the regolith and landscape of the Strathbogie-CanianboDookie area: the need for detailed interpretation of detailed geophysical data sets in the light of regional data and models. In Victoria Undercover: Benalla 2002 Conference Proceedings and Field Guide. (Eds GN Phillips and KS Ely) pp. 235–248. CSIRO Publishing, Melbourne. Griffith DH and Barker RD (1993). Two-dimensional resistivity imaging and modeling in areas of complex geology. Journal of Applied Geophysics 29, 211–226. Gunn PJ, Maidment D and Milligan PR (1997). Interpreting magnetic data in areas of limited outcrop. AGSO Journal of Geology and Geophysics 17, 175–185. Hatch M, Barrett B, Bennetts D, Heinson G, Telfer A and Roberts C (2002). Improved near surface mapping in groundwater studies: Application of fastsampling time-domain EM surveying methods: Preview 96, 25–29. Hatch M, Telfer A, Berens V, Heinson G and Munday TJ (2006). Fast-sampling surface EM in the Riverland of South Australia. In Proceedings of the Australian Earth Sciences Convention 2006, 2–6 July, Melbourne. (Extended Abstract on CD). Geological Society of Australia, Sydney. Hatch M, Berens V, Fitzpatrick A, Heinson G, Munday TJ and Telfer A (2007). Fast-sampling EM applied to the River Murray and surrounding floodplains in Australia. In 13th European Meeting of Environmental and Engineering Geophysics, 3–5 September 2007, Istanbul, Turkey. (Extended Abstract on CD). European Association of Geoscientists and Engineers, Houten, Netherlands Huang H and Fraser DC (2001). Mapping of the resistivity, susceptibility and permittivity of the earth using a helicopter-borne electromagnetic system. Geophysics 66, 148–157. Jackson J (2005). The use of sub-audio magnetics (SAM) in gold exploration—Examples from the
Yilgarn Craton, WA. Exploration Geophysics 36, 163–169. Jones LC, Papp E, Wake-Dyster K, Deen TJ and Gohl K (2000). High-resolution seismic imagery of palaeochannels near West Wyalong, NSW. Exploration Geophysics 31, 383–388. Kenyon WE (1997). Petrophysical principles of applications of NMR logging, The Log Analyst March– April, 21–43. Kew GA and Gilkes RJ (2007). Properties of regolith beneath lateritic bauxite in the Darling Range of south Western Australia. Australian Journal of Soil Research 45, 164–181. Killeen P (1997). Borehole geophysics: Exploring the third dimension. In Proceedings of Exploration 97: Fourth Decennial International Conference on Mineral Exploration. (Ed. AG Gubins) pp. 31–42. Prospectors and Developers Association of Canada, Toronto, Canada. Koulmametiev BA and Matveev BV (2002). ‘Ground acoustic penetration’. Open File Report 144.CRC LEME, Perth. Lane R (2000). Conductive Unit Parameters: Summarising Complex Conductivity Distributions. 70th Meeting, SEG, Calgary, Volume 1, Section EM 4.2, 328–331. Society of Exploration Geophysicists, Tulsa, Oklahoma. Lane R (2002). ‘Ground and airborne electromagnetic methods.’ Open File Report 144. CRC LEME, Perth. Lane R (Ed.) (2004). Airborne gravity 2004. In Abstracts from the ASEG-PESA Airborne Gravity 2004 Workshop. Geoscience Australia Record 2004/18. Lane R and Pracilio G (2000). Visualisation of subsurface conductivity derived from airborne EM. In Proceedings of the Annual SAGEEP 2000. (Extended Abstract on CD). Environmental and Engineering Geophysical Society, Denver, Colorado. Lane R, Green AA, Golding C, Owers M, Pik, P, Plunkett C, Sattel D and Thorn B (2000). An example of 3D conductivity mapping using the TEMPEST airborne electromagnetic system. Exploration Geophysics 31, 162–172. Lane R, Brodie R and Fitzpatrick A (2004). ‘Constrained inversion of AEM data from the Lower Balonne area, Southern Queensland, Australia’.
245
246
Regolith Science
CRC LEME Open File Report 163. CRC LEME, Perth. Lawrie KC, Chan RA, Gibson DL and de Souza Kovacs N (1999). Alluvial gold potential in buried palaeochannels in the Wyalong district, Lachlan Fold Belt, New South Wales. AGSO Research Newsletter 30, 1–5. Lawrie KC, Munday TJ, Dent DL, Gibson DL, Brodie RC, Wilford J, Reilly NS and Chan RA (2000). A ‘Geological Systems’ approach to understanding the processes involved in land and water salinisation in areas of complex regolith – the Gilmore Project, central-west NSW. AGSO Research Newsletter , 26–32. Lawrie KC, Munday TJ, Gibson DL, Mernagh T, Wilford J, Williams NC, Brodie RC and Apps H (2002). The role of airborne electromagnetics in a multidisciplinary approach to mapping mineral systems under cover – an example from the eastern Lachlan Fold Belt, central west NSW. In Victoria Undercover: Benalla 2002 Conference Proceedings and Field Guide. (Eds GN Phillips and KS Ely) pp. 107– 120. CSIRO Publishing, Melbourne. Lawrie KC, Gray M, Fitzpatrick A, Wilkes P and Lane R (2003). Reducing the acquisition costs of airborne electromagnetic surveys for salinity and groundwater mapping. Preview 31–37. Lee JB (2001). Falcon gravity gradiometer technology. Exploration Geophysics 32, 247–250. Legchenko A, Baltassat JM, Bobachev A, Martin C, Robain H and Vouillamoz JM (2004). Magnetic resonance soundings applied to characterization of aquifers, Ground Water 42, 436–473. Leslie C, Jones L, Papp É, Wake-Dyster K, Deen TJ, Gohl K (2000). High-resolution seismic imagery of palaeochannels near West Wyalong, New South Wales. Exploration Geophysics 31, 383–388. Lines K (2007). Application of the REPTEM airborne EM system to palaeo-drainage mapping at Marree. In Program and Presentations. Mineral Exploration through Cover 2007: Into New Horizons. 22 June 2007, Adelaide. University of Adelaide, Adelaide. Lo B, Johnson P, McConnell TJ, Ryder-Turner A (1999). HeliGravTM helicopter long-line transported ground gravity versus traditional ground gravity acquisition – survey efficiency and results
from an Australian regional survey. In SEG 1999 Annual Meeting. (Extended Abstract on CD). Society of Exploration Geophysicists, Tulsa, Oklahoma. Lubczynski M and Roy J (2004). Magnetic Resonance Sounding: new method for ground water assessment, Ground Water 42, 291–303. Mackey T, Lawrie K, Wilkes P, Munday TJ, de Souza Kovacs N, Chan R, Gibson D, Chartres C and Evans R (2000). Paleochannels near West Wyalong, New South Wales: A case study in delineation and modelling using aeromagnetics. Exploration Geophysics 31, 1–7. Macnae JC (2007). Developments in broadband airborne electromagnetics in the past decade In Proceedings of Exploration 07: Fifth Decennial International Conference on Mineral Exploration. 9–12 September, Toronto, Canada. (Ed. B Milkereit) pp. 387–398. Prospectors and Developers Association of Canada, Toronto, Canada. Macnae JC and Bishop J (2001). Simplified electrical structure models at AEM scales, Lawlers, Western Australia. Exploration Geophysics 32, 29–35. Macnae JC, King A, Stolz N, Osmakoff A and Blaha A (1998). Fast AEM data processing and inversion. Exploration Geophysics 29, 163–169. McCrea AF, Anand RR and Gilkes RJ (1990). Mineralogical and physical properties of lateritic pallid zone materials developed from granite and dolerite. Geoderma , 33–57. McMonnies B and Gerrie V (2007). Ground geophysics and borehole logging – a decade of improvements. In Proceedings of Exploration 07: Fifth Decennial International Conference on Mineral Exploration. 9–12 September, Toronto, Canada. (Ed. B Milkereit) pp. 39–49. Prospectors and Developers Association of Canada, Toronto, Canada. McNeill JD (1980). Electromagnetic terrain conductivity measurement at low induction numbers: Geonics Technical Note TN-6. < http://www.geonics.com/pdfs/technicalnotes/tn6.pdf > McNeill JD (1990): The use of electromagnetic methods for groundwater studies. In Geotechnical and Environmental Geophysics, Vol 1. (Ed. SH Ward) pp. 191–218. Society of Exploration Geophysicists, Tulsa, Oklahoma.
Regolith geophysics
Mayes KA (1992). ‘Applications of geophysical electrical and magnetics methods to regolith mapping at Lawlers, Western Australia’. Report 384R, CSIRO Division of Exploration Geoscience, Perth. Mendoza JA and Dahlin T (2008). Resistivity imaging in steep and weathered terrains. Near Surface Geophysics 6, 105–120. Meyers JB and Cooper M (2004). Sub-audio magnetics (SAM) for mineral exploration. In Proceedings of the Leading Edge Geophysical Technologies for Mineral Exploration Geologists. 6 September, Perth. (Extended Abstract on CD). Australian Institute of Geoscientists, Perth. Meyers J, Worrall L, Lane R and Bell B (2001). Exploring through cover – the integrated interpretation of high resolution aeromagnetic, airborne electromagnetic and ground gravity data from the Grant’s Patch area, Eastern Goldfields Province, Archaean Yilgarn Craton. Part C: Combining geophysical methods for a holistic exploration model: Exploration Geophysics 32, 198–202. Meyers JB, Cantwell, N, Nguyen P and Donaldson M (2005). Sub-audio magnetic survey experiments for high-resolution, subsurface mapping of regolith and mineralisation at the Songvang Gold Mine near Agnew, Western Australia. Exploration Geophysics 36, 125–132 Miller RD and Xia J (1997). Delineating palaeochannels using shallow seismic reflection. The Leading Edge 11, 1671–1674. Milligan PR, Reed G, Meixner T and FitzGerald D (2004). Towards automated mapping of depth to magnetic basement – examples using new extensions to an old method. In Proceedings of the 17th ASEG Geophysical Conference and Exhibition. Sydney. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Mills HH and Speece MA (1997). Ground-penetrating radar exploration of alluvial fans in the southern Blue Ridge province, North Carolina. Environmental and Engineering Geoscience 3, 487–499. Munday TJ and Sumpton J (1999). Regolith electrical structures associated with kimberlite dykes – an example from the Archaean Yilgarn Craton, Western Australia. In Proceedings of the South African Geophysical Association, 6th Bi-Annual Conference.
Cape Town, South Africa. (Extended Abstract on CD). South African Geophysical Association, Parkview, South Africa. Munday TJ, Mathews LR, Sotiroff KJ, Hunter DR and Worrall L (1998). ‘Regolith electrical structures: the Ida area, north of Lake Ballard, Yilgarn Craton, Western Australia’. CRC AMET/CSIRO EM Report 457R, Volumes I and II. CSIRO Exploration and Mining, Perth. Munday TJ, Macnae JC, Bishop J and Sattel D (2001). A geological interpretation of observed electrical structures in the regolith: Lawlers, Western Australia. Exploration Geophysics 32, 36–47. Munday TJ, Sumpton J and Fitzpatrick A (2004). Exploration for kimberlites through a complex regolith cover - a case study in the application of AEM in the deeply weathered Archaean Yilgarn Craton, Western Australia. In Proceedings of the SEG 74th Annual Meeting. 10–15 Oct 2004, Denver, Colorado. (Extended Abstract on CD of Conference Proceedings). Society of Exploration Geophysicists, Tulsa, Oklahoma. Munday TJ, Rutherford JL, Sattel D and Fitzpatrick A (2006). Modelling the subsurface distribution of salt in dryland catchments of southwestern Australia using AEM data – a comparison of EM interpretation techniques. In Proceedings of the 19th Annual SAGEEP 2006, 2–6 April, Seattle, Washington. (Extended Abstract on CD). Environmental and Engineering Geophysical Society, Denver, Colorado. Olhoeft GR (2004). Ground Penetrating Radar (Ground Probing Radar, Subsurface Radar, Georadar, Earth Sounding Radar) www.g-p-r.com> Palacky GJ (1981). The airborne electromagnetic method as a tool for geological mapping. Geophysical Prospecting 29, 60–88. Palacky GJ (1987). Clay mapping using electromagnetic methods. First Break 5, 295–306. Palacky GJ (1989). Advances in geological mapping with airborne electromagnetic systems. In Proceedings of the 3rd Decennial International Conference on Geophysical and Geochemical Exploration for Minerals and Groundwater. pp. 137–152. Ontario Geological Survey Special Volume 3. Ontario Geological Survey, Canada.
247
248
Regolith Science
Palacky GJ and Kadekaru K (1979). Effect of tropical weathering on electrical and electromagnetic measurements. Geophysics 44, 69–88. Papp É (Ed.) (2002). ‘Geophysical and remote sensing methods for regolith exploration.’ CRC LEME Open File Report 144.CRC LEME, Perth. Peters WS (1996). A rough guide to transient electromagnetics (TEM). Australian Institute of Geoscientists Bulletin 19, 21–47. Peters WS (2001). Ground electromagnetics: the basics and recent developments. Australian Institute of Geoscientists Bulletin 33, 15–32. Pracilio G, Street GL, Nallan Chakravartula P, Angeloni JR, Sattel D, Owers M and Lane R (1998). ‘National dryland salinity program, airborne geophysical surveys to assist planning for salinity control, 3. Lake Toolibin SALTMAP Survey – Interpretation Report August 1998’. National Airborne Geophysics Project, World Geoscience Corporation, Perth. Pride SR (1994). Governing equations for the coupled electromagnetics and acoustics of porous media. Physical Review B50, 15678–15696. Rajagopalan S, Carlson J and Wituik D (2007). Kimberlite exploration using integrated airborne geophysics. In Proceedings of the 19th ASEG Geophysical Conference and Exhibition, Perth. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Ritz M, Parisot J-C, Diouf S, Beauvais A, Dione F and Niang M (1999). Electrical imaging of lateritic weathering mantles over granitic and metamorphic basement of eastern Senegal, West Africa. Journal of Applied Geophysics 41, 335–344 Robineau B, Join JL, Beauvais A, Parisot J-C and Savin C (2007). Geoelectrical imaging of a thick regolith developed on ultramafic rocks: groundwater influence. Australian Journal of Earth Sciences 54, 773–781. Rosid MS and Kepic AW (2003). Electrokinetic sounding method to map hydrogeological boundaries. In Advances in Regolith Proceedings of the CRC LEME Regional Regolith Symposia. 27–28 November, Perth. (Ed. IC Roach) pp. 364–368. CRC LEME, Perth.
Rutherford JL, Munday TJ, Meyers J and Cooper M (2001). Relationship between regolith materials, petrophysical properties, hydrogeology and mineralisation at the Cawse Ni laterite deposits, Western Australia: Implications for exploring with airborne EM. Exploration Geophysics 32, 160–170. Rutley A, Oldenburg DW and Shekhtman R (2001). 2-D and 3-D IP/resistivity inversion for the interpretation of Isa-style targets In Proceedings of the 15th ASEG Geophysical Conference and Exhibition, August, Brisbane. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Sattel D (1998). Conductivity information in three dimensions. Exploration Geophysics 29, 157–162. Sattel D (2006). A brief discussion of helicopter timedomain EM systems. In Proceedings of the Australian Earth Sciences Convention 2006. 2–6 July, Melbourne. (Extended Abstract on CD). Geological Society of Australia, Sydney. Savin C, Robineau B, Monteil G, Beauvais A, Parisot J-C and Ritz M (2003). Electrical imaging of peridotite weathering mantles as a complementary tools for a nickel ore exploration in New Caledonia. In Proceedings of the 16th ASEG Geophysical Conference and Exhibition. February, Adelaide. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Schmidt PW and Clark DA (2000). Advantages of measuring the magnetic gradient tensor. Preview 85, 26–30. Seigel HO and Howland-Rose AW (1990). Magnetic induced polarisation method. In Induced Polarisation: Applications and Case Histories. (Eds JB Fink, BK Sternberg, EO McAlister, WG Wieduwilt and SH Ward) pp. 23–56. Society of Exploration Geophysicists. Tulsa, Oklahoma. Shevchenko S, Morris P and Howard D (2001). Regional gravity and regolith geochemistry as an integrated tool for mineral exploration. In Proceedings of the 15th ASEG Geophysical Conference and Exhibition. August, Brisbane. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Smith RC and Sjogren DB (2006). An evaluation of electrical resistivity imaging (ERI) in Quaternary
Regolith geophysics
sediments, southern Alberta, Canada. Geosphere 2, 287–298. Smith RJ and Pridmore DF (1989). Exploration in weathered terrains: 1989 perspective. Exploration Geophysics 20, 411–434. Smith RS, Annan AP, Lemieux J and Pedersen RN (1996). Application of a modified GEOTEM system to the reconnaissance exploration for kimberlites in the Point Lake area, NWT, Canada. Geophysics 61, 82–92. Smyth EL and Barrett DM (1994). Geophysical characteristics of the tertiary palaeochannels in the Yilgarn Block, Western Australia. In Geophysical Signatures of Western Australian Mineral Deposits. Geology and Geophysics Department (Key Centre) and UWA Extension, Publication 26. pp 417–425. University of Western Australia, Perth. Sørensen KI and Auken E (2004). SkyTEM – a new high-resolution helicopter transient electromagnetic system. Exploration Geophysics 35, 191–199. Spies B and Woodgate P (2005). ‘Salinity mapping methods in the Australian context’. Land and Water Australia, Canberra. Stolz EM (2000). Electromagnetic methods applied to exploration for deep nickel sulphides in the Leinster area, Western Australia. Exploration Geophysics 31, 222–228. Stolz EM (2005). Regolith mapping in hypersaline environments: a comparison of SAM with helicopter TEM. Exploration Geophysics 36, 157–162. Stolz EM and Roache TJ (2007). A thoroughly modern gold deposit: integrating multiple technologies at St Ives. In Proceedings of the Kalgoorlie ’07 Conference. 25–27 September 2007, Kalgoorlie, Western Australia. (Eds FP Bierlein and CM Knox-Robinson) pp.145–148. Record 2007/14. Geoscience Australia, Canberra. Street GJ (1989). MMR Surveys for the location of palaeochannels in the Eastern Goldfields, Western Australia. Bulletin of the Australian Society of Exploration Geophysicists 20, 123–126. Taylor MJ, Smettem K, Pracilio G and Verboom W (2002). Relationships between soil properties and high-resolution radiometrics, central eastern
wheatbelt, Western Australia. Exploration Geophysics 33, 95–102. Telford WM, Geldhart LP and Sheriff RE (1990). Applied Geophysics. 2nd edn. Cambridge University Press, Cambridge, UK. Thompson AH and Gist GA (1993). Geophysical applications of electrokinetic conversion. The Leading Edge 12, 1169–1173. Thompson AH, Hornbostel S, Burns J, Murray T, Raschke R, Wride J, McCammon P, Sumner J, Haake G, Bixby M, Ross W, White BS, Zhou M and Peczak P (2007). Field tests of electroseismic hydrocarbon detection. Geophysics 72, N1–N9. Thomson S, Fountain D and Watts T (2007). Airborne geophysics – evolution and revolution. In Proceedings of Exploration 07: Fifth Decennial International Conference on Mineral Exploration. 9–12 September, Toronto, Canada. (Ed. B Milkereit) pp. 19–37. Prospectors and Developers Association of Canada, Toronto, Canada Timms W and Acworth RI (2002). Origin, lithology and weathering characteristics of Upper Tertiary – Quaternary clay aquitard units on the Lower Murrumbidgee alluvial fan. Australian Journal of Earth Sciences 49, 525–537. Tracey R and Direen N (2002). ‘Gravity.’ Open File Report 144. CRC LEME, Perth. Valleau NC (2000). HEM data processing – a practical overview. Exploration Geophysics 31, 584–594. Waring C, Stone D, Hankin S and Baird D (2002). An integrated geophysical and geochemical approach to solving groundwater flow problems. In Victoria Undercover: Benalla 2002 Conference Proceedings and Field Guide. (Eds GN Phillips and KS Ely) pp. 261–264. CSIRO Publishing, Melbourne. Webb D and James B (2001). The application of electrical geophysics to gold exploration at Mt Wright, North Queensland. In Proceedings of the ASEG 15th Geophysical Conference and Exhibition. August, Brisbane. (Extended Abstract on CD). Australian Society of Exploration Geophysicists, Perth. Whitford M, Meyers JB and Stolz N (2005). The SAM EQMMR response of the regolith at East Victory, St Ives Gold Mine, Western Australia. Exploration Geophysics 36, 133–139.
249
250
Regolith Science
Wildman JE and Compston, D (2000). Magnetic expression of palaeodrainage systems in the Yandal greenstone belt: implications for exploration. In Yandal Greenstone Belt. (Eds N Phillips and R Anand). Bulletin 32. pp. 135–144. Australian Institute of Geoscientists, Perth. Wilford JR (1992). Regolith mapping using integrated Landsat TM imagery and high resolution gammaray spectrometric imagery – Cape York Peninsula. Bureau of Mineral Resources Record 1992/78. Canberra. Wilford JR (1995). Airborne gamma-ray spectrometry as a tool for assessing relative landscape activity and weathering development of regolith, including soils. AGSO Research Newsletter 22, 12–14. Wilford JR (2002). ‘Airborne gamma-ray spectrometry.’ Open File Report 144. CRC LEME, Perth. Wilford JR (2004). ‘3D Regolith architecture of the Jamestown area – implications for salinity’. Open File Report 178. CRC LEME, Perth. Wilford JR, Bierwirth PN and Craig MA (1997). Application of airborne gamma-ray spectrometry in soil/regolith mapping and applied geomorphology. AGSO Journal of Australian Geology and Geophysics 17, 201–216. Wilford JR, Gibson DL, Lawrie KC and Tan KP (2002). Extending regolith-landform maps into the third dimension – unravelling the palaeogeographic story for mineral exploration and environmental applications. In Victoria Undercover: Benalla 2002 Conference Proceedings and Field Guide. (Eds GN Phillips and KS Ely) pp. 193–202. CSIRO Publishing, Melbourne. Wilkes P, Abbott S, Harris B and Dogramachi S (2004). Ground geophysics for rapid location of buried valleys below high-value salt-affected land at Tammin, WA. In Conference Proceedings Engineering Salinity Solutions 1st National Salinity Engineering Conference 2004. (Eds S Dogramachi and
G Waterhouse) pp. 400–405. Institution of Engineers, Canberra. Wilkes P, Harris BD and Kepic A (2006). Geophysical investigations in rural towns of Western Australia. In Proceedings of the 19th Annual SAGEEP. 2–6 April, Seattle, Washington. (Paper on CD). Environmental and Engineering Geophysical Society, Denver, Colorado. Witherly K, Irvine R and Morrison E (2004). The Geotech VTEM time-domain helicopter EM system. In Proceedings of the SEG 74th Annual International Meeting. 10–15 October, Denver. (Extended Abstract on CD). Society of Exploration Geophysicists, Tulsa, Oklahoma. Worrall L and Gray D (2004). Regolith in the central Gawler, through and through. In Gawler Craton: State of Play 2004 - Conference Abstracts. (Compiler P Lyons) p. 27. Department of Primary Industries and Resources, South Australia, Report Book 2004/18. Office of Minerals and Energy Resources, South Australia. Department of Primary Industries and Resources, Adelaide. Worrall L, Lane R, Meyers J and Whitaker A (2001). Exploring through cover – the integrated interpretation of high resolution aeromagnetic, airborne electromagnetic and ground gravity data from the Grant’s Patch area, Eastern Goldfields Province, Archaean Yilgarn Craton, Part A: Mapping geology using airborne electromagnetics (TEMPEST). Exploration Geophysics 32, 190–193. Yaramanci U and Hertrich M (2006). Magnetic resonance sounding. In Groundwater Geophysics – A Tool for Hydrogeology. (Ed. R Kirsch) pp. 253–273. Springer, Berlin, Heidelberg, New York. Zhou W, Beck BF and Stephenson JB (1999). Reliability of dipole–dipole electrical resistivity tomography for defining depth to bedrock in covered karst terranes. Environmental Geology 39, 760–766.
10 Regolith and water Richard G Cresswell and Paul Shand
10.1
INTRODUCTION
As indicated throughout this volume, water is a critical factor in weathering. Regolith processes form the most complex part of the water cycle: representing the interface between rock, air and water, and also incorporating biogeochemical cycling (Figure 1.3). This chapter emphasises the ways in which the regolith controls water flow and quality. The principles of groundwater flow and hydrogeochemical principles are presented and selected case studies are used as examples of the importance of water in the regolith. (The use of groundwater as a sampling medium for mineral exploration is discussed in Section 13.13.)
10.2
THE WATER CYCLE
Of the world’s 1.5 ×1014 M tonnes of water, 97% resides in the oceans and 2% is locked up in icecaps and glaciers: leaving only 1% within, or on the landmasses. Almost all of this is stored beneath the land’s surface as groundwater; only about 0.01% lies on the Earth’s surface and only 0.001% of the global budget resides in the atmosphere. The water cycle describes the movement of water on and within the Earth and may be represented as a system of storages, which are linked by processes and
driven by the global climate and biosphere through precipitation, evaporation and transpiration (Figure 10.1). Processes are characterised by rates of movement, which vary from hours (such as atmospheric water cycling) to millennia (such as confined groundwater flow through an aquifer). We may estimate global average residence times for water in each reservoir by assuming the volumes are in dynamic equilibrium, which gives: T = V/Q
(Eqn 10.1)
where T is the average residence time; V is the reservoir volume (×1020 g H2O) and Q is the discharge into or out of the reservoir (×1020 g H2O)/year). Thus, ocean water has an average residence time of 13 700 / (0.34 + 0.01) 5 39 000 years, based on the influx from surface and groundwaters, while the atmospheric water has a mean residence time of about 10 days. By this reckoning, groundwater has a residence time – based on direct discharge to the oceans (0.01 × 1020 g H2O/year) – of 9500 years. Most groundwater discharge, however, is to river systems, which contributes to the run-off flux to the oceans and thus reduces the mean residence times to around 300 years. From this broad, global, approach to groundwater residence time, we can evaluate specific rates of discharge for
252
Regolith Science
Evapotranspiration
Evaporation
Precipitation
1.1 0.75
interception
3.95
3.6 throughfall
Interception storage
overland flow Surface storage
infiltration
Soil moisture
runoff
0.34
Oceans
interflow 0.01
groundwater recharge Unconfined groundwater
baseflow
confinement Confined groundwater
discharge
Figure 10.1: The hydrological cycle may be represented as a system of storages and processes that determine the rates of water movement between the stores. The shaded area represents the region where the regolith has its strongest interaction. Rectangles represent water storages; ovals represent inputs and outputs to the system; rate determined processes link the storages; entities in circles represent the volumes of water transferred each year ×108 M tonnes (data from Budyko 1982).
groundwater flow systems from local to regional scales. The dynamic nature of this cycle is illustrated in Figure 10.2, which shows the inter-relatedness of different compartments. Importantly in the regolith, groundwater storages – or aquifers – that are unconstrained by overlying materials must be distinguished from those whose volume is constrained by boundaries that impede the vertical movement of water. Shallow, unconfined aquifers are connected to the atmosphere through the pore and fracture network of the host rocks, and hence occur within the zone in which regolith processes are most likely to occur (Figure 10.2). Although some confined systems may also occur in the regolith, the majority occur within host formations that are below the zone of weathering. As the bulk of groundwater use comes from shallow, unconfined, systems, it is vital to understand the properties of, and processes occurring in, the regolith if we are to understand the movement of water
in it and hence evaluate its potential to host water resources. This is especially true in countries such as Australia, where regolith processes may extend more than 100 m beneath the surface (Section 1.1). The processes illustrated in Figures 10.1 and 10.2 are distinguished by their respective propensity to transmit and store water. The potential for a fluid to flow from one region to another was classically described by Henri Darcy (1856) from experiments used to determine flow of water in sands as part of his treatise on the water supply of Dijon, France. A mathematical and physical explanation of this process was outlined in some detail by Hubbert (1940) and is repeated in most groundwater texts (for example, Freeze and Cherry 1979). Darcy’s Law has borne the test of time, but, as noted below, does not apply in some important regolith environments. Darcy’s Law describes fluid flow in a macroscopic flow environment. That is, flow occurs through a
Regolith and water
Evaporation transpiration
Overland flow Unsaturated zone
Precipitation Evaporation from oceans
Water table Stream flow
Saturated zone of aquifer Unc
onf in
Co nf
ed Aqu ground itar wat d er
ine
dg
Sea
rou
ndw
ate r
Upward leakage through aquitard
Figure 10.2: Schematic representation of the hydrologic cycle. The dashed line encloses the zone in which regolith processes are most likely to occur (after Shand et al. 2007a).
large-enough volume such that it represents the average flow through the entire medium. Thus, Darcy’s Law pertains to the discharge of water through a volume of material and the relationship between the measured parameters of hydraulic gradient and hydraulic conductivity (K – the constant of proportionality characteristic of the medium through which the fluid flows). This law may be expressed as: v = - Kdh/dl = Q/A
(Eqn 10.2)
where: v is known as the Darcy velocity, Darcy flux, specific discharge or volume flux; h is the hydraulic head; and l is the distance between two measurements of hydraulic head on a similar flow path. Thus, dh/dl is the hydraulic gradient; Q is the discharge across an area, A. In most natural systems, hydraulic conductivity becomes largely a function of the ability of a material to allow a fluid to flow; that is, permeability. Permeability, in turn, is a function of the porosity of a material, and the way in which those pores are connected. While the porosity of natural materials varies by one order of magnitude (15–70%), permeability – and
hence hydraulic conductivity – can vary by 13 orders of magnitude (Figure 10.3). The driving force for water movement is the hydraulic head, whereby water will always attempt to move from high to low energy environments (Hubbert 1940). The water table sits at a pressure head in equilibrium with the ambient atmosphere (and this explains subtle variations in water table height as weather conditions vary). Below the water table, water movement is controlled by gravity and is subject to hydrostatic pressure; that is, the weight of the overlying water. Surface tension holds moisture to grain surfaces and forms a meniscus between grains. Thus, an unsaturated zone is that zone where air replaces some of the pore space that would be occupied by water in a saturated zone. Due to surface tension forces, this occurs at a negative pressure head that is dependent on the grain size and separation of the medium. This point is known as the air-entry point, ya, and has a more negative value for clays (up to several metres) than for sands (generally only a few mm). Hence clayrich soils develop thicker ‘tension-saturated’ zones
253
Regolith Science
IGNEOUS AND METAMORPHIC ROCKS (0-10%) Unfractured
Fractured
BASALT (5-50%) Unfractured
Lava flow
Fractured
SANDSTONE (5-30%) Semi-consolidated
Fractured
SHALE (0-10%) Unfractured
Fractured
CARBONATE ROCKS (0-50%) Cavernous
Fractured
SILT, DUST (35-50%) CLAY (40-70%)
SILTY SAND CLEAN SAND (25-50%) Fine
Coarse
GLACIAL TILL (25-60%)
GRAVEL (25-40%)
K (m d-1) 10-8
10-7
10-6
10-5
10-4
10-3
10-2
10-1
1
102
10
103
104
k (darcy) 10-8
10-7
10-6
10-5
10-4
10-3
10-2
10-1
1
10
102
103
104
Figure 10.3: Range of hydraulic conductivities (K), permeabilities (k) and porosity values (%) for common aquifer materials (summarised from Freeze and Cherry 1979).
this case, flow is so slow that it is only of academic interest. Darcy’s Law also does not apply under very high flow conditions. Surface Dry soil; air fills voids Discrete moisture films at grain contacts Continuous moisture films forming an interconnected system of capillary films. Moisture content increases with depth. Continuous moisture films with entrapped air pockets. Moisture content at capillary saturation.
Tension in soil water film
than do sandy soils. This zone is more commonly known as the capillary fringe, and separates the saturated from the unsaturated zones in an unconfined environment. A continuum in water content exists, however, and we must further distinguish zones within the capillary fringe (Figure 10.4). Thus, we may characterise the capillary fringe as the zone in which a continuous meniscus of water coats the host grains. A zone of capillary saturation is located immediately above the water table, and for a thickness determined by the size of pores (being greater for smaller pores). This zone has a constant moisture content and generally contains some air bubble inclusions. Above this, the moisture content decreases and air bubble content increases until a level is reached whereby a continuous film can no longer be sustained. This marks the upper surface of the capillary fringe. The vadose zone includes both the unsaturated zone and the capillary fringe down to the water table (Figure 10.5). Properties of the saturated and unsaturated zones are summarised in Table 10.1. Darcy’s Law is applicable for most geological situations but, for low permeability materials under very low hydraulic gradients, it is inapplicable. However, in
Capillary fringe
254
Water table Hydrostatic pressure
Figure 10.4: The zones of capillary moisture in regolith and soils (after Carson 1969).
Regolith and water
Table 10.1:
Properties of saturated versus unsaturated regolith. Saturated zone
Unsaturated zone
Occurrence
Below the water table
Above the water table; above the capillary fringe
Saturated pores (e)
Moisture content (q) = e
q<e
Fluid pressure (p)
p > 1 atmosphere
p < 1 atmosphere
Pressure head (y)
y>0
y < ya
Measurement of hydraulic head
Piezometer
Tensiometer
Hydraulic conductivity K
K = K0 (constant)
K = K(y)
Moisture content
y=e
y = q(y)
Darcy’s Law describes what is known as a direct flow phenomenon: hydraulic head gradient drives water flow. This is similar to a chemical concentration gradient driving solute flow, temperature gradients driving heat flow and electrical gradients driving the flow of electrons. In addition to direct flow, there can be coupled flow: whereby an indirect gradient drives water flow (Table 10.2). Thus chemical gradients cause osmotic flow and electrical currents can drive electroosmosis and streaming currents. Streaming currents occur where the flow of an electrolyte-bearing ground-
water generates electrical anomalies that reveal subsurface flow. Electro-osmosis can dewater fine-grained materials and chemical osmosis can be important in some clays, which can act as semi-impermeable membranes, leading to water flow from low to high concentration regions (Ingebritsen et al. 2006). Groundwater flow can be analysed analytically (Cooper 1966), but in general it is modelled numerically (Freeze and Cherry 1979), with models such as MODFLOW (McDonald and Harburgh 1988) becoming industry standards. Piezometer
Tensiometer
Porous cap
Unsaturated zone Vadose zone Water surface Capillary fringe Water table (pressure = 1 atmosphere)
-
Saturated zone h
Groundwater zone z
z h
datum level
Figure 10.5: Different equipment is required to measure hydraulic head for saturated (piezometers) and unsaturated (tensiometer) conditions. A distinction must be made between the saturated zone and the groundwater zone; the former includes a variable thickness of saturated, but negative pressure head, water: the capillary zone (see text for further details). (modified from Rose 2004; Thomasson and Youngs 1975).
255
Regolith Science
Table 10.2: Direct and coupled flow phenomena (after Ingebritsen et al. 2006). Direct flow phenomena appear along the indicated diagonal. Gradient
Flow
256
10.3
Hydraulic head
Chemical concentration
Temperature
Electrical
Fluid
Darcian flow (Darcy’s Law)
Chemical osmosis
Thermo-osmosis
Electro-osmosis
Solute
Ultra-filtration
Diffusion (Fick’s First Law)
Thermal diffusion
Electrophoresis
Heat
Isothermal heat transfer
Dufour effect
Heat conduction (Fourier’s Law)
Peltier effect
Charge
Streaming current (Rouss effect)
Diffusion and membrane potential
Thermoelectricity (Seebeck effect)
Electrical conduction (Ohm’s Law)
REGOLITH AND GROUNDWATER
We may broadly define the water beneath the surface of the Earth into two compartments: soil moisture and groundwater. Soil moisture represents less than 0.1% of this mix. The groundwater realm may be divided into a zone that is in contact with the atmosphere (through the pore network of the aquifer deposits), and one that is isolated from the atmosphere by a layer through which water flow is sufficiently slow, or non-existent, to restrict effective flow to within the aquifer only. This low-flow layer, termed an aquitard, may be saturated, but the connectivity of any pores is so low that flow in the system is dominated by flow in the aquifer. If this layer represents a true no-flow zone, such as a crystalline igneous body, it is referred to as an aquiclude. If the water table – the level at the top of the saturated zone – lies within the aquifer, the aquifer is termed un-confined. If it is open to the atmosphere, it is a phreatic, unconfined aquifer. If the aquifer is bounded by aquitards and is saturated, water in the pores become over-pressured and a bore penetrating the aquifer will allow water to rise within it to a level that is related to the pressure of water in the aquifer. This type of monitoring bore is known as a piezometer and measures the potentiometric level of the water in the aquifer – in contrast to a bore drilled into an unconfined aquifer, which generally, but not always, measures the level at which the water table sits in the aquifer material. Aquifers that exhibit a potentiometric surface – that is, above the top of the aquifer boundary – are termed confined. Confined aquifers are generally bound between two
such aquitards. Flow is determined by pressure gradients from remote sites. Recharge to confined layers is generally via an unconfined layer or zone. Thus a zone of sub-confinement may exist, which may temporally vary in extent (horizontally and vertically) – depending on climatic and anthropogenic impacts. Also, the simple concept of an unsaturated zone overlying a saturated one may not be so simple in reality. A common feature of Australian unconfined systems is the presence of discontinuous, low permeability lenses within a broader, higher permeability system, such as occurs in most alluvial systems. In this instance, vertical recharge will be impeded in the vicinity of the lens and a perched saturated zone will result. This gives rise to both a perched water table overlying the lens and an inverted water table within the low-permeability layer. This overlies a zone of unsaturated material above the true water table as shown in Figure 10.6. These conditions are often discontinuous in time as well as space. Thus, when sufficient recharge infiltrates to the lens, the saturated layer on the lens will persist longer than the saturated zone created by the recharge event in the higher permeability zone. Thus, clay-rich sub-soils leave a region prone to water-logging in wet periods, but may allow prolonged seasonal growth during dry periods due to the retarded movement of water through the profile.
10.4 HYDROGEOLOGICAL PROPERTIES OF THE REGOLITH There are six basic properties of fluid and porous media that must be quantified to evaluate water
Regolith and water
Sand
Perched water table
B A
C
Clay D Unsaturated zone E
F
Water table Saturated zone
KSf040-08
Figure 10.6: A perched water table (ABC) may occur where a low permeability lens impedes downward leakage, or recharge, en route to a deeper ‘true’ water table (EF). An inverted water table (ADC) is created in the lowpermeability layer (after Freeze and Cherry 1979).
movement in the regolith: three fluid parameters and three medium parameters (Table 10.3). Fluid parameters are for water with salinities ranging from fresh to sea water, and for a range of materials from unconsolidated gravels to unfractured igneous rocks (after Freeze and Cherry 1979). Regolith materials cover the entire range of parameter values. From these basic properties, we may describe all necessary parameters relevant to the storage and transport of water in the regolith. As can be seen from Table 10.3, the parameters relevant for the media can be highly variable and consequently have a substantial effect on rates of movement and quantities of storage. In particular, permeability is the most critical hydroTable 10.3:
logic parameter – with a variability that ranges from 10 –23 m2 to 10 –7 m2 (13 orders of magnitude), and varies with scale and the method of measurement, as well as with time and depth (Hsieh 1998); that is, the porosity of a medium does not necessarily relate to its permeability. Furthermore, the variability in permeability in any given material in three dimensions may be six orders of magnitude – depending on grain size, orientation and pore alignment or fracture spacing. It is, therefore, also common to define a depth-averaged property called transmissivity (T = Kb; where b is the thickness of the aquifer, or system and K is hydraulic conductivity: see also Figure 10.3), to describe flow resistance for an aquifer. This is a useful concept to embrace where only a few in situ measurements are to be used to describe an entire aquifer. Two important caveats arise from this information: 1. Scale is an important determinant when considering any fluid rock interaction: we must be able to define a representative elementary volume 2. For most applications, an order of magnitude in accuracy of hydraulic conductivity may be adequate for the description of water transport in the regolith. 10.4.1 Sources of groundwater Rainwater is the primary source for groundwater, and is predominantly distilled from the oceans (Figure 10.1), but contains sea spray and aerosols, together with continental dusts and gases. The transfer of water – from evaporation, through condensation and precipitation and finally flowing back to the oceans
Basic properties of fluids and porous media at 15.5°C.
Fluid properties
Range#
Units
p – density
1 – 1.035
g/cm3 (kg/m3 × 103)
m – viscosity
1.124
cP (kg/m.s = N.s/m2×10 –3)
B – compressibility
~4.6 × 10 –10
m2/N (m.s2/kg)
Reference Weast 1985
Medium properties n – porosity
15–70%
k – permeability
10 -8 – 105
darcy (~10 –8 cm2)
a – compressibility
10 –11
m2/N (bedrock – clays)
#
–
10 –6
Freeze and Cherry 1979
For waters up to seawater salinity and covering all regolith materials.
Domenico and Mifflin 1965; Johnson et al. 1968
257
258
Regolith Science
through the water cycle (Figure 10.1) – encounters a wide range of geochemical environments, which gives rise to a wide range of hydrochemical characteristics as water moves through the upper layers of the Earth. Water can thus be divided into several types – both spatially and chemically – depending on its origin. Most of the groundwater on the continents is derived from meteoric water: that is, it is derived from the atmosphere as rain or snow, during relatively recent geological time. Meteoric waters may become saline through evaporation at the Earth’s surface, through dissolution reactions, or through mixing with waters that represent trapped sea water in the pores of deposits laid down under marine conditions: these are known as connate waters. The mixing and flushing of aquifers means that it is unlikely that any groundwater exists that meets the true definition of connate water and so the more generic term ‘formation water’ is more commonly used. In addition to these sources, water may also be transported in magmas from the Earth’s mantle: this is termed juvenile water, which is water that has not been previously involved in the water cycle. However, groundwater discharge in geothermal areas (such as geysers or hot springs), where young volcanic rocks occur, has generally been shown to be of meteoric origin. Juvenile water is considered to be an extremely small component of groundwater. 10.4.2 Groundwater storage An increase in porosity, fluid density and/or volumetric fluid saturation for a given medium also increases the amount of fluid that medium can store. In the case of a saturated confined aquifer, the volumetric fluid storage is 1 and the fluid density is constant. Specific storage, Ss, can thus be defined as a function of porosity – and the compressibility of the fluid and the medium – and as the change in the volume of fluid stored in a unit volume of porous medium per unit change in hydraulic head. In unconfined systems, the amount of fluid present will vary with water table elevation, as well as with porosity and fluid density. In these cases, additional pores can be filled and a unit response in hydraulic head can result in large volumes of water added or removed from the system – depending on the porosity.
High values for storativity reflect the fact that releases from storage in unconfined aquifers represents a delayed yield upon dewatering – or drainage – of the soil pores, whereas releases from storage in confined aquifers represent only the secondary effects of water expansion and aquifer compaction caused by changes in the fluid pressure (Freeze and Cherry 1979). 10.4.3 Groundwater flow systems Figure 10.2 shows a diagrammatic representation of water flow in a landscape. In reality, the situation is much more complicated, and is controlled not only by slope shapes and angles, but also by the nature of the regolith and the bedrock lithology. In order to take account of the spatial variability in the way water moves in different landscape and geological settings, a groundwater flow systems (GFS) framework has been set up in Australia to help in the management of salinity (Coram 1998; Coram et al. 2000; Walker et al. 2003). A GFS is defined as a set of real aquifers that share similar characteristics and where processes leading to salinity are similar. Rock types and landscape features are used to define individual GFS, which are grouped into three types (Figure 10.7: Coram et al. 2000): 1. Local flow systems extend only a few kilometres along flow paths. They respond rapidly to increased groundwater recharge, and salinity may appear within 20–30 years of land clearing. 2. Intermediate flow systems are about 5–50 km along flow paths. They may take 50–100 years to develop land salinity (perhaps less for river salinity) and have a greater storage capacity and permeability than local systems. They take longer to respond to increased recharge. 3. Regional flow systems are typically more than 50 km along flow paths. They may show no signs of land salinity for over 100 years. They have a high storage capacity and high permeability, and respond very slowly to increased recharge. Wilford et al. (2006, 2007) have studied GFS in more detail – incorporating detailed information on the regolith and bedrock characteristics of several study areas in eastern Australia. The results are maps
Regolith and water
Int
erm
80
Loca
ed
iat
l
e
60
40
Regi onal
Response to change (%)
100
20
0 0
50
100
150
200
Time (years) Figure 10.7: Generalised groundwater response for local intermediate and regional flow systems.
of regolith hydrogeomorphic units (RHUs) and regolith hydrogeomorphic associations (RHAs). RHUs are regions or areas that display similar soils/regolith, landforms, bedrock geology and therefore hydrological characteristics. The RHU represents detailed hydrological landscape subdivision and its purity is largely scale-dependent. Emphasis is placed on regolith materials and bedrock structures, both of which have a major control on surface and groundwater flow. RHUs are grouped into RHAs, which consist of a hydrologically interconnected RHU cluster with similar patterns of soils, regolith and bedrock materials and hydrological characteristics.
10.5 HYDROGEOCHEMISTRY OF THE REGOLITH The regolith environment represents a critical interface where organic and inorganic chemical reactions create a complex zone of weathering reactions and fronts. This is a region of intense geochemical gradients: for example, in terms of oxidising/reducing conditions, acid generation and consumption, temperature, biogeochemical reactions and concentrations of dissolved (solute) constituents. This is especially the case in the uppermost parts of the regolith where soil formation leads to the production of distinctive layers of weathered material that are typically associated with organic matter and biogeochemical cycling. The depth of regolith (that is, the
thickness of the weathered zone) is a function of the rate of production by weathering processes relative to the rate at which the products of weathering are removed by erosion. Water has a variety of distinctive properties, such as its polar nature, that make it a good solvent and important as a weathering agent. Chemical weathering, through the action of water, dissolved CO2 (forming carbonic acid, H 2CO3) and other acids, causes the alteration and breakdown of regolith minerals and changes in the chemistry of the regolith and surrounding groundwater. Chemical weathering without water is slow or non-existent (cf. weathering on the Moon: Section 14.2). Hence, water exerts a major control on the release of nutrients, such as metal cations, P, N and Si, to surface and subsurface ecosystems. The hydrochemical characteristics of waters can provide a wealth of information on the chemical processes taking place within the regolith. However, it should be borne in mind that geochemistry of the regolith is a product of ancient and modern biogeochemical processes – often covering millions of years and under different climatic and geomorphological conditions. Nevertheless, an understanding of water– regolith interactions is an important step to deciphering regolith evolution, solute transport and the buffering capacity of the Earth’s protective blanket. Chemical weathering is the term used to describe the transformation of minerals (formed at particular pressure–temperature (P–T) conditions) into more stable secondary minerals and solute species through interaction with dilute waters close to the Earth’s surface (Ollier 1984; Drever 1997; Bland and Rolls 1998). Natural waters evolve due to a complex sequence of mineral reactions – many aided by microbial activity, as well as through gaseous exchange, sorption reactions, mixing and dilution (Drever 1997; Hem 1992; Langmuir 1997; Appelo and Postma 2005). Initial inputs from rainfall are important for some solutes: particularly Cl– and Br– ions. Once in the subsurface, groundwater chemical compositions evolve through biogeochemical reactions in the soil zone, unsaturated zone and, ultimately, the saturated zone. Some solutes are also introduced as a result of human activity at, or near, the land surface. Geochemical modelling is often necessary to predict, in detail,
259
260
Regolith Science
Table 10.4: Classification of different water types based on total dissolved solids (TDS). Type
TDS (mg L–1)
Fresh
0–1000
Brackish
1000–10 000
Saline
10 000–100 000
Brine
> 100 000
changes in water chemistry due to the large number of variables, but is often limited by data input. The reader is referred to texts that deal specifically with modelling (such as Bethke 1996; Appelo and Postma 2005). This section discusses the main hydrogeochemical processes reactions occurring within the regolith environment and main controls on solute chemistry. The mineralogy and geochemistry have been discussed in Chapters 4, 5 and 6. 10.5.1 Chemical components of groundwater Groundwater can be classified according to the quantity of total dissolved solids (Table 10.4). The actual dissolved constituents in groundwater are numerous: most elements are soluble to some degree. However, relatively few are soluble enough, or abundant enough, in the Earth’s crust to make them quantitatively important. The elements in natural waters can be divided into major, minor and trace constituents, depending on typical concentrations found (Table 10.5). Many of the major elements are present as major components of common minerals. However, some elements present at high concentrations in the Earth’s crust (such as Al at around 6 wt. %) are only found at trace levels in most waters due to their limited solubility under natural weathering conditions. Table 10.5: Typical concentration ranges of some dissolved inorganic elements in groundwater. Major constituents (> 5 mg L–1)
Minor constituents (0.01–10 mg L –1)
Ca2+, Mg2+, K+, Na +, HCO3 –, Cl –, SO42–
B4+, Fe2+, Fe3+, Si4+, F – , NO3 – , Br–
Trace constituents (< 0.1 mg L–1) Other elements
10.5.2 Units of analysis The concentration of dissolved substances can be represented in different ways depending on the use and presentation of the data. Concentrations are usually given as an amount per unit volume. For groundwaters, the most common measure is mg L –1. This is a measure of the weight of an element and, because different elements have different atomic weights, it is not a measure of the number of atoms/molecules present. Therefore, the term mmol l–1 is useful. In order to convert from mg L –1 to mmol L –1, simply divide by the gram formula weight, for example: 80 mg L- 1 Ca = 80/gfw of Ca = 80/40 = 2 mmol L- 1 (Eqn 10.3) This is obviously useful when studying chemical reactions, for example: CaCO 3 + CO 2 + H 2 O = Ca 2 + + 2HCO3 (Eqn 10.4) In this example, for every mole of calcite dissolved, one mole of Ca2+ and 2 moles of HCO3 – are produced. The equivalent weight is useful when the chemical behaviour of the element is of interest in terms of charge transfer, and is extremely useful for checking a chemical analysis (due to its bipolar nature, water and, hence, the solutes it dissolves, remains electrically neutral, so the sum of positive charges and negative charges should be equal): meq L- 1 = mmol L- 1 # charge of the ion (Eqn 10.5) e.g. 2 mmol L –1 Ca2+ = 4 meq L –1 Ca. 10.5.3 Graphical display of groundwater A wide range of graphical displays can be used to display the characteristics of individual analyses or ranges present in different water bodies, aquifers or pore fluids. Standard pie charts (Figure 10.8) and bar plots are sometimes used, but it is difficult to visualise large amounts of data and their relationships. The Stiff plot is effective – as different water types plot as different shapes (Figure 10.9) – but is also inconvenient for
Regolith and water
15-1
10
Ca Mg Na + K
Na + K Cl SO4
17-3
HCO3 Na + K
Mg Mg
Ca
Ca Cl HCO3
HCO3
12-6
SO
4
Cl
Mg Ca
SO4
Na + K Cl
HCO3
0
1
5 10
50
Scale of radii (total of milliequivalents per litre)
SO4
100 KSf016-08
Figure 10.8: Water analyses plotted as pie charts, which are subdivided on the basis of concentrations of major elements in meq L–1 (after Hem 1992). Sample numbers in bold, see Figure 10.9.
visualising large numbers of samples. Stiff plots can be useful for displaying spatial variability on a map. Probably the most common plot used by hydrogeologists is the Piper plot (Figure 10.10a). Two triangular plots are first used to plot the cations and anions and these are then projected diagonally onto a diamond-shaped figure. The position of these samples on the triangular plots can be used to classify waters according to the dominant cation and anion, such as Ca-HCO3 or Na-Cl (Figure 10.10b). The Piper plot is useful to see the relationships between a relatively large number of samples, but it needs to be remembered that it only displays the relative proportions of ions and not absolute concentrations; it should therefore only be used as a guide. It is possible, however, to plot samples with different sized symbols: for example, according to TDS, which will allow evolutionary sequences to be seen more clearly. An additional feature is that mixing produces linear relationships on the diagonal plot: for
example, mixing between recharge (Ca-HCO3) and sea (Na-Cl) type waters. Figure 10.10b shows the trends observed for this mixing – as well as those for seawater intrusion and aquifer freshening, which are described below (Section 10.7.4). 10.5.4 Master variables in hydrogeochemistry The parameters temperature (T), pH and redox potential (Eh) are described as master variables in aqueous chemistry, and exert major controls on mineral stability, solubility, kinetics and speciation. Most chemical reactions are temperature dependent, and reaction rates typically double for each 10oC rise. The solubilities of most substances are also T dependent: in general, becoming more soluble at higher T, although there are important exceptions, such as calcite, which is less soluble as T increases. Temperature also has a significant effect on biological
261
262
Regolith Science
Na + K Ca Mg Fe
Cl HCO3 SO4 CO3 10
12-6
15-1
17-3
30
25
20
15
10
Cations, in milliequivalents per litre
5
0
5
10
15
20
25
Anions, in milliequivalents per litre
Figure 10.9: Water analyses plotted as stiff plots, which are subdivided on the basis of concentrations of major elements in meq L–1 (after Hem 1992). Sample numbers as in Figure 10.8.
activity, affecting concentrations of O2 and CO2, as well as organic acids in the regolith (Chapters 7 and 8). The solubilities of many elements, such as Al, Fe, Si and the rare earth elements, are strongly dependent on pH (Section 5.3.6; Figure 5.10). The redox potential is a major controlling factor on the solubility and speciation of elements that exist in different oxidation states, such as Fe2+ and Fe3+. Many redox reactions also involve the transfer of protons (H+), hence pH and Eh are often linked as a major controlling factor on redox sensitive species (see Chapter 5.3.4). 10.5.5 Groundwater sampling and hydrogeological considerations Several parameters may change rapidly once the groundwater is removed from an aquifer, and therefore
in situ or on-site measurements are required. The unstable parameters include temperature (T), pH, specific electrical conductivity (SEC), redox potential (Eh) and dissolved oxygen (DO). The sample is usually also titrated with acid to determine alkalinity (as HCO3 –). Contact with the atmosphere causes changes in pH, Eh and DO and these should be measured either in situ or using a flow-through cell to avoid contact with air. In order to discard any particulate matter from groundwater, samples are normally filtered: the commonly accepted standard is to use a 0.45 µm filter. The substances that pass through this filter are then classed are being in solution. However, there is a range in molecular sizes of solutes and although the vast majority of truly dissolved species will pass through this filter, some organic macromolecules may be larger than this arbitrary cut-off point (Figure 10.11). The range of colloidal matter also spans across this value and it is possible that fine colloids of Fe, Mn or Al may also be present. Different samples are collected for specific types of analysis. Those for cation analysis are normally filtered and acidified to stop adsorption of metals onto container walls. Anion analysis is completed on filtered, but unacidified, samples. Some anion species are unstable (such as NH4+ and NO2–) and should be analysed as soon as possible following sampling. Groundwater samples are usually obtained as pumped discharges from wells and boreholes or from springs. Such samples may not necessarily be derived from one discrete aquifer unit, but represent an integrated discharge from several horizons (Figure 10.12). In such cases, the sample may represent a mixture of groundwaters drawn from different horizons and, in extreme cases (such as in aquifers dominated by fracture flow), the discharge may not be representative of any groundwater in the aquifer. Although depth-specific sampling from a borehole with long or multiple screens may overcome these problems, head differences with depth may cause downward or upward flow in the borehole itself, and short-circuiting between aquifer units. Ideally, sampling should be completed using multiple piezometers or packer sampling where fundamental studies are required to understand hydrochemical processes. Even in porous aquifers, chemical changes often occur with depth and it is therefore essential that the
263
Regolith and water
(a)
100 100
80 80
60 4
SO
Mg
Cl +
+ Ca
60
40
40
20 20
0 0
Mg
0
100
20
80
40
60
60
60
80
20
40
80
20
100
0 100
80
60
40
20
0
Ca
Na + K 100
4
O +S
4
+S O
Ca +M +M g g, Na +K
n
Ca
Mg a+ +K ,C
80
+K
Na
100
fresh
0
SO
0 1004
0
Sulfate type
60
Chloride type 60
80
40
60 40
60
40 20
Cl
0 100
Ca
80
40
60
60
40
80
80
20
0 100
20
g nin
No dominant type
40
20
80
she
Na
sio
Mg
g
+M
Cl
Ca
u ntr
20
Seawater mixing sea
fre
20
40
er
0
HCO3
ei lin
SO
Bicarbonate type
100 0
Na + K
60
uif
20
40
Aq
40
80
80
60
Sa
g 0
0 1004
40
60
60 No dominant type Sodium or 80 Calcium potassium type type 60
20
20
40
80
+M
O4 +S CO 3 ,H O4 +S
Cl
40
SO 4 l+ ,C O3 O3 HC
60
Magnesium type
100
20
0
0
20
80
40
100
Cl
Ca
60
80
HCO3
80
HC
100
60
80
60
0
Mg
40
100
Cl
20
20
100
80
40
0 0
Cl
(b)
0 100
80
40
40
Ca
100
20
60
20
SO4
0
20
100 80
60
40
20
0
Na + K
0
HCO3
20
40
60
80
0 100
Cl
Figure 10.10: (a) PIPER diagram showing the method for plotting water analyses (b) PIPER diagram showing different water types (left) and use in determining hydrochemical processes (right).
264
Regolith Science
0.45 micron boundary
Zooplankton
Fatty Acids Fulvic Acid
Phytoplankton
Carbohydrates
Humic Acid
Bacteria
Amino Acids
Viruses
Hydrocarbons
Colloidal clay-humic -metal complexes
Particulate 10-2
10-3
10-4
Hydrophilic Acids
Size of water molecules
Dissolved 10-5
10-6
10-7
10-8
10-9
10-10
10-11
Metres Figure 10.11: Dissolved constituents are normally taken as those that pass through a 0.45 µm filter. However, this is an arbitrary value and filtered samples may include small colloidal particles (after Shand et al. 2007a).
local hydrogeology, groundwater flow patterns and the borehole construction details be reasonably well understood. Such knowledge may have extremely useful practical implications because poor quality water is often limited to only part of the aquifer (such as high NO3 – in the shallow, oxidised part of the aquifer or high Fe in deeper, more-reducing units, Figure 10.12). Another potential problem is that stagnant water will be present in the borehole column, especially if it has not been producing for some time. It is therefore necessary to pump out this water and allow fresh groundwater to flow from the aquifer into the borehole column. Estimates for the number of well volumes to be pumped vary from between two and ten depending on local hydrogeological conditions (Figure 10.13). This can be determined by waiting for the master variables to stabilise before sampling. An alternative is to pump at very low rate precisely at the slotted section of the bore: thus sampling water as it is drawn in only from the adjacent aquifer. Extreme care needs to be taken, especially where an aquifer’s hydraulic conductivity is very low or there are large contrasts in hydraulic conductivity or salinities within the sequence.
experimentally and theoretically (see also Section 4.6). The dissolution rates of minerals vary over several orders of magnitude (Table 10.6). Therefore, it is to be expected that equilibrium will be achieved rapidly for some minerals, but not for others. An important deduction from this is that groundwater chemistry is controlled by only a few
10.5.6 Equilibrium assemblages The rates at which minerals react (reaction kinetics) have been established by a variety of means: both
Figure 10.12: Groundwater from a long-screened interval may contain waters of different quality. The use of piezometers is preferred for hydrochemical sampling (after Shand et al. 2007a).
Surface
Oxidising high NO3 high Fe, Mn
Water table
Piezometers 1
Piezometers 2
Abstraction borehole
Reducing low NO3 high Fe, Mn
Regolith and water
% conductivity change from last measurement % pH change from last measurement
30 20
6 4
10
2
0
0
pH change %
conductivity change %
Table 10.6: Approximate time calculated for a hypothetical sphere of 1 mm diameter to dissolve in a dilute solution at pH 5 (modified from Lasaga et al. 1994).
8
40
-2
-10
-4
-20 0
4
8
12
16
20
number of casings purged Figure 10.13: Stabilising groundwaters before sampling. Acceptable limits are shown by the heavy dashed lines. Standard procedure calls for three casings to be purged. In this instance, up to 12 casings of water needed to be purged before the system stabilised.
relatively reactive minerals such as carbonates and evaporite minerals (gypsum, halite). The relative proportions of the dissolving mineral phases may also change if the more soluble phases become depleted from parts of an aquifer (some carbonate-bearing aquifers are already de-calcified at shallow depth – leading to problems of acidification). If a rock contains a mere 1% of calcite or evaporite minerals, their dissolution will control groundwater chemistry. Table 10.6 also indicates that silicate aquifers should contain groundwaters low in TDS unless minor reactive mineral phases are present. Such areas are likely to be sensitive to environmental change such as acidification related to acid rain.
10.6 HYDROGEOCHEMICAL PROCESSES IN THE REGOLITH 10.6.1 Rainfall inputs The primary source of groundwater is rain or snow, which generally contains less than 10 mg L –1 dissolved material. Most of the naturally derived solutes in rainfall are derived from sea spray incorporated into the atmosphere over the oceans (Section 10.4.1). In arid zones, many non-marine solutes are derived from the dissolution of dust particles. Although most solutes in groundwater are derived predominantly from mineral reactions in the sub-surface, some do not form major constituents of rocks and minerals and so initial
Mineral
Lifetime (years)
Quartz
34 000 000
Kaolinite
6 000 000
Muscovite
2 600 000
Epidote
923 000
Microcline
921 000
Albite
575 000
Sanidine
291 000
Gibbsite
276 000
Enstatite
10 100
Diopside
6800
Forsterite
2300
Anorthite
112
Dolomite
1.6
Calcite
0.1
Gypsum
Days
Halite
Seconds
atmospheric inputs can provide a significant component of the total input. This is particularly the case for some of the halogen elements (such as I, Cl and Br). Rainfall may undergo evaporation or transpiration, which has the effect of increasing the concentration of most solutes. The high salinities of surface waters and groundwaters in arid regions (such as Australia) are ultimately derived from atmospheric inputs as a consequence of long-term high evaporation and transpiration rates. Rainwater is a particularly good solvent because of its acidity, which is due to the incorporation of atmospheric CO2, which dissolves to form carbonic acid: H 2 O + CO 2 _ g i = H 2 CO 3
(Eqn 10.7)
Rainwater in equilibrium with the atmosphere has a pH around 5.7. Rainfall containing strong acids (acid rain) from industrial pollutants may, however, be significantly lower (pH = 2–5) (Norton and Veselý 2005). 10.6.2 Soil processes The shallowest parts of the regolith are often covered by soil: unconsolidated materials at, or near, the
265
266
Regolith Science
Earth’s surface, which are composed of mineral and organic components capable of supporting plant growth. Soils are formed in response to many environmental factors, such as climate, biological organisms, topography and water availability. They typically display layers or horizons (Figure 6.3) as a result of pedogenic processes, including additions of materials such as organic material and dust, dissolution and precipitation of minerals and organic matter, mineral transformations and physical translocations of materials (see Section 6.2 and Chapter 8). Slightly acidic rainfall undergoes significant changes in the soil zone in terms of pH. The amount of CO2 (g) (quantified as the partial pressure of CO2) typically increases 10–100×, due to root respiration and organic matter decomposition (and thus organic acid production: see Section 8.2.4) lowering pH to 4–5 (or even less in peat). Soil water is critical to the transport of nutrients from the labile (exchangeable cations and anions, mineralisable organic matter) pool of elements to plants and microorganisms. A number of elements are considered as essential to plant growth and can be classed as macronutrients or micronutrients (Table 10.7; see also Chapters 7 and 8). However, excesses of some elements (as well as deficiencies) may also limit plant growth due to toxicity effects. Because many nutrients exist in solution as charged species, ion-exchange and adsorption-desorption reactions play a dominant role in controlling bioavailability. Soil pH thus plays an important role in sustaining nutrient status and availability in the shallow regolith. At low and high pH values, elements such as Al, which are toxic to plants at high concentration, become soluble. Other metals, such as Cd, Cu, Ni and Zn, form cations that are typically bound to clays and organic matter at higher pH, but become more labile at low pH. Soils with high pH (> 9) often contain high concentrations of Na, which may be detrimental to soil structure and limit many micronutrients. The optimum pH for plant growth is therefore typically considered to be around 6–7. Other potentially toxic elements are redox sensitive. Thus, Cr in its oxidised (+6) form is highly mobile and toxic but, fortunately, even small amounts of organic matter limit Cr mobility by adsorption.
Table 10.7:
Nutrients present in soils.
Macronutrients
Micronutrients
C, H, O, N, P, K, S, Ca, Mg
B, Cl, Co, Cu, Fe, Mn, Mo, Ni, Zn
As soil water moves towards the water table it is clear that much of the water chemistry is already established. Two of the main controls on groundwater chemistry are lithology (mineralogy) and residence time. Initial stages of soil development are typified by biogenic processes; as the profile develops, so permeabilities of soils tend to decrease and hardpans can develop, particularly in argillaceous (fine-grained) horizons where evapotranspiration causes a limit to water loss and hence secondary mineral precipitation, such as carbonates and evaporates (White et al. 2005).
10.7 MAIN GEOCHEMICAL REACTIONS IN GROUNDWATERS 10.7.1 Solution–precipitation reactions Mineral dissolution reactions account for the primary geochemical characteristics of most groundwaters. As discussed previously, all minerals are soluble to a greater or lesser degree, but the kinetics of dissolution vary greatly. Many reactions are reversible (such as carbonates and sulfates) while others are for the most part irreversible (such as dissolution of silicate). Minerals may dissolve either congruently (dissolve completely) or incongruently (dissolution accompanied by formation of secondary phase). For example, gypsum and quartz dissolve congruently and the reactions are reversible: CaSO 4 .2H 2 O (s) = Ca 2 + + SO 24 - + 2H 2 O (Eqn 10.8) SiO 2 (s) + 2H 2 O = Si(OH) 04 (Eqn 10.9) Silicate minerals can also dissolve either congruently or incongruently through acid hydrolysis reactions, such as for olivine (Section 5.3):
Regolith and water
and additionally for calcite dissolution: 4. Dissolution of calcium carbonate:
MgSiO 4 (s) + 4H 2 CO 3 = 2Mg 2 + + 4HCO3 + H 4 SiO 4 (Eqn 10.10)(Eqn 5.4) Most complex aluminosilicate minerals dissolve incongruently, leaving a residue of a secondary clay or oxide mineral, such as for anorthite (Section 5.3): CaAl 2 Si 2 O g (s) + 2H 2 CO 3 + H 2 O = Ca 2 + + 2HCO3 + Al 2 Si 2 O 5 (OH) 4 (s) (Eqn 10.11 cf. Eqn 5.9) or biotite (Section 5.3): 2K (Mg 2 Fe) (AlSi 3) O 10 (OH) 2 (s) + 10H + + 0.5O 2 + 7H 2 O = 2K+ + 4Mg 2 + + 2Fe(OH) 3 (s) + 4H 4 SiO 4 + Al 2 Si 2 O 5) (OH) 4 (s) (Eqn 10.12 cf. Eqn 5.30) Most silicate reactions are strongly pH dependent (see Chapter 5.3) with minimum dissolution rates in the pH range 6–8; that is, those typical of natural waters. Clays are the typical by-products of aluminosilicate weathering reactions: they not only provide an important control on water chemistry, but they are also an indicator of climatic conditions that have led to their development (see Chapters 4 and 5). 10.7.2 Carbonate equilibrium Equilibria involving the carbonate system and calcium carbonate dissolution represent some of the most important controls on groundwater chemistry. Several reactions need to be considered for these equilibria: 1. Solution of CO2 to form carbonic acid: H 2 O + CO 2 = H 2 CO 3
(Eqn 10.13)
2. Dissociation of carbonic acid in water to form bicarbonate: H 2 CO 3 = H + + HCO-3
(Eqn 10.14)
3. Dissociation of bicarbonate to form carbonate: HCO-3 = H + + CO 23 -
(Eqn 10.15)
CaCO 3 = Ca2 + + CO 23 -
(Eqn 10.16)
Examination of these equations shows that the activity of H+ (that is, the pH) plays a dominant role in determining which species of carbonate is present. The speciation of C in the carbonate system is thus pH dependent and the relative proportions are shown in Figure 10.14. In most natural systems, the presence of HCO3 – is an important buffering control on pH and the dissolution of carbonate minerals in an aquifer (providing HCO3 –) provides a good mechanism neutralising acidic inputs, as well as minimising metal mobility, many metals having only moderate solubility at nearneutral pHs (Section 5.3.3; Figure 5.10). We deduce from Equation 10.13 that the partial pressure of CO2 controls the amount of carbonic acid in waters. Therefore, there should be a relationship between the initial pCO2 (derived from the atmosphere and soils) and the amount of carbonate dissolved at equilibrium. The evolution of groundwater is also, in a similar way, dependent on whether the system is open or closed to CO2 : in a closed system Ca and HCO3 will be much lower due to CO2 consumption, whereas in an open system CO2 will continue to dissolve in water and be available for further carbonate dissolution (Figure 10.15). As discussed previously, groundwaters typically contain higher pCO2 than the atmosphere due to soil processes during infiltration. Once these groundwaters come into contact with the atmosphere, degassing of CO2 will occur. This means a change in equilibria and, because the kinetics of the system are rapid, the effects are often seen where waters discharge or are brought to the surface by pumping. The change to a lower pCO2 will increase the pH of the solution and may induce precipitation of calcite if the waters are at, or close to, saturation: as in cave systems. The mixing of waters at equilibrium with calcite may produce counter-intuitive results: in general, if two waters are saturated with respect to calcite, the mixture is under-saturated (Figure 10.16). This effect may be important in the formation of karst terranes (see also Section 6.3.5).
267
268
Regolith Science
_
H2CO3
_
CO32
HCO3
3
80 60
Ca2+ (mN)
Percent of indicated species
100
40
B 2
s
20
m
0 5
4
6
7
8
9
10
11
12
1
A
pH Figure 10.14: Relative proportions of C-species in the carbonate system. 0
10.7.3 REDOX reactions Redox reactions can have a significant impact on groundwater chemistry as they affect the solubility and speciation of a number of elements, including dissolved oxygen, organic carbon, S, N, Fe and Mn, as well as a number of other trace elements. The oxidation of organic matter is often a significant driver of chemical changes by redox reactions. Organic matter
PC
= 1 O2 0 -1.5
500
120C Do
lom
200
=
lcit
10
no CO 2
e
added
-2
P
2 CO
=
10
-3
P 50
no CO 2
1.0
1.5
CO2 (mN) Figure 10.16: The mixing of two waters with different pCO2 and saturated with respect to calcite (A and B) may result in under-saturation and an ability of groundwater to dissolve calcite (after Langmuir 1997)
may be present in aquifers naturally as solid or dissolved organic carbon, or introduced as a surface pollutant from landfills, slurry or septic tanks. Other reductants commonly present in the regolith include Fe2+ (either in mineral or dissolved forms) and sulfide. For example, oxidation of pyrite can have an impact on groundwater chemistry – leading to potentially large increases in concentrations of dissolved Fe, SO4 and some trace elements, as well as release of protons (acidity)(see Section 5.3.1):
CO
2
=
Fe(OH) 3 (s) + 8SO 24 - + 16H +
.0
10
(Eqn 10.17 cf. Eqn 5.22)
added 5
. -3
P
2 CO
=
10
7.5
However, not all oxidation reactions involve oxygen. For example, the oxidation of pyrite can also occur through the reduction of nitrate: + 5FeS 2 (s) + 14NO3 + 4H =
20 7.0
0.5
4FeS 2 (s) + 15O 2 + 14H 2 O =
.5
100
_
HCO3 (ppm)
P
CO
2
ite
Ca
-2
0
8.0
7N 2 + 5Fe 2 + + 10SO 24 - + 2H 2 O (Eqn 10.18)
pH Figure 10.15: Dissolution of calcite at 12oC and 1 bar pressure illustrating differences in HCO3 – concentration at saturation, for open and closed systems with respect to CO2 for a range of CO2 activities. Closed systems are marked with ‘no CO2 added’ (after Langmuir 1971).
10Fe 2 + + 2NO3 + 14H 2 O = 10FeOOH + N 2 + 18H+ (Eqn 10.19)
Regolith and water
Many redox reactions are microbially mediated, and microbial catalysis can speed up many reactions that would otherwise be immeasurably slow (Chapter 7.4.2) Some of the most significant redox changes in aquifers are observed in groundwaters as they flow from unconfined to confined conditions. Redox reactions tend to follow a well-established sequence (Figure 10.17) as conditions, including those in aquifers, become anoxic (Berner 1981). This sequence begins with loss of O2, followed by loss of NO3 – by reduction to metastable NO2– and then to the gases N2O and N2. Thereafter, Mn4+ in Mn oxides reduces to soluble Mn2+, followed by reduction of Fe3+ in Fe oxides to soluble Fe2+. Under increasingly reducing conditions, SO42– is reduced to H2S, then CH4 production occurs by fermentation and methanogenesis (see Section 7.3.1). Some of the most important of these redox reactions in groundwater are listed below: + 5C organic + 4NO3 + 4H = 2N 2 (g) + 5CO 2 + 2H 2 O
Denitrification (Eqn 10.20) C organic + 4Fe (OH) 3 + 8H+ = CO 2 + 4Fe 2 + + 2H 2 O Iron3+ reduction (Eqn 10.21) C organic + SO 42 - + 2H 2 O = H 2 S + 2HCO3 Sulfate reduction (Eqn 10.22) High Oxic
Methanogenesis (Eqn 10.23) These reactions also control the environmental speciation, and hence mobility, of a number of trace elements, some of which are toxic: As3+/As5+, Cr3+/ Cr6+, Se 4+/Se6+ and U 4+/U6+. It is possible to measure the redox potential of a water either by using an Eh electrode or by analysing the concentrations of the different species of a redox couple, such as SO42– and H2S, or Fe2+ and Fe3+. There are many problems with the measurement of Eh and redox couples: as redox reactions are typically very slow and most occur only under conditions of microbial catalysis. The sluggish kinetics of these reactions makes direct Eh measurement difficult because the electrode measures a mixed potential of all the redox reactions, and these occur at different rates. Added to this is the fact that redox couples may not all be at equilibrium. The Fe2+/Fe3+ couple is probably the most important in natural waters. Reliable redox measurements may therefore be obtained where the Fe2+/Fe3+ system is dominant. Despite the problems of measurement, field Eh is inexpensive to measure compared with analysing redox couples and does provide a general sense of the redox condition and changes within the aquifer. Most redox reactions involve the transfer of H+ as well as electrons, and the stability fields of solid and/ or aqueous redox species can be displayed very effectively on pH-Eh diagrams. The range in pH and Eh in aqueous systems is limited by the stability field of
Redox Potential
Suboxic -
NO3 Concentration of species
2C organic + 2H 2 O = CO 2 + CH 4
2+
Mn
Fe
Low Reducing
2+
SO4
2-
H2S
CH4
O2 Fe2+
Reaction Progress Figure 10.17: Sequential changes in redox-sensitive species as waters move from oxidising (left) to reducing conditions. The extent of reaction progress is dependent on the initial condition of the water and the amount of reducing agents (after Shand et al. 2007a).
269
270
Regolith Science
water. Typical conditions of different waters are shown on Figure 5.7 and an example of the stability fields for Fe species is shown on Figure 10.18. It should be noted that these fields are dependent on the activities of species and components considered and can vary for different elements. The reduced form of iron, Fe2+, forms an aqueous species in reducing and/or acidic conditions. The oxidised species, Fe3+, is only stable as an aqueous species under limited conditions of pH and Eh (Figure 10.18). Over most of the plot, Fe3+ is unstable, forming Fe(OH)3 or other oxide or hydroxyl species, which are insoluble in water. The plot would therefore predict that if a reducing groundwater containing Fe2+ were to become oxidised by contact with the atmosphere or another oxidised groundwater during abstraction, then precipitation of Fe(OH)3 would occur – leading perhaps to well clogging or the requirement of treatment. Many trace metals are also adsorbed on Fe(OH) 3, and changes in redox condi-
10.7.4 Ion exchange reactions Most aquifers contain clay minerals, which hold exchangeable cations adsorbed to their surface. This is largely due to charge imbalances, which cause the surfaces of minerals to have a negative charge. For example, in clays, the replacement of Al3+ for Si4+ causes an excess negative charge of –1. Under some conditions, the cations attached to a clay mineral can be exchanged for other ions from the aqueous solution: a process termed ion exchange. This is different from sorption-desorption reactions, such as on Fe(OH)3 surfaces, where the surface charge is variable and a change in solution pH would affect the adsorption capacity of the surface. A general ordering of cation exchangeability for common ions in groundwater is: Na+ 2 K+ 2 Mg 2 + 2 Ca2 + that is, divalent cations tend to be more strongly adsorbed and replace monovalent ions (see also Figure 5.12). Exchange equilibrium can be represented by the law of mass action:
1.4 1.2 1.0
A-clay + B+ = B-clay + A+ (Eqn 10.24)
Fe3+ Fe(OH)2+
0.8
O2 H2 O
0.6
Eh (V)
tions may also lead to removal/addition of metals to the aqueous environment.
0.4
Fe2+ Fe(OH)2
0.2
H2 O H2
0 -0.2
FeCO3
-0.4
Fe(OH)2
-0.6 -0.8 0
2
4
6
pH
8
10
12 KSf020-08
Figure 10.18: Eh-pH diagram for Fe showing the stability fields for Fe and samples from a range of aquifer types in the UK (after Shand et al. 2007a).
The concept here is similar to that taking place with some domestic water softeners. The exchange complex is dominated by Na+ ions, and these become replaced by Ca2+ and Mg2+ ions present in the water supply (‘hard’ water is rich in Ca2+ and Mg2+). The exchange complex is then regenerated by a brine solution (Na+ Cl–). Ion exchange processes are important and can occur during salt-water intrusion and aquifer freshening (Figure 10.10b). The dominant waters in a fresh aquifer are typically of Ca-HCO3 type; ions adsorbed onto clay mineral surfaces are therefore dominated by Ca2+. When salt water of Na-Cl composition intrudes this aquifer, the Ca2+ on the clays is replaced by Na+: Na+ + 0.5Ca-X 2 = Na-X + 0.5Ca 2 + (Eqn 10.25) The water will then change from a Na-Cl type to a Ca-Cl type.
Regolith and water
0.5Ca 2 + + Na-X = 0.5Ca-X 2 + Na+ (Eqn 10.26) where Ca is exchanged for Na. The water in this case changes from Ca-HCO3 type to Na-HCO3 type. The direction of groundwater movement can thus be determined in such cases from the water types observed. Because most aquifers are of marine sedimentary origin, the latter process is likely to be important for a long time after the aquifer becomes part of the terrestrial environment. 10.7.5 Sorption reactions Iron, Mn and Al oxides are ubiquitous in oxidising soils and sediments – occurring as discrete grains and as coatings on mineral surfaces. Amorphous and poorly crystalline forms have large specific surface areas and a particularly large capacity to adsorb trace elements (Section 4.5.5). As such, they are commonly used to remove trace metals in wastewater treatment (for example, Edwards 1994). The adsorption capacity of these oxide minerals is pH-dependent, and such variable charge solids are able to sorb ions in solution without releasing ions in equivalent proportions. The outer surfaces of the common oxyhydroxides are oxygen and under acidic conditions they are protonated and have a positive charge. At a higher pH, the surfaces become negatively charged. The region between these pH ranges where the net surface charge is zero is called the point of zero charge (pzc; Section 5.3.7) and varies from mineral to mineral (Table 5.3). Fe and Al oxides are therefore particularly effective anion adsorbants at low pH and can have a marked impact on the speciation and mobility of elements such as As, Mo, P, Se and U. The amount of cation adsorption at pH higher than the zpc is dependent on pH and the concentration in solution, and is different for different metals. As an example, the sorption of selected trace elements onto ferrihydrite is shown on Figure 10.19. Changes in pH or Eh can have a profound influence on the stability of oxyhydroxide species (Figure 10.18), as well as the sorbed species; hence, changes in pH-Eh conditions can lead to loss or gain of dissolved metals in the environment.
100
mol % bound
The reverse process takes place in a freshening aquifer:
80 60
Cr3+
40
Cu2+
Pb2+
Cd2+
20 0
Ni2+
3
4
5
pH
Zn2+ 6
7
Ca2+
8
Figure 10.19: Sorption of trace metals onto the surface of ferrihydrite as a function of pH (after Appelo and Postma 2005).
The extent of adsorption can be described by a distribution (Kd) or partition (Kp) coefficient (Eqn 10.27): Kd =
C ads C aq
(Eqn 10.27)
where Cads is the adsorbed concentration and Caq the dissolved concentration of the element of interest. The Kd is an empirically measured constant. A plot of dissolved versus adsorbed concentration is termed an isotherm because it is measured at a fixed temperature. The data would plot as a straight line for the above equation and is termed a linear sorption isotherm (Figure 10.20a). This implies that the surface has unlimited capacity to adsorb solutes and that solute concentration does not affect adsorption. This may be the case at low solute concentrations, but at high concentrations these assumptions may not be valid and other isotherms are more realistic at describing or predicting trace element behaviour (Freundlich and Langmuir isotherms: Figure 10.20b, c). The Freundlich isotherm becomes a curve at higher concentrations – reflecting lower adsorption as sites become filled and the Langmuir flattens out reflecting the capacity of sorbent.
10.8 REGOLITH AS A DYNAMIC MEDIUM Weathering rates are dependent on a number of factors that can be classified as either intrinsic or extrinsic (White 2005a). Intrinsic factors are those inherent in the minerals themselves, such as composition, defects
271
Regolith Science
Kd = caq (mg l-1)
Cads (mg kg-1)
Cads (mg kg-1)
Cads Caq
(c) Langmuir isotherm KL
(b) Freundlich isotherm KF
(a) Linear sorption isotherm Kd Cads (mg kg-1)
272
Cads = KF C1/n aq caq (mg l-1)
Adsorption capacity Am
Cads =
KL Am Caq 1 + KL Caq
caq (mg l-1)
Figure 10.20: (a) Linear, (b) Freundlich and (c) Langmuir isotherms showing the relationships between dissolved and adsorbed element concentrations.
and surface area, while extrinsic factors are those indicative of the environment of weathering, such as temperature, solution composition and biological activity. The composition of the solute has the most direct extrinsic impact on weathering rates. This has been well documented in experimental studies (for example, Blum and Stillings 1995), but not so in natural systems (for example, Driscoll et al. 1989). Thus, the concentration of dissolved silica in groundwater is commonly used as an indicator of weathering, but this effect may be masked by other factors – even in situations where increased weathering rates are expected (Driscoll et al. 1989). See Section 6.8 for a discussion of evaluating the degree of weathering. Weathering rates generally slow as saturation of the dissolving species is approached. This, in turn, is controlled by the volume and residence times of the fluid moving through the weathering environment. The roles of hydrology and chemical weathering are commonly coupled in a weathering regolith and can be viewed as co-evolving as the intensity of weathering increases (White 2005b). The classic example of differential weathering of minerals is that of granite weathering to clays and residual quartz (see also Sections 4.3 and 6.3.1). Thus, White et al. (2001) modelled the coupled relationship between weathering rates and the development of secondary permeability during the kaolinisation of the Panola granite. Initially, the tight crystalline structure of the granite precludes effective water transport, and weathering is limited by the rate at which water can enter the system. As weathering progresses, however, plagioclase converts to kaolinite (with the dissolution and removal of sodium and silica in solution), and
porosity increases by up to 50% (White et al. 2001). This is accompanied by an increase in the flux of water through the regolith and the system shifts from a transport-limited reaction to a kinetic-limited one. A complete shift from transport to kinetic constraint occurs when the new flux is around 150 times greater than the original flux through the crystalline rocks. Meanwhile, weathering of potassic feldspars in the same rock did not initiate until almost all of the plagioclase has reacted: at around 50 times the initial water flux (White et al. 2001). This is due to the lower solubility, not reaction kinetics – an effect further enhanced by the dissolution of silica from the plagioclase, which inhibits K-feldspar dissolution by increasing the saturation state (White 2005b). Further weathering, however, often results in a decrease in permeability due to in situ secondary mineral formation and development of hard pans and clay horizons (White 2005b). This process may be enhanced by physical translocation and collapse of regolith structures (Torrent and Nettleton 1978). These secondary clay horizons commonly correspond to the maximum depth of effective evaporation, and thus occur within the top 2–3 m of the profile (Section 6.3.1). Water loss from evapotranspiration initiates precipitation of secondary clays and oxides (aided by any loss of dissolved organic species and de-complexation of soluble Al) and the resulting lower permeabilities further retard downward movement of water – leading to perched water tables. Periodic drying in this zone focuses additional secondary mineral precipitation – particularly halides and sulfates – which then leads to a lower permeability and further clay formation (White 2005b).
Regolith and water
Disentangling the role of weather, or climate, on the development of the regolith still proves to be problematic. This is particularly the case as the scale of observation increases, because climate (and in particular the amount of rainfall) is also responsible for other processes that enhance weathering and regolith development. Thus, increased rainfall also leads to an increase in physical erosion and exposure of new material for weathering, and can also lead to increased vegetation and biogenic effects on weathering and regolith formation. In summary, it is the amount of water that is available – either as a medium for chemical reactions or as a solvent in its own right, or through its ability to physically transport material between regions, or its effect on the growth of biomass – that ultimately determines how the regolith is formed, transformed and, eventually, removed.
10.9 ENVIRONMENTAL ISOTOPIC TRACERS Environmental isotopes are naturally occurring (or, in some cases, anthropogenically produced) isotopes, whose distributions in the hydrosphere can assist in the solution of hydrological and biogeochemical problems (Kendall and Doctor 2005). A wide variety of isotopic tracers are used by hydrogeologists to trace flow pathways and chemical evolution, and to provide age constraints on the water. The main environmental isotopes routinely analysed include: carbon (Vogel 1970; Fontes and Garnier 1979), oxygen, hydrogen (Dansgaard 1964); nitrogen (Heaton 1986; Mariotti et al. 1988) and sulfur (Taylor et al. 1984; Krouse and Mayer 2000). Less commonly used are the isotopes of: chlorine (Bentley et al. 1986; Cook et al. 1994; Cresswell et al. 1999a, b), strontium (Capo et al. 1998; Shand et al. 2007b) and noble elements (Loosli et al. 2000), among others. Good summaries of the uses of these isotopes in hydrogeologic studies can be found in Clark and Fritz 1997, Kendall and McDonnell 1998 and Cook and Herczeg 2000. Environmental isotopes are characterised by being present in sufficient natural abundance to be easily measured and are generally low-mass elements, which hence have a significant, and measurable, mass difference between different isotopes. Variations in the
concentrations of the different isotopes, relative to each other, derive from; 1. fractionation, or partitioning, processes that result from the difference in their masses 2. changes in concentration of radiogenic isotopes due to radioactive decay. Importantly, these processes can be mathematically described using formulae developed to describe the separation of gases during distillation (Rayleigh 1896), and can be applied to any isotopic system where kinetic processes dominate – such as evaporation from surface water bodies and precipitation of rain from air masses – where one substance is continuously removed from an open system under condition of a constant fractionation factor. Stable isotopes of water (2H/1H, 18O/16O) have been used to evaluate primary sources of water to the regolith (Kendall and Doctor 2005) as well as to determine relative importance of diffuse and direct recharge (Leaney and Herczeg 1995). Kendall and Doctor (2005) also report on methodologies using stable isotopes to evaluate surface water mixing, to separate surface and groundwater components in run-off and outline an approach to qualitatively estimate the mean residence time of sub-surface waters. The stable isotopic signature of rainfall is a function of six effects: altitude, latitude, continentality, seasonality, rainfall amount and temperature. Good summaries of these effects may be found in Clark and Fritz (1997) and Coplen et al. (2000). These effects can have competing, or enhancing, effects on the isotopic signature of precipitation, so all must be carefully evaluated for trends and processes to be discerned. Isotopes of carbon (12C, 13C and radiogenic 14C) have been used to unravel soil processes and the fate of carbon in the atmosphere, biosphere, hydrosphere and lithosphere (Clark and Fritz 1997). Dissolved inorganic carbon (DIC) is a measure of the ‘reactivity’ of the regolith, and reflects the neutralisation of the carbonic and other acids by reactions with silicate and carbonate minerals as water moves through the sub-surface (Garrels and Mackenzie 1971). Carbon isotopes of DIC can help understand the biogenic processes that control its alkalinity. Dissolved organic carbon (DOC) primarily comprises humic and fulvic acids and hence contributes to the acidity of waters. The carbon isotopes of
273
274
Regolith Science
DOC therefore provide insights into primary nutrient loads and hydrologic processes that affect the fluxes of DOC in the carbon cycle (Kendall and Doctor 2005). The 15N/14N of soil nitrogen is affected by many factors, including: soil depth, vegetation, climate, particle size and land use history. Drainage and the influence of land cover are by far the most important (Shearer and Kohl 1986). Both are influential in determining the nature of the nitrogen species present and its relative mobility in the regolith. The input of anthropogenic sources (such as fertilisers and animal wastes) and subsequent biogeochemical reactions (denitrification, ammonia volatilisation and nitrification) can significantly modify the d15N values before and during mixing of different sources. The combined use of d15N with d18O and comparison of different nitrogen species (N2, NO, NO3) (Böhlke and Denver 1995), combined with modelling (Rassam et al. 2006) allows resolution of these sources and processes. Weathering of sulfide and sulfate minerals does not significantly fractionate the sulfur isotopes (34S/32S), and precipitation of sulfate minerals generates only a slight enrichment of the heavy isotope. (This feature can be used to identify secondary sulfates derived from sulfide ores: Section 13.4) The primary fractionation process is via biologically mediated sulfur transformations – particularly the breakdown of sulfate to sulfuric acid by anaerobic bacteria (Mitchell et al. 1998). This process leads to a progressive enrichment of 34S of the dissolved sulfate in the water. This is accompanied by oxidation of organic material – leading to a concomitant shift in d13C to that of the organic source due to production of CO2 (Clark and Fritz 1997). Sulfur isotopes can also be used to identify water sources because the sulfur in the dissolved sulfate may be derived from parent evaporites, sulfides and organic material or from atmospheric deposition – all of which have different d34S signatures (Hoefs 1987).
10.10
HYDROCHRONOLOGY
Time is a vital component of any study of the regolith, and dating the waters it contains is as important as dating other regolith material. (Discussion of isotopic
methods for dating solid phases occurs in Chapter 2.) Water flow rates and patterns vary greatly between, and within, aquifers as a result of variations in porosity and permeability, aquifer dimensions, recharge and discharge areas, and the degree of confinement and abstraction. Rates of groundwater flow are typically: 1–500 m/year for granular aquifers; much slower for finer-grained materials; and essentially zero for waters trapped within tight clays or re-crystallised hard pans or isolated pores. These relatively low rates of flow result in residence times for groundwaters of decades, centuries or millennia: hence, considerably greater than for surface waters. As a result, changes in groundwater chemistry tend to be slower, and the effects of chemical changes remain for much longer periods. Several dating techniques exist to determine the ‘age’ of groundwater, but the large range in residence times (days to millennia) in most aquifers means that several different techniques are required to assess the age distribution within an aquifer (Clark and Fritz 1997) (Figure 10.21). The term ‘age’ is misleading, however, for two main reasons: 1. Of the dating tools used, only tritium is part of the water molecule, and can thus, ‘date’ the water, and this has a short range of applicability (Figure 10.21). All other dating methods rely on tracers of the solutes in the water, whose abundance is controlled by physicochemical and biochemical processes, which may, or may not, be ideally controlled or understood. 2. Convergence and mixing of waters from different sources, with different ages means that a date is an integration of the ages of the mixing waters. Only in well-defined (and commonly regional artesian aquifers) will age gradients along a flow path be preserved. Either complete characterisation of the system, including assumptions about the geochemical evolution of the waters is required, or knowledge of the potential end-member, input and sources must be obtained. Either way, the term ‘mean residence time’ (MRT) is preferable as a descriptor, despite the general usage of the shortened term: ‘age’ (Clark and Fritz 1997).
Regolith and water
222
Rn 3
He/4He
T, 85Kr, CFC, SF6 39
Ar 14
C
“Bomb pulse”
36
Cl, 81Kr 4
He
101
102
103
104
105
106
107
Dating range (years) Figure 10.21: A range of isotope and tracer techniques is required to date waters of different ages. Note the ability to date very young waters with isotopes generated from the nuclear bomb test, but note that this is (hopefully) a one-time event: the significance and applicability of which will wane with time. Also, CFC and SF6 dating relies on the increasing concentration in the atmosphere from anthropogenic sources. The applicability of these methods relies on knowledge of the input to the system, and restricts their range to when this is known and was changing. Thus, the moratorium on CFC use effectively restricted this system to the period 1955 to 1980.
10.11 WATER AND NATURAL RESOURCE MANAGEMENT In Australia, the majority of bores in use are constructed within shallow aquifers. This means that the majority are constructed in the regolith and their water is maintained largely through recharge that is local and diffuse. Thus, their compositions readily indicate their source – a feature used during mineral exploration (Section 13.13). The nature of the regolith also determines how fast the system can recover and how much water it can hold and release in a given time. Deeper, sometimes confined, systems may also reside in regolith materials, such as the, so-called, ‘deep leads’ of the Murray–Darling Basin in SouthEast Australia. Recharge may be at some distance from the point of extraction. Here the regolith determines the rate of transport of recharging waters to the bore and will control the path it takes to get there. Now that we have control over the amount of water being extracted from an aquifer, there is a greater need to understand the capabilities of the aquifers and the properties of any materials that water may be in contact with, as it winds its way through them. Under-
standing the ability of aquifers to release the water they hold allows us to moderate our extraction to a sustainable level. Rather than relying on gravity discharge to service our needs – as was the case for the qanat-reliant systems (see Box 10.1) – we now rely on technology to enhance yields and supplies and maintain a constant supply as our needs expand. As our needs increase with increasing stress on our water supplies, it becomes more crucial to understand not just where water is, but how much water there is and how fast it can be released. We can then decide whether we wish to extract water sustainably, or whether we are prepared to mine water faster than it is replenished. A further consideration is water quality. While aquifers in deeper, consolidated sediments, such as the Great Artesian Basin (GAB) aquifers, are commonly relatively fresh and clean (although, here, trace element concentrations can be important: F in many GAB waters make it undesirable; the elevated Na concentrations generally make GAB waters unsuitable for irrigation, despite the low salinities), shallow, commonly unconfined, or open, systems in unconsolidated sediments or weathered aquifers can exhibit variable and often high levels of solutes – particularly salts – that make them unusable. Water in unconsolidated deposits, or passing through weathered strata, has the facility to dissolve and/or exchange significant quantities of elements. Weathering releases these elements into solution, while the large surface areas afforded by unconsolidated materials enhance waterrock interactions. Environmental conditions in the near surface are also commonly more extreme than at depth – specifically with regard to temperature, acidity and oxygen availability. This further enhances the rates and propensity of ions to exchange, dissolve or precipitate. Anthropogenic sources of pollutants, such as fertiliser by-products, pesticides and herbicides, can also reach unacceptable levels in these systems that are commonly in connection with the surface and the atmosphere. Furthermore, the proportion of solutes found in surface waters and stores, including the oceans, is largely a function of the regolith through which shallow and deep ground waters pass. The inherent variability in regolith materials shows its most
275
276
Regolith Science
Box 10.1
Qanats: Providing water to a parched land
Long, underground tunnels, which are built into the sides of valleys, tap the shallow water tables that exist beneath the colluvial aprons of the Middle East, northern Africa and across into Asia. Providing water to irrigate cotton, fruits and oilseeds in country similar to Central Australia, these qanats today support nearly three-quarters of Iran’s population. Over 300 000 km of low-gradient, narrow tunnels tap unconfined aquifers perched on clay lenses in the colluvial fans of the dry valleys. Roughly 20 000 of these systems exist – with some up to 30 km long – providing an almost continuous water supply, sometimes up to 250 L/s, for the fertile alluvial soils of the valley floors (Wulff 1968). The specialist surveyor – hired to locate the water source, design and oversee construction of the tunnels and ensure water flow to villages and fields – was thus one of the most important and revered members of society. He was highly paid for his services – and he was a regolith scientist! Understanding how water moved through the landscape, how the landscape influenced its quality and quantity, how the juxtaposition of particular beds and materials controlled both transport of
pronounced expression in low-order streams (Miller 1961; Meybeck 1986), particularly in association with shallow regolith materials. Hence, stream chemistry can be used to indicate regolith type and characteristics. This has proved particularly useful when targeting catchments for salinity mitigation. Stream salinity surveys isolate the sub-catchments that are contributing to increasing down-stream salinity. Globally, from a representative sample of 1200 pristine and sub-pristine rivers and tributaries, over 89% can be classed as weathering-dominated (with 8% evaporation-controlled and the remainder rainfall-dominated) (Meybeck 2005), which underlines the prominent role of the regolith in global stream chemistry. Thus regolith plays a crucial, vital and dynamic role in the hydrologic cycle. It not only acts as a medium for water transport at the interface of surface and deep groundwater flow, but also provides a
water and the state of the tunnels was vital if an adequate supply was to sustain the communities the qanat served. Exploratory shafts were dug – up to 80 m deep – to expose the stratigraphy and locate suitable water reserves (that had to be underlain by suitably impervious clays); access and maintenance shafts were dug to allow construction of the tunnels; and clay hoops were used to bolster sections of the tunnels where friable materials threatened their integrity. Qanat tunnels generally had gradients of 1:500 to 1:1500. They are gravity-driven and so self-regulating. In dry periods they may cease to flow; in wetter periods, sluice gates may be installed to create artificial weirs and dams in the system to store water. The skill of the surveyor dictates whether a sustainable supply will be realised. This simple system is found across the globe. Persia appears to be the birthplace – with the technology spreading across northern Africa and southern Europe (and from there to Mexico – although there is evidence of similar systems in use in Chile and Peru before the Spanish conquests on the east coasts) and following the Silk Route to Afghanistan, Pakistan and on to China.
vibrant chemical environment for an array of chemical reactions and exchanges. It is, therefore, both a passive and dynamic medium: representing a complex interface between land, air and water as well as playing a dominant role in biogeochemical cycling. Indeed, the qanats of Iran, while providing life-giving water to the desert, are also known to have been major conduits for pathogens and diseases: perhaps a macroscopic analogy for stygofauna mobility?
10.12
REFERENCES
Appelo C and Postma D (2005). Geochemistry, Groundwater and Pollution. A.A. Balkema, Amsterdam. Berner R (1981). A new geochemical classification of sedimentary environments. Journal of Sedimentary Petrology 51, 359–365. Bentley HW, Phillips FM and Davis SN (1986). 36Cl in the terrestrial environment. In Handbook of Envi-
Regolith and water
ronmental Isotope Geochemistry 2B. (Eds P Fritz and J-C Fontes) pp. 427–480. Elsevier, Amsterdam. Bethke C (1996). Geochemical Reaction Modelling: Concepts and Applications. Oxford University Press, Oxford. Bland W and Rolls D (1998). Weathering: An Introduction to Scientific Principles. Arnold, London. Blum AE and Stillings LL (1995). Feldspar dissolution kinetics. Reviews in Mineralogy 31, 290–351. Böhlke J and Denver J (1995). Combined use of ground-water dating, chemical, and isotopic analyses to resolve the history and fate of nitrate contamination in two agricultural watersheds, Atlantic coastal plain, Maryland. Water Resources Research 31, 2319–2339. Budyko M (1982). The Earth’s Climate: Past and Future. Academic Press, Oxford, UK. Capo R, Stewart B and Chadwick O (1998). Strontium isotopes as tracers of ecosystem processes: theory and methods. Geoderma 82, 197–225. Carson MA (1969). Soil moisture. In Introduction to Physical Hydrology. (Ed. RJ Chorley) pp. 98–108. Methuen and Co Ltd, London. Clark I and Fritz P (1997). Environmental Isotopes in Hydrogeology. CRC Press Inc, Boca Raton, Florida. Cook PG and Herczeg AL (Eds) (2000). Environmental tracers in Subsurface Hydrology. Kluwer Academic Publishers, Norwell, Massachusetts. Cook P, Jolly I, Leaney F, Walker G, Allen G, Fifield LK and Allison G (1994). Unsaturated zone tritium and Cl-36 profiles from southern Australia: their use as tracers of soil water movement. Water Resources Research 30, 1709–1719. Cooper HH Jr (1966). The equation of groundwater flow in fixed and deforming coordinates. Journal of Geophysical Research 71, 4785–4790. Coplen TB, Herczeg AL and Barnes C (2000). Isotope engineering – using stable isotopes of the water molecule to solve practical problems. In Environmental Tracers in Subsurface Hydrology. (Eds PG Cook and AL Herczeg) pp. 79–110. Kluwer Academic Publishers, Norwell, Massachusetts. Coram JE (1998). ‘National classification of catchments for land and river salinity control: A catalogue of groundwater systems responsible for dryland salinity in Australia’. Publication No
98/78. Rural Industries Research and Development Corporation, Canberra. Coram JE, Dyson PR, Houlder PA and Evans WR (2000). ‘Australian groundwater flow systems contributing to dryland salinity’. Report for the National Land and Water Resources Audit. Bureau of Rural Sciences, Canberra. Cresswell RG, Jacobson G, Wischusen J and Fifield LK (1999a). Ancient groundwaters in the Amadeus Basin, Central Australia: evidence from the radioisotope Cl-36. Journal of Hydrology 223, 212–220. Cresswell RG, Wischusen J, Jacobson G and Fifield LK (1999b). Assessment of recharge to groundwater systems in the arid southwestern part of Northern Territory, Australia, using chlorine-36. Hydrogeology Journal 7, 393–404. Dansgaard W (1964). Stable isotopes in precipitation. Tellus 5, 461–469. Darcy H (1856). Les Fontaines Publiques de la Ville de Dijon. Victor Dalmont, Paris. Domenico P and Mifflin M (1965). Water from lowpermeability sediments and land subsidence. Water Resources Research 1, 563–576. Drever J (1997). The Geochemistry of Natural Waters. Prentice-Hall, New Jersey. Driscoll CT, Likens GE, Hedin LO, Eaton JS and Bormann FH (1989). Changes in the chemistry of surface waters. Environmental Science and Technology 23, 137–143. Edwards M (1994). Chemistry of arsenic removal during coagulation and Fe-Mn oxidation. Journal American Water Works Association 86, 64–78. Fontes J-C and Garnier JM (1979). Determination of the initial 14C activity of the total dissolved inorganic carbon: A review of existing models and a new approach. Water Resources Research 15, 39–413. Freeze R and Cherry J (1979). Groundwater. PrenticeHall, Inglewood Cliffs, New Jersey. Garrels R and Mackenzie F (1971). Evolution of Sedimentary Rocks. Norton, New York. Heaton THE (1986). Isotopic studies of nitrogen pollution in the hydrosphere and atmosphere. Chemical Geology 5, 87–102. Hem JD (1992). ‘Study and interpretation of the chemical characteristics of natural water. Third
277
278
Regolith Science
Edition’. US Geological Survey Water-Supply Paper 2254. United States Government Printing Office, Washington DC. Hoefs J (1987). Stable Isotope Geochemistry. 3rd edn. Springer-Verlag, Berlin. Hsieh P (1998). Scale effects in fluid flow through fractured geologic media. In Scale Dependence and Scale Invariance in Hydrology. (Ed. GT Sposito) pp. 335–353. Cambridge University Press, New York. Hubbert M (1940). The theory of groundwater in motion. Journal of Geology 48, 785–944. Ingebritsen S, Sanford W and Neuzil C (2006). Groundwater in Geologic Processes. Cambridge University Press, Cambridge, UK. Johnson A, Moston R and Morris D (1968). ‘Physical and hydrologic properties of water-bearing deposits in subsiding areas in central California’. US Geological Survey Professional Paper 497-A. US Geological Survey, Washington DC. Kendall C and Doctor DH (2005). Stable isotope applications in hydrologic studies. In Surface and Ground Water, Weathering and Soils. Treatise on Geochemistry 5. (Ed. JI Drever) pp. 319–364. Elsevier-Pergamon, Oxford. Kendall C and McDonnell J (Eds) (1998). Isotope Tracers in Catchment Hydrology. Elsevier, Amsterdam. Krouse H and Mayer B (2000). Sulphur and oxygen isotopes in sulphate. In Environmental Tracers in Subsurface Hydrology. (Eds PG Cook and AL Herczeg) pp. 195–232. Kluwer Academic Publishers, Norwell, Massachusetts. Langmuir D (1971). The geochemistry of some carbonate groundwaters in central Pennsylvania. Geochimica et Cosmochimica Acta 35, 1023–1045. Langmuir D (1997). Aqueous Environmental Geochemistry. Prentice-Hall, Upper Saddle River, New Jersey. Lasaga AC, Soler JM, Ganor J, Burch TE and Nagy KL (1994). Chemical weathering rate laws and global geochemical cycles. Geochimica et Cosmochimica Acta 58, 2361–2386. Leaney FW and Herczeg A (1995). Regional recharge to a karst aquifer derived from chemistry and isotopes. Journal of Hydrology 164, 363–387. Loosli H, Lehmann BE and Smethie WJ (2000). Noble gas radioisotopes: 37Ar, 85Kr, 39Ar, 81Kr. In Environ-
mental Tracers in Subsurface Hydrology. (Eds PG Cook and AL Herczeg) pp. 379–396. Kluwer Academic Publishers, Norwell, Massachusetts. Mariotti A, Landrau A and Simon B (1988). 15N isotope biogeochemistry and natural denitrification process in groundwater: application to the chalk aquifer of northern France. Geochimica et Cosmochimica Acta 52, 1869–1878. McDonald MG and Harburgh AW (1988). A Modular Three-Dimensional Finite-Difference Ground-water Flow Model. USGS–TWRI Book 6 Chapter A1, United States Geological Survey, Reston Virginia. Meybeck M (1986). Composition chemique des ruisseaux non pollués en France. Sciences Geologique Bulletin 39, 3–77. Meybeck M (2005). Global occurence of major elements in rivers. In Surface and Ground Water, Weathering and Soils. Treatise on Geochemistry 5. (Ed. JI Drever) pp. 207–223. Elsevier-Pergamon, Oxford. Miller JP (1961). ‘Solutes in small streams draining single rock types, Sangre de Cristo Range, New Mexico’. US Geological Survey Water Supply Paper 1535-F. United States Geological Survey, Reston Virginia. Mitchell MJ, Krouse HR, Mayer B, Stam AC and Zhang Y (1998). Use of stable isotopes in evaluating sulfur biogeochemistry of forest ecosystems. In Isotope Tracers in Catchment Hydrology. (Eds C Kendall and J McDonnell) pp. 489–518. Elsevier, Amsterdam. Norton SA and Veselý J (2005). Acidification and acid rain. In Environmental Geochemistry. Treatise on Geochemistry 9. (Ed. B Sherwood Loller) pp. 367– 406. Elsevier-Pergamon, Oxford. Ollier CD (1984) Weathering. Longman, London. Rassam DW, Fellows CRD, Hunter H and Bloesch P (2006). The hydrology of riparian buffer zones; two case studies in an ephemeral and a perennial stream. Journal of Hydrology 325, 308–234. Rayleigh JWS (1896). Theoretical considerations respecting the separation of gases by diffusion and similar processes. Philosophical Magazine 42, 493–498. Rose C (2004). An Introduction to the Environmental Physics of Soil, Water and Watersheds. Cambridge University Press, Cambridge, UK.
Regolith and water
Shand P, Edmunds WM, Lawrence AR, Smedley PL and Burke S (2007a). ‘The natural (baseline) quality of aquifers in England and Wales’. BGS Report. British Geological Survey, London. Shand P, Darbyshire DPF, Gooddy DC and Harice AH (2007b). 87Sr/86Sr as an indicator of flow paths and weathering rates in the Plynlimon experimental catchments, Wales, UK. Chemical Geology 236, 247–265. Shearer G and Kohl D (1986). N2 fixation in field settings, estimations based on natural 15N abundance. Australian Journal of Plant Physiology 13, 699–757. Taylor B, Wheeler M and Nordstrom D (1984). Stable isotope geochemistry of acid mine drainage: experimental oxidation of pyrite. Geochimica et Cosmochimica Acta 48, 2669–2678. Thomasson AJ and Youngs EG (1975). Water Movement in Soil. Soil Physical Conditions and Crop Production. Her Majesty’s Stationery Office, London. Torrent J and Nettleton W (1978). Feedback processes in soil genesis. Geoderma 20, 281–287. Vogel JC (1970). Carbon-14 dating in groundwater. In Isotope Hydrology 1970, IAEA Symposium 129. pp. 225–239, International Atomic Energy Agency, Vienna. Walker G, Gilfedder M, Evans R, Dyson P and Stuaffacher M (2003). ‘Groundwater Flow Systems Framework – Essential Tools for Planning Salinity Management’. MDBC publication 14/03. ISBN 1 876830 63 8. Weast R (1985). Handbook of Chemistry and Physics. CRC Press Inc, Boca Raton, Florida.
White AF (2005a). Extrinsic versus intrinsic controls on rates of silicate weathering and CO2 drawdown. Geochimica et Cosmochimica Acta 69, A724. White A (2005b). Natural weathering rates of silicate minerals. In Surface and Groundwater, Weathering and Soils. Treatise on Geochemistry 5. (Ed. JI Drever) pp. 133–168. Elsevier-Pergamon, Oxford. White A, Blum A, Stonestrom D, Bullen T, Schulz M, Huntington T and Peters N (2001). Differential rates of feldspar weathering in granite regoliths. Geochimica et Cosmochimica Acta 65, 847–869. White AF, Schulz MS, Vivit DV, Blum AE, Stonestrom DA and Harden JW (2005). Chemical weathering rates of a soil chronosequence on granitic alluvium: III. Hydrochemical evolution and contemporary solute fluxes and rates. Geochimica et Cosmochimica Acta 69, 1975–1996. Wilford J, James J, Halas L and Roberts L (2007). ‘Regolith hydrogeomorphic units and bedrock features within the Bet Bet Catchment area, Victoria: value-adding GFS and hydrogeological models for salinity management’. CRC LEME Restricted Report 231R. CRC LEME, Perth. Wilford J, James JM and Halas L (2006). Advancing GFS in upland regions: new approaches for old landscapes. In Abstracts, 10th Murray-Darling Basin Groundwater Workshop. September 06, Canberra. CD, Murray–Darling Basin Commission, Canberra. Wulff HE (1968). The qanats of Iran. Scientific American April, 94–105.
279
11
Regolith description and mapping Colin F Pain
11.1 INTRODUCTION Regolith–landform mapping is important in Australia because rocks crop out in less than 15% of the 7.7 million km2 of the Australian land area, with the rest being covered by regolith of variable thicknesses. In situ weathering mantles – particularly in the western part of the continent – can be several hundred metres thick. Moreover, much of the continent has been exposed to weathering and erosion since at least the Permian (see Chapters 2 and 3). The surface of the Earth, and other planets (see Chapter 14), is covered by regolith of various ages and thicknesses. This chapter reviews some methods that have been used for regolith mapping, and stresses a number of important points: 1. Regolith units differ from bedrock geology units, so geological mapping techniques are not appropriate. 2. The use of landscape as a primary indicator of soil and regolith characteristics is not new, but has a different emphasis from bedrock geological mapping, which uses landforms and soils to map bedrock structure and to identify bedrock units. 3. A landscape approach can be used for all scales of regolith mapping, but with differing emphases.
4. Regional models of landscape evolution, which are based firmly on factual observations, are a useful approach to extrapolation of regolith characteristics at most scales of mapping. It also considers the presentation of regolith information on maps, and then draws some conclusions about regolith mapping and scales. Considerable emphasis is placed on methods that have been developed to support the mineral exploration industry in Australia. However, the same methods are now being applied to environmental issues, some of which are discussed briefly. There is also a brief consideration of regolith description.
11.2 REGOLITH CHARACTERISTICS AND LANDFORM MAPPING PRINCIPLES The regolith has a number of characteristics that make a landform approach appropriate to its study at all scales. These are: 1. Regolith is very closely associated with modern, relict and buried landforms (see Chapter 3). This makes a landform approach to regolith mapping most appropriate. This is not to say, however, that there is a one-to-one relationship between specific
282
Regolith Science
landform types and specific regolith types. Such relationships vary a great deal because of other influences, such as geology, climate and landform evolution. 2. Regolith materials are commonly thin, and generally occur as discontinuous layers over the bedrock (Figure 11.1). This means that only some transported regolith materials can be placed into litho-stratigraphic or time-stratigraphic units. Detailed study is required to place most regolith materials into a stratigraphic framework, and bedrock geological mapping techniques are therefore inappropriate for mapping regolith. 3. Individual units of sedimentary regolith generally occur as discontinuous accumulations of sediment in the lower parts of the present landscape or in places that were once the lowest parts of the landscape. Notable exceptions are aeolian materials, volcanic tephra blankets and some glacial deposits, especially ground moraine. Units of sedimentary regolith can seldom be correlated between drainage basins, let alone over much wider regions. The law of superposition – the basis for stratigraphic mapping – does not generally apply to in situ regolith materials (Pain and Ollier 1995). This is well illustrated by layers produced by deep weathering. Weathering acts on whatever materials occur at, or near, the surface. Moreover, weathering produces layers that develop downwards at the weathering front, rather than accumulating upwards. Layers produced by weathering are related to depth from ground surface, and may cross from in situ weathered bedrock into transported material without a change in primary characteristics. Regolith layers produced by weathering are imposed on stratigraphic layers, and are not part of the stratigraphy (Figure 11.2). Nor is cementing part of the stratigraphy, with siliceous hardpans and silcretes in valley floors having formed in both alluvium and adjacent weathered bedrock (Figure 11.3). In any study of regolith, it is necessary to separate weathering and cementing effects from the distribution of regolith types. 4. Regolith formed by in situ weathering of bedrock occurs as layers of various kinds – differing with age of land surface and underlying bedrock, and
reflecting the evolution of the landforms on which it occurs. Many geomorphic surfaces are timetransgressive, and their accompanying units of weathered bedrock are also time-transgressive – even though their characteristics may be similar. A complex regolith pattern and distribution results when in situ weathering profiles are buried by terrestrial sediments, or truncated by erosion. For any given area, a model of landscape development is essential if the regolith is to be understood. Models are also essential for extrapolation of field observations, although it is important to understand the limitations of such models for extrapolation. 5. Despite past attempts to use particular regolith features, such as duricrusts, to correlate different land surfaces, it is becoming clear that this is rarely accomplished with any degree of confidence. For example, Fe can cement any materials to form ferricretes in any part of the landscape where Fe accumulates (Pain and Ollier 1992; Ollier and Galloway 1990). The presence of such ferricrete in no way allows surfaces or materials to be correlated. Similarly, it is unrealistic to speak of ‘periods of weathering’ in anything but a very broad sense (Taylor and Shirtliff 2003). Weathering continues from the moment a land surface is exposed up to the present – unless it is eroded. Deep weathering is primarily a product of a long period of stability in the landscape. Moreover, weathering processes are not sufficiently understood to allow assessment of what changes would take place with a change in environmental factors, such as climate or vegetation. These and other aspects of regolith distribution are discussed by Pain and Ollier (1995), who strongly recommend that the term ‘stratigraphy’ should not be used for regolith materials. ‘Distribution’ covers twoand three-dimensional aspects of regolith, and also allows discussion of variation.
11.3 EXAMPLES OF LANDSCAPE-BASED MAPPING Ecological landscape mapping (Bryan 2006) produces perhaps the broadest kind of landscape-based maps. Just about all characteristics of a landscape –
Regolith description and mapping
(a)
(c)
0
5 km
(b)
0
0.5
1 km
(b)
(c)
0
0.5
Residual sand
Alluvium
Gneiss
Shist
Shallow soil
Alluvium, colluvium
Quartzite
Saprolite
1 km
Granite
Figure 11.1: Cross sections showing the contrast between bedrock and regolith units (Ebagoola 1:250,000 sheet area, North Queensland) (after Pain 1992). (a) Bedrock section. (b) and (c) Regolith sections (locations shown on section a).
from geology to vegetation – are used as separate data layers to define regions that have more-or-less uniform characteristics in a wide range of variables (such as land systems maps; Section 11.3.3). Initially this mapping was carried out subjectively by the surveyors but, with the introduction of geographic information systems (GIS), there have been many attempts to automate the process and to make it more objective. Bryan (2006) provides a very useful review (with abundant references) (see also Martin-Duque et al. 2003), so no detail is included here. He makes the very important point that automatic
attempts to classify landscapes have been largely unsuccessful, and one of the main reasons is that environmental characteristics generally change gradationally over considerable distances, and class boundaries automatically generated along these gradients are not necessarily useful. For that reason, regolith mapping as developed in Australia uses landform as the main criterion for generating class boundaries – this is because landform boundaries tend to be sharper and better defined than many other landscape characteristics. Moreover, in any one region, there is generally a close relationship
283
284
Regolith Science
Bedrock litho-stratigraphy
Tertiary sediments
Permian sediments
Tertiary sediments
Permian sediments
Greenstone basement
Greenstone basement Weathering layers
Siliceous hardpan
Mottled saprolite
Bleached saprolite
Figure 11.2: Simplified diagram of weathering layers superimposed on stratigraphic layers, Yilgarn Craton, Western Australia.
between landform (landscape position) and other landscape features such as soil and regolith. 11.3.1 Geomorphic mapping There are a variety of types of geomorphic maps, ranging from those that show areas with similar landforms to those that show the distribution of a variety of geomorphic features such as terraces, hill slopes, scarps, drainage patterns and structural landforms. Several examples with references are given in Cooke and Doornkamp (1990), while Gustavsson et al. (2006) provide an up-to-date review of geomorphic mapping – and a suggested new system. There is also a Working Group on Applied Geomorphic Mapping supported by the International Association of Geomorphologists (http://www.appgema.org/). However, such maps do not provide much regolith information, as they are based mainly on form, although there may be interpretive elements that imply the genesis of landform materials. Geomorphic maps are a useful basis for regolith maps, but must be complemented with appropriate field observations to determine the landform–regolith relationships.
11.3.2 Soil mapping Pedologists use a landscape approach for mapping regolith materials (Ollier 1995), although they are generally concerned with only the upper part of the regolith. They recognised early that soils are closely related to their position in the landscape. This led to the development of the catena concept, defined by Milne (1935) as ‘a unit of mapping convenience … a grouping of soils which while they fall wide apart in a natural system of classification on account of fundamental and morphological differences, are yet linked in their occurrence by conditions of topography and are repeated in the same relationships to each other wherever the same conditions are met with’. The catena concept is important because it recognises the interaction of soils and landforms and, therefore, soil processes and geomorphic processes. Thus, in areas where catenary relationships are understood, soil types can be predicted from their position in the landscape (Figure 11.4) (see Brown et al. 2004 for a recent review). Soil maps can be prepared from aerial photographs – using landforms to subdivide the mapping area. At
Outline of siliceous hardpan 2
metres 0 0 Residual sand
Alluvium
Moderately weathered saprolite
Figure 11.3: Siliceous induration of alluvium and adjacent weathered bedrock on Cape York Peninsula, Queensland.
10
SOILS
Gently sloping plains and tablelands
Red earths
Yellow earths
Scarps
Upper and middle slopes
Lowlands and plains
Flood plains
Solodic soils with strong texture contrast
Deep cracking clay soils
Alluvial soils
LANDSCAPE POSITION
Skeletal soils
Regolith description and mapping
Laterite over mottled and pallid zones
Colluvial-alluvial deposits
Saprolith
Figure 11.4: Soil catena from northern Queensland, showing variations in soil types with position in landscape (after Gunn 1967).
detailed mapping scales, individual soil types (soils classified on profile characteristics) can be mapped. At regional scales, soil families (soil types grouped together on the basis of soil profile similarities) can be predicted. The essential link is the relationship between soil type and landform. There are two other important and related concepts: those of classification versus map units, and the purity of map units. There are many soil classification schemes that divide soils into theoretical classes on the basis of a wide range of soil attributes. Such classification schemes give details of soil taxonomy (for example, McDonald et al. 1990). However, the results of many studies of catenas show clearly that the theoretical association of soil types in a classification system does not indicate how those soil types are associated in the landscape. Thus, a soil map shows soil map units, which are associations of soils that occur together in the landscape. The more detailed the map, the more closely the soil map units will reflect the soil classification units. Purity of map units depends on mapping scale. At detailed scales, map units will be more homogeneous; at regional scales, they will be less so. This is because map units often contain small areas of material different from the label given to the map unit.
11.3.3 Land systems mapping A land system was defined by Christian and Stewart (1953) as ‘an area or group of areas throughout which a recurring pattern of topography, soils and vegetation can be recognised’. Thus, there are similarities between soil catenas, which relate soil types to particular positions in a landscape, and land systems, which relate land units (subdivisions of land systems) to particular positions in a landscape. An important overall concept is that the landscape is organised into natural associations of land attributes, and that these natural associations can be used in a predictive way to describe land, and to extrapolate from known to unknown areas. Equally important is the idea that landforms are used as the first criterion for defining land systems. Such ideas can be used very effectively in the study and mapping of regolith. Land system descriptions contain information about underlying geology, landforms, soils, vegetation and, in some cases, hydrology. Regolith information is generally missing or only inferred. Nevertheless, at the regional scale, a map of land systems will look similar to a map of soil associations for any area at the same scale. This is because both are prepared using landforms as the main criterion for drawing
285
286
Regolith Science
boundaries. This, in turn, implies that a soil or land system map will allow some inferences to be made about regolith, especially if field work can be used to check regolith characteristics in the mapped area. A number of organisations have developed methods for assessing a variety of natural resources at a regional scale (Ollier 1977; Stewart 1968; Speight 1988). In Australia, the CSIRO has mapped land systems and described land units nationally (for example, Galloway et al. 1974). Such resources should not be ignored by regolith mappers. Surficial geology maps In Australia, maps in the 1:250 000 geological series commonly show Cenozoic units (the so-called yellow wash) that, in many cases, can be equated with regolith. In areas of transported regolith the correspondence can be good, but in areas of in situ residual and weathered material problems arise because the units are assigned an age – generally Cenozoic. As pointed out above (Section 11.2), defining regolith from stratigraphy is not valid. Such geological maps provide useful information about the distribution of hard rock materials, but cannot be used to indicate regolith. The Canadian Geological Survey has produced a national map of surficial materials at a scale of 1:5 000 000 (Fulton 1995). This map shows a variety of surficial sediments, and areas of rock exposure. Residual material is not shown, and is probably uncommon in Canada. In the United States of America there are a number of maps of surficial materials. That of Soller and Reheis (2004) is an example at the national scale, and it contains many references to other surficial material maps produces by various state geological surveys. This map shows the sediments and the weathered rock materials at the land surface, and is in this respect very similar to a national regolith map. Materials are divided into sediments and residual materials. The sedimentary materials are further classified according to origin, while residual materials, soil and weathered material down to unweathered bedrock are subdivided according to rock type. These maps show some similarities to state and national regolith maps of Australia.
11.4
REGOLITH–LANDFORM MODELS
The geomorphic literature is full of models of landform evolution (for example, Ollier and Pain 1996). Selection of a particular model places constraints on the predictions that can be made about regolith–landform inter-relationships. A full understanding of regolith distribution in an area comes only with the development of a conceptual model of landform and regolith evolution. Such a model must be based on field evidence, and must account particularly for the distribution of landforms and regolith in the study area. The models may allow regional extrapolation of mapping results, and should be underpinned by regolith studies at a detailed scale. Regional regolith–landform models are developed from analysis of the broad landform framework of an area, including elevation distribution, location of major landform features, and drainage pattern and density analysis. The regional models can then be used to predict the details of regolith distribution in more local areas. Regional models will not predict the precise location of specific regolith materials, but they will allow a prediction of regolith–landform associations, and the complexity of those associations. The development of conceptual regolith–landform models becomes an iterative process, with detailed mapping results feeding back into the regional models.
11.5 REGOLITH–LANDFORM MAPPING 11.5.1 General concepts A regolith–landform map shows regolith–landform units as polygons. Regolith–landform units are land areas characterised by similar landform and regolith attributes. This definition – and the resulting mapping method – is made possible by the close relationship between landforms and regolith. Landforms and regolith are formed by essentially the same group of processes, and once the inter-relationships between regolith and landforms are understood, landforms can be used to predict regolith patterns in mapping areas. Thus, if the mapping is sufficiently detailed, landforms can be used as a proxy for the largely hidden regolith. This means understanding the dynamics of the present landscape. It also means understanding
Regolith description and mapping
BUer RSep
(a) ACaf
(b)
RSep
Alluvium
100
Approximate scale (metres)
SM
ec
Residual sand
0 500
0
Aap
Mottled saprolite
Moderately weathered saprolite Granite bedrock
Figure 11.5: Regolith toposequences on Cape York Peninsula (Pain et al. 1994). (a) A toposequence on erosional rises and plains consisting of bare rock, residual sand on saprolite, and channel alluvium. (b) A toposequence on a plateau edge consisting of residual sand, saprolite and alluvium on fans. See Tables 11.2 and 11.3 for explanation of symbols.
the dynamics under which relict landforms and regolith materials formed. Regolith–landform unit boundaries are drawn on the basis of landforms, and the resulting map polygons are described in terms of regolith types and landforms. This approach uses the soil catena concept, and the idea of regolith toposequences as a basis for polygon description. A regolith toposequence is a group of regolith types that are linked by their regular association with particular landforms. Examples are given in Figure 11.5, where three adjacent map units are illustrated. This procedure requires fieldwork to collect new data and to provide field control on map units compiled from secondary sources, such as soil and land system maps, and remote sensing. Fieldwork for regional mapping consists mainly of determining the regolith toposequences present in each map unit. This allows useful description of the regolith–landform relationships in mapping areas, and also allows interpolation of fieldwork results to areas that are not visited in the field, but are covered by remote sensing or other maps. Regolith toposequence descriptions should contain the following information:
1. The distribution of surface regolith materials in relation to landscape position. This information is particularly important at regional scale mapping because it will not be possible to show all surface regolith materials on the map. 2. The 3D distribution of regolith materials. This information is essential for characterising the regolith of an area, and is also an important basis for testing models of landform and regolith evolution. Chan (1988), Anand et al. (1989a, 1993), Pain et al. (1991, 2007) and Hocking et al. (2001, 2005, 2007) all contain discussions on regolith–landform mapping. There are also more general discussions in Ollier and Pain (1996) and Taylor and Eggleton (2001). 11.5.2 The importance of geomorphology Landform description is an important first step in regolith–landform mapping. However, regolith–landform mappers must also understand the basic geomorphic principles behind landform and regolith development. These principles include drainage pattern evolution, slope processes and evolution, fluvial processes and the development of floodplains,
287
Regolith Science
terraces and fans, and hypotheses of long term landform change. An understanding of weathering processes and products is also essential. At more detailed scales, the use of subdivisions, such as the nine-unit land surface model (Dalrymple et al. 1969; Conacher and Dalrymple 1977), allows correlations between slope form, slope processes and regolith types to be assessed (see also Phipps 2001). 11.5.3 Scale At a regional scale, regolith–landform maps have much in common with land system maps. At a local scale, a regolith–landform map is more like a soil map. However, at all scales except the most detailed, landforms provide the basis for mapping. Webb and Lilburne (2005) – who are concerned with soil maps – note that there can be considerable uncertainty in map unit composition, and that this results in considerable spatial variability in soil properties within map units. The same is true for regolith maps – especially those covering regional or national areas. The scale of the landscape is also important. In places, the intricate nature of the landscape may mean that homogeneous areas occur in areas of 1–10 km2, as in the dissected eastern highlands of Australia. Elsewhere, uniform regolith–landform units may cover 100–1000 km2, as in areas of western New South Wales. The scale of the landscape requires changing the mapping scale to suit the landscape, or recognising that landscape scale has a direct effect on the homogeneity of map units (Figure 11.6). Figure 11.7 shows the changes that can occur as mapping becomes more detailed. A regolith map can show either just the regolith, or regolith–landform units. A disadvantage of the former is that a considerable part of the context-related knowledge and data that are collected and used during a regolith survey is lost in the final map preparation. This can be overcome if the map and data are presented as part of a hierarchical system of map units in which smaller units are nested within larger ones. Wielemaker et al. (2001) named this method a multihierarchical land system approach. The multi-hierarchical approach places regolith and landscape information in a wider perspective. In such a system, the levels of the hierarchy must be well defined and
knowledge of units at each level formalised. Geology and geomorphic criteria are used for compilation of units at nearly all levels, except the most detailed. Johnston et al. (2003) describe the Australian Soil Resources Information System (ASRIS), which is in many ways similar to that of Wielemaker et al. (2001). The hierarchical nature of ASRIS is best described in the ASRIS Technical Specifications (McKenzie et al. 2005). Table 11.1 gives a hierarchical approach to regolith mapping scales, and is based in large part on ASRIS. Lawrie et al. (2005) present a hierarchical scheme for the three-dimensional (3D) distribution of landscape and regolith features, as follows: 1. First order features. Bedrock-influenced elements such as structure, variably weathered saprolith and limited fresh bedrock (such as resistive silicified ridges). These landscape elements provide the large-scale controls on regolith distribution, and have characteristic dimensions of 10 km across and greater than 20 m thickness. 2. Second order features. Basins developed through preferential erosion of weathered bedrock and their subsequent infill. These form discrete basins with characteristic dimensions of 1 km across and less than 20 m thickness. 3. Third order features. Sinuous paleochannels and their fill materials. The paleochannels have
Homogeneity
288
Coarse
Map Scale
Fine
Low
Number of Classes
High
Coarse
Scale of Landscape
Fine
Figure 11.6: Generalised relationship between scale and homogeneity of map units (after Bryan 2006). Landscape scale is as important as map scale, and can dictate the latter.
Regolith description and mapping
(a)
Broken Hill
(b)
Ser7
o
32 00’
ACat1
SSel1
Aap2
Aep1 Aap1 Ser7
SSel2 CHei1
CHfc2 CHfc2 Isu2
RLer1 Aep1
SSer1
Aep1
CHfc2
Aap1
SHer5 SMer1 Aap1 SSer1
SMer1
CHfc2 CHei1
Aep1
500m
Aaf1
SMer1 Aep1 Aep1 o
o
32 15’ o 141 15’
10 km
31 30’ o 141 00’
o
141 30’
Figure 11.7: An illustration of the differences in detail allowed by different scale maps (after Taylor and Eggleton 2001). (a) Part of the 1:500 000 scale Broken Hill regolith–landform map (Gibson and Wilford 1995). (b) Part of the 1:25 000 scale Balaclava regolith–landform map (Foster et al. 2000). See Tables 11.2 and 11.3 for explanation of symbols.
Table 11.1: Regolith-landform mapping hierarchy. The ASRIS order is from McKenzie et al. (2005). Level
Regolith hierarchy
Characteristic dimension
Descriptive or defining attributes
Appropriate map scale
ASRIS order
0
100 km
Very broad geology – only 3 units for Australia
1
30 km
Broad physiography (slope and relief) and geology
1:10 million
Division
10 km
Physiography, geology
1: 2.5 million
Province
2
Physiographic region
Mapping hiatus 3
Regolith–landform province
3 km
Broad landforms and regolith materials
1:1 million
Zone
4
Regolith–landform association
1 km
Groupings of local landforms and associated regolith related in toposequences
1:250 000
District
5
Regolith–landform unit
300 m 100 m
Local landforms and associated regolith
1:100 000 1:25 000
System
6
Regolith–landform facet
30 m 10 m 3m
Slope, aspect, regolith class
1:10 000 1:2500 1:1000
Facet
7
Site
10 m
Regolith properties, surface condition, microrelief
NA
Site
289
290
Regolith Science
characteristic dimensions of 500 m across and greater than 10 m thickness. 4. Fourth order features. Small-scale basins and tributary paleochannels with characteristic widths of 100 m and thicknesses of less than 10 m. 11.5.4 Image data Remote sensing has been used for regolith mapping since the early 1980s (see Chapter 9, also papers in Papp 2002). Aerial photographs play an important role – as do images from satellite systems such as Landsat (both MSS and TM), SPOT and NOAA (see Table 9.1 for explanation of acronyms). Imagery from airborne systems that record gamma-ray spectrometric and magnetic data, and hyperspectral data (such as Hymap®) may also be used. In the compilation of the 1:5 000 000 scale regolith map of Australia (Chan et al. 1986), NOAA images – with a ground resolution of 1 km – were used for mapping regolith units where soil maps were not available, because these images were the most useful for discrimination of soils types. Landsat MSS – with a ground resolution of 90 m – was also used. Landsat TM was used to map the regolith of Cape York Peninsula, northern Queensland, in the early 1990s, at a compilation scale of 1:100 000 and a final publication scale of 1:250 000 (for example, Pain et al. 1994). It had a ground resolution of 30 m, and was useful because of minimal human modification of the environment. In the more heavily populated areas of south-eastern Australia, Landsat TM was of little value in the compilation of the Bathurst 1:250 000 scale regolith–landform map (Chan 1995). ASTER is another satellite remote-sensing system that can be used to discriminate regolith materials on the basis of their mineralogy (for example, Gozzard 2006). Hymap® (airborne hyperspectral) data provides even more detail (for example, Dehaan and Taylor 2002) (see also Chapter 4). The use of hyperspectral data is also being used for regolith–landform mapping. Skwarnecki (2005) provides a recent example in which the spectra of regolith materials are used with hyperspectral data to discriminate areas with particular regolith minerals such as gypsum, Fe oxides and kaolin.
Airborne gamma-ray spectrometry (airborne radiometrics) is being used increasingly for regolith mapping in Australia (Wilford et al. 1997). Aircraftmounted instruments flown 60–100 m above ground measure the abundance of potassium (K), thorium (Th) and uranium (U) in rocks, soils and weathered materials by detecting gamma-rays emitted by natural isotopic radioactive decay of these elements (see Chapter 9.4). Ninety per cent of gamma rays emanate from the top 30–45 cm of dry rock or soil. The intensity of gamma rays emitted from the land surface relates to the mineralogy and geochemistry of the surficial material, and the nature of weathering. Interpretation of these data allows separation between different bedrock types, if exposed, and between different weathering patterns on the same bedrock type (Dickson and Scott 1997). It also allows discrimination of alluvium and colluvium. Present day geomorphic activity can be revealed – and this allows actively eroding surfaces with shallow regolith to be separated from stable surfaces with deeper regolith and paleolandforms. It also allows separation of depositional regolith materials derived from different sources, or of different ages. As Dickson et al. (1996) indicate, use of gamma-ray images is considerably enhanced when combined with knowledge of geomorphic and weathering models. Airborne magnetic data – generally collected at the same time as airborne radiometric data – are also useful for regolith mapping (see Chapter 9). In areas with suitable regolith mineralogy and geochemistry, modern drainage lines can be separated from paleodrainage, and buried paleochannels can be recognised in some areas. In addition, shallow high-frequency magnetic responses on airborne magnetic data can be used to delineate highly ferruginous regolith materials, such as ferricrete and pisoliths. Wildman and Compston (2000) provide a good example from the Yandal Belt in Western Australia, where they were able to map the distribution of deep paleochannels where the channels contain magnetic material, especially maghemite. Chan et al. (2004) also used airborne magnetic data to map paleochannels containing maghemite. Airborne electromagnetic (AEM) surveys of the regolith are also proving useful for studies of regolith stratigraphy and distribution (for example,
Regolith description and mapping
Street and Anderson 1993). In the Lower Balonne area of southern Queensland, an AEM survey provided information for assessing the 3D distribution of regolith materials (Kernich et al. 2004) (see also Chapter 9). Ground EM surveys also provide useful information about regolith in the subsurface (McKenzie et al. 1997). Digital elevation models (DEMs) have proved to be of great value for mapping landforms at a variety of scales (see Pain 2005 for a brief review). This is an area of continuing research (Gallant and Dowling 2003). 11.5.5 Ground geophysics The use of geophysics in regolith studies is covered in Chapter 9. However, for mapping purposes techniques such as ground EM, ground penetrating radar and shallow seismic can all be used to provide additional information about the regolith of a survey area (see Papp 2002). 11.5.6 Mapping methods Regolith–landform mapping has three aspects: data collection, interpolation, and compilation. Data collection consists of both image analysis and fieldwork. The range of attributes that can be described for both sites and map units is covered by Anand et al. (1993, 2002) and Pain et al. (1991, 2007), and terminology by Eggleton (2001) and (Appendix 1). Data collection
The use of image data for regolith mapping has already been discussed in Section 11.5.4; this section concentrates on field studies, which are aimed at collecting ground truth information about various regolith and landform attributes from specific field sites, and from general field observations. A major aim is to determine the regolith toposequence(s) for each regolith– landform unit. In some areas, there is sufficient local relief to define toposequences from aerial photographs. However, many parts of Australia have extremely low relief, and map units cannot be based on landforms in the normal manner. In these areas, fieldwork involves calibrating the regolith using the radiometric and Landsat TM responses. The first step in fieldwork should be a reconnaissance of the area of study. This can be made by driving
through a representative part of the study area. It is good practice to get to the higher points in the landscape to observe more than one regolith landform unit. An aerial overview gives a good idea of the general features of the landscape, and also allows the different regolith landform units to be considered in context. The reconnaissance should also be used to check the general relationships between regolith landform units and geology. For practical reasons, sites should be selected on or near roads and tracks. This will make fieldwork efficient, and less time consuming, while still allowing good field observations. Before going into the field, note the location of the different land units in relation to transport routes, and make field plans accordingly. In most cases, it is possible to extrapolate observations from accessible areas into areas that are far from roads or paths. Observations of the landscape (in its broadest sense) can give a lot of detail that is not available from maps and images. Look for explanations of the landscape – its origins, active and relict geomorphic and weathering processes – and any information that will allow extrapolation to inaccessible areas. Images give a general idea of landform characteristics, but field observations give the opportunity to get information about the details of landforms, whether the slopes are smooth, undulating or irregular, and whether there is any obvious control by underlying rocks. Make a note of the types of geomorphic processes within regolith landform unit boundaries. Obtaining information about the regolith at depth can be a problem. Deep exposures, especially those more than 2 m deep (road cuttings, stream banks, gullies, gravel pits, and so on), should be studied wherever possible. Alternatively, drilling may be used – and some information can be obtained from shallow seismic data. Ground-penetrating radar offers a new, and as yet largely untried, source of data about the regolith. Also, AEM surveys are providing data that show great promise for assessing regolith distribution and characteristics to depths in excess of 100 metres (see Chapter 9; also Worrall et al. 1999). Deep exposures are by far the best source of information about the third dimension of regolith. Core or loose samples from drill holes can only be
291
292
Regolith Science
identified with difficulty as a particular regolith type. However, much of the Australian continent has a very low relief, and drill holes are often the only way of assessing geophysical techniques such as AEM. It is therefore important to be familiar with material from drill holes. Fieldwork – and the rigorous description of field sites – generates a large amount of data that have to be handled properly to be accessible both during the survey and into the future. The RTMAP database (Pain et al. 1991) has a series of entry screens for entering site data, and there is a summary data form for field use. The first page of the site data form is for information about the site as a whole. The second page is for individual layers, or zones, at the site. A palmtop can be used to record data in the field. The RTMAP database structure can be loaded onto such a portable device – as can maps and images of the survey area. When linked to a GPS and a digital camera, this provides a powerful tool for fieldwork, and for collecting and storing field observations (for example, Robertson et al. 2006). RTMAP is described in more detail in Section 11.6.1. Compilation and interpolation
Mapping of boundaries (compilation) is best accomplished from aerial photographs and other remotely sensed data sets (see above). Map compilation should be at a more detailed scale than the scale at which the map is intended to be used – although the distinction between compilation and use scales can be blurred when using a GIS. For example, the 1:250 000 scale regolith–landform maps produced by Geoscience Australia (for example, Chan 1995; Pain et al. 1994) are compiled at 1:100 000 scale. The procedure is to compile regolith–landform unit boundaries derived from the images onto a draft map, which can then be taken into the field for checking. Information from sites and field observations can help with compilation and checking of regolith–landform units. To date, compilation is more efficient if done first by hand on to stable paper, and then digitised by scanning. Recent advances in software may allow compilation directly into a digital format. Map compilation inevitably involves interpolation between a patchwork of sites and observations; this is
generally reliable, but should be based on models of landscape evolution and knowledge of regolith and landform relationships. Extrapolation of regolith– landform units from one survey area to another – even if it is adjacent – is more problematical without field checking.
11.6 PRESENTATION OF REGOLITH INFORMATION Regolith mapping began in Australia in 1982, with most products released as simple paper maps or as figures in reports (Anand et al. 1989a, 1993; Chan 1988; Chan et al. 1992). By 1990, the widespread use of GIS necessitated a change in the way data were collected, stored and presented. Much of this section derives from experience at Geoscience Australia – although the Geological Survey of Western Australia has also worked on databases for regolith mapping and materials (Hocking et al. 2001, 2005, 2007; Riganti et al. 2003). 11.6.1 Databases Much of the regolith data gathered in the field before Geoscience Australia moved to digital storage and manipulation used ill-defined terminology. Moreover, the terminology was being used inconsistently, and many of the attributes were being recorded in verbose descriptions that limited valid comparisons of map units. Relationships between attributes were not properly established, and it was not possible to search the data. For these reasons, the first relational database for regolith information with the ability to interface with a GIS was set up at Geoscience Australia in 1992. In the ensuing analysis, it was realised that for some attributes – such as regolith types, landform types, geomorphic and weathering processes, drainage patterns and some tectonic structure elements – there were only a relatively limited number of distinct values. Where available, authoritative references such as the Australian Land and Soil Survey Handbook (McDonald et al. 1990) were used to derive lists of these values and their definitions. In other cases (such as regolith types), Geoscience Australia devised lists after discussions with numerous workers. These
Regolith description and mapping
authority lists formed the basis of the look-up tables that are used in the database for the description of field sites and map units. More recently, a Corporate Data Model (CDM) has been implemented at Geoscience Australia, and authority lists are being migrated from individual databases to the CDM. During this migration, the authority lists will be reassessed with help from state and territory agencies concerned with both mineral exploration and natural resource management. In this way, considerable effort is being spent on structuring regolith and landform attributes into distinct, well-defined categories that are suitable for use in a relational database. The resulting Geoscience Australia regolith–landform mapping database is called RTMAP (Pain et al. 1991; Lenz 1991; Lenz and Pain 1992). A number of basic guidelines were recognised: 1. The RTMAP database had to be compatible with a GIS. 2. Standard terminology had to be used – relying on existing, rather than new, terms. 3. Terminology had to be ‘protected’: agreed terms must be used, and new terms could not be introduced. 4. Terminology had to fit into a hierarchical classification, and to be assigned codes accordingly. This allowed more efficient use of the database, especially for searches. More recently, other regolith databases have been set up (for example, Riganti et al. 2003; Brough et al. 2006). Structured databases provide powerful tools for storing and searching data. However, once set up, they tend to lack flexibility. For example, although it is easy to add terms to a look-up list, it is more difficult to delete terms or to change their meaning. The RTMAP database and GIS links
Most earth science databases contain information about a number of attributes at sites. Sites are small areas of land considered to be representative of the land features associated with the observation being made. For example, soil scientists (McDonald et al. 1990) describe land and soil attributes at soil profile
sites. Site information is also important for regolith studies, so site data are collected for RTMAP. Sites can be spatially referenced, and thus become point data in a GIS. However, point data do not give the GIS user general information relating to broader areas. For regolith – as for most land related attributes – it is not possible to use site data to characterise map units. There is too much spatial variability, which results in varying degrees of impurity of map units, depending on mapping scale (Figure 11.6). For this reason a second set of attributes is entered into RTMAP. These data relate to the whole map unit, or regolith–landform unit. They can incorporate site data, but they also incorporate observations of the whole map unit, made both in the field and from maps and images. Among other things, variability is described. These data become the polygon attributes in the GIS, and support the great potential for spatial analyses provided by the GIS. Thus, both site and unit data are essential for full utilisation of the GIS. Standard terminology
The use of standard terminology allows results between areas to be reliably compared. This has long been recognised by those engaged in soil and land survey (McDonald et al. 1990). The two most important groups of attributes in RTMAP relate to regolith and landform types. The initial landform classification and terminology in RTMAP is the landform pattern of McDonald et al. (1990), which is well suited to regional mapping scales. More recently, landform elements have been added to allow subdivision of landform patterns. McDonald et al. (1990) was also used for terms relating to drainage, geomorphic processes and weathering processes. Regolith terminology presents a problem. Unlike soil materials, regolith had not been classified in any systematic way for use at a regional mapping scale – although very useful work on classifying ferruginous (lateritic) and related materials had been carried out by the CSIRO Exploration and Mining (Anand et al. 1989b, Robertson and Butt 1993), largely for purposes of mineral exploration. Hocking et al. (2001, 2005, 2007) provide a classification system of regolith for use in Western Australia, based in part on the work of Anand et al. (1993). A classification of
293
294
Regolith Science
regolith was developed by Pain et al. (1991) – keeping in mind its purpose, which, at that time, was for compiling regolith maps at 1:100 000 for a publication scale of 1:250 000. Regolith–landform units at this scale inevitably contain groupings of specific regolith materials, so the map units are three-dimensional entities that frequently contain a wide variety of specific materials. The classification used in RTMAP therefore contains the basic regolith types (Pain et al. 2007). The current (2008) regolith and landform tables are given in Tables 11.2 and 11.3. As noted above, these are being re-assessed before migration to Geoscience Australia’s CDM. 11.6.2 Regolith maps The presentation of information on modern regolith maps relies heavily on GIS technology. This allows on-demand printing of standard hard copy regolith maps, rather than limited lithographic print runs. GIS coverages may also be sold on CD, and entered into clients’ GIS databases. In addition, thematic maps can be produced for any attribute stored in the RTMAP unit tables. In Australia, the automation of mapping, field data manipulation, database interrogation and digital data presentation has gained substantial momentum since comments by Rattenbury et al. (1992). The pitfalls – for example, spatial accuracy, referred to by Craig et al. (1993) – are yet to be fully resolved. Major programs involving regolith mapping by GA and CRC LEME (see Craig and Churchward 1995; Wilford and Craig, 1997) in Australia, now routinely use the framework set out by the RTMAP database and field mapping system (Pain et al. 1991; Craig and Chan 1992; Craig 2006). Some state geological surveys also have regolith mapping programs (for example, Marnham and Morris 2003). (See Section 11.6.1) Standard maps
Practitioners of regolith–landform mapping in Australia are still dealing with the problem of how to display regolith information on maps. Unlike geology or soil maps, there is no standard for regolith maps. The first regolith map produced – the 1:5 000 000 Regolith Map of Australia – was a simple colourless polygon map, with numbers identifying polygons
with information on each polygon class given within an accompanying report (Chan et al. 1986). A similar approach was used on the Hamilton (Victoria) 1:1 000 000 map (Ollier and Joyce 1986). With time, colour schemes and a set of standard symbols for polygons were devised. Polygon identification symbols currently in use consist of one or two upper-case letters, followed by one or two lower-case letters and a numerical suffix (Figure 11.8). The upper-case letters refer to the major regolith type (Table 11.2), and lower-case to landform type (Table 11.3), for each polygon class – and the numerical suffix distinguishes between polygon classes with similar major regolith and landform, but with differing details or minor components. A hierarchy of letter symbols has been devised that allows for rapid interpretation of regolith and landform from the letters used. In addition, various overprint stipples are used to delineate certain features, such as induration modification of regolith materials (silicification, clay hardpans, and so on), and in some cases, multiple weathering horizons associated with Cainozoic basalt flows. Figure 11.7 gives an example of this type of product. Note that the map symbols differ from the RTMAP database codes. The map symbols are designed for easy interpretation of map polygons, whereas the database codes are longer, and allow more rigorous searches to be made of the database. In regolith–landform maps, polygon classes with similar major regolith types – that is, with the same upper-case letters in the map symbol – are depicted in the same colour. Several different colour schemes have been used. The main map reference consists of colour Regolith type
SMer 1
Modifier
Landform type Figure 11.8: Map symbol layout used on Geoscience Australia regolith–landform maps.
Regolith description and mapping
Table 11.2: symbols.
Regolith types, RTMAP codes and map
Regolith type Unweathered bedrock Evaporite Halite Gypsum Alluvial sediments Channel deposits Overbank deposits Colluvial sediments Scree Landslide deposit Mudflow deposit Creep deposit Sheet flow deposit Fanglomerate Aeolian sediments Aeolian sand Loess Parna Fill Glacial sediments Lacustrine sediments Marine sediments Peat Coastal sediments Beach sediments Estuarine sediments Coral Terrestrial sediments Clay (unknown origin) Weathered material (unknown origin) Sand (unknown origin) Volcanic sediments Lava flow Tephra Saprolite Slightly weathered bedrock Moderately weathered bedrock Highly weathered bedrock Very highly weathered bedrock Completely weathered bedrock Residual material Lag Residual sand Residual clay
RTMAP code
Map symbol
BU00 EVA00 EVA01 EVA02 SDA00 SDA10 SDA20 SDC00 SDC01 SDC02 SDC03 SDC04 SDC05 SDC06 SDE00 SDE01 SDE02 SDE03 SDF00 SDG00 SDL00 SDM00 SDP00 SDS00 SDS01 SDS02 SDS03 SDT00 UOC00 UOM00
BU E EH EG A AC AO C CS CL CM CC CH CF I IS IL IP F G L OM P O OB OE OC T UC UW
UOS00 VOL00 VOL01 VOL02 WIR10 WIR11 WIR12 WIR13 WIR14 WIR15 WIR20 WIR21 WIR22 WIR23
US V VF VT S SS SM SH SV SC R RG RS RC
boxes and unit descriptions, with a major grouping into transported and in situ regolith. Transported regolith is divided into genetic groups (alluvium, colluvium, aeolian sediments, and so on). In situ regolith is divided into groups of different residual materials and saprolite (used to describe bedrock with any degree of weathering) of varying degrees of weathering. There is also a set of symbols for regolith of undefined origin. Various geomorphic information – such as fault and erosion scarps, drainage divides, sites of river capture and reversal, wind gaps, antecedent and superimposed drainage, locations and trends of paleodrainage – are also depicted on some maps. Mineral deposits and prospects are shown on some maps as coloured dots, with each colour denoting the major mineral present. A separate key to lower-case symbols, and the landforms to which they correspond, is also included. Roads, rivers, railways and towns are shown, along with an appropriate metric grid and latitude/ longitude data. Keys to the location of the map area are included and, in many cases, a small marginal map showing areas of different reliability of the main map, and field site locations. Plots of radiometric data and digital terrain models may also be included in the map surrounds, as well as photos of various features of regolith and landscape. Regolith polygon descriptions form an essential part of Geoscience Australia’s national regolith information. Digital versions of these regolith maps, and increasingly their attribute databases, are publicly available from GA as the field mapping is finalised. The database attributes are recognised as essential for the later construction of thematic regolith maps. Regolith map production within CRC LEME follows the same model (for example, Anand and Paine 2002). The Western Australia Geological Survey and Geoscience Victoria also produce regolith–landform maps as standard products: for example, the Regolith Landform Resources map of the Karridale–Tooker and Leeuwin 1:50 000 sheets ( Hall and Marnham 2002) and the regolith map of the Ballarat–Creswick area (Bibby and Radojkovic 2002; Radojkovic and Bibby 2003), respectively. Geoscience Victoria have adopted the RTMAP terminology, whereas the Western Australia Geological Survey use a system modified from
295
296
Regolith Science
Table 11.3:
Landform types, RTMAP codes and map symbols.
Landform type
RTMAP code
Map symbol
Landform type
RTMAP code
Map symbol
Alluvial landforms
AL00
a
Low hills
ER30
el
Alluvial plain
AL10
ap
Residual low hill
ER31
es
Flood plain
AL11
af
Hills
ER40
eh
Anastomatic plain
AL12
aa
Mountains
ER50
em
Bar plain
AL13
ab
Escarpment
ER60
ec
Covered plain
AL14
ac
Badlands
ER70
eb
Meander plain
AL15
am
Drainage depression
ER80
ed
Floodout
AL16
ao
Fan
FA00
f
Alluvial terrace
AL20
at
Alluvial fan
FA01
fa
Stagnant alluvial plain
AL30
as
Colluvial fan
FA02
fc
Terraced land
AL40
al
Sheet-flood fan
FA03
fs
Alluvial swamp
AL50
aw
Glacial features
GL00
g
Coastal lands
CO00
c
Depositional glacial features
GL10
gd
Beach ridge
CO01
cb
Erosional glacial features
GL20
ge
Chenier plain
CO02
cc
Karst
KA00
k
Coral reef
CO03
cr
Made land
MA00
m
Marine plain
CO04
cm
Meteor crater
ME00
t
Tidal flat
CO05
ct
Plain
PL00
p
Coastal dunes
CO06
cd
Depositional plain
PL01
pd
Coastal plain
CO07
cp
Lacustrine plain
PL02
pl
Beach
CO08
cc
Playa plain
PL03
pp
Delta
DE00
d
Sand plain
PL04
ps
Dunefield
DU00
u
Plateau
PT00
l
Longitudinal dune field
DU01
ul
Plateau edge
PT01
le
Erosional landforms
ER00
e
Plateau surface
PT02
ls
Erosional plain
ER10
ep
Volcano
VO00
v
Pediment
ER11
ei
Caldera
VO01
vc
Pediplain
ER12
ea
Cone (volcanic)
VO02
vv
Peneplain
ER13
en
Lava plain
VO03
vl
Etchplain
ER14
ee
Ash plain
VO04
va
Rises
ER20
er
Lava flow
VO05
vf
Residual rise
ER21
eu
Lava plateau
VO06
vp
CSIRO Exploration and Mining, but also indicate the RTMAP symbols on some products. Interpretive maps
Regolith–landform maps can be used as the basis for a number of interpretive products. Perhaps the best
known interpretive regolith–landform maps are those produced by CSIRO Exploration and Mining (for example, Churchward et al. 1992; Anand et al. 1993). These workers note that regolith–landform units can be broadly grouped into three major regimes: relict, erosional and depositional (RED). The RED scheme is
Regolith description and mapping
a means of interpreting factual regolith maps of deeply weathered terrain, which was initially developed for application to geochemical exploration on the Yilgarn Craton. It is based on the concept of a landscape that was covered by an extensive blanket of ferruginous duricrust that has been modified by erosion and deposition (Figure 11.9, page 168). The relict regime represents a grouping of regolith mapping units in areas that, in the Yilgarn Craton, are characterised by lateritic residuum at, or close to, the surface. An erosional regime represents a grouping of regolith mapping units in partly eroded terrain characterised by saprolite and/or bedrock. A depositional regime is characterised by widespread sediments that may be many metres thick and may overlie lateritic residuum, saprolite or bedrock. This concept is a simplification because the relict regime did not form a widespread, continuous unit on a deeply weathered surface, but rather a discontinuous cover on a broadly undulating plateau, with several cycles of weathering and erosion. Furthermore, the regolith distribution and composition of sediments suggest that the landscape not only included ferruginous duricrust, but also uniformly Fe oxide-stained, red clay soils (Anand et al. 1993). Because of their susceptibility to erosion, red clays have probably contributed much to the sediments in the depositional regimes. Despite these conceptual limitations, the scheme provides a practical guide for geochemical sampling and interpretation and has application in equivalent terrains elsewhere. The RED scheme differs from the RTMAP scheme in that the latter includes the relict regime with the erosional regime, and does not assume the former existence of areas with deep and intensely weathered in situ regolith. Both make interpretations about the nature of geomorphic activity that led to the formation of the present landscape and its regolith, and both can be used as guides to the distribution of regolith, and sampling strategies for geochemical exploration. These interpretations can also be incorporated into models of landscape evolution for survey areas, although equating particular regimes with purported ancient land surfaces, such as the Old and New Plateaux of Jutson (1914, 1934), is problematical. In Western Australia, Sanders et al. (1998) evaluated the use of widely spaced regional regolith geochemistry as an adjunct to geological mapping, and to
determine the usefulness of regolith geochemistry in suggesting areas with mineralisation potential. The Leonora 1:250 000 Regolith Materials Series Map (Bradley and Storey 1995) is an example of the maps produced. Wilford (2002) compiled an Interpretive Geochemical Sampling Strategy Map for the Selwyn area near Mt Isa in Queensland, in which he grouped regolith–landform units into geochemical sampling units on the basis of geochemical and landscape attributes. In the Half Moon area on the Gawler Craton in South Australia, Wilford et al. (2001) produced the following thematic maps: 1. Regolith–landform units with gold contours superimposed 2. Regolith–landform boundaries superimposed over a Landsat TM image 3. Regolith–landform boundaries superimposed over three-band gamma-ray spectrometry 4. Regolith–landform boundaries superimposed over total count gamma-ray spectrometry 5. Regolith–landform boundaries superimposed over a hill-shaded digital elevation model 6. Regolith–landform boundaries superimposed over gridded Au-in-calcrete concentrations. These few examples show the enormous potential for using regolith–landform maps to produce a number of on-demand maps. 11.6.3 New directions for regolith–landform mapping Regolith mapping is now entering a new phase of development in Australia – going beyond the identification and refinement of the basic processes and techniques. Mapping ideas are constantly growing in sophistication and are now addressing the pressing need for multi-data-set integration and analysis within a GIS environment. A new direction for regolith maps is the design and development of thematic maps derived from multiple data sets. Such maps are based on thorough field observations enhanced by combinations of additional information from other sources: for example, Landsat TM, airborne radiometrics and magnetics, air- and satellite-borne radar imagery, geochemical and geological data sets and digital elevation models (DEM). New remote-sensing
297
298
Regolith Science
systems have also been developed. These include spacecraft or aircraft-mounted hyperspectral sensing systems (with in excess of 100 narrow sensing bands), which have the capability to identify and map the distribution of many varieties of clay and other minerals exposed at the land surface. These types of systems will assist in further developing and refining rapid assessment and predictive capabilities. Currently, fundamental regional- and local-scale regolith maps show the potential of the basic regolith attribute data when coupled to the GIS environment, but the greatest potential of this data is still to be fully realised by exploiting multi-data-set interactions in the GIS environment. Linking regolith models, mineralisation models, geology and geophysics with a preferred exploration model to derive thematic maps is where the greatest advances in our understanding will occur. Exploration industry members are keen to see maps tailored to meet their specific needs, which can be best met by thematic regolith maps that address specific exploration models. Geochemical sampling strategy maps, or geochemical sampling media maps, can be created from basic regolith data and can be made even more robust and informative by incorporating a range of other data sets. For example, in Western Australia regolith image maps showing gold occurrences as point data, and regolith polygons coded as saprolite (in situ weathered rock) derived from either greenstone or granite (of significance for gold exploration), superimposed on a backdrop of a Total Magnetic Intensity (TMI) image derived from airborne magnetics data are being produced by Geoscience Australia (Craig and Wilford 1995; Craig 1995) (Figure 11.10, page 169). Digital regolith maps can be created employing a number of data layers. For example, a Landsat TM scene may be processed to highlight Fe oxides, clays and silica. This product can be used as a backdrop on which to drape regolith polygons selected to highlight Fe oxide-dominant geochemical sampling media (see Chapter 13) associated with greenstones. The resulting image may then be applied as a draped layer over a detailed digital terrain model for an area considered to be prospective for gold on the basis of, for example, regional geochemical results. The geochemical data can be added, and then presented in ARCVIEW© as
‘hot-linked’ point data along with ‘hot-linked’ graphic logs extracted from the results of drilling programs, and landscape photographs of the field sites visited. Increasingly, the limitation of paper maps for presenting the complex combinations arising from multidata analysis is being recognised. Individual paper copies can be printed at any desired step during the multi-layer integration process. Predictive soil mapping (PSM) is the development of a numerical or statistical model of the relationship among environmental variables and soil properties, which is then applied to a geographic database to create a predictive map (Scull et al. 2003). Bui et al. (1999, see also Bui 2004) point out that soil surveys are expensive, labour-intensive, and time-consuming. Moreover, the mapping rules and mental models used by soil surveyors to predict soil distribution in the landscape are not usually recorded, and are difficult to quantify. The same applies to regolith mapping. However, as with soil mapping, it should be possible to develop a set of decision rules that would allow the use of a variety of spatial data layers to be used in a GIS environment to produce a regolith map. This was carried out by John Wilford (pers comm.) in the Ebagoola area of Cape York Peninsula (Figure 11.11, page 170). In soil mapping, there is considerable effort being put into mapping soils characteristics using digital mapping and pedotransfer functions (for example, Minasny and McBratney 2000; Pachepsky and Rawls 2003; Lagacherie et al. 2007). This involves determining the associations between the attribute to be mapped (such as hydraulic conductivity) and other attributes about which there are more data (such as particle size distribution). These ideas are being transferred to regolith mapping. For example, Laffan and Lees (2004) discuss the utility of mapping the regolith as continuously varying fields of properties using environmental correlation, rather than as discrete entities (classes). In another example, Thwaites (2007) discusses the application of Regolith-Catenary Units to predicting soil and regolith attributes. This is certainly a way of the future, but it requires very firm knowledge about the relationship between regolith properties and land surface characteristics, and application at appropriate scales.
Regolith description and mapping
Three-dimensional Regolith Maps
Regolith consists of three-dimensional (3D) bodies of material, and therefore it is important to provide some understanding of the 3D distribution of regolith materials. The regolith toposequences mentioned above go some way towards this, but true 3D mapping of regolith is still in its infancy, and should not be confused with simply draping regolith information over a digital terrain model. Contours showing the depth to a particular layer, or the thickness of regolith units (for example, Robertson et al. 2001), are also not true 3D displays, but rather 2½D. Some areas that have been intensively drilled have allowed 3D regolith maps and models to be produced. In the Lower Balonne area a combination of AEM and drilling allowed the base of the transported regolith to be mapped (Figure 11.12, page 171). The Yandal Greenstone Belt in Western Australia is another area that has been intensively drilled, and the 3D distribution of various regolith materials has been illustrated by Anand (2000).
11.7 USER GROUPS Defining the purpose of regolith maps is important because the purpose will, at least to an extent, dictate the character of the map. As Bryan (2006) points out for ecological maps, a regolith map may be general purpose, in that it is aimed at no specific user group, but rather attempts to provide information of a general nature that will be useful for most users. On the other hand, a map aimed at a specific user group will be narrower in the amount of information portrayed on the map face and in the accompanying notes. Many examples have already been mentioned in passing, and some more specific examples are now briefly discussed. 11.7.1 The mineral exploration industry The impetus for regolith mapping in Australia was developed from the requirements of the mineral exploration industry (Murrell 1983). As areas of outcrop and shallow regolith cover became intensively explored, the mineral exploration industry turned to areas of deep regolith cover. This encouraged industry-funded research into the geochemistry of weath-
ering and its application to mineral exploration. CSIRO Exploration and Mining carried out multiclient research projects through AMIRA (Australian Mineral Industries Research Association), beginning in the 1980s. These had a significant impact on the mineral exploration industry, not only on the way they explore, but also on their success rate. The formation of CRC LEME in 1995 was also based on the need for continuing research into regolith and mineral exploration; much of this research is referenced above. Geoscience Australia and Australian state geological surveys have also aimed their own regolith mapping programs at the mineral exploration industry (for example, Taylor and Joyce 1996; Kojan and Faulkner 1994; Craig 2001). As exploration models evolve, and the roles played by geomorphology and landscape evolution in the development of the regolith are better understood, the need to re-process geochemical and regolith data in different ways will become an essential activity. The in-built flexibility that comes with having a digital map environment will prove to be of enormous significance to the exploration industry not only because of the relative ease of database interrogation and thematic map construction but also because a number of models can be examined in a shorter time span. New regolith and landscape evolution models, together with geochemistry data, can be tested from linked attribute databases. The results may well reduce the need for new, more extensive field programs. Further collaborative work with industry is vital to develop mineral exploration-oriented, more-specialised, thematic regolith maps. 11.7.2 Environmental applications of regolith maps and information By their very nature, regolith maps contain valuable information for a number of natural resource management issues. The comprehensive Cape York Peninsula survey, for example, was funded in large part by Federal and Queensland State governments as part of the Natural Resource Assessment Program for the Cape York Peninsula Land Use Strategy (Pain et al. 1995). The resulting map and report were used as part of the background for land-use planning in the Peninsula. Polygons produced for the regolith map formed
299
300
Regolith Science
the basis for a soil map of the same area. Similarly, the Bathurst regolith–landform map (Chan 1995) has been used by the New South Wales Department of Land and Water in their soil landscape mapping program. More recently, research by CRC LEME has demonstrated that regolith information can characterise areas affected by dryland salinity, and give information on the movement of water through the regolith (Lawrie et al. 2005, 2006; Pain et al. 2003). A regolith– landform map was also an important input into the assessment of dryland salinity and groundwater characteristics in the Lower Balonne area, Qld (Kernich et al. 2004). Much more detail of this kind of work is provided in Wilford et al. (2007). 11.7.3 Modellers Webb and Lilburne (2005), who are concerned with soil maps, note that there can be considerable uncertainty in map unit composition, and that this results in considerable spatial variability in soil properties within map units. The same is true for regolith maps – especially those covering regional or national areas. Yet these regolith maps are frequently used in computer models developed to test impacts of changing land use, or to estimate various physical parameters. This is an emerging research area in regolith mapping and, as noted above, will rely heavily on a robust understanding of the relationships between landforms, geomorphic processes and evolution, and regolith characteristics.
In addition to aerial photographs, there is a broad and increasing range of orbiting satellite and airborne remote-sensing systems with increasingly higher resolutions, ranging from those that rely on reflected energy (such as Hymap®) to active systems that can image regolith materials at depths up to 100 m below the ground surface. Regolith–landform mapping is now seen as an essential part of both mineral exploration and natural resource management. Increasingly a regolith–landform map is seen as a basic data layer for any survey area – whatever the purpose of the survey. For regolith–landform mapping, geoscientists thus require knowledge of geomorphology and the relationships between regolith and landforms. They also require familiarity with remote-sensing techniques, and to be conversant with GIS and database concepts. Moreover, it requires skill to read a regolith–landform map, in the same way as skills are required to read a geology map. This means that some training in landforms and regolith is required before full use can be made of regolith–landform maps. This chapter covers a large topic and for this reason touches only lightly on some aspects of regolith mapping. The references will, however, lead readers to more details on many aspects of the subject. It is worth noting that Australia leads the way with regolith– landform mapping and that the concepts and methods discussed in this chapter can be applied worldwide and even during the exploration of planets and other bodies in the Solar System (see Chapter 14; Clarke and Pain, 2004).
11.8 CONCLUSIONS Regolith materials are much more closely associated with landforms – both relict and present – than are bedrock materials. This means that regolith mapping is not simply another form of geological mapping, and different methods are required at both regional and local scales. The landscape mapping methods discussed in this chapter provide a suitable backdrop to regolith mapping. A landscape approach to regolith mapping similar to that used for mapping land systems and soils is therefore entirely appropriate. Many regolith–landform maps have been produced using remote sensing supported by field observations.
11.9 REFERENCES Anand RR (2000). Regolith and geochemical synthesis of the Yandal Greenstone belt. Australian Institute of Geoscientists Bulletin 32, 79–111. Anand RR and Paine M (2002). Regolith geology of the Yilgarn Craton, Western Australia: implications for exploration. Australian Journal of Earth Sciences 49, 3–162. Anand RR, Smith RE, Innes J and Churchward HM (1989a). ‘Exploration geochemistry about the Mt Gibson gold deposits, Western Australia’. CSIRO Division of Exploration Geoscience Restricted
Regolith description and mapping
Report 20R. (Reissued as CRC LEME Open File Report 35, 1998. CRC LEME, Perth.) Anand RR, Smith RE, Innes J, Churchward HM, Perdrix JL and Grunsky EC (1989b). ‘Laterite types and associated ferruginous materials, Yilgarn Block, WA. Terminology, classification and atlas’. Report 60R. CSIRO Division of Exploration Geoscience, Perth. Anand RR, Churchward HM, Smith RE, Smith K, Gozzard JR, Craig MA and Munday TJ (1993). ‘Classification and atlas of regolith-landform mapping units – exploration perspectives for the Yilgarn Craton’. CSIRO Division of Exploration Geoscience Restricted Report 440R. (Reissued as CRC LEME Open File Report 2, 1998. CRC LEME, Perth.) Anand RR, Paine MD and Smith RE (2002). ‘Genesis, classification and atlas of ferruginous materials, Yilgarn Craton’. Open File Report 73. CRC LEME, Perth. Bibby LM and Radojkovic AM (2002). ‘Ballarat-Creswick special 1:50,000 regolith-landform map’. Geological Survey of Victoria, Melbourne. Bradley JJ and Storey JM (1995). ‘Leonora, W.A. (preliminary edition)’ 1:250,000 Regolith Materials Series,. Geological Survey of Western Australia, Perth. Brough DM, Claridge J and Grundy MJ (2006). ‘Soil and landscape attributes: a report on the creation of a soil and landscape attribute information system for Queensland Natural Resources, Mines and Water, Brisbane, Australia.’ QNRMW, Brisbane. Brown DJ, Clayton MK and McSweeney K (2004). Potential terrain controls on soil color, texture contrast and grain-size deposition for the original catena landscape in Uganda. Geoderma 122, 51–72. Bryan BA (2006). Synergistic techniques for better understanding and classifying the environmental structure of landscapes. Environmental Management 37, 126–140. Bui EN (2004). Soil survey as a knowledge system. Geoderma 120, 17–26. Bui EN, Loughhead A and Corner R (1999). Extracting soil-landscape rules from previous soil surveys. Australian Journal of Soil Research 37, 495–508.
Chan RA 1988. Regolith terrain mapping for mineral exploration in Western Australia. Zeitschrift für Geomorphologie Supplementbände 68, 205–221. Chan RA (1995). ‘Bathurst Regolith-Landforms. 1:250,000 scale map’. Australian Geological Survey Organisation, Canberra. Chan RA, Craig MA, D’Addario GW, Gibson DL, Ollier CD and Taylor G (1986). ‘The regolith terrain map of Australia 1:5,000,000’. Record 1986/27. Bureau of Mineral Resources, Geology and Geophysics, Canberra. Chan RA, Craig MA, Hazell MS and Ollier CD (1992). ‘Kalgoorlie regolith terrain map commentary, sheet SH51, Western Australia’, 1:1,000,000 Regolith Series. Record 1992/8. Australian Geological Survey Organisation, Canberra. Chan RA, Greene RSB, Hicks M, Le Gleuher M, McQueen KG, Scott KM and Tate SE (2004). ‘Regolith architecture and geochemistry of the Byrock area, Girilambone region, north-western NSW’. Open File Report 159. CRC LEME, Perth. Christian CS and Stewart GA (1953). ‘General report on survey of Katherine-Darwin region, 1946’. CSIRO Land Research Series 1, CSIRO, Melbourne. Churchward HM, Butler IK and Smith RE (1992). ‘Regolith–landform relationships in the Bottle Creek orientation study, Western Australia’. CSIRO Division of Exploration Geoscience Restricted Report 247R. (Reissued as CRC LEME Open File Report 51, 1998. CRC LEME, Perth.) Clarke JDA and Pain CF (2004). From Utah to Mars: regolith-landform mapping and its application. In Martian Expedition Planning. (Ed. CS Cockell) pp. 131–159, American Astronautical Society Science and Technology Series 107, American Astronautical Society, Springfield, Virginia Conacher AJ and Dalrymple JB (1977). The nine-unit land surface model: an approach to pedogeomorphic research. Geoderma 18, 1–154. Cooke RU and Doornkamp JC (1990). Geomorphology in Environmental Management. 2nd edn. Clarendon Press, Oxford. Craig MA (1995). ‘Sir Samuel – regolith landforms image map, 1:250,000 regolith landforms map’. Australian Geological Survey Organisation, Canberra.
301
302
Regolith Science
Craig MA (2001). Regolith mapping for geochemical exploration in the Yilgarn Craton, Western Australia. Geochemistry: Exploration, Environment, Analysis 1, 383–390. Craig MA (2006). ‘Regolith map of the Northern Territory, 1:2,500,000 scale’. Northern Territory Geological Survey, Darwin. Craig MA and Chan RA (1992). ‘RTMAP BMR regolith database – Kalgoorlie Unit Data Surficial Geology Series’. Record 1992/32. Bureau of Mineral Resources, Geology and Geophysics, Canberra. Craig MA and Churchward HM (1995). ‘Wiluna regolith landforms. 1:250,000 Regolith Landforms Map’. Australian Geological Survey Organisation, Canberra. Craig MA and Wilford J (1995). ‘Leonora regolith landforms image map. 1:250,000 Regolith Landforms Map’. Australian Geological Survey Organisation, Canberra. Craig MA, Anand RR, Churchward HM, Gozzard JR, Smith RE and Smith K (1993). ‘Regolith-landform mapping in the Yilgarn Craton, Western Australia: towards a standardised approach’. CSIRO Division of Exploration Geoscience Restricted Report 338R. (Reissued as CRC LEME Open File Report 72, 1999. CRC LEME, Perth.) Dalrymple JB, Blong RJ and Conacher AJ (1969). An hypothetical nine unit landsurface model. Zeitschrift für Geomorphologie 12, 60–76. Dehaan RL and Taylor GR (2002). Field-derived spectra of salinized soils and vegetation as indicators of irrigation-induced soil salinization. Remote Sensing of Environment 80, 406–417. Dickson BL and Scott KM (1997). Interpretation of aerial gamma-ray surveys – adding the geochemical factors. AGSO Journal of Australian Geology and Geophysics 187-200. Dickson BL, Fraser SJ and Kinsey-Henderson A (1996). Interpreting aerial gamma-ray surveys utilising geomorphological and weathering models. Journal of Geochemical Exploration 57, 75–88. Eggleton RA (Ed.) (2001). The Regolith Glossary: Surficial Geology, Soils and Landscape. CRC LEME, Canberra and Perth. Foster KA, Shirtliff GJ and Hill SM (2000). ‘Balaclava 1:25,000 regolith landform map, sheet 7133-1-5’.
Australian Geological Survey Organisation, Canberra. Fulton RJ (Compiler) (1995). ‘Surficial materials of Canada’. Geological Survey of Canada Map 1880A, scale 1:5,000,000. Geological Survey of Canada, Ottawa. Gallant JC and Dowling TI (2003). A multiresolution index of valley bottom flatness for mapping depositional areas. Water Resources Research 39, 1347–1360. Galloway RW, Gunn RH, Pedley L, Cocks KD and Kalma JD (1974). ‘Lands of the Balonne-Maranoa area, Queensland’. CSIRO Land Research Series 34. CSIRO, Melbourne. Gibson DL and Wilford JR (1995). ‘Broken Hill regolith landforms. 1:500,000 scale map’. CRC LEME, Perth and Canberra. Gozzard JR (2006). ‘Image processing of ASTER multispectral data’. Record 2006/9. Western Australia Geological Survey, Perth. Gunn RH (1967). A soil catena on denuded laterite profiles in Queensland. Australian Journal of Soil Research 5, 117–132. Gustavsson M, Kolstrup E and Seijmonsbergen AC (2006). A new symbol-and-GIS based detailed geomorphological mapping system: Renewal of a scientific discipline for understanding landscape development. Geomorphology 77, 90–111. Hall GJ and Marnham JR (2002). ‘Regolith–landform resources of the Karridale–Tooker and Leeuwin 1:50,000 sheets’. Record 2002/10. Western Australia Geological Survey, Perth. Hocking RS, Langford RL, Thorne AM, Sanders AJ, Morris PA, Strong CA and Gozzard JR (2001). ‘A classification system for regolith in Western Australia’. Record 2001/4. Geological Survey of Western Australia, Perth. Hocking RM, Langford RL, Thorne AM, Sanders AJ, Morris PA, Strong CA, Gozzard JR and Riganti A (2005). ‘A classification system for regolith in Western Australia – an update’. Record 2005/10. Western Australia Geological Survey, Perth. Hocking RM, Langford RL, Thorne AM, Sanders AJ, Morris PA, Strong CA, Gozzard JR and Riganti A (2007). ‘A classification system for regolith in Western Australia (March 2007 update)’. Record
Regolith description and mapping
2007/8. Western Australia Geological Survey, Perth. Johnston RM, Barry SJ, Bleys E, Bui EN, Moran CJ, Simon DAP, Carlile P, McKenzie NJ, Henderson BL, Chapman G, Imhoff M, Maschmedt D, Howe D, Grose C, Schoknecht N, Powell B and Grundy M (2003). ASRIS: the database. Australian Journal of Soil Research 41, 1021–1036. Jutson JT (1914). ‘An outline of the physiographical geology (physiography) of Western Australia’. Bulletin 61, Geological Survey of Western Australian, Perth. Jutson JT (1934). ‘The physiography (geomorphology) of Western Australia, 2nd edn’. Bulletin 95, Geological Survey of Western Australia, Perth. Kernich A, Pain CF, Kilgour P and Maly B (2004). ‘Regolith landforms in the Lower Balonne area, Southern Queensland, Australia’. Open File Report 161. CRC LEME, Perth. Kojan CJ and Faulkner JA (1994). ‘Geochemical mapping of the Menzies 1:250,000 Sheet’. Explanatory Notes, Geological Survey of Western Australia, Perth. Laffan SW and Lees BG (2004). Predicting regolith properties using environmental correlation: a comparison of spatially global and spatially local approaches. Geoderma 120, 241–258. Lagacherie P, McBratney AB and Voltz M (Eds) (2007). Digital Soil Mapping: an Introductory Perspective, Developments in Soil Science. Vol. 31. Elsevier, Amsterdam. Lawrie K, Wilford J and Pain C (2006). Value adding to GFS frameworks for managing dryland salinity in Australia. Focus on Salt 37, 14–15. Lawrie KC, Pain C Gray M, Fitzpatrick A and Clarke J (2005). Salinity and groundwater mapping: a multi-scale, hierarchical approach to identify and map key functional elements in Australia’s complex depositional regolith landscapes. Geophysical Research Abstracts 7, 05780. Lenz SL (1991). ‘RTMAP, BMR regolith database user’s manual’. Record 1991/30. Bureau of Mineral Resources, Geology and Geophysics, Canberra. Lenz SL and Pain CF (1992). Categorising descriptive data with reference to the regolith and environmental geoscience databases. In Geographic Infor-
mation Systems, Cartographic and Geoscience Data Standards, Workshop Proceedings. pp. 173–179. Record 1992/27. Bureau of Mineral Resources, Geology and Geophysics, Canberra. Marnham J and Morris PA (2003). A seamless digital regolith map of Western Australia: a potential source for mineral exploration and environmental management. Geological Survey of Western Australia Annual Report 2002-2003, 27–33. Martin-Duque JF, Pedraza J, Sanz MA, Bodoque JM, Godfrey AE, Diez A and Carrasco RM (2003). Environmental assessment; landform classification for land use planning in developed areas: an example in Segovia Province (Central Spain). Environmental Management 32, 488–498. McDonald RC, Isbell RF, Speight JG, Walker J and Hopkins M (1990). Australian Soil and Land Survey Field Handbook. 2nd edn. Inkarta Press, Melbourne. McKenzie NJ, Jacquier DW, Maschmedt DJ, Griffin EA and Brough DM (2005). ‘The Australian Soil Resource Information System Technical Specifications, Version 1.5’. Australian Collaborative Land Evaluation Program, on behalf of the National Committee on Soil and Terrain, 89 pp. <www. asris.csiro.au> Australian Collaborative Land Evaluation Program, Canberra McKenzie RC, George RJ, Woods SA, Cannon ME and Bennett DL (1997). Use of the electromagnetic-induction meter (EM38) as a tool in managing salinisation. Hydrolgeology Journal 5, 37–50. Milne G (1935). Some suggested units of classification and mapping, particularly for east African soils. Soil Research 4, 183–198. Minasny B and McBratney AB (2000). Evaluation and development of hydraulic conductivity pedotransfer functions for Australian soil. Australian Journal of Soil Research 38, 905–926. Murrell B (1983). The regolith in economic geology, keynote address. In Abstracts, Regolith Symposium, 17 November, Canberra, Australia. Ollier CD (1977). Terrain classification: methods, applications and principles. In Applied Geomorphology. (Ed. JR Hails) pp. 277–316. Elsevier, Amsterdam.
303
304
Regolith Science
Ollier CD (1995). An unreliable history of regolith mapping. Centre for Australian Regolith Studies Occasional Paper 3, 1–16. Ollier CD and Galloway RW (1990). The laterite profile, ferricrete and unconformity. Catena 17, 97–109. Ollier CD and Joyce EB (1986). ‘Regolith terrain units of the Hamilton 1:1,000,000 Sheet Area, Western Victoria’. Record 1986/33. Bureau of Mineral Resources, Geology and Geophysics, Canberra. Ollier CD and Pain CF (1996). Regolith, Soils and Landforms. John Wiley & Sons, Chichester, UK. Pachepsky YA and Rawls WJ (2003). Soil structure and pedotransfer functions. European Journal of Soil Science 54, 443–451. Pain CF (1992). The landscape approach to mapping regolith. In Surface with Geology: Australian Landscapes and Economic Implications, The Fifth Edgeworth David Day Symposium. (Eds DF Branagan and KL Williams) pp. 45–60. Department of Geology and Geophysics, University of Sydney. Pain CF (2005). Size does matter: relationships between image pixel size and landscape process scale. In Abstracts, MODSIM 2005 International Congress on Modelling and Simulation. December, Melbourne. (Eds A Zerger and RM Argent) pp. 1430–1436. Modelling and Simulation Society of Australia and New Zealand, Canberra. Pain CF and Ollier CD (1992). Ferricrete in Cape York Peninsula, North Queensland. BMR Journal of Australian Geology and Geophysics 13, 207–212. Pain CF and Ollier CD (1995). Regolith stratigraphy: principles and problems. AGSO Journal of Australian Geology and Geophysics 16, 197–202. Pain CF, Chan R, Craig M, Hazell M, Kamprad J and Wilford J (1991). ‘RTMAP BMR regolith Database Field Handbook’. Record 1991/29. Bureau of Mineral Resources, Geology and Geophysics, Canberra. Pain CF, Wilford JR and Dohrenwend JC (1994). ‘Regolith-landforms of the Ebagoola 1:250,000 sheet area, (SD54-12), North Queensland’. Record 1994/7. Australian Geological Survey Organisation, Canberra. Pain CF, Wilford JR and Dohrenwend JC (1995). ‘Regolith-terrain mapping of Cape York Peninsula’. Cape York Peninsula Land Use Strategy, Office of
the Coordinator-General of Queensland, Brisbane, Department of the Environment, Sport and Territories, Canberra, and Australian Geological Survey Organisation, Canberra. Pain CF, Wilford JR and Lawrie K (2003). The role of geomorphology in assessing salt and water movement in upland landscapes. In Integrated Catchment Management Conference Proceedings, Sydney 26-28 November, Theme 1 Paper 1. (on CD). Australian Water Association, Sydney. Pain CF, Chan R, Craig M, Gibson D, Kilgour P and Wilford J (2007). ‘RTMAP regolith database field book and users guide’. (2nd edn). Open File Report 231. CRC LEME, Perth. Papp É (Ed.) (2002). ‘Geophysical and remote sensing methods for regolith exploration’. Open File Report 144. CRC LEME, Perth. Phipps PJ (2001). Terrain systems mapping. In Land Surface Evaluation for Engineering Practice. (Ed. JS Griffiths) pp. 59–61. Engineering Geology Special Publications 18, Geological Society, London. Radojkovic AM and Bibby LM (2003). ‘The regolith of the Ballarat-Creswick area’. Victorian Initiative for Minerals and Petroleum Report 76, Department of Primary Industries, Melbourne. Rattenbury M, Craig MA, Oversby B and Whitaker A (1992). ‘Adding layers towards a GIS data package in Geographic Information Systems, Cartographic and Geoscience Data Standards’. Record 1992/27. Australian Bureau of Mineral Resources, Geology and Geophysics, Canberra. Riganti A, Groenewald PB, and McCabe M (2003). ‘East Yilgarn Geoscience Database – a 1:100,000 reinterpretation of the Eastern Goldfields regolith’. Record 2003/11. Western Australia Geological Survey, Perth. Robertson IDM and Butt CRM (1993). ‘Atlas of weathered rocks’. CSIRO Division of Exploration Geoscience Restricted Report 390R. (Reissued as CRC LEME Open File Report 1, 1997. CRC LEME, Perth.) Robertson IDM, Craig MA and Anand RR (2006). ‘Atlas of regolith materials of the Northern Territory’. Open File Report 196. CRC LEME, Perth. Robertson IDM, King JD, Anand RR and Butt CRM (2001). ‘Regolith geology and geochemistry, Mt
Regolith description and mapping
Magnet District, Geochemical orientation studies, Stellar and Quasar deposits, Volumes 1 and 2’. Open File Report 92. CRC LEME, Perth. Sanders AJ, Davy R, Pirajno F and Morris PA (1998). Regolith geochemical mapping as an adjunct to geological mapping and mineral exploration. Geological Survey of Western Australia Annual Review 1997–98, 104–111. Scull P, Franklin J, Chadwick OA and McArthur D (2003). Predictive soil mapping: a review. Progress in Physical Geography 27, 171–197. Skwarnecki M (2005). ‘Regolith–landform mapping and hyperspectral data for the Kalgoorlie– Kanowna area’. Record 2005/7. Western Australia Geological Survey, Perth. Soller DR, and Reheis MC (Compilers) (2004). ‘Surficial materials in the conterminous United States’. United States Geological Survey Open-file Report 03-275, scale 1:5,000,000. USGS, Reston, Virginia Speight JG (1988). Land classification. In Australian Soil and Land Survey Handbook, Guidelines for Conducting Surveys. (Eds RH Gunn, JA Beattie, RE Reid and RHM van der Graaff) pp. 38–59. Inkarta Press, Melbourne. Stewart GA (Ed.) (1968). Land Evaluation. Macmillan, Melbourne. Street GJ and Anderson A (1993). Airborne electromagnetic surveys of the regolith. Exploration Geophysics 24, 795–800. Taylor DH and Joyce EB (1996). ‘Ballarat 1:100,000 regolith-exploration map report’. Technical Record 1996/4. Geological Survey of Victoria, Melbourne. Taylor G and Eggleton RA (2001). Regolith Geology and Geomorphology. John Wiley and Sons, Chichester, UK. Taylor G and Shirtliff G (2003). Weathering: cyclical or continuous? An Australian perspective. Australian Journal of Earth Sciences 50, 9–17. Thwaites RN (2007). Conceptual and digital soil-landscape mapping using regolith-catenary units, In Digital Soil Mapping: an Introductory Perspective.
(Eds P Lagacherie, AB McBratney and M Voltz) pp. 257–268. Developments in Soil Science 31, Elsevier, Amsterdam. Webb TH and Lilburne LR (2005). Consequences of soil map unit uncertainly on environmental risk assessment. Australian Journal of Soil Research 43, 119–126. Wielemaker WG, de Bruin S, Epema GF and Veldkamp A (2001). Significance and application of the multi-hierarchical landsystem in soil mapping. Catena 43, 15–34. Wildman JE and Compston D (2000). Magnetic expression of palaedrainage systems in the Yandal greenstone belt: implications for exploration. Australian Institute of Geoscientists Bulletin 32, 135–144. Wilford JR and Craig MA (1997). ‘Half Moon Lake regolith landforms (1:100,000 map)’. CRC LEME, Perth. Wilford JR, Craig MA, Tapley IJ and Mauger AJ (2001). ‘Regolith-landform mapping and its implicatins for exploration over the Half Moon Lake region, Gawler Craton, South Australia’. Open File Report 80. CRC LEME, Perth. Wilford JR, James J and Halas L (2007). ‘Upper Loddon groundwater flow system (GFS) map (1:100,000)’. Restricted Report 273R. CRC LEME, Perth. Wilford JR (2002). ‘Regolith-landform characteristics, evolution and implications for exploration over the Selwyn Region, Mt Isa’. Open File Report 131. CRC LEME, Perth. Wilford JR, Bierworth PN and Craig MA (1997). Application of airborne gamma-ray spectrometry in soil/regolith mapping and applied geomorphology. AGSO Journal of Australian Geology and Geophysics 17, 201–216. Worrall L, Munday TJ and Green AA (1999). Airborne electromagnetics – providing new perspectives on geomorphic processes and landscape development in regolith-dominated terrains. Physics and Chemistry of the Earth (A) 24, 855–860.
305
12
Soils and natural resource management Robert W Fitzpatrick
12.1 INTRODUCTION
12.2
This chapter examines the role of soil processes in forming saline, sodic, acid, eroded and acid sulfate soils. (The implications of soil processes in mineral exploration are discussed in Section 13.2.) Simplified schematic diagrams are used to illustrate the major soil–regolith processes involved and how specific soils susceptible to land degradation may be recognised. The chapter discusses differences between: (i) groundwater-associated salinity (GAS) or primary, secondary and seepage salinity and non-groundwater-associated salinity (NAS) and (ii) coastal and inland acid sulfate soils (ASS). It also highlights how coloured cross-sectional diagrams and photographs of soil–regolith and vegetation help local community groups to understand complicated scientific processes and terminology, and how this information can be used to underpin best management practices for saline, sodic, acid, eroded and acid sulfate soils. This chapter therefore has two principal objectives:
Regolith processes leading to the development of soils have been described in previous chapters (especially Section 6.2). Soils mean different things to different people. Soil scientists (pedologists) view soils as being made up of different size mineral particles (sand, silt and clay) and organic matter. Soils have complex biological, chemical, physical, mineralogical and hydrological properties that are always changing with time. Agronomists, farmers and gardeners, on the other hand, see soil as a medium for growing crops, pastures and plants – primarily in the top 50 cm of the Earth’s surface. Engineers regard soil as material to build on and excavate, and are usually concerned primarily with moisture conditions and the capacity for soil to become compacted and support structures. All agree, however, that soils are strongly affected during land degradation (for example, McKenzie et al. 2004) and, furthermore, that pedological approaches are crucial to understanding land-degradation processes, and how knowledge of these processes contribute to effective NRM. Identification of soil differences by using various soil attributes is the first step in using soil information to help NRM staff and environmental investigators at land-degraded sites and polluted sites, respectively. Unfortunately, pedologists often use quite difficult
s s
to review some established concepts and standard terminologies used in pedology that have practical relevance to natural resource management (NRM) to provide brief examples/case studies of the use of some pedological and related regolith and hydrological methods in NRM.
SOIL–REGOLITH PROCESSES
308
Regolith Science
and convoluted terminology in soil classification (taxonomy) and soil mapping, which are often hard to understand or will have little apparent relevance in NRM investigations. Pedological terminology on maps and in reports involves many unique terms that are difficult to understand for the ‘non-pedologist’. However, work in the field and in the laboratory traditionally carried out by pedologists involves an assessment of a wealth of mainly soil morphological features that can readily be interpreted to show soil–regolith processes, and so allows soil information to underpin best management practices for salinity, sodicity, acidity, erosion and acid sulfate soils. This applied aspect of the work of pedologists is often obscured by pre-occupation with using different national and international soil classification systems. However, in recent years pedologists have developed several user-friendly special-purpose classification systems, catering for example, for the following variety of practical purposes: 1. engineering applications (such as optical fibre cable and pipe installations) 2. mine sites and rehabilitation 3. soils used for viticulture and forestry 4. saline and acid sulfate soils (including supporting land planning policy and jurisdiction) 5. topdressing materials (such as clay spreading and liming) 6. urban planning 7. mineral exploration (for example, Fitzpatrick et al. 2003a). These special-purpose classification systems all apply traditional soil-assessment criteria and supply recommendations for soil-management practices to end-users. Typically, maps are used to provide pictorial representations of the distribution of soils, and each map varies according to the specific soil classification scheme used. As the mapping scale changes from fine scale (for example, 1:1000 – or paddock or landscape scale) to coarser scales (for example, 1:250 000 – or regional scale, and coarser ), the soil information contained in the map becomes more and more generalised, and so soil support for NRM purposes also become more generalised. Hence map units at scales coarser than 1:50 000 are not able to represent a single
kind of soil within landscapes (for example, Dent and Young 1981). As such, applying soil maps at the coarse scales described – especially for saline, sodic, acid, eroded and acid sulfate soils – should only be used for providing information for broad considerations. The reliability of such soil assessments depends on the density and quality of soil data collected at finer scales ( Section 11.5.3).
12. 3 APPROACHES AND METHODS FOR MAKING COMPARISONS BETWEEN SOIL SAMPLES The following list of six key soil morphological descriptors has been compiled from standard techniques used in soil science (McDonald et al. 1990; Schoeneberger et al. 2002) for assessing the soil properties for NRM. These include observations of depth changes in colour, consistence, texture, structure, segregations/coarse fragments (carbonates and ironstone) and abundance of roots in the different layers or horizons (Fitzpatrick et al. 1999). Morphological descriptors are useful in assessing soil conditions because:
s
s
they are rapid field and laboratory assessments, and may serve as substitutes for other methods, such as mineralogy (see below) and geochemistry, which are generally complex and more costly to carry out they can be used in research to evaluate causes for variation in soil condition induced by anthropogenic activities, land management, hydrology and weather conditions.
12.3.1 Soil colour Soil colour is usually the first property recorded in a morphological description of soils (and may be the only feature of significance to a layperson). Colour provides an indicator of redox status because soil colour depends upon the type of Fe oxides present and the soil’s position in the landscape relates to soil aeration and organic matter content (Bigham and Ciolkosz 1993; Fitzpatrick et al. 1999; Bigham et al. 2002). Ideally, soil colour should be determined on dry and moist samples using Munsell soil colour notation (Munsell Soil Color Charts 1994). This objective notion of soil colour uses three coordinates: hue
Soils and natural resource management
Hue (Colour based on spectral frequency) (usually red-yellow)
Munsell Soil Colour
Value (Lightness: white-black) (Range of values 1-9)
Chroma (Colour saturation: intensity of hue) (Range of values 1-20)
Figure 12.1: Components of the Munsell soil colour notation.
(shade), value (lightness) and chroma (intensity) (Figure 12.1). Hue is the colour (that is, spectral wavelength), which in most soils ranges from red to yellow. Value or tone refers to the lightness in the range from white to black, and chroma defines the degree of colour Relative accumulation of Fe and Al
saturation or intensity of hue. Red soil matrices are generally described with hues 5 YR or redder (and chroma greater than 1), ‘reddish’ with hues 7.5 YR (and chroma greater than 1) and yellow with hues 7.5 YR or yellower. Dark colours have low value (<3) and low chroma (<2). Training is recommended before consistent colour matching is made (Post et al. 1993). Uniform high chroma red and yellow colours (hues) indicate oxidising conditions, and are usually present in the upper parts of landscapes (Table 12.1, Figure 12.2). The Fe oxides dominating such aerobic soils are mainly hematite and goethite (Sections 5.4.3 and 6.2). Red soils are nearly always better drained than yellow soils. Under wet and anaerobic conditions, hematite and goethite can become soluble through action of anaerobic bacteria (Section 7.4.1). Soluble Fe2+ may move to aerobic zones either lower in the profile, or down slope to re-oxidise and form a new Fe oxide. Consequently, distinctive colour patterns are formed. In some cases, the re-oxidised Fe may form various
Note: B horizon goethites at (a) and (b) contain 20-28 and 25-30 mole % Al respectively
(a) A1 Yellowish B2 clay
Absolute accumulation of Fe
B3 Red clay
(b) A1 B
Weathered yellow sandy loam
Mottled zone
(c) A1 Grey
A2 sand Plasmic zone (kaolinite-rich)
Range of water table fluctuations
Saprolith
Unweathered Fe-rich rock Massive brick-like ferricrete (goethite 15-25% mole % Al; hematite 5-10% mole % Al) Ferricrete fragmentation Nodular ferricrete Seasonally wet soils and land surface
Sandstone
Recemented ferricrete with inclusions from upslope (goethite 15-25% mole % Al) Uniform ferruginous sandstone with hematite and goethite (10-15% mole % Al) Massive vesicular ferricrete (goethite 0-5% mole % Al) 2+ Iron (Fe ) mobilisation
Figure 12.2: Generalised cross section illustrating the multi-process formation of red–yellow–grey soils and ferricretes down a hillslope. The schematic representation demonstrates the relationship between hydrology, topography, geology and dominant soil profile horizons. It also indicates some of the different modes and forms of Fe oxide accumulation and/or transformation in relation to landscape features (after Fitzpatrick 1988; McKenzie et al. 2004).
309
310
Regolith Science
Table 12.1: Summary of the occurrence and distribution of secondary Fe oxides, sulfides, carbonates and salts in soil environments (modified from Bigham et al. 2002; see also Sections 4.4.3 and 5.4.3) Mineral
Dominant colour
Soil environment
†Landscape
Hematite [a-Fe2O3]
Red (< 0.5 mm) Reddish-purple (>0.5 mm nodules, mottles, ferricretes)
Aerobic soils of the tropics, subtropics, arid/ semi-arid zones, and Mediterranean climates; greater amounts with warmer temperatures and lower organic matter contents.
Well drained upper parts
Goethite [a-FeO(OH)]
Yellow (< 0.5 mm). Strong brown (>0.5 mm nodules, ferricretes)
All weathering regimes; greater amounts with cool, wet climates (including higher altitudes and moist/cool aspects) and elevated organic matter.
Well drained upper parts and mottles in mid-slopes.
Lepidocrocite [g-FeO(OH)]
Orange (<0.5 mm) Reddish-purple (>0.5 mm)
Seasonally anaerobic, non-calcareous soils of cool-temperate climates (including higher altitudes and moist/cool aspects on mid- to lower slopes).
Seasonally wet midslopes.
Ferrihydrite [5Fe2O3.9H2O]
Reddish-brown
Soils subject to rapid oxidation of Fe in the presence of organic matter.
Seasonally wet footslopes and seeps
Maghemite [g-Fe2O3]
Brown
Highly weathered soils of the tropics and subtropics derived from mafic rocks rich in precursor magnetite and/or soils subjected to burning in the presence of organic matter.
Well drained upper parts and foot-slopes after burning.
Schwertmannite [Fe8O8 (OH) 4.6 (SO4)1.7]
Reddish-orange
Sulfuric material in acid sulfate soils of both coastal and inland areas; anthropogenic sites including mines, spoils and tailings. pH between 3.5 and 4.5.
Poorly drained footslopes, seeps and bottom lands
Jarosite [KFe3 (SO4) 2 (OH) 6] Natrojarosite [NaFe3 (SO4) 2 (OH) 6]
Pale yellow
Sulfuric material in acid sulfate soils of both coastal and inland areas; anthropogenic sites including mines, spoils and tailings. pH between 2.5 and 3.5.
Poorly drained footslopes, seeps and bottom lands.
*Sideronatrite [Na2Fe(SO4) 2. OH.3H2O]
Pale yellowish green
Sulfuric material in acid sulfate soils in mostly inland or coastal back swamp areas; anthropogenic sites including mines, spoils and tailings. pH between 2.0 and 3.5.
Poorly drained footslopes, seeps and bottom lands.
Green rust [Fe(OH) 2]
Greenish-blue
Strongly hydromorphic soils.
Poorly drained footslopes, seeps and bottom lands.
Iron monosulfides [FeS]
Black
Strongly hydromorphic and sub-aqueous soils.
As above and in rivers and lakes.
Iron disulfides or pyrite [FeS2]
Black
Strongly hydromorphic and sub-aqueous soils.
As above and in rivers and lakes.
Calcite and dolomite [CaCO3] and [CaMg(CO3) 2]
White
Calcareous soils
Low rainfall regions
Gypsum [CaSO4.2H2O]
Very pale brown
Saline soils
Low rainfall regions
Quartz [SiO2]
Light grey
Sandy soils
All landscape positions
†Occurring only in specific soil horizons or sedimentary units *Widespread occurrences in sandy and peaty sulfuric materials in South Australia (Fitzpatrick et. al., 2000b; 2008a,b).
position
Soils and natural resource management
types of ferricretes as indicated in Figure 12.2. When the Fe oxides have been fully depleted from soils, and organic matter levels are low, then the soil colours will have low chroma colours, grey and blue tints (or even, white) in an A2 horizon (Figure 12.2; Section 6.2) – indicating reducing, or waterlogged conditions. In the same soil layer, mixtures of bright red or yellow soil matrices containing dark grey or bluish blotches (mottles) indicate periodic conditions of water saturation (Vepraskas 1992; Schoeneberger et al. 2002).
3. Attempt to make a ribbon by progressively shearing the ball between thumb and forefinger, and recording the characteristics of the soil ribbon formed. The behaviour of the worked soil and the length of the ribbon produced by pressing out between thumb and forefinger characterises ten soil texture grades as shown in Table 12.2 (McDonald et al. 1990; Fitzpatrick et al. 1999). This surrogate is used to estimate:
s 12.3.2 Soil consistence Soil consistence or consistency is also called rupture resistance, and is a measure of the strength and coherence of a soil. It can be measured in the field by simply manipulating a piece of soil in the hand and determining the magnitude of force needed to cause disruption or distortion (for example, squeezing by hand, or crushing under foot). Consistence is expressed as loose, soft, firm, very hard and rigid (Schoeneberger et al. 2002). Terms used to describe consistence vary depending on the moisture content of the sample tested (for example, soft when dry versus friable when moist). Obvious factors that influence consistency include soil texture, mechanical compaction, organic matter content, cementing agents and water content. Changes in soil consistence is a useful surrogate measure for identifying restrictive layers because soil texture and structure are often not measured consistently by inexperienced researchers.
s
s
Water and nutrient retention or leaching capacity. Coarse grained sands have larger pores than those found in finer textured soils. Consequently, coarse sands are typically drained rapidly and have a poor ability to hold water and nutrients. Loamy sands hold more water and nutrients, while the available water capacity and nutrient-retention ability of clays is high. Depth to restricting layers or sub-surface compaction that may affect root growth or water movement (such as sub-surface compaction or structure decline). Sandy soils are generally more prone to sub-surface compaction than finer textured soils. Erodibility. Grain size affects susceptibility to erosion. For example, fine sand grains are more easily transported by the wind than coarser (heavier) grains which require more force to be moved. Clay
Clay 100
12.3.3 Soil texture Soil texture is determined from the relative proportions of sand (2–0.02 mm), silt (0.02–0.002 mm) and clay (<0.002 mm) in a soil (Figure 12.3). However, in the field, the field (or hand) soil texture can be determined by the following procedure: 1. Take a sample of soil sufficient to fit comfortably into the palm of the hand (separate out gravel and stones). Moisten soil with water, a little at a time, and work until it just sticks to one’s fingers and is not mushy; that is, the soil is at ‘field capacity’. 2. Continue moistening and working until there is no longer any apparent change in the ball (bolus) of soil. This usually takes 1 to 2 minutes.
clay
50 sandy clay
silty clay clay loam
sandy clay loam loam loamy sand sand
100
Sand
sandy loam
50
silty clay loam
silt loam silt
50
100
Silt
Figure 12.3: Soil texture determined using the proportions of sand, silt and clay (after Saxton et al. 1986).
311
Code
S
LS
CS
SL
L
ZL
SCL
CL
Sand
Loamy sand
Clayey sand
Sandy loam
Loam
Silty loam
Sandy clay loam
Clay loam
40–50 mm
25–40 mm
25 mm
25 mm
15–25 mm
5–15 mm
5 mm
Nil
Ribbon
Coherent and plastic
Strongly coherent
Coherent and rather spongy
Coherent and rather spongy
Coherence slight
Coherence very slight
Coherence nil to very slight
Coherence nil to very slight
Ball
Smooth to manipulate. Clay is about 30% –35%.
Sandy to touch; medium size sands grains visible in finer matrix. Clay is about 20% –30%.
As above but more silky feel
Smooth feel when manipulated but with no obvious sandiness; may be greasy to touch if organic matter is present. Clay is about 25%.
Sandy to touch. Clay is 10–20%
Cannot be moulded. Clay is 5–10%.
Cannot be moulded. Clay is 5–10%.
Cannot be moulded. Clay is < 5%.
Feel and approximate clay content
As above.
As above.
As above
Root growth of annuals and perennials is not restricted with moderate susceptibility to mechanical compaction.
Root growth of annuals and perennials is not restricted, but has a high susceptibility to mechanical compaction. Very slight restriction on water movement; soil water is available to most crops and trees. Water drains from the soil readily but not rapidly.
As above.
As above.
Minimal physical restriction to root growth for annuals and perennials but has a moderate susceptibility to mechanical compaction. No restriction on water movement but periodic soil moisture stress is common because water is drained very rapidly.
Interpretation
Interpreting soil texture from behaviour of a moist bolus (ball) (after McDonald et al. 1990 and Fitzpatrick et al. 1999).
Texture
Table 12.2:
312 Regolith Science
MC
HC
Medium clay
Heavy clay
Ribbon
>75 mm
>75 mm
50–75 mm
Ball
Smooth plastic
Smooth plastic
Plastic
Handles like stiff plasticine; can be moulded into rods without fracture; has firm resistance to ribboning shear. Clay is >55%
Handles like plasticine and can be moulded into rods without fracture; has some resistance to ribboning shear. Clay is about 45% –55%.
Smooth to touch; slight to shearing between thumb and forefinger. Clay is about 35% –40%.
Feel and approximate clay content
Texture qualifiers: used as a prefix to refine texture description: Coarse sandy Coarse to touch; sand grains can be seen with the naked eye. Fine sandy Can be felt and often heard when bolus is manipulated; sand grains seen under hand lens of 10 times magnification. Gritty More than 35% very coarse sand and very fine (1–3 mm) gravel. Gravelly 35–70% of gravel by volume. Stony 35–70% of stones by volume.
Code
LC
Texture
Light clay
Interpretation
The texture groups according to Northcote (1979): 1. The sands = sand (S), loamy sand (LS), clayey sand (CS). 2. The sandy loams = sandy loam (SL). 3. The loams = Loam (L); sandy clay loam (SCL); silty loam (ZL). 4. The clay loams = clay loam (CL). 5. The light clays = light clay (LC). 6. The medium–heavy clays = medium clay (MC), heavy clay (HC).
As above.
Root growth of most species is moderately to severely restricted but with low susceptibility to mechanical compaction. Water drains very slowly. This does not apply to self-mulching or sub-plastic clays.
Root growth of annuals and perennials is frequently restricted with moderate susceptibility to mechanical compaction. Some restriction on water movement. Soil water is available to most crops and trees. Water flow is restricted contributing to periodic waterlogging.
Soils and natural resource management 313
314
Regolith Science
particles – although light and easy to transport – are often difficult to detach because they are bound together. Consequently, a trend in the change of texture down a profile is a diagnostic property of soils frequently used to classify soils (Northcote 1979; Isbell 1996) because of the importance of texture for plant growth and water movement. 12.3.4 Soil structure Soil structure relates to the way soil particles are arranged and bound together (Schoeneberger et al. 2002). Soil structure can easily be described from the visible appearance of in situ soil in a dry to slightly moist state by the presence or absence of:
s s s s
peds (granular, lenticular, platy, blocky, polyhedral, columnar and prismatic) single grain or structureless massive slickensides (shiny, cracked or grooved clay surfaces).
The size, shape and nature of soil aggregates, peds or slickensides play a major role in determining profile hydrology and the ease of root penetration. Soil that have peds separated by irregular spaces are described as having structure. The degree and nature of structural development is largely determined by clay mineralogy and organic matter content. Peds result from the natural subdivision of the soil by fine cracks to form either small (granular or polyhedral) or large blocks (columnar, prismatic and platy). The cracks separating these peds do not usually have slickensided surfaces (see below), and ped size and development may range from weak to strong. Where peds are largely absent, the soil is described as being structureless. In a single-grained material, two-thirds of a soil is composed of individual particles, which are not bound together (loose and incoherent). In a massive material, two-thirds of the soil occurs in one large block with the particles being bound together (coherent). Slickensides are easily observable shiny planes of weakness along which movement occurs in medium to heavy clay soils that have shrink–swell properties. These are shearing faults, which exist permanently in
wet or dry expansive clays. They take the form of cracked, polished or grooved surfaces, ranging from 10 mm to 200 mm across. Slickensides often run through the soil mass in many directions and may break the structure up into bowl-shaped blocks. They can move up to 25 mm/year. Hence, the frequency and size of slickensides present can quantify the potential capacity of the soil to shrink and swell (that is, the soil develops cracks when dry). Soils or soil layers with slickensides are highly impermeable to water movement, especially when they are moist, and so root growth is restricted. 12.3.5 Segregations and coarse fragments Segregations are accumulations of distinct mineral particles such as Fe oxides (Figure 12.2; Table 12.1), calcrete and gypsum, which have formed in soil. They occur in a variety of sizes, shapes and forms and can be either soft or hard. In many parts of the world, these segregations are common and can have a major influence on soil chemical and physical properties. Fe oxide nodules may accumulate significant abundances of other elements within them (Section 5.4.3) and hence be particularly important as a sample medium during mineral exploration (Section 13.6) Calcrete commonly occurs either as masses and nodules (Section 13.7). Hydrochloric acid (1M HCl) is commonly used in the field to confirm its presence in the soil (Schoeneberger et al. 2002). The reaction between acid and soil CaCO3 causes effervescence (‘fizz’ test). Gypsum is often present as crystals that glisten in sunlight. These crystals generally occur in lower rainfall areas and often indicate high electrolyte concentrations in soils. Coarse fragments include rock fragments, strongly cemented soil materials and hard segregations, which are larger than 2 mm.
12.4 SALINE SOILS Saline (or salt-affected) soils are those with relatively large amounts of soluble salts, such as NaCl. When such soils occur naturally, this is referred to as primary salinity (that is, primary groundwater-associated salinity (GAS); see Figures 12.4; 12.5). Secondary salinity results from human activities such as irrigation or land clearing in areas that are not irrigated
Soils and natural resource management
(this is also referred to as ‘dryland salinity’) (National Land and Water Resources Audit 2001b; Ghassemi et al. 1995; Keren 2000; Tanji 2002). Both primary and secondary salinity affect plant growth by causing dehydration and toxic conditions. Before salt-affected landscapes can be managed, the type of saline land must be determined using hydrological characteristics. In addition, the category of salt-affected soil must be assessed from dominant geochemical properties (Figures 12.4 and 12.5). Saltaffected soils form under vastly different environmental conditions under the influence of diverse hydrological, morphological, geochemical, mineralogical and physical processes. Types of saline soils inlcude:
s
s
'ROUNDWATER ASSOCIATED SALINITY '!3 comprises salt-affected soils in rain-fed areas that have direct or capillary contact with saline groundwater water tables, and categories defined by the following hydrological and geochemical environments: (i) primary (natural) or secondary (anthropogenic), (ii) alkaline (Na2CO3-dominant, pH >9), (iii) halitic (NaCl-dominant), (iv) gypsic (CaSO4-dominant), (v) sulfidic (pyrite-dominant, pH >4.0), (vi) sulfuric (sulfuric acid dominant, pH <4.0), and (vi) sodic, which posess elevated exchangeable sodium percent (ESP) on clay surfaces. .ON GROUNDWATER ASSOCIATED SALINITY .!3) comprises salt-affected soils in rain-fed areas that
s
have no direct contact with saline groundwater water tables, and with categories defined by the following soil chemical environments: (i) sodic (ESP *5), and (ii) saline (electrical conductivity, ECse *2 dS/m; where SE = saturation extract) conditions in the solum (A- and B-horizons, typically <1.2 m deep). )RRIGATION ASSOCIATED SALINITY )!3 comprises salt-affected soils in irrigated areas with shallow (surface IAS) or deep (subsoil IAS) saline water tables.
Various terms are in use to describe saline land. The classification scheme presented in Figure 12.5 features a unified, process-based scheme to promote clear communication and management of types of saline land. Table 12.3 shows terms in common use in Australia. Saline soils form under different environmental conditions and thus have diverse hydrological, morphological, chemical, physical and biological properties (Figure 12.4). There is no universally accepted definition for saline soils. The definition used depends on the discipline and the type of measurements taken. For example:
s s
Hydrogeologists distinguish primary and secondary saline soils (for example, Coram et al. 2001; George et al. 1997). Plant and soil scientists use the distribution of salt-tolerant plant species and/or the approximate
Saline land
Groundwater Associated Salinity (GAS) (ECse ≥ 4 dS/m)
Non-groundwater Associated Salinity (NAS) (ECse ≥ 2 dS/m)
Solum affected Primary GAS
Secondary GAS
Surface NAS (ECse 4–60 dS/m) Figure 12.4: Categories of saline land
Irrigation Associated Salinity (IAS) (ECse ≥ 4 dS/m)
Sub-solum affected “Deep NAS” Surface IAS (ECse 20–60 dS/m)
Subsoil NAS (ECse 20–60 dS/m)
Subsoil IAS
RFf002-08
315
316
Regolith Science
Mangrove swamps
Primary GAS (natural) Groundwater Associated Salinity (GAS) Salinity process driven by groundwater
Sulfidic or gypsic
Salt pan
Halitic of sulfidic
Seepage scald
Halitic of sulfidic
Salt seepage
Halitic of sulfidic
Mangrove swamps/ drained
Secondary GAS (dryland salinity)
Saline Land
Shallow NAS Solum layers
Non-groundwater Associated Salinity (NAS) Salinity process driven by seasonal flushing of rainwater
Deep NAS Below Solum
Sulfidic
Coastal marsh
Sulfidic Sulfuric
Coastal marsh/ drained
Sulfidic Sulfuric or gypsic
Salt pan
Halitic or gypsic
Salt seepage/ drained
Halitic, sulfidic or sulfuric
Eroded seepage scald
Halitic or sulfuric
Topsoil NAS (Magnesia patches)
Halitic, sodic
Subsoil NAS
Sodic, halitic
Deep NAS (deep in the regolith usually >2m)
Sodic, halitic
Figure 12.5: Categories of saline land as defined by hydrology, soil water status and soil chemistry (after SCAV 1982; Williams and Bullock 1989; Fitzpatrick et al. 2003c). Where GAS = groundwater-associated salinity and NAS = nongroundwater-associated salinity.
s s
range of soil electrical conductivity (EC) levels to distinguish slightly, moderately or severely affected saline soils (for example, Allan 1994). Scientists in other disciplines may use: measurements of pH (3.5–8.5), exchangeable Na percentage, the Na adsorption ratio and EC to identify sodic–saline soils (Soil Survey Staff 1987)
Table 12.3:
s s s
measurements of pH (> 9), presence of Na2CO3 and high EC to distinguish alkaline saline soils pH (< 3.5 or 4), presence of sulfur and high EC to distinguish acid sulfate saline soils (Fitzpatrick et al. 1996; Fanning 2002) apparent electrical conductivty (ECa) value intensities and patterns using geophysics (Rhoades et al. 1999; Corwin and Lesch 2003).
Terminology for saline land in use in Australia.
Process-based term
Alternative term (with key reference)
Groundwater-associated salinity
Dryland salinity (Spies and Woodgate 2005); seepage salinity
Surface non-groundwater-associated Salinity
Magnesic/magnesia patches; saline scald (Herriot 1942; Kennewell 1999)
Subsoil non-groundwater-associated salinity
Transient salinity (Rengasamy 2002); dry saline land (Kennewell 1999; Maschmedt 2002)
Deep non-groundwater-associated salinity
‘Salt bulge’ (e.g. Figure 12.6)
Soils and natural resource management
0 0.5
Depth (m)
1.0
4
15
Soil Surface
Section 12.8.2). Saline soils with high amounts of Na2CO3 may also occur and are usually associated with coarse-textured soils. These saline soils often exhibit a whitish surface crust when dry. High salt concentrations (as defined by high EC) dehydrate plant cells because the dissolved salts decrease the osmotic potential of soil water. Water flows from the high osmotic potential (low salt concentration in the plant cell) to low osmotic potential (high salt concentration in soil). Thus, plants cannot extract water from soil when the soil solution has a lower osmotic potential than the plant cells. The effect on plants is similar to drought stress: causing impaired plant growth and productivity, and even death may result if excessive. For many crops, yields are reduced when the soil extract EC (ECse) reaches 4 dS/m (US Salinity Laboratory Staff 1954) and decline proportionately as EC levels increase above that concentration. Some crops, such as sugar beets, are tolerant to
No agricultural productivity
Sodicity/alkalinity Acidity
Subsoil Non-groundwater Associated Salinity (NAS) (ECse 2–16 dS/m) Surface Non-groundwater Associated Salinity (NAS) (magnesia patch) (ECse 40–60 dS/m) Secondary Groundwater Associated Salinity (GAS) (dryland salinity) (ECse >4 dS/m) Deep Non-groundwater Associated Salinity (NAS) 0 (Salt Bulge) Salt accumulation (ECse 20–60 dS/m) Primary Groundwater Associated Salinity 2 (GAS)
Depth (m)
The definition is further complicated by the fact that soil salinity may also not be associated with a permanent saline groundwater table. In Australia, most studies of salinisation processes focus on primary and secondary salinity (for example, George et al. 1997; Coram et al. 2001; Salama et al. 1999, Clarke et al. 2002) or the processes occurring in sodic soils (described below) (for example, Isbell et al. 1983; Rengasamy and Sumner 1998; Shaw et al. 1998; Fitzpatrick et al. 1994). The soluble salts found in saline soils are generally of three types: chlorides, sulfates and carbonates. Most saline soils in Australia have high amounts of chloride salts (Gunn and Richardson 1979, Isbell et al. 1983). However, in many parts of the Australia (such as the Mount Lofty Ranges, South Australia; Fitzpatrick et al. 1996; the Dundas Tablelands, Victoria; and the River Murray system) and in Iraq (Fitzpatrick 2004) extensive areas of saline soils also contain sulfate salts and sulfides at depth (see
>10 000 mg/L >40
Groundwater Salinity ~1500 mg/L RFf004-08
Figure 12.6: Schematic cross section showing various categories of saline land as defined by hydrology: (i) nongroundwater-associated salinity (NAS), or dry saline land or transient salinity, which is not hydrologically connected to a saline water table. (ii) Deep non-groundwater-associated salinity (NAS) or ‘salt bulges’, which occur well below the root zone of former native vegetation (usually > 2 m from soil surface). (iii) Primary non-groundwater-associated salinity (NAS) or primary salinity, caused by rising saline groundwater. (iv) Secondary non-groundwater-associated salinity (NAS), or secondary salinity, caused by rising saline groundwater and salt accumulation in soils due to evaporative water loss in saline seeps (after Fitzpatrick et al. 2003d).
317
318
Regolith Science
EC levels between 4 and 8 dS/m. At an EC of 16 dS/m the growth and yields of most crops are affected. Figure 12.6 shows the hydrological processes and salinity development commonly found in cropping systems in Australia. Previous reviews on saline and sodic soils have detailed their wide range of occurrences and properties (for example, Wood 1924; Peck and Hurle 1976; Macumber 1991; Nulsen 1993; Tickell 1994; Fitzpatrick et al. 1994; Naidu et al. 1995; Semple et al. 1996; Bui et al. 1998; Semple and Williams 2002; NLWRA 2001b; Heng et al. 2001; Coram et al. 2001; Rengasamy 2002; Peck and Hatton 2003). Drainage and disturbance of saline soils
Generally speaking, most drainage and disturbance of saline soils is caused by human action, though some erosion and vegetation changes can result from natural processes and periods of low recharge (drought). Processes resulting in changed groundwater levels can be summarised as follows:
s s
s
s
engineering – groundwater pumping, construction of drains (Sinclair Knight Merz 2001). IMPROVED FARMING SYSTMES USING VEGETATION – reducing recharge, lowering of water tables by using deep-rooted plants (for example, BarrettLennard et al. 2003). erosion – such as local lowering of the water table by gully formation and deepening; removal of surface soil layers by wind or sheet erosion to produce scalds and expose hardpans. AGRICULTURE – such as tillage, pugging by cattle creating densipans or introducing oxygen to sulfidic discharge areas.
Understanding water table hydrology, soil properties and processes is fundamental to selecting the best options for drainage and the most appropriate management of the soils when they are drained.
12.5 SOILS AFFECTED BY GROUNDWATER-ASSOCIATED SALINITY (GAS) Approximately 5.7 million hectares of Australia’s agricultural and pastoral zone have a high potential for developing GAS through shallow groundwater
(NLWRA 2001b). GAS occurs primarily in agricultural zones that are in hydrological disequilibrium. This has generally occurred since European settlement and extensive land clearing, and the replacement of deeply rooted native vegetation by shallow rooted, water-inefficient annual crops and pastures (NLWRA 2001b). A critical combination of hydrologic disequilibrium and a source of salt are necessary for a landscape to be affected by GAS. Key physiographic factors govering the formation of GAS include hydrogeology, land use, climate and landform. Predictions based on the distribution of these factors indicate that, unless effective solutions are implemented, the area of dryland salinity could increase to 17 million hectares by 2050 (NLWRA 2001b). However, with improved modelling techniques and the projected impacts of climate change, the Audit’s 17 million ha estimate now appears to be too high because groundwater tables have dropped in many regions, shrinking saline discharge areas.
12.6 NON-GROUNDWATER-ASSOCIATED SALINITY (NAS) Non-groundwater-associated salinity (NAS) is associated with the presence of perched groundwater systems, and is confined to the upper hillslopes (Figures 12.4,12.5 and 12.6). There are three expressions of NAS that are governed by soil–regolith depth: 1. deep NAS (ECse 20–60 dS/m) deep in the soil profile (for example, 5 m), and corresponds to the maximum depth of root penetration / upper limit of phreatic water table 2. surface NAS (> ECse 16 dS/m) at the soil surface 3. subsoil NAS (ECse 2–16 dS/m) in the subsoil (< 1 m). Collectively, surface and subsoil NAS are called shallow NAS (Figures 12.4 and12.5). The presence of deep NAS indicates no hydraulic connection between the deep (for example, deeper than 2–5 m) phreatic water table and the soils directly above, and therefore confirms groundwater associated with shallow NAS to be perched. While 16% of Australia’s dryland cropping area has the potential to be affected by GAS, 67% of the area
Soils and natural resource management
has a potential to be affected by NAS – at a cost to the farming comminity of approximately A$1330 million per year (Rengasamy 2002). In these soils, where the upper layers of soil are sodic, water infiltration is very slow because dispersed clay clogs the soils pores. If the subsoils are also sodic, the downward movement of water is restricted, thus causing temporary waterlogging in the subsoil, and the development of ‘perched water tables’. Salts accumulate above perched water tables during the wet winter and accumulate in the sodic subsoils following drying by water uptake by plant roots and evaporation, when the accumulated salts concentrate. Not all shallow NAS is associated with the presence of perched water tables, or soils with sodic B horizons. In some landscapes clay layers or poorly permeable, near-surface regolith materials may impede the downward leaching of salts, resulting in the local accumulation of subsoil salts in the perched water tables. If salt is blown in by the wind, its rate of
accumulation is not necessarily large, but, over time, it can be detrimental to crops (Gunn and Richardson 1979). This shallow subsoil NAS fluctuates with depth and spatial distribution in the landscape, and seasonally as the balance between downward and upward soil profile fluxes change. Shallow NAS is strongly associated with texture contrast soils that feature a sandy/loamy A horizon over a clay-rich and sodic B horizon. These soils are agriculturally very important in southern Australia, and feature extensive areas of dryland cropping and grazing rotations. 12.6.1 Subsoil shallow NAS Figure 12.7 illustrates the conceptualisation of subsoil shallow NAS formation processes on hillslopes (Thomas 2007). Before the cultivation of crops (Figure 12.7a), native vegetation formed hydraulic conditions in which salts were held in near-equilibrium deep in
0.5 m
Pre-cultivation
Cultivated, winter
A horizon (loamy texture)
B horizon (heavy clay)
(a)
(b)
~ 20 m
Cultivated, spring
(c)
Cultivated, summer
(d)
Figure 12.7: Conceptualisation of shallow subsoil NAS formation processes on hillslopes. Small open circles indicate relative salt concentrations and arrows indicate predominant water movements (after Thomas 2007).
319
320
Regolith Science
the soil–regolith profile (that is, deep NAS) – beyond the root zone. On agricultural land clearance, however, native vegetation is replaced by a matrix of annual vegetation (crops) and shallow-rooted grasses, which result in lower water use efficiencies, landscapewide. Consequently, during winter (Figure 12.7b) when up slope soils are wettest, subsoil water movement is transmitted laterally (that is, down-slope) above the sodic B horizon, via the coarser textured A horizon. Salts are mobilised in the solum in association with through-flow. This combination results in the local formation of dilute saline wet patches (that is, perched water tables; see also Section 10.3) particularly in the low points of wavy/irregular B horizon boundaries (Schoeneberger et al. 2002), in landform depressions, breaks of slopes, and where upper layer soil textures become more clayey down-slope at soil unit transitions. However, where subsoil freshwater flow rates are sufficiently high (such as in drainage zones on steep slopes), subsoil salts are washed out to down-slope areas. Over time, the perched water tables accumulate salts, which become increasingly concentrated (Figure 12.7c). Subsequently, during the warm dry summer, the combination of surface evaporation and subsoil evapotranspiration creates dynamic moisture conditions in the subsoil profile. These conditions cause dissolved salts to mobilise through the subsoil profile in the wetting front, and concentrate in topsoil and subsoil layers as they dry (Rengasamy 2002) (Figure 12.7d). However, leaching and saturation within the rooting zone causes a number of chemical, biological and physical changes, including: (i) acidity, (ii) sodicity, (iii) sodicity and salinity and (iv) sodicity and alkalinity. 12.6.2 Surface shallow NAS The most extreme case of salt accumulation is where ECse values are very high at the surface (ECse >16 up to 60.0 dS/m), and is often noted due to surface salt efflorescences. These high levels of salt prevent crops – and even halophytes – from growing and can cause the soil to be susceptible to scalding and erosion. The cause of this salinity is the localised mobilisation of salts above slowly permeable sodic B horizons by through-flow to topographic depressions and salt accumulation by evaporation. This surface shallow NAS can occur in a
variety of soil types and at all positions within undulating landscapes, and was first reported in South Australia by Herriot (1942). Approximately 45 000 ha of marginal cropping land in South Australia are affected by this problem (Kennewell 1999). This form of NAS is commonly referred to in South Australia as ‘magnesia patches’ because of the apparent dominance of Mg when first documented. Subsequently, however, Na has been shown to generally dominate as a natural part of the salt evaporation sequence. When surface shallow NAS soils are drained, soils are leached and salt efflorescences on the soil surface are dissolved. Salt crystals develop at depth in sodic soils where salt is leached through the subsoil clay layers on edges of gullies or drains. This causes stream banks to erode by salt weathering. If these processes are expressed on the surface of the soil, bare eroded saline scalds are evident (Figure 12.8).
12.7 SOIL SODICITY AND ACIDITY 12.7.1 Sodicity Sodic soils, like some saline soils, contain relatively large amounts of Na+. A soil is considered to be sodic when the adsorbed Na+ reaches a concentration where it starts to affect soil structure (Rengasamy and Sumner 1998). Generally speaking, in Australia this occurs when the exchangeable sodium percent (ESP) exceeds 4 (Northcote and Skene 1972). Sodic soils can result from drainage of the soil profile by erosion gullies, or directly from the weathering of parent materials over thousands of years (Isbell et al. 1983). Sodic soils have poor structure and low permeability, with adverse effects on plant growth. Permeability allows water, gases (O2 and CO2) and solutes to circulate easily to and from plant roots, which promotes plant growth. However, if a hard sodic dispersed clay layer occurs on the soil surface (for example, Figure 12.7) or close to the soil surface (such as the B horizon in Figures 12.5 and 12.7) it can act as a barrier to root development. The hard soil restricts root growth to either the cracks or topsoil above the claypan because movement of water, nutrients, and gases is too slow in sodic B horizons. In fact, when dry, the B horizon can be so hard that it is also a physical barrier to root penetration (see also Section 8.3.2). The overall effect on
Soils and natural resource management
(a) Virgin Soil
(b) Agricultural Soil Subsoil Non-groundwater Associated Salinity (NAS)
Weakly saline
Sodic Loam
(c) Drained or Eroded Soil
Surface Non-groundwater Associated Salinity (NAS) Salts on surface
Sodic Loam
Sodic
Saline layer
Sodic Columnar Clay
0m
0m
0m
1m
1m
1m
>15m
>15m
Saline groundwater
Salt leaching and clay dispersion
Saline drainage
Drained
>15m
Salt leaching and clay dispersion
Salt weathering
Cropping
Clay
No salts
Figure 12.8: Schematic soil–regolith model showing salt transport and erosion processes leading to formation of shallow subsoil and surface non-groundwater-associated salinity (NAS). Note: A sodic duplex soil is used here as an example. However, these processes do also occur in gradational soils or in soils with thin A horizons directly overlying saprolite (after Fitzpatrick et al. 2003c).
plant growth is one of stress similar to that caused by extremely dry or saline conditions. The rate and amount of downward percolation of salts are primarily controlled by soil texture and subsoil layer permeability. In coarse-textured horizons, water flows more quickly and the average pore diameter is larger than in fine-textured soils. Decreased water storage is directly related to greater pore diameter. As a result, deep percolation of water and salts is more likely to occur in coarse-textured soils. In some localities in Australia, relatively coarse-textured soils overlay impermeable sodic clay horizons. Under these circumstances, percolation leads to lateral flow of water and solutes along the surface of the impermeable layers (Fitzpatrick et al. 1994). If the contact between the two different layers approaches the soil surface along a hillslope, as often happens, the laterally moving water will create a wet spot that eventually becomes saline as the water is evaporated (Figure 12.7). Most soils that have a high potential for developing NAS through shallow water tables are sodic and/or are duplex (Northcote and Skene 1972). Duplex soils have a definite and marked difference in texture between A and B horizons, and have a visually distinct boundary exists between these horizons. More than 60% of the 20 million ha of cropping soils in
Australia are sodic and dryland farming is mainly practiced on these soils. More than 80% of sodic soils in Australia have dense clay subsoils with both high sodicity and alkalinity pH (>8.5). Perched water tables may form within many of these subsoils in some years (Figure 12.8). Depending on landscape position and the dominant flow pathways within these soils, salt and other chemicals that form or accumulate within the rootzone under redoximorphic conditions, and can be redistributed within the catchment. Where interceptor drains have been installed to lower shallow water tables, drainage waters from duplex soils can also be high in solutes. 12.7.2 Acidity Soil acidity is a severe soil degradation problem in Australia that can greatly reduce the production potential of farming systems (for example, Helyar and Porter 1989; Helyar 1990; Scott et al. 2000; White et al. 2000). The National Land and Water Resources Audit (NLWRA 2001a) estimated that 50 million hectares of the agricultural zone are already suffering from acidification of soil surface layers and 20 million ha from subsoil acidification, and that these are ‘probably markedly affecting yields’. Much of this problem occurs in productive agricultural zones. It causes
321
322
Regolith Science
production losses within paddocks, and also longterm and off-site effects, including:
12.8 SUB-AQUEOUS AND ACID SULFATE SOILS (ASS)
s
12.8.1 Sub-aqueous soils In the last decade, the US Department of Agriculture’s (USDA) definition of soils has completely changed to include environments that are permanently submerged (Soil Survey Staff 1999). Sub-aqueous soils form in sediments found in shallow permanently flooded environments. Excluded from the definition of sub-aqueous soils are any areas ‘permanently covered by water too deep (typically greater than 2.5 m) for the growth of rooted plants’ (Stolt 2006). Pedologists have been studying these sub-aqueous soils in shallow sub-tidal lagoons and so have described their benthic substates from a pedological perspective: that is, characterising the morphological (colour, structure), physical (texture, compressibility), chemical (pH, salinity), mineralogical (layer silicates, Fe oxides) and biological (roots, plants) properties. Once the benthic materials and underlying sediments are described as soils, investigators can easily identify the relationship between the soils and their position on the landscape (for example, Bradley and Stolt 2003; Demas and Rabenhorst 1999). An understanding of these relationships enables land managers to identify the best location for specific land uses and to better predict the potential impact of proposed changes (such as dredging or drainage) on subaqueous soils and the ecosystems they support.
s s
poor water use by plants (leading to higher recharge and erosion) increased leaching of nutrients and Al the irreversible breakdown of layer silicate minerals in soils.
Soil acidification can be determined by assessing the pH of a soil. Soil pH can be measured in water (pHw) or calcium chloride (pHCa). A 1:5 mix of soil:CaCl2 solution (0.01M strength calcium chloride) strength is used to estimate the concentration of hydrogen ions in the soil solution. The pHCa method is the international standard because the pH is less affected by soil salts. Soil acidity is not thought to restrict the growth of most crops or pasture until the pH drops to below 5.5–6.0 (pHw) or below 5.0–5.5 (pHCa). Development of acidity in soils is a natural process, especially in the high rainfall regions of southern Australia. Some soils are inherently acidic because of the high rates of leaching in these regions. Even in lower-rainfall cropping areas, some soil types have become acidic because they have no free lime in the profile. As soils become more acidic, plants and crops that cannot tolerate acidic conditions do not flourish, so productivity and yields decline. When conditions become severely acidic (pH<4, such as in acid sulfate soils), biogeochemical processes start to break down the layer silicates in the soil, releasing Al, Fe and Mn. This may lead to mineral toxicities and nutrient imbalances. The natural rate of acidification is accelerated by the use of acidifying fertilisers, nitrogen fertilisers, the removal of agricultural products and nitrate leaching. The management of soil acidity involves the following requirements at the farm level:
s s s s
recognising paddock indicators of soil acidity monitoring soil pH knowing crop and pasture tolerances to acidity treating paddocks that have acidity problems.
Acid soils can be ameliorated by applying liming material or other types of neutralising agent, growing acid-tolerant plants, or reducing the rate of acidification.
12.8.2 Acid sulfate soils Acid sulfate soils (ASS) are all those soils in which sulfuric acid may be produced, is being produced, or has been produced, in amounts that have a lasting effect on main soil characteristics (Pons 1973). In general, three broad genetic soil types of ASS are recognised (for example, Fanning 2002), including: (i) potential, (ii) actual (or active) and (iii) post-active ASS. ASS form in coastal, estuarine, mangrove swamp and marsh environments because these waterlogged or highly reducing environments are ideal for the formation of sulfide minerals – predominantly pyrite (FeS2). Iron sulfide minerals are one of the end products that form as part of the process of sulfate reduction (that is, the use of SO42– instead of O2 during microbial
Soils and natural resource management
SO42-
Organic matter
Bacteria
Bacteria
Fe minerals
H2S
Saturated, organic-rich soil promotes reducing conditions and the activation of the SO42– / S2– redox couple (see especially Section 7.7.1). This process produces pyrite [FeS2] that is, sulfidic material (Soil Survey Staff 2003; Isbell 1996), in reactions that are the reverse of those discussed in Section 5.3.8:
s
Sulfate reduction:
Ba
cte
ria
2H + + SO 24 - + 2 (CH 2 O) = 2CO 2 + H 2 S + 2H 2 O (Eqn 12.1)
S0
FeS
s
H2S reacts with Fe2+ to precipitate FeS, which can be converted to pyrite:
FeS2 Pyrite Figure 12.9: Schematic diagram for the formation of pyrite in anoxic sediments (after Berner 1984).
s
H 2 S + Fe 2 + = FeS + 2H +
(Eqn 12.2)
H 2 S + FeS = FeS 2 + H 2
(Eqn 12.3)
H2S also reduces Fe3+ minerals to form pyrite: 4H 2 S + 2Fe(OH) 3 = 2FeS 2 + 6H 2 O + H 2
respiration; Section 7.4.1). Sulfate reduction is a natural process that occurs in virtually all lakes, rivers, wetlands and oceans. However, the quantities of sulfidic material that will accumulate in a given environment are a function of many factors. The key general requirements for high rates of sulfate reduction and sulfide accumulation are:
s s s s
high concentrations of sulfate in surface or groundwater saturation of soils and sediments for periods long enough to favour anaerobic conditions availability of labile carbon to fuel microbial activity availability of Fe minerals (Table 12.1 and Figure 12.9).
To form sulfidic materials, the bicarbonate produced by the reduction reactions must be flushed from the sediment, for example, by tides. Sulfidic material
Soil horizons that contain sulfides are called sulfidic materials (Isbell 1996; Soil Survey Staff 2003) and can be environmentally damaging if exposed to air by disturbance.
(Eqn 12.4) Pyrite-enriched soil materials are termed sulfidic material because they have the ingredients necessary to produce sulfuric materials (for example, Pons 1973; Fanning 2002; Fitzpatrick et al. 1996). They are waterlogged, mineral or organic subsoil material that contains oxidisable sulfur compounds, usually pyrite, that has a field pH of 4 or more, but which will become acid (pH less than 4) when drained (Isbell 1996). Sulfidic material is identified by a drop in pH by at least 0.5 unit to 4 or less (1:1 by weight in water, or in a minimum of water to permit measurement) when a 10 mm thick layer is incubated wet (that is, at field capacity) for 8 weeks. Sulfuric material
Sulfuric material has a pH less than 4 (1:1 by weight in water, or in a minimum of water to permit measurement). Exposure of pyrite to air results in the oxidation of pyrite with each mole of pyrite yielding 4 moles of acidity: FeS 2 +
15 7 O2 + 2 H 2 O 4
= Fe(OH) 3 + 2SO 24 - + 4H + (Eqn 12.5)
323
324
Regolith Science
This process transforms sulfidic material to sulfuric material (Isbell 1996) because the material has a pH of less than 4 (that is, a sulfuric horizon with pH of less than 3.5 accoring to Soil Survey Staff 2003). Acidification occurs if the amount of acidity produced exceeds the buffering capacity of soil. Evidence that low pH is caused by oxidation of sulfides within a soil profile is provided by one of the following:
s s
yellow mottles and coatings of jarosite or natrojarosite underlying sulfidic material.
If the amount of acidity produced exceeds the buffering capacity of soil then, acidification occurs. Once sulfidic material is drained it may transform to sulfuric material. Sulfuric material forms in the hydromorphic zone following dewatering or drainage as water levels drop (such as those caused by either drought conditions, cattle pugging or dredging operations), which exposes pyrite in the upper layers of the soil profile to oxygen. When this happens, pyrite is oxidised to sulfuric acid (Eqn 12.5) causing the soil pH to drop from neutral (pH 7) to below 4, which dissolves layer aluminosilicates (clay) in the soil. The dissolution of the pyrite and associated rise in acidity also causes cations and associated anions (such as Na+, Mg2+, Ca2+, Ba2+, Cl–, SO42– and SiO44–), trace elements, and metal ions (such as Fe3+ and Al3+) to be released on the soil surface. This acid, together with toxic elements that are leached from ASS, can kill fish and oysters, contaminate groundwater, and may corrode concrete and steel in built structures. These impacts can be measured in terms of:
s s s s
poor water quality, with loss of amenity, damage to estuarine environments and reduction of wetland biodiversity the need for rehabilitation of disturbed areas to improve water quality and minimise impacts loss of fisheries and agricultural production additional maintenance of community infrastructure affected by acid corrosion.
Infrastructure development and primary industries around Australia are facing a $10 billion legacy of acid sulfate soils (National Working Party on Acid Sulfate Soils 2000). Public recognition of this serious problem
BOX 12.1: Monosulfidic black ooze material Monosulfidic black ooze (MBO) material comprises sub-aqueous or waterlogged minerals or organic materials that contain mainly oxidisable monosulfides. This material has a field pH of 4 or more, although it will not become extremely acid (pH <4) when drained. The recognition of the occurrence and importance of monosulfides in soil materials led in 2005 to the inclusion of monosulfidic materials as a distinguishing property within mapping units of the Atlas of Australian Acid Sulfate Soils (Fitzpatrick et al. 2006). High nutrient environments together with the activity of algae and micro-organisms generate redoximorphic conditions, which result in the formation of black, smelly, Fe monosulfides (see also Section 7.7.1). When sub-aqueous materials rich in monosulfides are resuspended – for example during the flushing of drains by high run-off events – they rapidly oxidise to potentially remove most of the oxygen from the water column (Sullivan et al. 2002). This can lead to fish kills, especially in enclosed areas such as aquaculture ponds or estuaries. Hence, MBO is reactive if exposed to oxygen but is harmless if left undisturbed. Monosulfidic soil materials have the ability to favourably affect surrounding environments by immobilising potential metal pollutants (Simpson et al. 1998). However, when a drain is cleaned, alunite supergroup minerals (especially alunite and jarosite) and Fe oxyhydroxy-sulfate salts (such as schwertmannite) precipitate on the soil surface along the drain edges. These soluble salts dissolve during rain events and contribute to MBO formation, acidity and metal content in drainage waters.
has been reflected in government building legislation in several Australian states. In addition, there is gathering support from local government and industries to develop statutory requirements for rehabilitation. Occurrence of acid sulfate soils
Australia contains a wide range of types of ASS in different physical settings (coastal, estuarine, mangrove swamp, back swamp and inland lake environments), which occur because of changing hydrological and biogeochemical conditions. Various sources of organic
Soils and natural resource management
BOX 12.2: Formation of sulfuric materials in bunded tidal zones In several parts of Barker Inlet near Port Adelaide, bund walls were constructed across tidal zones (mangrove and samphire swamps) nearly 50 years ago to cut off tidal flushing, which effectively disturbed (drained) these areas causing mangrove trees and samphire vegetation to die (Figure 12.10). Excluding sea water from the original sulfidic material caused the surface to dry and oxidise sulfide to produce sulfuric acid (pH commonly between 2.5–3.5) and bright yellow mottles of jarosite were identified (Fitzpatrick et al. 2008a). The schematic cross-section in Figure 12.10 illustrates how the former back barrier sand ridge at Gillman has devel-
Barker Inlet High tide mark
oped a 2 m thick soil profile with sulfuric material because pyrite framboids in and surrounding decomposed mangrove pneumatophores have oxidised to form jarosite mottles and acidity where neutralising by alkaline materials is limited. Coatings of jarosite and Fe oxides form rapidly along large root channels during periods of drying. Some small, unoxidised pyrite framboids still occur in the underlying sandy, sulfuric horizons. In the upper horizons (0–58 cm), the oxidation of pyrite in organic residues caused precipitation of Fe oxides and lenticular gypsum crystals, which are now being leached out of the profile (Fitzpatrick et al. 2008a).
Acidic, contaminated water discharge (minimal) to Barker Inlet via a tidal gate
Undrained tidal wetland Mangroves
Salt bush
20% Carbon
Greenhouse emissions Evaporation>Recharge
0.7 m of land subsidence
Dead mangroves
Bund wall
Mangrove peat (sulfidic material)
5% Carbon
Low tide mark Sea water
Acid and toxic oxidation products
Drained mangrove peat (sulfuric material)
Tidal stream (subaqueous soil filled with sapric material) Tidal stream - blocked by bund wall (subaqueous soil with bottom sediment which contains MBO material )
Figure 12.10: Schematic soil–landscape cross section at Gillman in Barker Inlet, Port Adelaide. Normal tidal dynamics are interrupted by a bund wall (levee bank built in 1965) causing oxidation of sulfidic materials and monosulfidic black ooze (MBO) to occur, which contributes to degraded acidic saline land (sulfuric materials), denuded vegetation, reduction of wetland biodiversity, poor estuarine and stream water quality, ground subsidence, increase in greenhouse emissions and loss of amenity (after Fitzpatrick et al. 2008a).
matter fractions (that is, sapric and hemic materials), minerals (such as pyrite, jarosite and gypsum), and micro-scale weathering pathways and mechanisms occur under drained (such as through levee bank construction) and undrained (ranging from natural tidal to intertidal, to supratidal zones) conditions. Ultimately, they pose different environmental hazards requiring tailored management options. The purpose of this section is to summarise available information on these fascinating and critically important soils. In Australia, coastal ASS occupy an estimated 95 000 km2, of which 74 000 km2 are exposed at some point during the tidal cycle. This area contains well
over two billion tonnes of potentially dangerous sulfidic material (Fitzpatrick et al. 2006). These soils underlie coastal estuaries and tidal flats, much of which are close to major population centres in Australia. At time of writing, considerable effort is being expended in Australia to document and map the occurance of inland ASS. For example, major occurances have been documented in river and wetland systems (such as the Murray River, Lower Lakes and Coorong in South Australia; Fitzpatrick et al. 2008b), artifically drained landscapes (such as the Western Australian wheatbelt cropping region; Fitzpatrick et al. 2005) and upland landscapes experiencing altered
325
326
Regolith Science
Table 12.4: Case studies from geochemically variable saline and/or acid sulfate environments where soil indicators and hydro-pedologically based toposequence models have been incorporated in land management manuals or reports. Locality of case study Key references
Hydrological type GAS
NAS
Herrmanns and Keynes, Mt. Lofty Ra., SA. (Fitzpatrick et al. 1996, 1997, 2003b; Skwarnecki et al. 2002)
D (s)
M
Dairy Creek, Mt. Lofty Ra., SA. (Fitzpatrick et al. 1996,1997, 2003b; Skwarnecki et al. 2002)
D (s)
Strathalbyn, Mt. Lofty Ra., SA. (Skwarnecki and Fitzpatrick 2003)
D (s)
M
Loveday disposable basin, River Murray system, SA (Lamontagne et al. 2004)
D (s)
M
Gillman, SA (Fitzpatrick et al. 2008a)
M (p)
Lakes Albert and Alexandrina, River Murray below Lock 1, SA (Fitzpatrick et al. 2008b)
M
Woorndoo, Vic. (Cox et al. 1999; Fitzpatrick et al. 2003b)
D (s) (p)
M
Gatum, Dundas Tablelands, Vic. (Brouwer and Fitzpatrick 2002)
D (s)
Merriefields, Dundas Tablelands, Vic. (Gardner et al. 2004a, b)
IAS
Geochemical category of salt-affected soil Alk
Hal
Gyp
Sulfidic
Sulfuric
Sod
M
S
D
M
S
M
S
S
D
S
M
S
D
M
S
D
M
M
M
D
D
M
M
D
D
S
S
S
S
M
M
M
S
S
D (s)
M
M
S
D
Rouse Hill, NSW. (Cox et al. 2002)
D (s)
M
M
S
M
M
S
Mesopotamian marshlands, Iraq (drained, burned, reflooded). (Fitzpatrick 2004).
D (s)
S
S
S
D
S
D
North China Plain, China. (Fitzpatrick et al. 2002)
D (s)
M
S
S
M
M
S
Jamestown, SA. (Fitzpatrick et al. 2003d; Thomas et al. 2008a, b)
M
D
M
M
M
D
S
M
M
D
S
M
S
Kyndalyn Park, Riverland, Vic. (Cass and Fitzpatrick 1999).
D
D
S
S
Barossa Valley, SA, (Clark et al. 2002)
D
D
M
D
Wheat belt, WA (Lee, 2002, Fitzpatrick et al. 2005)
D (s) (p)
M
M
D
M
D
M
GAS = groundwater-associated salinity, (p) = primary (naturally saline); (s) = secondary (anthropogenic salinity). NAS = non-groundwater-associated Salinity. IAS = irrigation associated salinity. D = dominant; S = sub-dominant; M = minor. Alk = alkaline (sodium carbonate dominant, pH >9); Hal = halitic (sodium chloride dominant); Gyp = gypsic (gypsum dominant) or Mg-sulfate salts; sulfidic = pyrite rich and pH >4.0; Sulfuric = sulfuric acid dominant, pH <4.0); Sod = sodic (high ESP), Ra = ranges, SA = South Australia; Vic = Victoria; NSW = New South Wales, WA = Western Australia.
Soils and natural resource management
BOX 12.3: Formation of sulfuric materials in inland River Murray and Lower Lake Systems The River Murray, adjacent wetlands and the Lower Lakes (Alexandrina and Albert) – which are close to the Murray Mouth in South Australia – are being seriously impacted by a combination of low water levels and the presence of acid sulfate soils (Fitzpatrick et al. 2008b). The Lower Lakes and the floodplains below Lock 1 at Blanchetown are undergoing their first major drying phase since the introduction of barrages more than 50 years ago. Anaerobic environments are ideal for the buildup of Fe sulfide minerals, which have been collecting in submerged and waterlogged soils of the Lower Murray region’s waterways since the construction of locks, weirs and barrages over 50 years ago, and have led to the retention of water in the river system. The current drought conditions (February 2008) – the worst on record – have resulted in these accumulated sulfide minerals being
exposed to air for the first time. This has lead to the development sulfuric material (Figure 12.11), with pH levels dropping below 4 because of the formation of sulfuric acid. Water levels in the River Murray have dropped particularly between Lock 1 and the Lower Lakes. These low water levels have exposed submerged or sub-aqueous soils, wetlands, areas of riverbank and parts of the lower lakes that contain high levels of pyrite (Figure 12.11). Dredging operations have also heaped and exposed sulfides to oxygen, resulting in the formation sulfuric material. Fitzpatrick et al. (2008b) have identified four sequential phases (or classes of ASS) that form depending on drainage conditions. Soils range from deep submerged sediments to sub-aqueous soils to waterlogged/saturated (all anaerobic) to unsaturated (aerobic) drained soils (Figure 12.11).
Deep water ASS material below a water depth of 2.5m Sulfidic or MBO (monosulfidic black ooze) materials
Lowering water levels
Lowering of water levels to depths shallower than 2.5m due to drought conditions and evapotranspiration Formation of subaqueous ASS with sulfidic material or MBO in shallow water
Subaqueous ASS in water at depths shallower than 2.5m Sulfidic or MBO materials Lowering of water levels until the soil surface is no longer under water but still saturated Increased formation of sulfidic or MBO materials due to higher organic matter accumulation and temperatures
Waterlogged and saturated ASS in upper parts of soil with anaerobic conditions Sulfidic or MBO materials Lowering of water levels and water tables resulting in upper parts of the soil becoming drier and aerobic Progressive exposure of sulfidic material to air Formation of sulfuric acid because pyrite in sulfidic material reacts with oxygen Development of sulfuric materials (pH drops below 4)
Drained and unsaturated ASS in upper parts of soil with aerobic conditions Sulfuric material (pH less than 4) or MBO material with desiccation cracks Figure 12.11: Generalised schematic diagram showing the sequential transformation of four classes of ASS due to lowering of water levels from ‘Deep water ASS’ ® ‘Sub-aqueous ASS’ ® ‘Waterlogged and saturated ASS’ (all containing sulfidic material with high sulfide concentrations and pH greater than 4) ® ‘Drained and unsaturated ASS’ containing sulfuric material (pH<4) in the upper soil layers (after Fitzpatrick et al. 2008b).
327
328
Regolith Science
groundwater drainage where mineralisation occurs (such as the Mount Lofty Ranges, South Australia and Dundas Tablelands, Victoria) (see examples listed in Table 12.4). Acid sulfate soil management: Summary of principles
The environmental risk is present because draining river or wetland systems involves the disturbance of material that was not previously in contact with the oxygen in the atmosphere. During the lowering of the water table or drying of sub-aqueous soils, sulfidic materials may be exposed and sulfides within the sub-aqueous soil horizons will begin to oxidise because they are exposed to air (Figures 12.10 and 12.11). As discussed, this will produce sulfuric acid and potentially release toxic quantities of Fe, Al and heavy metals if the soils do not contain sufficient acid neutralising capacity to maintain a sufficiently high pH (> 5). The acid, Al and heavy metals can leach into waterways, kill fish, other aquatic organisms and vegetation, and can even degrade concrete and steel pipes and structures to the point of failure. However, appropriate management of ASS during development can improve discharge water quality, increase agricultural productivity, and protect infrastructure and the environment. Such improvements can generally be achieved by applying low-cost land management strategies (for example, Dear et al. 2002) based on:
s
s
Identification and avoidance of ASS materials. Slowing or stopping the rate and extent of pyrite oxidation. This can be achieved either by keeping sulfidic material anaerobic under saturated conditions, or by rapid drying of sulfidic material to slow the biological process, which are responsible for the formation of acid. Retaining existing acidity within the ASS landscape. Acidity and oxidation products that cannot be retained on-site may be managed by other techniques such as acidity barriers or wetlands that intercept and treat contaminated water before discharge into rivers or estuaries.
The selection of management options will depend on the nature and location of the ASS materials, the assets to be protected, and the position of ASS in the landscape. This is why reliable ASS risk maps (at
appropriate scales) and characterising ASS landscapes are so important.
12.9 SOIL AND LANDSCAPE FIELD INDICATORS This section summarises approaches and procedures developed at the CSIRO Land and Water over two decades to:
s s
s
identify the best set of soil and landscape field indicators of soil–landscape condition for a region. construct appropriate 3D and 4D mechanistic models of soil–regolith and water processes that explain and predict the processes giving rise to geochemically variable salt-affected and acid sulfate soils using the toposequence approach (soil landscape cross-sections), which integrates pedological, hydrological, geological, biogeochemical and mineralogical information. publish easy-to-use pictorial manuals that incorporate field indicators and mechanistic models to be used by land managers and which provide landuse options that help prevent the spread of soil salinity.
Field indicators linked to landform elements are useful for identifying salt-affected soils and increasing awareness of the extent of salinity among landholders and regional advisers. Standard descriptive soil indicators, such as visual indicators (such as colour) and consistency, are often used by farmers, regional advisers and scientists in the field to identify and report attributes of soil quality (Fitzpatrick et al. 1999). For example, as discussed, soil colour can provide a simple means to recognise or predict salt-affected and ASSaffected wetlands caused by poor drainage, which provide a low-cost alternative to difficult and costly methods to document saline water table depths and to estimate water duration in soils. Visual indicators of salinity and ASS may be obvious (such as white or yellow salt accumulations on soil surfaces) or subtle (for example, subsoil mottling patterns, strong pedality). Analytical indicators include electrical conductivity (salinity) and dispersion (sodicity). Combining descriptive and analytical indicators has provided vital information about soil–water processes, leading to improve management and remediation of saline land,
Soils and natural resource management
as demonstrated in several case studies from Australia, China and Iraq (Table 12.4).
12.10 SOIL–REGOLITH TOPOSEQUENCE MODELS Conceptual process models enable researchers to develop, refine and present mechanistic understanding of complex soil–regolith environments (Fritsch and Fitzpatrick 1994). These models are graphic, cross-sectional representations of soil–regolith–bedrock profiles that illustrate vertical and lateral changes that occur down toposequences. They are used to explain the complex pedological, hydrological and biogeochemical interactions that occur in the regolith environment (Fitzpatrick and Merry 2002). Three categories of conceptual toposequence models have been described, which are: 1. descriptive soil–regolith models 2. explanatory soil–regolith models 3. predictive soil–regolith models. The descriptive soil–regolith process model shown in Figure 12.12 (page 172) characterises relict (past geomorphological processes in development of deep weathering and erosion) and current saline, alkaline, sodic, sulfidic or sulfuric soil-forming processes. Such models help to develop practical solutions for ameliorating soils at the farm scale. The descriptive soil– regolith model is used as the precursor or framework for developing the explanatory soil–regolith model (3D) shown in Figures 12.13 and 12.14, which represents current soil salinity (hatching), salt groundwater flow (dark blue arrows) and freshwater flow (light blue arrows). If required, the explanatory soil–regolith model in turn is used to develop the predictive soil–regolith model (4D) shown in Figure 12.15. Consequently, the predictive soil–regolith model (4D) consists of a collage of figures, which illustrates several evolutionary cycles of soil–regolith events. 12.10.1 Descriptive soil–regolith models To understand the lateral linkages and relationships between soil and landscape indicators (soil profile features), a systematic structural approach can be used to characterise soil–regolith features at different points along toposequences (Fritsch and Fitzpatrick
1994; Brouwer and Fitzpatrick 2002). Colour photographs of typical profiles at different parts down the toposequence are used (Figure 12.12, page 172). Briefly, these authors identified and described in the field – by depth interval in all profiles along the toposequence – all relevant soil properties, including texture, coarse fragments, structure, matrix colour and mottling. In the laboratory, chemical and mineralogical properties were determined. Toposequence cross-sections were then drawn that identified uniform layers that contain individual, or sometimes several, soil–regolith properties. Subsequently, boundaries were drawn around these layers. Each cross-section mapping unit or layer delineated is called a soil feature. A soil feature thus represents a limited range of one or more soil–regolith properties. The key soil– regolith features that help recognise and explain soil formation and interactions between different parts of the toposequence were grouped into the same soil systems using concordant relationships: that is, where there is a concordant relationship spatial distributions and boundaries mostly coincide – and hydrological processes, geochemical processes and/or parent material will be the same. Soil features were separated into different soil systems using discordant relationships; in such cases spatial distributions show no or only partial overlap, boundaries do not coincide but touch or cut cross each other, and processes and/or parent material will be different (Figure 12.12, page 172). In summary, these workers were able to group similar soil features into fewer soil layers, which were linked down the toposequence and mapped in crosssection. Each of these soil layers were linked to hydrological processes (water flow paths, salinity and sodicity) by using soil colour (together with other morphological, chemical and mineralogical indicators) and hydrology measurements (Cox et al. 1996; Fitzpatrick et al. 1996). This enabled the construction of 2D linkages that described water flow paths and the development of salinity in the Herrmann catchment in the Mount Lofty Ranges, South Australia (Figure 12.12, page 172). Fitzpatrick and Skwarnecki (2005) explain how these descriptive process models can be used to characterise catchment-scale variability of relict (past geomorphological processes in development of deep weathering and erosion) and current (saline, sodic
329
330
Regolith Science
Perched fresh water
Anaerobic Permanently wet (pH >6)
Aerobic
.!%,%00%+*+"/!+* .5 !*/* 1/1("% !/ 5%+)%*!.(%/0%+** /1("1..! 1%*#0!.%
Fluctuating wet and dry conditions 74% 0%+*+"/!+* .5/1("% !/* ,.!%,%00%+*+" "!..%$5 .%0!5!+4% %6%*#0!.%, /$3!.0)**%0!, &.+/%0!*0.+&.+/%0!* ,(1)+&.+/%0!, /% !.+*0.%0!* 0).1#%0!, 01(% /1("0!/+%(/1("1.%$+.%6+*, /1("1.%% "+.)! 5/1("1.+4% %/%*#0!.% 72!*#%*#+"1*/* +0$!.!(!)!*0/ 5!,.!%,%00!/
* !.+
,.+(%0!
.01.! ! .+' +*0%*%*# %//!)%*0! ,.%).5!/1("% !/
Halite and gypsum
Saline seepage
%
/! +* .5 /1
!//,!.)!(! /+ %(5(5!.
("% ! /%* /
1("%
%)
0!.
Dispersed clay, Fe oxides, acidic water
%(
,
,3. ,!.+(0%+*+"/1("0!* /(%*! #.+1* 30!./%*+*"%*! -1%"!./1* !.,.!//1.!
Mineralised zone with primary Zn, Pb, Fe and Cu sulfides
Figure 12.13: Explanatory soil–regolith model showing geochemical dispersion and erosion processes in saline seepages and formation of secondary sulfides in sulfidic material in a perched wetland and sulfuric materials along eroded drainage lines (after Fitzpatrick et al. 1996).
and acid sulfate soils) soil-forming processes to develop practical solutions for ameliorating soils at the farm scale, and for possible use in mineral exploration. 12.10.2 Explanatory soil–regolith models Fitzpatrick et al. (1996) used the descriptive soil–regolith toposequence model (Figure 12.12, page 172) to construct an explanatory soil–landscape process model to explain contemporary geochemical dispersion and erosion processes present in the lower parts of a toposequence (Figure 12.13). This model explains the formation and degradation of acid sulfate soils (ASS) in a single diagram that illustrates the pedological, geological, biogeochemical, mineralogical and hydrological processes occurring in the eastern Mount Lofty Ranges. Fitzpatrick et al. (1996) illustrate that a combination of: (i) saline groundwaters enriched in sulfate (with other elements sourced from mineralised zones, such as Pb and Zn) seeping up through soils, (ii) anaerobic conditions and (iii) organic carbon
in saturated soils yield sulfidic material containing pyrite framboids through anaerobic bacterial reduction of sulfate. Thus, when these sulfidic materials are eroded and exposed to air, pyrite is oxidised producing sulfuric acid, which dissolves soil minerals and leads to precipitation of mineral combinations:
s
s s
sideronatrite [Na2Fe(SO4)2 (OH).3H2O], tamarugite [NaAl(SO4)2.6H2O], copiapite [Fe5 (SO4) 6 (OH)2.20H2O], halite and gypsum (see also Table 12.1) in sandy sulfuric horizons with a pH of less than 2.5 natrojarosite, jarosite and plumbojarosite in clayrich sulfuric horizons with a pH of 3.5–4 schwertmannite (orange; pH of 4), ferrihydrite (reddish-brown; pH greater than 6), akaganéite (reddish-orange) and white, poorly-crystalline Al oxyhydroxide precipitates (see also Section 4.5.5).
The formation of the complex suite of sulfate salts (of Fe, Al, Na, Pb, Ca, As and Zn), jarosites, oxyhydroxysulfates and Fe oxides are indicative of rapidly
Soils and natural resource management
100 m
Water table
Precipitation of secondary base metal sulfides by biomineralisation (including bacterial reduction) in saline mounds and wetlands Saline seepage mound springs Creek Wetland
50 m
Thin skeletal soil (<50cm)
Colluvium and alluvium with acid sulfate soils
Through flow
Gossanous zone with enhanced Ag and Au
Fractured bedrock
Groundwater flow
Orebody or zone of mineralisation (primary sulfides with Zn, Pb, Fe, Cu, Ag and Au) Figure 12.14: Explanatory soil–regolith model showing geochemical dispersion from mineralised zones in sulfidic/sulfuric materials from seeps, springs and wetlands, eastern Mount Lofty Ranges, South Australia (after Skwarnecki and Fitzpatrick 2003).
changing local environments and variations in Eh (redox), pH and rates of availability of Fe, S and other elements (Skwarnecki and Fitzpatrick 2003). Regional sampling by Skwarnecki and Fitzpatrick (2003) has shown that a range of materials (sulfidic materials, sulfuric horizons, salt efflorescences, and Fe- and Al-rich precipitates) are anomalous in elements such as As, Bi, Cd, Cu, Pb, Tl and Zn, especially where they are spatially related to sulfide mineralisation (cf. elements likely to be present in gossans; Section 13.4). Thus, the sulfidic/sulfuric material may carry indications of the presence of blind or concealed ore deposits, making these sediments a potential sampling medium for mineral exploration (Figure 12.14). 12.10.3 Predictive soil–regolith models: landscape evolutionary processes Fitzpatrick et al. (2000a) used the information contained in Figures 12.12 (page 172), 12.13 and 12.14 to construct a predictive soil model showing the hydrogeochemical processes, which transform sulfidic material in a perched wetland to highly sulfuric material (Figure 12.15). Stage 1
Saline groundwater enriched in sulfate (SO42–) seeps up through the soil, along with other ions in solution
such as Na+, Ca2+, Mg2+, AsO43–, I– and Cl–, and concentrates by evaporation to form various mineral precipitates within and on top of the soil surface (Figure 12.15a). The combination of: (i) rising sulfatic groundwater, (ii) anaerobic conditions associated with saturated soils, (iii) agricultural activity and (iv) fractured rocks relatively enriched in, for example, Fe, S, Pb and Zn leads to the formation of sulfidic material and precipitation of anomalous concentrations of Pb and Zn. If the soil is wet and contains sufficient organic carbon, anaerobic bacteria use the oxygen associated with the sulfate (SO42–) ions during the assimilation of carbon from organic matter. This process produces pyrite and forms sulfidic materials (Figure 12.15a) (Fitzpatrick and Skwarnecki 2005). Stage 2
Sulfuric materials result when pugging from animals, drainage works or other disruptions expose the pyrite in previously saturated soils to oxygen in the air. Thus pyrite is oxidised to sulfuric acid and various Fe sulfate-rich minerals, and sulfuric material forms (Figure 12.15b). When sulfuric acid forms, the soil pH can drop from neutral (pH 7) to below 4; locally pH may attain values as low as 2.5 to form a sulfuric horizon (Figure 12.15b). The sulfuric acid dissolves the clay particles in soil, causing basic cations and
331
332
Regolith Science
(a) Stage 1
(b) Stage 2
(c) Stage 3
Perched wetland (undrained) - Acid Sulfate Soil with sulfidic material
Downslope erosion (drained) - Acid Sulfate Soil with sulfuric material
Upslope erosion (drained) - Remnant Post-active Acid Sulfate Soil
sedges and reeds
Gypsum Mg-calcite Ferrihydrite gels
salt tolerant grasses
Gypsum Schwertmannite crust Ferrihydrite Few salts
Sulfidic Material (saline)
pH > 6.5
Sideronatrite pH < 2.5
Sulfuric Horizon (saline)
no vegetation
Hard, cemented, impermeable layer (Sodic)
Natrojarosite pH 3.5
Gleyed Clay Saline and Sulfate rich 2–
pH > 7.0
Gleyed Clay Saline and Sulfate rich 2–
+
SO , Na , _ 4 2+ 2+ Cl , Mg , Fe
pH > 7.0
Gleyed Clay Saline and Sulfate rich 2–
+
SO , Na , _ 4 2+ 2+ Cl , Mg , Fe
pH > 7.0
+
SO , Na , _ 4 2+ 2+ Cl , Mg , Fe
Potential ASS expands upslope
ASS clogged
Fractured rock
Perched wetland/ saline seepage or spring
Saline discharge water with _ + 2+ 2+ Na , Cl , Mg and Fe
Marker used as reference to show position of movement of erosion and expansion of potential ASS seepage area.
Figure 12.15: Predictive soil–regolith model showing the hydrogeochemical processes, which transform sulfidic material in a perched wetland to highly saline sulfuric material (after Fitzpatrick et al. 2000a).
associated anions (such as Na+, Mg 2+, Ca 2+, Ba 2+, Cl–, SO42– and SiO44–), trace elements and metal ions, such as Fe3+ and Al3+, to be released onto the soil surface and into stream waters. As the regolith structure degrades due to the accompanying sodicity, soils become clogged with dispersed clay and Fe precipitates and they lose their permeability and groundcover. This prevents the groundwater below from discharging and forces it to move sideways or upslope (Figure 12.15b). Soil around the clogged area eventually erodes, sending acid, metal ions and salts into waterways and dams, and an extended area with sulfidic material progresses up-slope or adjacent to the original area with sulfidic material. If the actions of cattle, or other activities, continue to disturb the
soil around the newly created sulfidic material, the area affected continues to expand up-slope (Figure 12.15b) (Fitzpatrick and Skwarnecki 2005). Stage 3
If these processes express on the surface of the soil, bare eroded saline scalds surrounding a core of slowly permeable, highly saline, eroded sulfuric material may result (Figure 12.15c). These saline landscapes are characterised by slimy red or white ooze and scalds with impermeable Fe-rich crusts. As shown in Figure 12.15a, b when the sulfidic materials undergo changes, different salt and Fe minerals form because of differences in pH and salt concentrations. In the final stage of formation, a hard soil layer remains, with only few
Soils and natural resource management
salts (Figure 12.15c). The acidification process accelerates the decomposition and formation of minerals in the soils and underlying rocks and can cause an increase in salinity and carbonate formation. 12.10.4 Descriptive 3D whole-of-landscape process models Fitzpatrick et al. (2003d) constructed the descriptive 3D whole-of-landscape process model (Figure 12.16, page 173) for a regionally representative upland hillslope near Jamestown, South Australia. The model characterises the catchment-scale variability of relict (past geomorphological processes in development of rock weathering and erosion) and current (saline, sodic and sulfidic soils) soil-forming processes. The model also explains the contemporary geochemical dispersion and erosion mechanisms present in the lower parts (erosion gully) of the toposequence (Figure 12.16, page 173) and, in particular, explains salt storage and salt mobilisation in this complex landscape dominated by both NAS and GAS (that is, groundwater induced, occurring in the lowest part of the landscape/ in the erosion gully with stream salinity). The model identifies a complex palaeovalley system derived from alluvium, which provides new insights into the soil– regolith, geological and hydrological features associated with salt stores in both upland soil surface features and in low-lying valley-fill sediments. These observations are placed in a regional 3D regolith-landform evolution model derived from the interpretation of airborne magnetics and gamma-ray spectrometry, digital terrain analysis, airborne EM and drilling (Wilford 2004a; Section 9.5). Electromagnetic (EM-38, EM-31) and volume magnetic susceptibility (VMS) surveys were used to rapidly characterise complex landscape patterns (Thomas et al. 2008a, Fitzpatrick et al. 2003d). These survey methods in conjuction with terrain analysis and 3D GIS terrain visualisation showed strong promise for obtaining high-intensity, non-intrusive, spatially continuous soil information that revealed salt accumulation and other pedological processes. Using the combination of approaches, the authors: (i) produced maps showing the aerial extent of shallow NAS and (ii) constructed a colour cross-sectional diagram or model to show the various saline and sodic soil horizons/layers and water
flow pathways (Figure 12.16, page 173). Furthermore, the model was then used to underpin the development of GIS method (upscaling) to predict and map the distribution of soil types and shallow NAS for the small region (2300 ha) surrounding the hillslope under study (Thomas et al. 2008b). These detailed toposequence descriptions and processes have the potential to be integrated into broader scale regolith–landscape models defined by airborne geophysical and terrain modelling techniques (Wilford 2004b; Section 9.5).
12.11 PICTORIAL MANUALS FOR LAND MANAGEMENT PLANNING Figure 12.17 shows the sequence of steps used to develop easy-to-follow pictorial manuals for identifying soil indicators, land-use options and best-management practices. Steps 1–5 describe soil layers and construct them in toposequences (descriptive, explanatory or predictive models), which are used to help map soil types in areas with variable geochemisty (Fitzpatrick et al. 2003b). STEPS 1 - 3
Describe landscape and soils down toposequence
STEPS 4 - 5
Identify and locate landscape degradation using visual and chemical indicators
STEP 6
STEP 7
Develop land management options Produce draft manual Consult with Landcare groups
STEP 8
Produce final manual (for farmers and advisers) Use recording sheet
STEP 9
Use aerial photograph and plasticised overlay
Figure 12.17: Flow diagram showing steps involved in developing manuals for land management (after Fitzpatrick et al. 2003b).
333
334
Regolith Science
Steps 6–9 involve local communities in developing the manual by integration and adoption, where knowledge of the hydrological and soil–regolith processes models (bottom half of Figure 12.18, page 174) and production systems are bought together in recommendations for appropriate best-management practices (top half of Figure 12.18, page 174). Several case studies from geochemically variable saline environments are listed in Table 12.4.
12.12 REFERENCES Allan MJ (1994). ‘An assessment of secondary dryland salinity in Victoria’. Technical Report No.14. Centre for Land Protection Research, Department of Conservation and Natural Resources, Bendigo, Victoria. Barrett-Lennard EG, George R, Hamilton G, Norman H and Masters D (2005). Multi-disciplinary approaches suggest profitable and sustainable farming systems for valley floors at risk of salinity. Australian Journal of Experimental Agriculture 45, 1415–1424. Berner RA (1984). Sedimentary pyrite formation: an update. Geochimica et Cosmochimica Acta 48, 605–615. Bigham JM and Ciolkosz EJ (Eds) (1993). ‘Soil color’. Soil Science Society of America Special Publication .O . 159 pp. Bigham JM, Fitzpatrick RW and Schulze DG (2002). Iron oxides. In Soil Mineralogy with Environmental Applications. (Eds JB Dixon and DG Schulze) pp. 323–366. Soil Science Society of America Book Series No 7. Soil Science Society of America, Madison, Wisconsin. Bradley MP and Stolt MH (2003). Subaqueous soillandscape relationships in a Rhode Island estuary. Soil Science Society of America Journal 67, 1487–1495. Brouwer J and Fitzpatrick RW (2002). Interpretation of morphological features in a salt-affected duplex soil toposequence with an altered soil water regime in western Victoria. Australian Journal of Soil Research 40, 903–926. Bui EN, Krogh L, Lavado RS, Nachtergaele FO, Toth T and Fitzpatrick RW (1998). Distribution of sodic soils: the world scene. In Sodic Soils: Distribution,
Properties, Management and Environmental Consequences. (Eds ME Sumner and R Naidu) pp. 19–33. Oxford University Press Inc., New York. Cass A and Fitzpatrick RW (1999). ‘Soil assessment for almond production on the Weman Development’. Confidential Report to Kyndalyn Park Almonds Pty Ltd. CSIRO Land and Water Client Report 99/40. Clark L, Fitzpatrick RW, McCarthy M, Murray R and Chittleborough D (2002). Vineyard soil degradation following irrigation with saline groundwater for twenty years. In Soil Science: Confronting new realities in the 21st century. Transactions of IUSS 17th World Congress of Soil Science. Bangkok, Thailand. (on CD). International Union of Soil Sciences, Wageningen, Netherlands. Clarke CJ, George RJ, Bell RW and Hatton TJ (2002). Dryland salinity in southwestern Australia: its origins, remedies, and future research directions. Australian Journal of Soil Research 40, 93–113. Coram JE, Dyson PR, Houlder PA and Evans WR (2001). ‘Australian groundwater flow systems contributing to dryland salinity’. A Bureau of Rural Sciences Project for the National Land and Water Resources Audit’s Dryland Salinity Theme. Bureau of Rural Sciences, Canberra. Corwin DL and Lesch SM (2003). Application of soil electrical conductivity to precision agriculture: theory, principles, and guidelines. Agronomy Journal 95, 455–471. Cox JW, Fritsch E and Fitzpatrick RW (1996). Interpretation of soil features produced by ancient and modern processes in degraded landscapes: VII. Water duration. Australian Journal of Soil Research 34, 803–824. Cox JW, Fitzpatrick RW, Davies PG and Forrester S (2002). ‘Salinity investigation at Second Ponds Creek’. CSIRO Land and Water Confidential Report to Rouse Hill Infrastructure Pty Ltd. Dear SE, Moore NG, Dobos SK, Watling KM and Ahern CR (2002). Soil management guidelines. In Queensland Acid Sulfate Soil Technical Manual: (version 3.8). Department of Natural Resources and Mines, Indooroopilly, Queensland. Demas GP and Rabenhorst MC (1999). Subaqueous soils: Pedogenesis in a submersed environment. Soil Science Society of America Journal 63, 1250–1257.
Soils and natural resource management
Dent D and Young A (1981). Soil Survey and Land Evaluation. George Allen and Unwin, London. Fanning DS (2002). Acid sulfate soils. In Encyclopedia of Soil Science. (Ed. R Lal) pp. 11–13. Marcel Dekker, New York. Fitzpatrick RW (1988). Iron compounds as indicators of pedogenic processes: Examples from the southern hemisphere. In Iron in Soil and Clay Minerals. (Eds JW Stucki, BA Goodman and U Schwertmann) pp. 351–389. NATO ASI Series C. Vol. 217. D. Reidel, Dordrecht, Netherlands. Fitzpatrick RW (2004). ‘Changes in soil and water characteristics of natural, drained and re-flooded soils in the Mesopotamian marshlands: Implications for land management planning.’ CSIRO Land and Water Client Report. 102 pp. (plus Appendices 6 and 7: 182 pp). Restricted Report for The Iraq Marshlands Restoration Program funded by the U.S. Agency for International Development (USAID) contracted under the Development Alternatives, Inc. (DAI) Water Indefinite Quantity Contract (IQC). Fitzpatrick RW and Merry RH (2002). Soil-regolith models of soil-water landscape degradation: development and application. In Regional Water and Soil Assessment for Managing Sustainable Agriculture in China and Australia. (Eds TR McVicar, L Rui, RW Fitzpatrick and L Changming) pp. 130–138. Australian Centre for International Agricultural Research, Canberra. Fitzpatrick RW and Skwarnecki MS (2005). Mount Torrens, eastern Mount Lofty Ranges, South Australia: regolith models of soil-water landscape degradation. In Regolith Landscape Evolution Across Australia: A compilation of Regolith-landscape Case Studies and Landscape Evolution Models. (Eds RR Anand and P de Broekert) pp. 220–225. CRC LEME, Perth. Fitzpatrick RW, Boucher SC, Naidu R and Fritsch E (1994). Environmental consequences of soil sodicity. Australian Journal of Soil Research 32, 1069–1093. Fitzpatrick RW, Fritsch E and Self PG (1996). Interpretation of soil features produced by ancient and modern processes in degraded landscapes. V. Development of saline sulfidic features in nontidal seepage areas. Geoderma 69, 1–29.
Fitzpatrick RW, Cox JW and Bourne J (1997). Managing Waterlogged and Saline Catchments in the Mt Lofty Ranges, South Australia: A Soil-landscape and Vegetation Key with on-farm Management Options. Catchment Management Series No. 1. CSIRO Publishing, Melbourne. Fitzpatrick RW, McKenzie N and Maschmedt DJ (1999). Soil morphological indicators and their importance to soil fertility. In Soil Analysis: An Interpretation Manual. (Eds KI Peverell, LA Sparrow and DJ Reuter) pp. 55–69. CSIRO Publishing, Melbourne. Fitzpatrick RW, Merry RH and Cox JW (2000a). What are saline soils? What happens when they are drained? Journal of the Australian Association of Natural Resource Management (AANRM). 3PECIAL )SSUE *UNE , 26–30. Fitzpatrick RW, Raven M, Self PG, McClure S, Merry RH and Skwarnecki M (2000b). Sideronatrite in acid sulfate soils in the Mt. Lofty Ranges: first occurrence, genesis and environmental significance. In New Horizons for a New Century. Australian and New Zealand Second Joint Soils Conference. Volume 2. 3–8 December, Lincoln University, New Zealand. (Eds JA Adams and AK Metherell) pp. 109–110. New Zealand Society of Soil Science, Christchurch, New Zealand. Fitzpatrick RW, Powell B, McKenzie NJ, Maschmedt DJ, Schoknecht N and Jacquier DW (2003a). Demands on soil classification in Australia. In Soil Classification: A Global Desk Reference. (Eds H Eswaran, T Rice, R Ahrens and BA Stewart) pp. 77–100. CRC Press, Boco Raton, Florida. Fitzpatrick RW, Cox JW, Munday B and Bourne J (2003b). Development of soil - landscape and vegetation indicators for managing waterlogged and saline catchments. Australian Journal of Experimental Agriculture 43, 245–252. Fitzpatrick RW, Merry RH, Cox JW, Rengasamy P and Davies PJ (2003c). ‘Assessment of physico-chemical changes in dryland saline soils when drained or disturbed for developing management options.’ CSIRO Land and Water Technical Report 02/03. CSIRO, Adelaide. Fitzpatrick RW, Thomas M, Davies PJ and Williams BG (2003d). ‘Dry saline land: an investigation using ground-based geophysics, soil survey and
335
336
Regolith Science
spatial methods near Jamestown, South Australia.’ CSIRO Land and Water Technical Report 55/03. CSIRO, Adelaide. Fitzpatrick RW, Baker AKM, Raven M, Rogers S, Degens B, George R and Kirby J (2005). Mineralogy, biogeochemistry, hydro-pedology and risks of sediments, salt efflorescences and soils in open drains in the wheatbelt of Western Australia. In Regolith 2005 – Ten Years of CRC LEME. (Ed. IC Roach) pp. 97–101. CRC LEME, Perth. Fitzpatrick RW, Powell B and Marvanek S (2006). Australian coastal acid sulfate soils – a national atlas. In Proceeding of the 18th World Congress of Soil Science, July 9–15 2006, Philadelphia, Pennsylvania. International Union of Soil Sciences. Fitzpatrick RW, Thomas BP and Merry RH (2008a). Acid sulfate soils. In Natural History of Gulf St Vincent. (Eds SA Shepherd, S Bryars, IR Kirkegaard, P Harbison and JT Jennings) pp. 106–120. Royal Society of South Australia Inc., Adelaide, South Australia. Fitzpatrick RW, Marvanek S, Shand P, Merry R, Thomas M and Raven M (2008b). ‘Acid sulfate soil maps of the River Murray below Blanchetown (Lock 1) and Lakes Alexandrina and Albert when water levels were at pre- drought and current drought conditions’. CSIRO Land and Water Science Report 12/08. CSIRO, Adelaide. Fritsch E and Fitzpatrick RW (1994). Interpretation of soil features produced by ancient and modern processes in degraded landscapes. I. A new method for constructing conceptual soil-water-landscape models. Australian Journal of Soil Research 32, 880–885, 889–907. Gardner WK, Fawcett JD, Fitzpatrick RW and Norton RM (2004a). Chemical reduction causing land degradation: I Overview. Plant and Soil 267, 51–59. Gardner WK, Fawcett JD, Fitzpatrick RW, Norton RM and Trethowan M (2004b). Chemical reduction causing land degradation: II Detailed observations at a discharge site in the Eastern Dundas Tablelands, Victoria, Australia. Plant and Soil 267, 85–95.
George RJ, Speed R and Commander P (1997). Hydrogeological models used for dryland salinity research and management: Western Australia. In Conceptual Models Workshop. 7th October. Australian Geological Survey Organisation, Canberra. Ghassemi F, Jakeman AK and Nix HA (1995). Salinisation of Land and Water Resources: Human Causes, Extent, Management and Case Studies. CAB International Publishing, Wallingford, UK. Gunn RH and Richardson DP (1979). The nature and possible origins of soluble salts in deeply weathered landscapes of eastern Australia. Australian Journal of Soil Research 17, 197–215. Heng LK, White RE, Helyar KR, Fisher R and Chen D (2001). Seasonal differences in the soil water balance under perennial and annual pastures on an acid Sodosol in southeastern Australia. European Journal of Soil Science 52, 227–236. Helyar KR (1990). Soil acidity in NSW, current pH values and estimates of acidification rates. Australian Journal of Soil Research 28, 523–537. Helyar KR and Porter WM (1989). Soil acidification. In Soil Acidity and Plant Growth. (Ed. AD Robson) pp. 61–101. Academic Press, Marrickville, NSW. Herriot RI (1942). The reclamation of highland ‘Magnesia’ patches, a preliminary note on work being conducted at Mt Bryan East. South Australian Journal of Agriculture 1942, 94–95. Isbell RF (1996). The Australian Soil Classification System. CSIRO Publishing, Melbourne. Isbell RF, Reeve R and Hutton JT (1983). Salt and sodicity. In Soils: An Australian viewpoint (CSIRO Division of Soils) pp. 107–111. CSIRO, Melbourne/ Academic Press, London. Keren R (2000). Salinity. In Handbook of Soil Science. (Ed. ME Sumner) G3-25. CRC Press, Boca Raton, Florida. Kennewell BK (1999). ‘Investigations into the management of dry saline land’. Primary Industries and Resources South Australia Technical Report 272. Adelaide. Lamontagne S, Hicks WS, Fitzpatrick RW and Rogers S (2004). ‘Survey and description of sulfidic materials in wetlands of the Lower River Murray floodplains: Implications for floodplain salinity management’. CSIRO Land and Water, Technical
Soils and natural resource management
Report 28/04 / CRC LEME Open File Report 165. CRC LEME, Perth. Lee SY (2002). Groundwater geochemistry and associated hardpans in southwestern Australia. MSc thesis. University of Western Australia, Perth. McDonald RC, Isbell RF, Speight JG, Walker J and Hopkins MS (1990). Australian Soil and Land Survey Field Handbook. 2nd edn. Inkata Press, Melbourne. McKenzie N, Jacquier D, Isbell R and Brown K (2004). Australian Soils and Landscapes. An Illustrated Compendium. CSIRO Publishing, Melbourne. Macumber PG (1991). Interactions between Groundwater and Surface Systems in Northern Victoria. Department of Conservation and Environment, Victoria. 345 pp. Maschmedt D (2000). ‘Assessing agricultural land: agricultural land classification standards used in South Australia’s land resource mapping program’. Primary Industries and Resources South Australia, Adelaide. Munsell Soil Color Charts (1994). Munsell – A Color Notation. 15th edn. Macbeth, Baltimore, Maryland. Naidu R, Sumner ME and Rengasamy P (Eds) (1995). Australian Sodic Soils. Distribution, Properties and Management. CSIRO Publishing, Melbourne. National Working Party on Acid Sulfate Soils (2000). ‘National Strategy for the Management of Coastal Acid Sulfate Soils’. NSW Agriculture, Wollongbar Agricultural Institute, Wollongbar, NSW. NLWRA (National Land and Water Resources Audit) (2001a). Australian Agricultural Assessment 2001. Volume 2. National Land and Water Resources Audit, Land and Water Australia, Canberra. NLWRA (2001b). Australian Dryland Salinity Assessment 2000. National Land and Water Resources Audit, Land and Water Australia, Canberra. Northcote KH (1979). A Factual Key for the Recognition of Australian Soils. 4th edn. Rellim Technical Publishers, Glenside, South Australia. Northcote KH and Skene JKM (1972). ‘Australian soils with saline and sodic properties’. CSIRO Soil Publication No. 27. CSIRO, Melbourne.
Nulsen RA (1993). Changes in soil properties. In Reintegrating Fragmented Landscapes. pp. 107–145. Springer-Verlag, New York. Peck AJ and Hatton T (2003). Salinity and the discharge of salts from catchments in Australia. Journal of Hydrology 272, 191–202. Peck AJ and Hurle DH (1976). Chloride balance of some farmed and forested catchments in south-western Australia. Water Resources Research 9, 648–657. Pons LJ (1973). Outline of the genesis, characteristics, classification and improvement of acid sulphate soils. In Proceedings of the 1972 (Wageningen, Netherlands) International Acid Sulphate Soils Symposium, Volume 1. (Ed. H Dost) pp. 3–27. International Land Reclamation Institute Publication 18. Wageningen, Netherlands. Post DF, Bryant RB, Batchily AK, Heute AR, Levine SJ, Mays MD and Escadafal R (1993). Correlations between field and laboratory measurements of soil color. In Soil Color. (Eds JM Bigham, EJ Ciolkosz and RJ Luxmoore). Special Publication No. 31. pp. 35–49. Soil Science Society of America, Madison, Wisconsin. Rengasamy P (2002). Transient salinity and subsoil constraints to dryland farming in Australian sodic soils: an overview. Australian Journal of Experimental Agriculture 42, 351–361. Rengasamy P and Sumner ME (1998). Processes involved in sodic behaviour. In Sodic Soils: Distribution, Properties, Management and Environmental Consequences. (Eds ME Sumner and R Naidu) pp. 35–50. Oxford University Press. Oxford. Rhoades JD, Chanduvi F and Lesch SM (1999). ‘Soil salinity assessment: methods and interpretation of electrical conductivity measurements.’ FAO Irrigation and Drainage Paper 57, Food and Agriculture Organization of the United Nations, Rome. Salama RB, Otto CJ and Fitzpatrick RW (1999). Contributions of groundwater conditions to soil and water salinisation. Hydrogeology Journal 7, 46–64. Saxton KE, Rawls WJ, Romberger JS and Papendick RI (1986). Estimating generalized soil-water characteristics from texture. Soil Science Society of America Journal 50, 1031–1036. Schoeneberger PJ, Wysocki DA, Benham EC and Broderson WD (Eds) (2002). Fieldbook for Describing
337
338
Regolith Science
and Sampling Soils. Version 2.0. Natural Resources Conservation Service, National Soil Survey Center, Lincoln, Nebraska. Scott BJ, Ridley AM and Conyers MK (2000). Management of soil acidity in long-term pastures of south-eastern Australia: a review. Australian Journal of Experimental Agriculture 40, 1173–1198. Scott DI (1990). Hydrogeological, geophysical and water balance study of dryland salinity at Dick’s Creek, Yass Valley, NSW. MAppliedSc thesis. University of NSW, Sydney. SCAV (1982). Salting of non-irrigated land in Australia. Soil Conservation Authority of Victoria. Government Printer, Melbourne. Semple WS and Williams BG (2002). Saline seepage scalds. Can they be usefully classified? Natural Resource Management 5, 13–21. Semple WS, Koen TB, Williams BG, Murphy BW and Nicholson AT (1996) ‘Saline seepage scalds in the Central West of NSW’. Technical Report No 29. Department of Land and Water Conservation, Sydney. Shaw RJ, Coughlan KJ and Bell LC (1998). Root zone sodicity. In Sodic Soils: Distribution, Properties, Management and Environmental Consequences. (Eds ME Sumner and R Naidu) pp. 95–106. Oxford University Press, New York Sinclair Knight Merz (2001). ‘Assessment of efficacy of engineering options for the management of dryland salinity’. Research Report Final–December 2001. National Dryland Salinity Program. Simpson SL, Apte SC and Batley GE (1998). Effect of short-term resuspension events on trace metal speciation in polluted anoxic sediments. Environmental Science and Technology 32, 620–625. Skwarnecki MS and Fitzpatrick RW (2003). ‘Regional geochemical dispersion in acid sulfate soils in relation to base-metal mineralisation of the Kanmantoo Group, Mt Torrens-Strathalbyn region, eastern Mt Lofty Ranges, South Australia’. CRC LEME Restricted Report 185R, CRC LEME, Perth. Skwarnecki M, Fitzpatrick RW and Davies PJ (2002). ‘Geochemical dispersion at the Mount Torrens lead-zinc prospect, South Australia, with emphasis
on acid sulfate soils’. CRC LEME Restricted Report No. 174. pp. CRC LEME, Perth. Soil Survey Staff (1987). ‘Sodic, sodic-saline, and saline soils of North Dakota’. Miscellaneous Publications Soil Conservation Service, US Department of Agriculture, Bismarck, North Dakota. Soil Survey Staff (1999). Soil Taxonomy - a basic system of soil classification for making and interpreting soil surveys. 2nd edn., USA Agriculture Handbook No. 436. United States Department of Agriculture, Natural Resources Conservation Service, Lincoln, Nebraska. Soil Survey Staff (2003). Keys to Soil Taxonomy. 9th edn. United States Department of Agriculture, Soil Conservation Service, Blacksburg, Virginia. Spies BR and Woodgate P (2005). Salinity Mapping Methods in the Australian Context. Commonwealth of Australia, Canberra. Stolt MH (2006). Glossary of Terms for Subaqueous Soils, Landscapes, Landforms, and Parent Materials of Estuaries and Lagoons. Department of Natural Resources Science, University of Rhode Island, Kingston, Rhode Island. (accessed 6/6/06). Sullivan LA, Bush RT and Fyfe D (2002). Acid sulfate soil drain ooze: distribution, behaviour and implications for acidification and deoxygenation of waterways. In Acid Sulfate Soils in Australia and China. (Eds C Lin, MD Melville and LA Sullivan) pp. 91–99. Science Press, Beijing. Tanji KK (2002). Salinity in the soil environment. In Salinity: Environment – Plants – Molecules. (Eds A Läuchli and U Lüttge) pp. 21–51. Kluwer Academic Publishers, Netherlands. Thomas M (2007). Multiscale prediction of salinesodic land degradation processes in two South Australian regions. PhD thesis. University of Adelaide, Adelaide Thomas M, Fitzpatrick RW and Heinson GS (2008a). Distribution and causes of intricate saline-sodic soil patterns in an upland South Australian hillslope. Australian Journal of Soil Research (in press). Thomas M, Fitzpatrick RW and Heinson GS (2008b). Upscaling saline-sodic subsoil patterns from a
Soils and natural resource management
hillslope to a region in upland South Australia. Australian Journal of Soil Research (in press). Tickell SJ (1994). ‘Dryland salinity hazard map of the Northern Territory’. Report 54/94D. Power and Water Authority, Northern Territory of Australia, Darwin US Salinity Laboratory Staff (1954). Diagnosis and Improvement of Saline and Alkali Soils. US Department of Agriculture. US Government Printing Office, Washington DC. Vepraskas MJ (1992). ‘Redoximorphic features for identifying aquic conditions’. North Carolina State University Technical Bulletin 301. Raleigh, North Carolina. White RE, Helyar KR, Ridley AM, Chen D, Heng LK, Evans J, Fisher R, Hirth JR, Mele PM, Morrison GR, Cresswell HP, Paydar Z, Dunin FX, Dove H and Simpson RJ (2000). Soil factors affecting the sustainability and productivity of perennial and annual pastures in the high rainfall zone of southeastern Australia. Australian Journal of Experimental Agriculture 40, 267–283.
Wilford JR (2004a). ‘Regolith-landforms and salt stores in the Angas-Bremer Hills’. CRC LEME Open File Report 177. CRC LEME, Perth. Wilford JR (2004b). 3D Regolith architecture of the Jamestown area – implications for salinity. CRC LEME Open File Report 178. CRC LEME, Perth. Williams BG and Bullock PR (1989). ‘The classification of salt-affected land in Australia’. Technical Memorandum 89/8. CSIRO Division of Water Resources, Canberra. Williams BG and Semple WS (2001). Implications of the chemical composition of saline seepage scalds. In Wanted Sustainable Futures for Saline Lands, Seventh National PURSL (Productive Use and Rehabilitation of Saline Land) Conference. 20–23 March, Launceston, Tasmania. (Eds C Grose, L Bond and T Pinkard) pp. 1–6. CRC for Plant-Based Management of Dryland Salinity, Perth. Wood WE (1924). Increase of salt in soil and streams following the destruction of the native vegetation. Journal of the Royal Society of Western Australia 10, 35–47.
339
13
Regolith sampling for geochemical exploration Charles R M Butt, Keith M Scott, Matthias Cornelius and Ian D M Robertson
13.1
INTRODUCTION
In many landscapes, well-developed weathering profiles show vertical zonation from the fresh rock through increasingly weathered saprolite, in places to lateritic and other duricrusts, and into soil (see Chapter 6). In low-lying or depositional areas, a cover of transported overburden may be present, forming the direct parent material of the soil. Each of these components of the regolith may be used as a geochemical sample medium, as may the vegetation that grows on it or the water that drains it. When mineralised rocks are weathered, many of the ore-associated elements are leached, but, nevertheless, the regolith, vegetation or water may contain anomalously high abundances, either of the ore elements themselves or of the accessory or pathfinder elements. Geochemical techniques seek to increase exploration success by sampling and analysing those materials that provide the largest potential target. Although some broadening of the target by chemical dispersion may be expected at depth, this is surprisingly rarely significant in the saprolite. Dispersion that increases target size is generally confined to the uppermost few metres. In outcrop, the usually Fe-rich gossan (a specific type of saprolite) formed by weathering of sulfide-rich mineralisation is commonly highly leached, degraded,
difficult to recognise and only small in size. However, a much more extensive geochemical target may be presented by residual and semi-residual soils (Figure 13.1), which may contain gossan fragments dispersed by physical erosion, or host elements dispersed chemically as the gossan itself weathered. Gossanous material may be also mechanically (or chemically) transported into a stream, so that even larger, but more distant, targets may be detected by sampling stream sediments or stream water (Figure 13.1). Mineralisation that has a lower sulfide content (such as Au mineralisation) is much less likely to develop a gossan, but may nevertheless give broad anomalies in surface media such as soils and stream sediments (Figure 13.2). Less-mobile elements hosted by resistant minerals (for example, Au, Sn in cassiterite) and elements that concentrate in Fe oxides (such as As) are also likely to be associated with lateritic pisoliths and ferruginous lag, and thus give widespread, near-surface anomalies (Section 13.6). Some elements concentrate in specific sample media: one of the most significant being the association of gold with pedogenic calcrete (Section 13.7). Where there is transported overburden, soil sampling is commonly ineffective, even if the cover is as thin as 2 metres. Partial and selective analytical techniques have been widely used to try to
342
Regolith Science
Truncated gossan - rich in Pb, Cu, Fe oxides + secondary carbonates and sulfides
Localised multi-element anomalies in lithosols - Fe oxides shedding from gossans
Anomalous stream sediments
Top of sulfides Massive sulfides Disseminated sulfides Supergene enrichment (Cu, Ag) Figure 13.1: A generalised dispersion model for weathering base metal sulfides (after Butt 2005a).
enhance any signature due to upwards chemical dispersion through overburden into soil, but with very limited success. The best option for detecting a lateral dispersion halo is to sample across the unconformity (interface) between the residual and transported material (interface sampling, see Section 13.10), although vegetation and groundwater sampling also have a role. This chapter provides an understanding of the mechanisms affecting the formation of geochemical anomalies in complex landscapes so that the appropriate selection of the possible sampling media can be made.
Soils are widely used during exploration because their geochemical signature largely reflects that of the parent material, but may be greater in area, due to additional mechanical and chemical dispersion during pedogenesis. However, some soils have developed partly or wholly from transported overburden, which, with rare exceptions, will show no geochemical relationship with the underlying bedrock. Nevertheless, not all anomalies in residual soils reflect mineralisation. Some may be due to secondary concentrations in ironstones or weathered bedrock that form the parent material; such possibilities must be considered in data interpretation.
13.2
13.2.2 Soil sampling procedures Soils are generally sampled from shallow pits or by augering – depending on the depth of soil and the horizon sought. Where soils are strongly differentiated (see Figure 6.3), care must be taken to sample a consistent soil horizon, but, except where a clean, deep pit is dug, some smearing or cross-contamination between
SOIL
13.2.1 Soil as a sampling medium Soils are almost ubiquitous across the landscape – being absent only on rock outcrops, actively eroding slopes, some ferruginous, siliceous or calcareous duricrusts, and active depositional areas such as talus, sand dunes and playa surfaces.
Regolith sampling for geochemical exploration
(a) Dominantly erosional regime - Calcareous soil, saline groundwater Au in pedogenic carbonate ± As, Sb, W in associated Fe oxide and lag (50->200m) Lateritic residuum: leached of base metals; retains some Au, pathfinders
Au anomaly in pedogenic carbonate in shallow colluvium
As, Sb, W, Mo, Bi, K retained through Au depletion zone
Au depletion zone
Supergene Au mineralisation
Groundwater anomalous in Au only
Supergene Au mineralisation
More sulfide-rich ore K-alteration halo + low grade Au mineralisation Broad, Fe-Mg (phengite) alteration halo
(b) Depositional regime Anomalies in lag potentially false
Au concentrated in carbonate horizon >5 - >200ppb in cover <10m thick
c
>100m c c
c
c
c
c c
c c
c
c c
c c c
c
c
c c c
c
c
c
Au depleted in upper saprolite but pathfinder elements remain anomalous
Primary multi-element (not Au) and alteration mineral signature present in saprolite
Dispersion of pathfinder elements in Fe nodules or at interface ->300m
Figure 13.2: A generalised dispersion model for weathering sulfide-poor Au mineralisation (after Butt 2005a).
343
344
Regolith Science
horizons is inevitable. Being least affected by recent pedogenetic processes, the C horizon (closest to the underlying saprolite) represents the most consistent sampling horizon. In poorly differentiated soils – for example, in some arid regions – it may be appropriate to simply sample below 100–150 mm, thus avoiding much of the transported component at the surface. Elsewhere, a more targeted approach is necessary – for example, to collect (or avoid) calcrete horizons. Litter and organic A horizon sampling is effective where vegetative cycling of metals is suspected, especially if the soil is developed in transported overburden (Anand et al. 2007). Sampling of the A horizon provides a readily obtainable, consistent sample type for regional sampling (Farrell and Orr 1980). Depending on the results from orientation or previous experience, either the whole sample may be split and crushed for analysis, or a particular size fraction chosen. Within any particular soil horizon – other than those developed from lateritic duricrust and gravels – most elements tend to be fairly uniformly dispersed regardless of size fraction. This can be more apparent than real, because dry sieving is commonly ineffective in separating fine fractions. Clay particles aggregate and can coat coarser particles so that clean separations are rarely possible and analytical results are accordingly modified. For example, at Higginsville, WA, Lintern (2005) found that the amount of -63 µm material was under-estimated by at least a factor of two when dry sieved and + 1 mm material over-estimated by a factor of up to six when dry sieved. Thus chemical abundances in coarse dry sieved material could be under-estimated by up to a factor of six. Nevertheless, fine soil fractions are commonly convenient because metal contents are enhanced by the removal of barren quartz, and no further sample preparation is necessary prior to analysis. Where there is substantial input of transported components, the soil response may be reduced by dilution. This may affect the whole soil or be biased towards particular (usually finer) fractions. Dilution by barren colluvium restricted the response of Au mineralisation at Beasley Creek, WA (Robertson 1999); in residual soil, a better contrast was obtained in the 710–4000 µm fraction than the whole soil, because of the high content of barren aeolian dust in
the 75–710 µm fraction. Similarly, in South-East Australia, Dickson and Scott (1998) and Tate et al. (2007) noted that the minus 100 µm fraction contained abundant aeolian material, which significantly diluted the geochemical signature of the commonly used minus 180 µm (80 mesh) soil fraction. Soil-sampling grids should be run across strike – using spacings appropriate to the expected target size. Provided that a consistent sample medium is used, the data can be contoured or presented as stacked profiles to highlight zones of interest. The use of a constant sample depth is not recommended because profile conditions can change markedly along a slope. Where such consistency is not possible, the sample types collected must be documented for use in subsequent interpretation. The need to determine optimum sampling horizons and size fractions by orientation surveys cannot be over-emphasised, nor can the need for adequate training and supervision of field samplers. (NB: Cuttings from a 0–1 m composite of an existing deep auger, RAB or RC hole as a surrogate for shallow soil augering during an orientation survey may give misleading results due to cross-hole contamination. The intended sampling procedure itself should be used.) Multi-element analysis of a soil or specific soil fraction is of great value in determining the significance of anomalies in areas the target elements may have been leached or, conversely, where hydromorphic dispersion may have led to secondary concentration. Relatively elevated concentrations of less mobile pathfinder elements (such as Sn, W, Pb or Pt) can indicate leaching of target elements, whereas their absence may indicate an anomaly is displaced or even spurious. For example, at the Mount Lindsay Sn prospect, Tasmania, As and Cu gave broader soil anomalies (B horizon) than Sn and W due to active hydromorphic dispersion (Ross and Schellekens 1980). At Lady Loretta, Queensland, shallow (100 mm) soil sampling reflects past widespread leaching and dispersion of Zn, compared with the relative immobility of Pb and Ag, which have been retained in gossans and little dispersed (Cox and Curtis 1977). Multi-element surveys tend to emphasise target and pathfinder elements; major element soil geochemistry is comparatively little used, but can be useful for:
Regolith sampling for geochemical exploration
s s s s
s
defining lithological changes in soil-covered terrains recognising geological marker horizons in monotonous sequences mapping alteration haloes associated with mineralisation normalising data when minor element abundances are partly controlled by the presence of specific minerals, such as Fe oxides, Ca and Mg carbonates; see also Section 5.4.3 determining the degree of surface leaching or evolution towards lateritisation, for example, by comparing the ratios of Si or Fe contents with those of the alkali or alkaline-earth elements (see also Section 6.8).
13.3 WEATHERED BEDROCK (SAPROLITE) The chemical alteration responsible for the transformation of bedrock to saprolite includes:
s
s
leaching of soluble alkalis, alkaline earths and some transition metals from less-stable minerals (sulfides, carbonates, ferromagnesian minerals and alumino-silicates such as feldspar) by groundwater, transforming them to stable secondary minerals – principally clays and Fe oxides. Resistant minerals (quartz and heavy minerals) are retained and less soluble constituents (Ba, Ti, Cr and Zr) are relatively enriched (Chapters 5 and 6). precipitation, in specific environments, of silica, Fe oxides and Ca-Mg carbonates as absolute enrichments.
Through these processes, primary minerals are pseudomorphically replaced and the precipitated oxides and carbonates form cements that preserve rock fabrics. The principal secondary minerals of the saprolite are Fe oxides, kaolinite and smectite. Uncemented saprolite is mostly soft – especially towards the top, where it may become increasingly clay rich. It has a wide range of colours: green, grey, yellow, brown, red and black. When pale grey or white, it is commonly referred to as the ‘pallid’ or ‘leached zone’ (see also Section 6.2); however, these terms are best used as descriptors, because other units, including sedi-
ments, may also be pallid or leached but have quite different properties. 13.3.1 Weathered rock as a sampling medium The expected target size for a given ore deposit type in saprolite depends largely upon the chemical mobility of the elements during intense weathering under humid conditions and during subsequent, commonly arid, episodes. In general, the less mobile the element and the deeper (that is, the nearer to fresh rock) the sample is taken, the smaller the anomaly size. Of the usual indicator elements, those immobilised in Fe oxides, clay minerals or stable secondary minerals (Cu, Mo, As, Sb and Pb) form strong anomalies in saprolite derived from mineralised rocks. Conversely, weathering can leach mobile elements such as Zn almost entirely, especially from the near-surface (Teutonic Bore: Greig 1983; Butt 1992), so that the detectable anomaly is again restricted. Primary alteration haloes are preserved if the elements are held in resistant minerals such as rutile, tourmaline, muscovite or barite (for example, Scott et al. 1994; Appendix 2). Secondary enrichment of target metals – derived largely from bedrock with only background abundances (whether hosted by silicates or minor sulfides) – may give false anomalies; for example, concentration of Cu along faults and contacts (Killara: Butt 1979, 2005b) and of olivine-derived Ni over ultramafic rocks. Such enrichment may, however, become secondary mineralisation in its own right – importantly including bauxite, Ni+Co laterite and some Mn deposits. Supergene Au and Cu deposits are also enrichments in saprolite, which have developed over primary mineralisation that itself may be sub-economic. 13.3.2 Sampling procedures Outcropping saprolite can be chip-sampled as an adjunct to gossan, ironstone or laterite sampling. However, as a survey technique, this is liable to bias because only the most resistant or secondarily hardened rocks are exposed, and these may not be representative of the unexposed rocks. Trenching (costeaning) – if logistically possible – is useful for local investigations where the saprolite is soft. Elsewhere, drilling is used.
345
346
Regolith Science
Rotary air blast (RAB), reverse-circulation blade, roller, percussion or air-core drilling are routinely used for saprolite sampling. Auger drilling can be used to sample the upper saprolite – generally only within about 5–15 m of the surface – although this may not be possible where there is strong induration (ferruginous, siliceous or calcareous duricrusts), or where the surface horizons are very wet. A number of strategies can be adopted. These include drilling: 1. to a set depth, to fresh bedrock and/or until the drill cannot penetrate farther 2. until ‘recognisable saprolite’ is visible in the cuttings 3. in areas of deep transported overburden, to base of the overburden and collecting either an ‘interface sample’ across the unconformity (see Section 13.10) or to the uppermost in situ material, whether having recognisable rock fabrics or not. The drill spacing is critical, especially where little or no secondary dispersion into saprolite is expected. Where samples are collected deep in the saprolite, dispersion may be so restricted that the anomaly is of much the same width as the primary mineralisation itself. For narrow targets – such as massive sulphide lenses, where there is little lateral dispersion – vertical drilling on centres of 10 or even 5 m across strike is commonly used. Rather wider intervals may be adequate if overlapping angle drilling is used, but care must be taken to allow for the effects of strong leaching, depletion and, potentially, transported overburden. Samples obtained by these methods are generally composites of cuttings from 1–2 m intervals. It is best if the cuttings are passed through a cyclone to minimise dust-loss, and collected either into plastic bags or into spoil heaps on the ground. Wet samples are collected in calico bags, although these may leak (for example, Smee and Stanley 2005). Analyses are generally on complete samples; size fractionation in the cuttings is likely to be an artefact of the drilling procedure. Analytical samples are generally collected by:
s s s
grab sampling from individual spoil heaps splitting the cuttings as they leave the cyclone pipe- or spear-sampling cuttings in plastic or calico bags (Barnes, 1987). The bag is laid on its side and
a PVC pipe, 50–60 mm in diameter, is inserted diagonally through the cuttings; one or more samples are taken from each bag and combined. Ideally samples should be collected throughout the profile, to take advantage of the dispersion haloes of the target and pathfinder elements in different horizons of the regolith. However, for reasons of economy, aliquots of several samples are commonly composited for analysis and the individual sub-samples only analysed separately if anomalous concentrations are found. Alternatively, samples may be collected:
s s s s s
from one or more ‘recognisable horizons’ from several set depths of specific features, such as redox fronts from where the strongest ‘colour’ occurs (red, yellow or brown) to target Fe oxides at the ‘bottom-of-hole’, especially if this is close to fresh rock.
Accurate logging of drill material is essential for correct selection of analytical samples and for data interpretation. This important, generally complex, aspect of saprolite sampling is often neglected or done poorly or inconsistently. Too commonly, logging is done in a very cursory manner during the supervision of a drilling program, and not subsequently checked or repeated. It is not, however, an easy task, and it can be very difficult to subdivide saprolite, in particular, into useful or consistently recognisable units, especially where it is powdered by drilling or features are obscured by strong colour variations. Commonly, the best that can be achieved is to establish boundaries such as ‘top of fresh rock’ (weathering front), ‘base of complete oxidation’, and ‘base of alluvium’ (unconformity between transported and residual units). Some consistency can be achieved by displaying small samples of drill cuttings from representative drill profiles in ‘chip trays’ or other containers. Comparison of an array of chip trays representing a drill traverse can assist in the recognition of larger-scale features. If the chips are wet sieved from a grab sample of drill cuttings, there is a danger that they may not be representative of the whole interval – especially from middle and upper parts of the saprolite. The resistant fragments may be from a specific small feature, such as a quartz vein or minor Fe
Regolith sampling for geochemical exploration
oxide mottling, whereas the bulk of the sample is clay and has not been retained. However, wet sieving is a useful procedure; indurated or less-weathered remnants form coherent, coarser materials that preserve fabrics useful for lithological logging. Identification of bedrock lithology from drill cuttings can be difficult, but is assisted by maintaining a library of representative, properly characterised, samples, and by consulting appropriate reference documents (for example, Robertson and Butt 1993). Instrumental techniques for the rapid logging of weathered and fresh rocks are being developed and should increase the information available from drill cuttings and core. Saprolite samples should be analysed for a range of ore and pathfinder elements, including those that might represent the wider exploration target presented by primary alteration haloes (such as potassic alteration associated with orogenic Au mineralisation) and even lithological indicators. This is especially important where drilling intervals are wide and for deep samples close to bedrock. Total analysis is preferred. Chemical analysis can usefully be supplemented by rapid mineralogical analysis using visible to infra-red spectral techniques (such as PIMA® or ASD® ; Section 4.8), targeting alteration haloes represented by specific mineralogical or lithological associations, such as potassic alteration, indicated by muscovite, which are readily detected by these techniques (Scott 1996).
13.4
GOSSANS AND IRONSTONES
A gossan is the weathered expression of a rock that contains substantial amounts of sulfide mineralisation: that is, it is the ferruginous saprolite developed from weathered sulfides (see also Section 5.3.8). The term has no economic connotation and may be applied to the weathered product of any sulfide, including barren pyrite. Some gossans may have boxworks (casts after the original sulfides), but, commonly, textural indications of parentage are absent. Gossans derived from Fe-rich sulfide assemblages typically consist largely of Fe oxides, whereas gossans formed from the weathering of Fe-poor sulfides (for example, galena and sphalerite in carbonate hosted Pb-Zn deposits)
may be siliceous or have high Mn contents. A more full discussion of different gossan types is given by Butt et al. (2005b) and Taylor and Thornber (1992). Ironstone is a general term that describes any highly ferruginous weathered outcrop, including gossans. Ironstones may or may not be derived from sulfides, but gossans are all derived from sulfides. Some ironstones are essentially linear in outcrop – following an underlying geological unit or structure. Stratigraphic ironstones are largely residual accumulations on Fe-rich lithologies, such as Fe-rich carbonates (Taylor 1973). Ironstones along faults, lithological contacts, bedding planes and other channel ways are formed by precipitation of Fe – derived from weathering of country rocks – in these more oxidising zones of water flow (Figure 13.3). All these ironstones may resemble gossans, and may have concentrated one or more base metals (for example, Killara Cu pseudogossan; Butt 1979, 2005b). Fault ironstones can contain a high proportion of quartz breccia. Siliceous or ferruginous boxworks after primary or secondary minerals, such as carbonates, garnet or olivine, may superficially resemble sulphide boxworks. Strongly jointed and textured rocks also give rise to similar cellular structures. Because it is commonly not possible to recognise a gossan in a hand specimen, it is better to refer to any very ferruginous field sample as an ‘ironstone’ until definite evidence of a sulfide parent is obtained by petrographic or geochemical study, or by drilling. Ironstones with a fabric and/or composition suggestive of a gossan, but not developed over sulfides, are referred to as false gossans or pseudo-gossans. Other ironstones may be more-or-less conformable with the land surface, as an horizon in the regolith. Examples include lateritic duricrusts – either at surface or buried beneath transported overburden – and ferruginous pans, which are precipitated at the level of past water tables or redox fronts; in each of these, the original rock fabrics may be partly preserved or completely destroyed. Ferricretes and seepage ironstones have formed by the precipitation of Fe oxides (and, in places, Mn oxides) at breaks of slope, swamps and along watercourses. Botryoidal, cellular, pisolitic and nodular structures are common in lateritic duricrusts and ferricretes.
347
348
Regolith Science
Lateritic gravels and duricrust
(a)
Soil or alluvium
# #
Mottled zone
#
#
# #
#
#
#
#
Oxide zone
Saprolite
Base of weathering
Carbonate and sulfate zone Supergene sulfide zone Transition zone Unweathered rock Primary sulfide zone
Lateritic ironstone
(b)
Translocated laterite Ferruginised wallrock Gossan Translocated gossan Solution-deposited gossan
#
#
#
Oxide zone
Saprolite
Leakage gossan Stratigraphic ironstone Fault ironstone
#
Carbonate zone Sulfate zone Water table
Drainage ironstone
Supergene zone Fresh bedrock
Fresh sulfide Fault
Iron-rich formation
Figure 13.3: (a) Generalised profile of a mature gossan above massive sulfides in a semi-arid terrain. (b) Types of gossans and ironstones (after Butt et al. 2005b).
Regolith sampling for geochemical exploration
13.4.1 Gossans as a sampling medium Many of the major sulfide ore deposits of Australia (such as Broken Hill, Mt Lyell, Mt Morgan, Mt Isa and Kambalda) were found by recognition of outcropping gossans (Taylor et al. 1980). However, today, recognition of gossanous material is more likely to be needed when drilling through transported overburden. Once an ironstone has been identified as a gossan, attempting to determine whether the sulfides from which it is derived are likely to be economic is difficult because the composition of the gossan does not necessarily reflect that of the parent sulfides. For example, Zn may be highly depleted (for example, Cox and Curtis 1977) and Au can be enriched (Scott et al. 2001) during gossan formation. Thus, a thorough understanding of the weathering processes, and any influence of the wall rocks, is needed before attempting a prediction with confidence. Multi-element geochemistry
Analysis for target elements alone is commonly inadequate for correct identification of gossans: either because these may be leached from true gossans or, conversely, concentrated in ironstones with no sulfide precursor. Specific pathfinder elements or suites of elements may be diagnostic. For example, Cu-rich ironstones are common, but most are unrelated to mineralisation and lack associated pathfinder elements; examples include the Cu-rich ironstone on a shale–dolerite contact (Killara pseudogossan; Butt 1979, 2005b), anomalous Cu in stratigraphic ironstones (Taylor 1973) and Cu in fault ironstones (even when partially derived from weathering sulfides: Bampton et al. 1977). Conversely, the presence of As is a good indicator of the significance of Cu anomalies, because it is derived from pyrite associated with the Cu mineralisation (Glasson 1973; Taylor 1973). General multi-element suites for various mineralisation types are given in Table 13.1. The geochemical data are best analysed using multivariate statistical techniques (see also Section 5.5.3). These are more effective if the variables have normal distributions; however, depending on the distribution of each element, log or power transformations may be needed. Well-characterised reference data sets are required – representing the principal types of gossans
Table 13.1: Elemental associations in various mineralisation types for gossan identification (after Taylor and Thornber 1992).
Host rock
Target mineralisation
Target and pathfinder elements
Mafic-ultramafic volcanics
Ni-Cu
Ni, Cu, Co, Pt, Pd, Ir, Se, Te. (Cr, Mn, Zn)*
Felsic volcanics
Massive sulfides
Cu, Pb, Ag, Au, As, Sb, Bi, Se, Hg, Sn, Ba.
Cu-Mo-Au
Cu, Mo, Au, Re.
Pb-Zn-Ag
Pb, Zn, Ag, Cu, As, Hg, Sb. (Mn, Ba, Co, Ni)
Cu
Cu, As, Pb, Sb, Ag Hg. (Mn, Zn, Co, Ni)
Carbonates
Pb-Zn
Pb, Zn, Cd.
Skarns
Cu-Zn, Pb-Zn, Sn, W, Au-Ag
Cu, Pb, Zn, Sn, W, Au, Ag, Mn.
Sedimentary rocks
*Elements in parentheses are negative indicators. Al, Si Cr, V, Ti, Mn and P are enriched in lateritic ironstones and, hence, are also generally negative indicators.
and other ironstones in the survey area, based upon major and trace element data. Some of most effective procedures include stepwise discriminant analysis (for example, Taylor and Scott 1982) and canonical variate analysis (for example, Smith et al. 1983; see also Section 5.5.3). Isotopic ratios
Positive identification of gossans may also be made by using with Pb and S isotopic ratios, due to the differences in isotopic signatures of country rocks, barren sulfides and potentially economic mineralisation. Separation of Pb from U and Th to form an ore body at a particular time gives it a unique signature of 208Pb/206Pb, 207Pb/206Pb and 206Pb/204Pb isotopic ratios. In Pb-rich deposits, the signature is homogeneous and does not change appreciably after formation, due to the high initial Pb/Th and Pb/U ratios. However, in Pb-poor orebodies, the isotopic ratios are not homogeneous and change with time, due to the formation of radiogenic Pb. The ratios may also vary markedly within the deposit, because of variations in
349
350
Regolith Science
initial U/Pb and Th/Pb ratios. The isotopic signatures are not influenced by weathering. Because barren ironstones contain Pb (and other base metals) derived from a variety of sources during weathering and have a random Pb isotope signature, gossans and ironstones are readily differentiated isotopically in the Mount Isa region (Gulson 1986). Similarly, studies of numerous sulfide deposits and their gossans have shown that individual metallogenic provinces have distinctive S isotopic signatures that are not affected by weathering (for example, Taylor et al. 1984). These signatures may be significantly different from those sulfates formed by precipitation from groundwaters (see Section 13.13). 13.4.2 Gossan sampling procedure Because the formation of gossans commonly involves multiple events during an extended period – leading to variable amounts of ferruginisation, silicification and calcification – individual samples are chemically highly variable. Thus, it is desirable to collect a number of discrete samples from each occurrence to fully evaluate it (N.B: more information is gained from a group of individually analysed samples than from bulking the samples prior to analysis). However, in many cases where the ferruginous material is obtained by drilling, multiple samples may not be available and identification of the material as a gossan must rely on comparison with known gossans in the area and determining whether the composition could have a sulfide source. A list of the more commonly occurring minerals found in gossans is given in Table 4.2.
13.5
LATERITIC RESIDUUM
Lateritic residuum is a collective term for the upper ferruginous zone of the lateritic profile (Figure 13.4; see also Section 6.2), and consists of: -
s s
lateritic gravel: an unconsolidated horizon of pisoliths and nodules (2–64 mm in size), and/or lateritic duricrust: an indurated horizon that may retain primary structural or lithic fabrics, or have secondary fabrics, including pisoliths and nodules.
Where both horizons occur together, the lateritic gravel commonly overlies the lateritic duricrust. Lateritic residuum is composed dominantly of secondary
Fe oxides (goethite, hematite and maghemite), with minor Al hydroxides (such as gibbsite and boehmite) and kaolinite, with or without quartz (Anand et al. 2002). Exotic materials of colluvial or aeolian origin may be intermixed through lateritic residuum by soilforming processes during and after lateritisation. The distinction between pisoliths and nodules is based on shape: pisoliths are spherical or ellipsoidal, whereas nodules are irregular. They may have one or more coatings or skins (cutans). An intact cutan on a pisolith or nodule implies that physical transport has been minimal. Pisoliths and nodules may be lithic or non-lithic. Lithic forms preserve rock fabrics and consist dominantly of hematite, goethite, kaolinite and gibbsite. Non-lithic forms have a clay- or sandrich core, with one or more generations of Fe oxides (hematite, goethite and/or maghemite) disseminated through or replacing it. Both forms may retain resistant minerals in the cores. Lateritic duricrust exhibits a variety of gross fabrics, including:
s
s s s s
lithic – essentially as intensely ferruginised saprolite and/or as a brecciated saprolite, showing the first stages of collapse. Primary structures, including shears and lithological contacts, are retained FRAGMENTAL – with angular clasts massive – generally with small, partly filled voids; VERMIFORM – having numerous filled or partly filled tubes or pipes (possibly after roots) in duricrust with massive or lithic fabrics PISOLITIC OR NODULAR – both clast-supported and matrix-supported. The cores of pisoliths and nodules may be lithic or non-lithic.
Similarly, lateritic gravels may have lithic and nonlithic components. Over Fe-poor rocks, such as sandstones and some granitoids, lateritic residuum may be poorly developed – represented only by mottles, nodules or pisoliths spaced within a sandy matrix – or absent. 13.5.1 Lateritic residuum as a sample medium The basis for the use of lateritic residuum as a sample medium (commonly loosely referred to as ‘laterite sampling’) is its close genetic relationship with underlying bedrock and its high content of Fe oxides. This relationship is demonstrated by the preservation of
Regolith sampling for geochemical exploration
0
2
4
6 Depth (m) 8
Lateritic residuum fragmental, nodular, pisolitic or massive duricrust and/or close packed nodules and pisoliths Ferruginous breccia (collapsed mottled saprolite) Mottled saprolite Ferruginous saprolite Saprolite
~50
Figure 13.4: A typical weathered profile – including lateritic residuum at the top – passing down through ferruginous breccia of mottles, mottled and ferruginous saprolite and saprolite (after Anand et al. 2002).
primary rock fabrics in duricrusts, nodules and lags, which results in proximal geochemical anomalies. Where there has been lateral movement in the surface horizons during weathering, the value of the lateritic residuum as a sample medium may be increased, so long as the movement has been minor in the context of the whole landscape and the scale of exploration. Thus, cassiterite grains and gossan fragments in nodules, several hundred metres from the gossan outcrop and sub-crop at the Golden Grove-Scuddles Cu-Zn-Au deposit (Smith and Perdrix 1983), are reflected by a widespread geochemical anomaly. A geochemical atlas of the western Yilgarn Craton, based on laterite sampling on a 9 km triangular grid demonstrates the effectiveness of the technique at a regional scale – potentially extending the prospective areas and identifying greenstone remnants in granite-dominated areas (Cornelius et al. 2007). 13.5.2 Lateritic residuum sampling procedure Sampling of lateritic gravel and duricrust should always be in a regolith–landform context to avoid inadvertent sampling of ferricrete or other ‘non-lateritic’ material. Plateau surfaces on lateritic residuum
adjacent to small erosional scarps (‘breakaways’) provide much geological information and good sample sites, even in areas of thick sandplain. Suitable sample sites in sand-covered terrain are may be relatively small (5–10 m across) and difficult to find except from the air. Here, a regolith map, Landsat (TM) image or airborne radiometrics may assist identification of suitable sample sites. Where shallow soil or sand covers lateritic residuum, it has to be excavated, which requires a good understanding of the landform setting to ensure success. Where the profile is stripped to saprolite, remnants of lateritic residuum may be preserved in shallow pockets 0.5–1 m in diameter. Locally derived (colluvial) ferruginous detritus, or lateritic lag, may be trapped in cracks and depressions. Where truncation is severe, however, even this may be lost, so that laterite sampling cannot be used. Regional sampling should be without regolith or bedrock bias – unless there is an established connection with the targeted commodity. For example, preferential sampling of ferruginous material near quartz veins could bias the samples towards sub-economic vein-hosted mineralisation because their geochemical signature could obscure a weak, regional, multielement signature of a larger ore body of a different type. Similarly, lateritic gravel should be collected without bias towards specific types of pisolith or nodules. However, those with well-developed cutans are preferred because they imply restricted transport. Physical characteristics (shape, colour and size) should be documented to assist with interpretation; Smith et al. (2000) discuss preferred sample hierarchies for different terrains (exposed lateritic residuum, lateritic residuum buried beneath terrestrial sediments and beneath sedimentary basins). Laterite samples need to be representative, and a minimum of 1 kg should be collected over a 10 m radius, if sufficient material is available. Where lateritic gravel is scarce, the search radius may have to be larger. Drill spoil from reconnaissance or past exploration drilling may also yield samples in extensive depositional areas. Because less mechanical dispersion has generally occurred during their formation, duricrusts are more suited to later stages of exploration and the definition of drill targets than lateritic gravel. Analysis by a variety of sensitive total analytical techniques is essential to ensure detection of elements hosted by
351
352
Regolith Science
resistant minerals. Sampling strategies for regional surveys are reviewed by Cornelius et al. (2001).
13.6
LAG
Lag is the accumulation of coarse, usually hard, fragments of diverse origins or compositions that forms a veneer or pavement on the land surface. In an exploration context, lag commonly refers to highly ferruginous materials, including pisoliths, nodules, fragments of ferruginous saprolite, mottles and other segregations. In some circumstances, other materials such as calcrete fragments, could be targeted. 13.6.1 Lag as a sample medium Regolith–landform control is very important when using lag as a sample medium. Lag is effective in relict and erosional regimes: that is, areas characterised by widespread preservation of the lateritic mantle and those where bedrock, saprolite or mottled zone are exposed at surface or covered by thin soil or locally derived colluvium. Surface lag is very unlikely to show the geochemical signature of the residual regolith or bedrock in depositional regimes where there is a cover of exogenous sediments several metres thick. The use of buried lag – along the unconformity – may be used in the same manner as surface lag (see interface sampling: Section 13.10). Geochemical anomalies in fine lag are generally broader than those in soil, due to enhanced mechanical dispersion at the surface (Robertson 1996). Lag geochemistry may show a better anomaly contrast than soil, such as for Ni sulfides in the Eastern Goldfields WA, (Carver et al. 1987). Where the composition of lag varies strongly – for example, due to dilution by coarse quartz sand – normalisation of trace element concentrations based on the Fe content is recommended. For some elements (such as Cu and Zn), however, at higher Fe concentrations, changes in the type of Fe oxide mineral can result in lower abundances and affect this procedure (McQueen et al. 2004). The interior fabric of lag can provide evidence of the lithology of underlying rocks and their weathering history. Lag derived from ferruginous saprolite is considered most useful, although duricrust-derived lag may also retain lithic remnants (Robertson 1995).
13.6.2 Lag sampling procedure Lag is sampled by simply sweeping unconsolidated material from the surface and screening particles (Carver et al. 1987). Both coarse (10–50 mm) and fine (0.5–10 mm) lag are effective sampling media, but the use of coarse lag in Au exploration requires large (much greater than 1 kg) samples to minimise any potential nugget effect. Coarse lag tends to resist aeolian and sheet-wash action better than fine lag and therefore may show less lateral dispersion – making it more suitable for follow-up (Robertson 1996). Fine lag can be obtained even in areas of dune cover, such as the Paterson Province within the Great Sandy Desert, Western Australia (WA), where termite mounds are abundant and termite activity has transported fine lag to the surface where it is redistributed (Carver et al. 1987). An orientation survey will determine the best size fraction for a given commodity and specific area. Where lag is sparse, the relatively coarse, ferruginous fraction (0.5–10 mm) of the soil – sieved out on site – may provide a sample equivalent to fine lag. Commonly, some of lag may be magnetic, due to maghemite which forms on the surface, at least in part, as the result of heating by bush fires (Anand and Gilkes 1987). Maghemite lag can be separated using a handheld magnet and provides a very consistent sample. However, in general, it may not be an ideal sample medium because non-magnetic gossan fragments, enriched in pathfinder elements are specifically excluded, as noted at Bottle Creek (Robertson and Wills 1993; Anand 2001) and Beasley Creek (Robertson 1989). As no significant advantage can be gained by separating and using the major non-magnetic component, analysis of the total lag is recommended. Nonetheless, use of magnetic lag may be the only practical method if ferruginous components are sparse, such as in regions where aeolian cover is widespread.
13.7
PEDOGENIC CALCRETE
Calcrete is a generally indurated material formed by the in situ cementation or replacement, or both, of pre-existing regolith by authigenic calcite and/or dolomite (or, rarely, aragonite) precipitated from soil water or groundwater (the term ‘calcrete’ is commonly used to refer to both indurated and friable or powdery
Regolith sampling for geochemical exploration
forms; Netterburg, 1967; Chen et al. 2002). There are two principal genetic types:
s s
pedogenic or vadose calcrete groundwater or phreatic calcrete, with numerous varieties between these end members.
Pedogenic calcretes have considerable significance in exploration in semi-arid regions, due to their ability to concentrate Au preferentially (Lintern et al. 1992; Lintern 2002; see below); conversely, they may also act as a diluent to base metals (Garnett et al. 1982). Groundwater carbonates and associated sediments host U mineralisation – predominantly as carnotite (K 2 (UO2)2 V2O8.3H2O). A detailed discussion of the characteristics and distribution of calcretes in Australia, and their use in mineral exploration, is given in Chen et al. (2002). Pedogenic (vadose) calcrete
Carbonates are precipitated in the unsaturated zone of soil profiles – commonly within 1–2 m of the ground surface, but deeper in some locations. They occur across southern Australia south of 30°S. In the Yilgarn Craton, this boundary is referred to as the ‘Menzies Line’: a narrow gradational zone having quite different vegetation, soil and groundwater characteristics to the north and south (Butt et al. 1977). Gold is associated with both calcareous soils and nodular calcretes south of the Menzies Line in the Yilgarn Craton, and with massive calcretes in the equivalent region of the Gawler Craton, SA (Lintern 2002). Some pedogenic carbonates occur farther north, such as in dune sands in central Australia, but no Au-calcrete association has yet been recorded there. Where pre-existing weathering profiles have been partly or completely stripped, the Ca is largely derived from the weathering of primary minerals; in such areas, there can be a marked lithodependence, with calcrete development strongest on mafic rocks. However, calcrete may also develop in highly leached, fully preserved, lateritic profiles, even in upland sites. Here, the derivation of the Ca is less clear, but it is probably from:
s
aeolian accession of CaCO3 deflated from coastal deposits or calcareous soils and sediments lower in the landscape
s s s
accession, with other ions, as aerosols with rainwater capillary rise and surface concentration under an evaporation gradient uptake by plants from deep in the profile and released at the surface when they decay.
Groundwater (phreatic) calcrete
Carbonates are precipitated in the saturated zone, at or below the water table, in colluvial and alluvial sediments that infill broad valley systems (Butt et al. 1977; Deutscher et al. 1980). They typically occur as massive bodies varying in width from a few hundred metres to several km, and to over 100 km in length, with longitudinal gradients generally less than 1:1000. Thickness is commonly 5–10 m but may exceed 30 m along valley axes. They are major aquifers and, where they debouch into playas, they broaden out to deltaic platforms. Groundwater calcretes commonly form positive features in valley axes, forming mounds three metres or more above flanking alluvial plains. Upward growth is due to active carbonate precipitation at or below the water table. The calcrete may show weak layering and a variety of heave structures, including pseudo-anticlines and diapirs, with karst and collapse structures near the surface. Exploration case histories for carnotite uranium deposits associated with groundwater calcretes are described by Mann and Deutscher (1978), Butt et al. (1977), Deutscher et al. (1980) and Butt (1988) (see also Section 13.13). 13.7.1 Pedogenic calcrete as a sampling medium In some environments, the calcrete horizon represents a pH contrast to underlying neutral to acid regolith and may cause the precipitation and concentration of leached metals, hence forming epigenetic anomalies at the base of the calcrete horizon. On the Yorke Peninsular, SA, there is a significant Cu response at the calcrete–clay interface to partial extraction (acid ammonium acetate) analyses (Kadina: Mazzucchelli et al. 1980) and total analyses (Poona: Keeling and Hartley 2005). However, in other cases, pedogenic calcrete may represent an absolute addition to soil, diluting and
353
354
Regolith Science
depressing anomaly contrasts. The effects of dilution were demonstrated in South Africa by Garnett et al. (1982), who found that abundances and anomaly contrasts at the Putsberg Cu deposit could be increased by removal of the carbonate by an ammonium acetate leach and analysing the residue. Similarly, strong dilution of Ni and Cu in the calcrete horizon, compared with underlying saprolite, was noted at Pioneer, near Kambalda, WA (Cox 1975). However, in both these cases, the calcrete was developed in residual or semiresidual material and did retain a response and, hence, was considered a more reliable sample medium than overlying soil, which has a much higher proportion of transported components. More importantly, gold can be enriched in the calcareous horizons of soils and may give rise to or enhance a near-surface expression to concealed primary or secondary mineralisation. This property has led to pedogenic calcretes and carbonates being specifically targeted as a preferred sample medium for Au in southern Australia. Many discoveries have been made in areas having residual soil or shallow transported overburden in the Yilgarn and Gawler Cratons (Lintern 2002). Such discoveries illustrate that carbonate sampling is effective even where developed in transported overburden, if this is no greater than 5–10 m thick. Although the limited dispersion suggests that calcrete sampling is most suited to local exploration and target definition, it has been used for regional surveys in the Gawler Craton with sample intervals as great as 1.6 km. However, concentrations are very low – with anomaly thresholds of 3 ppb Au or less (Lintern 2002). 13.7.2
Calcrete sampling procedures
Base metals
For exploration for base metals in terrains where the carbonate horizon may form a pH barrier and potentially cause the precipitation of chemically dispersed metals, the carbonates and the underlying material – whether residual or transported – should be sampled by drilling. Ideally, all samples should be analysed; but, failing that, those across the lower contact should be selected for analysis, preferably by both total and partial extraction methods.
Gold
The carbonate horizon should be preferentially targeted for Au exploration, because adjacent non-calcareous horizons will be essentially devoid of gold. Sampling procedures vary according to the type and depth of carbonate, and the presence of other material (Lintern 2002). Indurated laminar calcretes – whether outcropping or buried – can be sampled in small pits and breaking with a crow-bar or hammer, or by drilling using a robust auger. Fine material, such as clay and sand, can be removed by sieving. Where silcrete is present, the calcrete horizon immediately above is commonly the site of highest Au abundance. Calcareous soils and powdery calcretes are best sampled by power auger – collecting the whole carbonate horizon (commonly, in the top 1.5 m), excluding any non-calcareous topsoil. Accurate identification of the calcrete is essential at the time of sampling; testing for effervescence with dilute HCl is generally adequate, although more dolomitic samples can be slow to react. Because Au is dispersed throughout the carbonate, a large sample is not essential, although conventionally a 30 g aliquot of a 1 kg sample is analysed. Samples should be crushed and analysed for Ca, Fe, Au and pathfinders such as As, Sb and W. An acid digestion (such as with aqua regia) is generally adequate, if there is sufficient acid to dissolve the carbonate fully. Cyanide digestion gives good results, but for Au only. A detection limit of 0.1 ppb is recommended where abundance is low. The Ca data are useful to confirm that the sample is appropriate (for example, lithic fragments with carbonate coatings may be mistaken for carbonate nodules) although gypsum may give misleadingly high Ca abundances. However, normalisation of the data to the Ca content has not been shown to be effective. The Fe analyses will assist in interpretation of data for elements such as As, which are known to be scavenged by Fe oxides (Table 5.4; Appendix 2). If the total depth of transported overburden is less than 10 m, it is possible that preferential concentration of Au may give a surface expression to concealed mineralisation. However, the contrast may be diminished. Caution should be exercised before interpreting a nil response as negative (see also transported overburden: Section 13.11).
Regolith sampling for geochemical exploration
13.8
VEGETATION
The use of vegetation in exploration programs involves either:
s
s
'EOBOTANY – in which the distributions and/or morphological changes of specific indicator plants or plant associations are related to their preferential growth in soils having anomalous metal contents, nutrient status or other properties. "IOGEOCHEMISTRY – in which plant tissues or organs are chemically analysed to detect biological concentrations of elements that may reflect mineralisation.
These and associated topics are discussed in detail in Chapters 7 and 8 and by Brooks et al. (1995) and Dunn (2007). Geobotany
Climate – especially temperature and rainfall – is the principal controls on plant distributions and assemblages. Bedrock, regolith geology and land use mostly have a subsidiary influence at regional scales, but can be important at district to local scales. Consequently, the principle of mapping plant assemblages as a surrogate for underlying bedrock has long been applied in photogeology, and similarly contributes to the textural patterns used for regolith landform mapping by aerial photography, satellite imagery and hyperspectral surveys (in which there is commonly a vegetational component) (see Chapter 11). Thus, in a general sense, geobotany is widely used as a geological tool. However, it has been little used directly for exploration in Australia, although it has been recognised that a number of plants may indicate mineralisation:
s
s
s
Polycarpaea glabra, P. spirostylis and Tephrosia spp. indicate Cu mineralisation at Dugald River and other sites in Queensland (Nicolls et al. 1965; Brooks and Radford 1978) Gomphrena canescens, Polycarpaea spp. and Tephrosia spp. were said to indicate carbonate hosted Pb-Zn mineralisation at Bulman, Northern Territory (NT), at its discovery in 1952 (Cole et al. 1968) Solanum linearifolium grows preferentially near Pb-Zn mineralisation at Woodlawn, New South Wales (NSW) (Ryall and Nicholas 1979)
s
Hybanthus floribundus indicates Ni-rich soils on ultramafic bedrock in the Kalgoorlie region, WA (Severne 1974).
These plants are all tolerant to metal-rich soils; Hybanthus floribundus is also a Ni hyper-accumulator. Each of the above species is part of broader associations in which either the plant assemblages may differ from those on surrounding soils with ‘normal’ metal contents – or specific species may be stunted, as illustrated by the detailed studies by Nicolls et al. (1965) and Cole et al. (1968). These effects are well developed in areas of outcrop and residual soil, but have little direct application where there is a thick, leached regolith or transported overburden. Biogeochemistry
Because of the cover of vegetation across nearly all climatic, geological and regolith–landform environments in Australia, there is a wide choice of plants for biogeochemical sampling. However, this choice is constrained by the need for the plants to have appropriate root qualities and a sufficiently extensive distribution at the scale of the survey. Plants have evolved to mitigate the stresses caused by the dry climates and nutrient-poor soils that prevail over much of Australia. One adaptation is the dimorphic root systems with shallow lateral and deep tap (or sinker) roots. The latter may reach 10 to >40 m and are able to access water and nutrients deep in the regolith, especially during dry periods and drought (Pate et al. 1998; Aspandiar et al. 2004). However, many plants are dormant at these times, so nutrient demand and transfer rates are potentially low. Some plants are phreatophytes, which are able to access water from the saturated zone or from the capillary fringe. The degree to which tap roots can tolerate highly saline groundwater is uncertain, and is a potential restriction on this mechanism of element transfer in much of semi-arid southern Australia and some environments elsewhere. Vegetation takes up a wide range of elements – from groundwater or mineral surfaces using some elements in the metabolic processes, but storing or rejecting others. Biologically essential elements (Ca, K, Mg, Na, S, Cu, Fe, Mo, Se and Zn,) are selectively taken up by vegetation (see Chapter 8). The elements
355
356
Regolith Science
may be disproportionately concentrated in plant tissue, even where their abundances are very low in regolith and rocks (for example, high Cu contents in ashed Eremophila at Panglo: Lintern et al. 1997). Nonessential elements, including those that are potentially toxic (such as Ag, As, Au, Cr, Ni, Pb, Sb and U), are also taken up and may reflect more closely the composition of the substrate. These elements are subsequently released to the soil by decay after leaf-fall or the death of the plant – contributing initially to the litter horizon and ultimately to underlying soil horizons. Products derived from vegetation, such as litter and mull (decaying plant material) may have element concentrations greater than the original plant due to: (i) nutrient withdrawal during the senescence period immediately prior to leaf fall; or (ii) preferential leaching of more mobile constituents, including Ca, Cl, Mg and Na. Such mechanisms may account for high concentrations of base metals, including Sn, in litter over the buried Jaguar base metal deposit, WA (Anand and Cornelius 2004). Although most will be recycled, there is potential for metals to accumulate gradually in surface soil horizons over time. If the soil is developed on transported overburden, this metal could be derived from depth via taproots and give surface expression in soil to buried mineralisation. Uptake by vegetation may transform the chemical nature of elements in the regolith by, for example, forming organo-metallic complexes that – on release to the soil after plant decay – may impart a different mobility compared with the same metal bound in a mineral. These transformations may, in turn, influence partial extraction analyses. Cohen et al. (1998) found only a poor correlation between vegetation and partial extraction soil anomalies (cold HCl, enzyme leach) in the Cobar Region, NSW. This implied that the contributed metals had been flushed from the upper soil horizons and they concluded that cycling by plants is not sufficiently active to cause significant surface enrichment. This conclusion needs further testing. A selectively sampled and/or analysed soil horizon would have advantages over vegetation in terms of increased availability, reduced variability and cumulative response: that is, contributions by all parts of all plants over many generations, rather than by one organ of an individual plant at a specific time.
13.8.1 Vegetation as a sample medium Biogeochemistry – using various portions of trees has been widely used in Canada to successfully identify mineralised zones in areas of transported cover (for example, Cohen et al. 1987; Dunn 1989, 2007). The methodology has even been adapted to sample the tops of Douglas-fir trees by helicopter (Dunn and Scagel 1989). However, in Australia, biogeochemistry has been tested in a variety of environments and scales with mixed success. Cohen et al. (1999) compared vegetation and stream sediment surveys over a large area of north-eastern NSW. Overall, vegetation was found to depend more on hydromorphic dispersion and to give broader dispersion trains from mineralisation. Both media reflected regional variations, but there was little correlation on a site-by-site basis. Although known mineralisation and new Au targets were indicated, these were generally in one sample medium only. At a district to prospect scale, there are a number of examples where metal contents of various plant species and plant organs have been shown to reflect mineralisation beneath shallow residual soils, for base metals (Nicolls et al. 1965; Cole et al. 1968) and for Au (Lintern et al. 1997; Arne et al. 1999, 2001). However, where the residual regolith cover is deeper and more strongly leached, plants may give poor responses (Panglo: Lintern et al. 1997). In general, sampling of residual soils and other shallow regolith materials is preferred, because it is easier and gives more consistent results. The greatest potential for biogeochemistry lies in areas of transported overburden, where tap roots and – if the cover is shallow – some lateral roots may access weathered bedrock and deep groundwaters. There have been promising results at a number of such sites in arid to semi-arid regions of Australia:
s
s
s
Gold and base metal mineralisation, Cobar NSW; shallow colluvium and skeletal residual soil. Multielement anomalies in needles of white cypress pine (Callitris columellaris) (Cohen et al. 1998). Flying Doctor Ag-Pb-Zn prospect, Broken Hill, NSW (Hill et al. 2005); <5 m mainly coarse colluvium. Response in Maireana leaves, and Acacia aneura and A. victoriae phyllodes. Teilta, NSW; variable regolith cover in major drainage channels and alluvial outwash plains.
Regolith sampling for geochemical exploration
s
s
s
s
s
s
s
Copper, Zn, As and Au anomalies in Eucalyptus camaldulensis leaves, with high seasonal variability (Hulme and Hill 2004). Wyoming Au deposit, NSW; 20–40 m transported overburden. Anomalous Au and possibly As in Eucalyptus microcarpa and Callitris columellaris (Roach 2004). Curnamona Province, South Australia (SA); shallow calcareous soils on colluvium. Copper and patchy Au response in saltbush (Atriplex vesicaria), White Dam prospect (Brown and Hill 2005). Stuart Shelf, SA; 20–30 m sand and cover sequences. Copper-Fe-Zn anomalies in ashed Acacia twigs over porous non-calcareous cover rocks; no response through calcareous shales (Rattigan et al. 1977). Barns Au prospect, Gawler Craton, SA; 1–8 m of aeolian sand. Plant organs (Melaleuca and Eucalyptus spp.), litter and calcrete have anomalous Au contents (maxima 3.6 ppb in Melaleuca bark, 9 ppb in carbonate rhizomorphs) in dunes that are no more than 27,000 years old (Lintern 2007). Titania Au prospect, Tanami, Northern Territory (NT); 15 m of paleodrainage sediments. Spinifex hummock grass (Triodia pungens) delineated concealed mineralisation, with anomalous concentrations of Au, As and Sb in leaves (Reid et al. 2007). Jaguar base metal deposit, Yilgarn Craton, WA; 10–20 m alluvium and colluvium. Anomalous base metal and Sn contents in leaf litter and various organs of Acacia aneura over concealed, blind mineralisation (Anand and Cornelius 2004; Anand et al. 2007). Litter was also found to be anomalous over the Gossan Hill base metal deposits at Golden Grove, where gossans are exposed at surface. Gold deposits, northern Yilgarn Craton, WA; shallow colluvium and other sediments. Litter and, where sampled, various organs of Acacia aneura give Au and multi-element anomalies over concealed Au mineralisation (Anand et al. 2007).
In comparison, in the southern Yilgarn Craton, vegetation and mull sampling over several Au deposits gave equivocal results. Wherever there was an anomaly in these media, a similar or better response
was found in pedogenic carbonate, whether in erosional regimes (for example, Bounty; Lintern et al. 1997) or areas of shallow (<10 m) transported overburden (Butt et al. 1997). At Zuleika, the response to concealed paleochannel Au mineralisation at 20 m depth gave a broad response in Maireana spp. on one section but only a smaller single-point anomaly over higher grade mineralisation on another – Eucalyptus gave no response (Lintern et al. 1997). At Apollo, there were no anomalies in soil or Eucalyptus leaves above buried (60 m) mineralisation, but Maireana again gave a strong response, although contamination is possible (Butt et al. 1997); the adjacent Argo deposit gave no response in soil, Eucalyptus leaves, Eremophila plants or mull. It is possible that the advent of analytical techniques with lower detection limits may yield responses similar to those at Barns, Gawler Craton (above) but, at present, biogeochemistry appears to have little application for Au exploration in the southern Yilgarn Craton. 13.8.2 Vegetation sampling procedures Trace metal contents in vegetation can vary according to the genus, species, organ selected (that is, whole plant, leaves, twigs, bark, wood, roots, litter or mull), the age (for example, new growth or year-old twigs) and season. Accordingly, the success of biogeochemical surveys depends on the selection of suitable samples for analysis and assiduous attention to maintaining these selections to achieve a consistent sample medium. This may present a challenge because of the variable distribution within an area, similarities between species of the same genus and practical problems, such a collecting leaves from tall trees. In general, it is desirable to composite at least three or four samples from each tree or shrub – collected using gloved hands and clean, preferably plastic-coated, instruments. Contamination by dust is a potential problem in arid environments and washing is commonly essential. The dried samples (70°C) should then be cut, ground or milled and can be analysed directly (for example, by neutron activation, or acid digestion and ICP-MS) or after ashing. The latter offers considerable concentration, but there is potential for loss by volatilisation. See Anand et al. (2007) and Hill (2002) for more details on methodology.
357
358
Regolith Science
13.9
INDURATED MATERIALS
Cementation by Fe, Si and Ca is a common feature in near surface materials to form ferricretes, silcretes, hardpans and calcretes (Section 13.7) Ferricretes are indurated materials formed by the in situ cementation of transported regolith by Fe oxides: mainly goethite and/or hematite. The matrix may consist of a wide range of colluvial and alluvial sediments. Where these are transported lateritic gravels, the ferricrete may be difficult to differentiate from residual lateritic duricrust. A toposequence in lateritic terrain may show a continuum from lateritic residuum on the hillcrest through to a ferricrete in the valley floor. There is no genetic relationship between the ferricrete and the underlying weathered or fresh rock (Anand and Paine 2002). Silcretes are highly indurated and generally of low permeability; they commonly have a conchoidal fracture with a vitreous lustre. They appear to represent the complete or near-complete silicification of a precursor regolith by the infilling of available voids. The induration is such that fractures propagate across enclosed quartz grains, rather than around them. Silcretes may be classified broadly into pedogenic or groundwater types. Most are dense and massive, but some may be cellular, with boxwork fabrics. The fabric, mineralogy and composition of silcrete may reflect those of the parent (regolith) material and, hence, if residual, the underlying lithology. Thus, most silcretes over granites and sandstones have a floating or terrazzo fabric, with residual quartz grains separated by a matrix of cryptocrystalline quartz, anatase and zircon, reflected by enrichment in Ti and Zr (see also Section 6.3.1). Silcretes with lithic fabrics (for example, on dunites) are silicified saprolites with initial constituents diluted or replaced by silica. Red-brown hardpan is a weakly silica-cemented surficial unit consisting of a variety of transported or residual host materials including colluvium, nodular gravels and brecciated saprolite and found in central WA and parts of NT and SA. It is cemented by a porous matrix of silica (commonly hyalite), alumino-silicates and Fe oxides and fractures readily. Manganese oxide precipitates are commonly present on partings.
13.9.1 Indurated materials as sample media Ferricretes, including authigenic nodules and pisoliths formed in sediments, will commonly not show a geochemical signature that relates to the underlying or nearby bedrock – hence their use as a sampling media is restricted. However, some – for example, transported and recemented lateritic gravel – may have an application in widespread reconnaissance sampling in the absence of residual lateritic gravel. It is also possible that ferricretes in seeps and drainages, which are dominated by reprecipitated Fe (and Mn) oxides, might concentrate metals dispersed hydromorphically from concealed mineralisation. However, their use in exploration would require a good understanding of regional regolith–landscape evolution, paleotopography, and past and present drainage to determine probable directions of dispersion, and to account for false anomalies generated by the scavenging of metals leached from country rock. Silcrete is rarely deliberately selected as a sample medium, although it may form a component of other media. Where the silcrete has formed by silica flooding of residual regolith materials (Section 2.8.3) – whether these are soils or saprolite – it will retain original mineral and geochemical characteristics of the precursor, including evidence for mineralisation:
s
s
Over granite and some sandstones, Zr and Ti have been concentrated before, and possibly during, silicification (Butt 1985) to give a recognisable geochemical signature. Silcretes on dunitic rocks have Ni and Cr contents that indicate the parent lithologies: for example, where they outcrop as caprocks. However, this silicification results in dilution; in silicified saprolite at Mt. Keith (Butt and Brand 2005), the Ni content is inversely proportional to the silica content of the silicified saprolite (Butt and Sheppy 1975). This silicification is rarely complete on a bulk scale – hence the siliceous zones in oxide Ni laterite deposits, which, overall, have sub-economic Ni contents, can be upgraded by screening to remove the hard, coarse, low-grade siliceous component, and retaining the finer, higher-grade Fe oxide fraction.
Regolith sampling for geochemical exploration
s
Silcretes associated with calcrete U deposits are enriched in U (to >100 ppm), may fluoresce and give a radiometric response in outcrop (Butt et al. 1977).
Because of its widespread occurrence as a nearsurface unit in the Gawler Craton, SA, silcrete has been tested as a possible sample medium for Au exploration. Samples of silcrete lag from the Challenger Au deposit show a prominent Au anomaly (maxima 50–105 ppb) over outcropping mineralisation, but little or no response where there is a cover of transported overburden. This compares somewhat unfavourably with calcrete, which has a more consistent Au content (again at 50–100 ppb), but also has peak concentrations >2000 ppb over outcropping mineralisation and shows a response to one zone of shallowly buried mineralisation. Whereas Au appears to concentrate specifically in calcrete and is evenly distributed, in silcrete Au has a highly variable distribution, with some individual sub-samples having extreme Au concentrations. Silicification is probably retaining (and diluting) particulate Au in the precursor matrix. Lintern and Sheard (1998) suggest that silcrete has potential as a sample medium in the absence of calcrete, but will only be effective if developed in residual materials. Red-brown hardpan cementing colluvium at Broad Arrow, WA, contains Au anomalies (50 ppb versus a background of 10 ppb). These are thought to be due to both mechanical and hydromorphic dispersion (Mahizhnan 2004) and could be used as a useful sampling medium for Au exploration. Marshall and Goldsworthy (2006) reported the use of base of hardpan sampling for gold and base metal exploration in the Murchison district in the Yilgarn Craton. This was considered a cost-effective approach to the evaluation of broad geological and geophysical targets on the margins of colluvial/alluvial plains that flank ‘islands’ of greenstone outcrop. The basal 0.3–0.5 m of the hardpan were sampled by vacuum drilling on 50 × 200 m grids and analysed for Au, base and pathfinder elements after an aqua regia digest. Gold and base metal anomalies were considered to be dominantly hydromorphic, but were sharply defined and showed little or no lateral displacement. In this region,
the Au threshold is 20–25 ppb, with peak anomalies of 75–125 ppb in infill sampling providing direct targets for deeper drilling. This approach is similar to that of interface sampling (see Section 13.10), but will probably only be effective where the base of hardpan corresponds with the unconformity between colluvium/ alluvium or is in residuum. 13.9.2 Indurated material sampling procedure Ferricrete should be sampled in the same manner as lateritic gravels and duricrusts. If ferricretes are included within regional laterite surveys, it is imperative to be able to separate them from lateritic residuum during data processing and interpretation. It is possible that the detrital, ferruginous gravel and the matrix might distinguish between distal and local geochemical signatures, respectively; hence, ideally, these components should be analysed separately. Silcrete lag can be collected from the surface: preferably compositing numerous small fragments over a 10 m2 area. As with ferruginous lag (above), the use of coarse fragments (10–50 mm) requires large (much larger than 1 kg) samples to minimise any potential nugget effect. Outcropping and sub-cropping silcrete, including silicified saprolite, may be sampled from drill cuttings, again using large, well-mixed samples. Total analytical methods are necessary to ensure detection of Au particles encapsulated by silica. Red-brown hardpan is best sampled by shallow drilling, targeting the basal 0.3–0.5 m, where the hardpan is developed in residuum, or in thin transported overburden, down to, or below, the unconformity. The whole sample should be crushed and pulverised, and preferably analysed using a total digestion. Selective leach analysis might be appropriate to emphasise active hydromorphic dispersion.
13.10
INTERFACE SAMPLING
‘Interface’ sampling refers to sampling across an unconformity – generally that between weathered basement and cover. In most exploration geochemical sampling, a single, specific medium should be taken to ensure data are comparable. Interface sampling is unusual, in that it deliberately uses a mixed medium;
359
360
Regolith Science
that is, transported overburden above and weathered basement below. 13.10.1 Use of Interface material as a sample medium Interface sampling is useful in areas of extensive transported overburden, especially where the buried residual profile has undergone significant truncation. Where the pre-existing profile has been little eroded and lateritic residuum is widely preserved beneath the cover, this should be targeted instead (Smith et al. 2000). Both methods require drilling through the cover. Interface sampling is most successful where the sedimentary environment was of low energy, dominated by soil creep, sheet-wash or, at most, ill-defined, slow-flowing streams – depositing sediment as gently inclined fans. Such situations favour the preservation of detectable amounts of mechanically dispersed mineralised detritus, which may be supplemented by later, hydromorphic dispersion. In contrast, in high-energy fluvial environments, anomalous detrital material would be scoured away and diluted, perhaps to form localised dispersion trains higher in the sedimentary column, which are distant from the sub-crop of the mineralisation. Any proximal geochemical response would depend on post-depositional hydromorphic dispersion at the unconformity. 13.10.2 Interface sampling procedure The base of the cover can be a simple, sharp, erosive unconformity (Figure 13.5a) or a complex mixture of saprolite and colluvium, a metre or more thick (Figure 13.5b), possibly including a paleosol, which has later been buried (Robertson et al. 1999). In a clay-rich environment, a mixed zone as much as 10–15 m thick may have formed by post-depositional churning (for example, Mt. Keith; Butt and Brand 2005). Ideally, the position of the unconformity should be logged accurately. This is generally only possible from drill cuttings if there is a useful contrast (colour or texture), or the abrupt appearance of minerals or other components typical of either the sediments (such as rounded grains or abraded pisoliths) or the bedrock (such as mica). In the future, hyperspectral analysis of mineralogical characteristics, such as kaolinite crystallinity, may become used routinely to assist conventional logging (see also Section 4.8).
In low-energy depositional environments, if it is possible to locate the unconformity to the nearest metre, the metre interval crossing the unconformity that includes both basement and cover materials is an ideal choice (Figure 13.6a). If a mixed sample cannot be identified, the metre intervals above and below the unconformity can be collected (Figure 13.6b) and composited to ensure the unconformity is included. Where the contact cannot be clearly defined, collecting and analysing several separate samples across the contact is preferred because dispersion may be less well confined. Here, several individual samples are recommended, rather than compositing samples across the more than two metre contact zone, as this would dilute the response. In higher energy depositional environments with steeper depositional gradients, anomalous detritus may occur higher in the stratigraphic column and not necessarily at the base (for example, Pajingo; Robertson 2003). This necessitates sampling throughout the cover and using either a mean or a maximum value. In most circumstances, an aliquot of the whole sample is crushed and pulverised for ‘total’ analysis. However, in specific settings, particular components, such as ferruginous gravel or mottles, may be separated. Selective leach analysis might be appropriate to emphasise active hydromorphic dispersion.
13.11
TRANSPORTED OVERBURDEN
Transported overburden generally refers to exotic or redistributed material of continental origin that blankets weathered and fresh bedrock. In some cases, it is partly cemented by Fe oxides, silica or carbonates. The term usually excludes dominantly marine, lithified sequences in sedimentary basins, which themselves may host mineral deposits, and igneous units such as flood basalts that overlie Precambrian and Paleozoic rocks. Transported overburden commonly refers to:
s s s
aeolian clay (loess) and sand, in semi-arid and arid environments sheet-wash deposits, colluvium and alluvium Evaporites in playas (such as halite and gypsum) and paleodrainage channels (such as valley or groundwater calcrete)
Regolith sampling for geochemical exploration
Simple, erosive contact
a
Complex contact - paleosol
b Colluvium a
b
50% 50%
Colluvium
Basement
Saprolite
Figure 13.5: The base of the cover may form (a) a simple erosive contact or (b) a more complex mixture of cover and basement as a paleosol.
s s s s
Piedmont fan and outwash deposits of cobbles, gravel, sand and clay shedding from dissected plateaux, hills and mountains talus and landslip deposits in hills and mountains sediments in paleodrainage channels, including sand, alluvial and lacustrine clays, lignite and estuarine sediments glacial deposits.
The relative timing of continental sedimentation and the principal weathering events have considerable potential significance for exploration. Older sediments are more likely to contain chemical dispersion from concealed mineralisation. Those deposited before, during or just after the main phases of deep weathering in the Mesozoic and early Tertiary will have been subjected to more post-depositional alteration (and therefore dispersion) than younger
Figure 13.6: Interface sampling. (a) Where a mixed sample of both sides of the unconformity is available, the appropriate metre interval is selected. (b). Where a mixed sample is not apparent, a 50/50 mix of metre intervals above and below the interface is selected to ensure the interface is included.
sediments. The principal types of transported overburden and their distributions in a landscape are illustrated schematically in Figure 13.7. 13.11.1 Transported overburden as a sampling medium Although minor colluvial transport during deep weathering contributes to the effectiveness of soil, lag and laterite sampling by broadening the dispersion haloes, transported overburden is rarely an effective sample medium. In regions where bedrock and mineralisation are extensively weathered and leached,
Local
Sandy silty clay
Age Interpreted Late Tertiary to Quaternary
Massive, structureless, red clays with black hematite-maghemite-rich fine (2-10 mm) ferruginous granules
Interpreted Late Tertiary
Sand dunes, gypsum dunes, lunettes
Distal
Playa
Gravelly sandy clay
Clays Sand and gravel
Paleochannel sequence
Boulder clays, grits and sandstone
Late Eocene Middle Eocene Permian
Saprolite Bedrock
Figure 13.7: Schematic diagram illustrating the principal types of transported overburden and their distributions in the landscape.
361
362
Regolith Science
further dispersion is generally limited to mechanical smearing at the bedrock/overburden interface, as discussed above. Dispersion to the surface is generally precluded in arid regions and it is no coincidence that most mineral discoveries attributed to surface-sampling techniques in such terrain have been made on hills and low rises, with outcrop and residual soil – above the overburden-covered plains. Despite extensive testing of different sample media and using a variety of total and partial or selective analytical techniques, there are few unequivocal examples of surface geochemical anomalies being directly related to mineralisation concealed by transported overburden. The principal exceptions are some Au deposits in the southern Yilgarn Craton and the Gawler Craton, where hydromorphic Au anomalies occur in calcareous soils and calcrete developed in shallow sediments. Even here, however, carbonate sampling can only ‘see through’ a maximum of 5–10 m of sediment (Butt et al. 1997; Lintern 2002). In comparison, in non-calcareous soils, deposits may be blind where the sediment is only 2 m or less thick (for example, Fender; Butt and Dries 2005), probably depending on bioturbation as the principal near-surface dispersion mechanism. However, Scott and van Riel (1999) found that the coarse component of partially transported soils material reflected Au mineralisation in the saprolite at least 5 m below at Goornong South, Victoria. Deeper bioturbation is possible, but cannot be relied upon. In this context, it can be noted that a surface anomaly is reported from partial analyses of soils at Osborne, Queensland (Rutherford et al. 2005), which is concealed beneath 30–60 m of Mesozoic marine sediments. The metal enrichment of several sub-horizontal zones in these sediments may have occurred during marine depositional–diagenetic stages (Lawrance, 1999), or later sub-aerial weathering, with near-vertical fractures as fluid pathways. The soil anomaly possibly reflects one such enrichment zone. 13.11.2 Transported overburden sampling procedures Generally, overburden sampling is avoided and systematic grid drilling to residuum is preferred. Exceptions are where the overburden is very thin (less than 2 m) and, in gold exploration, where pedogenic carbonates occur and overburden is less than 10 m thick.
However, even in such environments there may be uncertainty because the thickness of overburden may increase without any obvious indication at surface. In both instances, shallow power augering is generally the most appropriate procedure. In environments where metal enrichment in deeper sediments is anticipated, present or past redox fronts may be specifically targeted during exploration drilling. Such fronts are associated with organic matter and/ or sulfides or Fe oxides. At some sites, the latter may themselves have been derived from oxidised sulfides; redox fronts represent trap sites for metal enrichment during sedimentation, diagenesis or weathering (for example, Osborne; Lawrance 1999). Where systematic drilling is employed, and especially if it is intended to collect buried lateritic residuum or interface samples, it is essential to be able to distinguish between transported and residual regolith. The presence of disordered kaolinite – as detected by infra-red reflectance spectroscopy (see Section 4.8) – may be effective in distinguishing between transported and residual units (Pontual and Merry 1996). However, the relationship is empirical and is not diagnostic everywhere. The following are characteristics of transported overburden: 1. polymictic 2. fractured ferruginous fragments 3. maghemite at depth – indicates the material was once at the land surface 4. no cutans on pisoliths – but pisoliths developed in situ in sediment have very well-defined and commonly multiple cutans 5. weatherable minerals in the near-surface, but absent deeper in the regolith 6. absence of lithic fabrics – although detrital lithic fragments, such as weathered boulders in Permian glacial sediments, may be misleading 7. lignite and organic matter 8. rounded quartz grains and gravels 9. change in resistate mineralogy – for example, abrupt downward appearance of mica or talc may indicate the basal unconformity. Paleochannel sequences commonly have massive, structureless clay that is mottled in the upper part, overlying a quartz-rich gravel and sand unit. The
Regolith sampling for geochemical exploration
latter will be approximately horizontal (distinguishing it from disaggregated quartz veining), have some rounded quartz, attenuate laterally (across channel) and be below the deepest section of the clays. Carbonaceous material may occur in reducing environments. Megamottles (Section 6.4; Figure 6.10) may also form in paleochannel infill material.
13.12
STREAM SEDIMENTS
Stream sediments are unconsolidated materials that are being mechanically transported in a confined, connected drainage channel by saltation, traction or suspension in flowing water – or that have been chemically precipitated from the stream water. Stream sediments commonly have both components. The detritus may range from boulders to clay in size, and include both mineral and organic matter. Active stream sediments are those being transported or reworked during stream flow under the present climatic environment – bearing in mind that many streams in arid regions only flow seasonally or intermittently after rain. Bank or over-bank sediments are those deposited during the waning stages of flood events. Connected drainage is commonly absent from many semi-arid regions of low relief. Following heavy rain, surface water is shed by sheet flow or sheet-wash – a more or less continuous cover of flowing water, either unconfined or in broad, ill-defined channels. Sheet-wash can transport clay, sand and gravel across slopes of less than 1°. Water flow is discontinuous forming networks of braided rills and wash channels that change from season to season. In places, the colluvial sediments deposited by sheet-wash may be used as a substitute for stream sediments, but their provenance is generally less certain. 13.12.1 Stream sediments as a sample medium Stream sediment sampling is a well-established procedure for mineral exploration at a wide variety of scales, from broad, province-scale reconnaissance surveys through to local surveys that may indicate outcropping mineralisation. Thus, sample densities may vary from one sample/several km2 to several samples/km2. The general principles are well described in Hale and Plant (1994), with specific application to arid terrains
by Mazzucchelli (1994). At a regional level, surveys may be (i) reconnaissance surveys, for which a representative proportion of streams of a specified (usually low) order is sampled, to identify broad mineral provinces, and (ii) regional surveys, in which all streams of a specified order are sampled, to obtain a ‘complete’ coverage. Low-density sampling (for example, 1/900 km2) of active or over-bank sediments from higher order drainages provide geochemical backgrounds, which have more application in environmental studies than in area selection or targeting for exploration (Lech and de Caritat 2007). There are numerous case histories containing stream sediment data. These are from a wide range of climatic and physiographic environments. They include examples from the temperate rainforests of western Tasmania (for example, Que River Pb-Zn deposit; Skey and Young 1980) to the arid northern Flinders Ranges, SA (Beltana willemite deposit; Moeskops and White 1980). The data show that, in more humid regions, hydromorphic dispersion may contribute significantly to the anomaly and result in long drainage trains. At Que River, hydromorphically dispersed Zn extends further down drainage (peak >1000 m) than clastically dispersed Pb (peak <400 m) (Skey and Young 1980). Dispersion trains of similar or greater length occur in other humid environments, such as Halls Peak Cu-Pb-Zn-Ag deposit, NSW (Ashley and Wolfenden 2005). In more arid regions, clastic dispersion dominates and drainage trains are generally quite short and sampling at close intervals (2/km2) – generally from streams 2–3 km in length must be undertaken (for example, Beltana willemite deposit; Moeskops and White 1980). Contamination due to mine tailings can be a major problem with stream sediment surveys (for example, mine contamination has contributed to dispersion of over 40 km at Halls Peak; Ashley and Wolfenden 2005). However, most contamination is relatively restricted and can commonly be avoided by careful sample site selection and data interpretation. Thus, the Khan’s Creek gossan (Palethorpe 1980) was found despite the general contamination in the Halls Peak area. However, contamination by past exploration activities is both more widespread and more difficult to detect. Sediments from streams in explored areas may contain drill cuttings; these can occur in all size
363
364
Regolith Science
fractions and may include fresh, sulfide-bearing rock fragments as well as weathered material. 13.12.2 Stream sediment sampling procedures Stream sediment surveys are commonly based on sampling first-, second- or third-order tributary streams; higher order streams are commonly too large – causing considerable dilution. The sample density varies according to the application – ranging from 1/15–20 km2 for regional reconnaissance to 3–5/km 2 for detailed follow-up. Active sediments are collected from close to the centre of the channel, preferably compositing several sub-samples to ensure that the sample is representative. For Au exploration, highenergy sand and cobble bars are recommended because, in most stream environments, Au is transported as a heavy mineral and accumulates in the fine fraction (<100 µm) at these sites (Fletcher and Loh 1996; Melo and Fletcher 1999). Bank and over-bank samples may be used in areas where mine contamination is possible because these may represent material deposited prior to the commencement of mining. Over-bank (floodplain) sediments are deposited as fine sediment trapped from receding floodwaters. For their low density, geochemical baseline study of the Riverina region of SE Australia, Lech and de Caritat (2007) collected samples from the top (0–10 cm) and bottom 10 cm (usually between 60–90 cm depth) of the sediment – choosing the <180 µm fraction for analysis. Sample sites were selected from the lowest point in each catchment using digital elevation and drainage data in a GIS; a similar procedure could be followed in higher density, exploration, surveys. The whole sample may be used for analysis, but, more commonly, one or more specific size fractions are selected. Samples may also be panned to separate heavy minerals. The most appropriate size fraction is best selected following an orientation survey. Judicious use of data from appropriate case histories (for example, Butt et al. 2005a), may obviate the need for a detailed orientation in some areas. Dilution by aeolian material and the derivation of sediments from strongly weathered, leached rocks are specific problems that can affect the application of stream sediment geochemistry in arid terrains, but these can be combated
by selection of appropriate sample fractions, analytical suites and data thresholds. Commonly, the coarse (>500 µm) or fine (<75 µm) fractions give the best results, because these concentrate Fe oxide rich fragments, and fine clays and oxides, respectively (see soils, Section 13.2), and exclude much of the aeolian material. The fine fraction may also include widely dispersed heavy minerals, including Au (as noted above). The intermediate fraction may be dominated by aeolian quartz in much of Australia. However, in South-East Australia, aeolian material is commonly more than 70 µm in diameter so, in that region, coarser fractions are a better sample medium (Dickson and Scott 1998; Tate et al. 2007). Partial extraction analysis can be used to distinguish hydromorphically dispersed metals, and has particular application in more humid regions. Bulk leach extractable gold (BLEG) analysis of large, unsieved samples has been applied in an attempt to overcome the ‘nugget effect’ in sampling and analysis for Au, and to provide a sensitive procedure that can detect widespread dispersion haloes. However, the technique seems to offer no advantages over the <75 µm fraction in much of Australia (for example, Mazzucchelli 1994).
13.13
GROUNDWATER
Groundwater refers to all sub-surface water (Chapter 10). Water in the unsaturated zone above the water table is referred to as vadose, whereas that in the saturated zone below the water table is phreatic. Hydrogeochemical exploration uses phreatic groundwater – generally from the uppermost, unconfined aquifer. However, it is not uncommon for drill holes to intersect confined aquifers deep in the regolith, especially in depositional landforms with impermeable sediments. Hydrogeochemical surveys aim to detect elements or their isotopes being actively dispersed from weathering mineral deposits, or from primary dispersion haloes associated with the deposits – with potential to reveal blind mineralisation concealed by cover sequences. A water sample inevitably contains suspended solids. The accepted definition of water is a solution that passes through a 0.45 µm filter (Brown et al. 1970; Section 10.5.5), but although smaller filter pore sizes
Regolith sampling for geochemical exploration
are now being used, fine colloidal particles, clays and oxides remain in the filtrate that is analysed. 13.13.1 Groundwater as a sample medium Hydrogeochemical surveys are best suited to elements that are relatively soluble in the near-surface environment, with the dominant dispersion mechanism being by groundwater flow. The pH of natural water may range from about 3.5 to 9, so there may be more than one ionic complex species for each element (for example, for Cu: Cu2+, CuCO30, CuCl+, CuSO40, Cu(OH)+: see also Table 5.2) The dominant species in any given solution is dependent on other solution parameters, especially the concentration of anionic species. Thus, analysis for total element concentrations, rather than ion-specific techniques, are preferred (Mann 1980). Groundwater sampling has also been applied to dissolved gases (for example, helium: Butt and Gole 1986), for which upward dispersion as bubbles is possible. Improved techniques of analysis and data interpretation have made hydrogeochemical surveys more attractive, but there appear to be few examples of such surveys leading directly to mineral discovery. The most significant advances have been increased knowledge (i) at the deposit scale, of the aqueous geochemistry and active dispersion of ore related elements in a variety of different geological, regolith and climatic settings, and (ii) at the district to regional scales, of variations in groundwater composition and solution characteristics and their impacts on the potential application of hydrogeochemical techniques. The greatest emphasis has been placed on Au and U exploration. Numerous studies on the Yilgarn Craton, WA (Gray 2001) have demonstrated there to be at least four major groundwater regimes, based on variations in salinity (<0.02 to >30% TDS), acidity (pH 3–8.5) and oxidation potential (Eh –200 to >800 mV). These variations have a profound effect on the concentrations of many elements:
s s
Al, Li, Y, REE and U – controlled by pH, with high concentrations in acid groundwaters; for example, in the Kalgoorlie region Mn, Co, Ni, Cu and Zn – less correlation with pH and may show lithological variation
s s s
Cr – correlates with ultramafic rocks, irrespective of pH As, Sb, Bi, Mo and W – low in acid groundwaters, but increase when pH>6.5 Au – enhanced concentrations in saline groundwaters, where it forms halide complexes under acid, oxidising conditions.
The implication for exploration is that Au is the best indicator of Au mineralisation in the acidic, oxidising and highly saline waters of the Kalgoorlie region and other areas, especially in the southern Yilgarn Craton, because of the low solubility of other potential pathfinder elements. Carey et al. (2003) also found a broad Au anomaly (threshold 11 ppt, maximum 52 ppt) in groundwaters around the Junction Au deposit, St. Ives, extending west to Lake Lefroy, based on sampling on a 1 × 1 km grid. Because of the strong lithological control, hydrogeochemical surveys for Ni tend to locate ultramafic rocks, rather than those that may contain Ni sulfides. Some preliminary work by Gray and Noble (2006) in the northeast Yilgarn Craton has shown that it may be possible to develop a series of multi-element indices that can distinguish barren ultramafic rocks (Ni and Cr) and proximal, barren, acid-generating Fe sulfide deposits (pH-Eh + Fe + Mn) and (Mo + Ba + Li + Al), from Fe-Ni sulfide-mineralised ultramafic rocks (Ni + Co + W + Pt). Further research is needed to refine this approach. Regional variations in groundwater geochemistry have also contributed to the distribution of calcrete U deposits in WA and NT. These are deposits of carnotite (K 2 (UO2)2V2O8.3H2O), which are associated with waters of low to moderate salinity in valley axes in the northern Yilgarn Craton and similar environments to the north and east, having dominantly granitic bedrock (Butt et al. 1977; Butt 1988). They do not occur in the southern Yilgarn Craton, where the waters are saline and more acidic – possibly due to lower V solubility. A groundwater survey of a 65 000 km2 region of the northern Yilgarn Craton – centred on the Yeelirrie deposit, and based on 575 samples, mainly from stock bores and wells – show a good correlation between hydrogeochemical data, geology and mineralisation (Cameron et al. 1980).
365
366
Regolith Science
Drainages with carnotite mineralisation were characterised by groundwater containing 100 ppb U, compared with background values of less than 5 ppb in catchments over greenstones and Proterozoic sediments, and less than 20 ppb on granitic bedrock. However, within many calcrete drainages, some carnotite deposits are dissolving, whereas others are actively precipitating, so that U content alone cannot be used as a predictive vector. Protocols based on groundwater sampling, determination of pH, K, U and V concentrations and calculation of carnotite solubility indices have been developed for hydrogeochemical exploration to indicate dissolution sites updrainage and the proximity to precipitation sites down-drainage (Mann and Deutscher 1978; Deutscher et al. 1980). In the Alligator Rivers area, NT, high grade U mineralisation in veins at Nabarlek, Ranger and Koongarra is indicated by the concentrations of dissolved U, but similar or higher concentrations occur in waters from barren felsic lithologies. However, ‘normalised Mg’ (NMg = Mg/(Mg +Na + Ca + K) provides a geochemical indicator, increasing towards mineralisation, with high values (NMg >0.8) coinciding with high U concentrations and reflecting the association of Mg-chlorite with U minerals (Giblin 2005; Giblin and Snelling 1983). Koongarra is also the only site where dissolved He concentrations are directly related to U mineralisation (Gole et al. 1986). Elsewhere, He concentrations appear controlled by hydrological conditions, such as residence time and the permeability of the upper aquiclude (Butt and Gole 1986). Hydrogeochemical surveys have their greatest potential where mineralisation is concealed by transported overburden or other cover materials. In the Lachlan Fold Belt of NSW and Victoria, extensions from known mineralisation are concealed by sediments of the Murray Basin and Mesozoic and Cainozoic sediments and volcanic rocks. Groundwater compositions may directly indicate mineralisation of the host sequences using specific elements, element ratios and calculations of theoretically saturated minerals (Giblin 2003). For example:
s
Au, As, possibly with Cu, U, Pb, Rb and Mo – associated with Au mineralisation in the Stawell, Bal-
s s
s
larat and Bendigo regions, with samples distant from known mineralisation indicating potential U and Mo – felsic igneous (probably granitic) rocks that host Mo deposits, near Bendigo normalised Mg – a guide to mafic Cambrian volcanic rocks that host Au mineralisation in Victoria and to alteration associated with U mineralisation in the Alligator Rivers region, NT (see above) modelled minerals – may indicate bedrock sequences associated with gold (muscovite ± talc, antigorite) and base metal (witherite) mineralisation, or calcic (strontianite) or ultramafic (talc, antigorite) rocks.
Water from cover sequences may, however, dilute these responses, and other indicators, such as isotopic ratios, may be necessary to identify the chemical signature of basement rocks. In Victoria, for example, Sr isotopic ratios appear to distinguish Cainozoic basalts from prospective mafic units beneath (Giblin 2003). At Abra, Western Australia, mineralisation can be detected in groundwaters by distinctive Pb, relatively high S and high Sr isotopic ratios (Whitford et al. 1998). A distinctive S isotopic ratio occurs in waters up to 1 km from the blind Menninnie Dam base metal mineralisation, Gawler Craton (Andrew et al. 1998). Similar studies in the Broken Hill region have recognised that some variations in S, O, Pb and Sr isotopic ratios can be attributed to contributions from oxidising Pb-Zn-Ag mineralisation and its host rocks (de Caritat et al. 2005; Kirste et al. 2003). Direct contact with mineralisation and pervasive distribution mean that groundwater is an attractive sample medium for exploration for concealed mineralisation. Nevertheless, despite the promise shown by results such as those discussed above, much work remains to be done before hydrogeochemical surveys can be used with confidence. It is essential to have a good knowledge of the chemistry and hydrology of specific regions to interpret data – hence, considerable preparatory work is required to optimise the approach. 13.13.2 Groundwater sampling procedures Except where it can be collected directly from seeps and springs, groundwater is sampled from exploration drill holes, water bores and wells. Ideally, holes
Regolith sampling for geochemical exploration
should be cased with PVC tubing slotted at regular intervals, samples taken at several depths by pump – with packers above and below to avoid mixing waters from different aquifers – and the pump run for some while to ensure the sample is fresh. Such a procedure is impractical in an exploration context for many reasons. Most hydrogeochemical surveys use standing water from open exploration holes and stock bores and wells (Cameron et al. 1980; Giblin and Snelling 1983; Gray 2001; Carey et al. 2003). Many drill-holes will be uncased, whereas others may be cased, slotted throughout or at some specific depth, or unslotted, so that water only enters from the bottom. It is generally not feasible to pump holes and only sample the recharge because of the small diameter of many holes, the risk of loss of the pump in uncased holes, pump damage by solids, slow pumping/recharge rates and bias towards the most active aquifers. Although the water in an open hole will differ from that in an enclosed aquifer, it is assumed that they will be in dynamic equilibrium; certainly, depth profiles from both uncased and slotted, cased holes show the water to be stratified (for example, in pH, Eh and salinity). Direct contact with the atmosphere affects the upper part of the water column, and distinct degassing trends are observed for He (Butt and Gole 1986), so that it is best to collect samples at fixed depths (at least 5 m) below the water table. Samples are generally collected with a ‘bailer’ – the most effective is one that has a small pump at the base which is turned on at the specified sampling depth to purge and fill the sample chamber above, and then switched off before being retrieved. This ensures that the sample comes from a known depth. Effective sample densities at regional scales may be as great as 1 sample per 170 km 2 using stock bores (Cameron et al. 1980), whereas for tenement scale exploration, specifically drilled holes on 0.5 to 4 km grids may be needed (Carey et al. 2003). After retrieval of the bailer, a procedure such as the following is recommended (Gray 2001). Temperature, pH, Eh and conductivity are determined at the time of sampling. Samples for HCO3– analysis are collected by overfilling a bottle to remove all air before sealing. A further 1.5 L is filtered in the field: 100 ml is acidified for later analysis by ICP-OES and ICP-MS, 50 ml is kept for analysis for Br, Cl and other anions, 1 L is
used for Au analysis by shaking with a 1 g sachet of activated carbon in an acid/saline medium (quantitative to <0.005 µL –1). Acidification is recommended to stabilise the solution and prevent precipitation/ adsorption of dissolved metals. However, it is essential that the water is filtered first, so as not to dissolve particulates. Filtration at 0.45 µm may be preferable over 0.2 µm, because the latter may remove a higher proportion of colloidal particles. Highly saline solutions require specific analytical treatment – commonly with higher detection limits. Data interpretation requires calculation of the solution species of many major and trace elements, and the degrees of mineral saturation in terms of the solubility indices from the compositions of the solutions.
13.14 SAMPLE PREPARATION AND CONTAMINATION Because preparation of rock and mineral samples for chemical analysis almost always involves crushing and grinding, contamination of samples by materials abraded from the surfaces of the equipment is unavoidable. The major components of common crushing and grinding equipment are shown in Table 13.2. Good sample preparation technique thus requires that this contamination be kept to a minimum and that it should not prejudice the required quality of the results of the proposed analysis. It is also necessary to prevent cross contamination of samples by adequate cleaning of equipment between each use. Analysis of various crushing methods and procedures has been carried out at CSIRO by Agus and Hesp (1974) who found that only agate vessels do not introduce strong Table 13.2:
Major components in crushing media.
Agate
SiO2 100%
Alumina
Al2O3 99.9% (some Si, Mg and Na)
Chrome steel
Fe 84%, Cr 13.2%, C 1.9%, Mn 0.5%, Si 0.4%
Manganese steel
Fe 96.7%, Mn 2%, C 1%, Si 0.3%
Tungsten carbide
W 87%, Co 6%, C 5.8%, Ti 0.5%, Ta 0.5%
Zirconia
ZrO2* 96.5%, MgO 3.3%
* Includes HfO2
367
368
Regolith Science
LEVEL OF CONTAMINATION (APPROXIMATE) GRINDING EQUIPMENT
TIME 0
Alumina hand pestle and mortar + motorised agate pestle and mortar
10 min
Motorised agate pestle and mortar
45 min
Agate rotary ball mill with stealite balls
15 min
Sturtevant plate mill + agate pestle and mortar
15 min
Steel rotary ball mill with stealite balls
5 min
1
10
100
1000
10,000 ppm
Fe Mn
Fe Mn Fe Mn Cr Fe Mn Cr Fe Mn Cr Fe
Siebtechnik mill - chrome steel container. 100 ml
20-50 sec
Siebtechnik mill - tungsten carbide container. 100 ml
15-45 sec
Mn Cr Fe
LEVEL OF CONTAMINATION
Mn Cr Co Negligible
Slight
Moderate
W
Strong
Very strong
KSf037-08
Figure 13.8: Summary of contamination tests of different grinding equipment and times.
contamination of Fe and /or potential pathfinder elements (Figure 13.8). More recent testing conducted by Davis and Hansen (1996) has considered contamination introduced during both the crushing and grinding phases of sample preparation. They used quartz during such testing to provide a worst case scenario for contamination. Passing 20 g of quartz though jaw crushers, composed of Mn-rich steel (12–16% Mn), three times, resulted in 50–310 ppm Fe contamination and probably up to 50 ppm Mn contamination. Crushing in a tungsten carbide hydraulic press resulted in lower Fe contamination, but significant W (and Co) contamination (Table 13.3). Grinding of 20 g aliquots of quartz for 1 minute in mills of different composition (5 minutes for a motorised agate mortar and pestle) shows the following results (Table 13.4):
These are worst case scenarios. As shown in Table 13.5, where softer materials are crushed, much less contamination occurs. Similarly, tests using partially stabilised zirconia (PSZ) plates to crush to <3 mm and then agate to reduce samples to <75 µm have shown Table 13.3: Composition of quartz (ppm) after crushing (×3) with jaw crushers or hydraulic press. Quartz control
High Mn steel jaw crushers Wedag Van Gelder
Tungsten carbide press
Cr
0.9
<0.3
<0.3
<0.3
Co
<0.2
<0.2
<0.2
1.0
Cu
2.8
<0.2
<0.2
Fe
<0.2
310
53
Mn
0.4
–
–
–
W
<0.05
0.2
–
38
<0.2 32
Regolith sampling for geochemical exploration
Table 13.4:
Composition of quartz (ppm) after grinding by various mills for 1 minute (5 minutes for agate). Quartz control
Tungsten carbide
Manganese steel
Chrome steel
Alumina
Agate
Al
<100
130
150
130
400
120
Cr
1
1
15
500
1
1
Co
<0.2
170
0.6
1
<5
<0.2
Cu
3
0.6
4
2
0.8
1
Fe
<0.5
<50
3880
2700
<50
50
Pb
0.4
1
0.4
0.4
0.5
0.6
Mn
0.4
0.5
65
16
0.5
0.4
Mo
<0.2
<0.2
2
1
0.6
0.6
Ni
0.3
0.7
9
5
0.3
0.3
W
<0.2
1630
14
7
0.2
0.3
V
3
2
2
7
1
2
Zn
3
7
3
1
2
2
that 230 ppm Zr is introduced into hard (quartz) samples, but lesser amounts for softer materials (Robertson et al. 1996). These results emphasise the importance of using non-metallic grinding equipment in sample preparation if metallic contamination has to be avoided. However, the lack of metallic contamination obtained by the use of non-metallic grinding equipment must be balanced against the longer times taken (Fig 13.8) and the fact that introduced Fe in Table 13.5:
ferruginous samples may be less that the uncertainty in the analysis. The effects of cross-contamination were investigated by crushing quartz in a tungsten carbide mill after massive sulfide ore had been crushed previously. Two cleaning procedures were used: sandblasting with glass beads and grinding with a sand/detergent mixture prior to washing in hot water. Table 13.6 indicates that significant contamination may occur with the latter procedure.
Effect of crushing materials of differing hardness in Mn-steel mill (ppm). Quartz control
Quartz in Mn-steel mill
Magnesite control
Magnesite in Mn-steel mill
Al
<100
150
–
–
Cr
1
15
3
3
Co
<0.2
0.6
0.3
0.2
Cu
3
4
<0.2
<0.2
Fe
<0.5
3880
38
46
Pb
0.4
0.4
–
–
Mn
0.4
65
–
–
Mo
<0.2
2
–
–
Ni
0.3
9
–
–
W
<0.2
14
0.2
<0.05
V
3
2
–
–
Zn
3
3
–
–
369
370
Regolith Science
Table 13.6: Compositions of quartz (ppm) after grinding for 1 minute in a tungsten carbide mill – different cleaning techniques. Quartz control
Quartz, after previously grinding a sulfide sample
Cleaning technique
–
Sand blasting
Sand/detergent/washing
Au (ppb)
<5
<5
79
Cr
1
1
2
Co
170
170
310
Cu
0.6
2
50
Fe
1
1
8
Pb
1
1
8
Mn
0.5
0.7
3.6
Mo
<0.2
<0.2
0.3
Ni
0.7
0.6
0.9
S
<50
<50
65
W
1630
2000
3200
V
2
2
1
Zn
7
3
65
Contaminants that can be introduced prior to crushing include Zn from galvanised iron core trays, and Au from gold ring(s) on the finger(s) of the sampler. Grease from drill rods has been known to result in Mo, W and Zn contamination. Sun screen can also result in Zn contamination. Cross-contamination occurring during drilling can also be significant. The most misleading effect is cross-hole contamination where retention of material from the basal samples from one hole results in its incorporation into the top samples of the next. This can lead to ‘surface anomalies’, which can not be reproduced, especially if the upper metre is collected as a surrogate soil sample (see Section 13.2).
13.15 REFERENCES Agus F and Hesp WR (1974). ‘Factors involved in mineral sample preparation’. Investigation Report 100. CSIRO Division of Mineralogy, Sydney. Anand RR (2001). Evolution, classification and use of ferruginous regolith materials in gold exploration, Yilgarn Craton, Western Australia. Geochemistry: Exploration, Environment, Analysis 1, 221–236.
Anand RR and Cornelius M (2004). Vegetation and soil expression of the Jaguar base metal deposit, Yilgarn Craton. In Regolith 2004. (Ed. IC Roach) pp. 7–8. CRC LEME, Perth. Anand RR and Gilkes RJ (1987). The association of maghemite and corundum in Darling Range laterites, Western Australia. Australian Journal of Soil Research 35, 303–11. Anand RR and Paine M (2002). Regolith geology of the Yilgarn Craton, Western Australia: implications for exploration. Australian Journal of Earth Sciences 49, 3–162. Anand RR, Paine MD and Smith RE (2002). ‘Genesis, classification and atlas of ferruginous materials, Yilgarn Craton’. Open File Report 73, CRC LEME, Perth. Anand RR, Cornelius M and Phang C (2007). Use of vegetation and soil in mineral exploration in areas of transported overburden, Yilgarn Craton, Western Australia: a contribution towards understanding metal transportation processes. Geochemistry: Exploration, Environment, Analysis 7, 267–288. Andrew AS, Carr GR, Giblin AM and Whitford DJ (1998). Isotope hydrogeochemistry in exploration
Regolith sampling for geochemical exploration
for buried and blind mineralisation. In The State of the Regolith. (Ed. RA Eggleton) pp. 220–225. Special Publication No 20. Geological Society of Australia, Sydney. Arne DC, Stott JE and Waldron HM (1999). Biogeochemistry of the Ballarat East goldfield, Victoria, Australia. Journal of Geochemical Exploration 67, 1–14. Arne DC, Hughes MJ, Walters B and Waldron HM (2001). The application of biogeochemistry to gold exploration in Central Victoria. Australian Institute of Geoscientists Bulletin 34, 53–62. Ashley PM and Wolfenden BJ (2005). Halls peak massive sulphide deposits, New England, NSW. In Regolith expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 163–164. CRC LEME, Perth. Aspandiar MF, Anand RR, Gray DJ and Cucuzza J (2004). Mechanisms of metal transfer through transported overburden within the Australian regolith. Explore 125, 9–12. Bampton KF, Collins AR, Glasson KR and Guy BB (1977). Geochemical indications of concealed copper mineralization in an area northwest of Mount Isa Queensland, Australia. Journal of Geochemical Exploration 8, 169–188. Barnes JFH (1987). Practical methods of drill hole sampling. Australian Institute of Geoscientists Bulletin 7, 145–170. Brooks RR and Radford CC (1978). An evaluation of the background and anomalous copper and zinc concentrations in the ‘copper plant’ Polycarpaea spirostylis and other Australian species of the genus. Proceedings of the Australasian Institution of Mining and Metallurgy 268, 33–37. Brooks RR, Dunn CE and Hall GEM (Eds) (1995). Biological Systems in Mineral Exploration and Processing. Ellis Horwood, London. Brown AD and Hill SM (2005). White Dam Au-Cu prospect, Curnamona Province, South Australia. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 392–394. CRC LEME, Perth. Brown E, Skougstad MW and Fishman FJ (1970). ‘Methods for collection and analysis of water sam-
ples for dissolved minerals and gases.’ Techniques of Water-Resource Investigation, United States Geological Survey , Washington, DC. Butt CRM (1979). Geochemistry of a pseudogossan developed on a black shale-dolerite contact, Killara, Western Australia. Journal of Geochemical Exploration 11, 131–156. Butt CRM (1985). Granite weathering and silcrete formation on the Yilgarn Block, Western Australia. Australian Journal of Earth Sciences 32, 415–432. Butt CRM (1988). Major uranium provinces: Yilgarn Block and Gascoyne Province. In Recognition of Uranium Provinces. pp 273–304. International Atomic Energy Agency, Vienna. Butt CRM (1992). Teutonic Bore Cu-Zn-Ag-Pb deposit, Western Australia. In Regolith Exploration Geochemistry in Tropical and Subtropical Terrains. Handbook of Exploration Geochemistry, Volume 4. (Eds CRM Butt and H Zeegers) pp. 336–338. Elsevier, Amsterdam. Butt CRM (2005a). Geochemical dispersion, process and exploration models. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 81–106. CRC LEME, Perth. Butt CRM (2005b). Killara copper pseudogossan, Western Australia. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 168–170. CRC LEME, Perth. Butt CRM and Brand NW (2005). Mt Keith nickel sulphide deposit, Western Australia. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 112–114. CRC LEME, Perth. Butt CRM and Dries S (2005). Fender gold deposit, Cue, Western Australia. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 247–249. CRC LEME, Perth. Butt CRM and Gole MJ (1986). Groundwater helium surveys in mineral exploration in Australia. Journal of Geochemical Exploration 25, 309–344. Butt CRM and Sheppy NR (1975). Geochemical exploration problems in Western Australia,
371
372
Regolith Science
exemplified by the Mt. Keith area. In Geochemical Exploration 1974, Developments in Economic Geology, Volume 1. (Eds IL Elliott and WK Fletcher) pp. 391–415. Elsevier, Amsterdam. Butt CRM, Horwitz RC and Mann AW (1977). ‘Uranium occurrences in calcrete and associated sediments in Western Australia’. Report FP 16. CSIRO Division of Mineralogy, Floreat Park, WA. Butt CRM, Gray DJ, Robertson IDM, Lintern MJ, Anand RR, Britt AF, Bristow APJ, Munday TJ, Phang C, Smith RE and Wildman JE (1997). ‘Geochemical exploration in areas of transported overburden, Yilgarn Craton and environs, Western Australia: Final Report’. CSIRO Exploration and Mining Restricted Report 333R, Perth. (Reissued as Open File Report 86, 2001. CRC LEME, Perth.) Butt CRM, Robertson IDM, Scott KM, and Cornelius M (Eds) (2005a) Regolith Expression of Australian Ore Systems. CRC LEME, Perth. Butt CRM, Scott KM, Robertson IDM and Cornelius M (2005b) Sample media. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 53–79. CRC LEME, Perth. Cameron E, Mazzucchelli RH and Robbins TW (1980). Yeelirrie calcrete uranium deposit, Murchison District, W.A. Journal of Geochemical Exploration 12, 350–353. Carey ML, McPhail DC and Taufen PM (2003). Groundwater flow in playa lake environments: impact on gold and pathfinder element distributions in groundwaters surrounding mesothermal gold deposits, St. Ives area. Eastern Goldfields, Western Australia. Geochemistry: Exploration, Environment, Analysis 3, 57–71. Carver RN, Chenoweth LM, Mazzucchelli RH, Oates CJ and Robbins TW (1987). ‘Lag’ – A geochemical sampling medium for arid terrain. Journal of Geochemical Exploration 28, 183–199. Chen XY, Lintern MJ and Roach IC (Eds) (2002). Calcretes: Characteristics, Distribution and Use in Mineral Exploration. CRC LEME, Perth. Cohen DR, Hoffman EL and Nichol I (1987). Biogeochemistry: A geochemical method for gold exploration in the Canadian Shield. Journal of Geochemical Exploration 29, 49–73.
Cohen DR, Shen XY, Dunlop AC and Rutherford NF (1998). Comparison of selective extraction soil geochemistry and biogeochemistry in the Cobar area, New South Wales. Journal of Geochemical Exploration 61, 173–189. Cohen DR, Silva-Santisteban CM, Rutherford NF, Garnett DL and Waldron HM (1999). Comparison of vegetation and stream sediment geochemical patterns in northeast New South Wales. Journal of Geochemical Exploration 68, 469–489. Cole MM, Provan DMJ and Tooms JS (1968). Geobotany, biogeochemistry and geochemistry in the Bulman-Waimuna Springs area, Northern Territory, Australia. Transactions of the Institute of Mining and Metallurgy 77, B81–B103. Cornelius M, Smith RE and Cox A (2001). Laterite geochemistry for regional exploration surveys – A review and sampling strategies. Geochemistry: Exploration, Environment, Analysis 1, 211–220. Cornelius M, Robertson IDM, Cornelius AJ and Morris PA (2007). ‘Lateritic geochemical database for the western Yilgarn Craton, Western Australia’. Record 2007/9. Geological Survey of Western Australia, Perth. Cox R (1975). Geochemical soil surveys in exploration for nickel copper sulphides at Pioneer, near Norseman, Western Australia. In Geochemical Exploration 1974. Developments in Economic Geology, Volume 1. (Eds IL Elliott and WK Fletcher) pp. 437–460. Elsevier, Amsterdam. Cox R and Curtis R (1977). The discovery of the Lady Loretta zinc-lead-silver deposit, northwest Queensland, Australia – a geochemical exploration case history. Journal of Geochemical Exploration 8, 189–202. Davis JJ and Hansen GW (1996). ‘Metal and trace element contamination during sample preparation’. CSIRO Division of Exploration and Mining Restricted Report 233R, Perth. de Caritat P, Kirste D, Carr G and McCulloch M (2005). Groundwater in the Broken Hill region, Australia: recognising interaction with bedrock and mineralisation using S, Sr and Pb isotopes. Applied Geochemistry 20, 767–787. Deutscher RL, Mann AW and Butt CRM (1980). Model for calcrete uranium mineralization. Journal of Geochemical Exploration 12, 158–161.
Regolith sampling for geochemical exploration
Dickson BL and Scott KM (1998). Recognition of aeolian soils of the Blayney district, NSW: implications for geochemical exploration. Journal of Geochemical Exploration 63, 237–251. Dunn C (2007). Biogeochemistry in Mineral Exploration. Handbook of Exploration and Environmental Geochemistry 9. Elsevier, Amsterdam. Dunn CE (1989). Reconnaissance-level biogeochemical surveys for gold in Canada. Transactions of the Institute of Mining and Metallurgy, Section B: Applied Earth Sciences 98, 153–161. Dunn CE and Scagel RK (1989). Tree-top sampling from a helicopter – a new approach to gold exploration. Journal of Geochemical Exploration 34, 255–270. Farrell BL and Orr DB (1980). Pinnacles Cu-Pb-Zn prospect, Dundas Trough, Tasmania. Journal of Geochemical Exploration 12, 281–284. Fletcher WK and Loh CH (1996). Transport of cassiterite in a Malaysian stream; implications for geochemical exploration. Journal of Geochemical Exploration 57, 9–20. Garnett DL, Rea WJ and Fuge R (1982). Geochemical exploration techniques applicable to calcrete-covered areas. In Proceedings of the 12th Commonwealth Mining and Metallurgical Institute Congress. (Ed. HW Glenn) pp. 945–955. Geological Society of South Africa, Johannesburg. Giblin AM (2003). Groundwater geochemistry from some major gold deposits and exploration targets in northern Victoria and southern NSW. In Victoria Undercover, Benalla 2002. (Eds GN Phillips and KS Ely) pp. 101–105. CSIRO Publishing, Melbourne. Giblin A (2005). Alligator Rivers uranium deposits (Koongarra, Nabarlek and Ranger One). In Regolith expression of Australian ore systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 411–414. CRC LEME, Perth. Giblin AM and Snelling AA (1983). Applications of hydrogeochemistry to uranium exploration in the Pine Creek Geosyncline, Northern Territory, Australia. Journal of Geochemical Exploration 19, 33–55. Glasson KR (1973). A study of anomalous copperlead-zinc values in relation to rock types and weathering profiles in an area north of Mount Isa.
The Australasian Institute of Mining and Metallurgy, Western Australian Conference. pp. 27–40. Gole MJ, Butt CRM and Snelling AA (1986). A groundwater helium survey of the Koongarra uranium deposits, Pine Creek Geosyncline, Northern Territory. Uranium 2, 343–360. Gray DJ (2001). Hydrogeochemistry in the Yilgarn Craton. Geochemistry: Exploration, Environment, Analysis 1, 253–264. Gray DJ and Noble RPP (2006). ‘Nickel hydrogeochemistry in the NE Yilgarn Craton, Western Australia’. CRC LEME Report 243/CSIRO Exploration and Mining Report P2006/524. CRC LEME and CSIRO Exploration and Mining, Perth. (Reissued 2007 as Open File Report 212, CRC LEME, Perth). Greig (DD) 1983. Primary and secondary dispersion at the Teutonic Bore deposit. In Geochemical Exploration in the Eastern Goldfields Region of Western Australia: Tour Guide. (Ed. BH Smith) pp. 73–87. Association of Exploration Geochemists, Perth. Gulson BL (1986). Lead Isotopes in Mineral Exploration. Elsevier, Amsterdam. Hale M and Plant JA (Eds) (1994). Drainage Geochemistry. Handbook of Exploration Geochemistry Volume 6. Elsevier, Amsterdam. Hill LJ (2002). Branching out into biogeochemical surveys: a guide to vegetation sampling. In Regolith and Landscapes in Eastern Australia. (Ed. IC Roach) pp. 50–53. CRC LEME, Perth. Hill SM, Thomas M, Earl K and Foster KA (2005). Flying Doctor Ag-Pb-Zn prospect, Northern Leases, Broken Hill, NSW. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 146–148. CRC LEME, Perth. Hulme KA and Hill SM (2004). Seasonal element variations of Eucalyptus camaldulensis biogeochemistry and implications for mineral exploration: an example from Teilta, Curnamona Province, western NSW. In Regolith 2004. (Ed. IC Roach) pp. 151– 156. CRC LEME, Perth. Keeling JL and Hartley KL (2005). Poona and Wheal Hughes Cu deposits, Moonta, South Australia. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 383–385. CRC LEME, Perth.
373
374
Regolith Science
Kirste D, de Caritat P and Dann R (2003). The application of the stable isotopes of sulfur and oxygen in groundwater sulfate to mineral exploration in the Broken Hill district of Australia. Journal of Geochemical Exploration 78-79, 81–84. Lawrance LM (1999). Multi-element dispersion in Mesozoic basin sediment over the Osborne Deposit, northern Queensland, Australia: Implications for regional geochemical exploration in buried terrain. Australian Institute of Geoscientists Bulletin 28, 73–81. Lech ME and de Caritat P (2007). Baseline geochemical survey of the Riverina region of New South Wales and Victoria, Australia: concentrations and distributions of As, Ba, Br, Cd, Co, Cr, F, Ga, Mo, Sb, U and V compared to national and international guidelines. Geochemistry: Exploration, Environment, Analysis 7, 233–247. Lintern MJ (2002). Calcrete sampling for mineral exploration. In Calcrete: Characteristics, Distribution and Use in Mineral Exploration. (Eds XY Chen, MJ Lintern and IC Roach) pp. 31–109. CRC LEME, Perth. Lintern MJ (2005). Higginsville palaeochannel gold deposits, Kambalda, Western Australia. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 264–266. CRC LEME, Perth. Lintern MJ (2007). Vegetation controls on the formation of gold anomalies in calcrete and other materials at the Barns gold prospect, Eyre Peninsular, South Australia. Geochemistry: Exploration, Environment, Analysis 7, 247–266. Lintern MJ and Sheard MJ (1998). Silcrete – a potential new exploration sample medium: a case study from the Challenger gold deposit. MESA Journal 11, 16–20. Lintern MJ, Downes PM and Butt CRM (1992). Bounty and Transvaal Au deposits, Western Australia. In Regolith Exploration Geochemistry in Tropical and Subtropical Terrains. Handbook of Exploration Geochemistry, Volume 4. (Eds CRM Butt and H Zeegers) pp. 351–355. Elsevier, Amsterdam. Lintern MJ, Butt CRM and Scott KM (1997). Gold in vegetation and soil – three case studies from the
goldfields of southern Western Australia. Journal of Geochemical Exploration 58, 1–14 Mahizhnan A (2004). Red-brown hardpan: Distribution, origin and exploration implications for gold in the Yilgarn Craton of Western Australia. Ph.D thesis, Curtin University of Technology, Perth. Mann AW (1980). Waters. Journal of Geochemical Exploration 12, 133–136. Mann AW and Deutscher RL (1978). Genesis principles for the precipitation of carnotite in calcrete drainages in Western Australia. Economic Geology 73, 1724–1737. Marshall AE and Goldsworthy J (2005). Hardpans and base of hardpan geochemical surveys. In From Tropics to Tundra, Program and Abstracts. 22nd International Geochemical Exploration Symposium, Perth. p. 79. Association of Applied Geochemists (Ottawa) and Promaco Conventions, Perth. Mazzucchelli RH (1994). Drainage geochemistry in arid regions. In Drainage Geochemistry. Handbook of Exploration Geochemistry, Volume 6. (Eds M Hale and JA Plant) pp. 379–414. Elsevier, Amsterdam. Mazzucchelli RH and James CH (1966). Arsenic as a guide to gold mineralization in laterite-covered areas of Western Australia. Transactions of the Institute of Mining and Metallurgy, Section B: Applied Earth Sciences 75, 286–294. Mazzucchelli RH, Chapple BEE and Lynch JE (1980). Northern Yorke Peninsular Cu, Gawler Block, S.A. Journal of Geochemical Exploration 12, 203–207. McQueen KG, Munro DC, Gray DJ and Le Gleuher M (2004). Weathering-controlled fractionation of ore and pathfinder elements, Part II: the lag story unfolds. In Regolith 2004. (Ed. IC Roach) pp. 112–116. CRC LEME, Perth. Melo G and Fletcher WK (1999). Dispersion of gold and associated elements in stream sediments under semi-arid conditions, northeast Brazil. Journal of Geochemical Exploration 67, 235–243. Moeskops PG and White AH (1980). Beltana and Aroona Zn deposits, Adelaide Geosyncline, S.A. Journal of Geochemical Exploration 12, 272–275. Netterburg F (1967). Some road-making properties of South African calcretes. In Proceedings of the Fourth Regional Conference of African Soil Mechanics and Foundation Engineering. 1, 77–81. Cape Town.
Regolith sampling for geochemical exploration
Nicolls OW, Provan DMJ, Cole MM and Tooms JS (1965). Geobotany and geochemistry in mineral exploration in the Dugald River Area, Cloncurry District, Australia. Transactions of the Institute of Mining and Metallurgy 74, 695–799. Palethorpe C (1980). Khan’s Creek Cu-Pb-Zn prospect, Nambucca Block, N.S.W. Journal of Geochemical Exploration 12, 320–322. Pate JS, Jeschke DW, Dawson TE, Raphael C, Hartung W and Bowen BJ (1998). Growth and seasonal utilisation of water and nutrients by Banksia prionotes. Australian Journal of Botany 46, 511–532. Pontual S and Merry N (1996). ‘An exploratory strategy to aid the differentiation of residual from transported kaolinites using field-based spectral analysis’. Restricted Report. Ausspec International, Melbourne. Rattigan JH, Gersteling RW and Tonkin DG (1977). Exploration geochemistry of the Stuart Shelf, South Australia. Journal of Geochemical Exploration 8, 203–217. Reid N, Hill SM and Lewis D (2007). Mineral expression and plant species differences at the Titania prospect: biogeochemical sampling in the Tanami region, Northern Territory, Australia. In Exploring our Environment, 23rd International Applied Geochemistry Symposium. Oviedo, Spain. (Ed. J Loredo) p. 122. Association of Applied Geochemists (Ottawa) and Departamento de Explotación y Prospección de Minas, University of Oviedo. Roach IC (2004). Results of a preliminary biogeochemical survey of the Wyoming gold deposit, Tomingley, NSW. In Regolith 2004. (Ed. IC Roach) pp. 307–310. CRC LEME, Perth. Robertson IDM (1989). ‘Geochemistry, petrography and mineralogy of ferruginous lag overlying the Beasley Creek Gold Mine – Laverton WA’. CSIRO Division of Exploration Geoscience Restricted Report 27R, Volumes 1 and 2, Perth. (Reissued as Open File Report 10, CRC LEME, Perth. 1998.) Robertson IDM (1995). Interpretation of fabrics in ferruginous lag. AGSO Journal of Australian Geology and Geophysics 16, 263–270. Robertson IDM (1996). Ferruginous lag geochemistry on the Yilgarn Craton of Western Australia; practi-
cal aspects and limitations. Journal of Geochemical Exploration 57, 139–151. Robertson IDM (1999). Origins and applications of size fractions of soils overlying the Beasley Creek gold deposit, Western Australia. Journal of Geochemical Exploration 66, 99–113. Robertson IDM (2003). Dispersion into the Tertiary Southern Cross Formation sediments from the Scott and Cindy Lodes, Pajingo, N.E. Queensland, Australia. Geochemistry: Exploration, Environment, Analysis 3, 39–50. Robertson IDM and Butt CRM (1993). ‘Atlas of weathered rocks’. CSIRO Division of Exploration Geoscience Restricted Report 390R, Perth. (unpaginated). (Also published as Open File Report 1, CRC LEME, Perth). Robertson IDM and Wills R (1993). ‘Petrology and geochemistry of surface materials overlying the Bottle Creek Gold Mine, WA’. CSIRO Division of Exploration Geoscience Restricted Report 394R, Perth. (Reissued as Open File Report 56, CRC LEME, Perth. 1998). Robertson IDM, Crabb JF and Hart MKW (1989). ‘Assessment of contamination from a PSZ armoured jaw crusher’. Restricted Report 9R. CSIRO Division of Exploration Geoscience, Perth. Robertson IDM, Phang C and Munday TJ (1999). The regolith geology around the Harmony gold deposit, Peak Hill, WA. In Regolith ’98, Australian Regolith and Mineral Exploration. (Eds G Taylor and C Pain) pp. 283–298. CRC LEME, Perth. Ross AF and Schellekens RR (1980). Mt. Lindsay Sn prospect, Dundas Trough, Tasmania. Journal of Geochemical Exploration 12, 293–296. Rutherford NF, Lawrance LM and Sparks G (2005). Osborne Cu-Au deposit, Cloncurry District, NW Queensland. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 380–382. CRC LEME, Perth. Ryall WR and Nicholas T (1979). Surface geochemical and biogeochemical expression of base metal mineralization at Woodlawn, New South Wales, Australia. Journal of the Geological Society of Australia 26, 187–195.
375
376
Regolith Science
Scott K (1996). Composition of white mica in weathered rocks: indicators of rock type and proximity to gold mineralisation, Western Australia. Explore 93, 3–5. Scott KM and van Riel B (1999). The Goornong South Gold Deposit and its implications for exploration beneath cover in Central Victoria, Australia. Journal of Geochemical Exploration 67, 83–96 Scott KM, French DH and Griffin WL (1994). The application of rutile geochemistry in base metal exploration. In Australian Research on Ore Genesis Symposium. pp. 17.1–17.5. Australian Mineral Foundation, Adelaide. Scott KM, Ashley PM and Lawie DC (2001). The geochemistry, mineralogy and maturity of gossans derived from volcanogenic Zn-Pb-Cu deposits of the eastern Lachlan Fold Belt, NSW, Australia. Journal of Geochemical Exploration 72, 169–191. Severne BC (1974). Nickel accumulation by Hybanthus floribundus. Nature 248, 807–808. Skey EH and Young CH (1980). Que River Zn-Pb deposit, Dundas Trough, Tas. Journal of Geochemical Exploration 12, 284–290. Smee BW and Stanley CR (2005). Technical note: Sample preparation of ‘nuggetty’ samples: dispelling some myths about sample size and sampling errors. Explore 126, 21–27. Smith RE and Perdrix RL (1983). Pisolitic laterite geochemistry in the Golden Grove massive sulphide district, Western Australia. Journal of Geochemical Exploration 18, 131–164. Smith RE, Campbell NA and Perdrix JL (1983). Identification of some Western Australian gossans by multi-element geochemistry. In Geochemical Exploration in Deeply Weathered Terrain. (Ed. RE Smith) pp. 75–90. CSIRO Division of Mineralogy, Floreat Park, Western Australia.
Smith RE, Anand RR and Alley NF (2000). Use and implications of paleoweathering surfaces in mineral exploration in Australia. Ore Geology Reviews 16, 185–204. Tate SE, Greene RSB, Scott KM and McQueen KG (2007). Recognition and characterisation of the Aeolian component in soils in the Girilambone Region, north-western New South Wales, Australia. Catena 69, 122–133. Taylor GF (1973). The geochemistry of siderite in relation to ironstones in the Paradise Creek Formation, Northwest Queensland. Journal of Geochemical Exploration 2, 367–382. Taylor GF and Scott KM (1982). Evaluation of gossans in relation to lead-zinc mineralization in the Mount Isa Inlier, Queensland. Australian Bureau of Mineral Resources, Geology and Geophysics Bulletin 7, 159–180. Taylor GF, Wilmshurst JR, Butt CRM and Smith RE (1980). Gossans. Journal of Geochemical Exploration 12, 30–32. Taylor GF, Wilmshurst JR, Togashi Y and Andrew AS (1984). Geochemical and mineralogical haloes about the Elura Zn-Pb-Ag orebody, western New South Wales. Journal of Geochemical Exploration 22, 265–290. Taylor GF and Thornber MR (1992). Gossan and ironstone surveys. In Regolith Exploration Geochemistry in Tropical and Subtropical Terrains. Handbook of Exploration Geochemistry, Volume 4. (Eds CRM Butt and H Zeegers) pp. 139–202. Elsevier, Amsterdam. Whitford DJ, Andrew AS, Carr GR and Giblin AM (1998). The application of isotope studies in Australian groundwaters to mineral exploration: The Abra Prospect, Western Australia. In Proceedings of the 9th International Symposium on Water-Rock Interaction. (Eds GB Arehart and JR Hulston) pp. 583–586. A.A. Balkema, Rotterdam.
14
Extraterrestrial regolith Jonathan D A Clarke
14.1 INTRODUCTION Why should a book on terrestrial regolith contain a chapter on the regolith found on other planets? There are several reasons for this. First, historically regolith geology has never been tied to one planet and has received considerable developmental impetus from studying the regolith of other planets. Secondly, terrestrial regolith can provide an analogue for that found on other planets, especially Mars. This is particularly true for the Australian regolith, which, with its long history, is increasingly attracting the attention of overseas planetary scientists (for example, Bourke and Zimbelman 2000, 2001; Benison and Bowen 2006). Detailed understanding of planetary regolith compositions and properties is also critical to many applied areas in planetary science. The designers of landing systems, such as pads, airbags, and braking rockets, need to understand how these technologies will interact with the regolith. For example, braking jets will disturb the surface – causing billowing of dust and other loose materials that may damage spacecraft and their systems. The wheels of rovers needed to be designed with reference to the bearing strength and roughness of the surfaces to be traversed, while protection of moving surfaces against regolith fines is a major challenge for operations on both the
Moon and Mars. Regolith may also present a health hazard to crews when regolith fines enter the spacecraft via the airlock during surface operations (Beaty et al. 2005). Lastly, planetary regolith may be a source of useful resources – in particular water and oxygen – as well as construction materials. Use of in situ resources, including those in the regolith, is a key to minimising the mass, and therefore cost, of human missions to the Moon and Mars (see Lewis et al. 1993) for a compendium of papers on possible extraction of resources such as oxygen and solar wind volatiles from the lunar regolith, water from the sub-surface of Mars, and the use of regolith materials for construction purposes). This chapter reviews features on planets, satellites and asteroids (Figure 14.1) relative to terrestrial regolith features.
14.2 THE MOON 14.2.1 Introduction The lunar regolith is the most extensively studied of all extraterrestrial examples, through six crewed and 13 robotic missions in the period 1966–1976. Because of the extensive research on the lunar regolith and the fact that it was carried out early on in the history of
Ch ar on an d
Pl ut o
ep tu ne N
Ur
an
us
an d Sa tu rn
Ju pi te r
Ea rth an d M M ar oo s n As te ro id s
M er cu Ve ry nu s
Ti ta n
Regolith Science
Su n
378
Figure: 14.1. The Solar System, showing the extraterrestrial bodies studied.
space exploration, models of lunar regolith formation have been very influential in the development of broader models for extra terrestrial surfaces – especially rocky, airless surfaces, such as asteroids, rocky satellites and Mercury and, to a lesser degree, the surfaces of airless icy bodies and Mars. The naked eye is able to distinguish the major units of the lunar crust. The light areas are anorthositic in composition and, on the near side of the Moon, are mostly found in highland areas. These represent the primordial lunar crust regolith. The dark areas, which are composed of various basaltic outflows, form most of the lunar lowland plains, and overlie the anorthosites (Taylor 2001). Various volcanic landforms have been observed on the lava plains, including wrinkle ridges formed through compression, lobe flow margins, sinuous rilles formed by lava channels, and a range of small volcanic features, such as domes (Greeley 1994). 14.2.2 Chronology of the lunar surface The dominant process shaping the lunar surface is impact – as the number of craters per unit area increases over time, the crater density of different geomorphological units allows a relative chronology to be constructed. Radiometric dating of specific bedrock
units provides a numerical constraint on the relative chronology. As a result, the lunar surface has been divided into units of different ages. These are named after type regions on the Moon and are termed (Cattermole 2001):
s s s s s
Copernican (<1.10 Ga) Eratosthenian (3.20–1.10 Ga) Imbrian (3.85–3.20 Ga) Nectarian (3.92–3.85 Ga) Pre-Nectarian (>3.92 Ga).
This methodology of relative age dating by crater densities has been extended to other planets and satellites. However, with the exception of the Earth and Moon, the crater fluxes are not constrained by absolute dating and are based on a combination of modelling and extrapolation from the lunar and terrestrial cratering records. While lunar stratigraphy has proved to be the basis of surface chronology throughout the solar system, correlations must be treated as very tentative. 14.2.3 Architecture of the lunar regolith The lunar regolith (see Taylor 1975 for a summary of the immediate post Apollo and Luna era science) averages several metres in thickness and blankets almost
Extraterrestrial regolith
Figure 14.2: Panorama of Shorty crater in the Taurus Littrow Valley. Despite the size of the crater, in situ bedrock is not exposed and the entire depth of the crater has been excavated in regolith (Note: Apollo 17 lunar rover at left for scale). Image courtesy moonpans.com.
the entire surface (Figure 14.2). It is thickest on the oldest (Nectarian) and thinnest on the youngest (Copernican) surfaces. In situ bedrock was observed only at a distance in the sides of Hadley Rille during the Apollo 15 mission, and not at any of the other areas visited by manned and unmanned missions. Because of the number of sites visited by manned and unmanned spacecraft during the 1960s and 1970s, and numerous orbital, fly-round and flyby missions continuing to the present, the lunar regolith is better characterised than that of any other extraterrestrial body and has greatly influenced conceptual models for their development. The Lunar regolith has a complex local structure composed of overlapping ejecta deposits. Despite this layering, the regolith is compositionally well mixed because of continuous reworking of the regolith by repeated impacts (‘impact gardening’; Iriyama and Honda 1979). The gardening results in deeper parts of the regolith being periodically exposed and redistributed across the surface. The absence of an atmosphere greatly limits the degree to which the regolith is reworked, except by impact. However, localised textures suggestive of creep have been reported and the trails of boulders dislodged by moonquakes have been found. The most important non-impact reworking of the regolith is by slides and slips along steep slopes, especially crater walls. Some of these can produce small gullies (Bart 2007). Beneath the surface regolith is the mega regolith (Hartmann 1973). The mega regolith consists of the
upper part of the lunar crust that has been fractured and otherwise disturbed by large impacts. The evidence for the mega regolith lies in the unusual seismic properties of the upper part of the crust suggesting extensive fracturing and localised cavities. Beneath the lunar plains the mega regolith is predicted to be several hundred metres thick; beneath the highlands, in the areas with the longest exposure histories, the mega regolith may extend to a depth of 2 km. 14.2.4 Composition of the lunar regolith The lunar regolith consists of several different components. They include bedrock fragments of all sizes that consist both of rocks from the immediate vicinity and clasts ballistically implanted by more distant impacts. Mixed in with the rock fragments are two other common components: glasses and agglutinates. Glasses consist of small (100 µm) droplets of chilled melt that take on a range of shapes. Most are of impact morphology, but some – such as the famous orange soil at the Apollo 17 Taurus-Littrow site – are derived from lava fountains. Agglutinates are glass-cemented aggregates of small rock fragments and mineral grains. The formation of agglutinates works against the fragmentation induced by bombardment, with an equilibrium being reached at the 60 µm size fraction (Figure 14.3). Other, minor contents of the lunar regolith are meteoritic fragments including nickel-iron particles, and implanted gases from the solar wind. In addition to impact gardening, the regolith is modified by space weathering. Modification of the
379
380
Regolith Science
a
c
b
Figure 14.3: Lunar regolith textures. (a) Thin section of Apollo 17 micro breccia under cross-polars. Field of view 1.3 mm. Image courtesy Marc Norman, ANU. (b) Green glass spherule (100 µm across) in a regolith sample from Mare Crisium, collected by Luna 24. Image courtesy Marc Norman, ANU. (c) Agglutinate fragment (about 3 mm across) from Apollo 17 mission. Reproduced with the permission of the Mineralogical Society of America.
regolith that are categorised as space weathering include: formation of micro craters (Horz et al. 1971) and glasses by micrometeorite impacts, welding through ion sputtering, fragmentation through thermal exfoliation, and mineral damage through exposure to ultraviolet radiation and cosmic rays. 14.2.5 Polar ice deposits? Permanently shadowed regions at the lunar poles have long been thought to be potential trap of volatiles released by passing comets (Feldman et al. 2001). Hydrogen has been detected from orbiting spacecraft in the permanently shadowed areas, which is consistent with the presence of ice in the regolith. However, attempts to detect water in the plumes of crashed spacecraft or in radar signals have proved unsuccessful (Hodges 2002). The presence of such deposits – if confirmed to occur in significant quantities – would greatly facilitate the establishment of manned stations on the lunar surface.
14.3 MARS 14.3.1 Introduction Smaller than the Earth but larger than the Moon, Mars should not be regarded as being like the Moon with the addition of a thin atmosphere and a trace of moisture, or a frozen version of an embryonic Earth – or even as a combination of the two. Mars has experienced its own history, with its own unique attributes, including the decay of a global magnetic field early in its history (see Cattermole 2001) and the presence of
two condensable volatiles (CO2 and H2O) in its atmosphere: a theme of a book by Kargel (2004). For the most part, Mars lacks fold-belt mountain ranges, and the surface is composed of a mosaic of overlapping sedimentary and volcanic deposits truncated by impact structures, major rift systems, and giant fluvial incisions. Five sites on the surface had been visited (at the time of writing) by various space probes. Compared with lunar landing missions, the Mars landers have sampled considerably fewer sites and those visited have been restricted by severe engineering constraints on factors such as roughness and slope. It is highly likely that these sites are unrepresentative of much of the Martian surface. 14.3.2 Martian surface chronology The complex surface of Mars has, like the Moon, been divided into a number of eras based on crater densities. Age limits on these are far more uncertain, due to the lack of absolute dating of units, the complex surface history of Mars that has seen burial and exhumation of many land surfaces, erosional modification and destruction of craters – especially smaller ones – and the presence of a greater proportion of secondary craters (Hartmann et al. 2001). The Martian eras, based on crater densities are (Cattermole 2001):
s s
L Amazonian (younger limit <0.25 Ga; older limit <0.70 Ga) M Amazonian (younger limits 0.25–0.70 Ga; older limits 0.70–2.30 Ga)
Extraterrestrial regolith
s s s s s s
E Amazonian (younger limits 0.70–1.80 Ga; older limits 2.3–3.55 Ga) L Hesperian (younger limits 1.80–3.10 Ga; older limits 3.55–3.70 Ga) E Hesperian (younger limits 3.10–3.55 Ga; older limits 3.70–3.80 Ga) L Noachian (younger limits 3.50–3.85 Ga; older limits 3.80–4.40 Ga) M Noachian (younger limits 3.85–3.92 Ga; older limits 4.40–4.60 Ga) E Noachian (younger limit >3.92 Ga; older limit >4.60 Ga).
The major events of these eras can be found in the excellent summary for the general reader by Hartmann (2003). The Noachian era is characterised by large impact basins, lava plains and intercalated sediment deposits – the oldest valley systems and dendritic valley networks apparently formed either by run-off or snowmelt. The largest drainage system, Uzboi Vallis, has its headwater in the Martian south polar region and discharged into the northern hemispherical depression of Vasitas Borealis. Similar, but smaller, drainages are associated with Ma’adim Vallis and Gusev Crater. Vasitas Borealis itself is likely to be the site of a former Martian ocean (Perron et al. 2007). The Hesperian appears characterised by the development of large basaltic shield volcanoes and tectonism. Major Hesperian volcanoes include Nilli and Tyrrhena Patera. Tectonic features include the Valles Marineris rift system and the Thaumasia fold belt. The igneous and tectonic activity may have given rise to large-scale melting of sub-surface ice, leading to chaotic terrain formed by collapse, sapping valley networks, and large outflow channels – most draining into large impact basins or the northern plains. Numerous sedimentary deposits have also been found, including those of Meridiani Planum (Grotzinger et al. 2005) studied by the Opportunity rover. The Amazonian era was marked by the construction of the youngest large volcanic edifices, such as Olympus Mons and other volcanoes of the Tharsis uplift, and the somewhat smaller peaks of the Elysium province. The polar ice caps, layered terrain and the many dune fields are all of Amazonian age (see Cattermole 2001). Very strong variability in the obliquity of
the Martian ecliptic has resulted in alternating build up of ice at the poles and at the equator (Head et al. 2006), and volcanism (and, by implication, volcanicdriven melting of ground ice) may have continued to within the last few million years (Neukum et al. 2004). Bibring et al. (2006) observed that, although much of the surface is obscured by mantling dust, there is an approximate correlation between rock units of different ages and different mineral assemblages detected by the OMEGA spectrometer in the Mars Express spacecraft. Phyllosilicates formed by aqueous alteration are confined to bedrock of Noachian age. Sulfates formed from acid brines are most common in rocks corresponding to the Hesperian era, and anhydrous ferric oxides formed by slow superficial weathering with little or no water are most common in the Amazonian (see Chapter 12 for terrestrial details). Bibring et al. (2006) proposed names for these periods: the ‘Phyllocian’ era for the period rich in clays; the ‘Theiikian’ era when sulfates were forming; and the ‘Siderikian’ era for the period dominated by anhydrous ferric oxides. While these terms are not yet generally accepted, the recognition that different phases of Martian history had distinctive weathering signatures has major implications for models of landscape evolution on the planet. Standard models of Martian weathering history (Hurowtiz and McLennan 2007) suggest a predominance of water-limited acidic conditions. However, the apparent scarcity in orbital spectral data of the clays and carbonates required by mass balance considerations, and the very limited ground checking (only five sites), shows that our understanding of Martian surface processes is still very limited. A recent suggestion that reconciles the apparently contradictory features is that of Halevy et al. (2007), who proposed that an early Martian greenhouse was sustained by volcanic SO2 – resulting in a sulfur cycle analogous to the terrestrial carbon cycle. This led to slightly acid surface water that inhibited carbonate deposition and precipitated sulfite (SO3) minerals. Reduction in the rate of volcanism led to breakdown of the cycle. Atmospheric SO2 was taken up the regolith – leading to cooling. Atmospheric CO2 was likewise taken up by regolith as the minor carbonate observed in meteorites. Breakdown of the
381
382
Regolith Science
greenhouse led to runaway global cooling and freezing of surface water. Surface sulfites were oxidised to sulfates – as observed by various Mars missions. 14.3.3 Surface features The surface of Mars can be divided into a number of physiographic regions (Greeley 1994). These include the southern highlands, the northern lowlands, large impact basins, volcanic plains and associated volcanoes, the Valles Marineris rift system, and the polar deposits (Hartman 2003). The hemispheric dichotomy of Mars into the northern lowlands and southern highlands is the largest and most ancient physiographic feature on the planet. Its origin is controversial, and the similarity to the lunar highland–lowland dichotomy is misleading. The two regions are not distinct in composition – with volcanic and sedimentary material in both areas. Although the lowlands are covered by Amazonian material, the presence there of large infilled impact basins suggests that ancient Martian crust continues beneath the lowland surface. The nature of the infill of the northern lowlands is also controversial. Some evidence exists for fossil shorelines along the highland–lowland boundary and it is possible that the northern lowlands may have been at least briefly filled by an ocean (Perron et al. 2007) – perhaps several times – and received sediment from the northerndraining valley systems. In other areas, the lowland plains appear to be composed of volcanic material. Ground ice is suspected beneath much of its surface (Mitrofanov et al. 2003). The large Noachian basins on Mars, like lunar basins, appear to be large impact structures. Hellas and Argyre are the two largest and best exposed, but other basins – partly to wholly infilled by the northern lowlands – are known, including Chryse and Isidus Planitia. The large impact basins, like the northern lowlands, appear to have acted as sinks for material transported down the valley systems and other sedimentary and volcanic material, albeit on a smaller scale. The volcanic plains and associated volcanoes range in age from Hesperian to Amazonian. The largest and most extensive of these are associated with the Tharsis region and include the edifice of Olympus Mons,
which rises 27 km above the surrounding plains. A slightly smaller volcanic region occurs on the other side of the planet at Elysium, and there are lesser areas elsewhere on Mars. The Valles Marineris rift system is large and complex tectonic structure of mostly Amazonian age that is more than 6 km below the surface of the surround plains in places and extends for more than 4000 km. Its formation may be related to the Tharsis volcanic province through stresses generated either by crustal loading by the Tharsis volcanics or by the supposed mantle plume that underlies the Tharsis volcanic province. The rift system has been extensively modified by collapse and sapping. Drainage systems from the Valles Marineris system form several large outflow valley systems into the northern lowlands, including Kasei and Simud Vallis. Neutron mapping from orbit has shown the presence of abundant hydrogen in the top metre of the regolith at latitudes higher than 60° north and south. This is interpreted as indicating polar water ice or H2O-CO2 clathrate, which is surrounded by, and partly built upon, thick successions of interlayered icy and dusty deposits and is consistent with observed medium- to small-scale geomorphic features indicative of periglacial processes (solifluction, polygonal fracturing and rock glaciers; Kargel 2004). 14.3.4 Martian regolith architecture The uppermost Martian regolith is composed of many different lithologies. Some of these are surface deposits of comparatively recent origin, including aeolian sand and silt deposits, colluvium and alluvium associated with small gully systems, and possible glacial deposits. Elsewhere, impact gardening forms a significant regolith blanket (Figure 14.4). However, unlike on the moon, even where present, the impact gardened regolith has undergone both erosion and burial – resulting in complex architectural relationships with surface sediments and underlying bedrock. Furthermore, the thin atmosphere on Mars reduces the impact velocity of smaller meteorites, so that they land intact, rather than with cosmic velocities. Several meteorites have been discovered by the Spirit and Opportunity rovers. As a result, the small craters that are characteristic of the lunar regolith are absent and
Extraterrestrial regolith
a
c
b
d
e
f
Figure 14.4: The surface of Mars. (a) Impact-gardened basaltic plain with aeolian drifts (Viking 1 image. (b) Deflated ejecta blanket from Mie crater, Viking 2 image. (c) Outwash deposit of Ares Vallis with possible boulder imbrication (part of Mars Pathfinder panorama released November 4, 1996). (d) Martian dunes, Meridiani Planum, Opportunity Pancam image 1P182165603EFF62JLP2393L2M1. (e) Large trough in the surface of Meridiani Planum (Anatolia site), along with array of wind ripples across deflated surface of residual lag (Opportunity Navcam image 1N134591492EFF1000P1803R0M1). (f) Spirit rover panorama released March 11th, 2004, showing Bonneville crater.
the degree of impact gardening is correspondingly less, although still important (Hartmann et al. 2001). Other deposits are much older and have been heavily dissected by erosion – especially wind action – but also by mass movement, spring sapping, and fluvial action, as well as by impacts. Whether these partly to wholly indurated units can be classed as regolith – as opposed to basin and valley-filling sediments and
major volcanic units – is as problematic on Mars as it is on Earth. Such formations occur in the oldest Martian terrains, making it unlikely that Mars has exposed primordial crust – at least not in the lunar sense. Early studies applied lunar concepts to Mars and postulated the presence of a mega regolith developed on the oldest crustal units of Mars (Cattermole 2001). However, the absence of clearly mappable primordial
383
384
Regolith Science
a
b
Figure 14.5: Martian megaregolith. (a) Impact-brecciated flat-lying evaporitic sediments, Meridiani Planum, imaged by the Opportunity Mars rover. NavCam image 1N181541865EFF62EIP1655R0M1. (b) Concentric and radial fracturing related to 86 km diameter impact structure being exhumed from under overlying sediments. MOC wide angle image E09-00186.
crust (despite the presence of the 4.5 Ga SNC meteorite ALH84001, thought to have originated on Mars) and the presence of complex volcanic–sedimentary units in the oldest Noachian terrains means that the characterisation of the Martian megaregolith is more difficult than on the Moon, although its properties are critical to the development of realistic Martian volatile inventories (Jankowski and Squyres 1993). Impact fracturing is locally evident at a range of scales and is especially well developed in sedimentary bedrock (Figure 14.5). 14.3.5 Current Martian surface processes Despite the low average temperatures and pressures (–63°C and 6–7 millibars), many small scale features of the Martian surface would be familiar to terrestrial regolith geologists (Thomas et al. 2005). The surface has gibber pavements, lags, rocky outcrops, dunes and aeolian mantles. Rocky surfaces show wind-polishing (ventifact textures) and dirt cracking, soft materials undergoing settling and slumping, and there are examples of tafoni, honeycomb, pachydermal and onion-skin weathering, and the spalling of rocks (Figure 14.6; see Appendix 1 for definitions). Dust
and frost deposits are built up and removed on a seasonal basis. At all five sites observed from the ground, the dominant ongoing processes are aeolian. Hard rocks at Chryse, Utopia, Ares Vallis and Gusev crater show ventifacts; soft rocks at Meridiani have been fretted by the wind. Deflation has resulted in the formation of deflation lags at Meridiani and gibber-armoured surfaces at Chryse and Utopia. Small, well-developed dunes are abundant at Meridiani and small drifts or dunes are locally present at Ares Vallis, Gusev crater and Chryse. None of the coarser deposits have been observed to move on at any of the lander sites; however, dust accumulates and is removed both seasonally and by local wind gusts and willy willies (dust devils; Greeley et al. 2006). Evidence of impact gardening is variable. The bedrock at Gusev crater and Chryse has been extensively fractured – apparently by impact gardening. In Meridiani, the bedrock is too soft to generate regolith breccias; however, the bedrock has been extensively brecciated by impact processes. The impact eject breccias and boulder-rich alluvial deposits at Utopia and Ares Vallis, respectively, are difficult to distinguish
Extraterrestrial regolith
a
e
d
g
c
b
h
f
i
Figure 14.6: Regolith features on the Martian surface. (a) Honeycomb weathering developed on possible volcaniclastic sediments at Husband Hill, Gusev crater (Spirit Pancam image 2P151574958EFF8987P2416R1M1). (b) Chemical weathering: eroded section though spheroidal weathering, near Missoula crater rim, Gusev landing site. Spirit Pancam image 2P135681301EFF3000P2387L7M1. (c) Slumping of aeolian sand triggered by Spirit rover, summit of Husband Hill, Spirit Navcam image 2N169692755EFFAAE0P0617R0M1 (d) Physical weathering: dirt cracking of sulfate-rich sediments at Meridiani Planum (Opportunity Pancam image 1P132540977EFF0600P2400R1M1). (e) Physical weathering: Disintegration of surface rock fragment at Gusev crater. (Spirit Pancam image 2P129908453EFF0500P2598L3M1). (f) Highly fretted laminae in fractured rocks on the rim of Eagle crater, Meridiani Planum (Opportunity Pancam image 1P131477488EFF0518P2392L5M1). (g) Deflated lag of resistant grains on the surface at Terra Meridiania (Opportunity image (h) Ventifact texture on boulder at top of Husband Hill, Gusev (Spirit Pancam image 2P178209533EFFAE03P2265L5M1). (i) Pachydermal weathering of boulder in Endurance crater, Meridiani (Opportunity Navcam image 1N150376755EFF3649P1968R0M1).
385
Regolith Science
9.5
10.6
0.14
10.9
Fe2O3%
9.5
0.27
Greyblack
17.0
MgO % 9.3
0.78
10.7
Fresh rock
18.4
K2O % 0.41
17.1
Brownochre
10.4
Al2O3 % 10.0
2100
Weathering rinds
19.2
Cr ppm 1600
Orangeochre
Sulfate
Rock surface
Olivine
Orangeochre
Zeolite
Soil
Silica
Colour
2000
Magnetite
Layer silicate
Pyroxene
Carbonate
Plagioclase
temperature changes which can exceed 70°C. Chemical weathering features are subtle, and include tafoni, honeycomb textures and pachydermal surfaces. Together these features point towards surface induration and the action of salt-crystal growth. This requires the presence of ephemeral moisture films. The same applies to the surface crusts developed on loose regolith. A Martian weathering profile on a basaltic substrate is shown in Figure 14.7, based on measurements by the Spirit rover on the plain basalts in Gusev crater (Gellert et al. 2004; Christensen et al. 2004). Such weathering promotes short range mobility of Al and K into the rock surfaces and depletion of Fe in weathering rinds (Figure 14.7) but, because the weathering is dominantly physical rather than chemical, the Martian soil is not enriched in Cr and Fe and depleted in Mg relative to underlying rock as occurs in the soils over terrestrial mafic rocks
2500
1 mm
5 - 20 cm Thickness
from impact-gardened regolith. At Ares Vallis, however, local preservation of apparent boulder imbrications indicates that some sedimentary fabrics have persisted and that impact gardening has not completely homogenised the regolith. Several mass-movement processes have been observed by the Mars landers during their observation, including spontaneous slumping of aeolian drifts and slumping of drifts triggered by the probes themselves. In other locations, open cracks and fissures point towards movement in the very recent past. Thermal expansion and contraction, which are associated with diurnal, seasonal and climatic temperature changes, may be the most likely explanation of these. Physical weathering features observed by Mars landers include fragmentation, spalling and exfoliation of rocks. The most likely cause of these is the diurnal
1-10 mm
386
KSf054-08
Figure 14.7: An idealised profile through weathered mafic rocks, Mars.
Extraterrestrial regolith
(Figure 6.6). This idealised Martian weathering profile is less than 25 cm thick, but the presence of spheroidal weathering of basalt at Gusev Crater (Thomas et al. 2005) shows that much greater weathering depths are locally present, even though their vertical and lateral extent remains to be established. 14.3.6 Rover observations of the Martian surface The most detailed studies of the Martian surface have been provided by the two Mars exploration rovers, Spirit and Opportunity, which landed at Gusev Crater and Meridiani Planum, respectively. Unlike previous Mars landers, which characterised single sites, the rovers have carried out traverses of several km (Wang et al. 2006). As of July 2007 Spirit had covered 7.1 km and Opportunity 11.3 km. These traverses provide the best data yet available on detailed regolith characteristics of the Martian surface. The Spirit rover landed on flat plains in Gusev Crater that were identified as being impact-modified basaltic plains with extensive aeolian deflation and minor deposition in ripples (Grant et al. 2004). Despite the modification, mappable bedrock and surficial units remain (Wang et al. 2006) – in particular the contact between the younger basalts and the older bedrock units of the Columbia Hills that rise up through the flows. Locally the basalts have undergone deep weathering and/or alteration (Thomas et al. 2005). The soils of the Columbia hills show signs of sub-surface precipitation of sulfates in the form of white powdery horizons (Crumpler et al. 2005). Dunes of basaltic sand occur in the lee of the Columbia Hills. The surface at Meridiani Planum was revealed by the Opportunity rover to consist of a lag of haematite granules and basaltic sand reworked into ripples and small dunes (Morris et al. 2006). The haematite was derived locally from the bedrock, where it occurs as concretions and vein fills; the basalt has been blown in from more distant regions and no specific source has been identified. The lag variably covers over sulfate-rich sedimentary bedrock (Squyres et al. 2006) and the entire landscape has complex history of burial and exhumation (Arvidson et al. 2006) that extends from the present back to the Noachian. The detailed surface observations at Gusev and Meridiani have a range of implications for remote
sensing of the Martian surface (Golombek et al. 2005). While pre-landing interpretations correctly predicted the physical properties and roughness of the surface, they failed to correctly identify the formative processes at each site. Arvidson et al. (2003) incorrectly interpreted Meridiani Planum as a series of volcanic deposits when in reality they have proved to be sedimentary (Grotzinger et al. 2005). Grin and Carbol (1997) interpreted Gusev Crater as a former lake with exposed sediments, when the surface is in fact largely of flood basalts (Wang et al. 2006). Leshin et al. (2004) argued spectroscopic mineralogy proved a better predictor for the actual nature of the surface than geomorphology. This reflects the immaturity of our understanding of the processes that shape the Martian landscape. It is important to note that, in spite of the success of remote sensing in identifying Gusev and Meridiani as a sites of interest, it was not able to predict the complexities of the sedimentary and volcanic architecture or regolith. In situ exploration resulted in revision of all previous hypotheses – highlighting the importance of detailed field studies (Leshin et al. 2004).
14.4 VENUS 14.4.1 Introduction Venus in size and mass is very similar to the Earth – with an equatorial diameter 94.9% of Earth and a mass of 81.5% of Earth. However, if Mars is regarded as an intermediate evolutionary stage between the Moon and the earth, then Venus presents an alternate evolutionary history – as in almost every respect Venus’s atmosphere, surface and planetary evolutionary history is radically different (Taylor 2001). The surface of Venus is obscured by high-altitude clouds of sulfuric acid droplets. The atmosphere is composed almost entirely of CO2, and the average surface pressure is about 60 bars – equivalent to an ocean depth on Earth of about 600 m. As a result of the dense CO2 atmosphere, the greenhouse effect is extreme and surface temperatures average about 400°C. There is a considerable range, however, with pressures and temperatures of 41 bars and 374°C and 96 bars 465°C typical of the highlands and lowlands, respectively. This range means that impact, volcanic, aeolian and weathering processes may operate very
387
388
Regolith Science
differently in the highlands and lowlands (Greeley 1994). Conversely, diurnal variations from the slow, retrograde rotation of the planet are minimal. A total of 14 probes have successfully reached the surface of Venus, with nine of these transmitting back data on the surface composition and four returned images of the surface. Because of the extreme conditions, none has lasted for much more than two hours before succumbing to the heat, pressure and corrosive atmosphere. However, these probes – together with orbiter radar missions that have mapped the surface (Pioneer Venus 1, Venera 15 and 16, Magellan) – provide some information on the surficial morphology and geological processes on Venus. 14.4.2 Surface processes Radar mapping and imaging from orbit (Figure 14.8) has shown that the surface of Venus is dominated by the internal forces of tectonics and volcanism (Greeley 1994). Impact craters, although present, are comparatively few – suggesting a cratering age for the surface of about 500 Ma (Phillips et al. 1992). This has prompted speculation that the planet may have undergone a global volcanic resurfacing event at the time. There is also a dearth of craters less than 5 km in diameter, which is attributed to the shielding effect of the deep, dense atmosphere (Taylor 2001). Impact ejecta deposits show very complex morphology, indicating that there has not only been ballistic emplacement of deposits but also complex base surge and pyroclastic flow deposition associated with impact events. In some cases impacts may have triggered volcanic flows. The volcanic-like processes associated with impact craters may be a reflection of the dense atmosphere and high surface temperature. Venus does not appear to have plate tectonics – at least not in the terrestrial sense. The circular coronae appear to be sites of possible mantle plume activity, and the tessarae appear to have undergone orthogonal fracturing. Large rift sequences similar in scale to the African rift valley occur between the Beta and Phoebe region areas. Ridged and grooved terrain indicate zones of crustal shortening, which are structurally (though perhaps not tectonically) similar to terrestrial fold belts (Greeley 1994). As a result of the different global tectonic system, fractionation of igneous rocks
has occurred somewhat differently to those on Earth (Kargel et al. 1993). Venus has a wide diversity of volcanic features, including very large shield volcanoes, plains and calderas of all sizes, along with smaller domes, lava pancakes, cones and fissures. Lava flows hundreds of kilometres in length have been observed – leading to speculation of komatiitic or carbonatitic composition (Greeley 1994). Some of these channels show signs of thermal erosion of their substrate (Oshigami et al. 2007), suggesting komatiitic compositions. All but one of the nine sites landing sites with surface chemistry information have, however, proved to be basaltic in composition – some enriched in alkali metals. Venera 8 reported high K, U and Th levels similar to terrestrial granites from the Hathor Mons region. These could be explained either as a fractionated alkaline basalt or by the presence of a rhyolitic lava dome (Taylor 2001). Secular variations in atmospheric SO2 suggest that volcanism is ongoing, although no volcanic eruption has been observed by any spacecraft. The corrosive atmosphere and high temperatures and pressures suggest that chemical weathering may be significant on Venus (McGill et al. 1983; Nozette and Lewis 1982). Furthermore, the proposed weathering reactions are often reversible over the temperature ranges experienced between the highest and lowest elevations of the planet. These reactions resemble more closely terrestrial thermal metamorphic processes than they do terrestrial weathering. Examples include: wollastonite + CO 2 = calcite + quartz (Eqn 14.1) forsterite + CO 2 = magnesite + enstatite (Eqn 14.2) 2 forsterite + diopside + 2CO 2 = 2 dolomite + 4 enstatite (Eqn 14.3) calcite + enstatite + CO 2 = 2 dolomite + quartz (Eqn 14.4) Trace gases in the atmosphere of Venus include H2O, HCl and HF. The following reactions may occur therefore occur:
Extraterrestrial regolith
a
b
c d
Figure 14.8: Magellan radar images of the surface of Venus. (a) Bright wind streaks associated with a 5 km-wide volcano at the western end of Parga Chasma (JPL press image P-38810). (b) Stacked lava flows forming volcanic edifice 100 km across and 1 km high superimposed on faulted terrain. Volcano lies between Artemis Chasma and Imdr Regio (JPL press image P-39916). (c) Sinuous lava channel (2 km wide and 200 km long), 40oS 273oE (JPL press release image P-39226). (d) Complex ejecta flows from the 69 km diameter crater Dickinson (JPL press release image P-39716).
2 diopside + 3 enstatite + quartz + H 2 O = tremolite + 2O 2 (Eqn 14.5) 2 anorthite + 5 enstatite + quartz + H 2 O = tremolite + 6 andalusite (Eqn 14.6) orthoclase + 2 enstatite + 2HF = phlogopite + 3 quartz + H 2 O (Eqn 14.7) 2 albite + 2HCl = 2 halite + andalusite + 5 quartz + H 2 O (Eqn 14.8)
Despite this, images from four locations (Figure 14.9) on the surface of Venus show no sign of chemical weathering. Whether this is because weathering proceeds more slowly than expected, or the four sites are fortuitously on young surfaces, or because the weathering does not produce mesoscopic degradation features familiar to terrestrial scientists, is not clear. The presence of loose regolith and angular rock surfaces in most of the images shows that some physical weathering and fragmentation is occurring even at these sites. The processes driving the fragmentation are not clear. Given the volcanic composition of the bedrock, some may be due to primary volcanic
389
390
Regolith Science
a
b
c
d
Figure 14.9: Regolith of Venus. (a) Venera 9 landing site showing angular layered gibbers and coarse interstitial regolith. (b) Venera 10 landing site showing slabby rocky outcrops and patches of dark fine-grained regolith. (c) Venera 13 landing site showing slabby rocky outcrops and patches of dark, fine to medium-grained regolith. Note the higher relief area near horizon in top left hand corner. (d) Venera 14 landing site, showing smooth plain of slabby, fractured rock and minimal fine regolith. All images 180º panoramas using semi-circular imaging scanner. Images reprocessed by Don P Mitchell, and used by permission.
processes. Exfoliation is unlikely, given the minimal diurnal change in temperature. One possibility is that the chemical reactions above may assist mechanical breakdown of the rock. Measured wind velocities on Venus are low (<2 m/s). These are sufficient to move sand particles. Radar observations indicate loose materials appear to comprise only a quarter of the surface, which is consistent with surface imagery from landers. Despite this, scattered dune fields have been observed, and streaks of bright material down-wind of craters are
not uncommon. Localised mass movement has been observed on the steeper slopes (Greeley 1994). In addition to limited sand supply, it is possible that radar look-angle effects may result in the under-representation of dunes in the radar images (Weitz et al. 1994). Another consequence of the very high surface temperatures and pressures on Venus is the inferred presence of metallic snow or frost in the highlands. Radar mapping by the Magellan spacecraft has shown that at altitudes at greater than 3 km above datum, rock surfaces appear coated with a material of low emissivity
Extraterrestrial regolith
resembled that of Venera 10, except the fine regolith was only slightly darker than the bedrock but more abundant. The higher resolution images returned by these missions showed that the loose regolith was composed of very poorly sorted angular fragments. Venera 14 imaged a flat plain almost completely covered by thin layers of slabby bedrock. Fractures were very irregular – almost scalloped – and there were only very minor amounts of loose regolith. Areas with looser regolith appear in the distance.
and high dielectric constant. This is consistent with the raining or freezing out of a range of heavy metal sulfides and sulfo-salts: probably galena and/or bismuthite. These minerals have been reported from sulfur-rich fumaroles on Earth (Schaefer and Fegley 2004). If weathering on Venus resembles terrestrial high temperature metamorphism, the condensable volatiles of Venus behave like magmatic gas vents on Earth. All four landers that have imaged the surface to date have unfortunately touched down at too low an elevation to observe this phenomenon, whose significance was not realised until after the conclusion of the Venus lander missions of the 1970s and 1980s.
14.5 TITAN
14.4.3 Surface imagery Black and white images of the surface of Venus were obtained by Venera 9 and 10 (Florenskiy et al. 1983) and higher resolution colour images were obtained by Venera 13 and 14 (Moroz 1983). These show that most of the surface consists of rocky material, with loose regolith in the minority (Figure 14.9). These observations are consistent with the radar data that shows that only a quarter of the surface consists of loose material (Greeley 1994). The Venera 11 and 12 landers failed because of the intense heat, although other instruments functioned as planned at their landing sites at Dzerassa Planita. The Vega 1 and 2 landers carried no cameras as they landed on the night side of the planet, on Aphrodite Terra. Venera 9 landed on the eastern flanks of the highlands of Rhea Mons and showed a surface consisting of scattered slightly flattened to equi-dimensional, angular to sub-angular rocks. Possible layering was seen in several rocks. The rocks were both resting on, or partly buried by, a poorly sorted granular regolith. The landing site for Venera 10 was to the south of Venera 9, on the south-western flanks of Theia Mons. The site proved to consist of large flat slabs of rock – almost flush with the surface – separated by patches of very dark granular regolith. The slabs showed wavy fracture patterns. Textures similar to terrestrial sandy corrosion and cellular weather can be observed (Florenskiy et al. 1977). The Venera 13 and 14 probes touched down on the eastern flanks of the highlands of Phoebe Regio, with Venera 14 to the east of Venera 14. The Venera 13 image
14.5.1 Introduction Titan is the largest satellite of Saturn, and one of the largest in the solar system. Only Jupiter’s satellite Ganymede is larger. Titan is larger than the planet Mercury and its diameter is 74% of Mars. Composed of an inferred sepentinite core overlain by icy mantle and crust (Fortes et al. 2007), the density of Titan is, however, much lower – resulting in a surface gravity only 14% of Earth’s. Despite this, Titan has a very dense atmosphere, with a surface pressure of 1.46 bars. The atmosphere is very cold, with an average temperature of –280°C. The atmosphere is composed of 98.4% nitrogen and 1.6% methane; it contains few clouds, but a thick haze layer. This layer almost completely obscures the surface at visible wavelengths, but less so at infra-red wavelengths. The haze is not present below altitudes of 40 km (Tomasko et al. 2005). Like that of Venus, the atmosphere is completely transparent to radio frequencies. The surface temperatures on Titan are such that methane-ethane can condense and fall as rain, acting in a manner analogous to water on earth (Tokano et al. 2006). Knowledge of Titan is in a state of considerable flux at the time of writing (2007–2008). Little has been deduced about the nature of the satellite from terrestrial observations and early planetary flybys were brief. Since 2004, the Cassini spacecraft in orbit round Saturn has experience multiple encounters with Titan. Infra-red imaging and spectroscopy has provided detailed information on the composition of the surface (for example, Soderblom et al. 2007a) and radar on the morphology (for example, Elachi et
391
392
Regolith Science
al. 2006). However, only a small part of the surface of Titan has been imaged in any detail and many unexpected features may well emerge in the course of this mission. Flyby data on the surface of Titan us supplemented by a single lander. The Huygens probe was released by Cassini during its first encounter with Titan and provided detailed in situ data on atmospheric properties and surface composition (Lebreton et al. 2005). It also imaged the surface at altitudes of several kilometres as it descended beneath its parachute. The final images returned by the Huygens probe were of the surface itself. 14.5.2 Remote sensing data Synthetic aperture radar studies of the surface of Titan by the Cassini mission have revealed a wide range of landforms (Figure 14.10). These included highland areas showing evidence of cryovolcanism (where volcanic features are constructed by eruption of water-based magmas) and dissected by valley systems that debouche onto lowland areas marked by fans. Lowland areas on Titan are characterised by dune seas and in high latitudes by lakes. Fans, channels and lakes are inferred to be the result of a surface liquid composed of ethane and methane. Cryovolcanic features consist of bright lobate flows (Elachi et al. 2005) and possible pancake domes. The magma consists of liquid water – probably with a high concentration of ammonium sulfate (Fortes et al. 2006). Geophysical data indicates that the source of this magma is a water ocean 153 km deep occurring between a water ice-methane clathrate crust. The ascending magma entrains methane from the crust, resulting in explosive volcanism at the surface. This provides a source for both the atmospheric methane and abundant fragmental ice material for transport by liquid methane and ethane on the surface. Atmospheric modelling shows that methane and ethane form a ‘hydrologic’ cycle on Titan (Tokano et al. 2006) with methane-ethane rain, run-off, lakes and as sub-surface floor fluid. The run-off has formed a range of features including depositional fans (Paganelli et al. 2005) and a range of channel features (Stofan et al. 2007). Over much of the surface, methane-ethane liquids are
predicted to evaporate rapidly, resulting in a desert landscape. In the polar regions, the temperature is cold enough for liquid methane and ethane to persist, forming large and small lakes (Stofan et al. 2007; Mitri et al. 2007). The lakes occupy craters and other depressions in the surface; the largest lakes have complex, ria-like coastlines, indicating drowning of pre-existing fluvial surfaces and limited coastal sedimentation. Some rivers have, however, built up small deltas. Large seas of longitudinal dunes were among the first landforms readily identified in the radar images. These features – comparable in scale to the longitudinal dunes of central Australia – dominate the low lying equatorial regions of Titan. The winds are thought to be tidal in nature, driven by the orbit of Titan round Saturn. The dunes indicate an abundance of sand-sized particles, winds of the order of at least 0.5 m/s, and the absence of persistent, free-standing liquid in the equatorial regions that would modify the dunes into coastal and sub-liquid deposits (Stofan et al. 2007; Elachi et al. 2006; Lorenz et al. 2006). As a result of cryovolcanism erosion and deposition by wind and flowing methane-ethane, the surface of Titan has extensively modified over time. Titan’s surface has few impact craters (Stofan et al. 2007) and, in this respect, resembles Venus and the Earth – both of which have relatively young surfaces. Both radar and infra-red measurements provide information on the surface materials (Stofan et al. 2006). The brightness of the radar reflections corresponds to surface roughness. Bright surfaces represent surfaces covered by pebble-sized and coarser materials, and sand-sized surface surfaces are darker. Highland areas are mostly bright in radar, and cannel fills can be either bright or dark, while sand dunes are always dark. The darkest surfaces are the surfaces of the lakes. Spectrometry from instruments on the Cassini spacecraft during the encounters with Titan have identified a number of different compositional units (Soderblom et al. 2007a; Porco et al. 2005). These have indicated that surface consists mostly of light coloured ‘blue’ highlands and darker ‘brown’ lowlands. The blue units have spectra that resembles contaminated water ice, while the brown units contain relatively little ice. They are interpreted to be either
393
Extraterrestrial regolith
a
c
b
d
Figure 14.10: The surface of Titan during flyby. (a) Probable methane/ethane lakes (dark) surrounded by uplands (bright) on Titan, note ria-like indented shoreline. Cassini radar image PIA09180, image width ~300 km. (b) Radar-dark channels (possibly filled by flowing methane/ethane) with delta features along shore of probable hydrocarbon lake. Cassini radar image PIA01942, image width ~200 km. (c) Radar-dark longitudinal dunes adjacent to radar-bright highlands. Cassini radar image PIA0918, image width ~400 km. (d) Radar-bright channel sediments and outwash plain. Cassini radar image PIA07366, image width ~350 km. Images NASA/ESA/University of Arizona.
composed of dark-reddish organic material (tholins) and carbon-nitrogen compounds (nitriles), or of contaminated water ice with tholin and nitrile coatings. The tholins and nitriles are believed to have been formed by photochemical reactions in the upper atmosphere of Titan. 14.5.3 Surface data During its descent through the atmosphere of Titan, the Huygens probe carried out a wide range of investigations into the nature of the atmosphere and surface (Lebreton et al. 2005). Most relevant to understanding the nature of the surface were the images taken after the probe descended below the lower haze layer at about 40 km (Tomasko et al. 2005). These images (Figure 14.11) showed a remarkable landscape, which is broadly consistent with what was determined from spacecraft images during encounter. These included
bright highlands and dark lowlands, with an abrupt contact between the two. The highlands were extremely rugged, with slopes of up to 30°. The lowlands were much smoother and almost flat (Soderblom et al. 2007b). Two distinct types of drainage were seen: one with short, stubby tributaries and low stream order resembled those formed by sapping on Earth and many Martian channels; the other had many long tributaries and formed a high-order system (Soderblom et al. 2007b). These features matched those observed at lower resolution by the Cassini mission during its encounter and allowed the Huygens landing site to be pin-pointed. The Huygens lander touched down on a smooth dark area composed of rounded pebbles of water ice 5–10 cm across (Tomasko et al. 2005; Soderblom et al. 2007b). These were on a low-relief, channel surface with distinct sediment tails behind many of the
394
Regolith Science
a
1 km
b
Figure 14.11: The surface of Titan as seen by the Huygens lander. (a) Two types of channels formed by flowing methane/ ethane, composite image from an altitude of about 40 km, varying resolution due to atmospheric haze and different altitude of composite components. Channels cut through optical bright terrain and filled with optically dark material and terminate in an optically dark plain. Stubby channels (left) possibly cut by sapping methane/ethane, contrast with dendritic channels that are possibly formed by run-off of methane/ethane rain. (b) Ground view from Huygens lander showing optically bright centimetre-scale pebbles on optically dark depositional plain. Note the tails of fine-grained sediment to pebbles in the distance and a low-relief ridge on the skyline.
pebbles in the middle distance (Figure 14.11). Penetrometry and accelerometry measurements of the surface the probe landed on revealed that it was neither hard nor very compressible– instead it was relatively soft. The surface properties were analogous to wet clay, lightly packed snow and wet or dry sand. The probe settled a few mm after landing, and other sensors indicated significant outgassing: possibly from the boiling of liquid methane heated by the lander’s lamp (Zarnecki et al. 2005). These results are consistent with the landing site being an alluvial lowland floodout.
14.6
ASTEROIDS
14.6.1 Introduction A number of asteroids have been imaged during flyby missions – including Gaspra, Ida and its moon, Dactyl – by the Galileo spacecraft (1991–1993), Mathilde by the NEAR-Shoemaker spacecraft (1997), Braille visited by Deep Space 1 in 1999, and Annefrank by the Stardust probe in 2002. Only two asteroids, however, have had landing attempts on them: the Mars-crossing asteroid, Eros, and the main belt asteroid, Itokawa.
14.6.2 Eros Eros (Figure 14. 12) is a small, highly irregular (34 × 11 km) near-Earth asteroid that was visited by the NEAR-Shoemaker spacecraft in 2000–2001. A series of close approaches culminated in a crash landing. Eros is an S-type asteroid, which is typical of many found in the inner asteroid belt (Veverka et al. 2000), whose surface is composed of pyroxene, olivine and metallic iron – similar to ordinary chondritic meteorites. The NEAR-Shoemaker mission showed that Eros is a solid body whose surface is characterised by well-formed impact craters a series of groves and lineations thought to be related to impact fracturing. Softening of crater outlines and scattered boulders is strongly suggestive of a fragmental regolith. One unexpected discovery from the mission is that, while most the surface is saturated with craters – indicating an ancient surface – the final image, taken when the probe was only a few metres above the surface. (Veverka et al. 2001), shows a paucity of small craters and contrasting rough and smooth areas. The smooth areas show small pits and sinuous furrows, whose origin is unknown. Despite the small size of Eros, its regolith is clearly not a simple passive blanket, but a complex and evolving unit.
Extraterrestrial regolith
a
b c
d
Figure 14.12: (a)The surface of Eros. Southern hemisphere of Eros, landing site in shadowed area, just left of centre. (Mosaic of images 0151025658, 0151025720, 0151025782, 0151025844, 0151025906, 0151025968, 0151026030, 0151026092, 0151026154, 0151026216, 0151026278 and 0151026340). (b) Eros surface, showing finer-grained crater bottom regolith. Image width 1 km (image 0155768823). (c) Surface of Eros showing an area 33 m across (image 0157416593). (d) Final image of the surface of Eros. Image width 6 m. Note the 4 m wide rock, the differentiation of the surface into areas of coarser and finer material, and the sinuous depression. Lines in the lower half represent a loss of signal on impact (image 0157417198).
14.6.3 Itokawa Asteroid Itokawa (Figure 14.13) provides a complete contrast to Eros. This tiny, sea-otter shaped asteroid is only 300 × 500 m across. Itokawa was visited by the Hayabusa probe in 2005. During its visit, the probe carried out a soft landing and attempted to collect a sample. The success of the collection attempt will not be known until the return capsule lands at Woomera in South Australia in 2010. Like Eros, Itokawa is an S-type asteroid (Fujiwara et al. 2006); however, unlike Eros, this tiny world lacks a solid structure and appears to be composed entirely of a loose aggregation of rubble. The entire asteroid is made up of regolith ranging from sub-millimetre dust up to boulders tens of metres across. The density of Itokawa is 1.3g/cm3, which – given the density of an ordinary chondrite is 3.2g/cm3 – means that the
asteroid must consist of 41% open space. Despite the small size, the surface of Itokawa is geomorphically differentiated: with rocky areas and a smooth plain known as the Muses Sea. The closest images of the surface (Yano et al. 2006) of the Muses Sea show that the surface consists of sorted centimetre- to millimetre-sized grains. Apparent landslide deposits are observed on the surface on the steepest slopes. The two lobes of Itokawa that form the ‘head’ and ‘body of the sea otter suggest that the asteroid may have formed from the coalescence of two smaller bodies during a low speed collision or the merging of a contact binary asteroid. The surface of Itokawa shows very few craters, which is: partly a reflection of the rubble pile nature of the surface being a poor preserver of craters; partly by the destruction of older craters through seismic activity caused by later
395
396
Regolith Science
a
b
c
d Figure 14.13: The surface of Itokawa, imaged by the JAXA Hayabusa mission. Itokawa has a maximum and minimum diameter of about 500 and about 300 m, respectively. (a) side view of Itokawa, showing the ‘sea otter’ shape (ISAS/JAXA release 051101-1). (b). Large (~10 m) boulders on the ‘neck’ and ‘upper back’ of Itokawa (ISAS/JAXA release 051101-3). (c) Shadow of Hayabusa during approach ((ISAS/JAXA release 051110-6.1). (d) Close up of the surface of Itokawa, mosaic imaged from an altitude to 63 to 68 m, white lines represent 1 m. Modified from Yano et al. (2006).
impacts; and partly because the body itself may have coalesced quite recently. Like the surface of Eros, that of Itokawa shows a very complex and active regolith history, with ongoing movement of particles (Miyamoto et al. 2007) that is completely different from that of Eros.
14.7 MAPPING EXTRATERRESTRIAL REGOLITH A wide range of mapping methodologies has been applied to planetary surfaces, such as geologic, soil, physiographic (geomorphic) and surficial mapping Greeley (1994). Integrated regolith landform mapping, as refined by CRC LEME (Chapter 11), has not been generally applied to extraterrestrial landscapes, even though it is eminently applicable. Indeed, regolith landform mapping is largely unknown to most plan-
etary researchers, who tend to map landforms and regolith materials separately. A trial of the application of regolith–landform mapping to Mars imagery was carried out by Clarke and Pain (2004). Both satellite (Figure 14.14) and ground images were interpreted, with readily applicable results. The greatest handicap for the interpretation of satellite imagery of other planets is the lack of information on the materials that make up the composition of the map units (see Section 14.3.6). However, the increasing number of surface observations from lander sites and the availability of multispectral systems and systems collecting data from radar, infra-red, X-ray and gamma ray wavelengths, as well as information on neutron emission, assists greatly in the interpretation of regolith composition. Images of planetary surfaces are widely available off the internet from US and European space missions.
Extraterrestrial regolith
2
1
2
4
3
7
4
4 4
6
5
5
5
4
3
4 6
8 8
8 9
8 1 km Figure 14.14: Example of regolith–landform mapping on Mars, base MOC image located 42ºS, 158ºW within Newton Basin, Sirenum Terra. Unit 1: A smooth plain, probably of finer-grained material. Unit 2: An irregular plain, probably of loose regolith, developed on the same material as Unit 3. Unit 3: Steep rough slopes in layered sediments. Steep valleys have been incised into this unit. Unit 4: Smooth slopes covered in fine-grained material. Locally the underlying material shows through. Unit 5: Steep, densely gullied slopes developed on a different stratigraphic succession to unit 3. Morphology suggests more weathered bedrock. Unit 6: Colluvial fans with incised channels on the upper parts and moderately rough surfaces at the lower ends. Unit 7: Dune field of slightly arcuate dune, aeolian sand. Unit 8: Bedrock outcrops. Unit 9: Dune field with low transverse dunes of aeolian sand. Modified from Clarke and Pain (2004
The number of images greatly exceeds the number of researchers and interpreting them will provide a fruitful field of research for decades to come. However, access to raw data as opposed to processed images, is restricted for researchers outside the source nations and, in the case of the United States, generally requires software not generally available outside the US (N Jackson, pers. comm. 2006).
14.8 REGOLITH PROCESSES ON ROCKY AND ICY BODIES 14.8.1 The importance of liquid water The extent of liquid water activity has a significant impact on the evolution of a regolith. Table 14.1 compares the rock and soil minerals found in anhydrous basalts on the Earth, Mars and the Moon. On Earth, where there is a continuously active hydrosphere, the basalt minerals are largely to completely converted to clays (especially kaolinite), oxides and (under arid conditions) sulfates and carbonates. On Mars, where the hydrosphere may have been significant early in the planet’s history, but it is now only intermittently and locally active, primary minerals are much more common in the regolith – the main secondary minerals are smectites and sulfates (see also Figure 14.7). On the Moon, which has no chemical weathering, the regolith is composed entirely of the minerals from the host rock. 14.8.2 Comparison of rocky and icy bodies In many respects the terrestrial environment is very different to that encountered on other bodies of the solar system. Important differences include:
Table 14.1: Mineralogy of basalts on Earth, Mars and the Moon. Minerals
Earth
Mars
Moon
Rock Olivine
xxx
xxx
xxx
Plagioclase
xxx
xxx
xxx
Pyroxenes
xx
xx
xx
Oxides
x
x
x
tr
xx
xx
Soil Olivine Plagioclase
x
xx
xxx
Pyroxenes
tr
x
xx
Kaolinite
xxx
x
Smectites
xxx
xx
Oxides
xxx
xx
Sulfates
xx
xxx
Carbonates
xx
tr
Silica
xxx
x
Note: xxx = abundant; xx = common; x = minor; tr = trace
x
397
398
Regolith Science
s
s s
s
s
s
the presence of plate tectonics for much or all of Earth history as a driver for the lithologic cycle and for cycling of volatiles through the crust and upper mantle the presence of continental crust, which is found nowhere else in the solar system and which may be a result of the presence of plate tectonics the presence of a continuously active hydrosphere through geologic history, unlike on Mars where the hydrosphere was active during the early history and has subsequently become frozen, with occasional interruptions through impact or volcanic heating the presence of an oxygen-releasing biosphere for most of Earth history, which has resulted in an oxidising atmosphere and extensive crustal sequestration of carbon. Organisms bioturbate the regolith and greatly accelerate many physical and chemical weathering processes an oxidising atmosphere and ocean being present for much of the planet’s geologic history, with a correspondingly major influence on composition of the crust large-scale sequestration of atmospheric CO2 carbon in the form of reduced organic carbon and as oxidised carbonate rocks through the activity of the biosphere.
Despite these differences, there are also many similarities. These include:
s s
s s s
mass movement and impact, which are common processes to the surface of all planets and moons with solid surfaces volcanism (not always silicate in composition), which occurs on all planet and moons with a solid surface that have undergone internal differentiation aeolian processes, which are found on all solid moons and planets with an atmosphere Evidence of liquid processes, which is common to all moons and planets with condensable volatiles on the surface and in the sub-surface Glacial and periglacial processes, which are present on rocky bodies with surface and sub-surface ice deposits
A comparison of the expression these processes on the surfaces of the Moon, Mars, Earth, Venus and Titan is shown in Table 14.2.
14.9 TERRESTRIAL ANALOGUES OF EXTRATERRESTRIAL REGOLITH 14.9.1 What is analogue research? Analogue research is investigations of terrestrial features that may provide analogues to extraterrestrial counterparts. In addition to comparisons, terrestrial analogues may also provide contrasts – illustrating how divergent superficially similar processes may be on other bodies in the solar system compared with on Earth. Most analogue research has focused on comparisons between Earth and Mars because of many similar processes that operate or are believed to have operated on the two planets as the action of both ice and liquid water, physical and chemical weathering and wind. Analogue research can also include investigation of extremophile habitats to constrain the possible locations for life elsewhere in the universe. This section will focus on analogue research that has been carried in Australia to illustrate the scope of research that is possible and the opportunities that exist. The reader is urged to investigate the cited literature for more details of this research. 14.9.2 General studies The Australian landscape has a diversity of impact craters: ranging in size from the smallest 6 m diameter crater at Henbury in the NT to the Acraman astrobleme in SA, which may be 90 km across. These craters range in age from a few thousand years (Henbury) to Mesoproterozoic (Teague Ring, WA). Target rocks vary from unconsolidated sediments (Yallallie, WA) to metamorphics (Darwin crater, Tasmania) and volcanics (Acraman). There is an extensive literature on impacts and astroblemes, which extends over more than 70 years: the two most recent and most important compendia being those of Glikson (1996) and Glikson and Haines (2005). These craters provide important insights not only into the consequences of terrestrial bombardment by asteroids and comets, but also more general clues about the
Moon
Exfoliation Micrometeorite pitting Agglutination UV crystal damage Cosmic ray crystal damage
Sputtering Solar wind volatile trapping Cometary volatile trapping
Not present
Impacts Mass movement
Impact ejecta Mass movement
Physical weathering
Chemical weathering
Biological weathering
Erosion
Deposition
Impact ejecta Mass movement Volcanism Wind Fluvial Solifluction
Impacts Mass movement Wind Fluvial Glacial Solifluction Groundwater processes Coastal processes?
Impact ejecta Mass movement Volcanism Wind Fluvial Solifluction Coastal processes Submarine/ sublacustrine processes
(Impacts) Mass movement Wind Fluvial processes Glacial Solifluction Groundwater processes Coastal processes
Acceleration of physical and chemical weathering Release of organic acids Burial of carbon Bioturbation
Solution and precipitation Low temperature hydration– dehydration Oxidation Fe oxidation/ hydrolysis Hydrolysis Carbonation
Solution and precipitation Low temperature hydration–dehydration Oxidation Fe oxidation/ hydrolysis Not known
Exfoliation Salt weathering Frost shattering
Earth
Exfoliation Salt weathering Frost shattering
Mars
Impact ejecta Mass movement Volcanism Wind
(Impacts) Mass movement Wind Lava (thermal erosion)
Not present
High temperature carbonation– decarbonation High temperature hydration–dehydration Gaseous mineral acid weathering
Not known
Venus
Regolith processes on the Moon, Mars, Earth, Venus and Titan. (Brackets indicate minor processes.)
Process
Table 14.2:
Impact ejecta Mass movement Cryo-volcanism Wind Methane-fluvial Methane-coastal processes Sub-methane processes
(Impacts) Mass movement Wind Methane-fluvial processes Ground-methane processes Methane-coastal processes Water melts (thermal erosion)
Not present
Solution and precipitation of methane-soluble organics Organic polymerisation
Not known
Titan
Extraterrestrial regolith 399
400
Regolith Science
impact process. In particular they provide insight into the post-impact evolution of craters, and how they are eroded, infilled, buried and exhumed. This is especially relevant to the study of crater evolution on other bodies with atmospheres, such as Mars and Venus, and those with past or present surface liquids, such as Mars. Specific examples of study into Australian analogues to extraterrestrial surface processes conclude the chapter. 14.9.3 Moon The unusually long Quaternary basaltic flows and associated lava tubes at Undara, northern Queensland, have been used as analogues for lunar lava flows, collapse features and fills (Steven and Atkinson 1972). Other Australian features that have been used as a lunar analogue are the Henbury crater group, which was formed by the low altitude break-up of a large iron meteorite. Because of the extreme youth of the craters, one preserves a ray structure (Milton and Michel 1965). As the only reported example of a terrestrial rayed crater, the Henbury craters provided a unique opportunity to examine the nature of such features. 14.9.4 Mars Because Mars has, in many respects, the most Earthlike of environmental conditions, the greatest potential for terrestrial planetary analogues is for Mars. Investigation of Australian–Mars analogues is at a very early stage: most published research has served to draw the initial parallel, with very little follow-on investigation. These studies are valuable not only for their extraterrestrial significance but also because they enable investigation of specific areas of terrestrial regolith that might not otherwise be studied. Proposed analogues include a range of aeolian deposits (Greeley and Williams 1994) and dunes (Bishop 1999). Central Australian paleoflood deposits were proposed as a Mars analogue by Bourke and Zimbelman (2000, 2001) and Dalhousie springs by Clarke and Stoker (2003), Nelson et al. (2007), Bourke et al. (2007) and Clarke et al. (2007). Acid salt lakes are a common feature of the more arid environments of southern Australia – in particular on the southern Yilgarn Craton, Eyre Peninsula and in the Wimmera region of Victoria (Chapter 12). Acid saline lakes are, however, rare elsewhere in the world and have been
proposed as an analogue for the arid evaporite deposits of Meridiani Planum. Initial research on these features has been started by Benison and Bowen (2006) and in Australia is probably the best location in the world to study these lakes. The diagenetic history of channel ironstones in the Pilbara region has also been proposed as a possible Mars analogue (Heim et al. 2006). Relief inversion is a common feature of terrestrial landscape evolution (Pain and Ollier 1995) and is also present in Mars (Pain et al. 2007). Lastly, in areas of minimum weathering, such as the Mount Painter (SA) and the Pilbara (WA) regions, hyperspectral studies of alteration associated with fossil hydrothermal systems have been carried out with the aim of developing recognition criteria for such features on Mars because of the their significance as possible habitats for past Martian life (Brown et al. 2005; Thomas and Walter 2002). Lastly, in addition to their scientific value, sites that provide analogues to the Moon and Mars can serve as training and test areas for instruments and methods that may one day be used on actual missions. Logistical constraints mean that areas that combine a diversity of analogue features are more attractive than those with one or two. One such area that has been identified is the Arkaroola region in the northern Flinders Ranges (SA) and the adjacent areas (Mann et al. 2004, Clarke et al. 2004). Research into the documenting the regolith geology of this area to provide the background for analogue research has just begun (Waclawik and Gostin 2006), infilling on the regional study of Gibson (1999). 14.9.5 Titan The hot deserts of Australia might, at first, seem unlikely analogues for the icy surface of Titan. However, the extensive linear dune fields of central Australia (for example, Nanson et al. 1995; Hollands et al. 2006) and their interactions with other topographic features may provide insights into the formation and environmental significance of the sand dunes of that moon.
14.10 SOLAR SYSTEM PROCESSES IN TERRESTRIAL REGOLITH However, just as terrestrial features and processes can serve as analogues for those on the Moon, Mars and
Extraterrestrial regolith
elsewhere, so an understanding extraterrestrial regolith and processes may inform understanding of terrestrial features and the processes that shape them. One such example that is already well developed is the use of cosmogenic isotopes – formed by space weathering from cosmic ray bombardment – to date the terrestrial regolith (Pillans 2004, 2005; Section 2.3.3; Figure 2.2). There are more than 30 known impact structures on the Australian continent (Glikson 1996). While comparatively uncommon, impact structures are locally very important. They intensely modify the local geology and influence groundwater flow systems, sulfide mineralisation and hydrocarbon migration (Grieve 2005), and locally control local biogeography (Cockell and Lee 2002). While less than 10 craters or crater groups less than 1 km across are presently known from Australia (Bevan 1996), many more are likely to exist and have been too degraded or infilled to be readily identified. Some may be confused with pits and depressions formed by other processes, such as deflation or collapse. The structural and geophysical expression of impacts may also be confused with those of diatremes and ring dykes, and of domes and basins – an important consideration when using air or space-borne imagery and geophysical surveys for reconnaissance and mapping. Another area for consideration is that, although the amount of meteoritic material that descends on Australia is small (about 78 tonnes per year, by extrapolation from Love and Brownlee 1993), not all the material descends in the form of particulates. A proportion of the cosmic material arrives as meteorites, some of which are Ni-Fe alloys. On weathering, their high Ni content may lead to confusion with gossans (see Section 13.4). Although the probability of finding a meteorite is small, several anecdotal examples of ‘gossans’ proving to be weathered meteorites are known to the author. Recognition of meteoritic material in both hand specimen and in lag (Tilley and Bevan 1998) could be critical in the correct assessment of geochemical sampling programs. Development of criteria for the recognition and elimination of such fragments in lag sampling could be valuable. A third area where understanding of that extraterrestrial processes may assist in terrestrial regolith studies is that of stratigraphy and dating. The 790 Ka age of the Australasian tektite strewn field provides a
potential marker horizon in the study of the stratigraphy of soil toposequences and Quaternary sediments and can provide helpful controls when other forms of dating are absent (Pillans 2004, 2005), provided appropriate care is taken in the interpretation of their stratigraphic context. The mere presence of tektites in a deposit means that the deposit cannot be older than 790 Ka. The distribution and preservation of tektites can also shed valuable light onto rates of erosion and deposition: for example, the scarcity of tektites in north-western Australia (Fudali et al. 1991) may provide supporting evidence for high levels of Quaternary denudation suggested by the absence of thick weathering profiles across parts of the region. Finally, it is worth noting that impact craters are often occupied by central lakes. As a result they can provide significant paleoenvironmental records. Significant examples in Australia are the mid-late Quaternary record in the Darwin crater in Tasmania (Colhoun and van der Geer 1988) and the Pliocene record in Yallalie crater, Western Australia (Dodson and Ramrath 2001). The Yallalie succession is particularly important for paleoenvironmental reconstruction given the paucity of detailed Pliocene paleoenvironmental records in Australia during the critical period when the climate shifted towards a more arid regime (Section 2.8.5).
14.11 CONCLUSIONS Our understanding of the range of extraterrestrial regolith will increase in the coming decades as the surfaces of more planets, satellites, asteroids and comets are studied in detail. Furthermore, the amount of data being returned by missions to date greatly exceeds the numbers and ability of the existing planetary research community to process and interpret. The Mars Reconnaissance Orbiter is imaging the surface of Mars at sub-metre resolution, and new rover missions to Mars are in the planning stage. Phoenix will land in the Martian polar regions in 2008. Further lunar rovers and possible new manned missions may be expected to the Moon within the next 20 years. Terrestrial applications are also important, with analogue studies providing valuable data on both terrestrial and extraterrestrial regolith, and it is high likely that there are still many impact
401
402
Regolith Science
structures at all scales waiting to be identified on the Australian continent. The sky is literally the limit for regolith studies, and Australian regolith scientists are very well placed to be at the forefront of research.
14.12 REFERENCES Arvidson RE, Seelos FP, Deal KS, Koeppen WC, Snider NO, Kieniewicz JM, Hynek BM, Mellon MT and Garvin JB (2003). Mantled and exhumed terrains in Terra Meridiani, Mars. Journal of Geophysical Research 108, DOI: 10.1029/2002JE001982. Arvidson RE, Poulet F, Morris RV, Bibring J-P, Bell J F, Squyres SW, Christensen PR, Bellucci G, Gondet B, Ehlmann BL, Farrand WH, Fergason RL, Golombek M, Griffes JL, Grotzinger J, Guinness EA, Herkenhoff KE, Johnson JR, Klingelhöfer G, Langevin Y, Ming D, Seelos K, Sullivan RJ, Ward JG, Wiseman S M and Wolff M (2006). Nature and origin of the hematite-bearing plains of Terra Meridiani based on analyses of orbital and Mars Exploration rover data sets. Journal of Geophysical Research 111, DOI: 10.1029/2006JE002728. Bart GD (2007). Comparison of small lunar landslides and martian gullies. Icarus 187, 417–421. Beaty DW, Snook K, Allen CC, Eppler D, Farrell WM, Heldmann J, Metzger P, Peach L, Wagner SA and Zeitlin C (2005). ‘An analysis of the precursor measurements of Mars needed to reduce the risk of the first human missions to Mars.’ Unpublished white paper, 77 pp, posted June, 2005 by the Mars Exploration Program Analysis Group (MEPAG) at HTUhttp://mepag.jpl.nasa.gov/reports/index. htmlUTH. Benison KC and Bowen BB (2006). Acid saline lake systems give clues about past environments and the search for life on Mars. Icarus 183, 225–229. Bevan AWR (1996). Australian crater-forming meteorites. AGSO Journal of Australian Geology and Geophysics 6, 421–430. Bibring J-P, Langevin Y, Mustard JF, Poulet F, Arvidson R, Gendrin A, Gondet B, Mangold N, Pinet P, Forget F and the OMEGA team (2006). Global mineralogical and aqueous mars history derived from OMEGA/Mars Express data. Science 312, 400–404.
Bishop MA (1999). Comparative geomorphology of seasonally active crescentic dunes: Nili Patera, Mars and Strzelecki Desert, Earth. Fifth International Conference on Mars Abstract #6069. LPI Contribution 972. Lunar and Planetary Institute, Houston. Bourke MC and Zimbelman JR (2000). Australian paleoflood systems: an analogue for Martian channel systems. Proceedings of the 31st Lunar and Planetary Science Conference. Abstract #1393. Bourke MC and Zimbelman JR (2001). The Australian paleoflood model for unconfined fluvial deposition on Mars. Proceedings of the 32nd Lunar and Planetary Science Conference. Abstract #1679. Bourke MC, Clarke J, Manga M, Nelson P, Williams K, Fonseca J and Fobar B (2007). Spring mounds and channels at Dalhousie, central Australia. Abstracts of the 37th Lunar and Planetary Science Conference. Abstract #2174. Brown AJ, Walter MR and Cudahy TJ (2005). Hyperspectral imaging spectroscopy of a Mars analogue environment at the North Pole Dome, Pilbara Craton, Western Australia. Australian Journal of Earth Sciences 52, 353–364. Cattermole P (2001). Mars: the Mystery Unfolds. Oxford University Press, New York. Christensen PR, Wyatt MB, Glotch TD, Rogers AD, Anwar S, Arvidson RE, Bandfield JL, Blaney DL, Budney C, Calvin WM, Fallacaro A, Fergason RL, Gorelick N, Graff TG, Hamilton VE, Hayes AG, Johnson JR, Knudson AT, McSween HY, Mehall GL, Mehall LK, Moersch JE, Morris RV, Smith MD, Squyres SW, Ruff SW and Wolff MJ (2004). Mineralogy at Meridiani Planum from the mini-TES experiment on the Opportunity Rover. Science 306, 1733–1739. Clarke JDA and Pain CF (2004). From Utah to Mars: regolith-landform mapping and its application. In Mars Expedition Planning. (Ed. CC Cockell). Science and Technology Series 107, pp. 131–160. American Astronautical Society, Springfield, Virginia. Clarke JDA and Stoker C (2003). Mound spring complexes in central Australia: an analog for Martian groundwater-fed outflow channels? Abstracts of the 34th Lunar and Planetary Science Conference. Abstract #1504.
Extraterrestrial regolith
Clarke JDA, Thomas M and Norman M (2004). The Arkaroola Mars analog region, South Australia. Abstracts of the 34th Lunar and Planetary Science Conference. Abstract #1029. Clarke JDA, Bourke MC, Manga M, Nelson PA and Fonseca J (2007).The Dalhousie Mound Spring Complex as a guide to Martian landforms and processes. Proceedings of the 7th Australian Mars Exploration Conference, in press. Cockell CS and Lee P (2002). The biology of impact craters – a review. Biology Review 77, 279–310. Colhoun EA and van der Geer G (1988). Darwin Crater, the King and Linda valleys. In Cainozoic Vegetation of Tasmania. (Compiler EA Colhoun) pp. 30–71. University of Newcastle, New South Wales, Australia. Crumpler LS, Squyres SW, Arvidson RE, Bell JF III, Blaney D, Cabrol NA, Christensen PR, Desmarais DJ, Farmer JD, Fergason R, Golombek MP, Grant FD, Grant JA, Greeley R, Hahn B, Herkenhoff KE, Hurowitz JA, Knudson AT, Landis GA, Li R, Maki J, McSween HY, Ming DW, Moersch JE, Payne MC, Rice JW, Richter L, Ruff SW, Sims M, Thompson SD, Tosca N, Wang A, Whelley P, Wright SP and Wyatt MB (2005). Mars Exploration Rover Geologic traverse by the Spirit rover in the Plains of Gusev Crater, Mars. Geology 33, 809–812. Dodson JR and Ramrath A (2001). An Upper Pliocene lacustrine environmental record from south-western Australia – preliminary results. Palaeogeography, Palaeoclimatology, Palaeoecology 167, 309–320. Elachi C, Wall S, Allison M, Anderson Y, Boehmer R, Callahan P, Encrenez P, Flamini E, Franceschetti G, Gim Y, Hamilton G, Hensley S, Janssen M, Johnson W, Kelleher K, Kirk R, Lopes R, Lorenz R, Lunine J, Muhleman D, Ostro S, Paganelli F, Picardi G, Posa F, Roth L, Seu R, Shaffer S, Soderblom L, Stiles B, Stofan E, Vetrella S, West R, Wood C, Wye L and Zebker H (2005). Cassini Radar views the surface of Titan. Science 308, 970–974. Elachi C, Wall S, Janssen M, Stofan E, Lopes R, Kirk R, Lorenz R, Lunine J, Paganelli F, Soderblom L, Wood C, Wye L, Zebker H, Anderson Y, Ostro S, Allison M, Boehmer R, Callahan P, Encrenaz P, Flamini E, Francescetti G, Gim Y, Hamilton G, Hensley S, Johnson W, Kelleher K, Muhleman D,
Picardi G, Posa F, Roth L, Seu R, Shaffer S, Stiles B, Vetrella S and West R (2006). Titan Radar Mapper observations from Cassini’s T3 fly-by. Nature 441, 709–713. Feldman WC, Maurice S, Lawrence DJ, Little RC, Lawson SL, Gasnault O, Wiens RC, Barraclough BL, Elphic RC, Prettyman TH, Steinberg JT and Binder AB (2001). Evidence for water ice near the lunar poles. Journal of Geophysical Research (E) 106, 23,231–23,251. Florenskiy KP, Basilevskiy AT and Pronin, AA (1977). The first panoramas of the Venusian surface – geological-morphological analysis of pictures. In Space research XVII. (Ed. MJ Rycroft) pp. 645–649. Pergamon Press, Oxford and New York. Florenskiy KP, Bazilevskiy AT, Burba GA, Nikolayeva OV, Pronin AA, Selivanov AS, Narayeva M.K, Panfilnov AS and Chemodanov VP (1983). Panorama of the Venera 9 and 10 landing sites. In Venus. (Eds DM Hunten, L Colin, TM Donahue and VI Moroz) pp. 137–153. The University of Arizona Press, Tucson. Fortes AD, Grindrod PM, Trickett SK and Vocaldo L (2007). Ammonium sulfate on Titan: possible origin and role in cryovolcanism. Icarus 188, 139–153. Fudali RF, Miller AJ and Bevan AWR (1991). Australites from northern Australia. Meteoritics 26, 153–155. Fujiwara A, Kawaguchi K, Yeomans DK, Abe M, Mukai T, Okada J, Saito J, Yano H, Yoshikawa M, Scheeres DJ, Barnouin-Jha O, Cheng OF, Demura H, Gaskell RW, Hirata N, Ikeda H, Kominato T, Miyamoto H, Nakamura AM, Nakamura R, Sasaki S, Uesugi K (2006). The rubble-pile asteroid Itokawa as observed by Hayabusa. Science , 1330–1334. Gellert R, Rieder R, Anderson RC, Brückner J, Clark BC, Dreibus G, Economou T, Klingelhöfer G, Lugmair GW, Ming DW, Squyres SW, d’Uston C, Wänke H, Yen A and Zipfel J (2004). Chemistry of rocks and soils in Gusev Crater from the alpha particle X-ray spectrometer. Science , 829–832. Gibson DL (1999). Explanatory notes for the Broken Hill and Curnomona 1: 500,000 regolith landform maps. Open File report 77. CRC LEME, Perth.
403
404
Regolith Science
Glikson AY (Ed.) (1996). Australian impact structures. AGSO Journal of Australian Geology and Geophysics 16, 371–619. Glikson AY and Haines PW (Eds) (2005). Shoemaker Memorial Issue on the Australian impact record: 1997–2005 update. Australian Journal of Earth Sciences 52, 475–798. Golombek MP, Arvidson RE, Bell JF, Christensen PR, Crisp JA, Crumpler LS, Ehlmann BL, Fergason RL, Grant JA, Greeley R, Haldemann AFC, Kass DM, Parker TJ, Schofield JT, Squyres SW and Zurek RW (2005). Assessment of Mars Exploration Rover landing site predictions. Nature 436, 44–48. Grant JA, Arvidson R, Bell JF, Cabrol NA, Carr MH, Christensen P, Crumpler L, Des Marais DJ, Ehlmann BL, Farmer J, Golombek M, Grant FD, Greeley R, Herkenhoff K, Li R, McSween HY, Ming DW, Moersch J, Rice JW, Ruff S, Richter L, Squyres S, Sullivan R and Weitz C (2004). Surficial deposits at Gusev Crater along Spirit Rover traverses. Science 305, 807–810. Greeley R (1994). Planetary landscapes. Chapman and Hall, New York, 2nd edition. Greeley R and Williams SH (1994). Dust deposits on Mars; the ‘Parna’ analog. Icarus 110, 165–177. Greeley R, Whelley PL, Arvidson RE, Cabrol NA, Foley DJ, Franklin BJ, Geissler PG, Golombek MP, Kuzmin RO, Landis GA, Lemmon MT, Neakrase LDV, Squyres SW and Thompson SD (2006). Active dust devils in Gusev crater, Mars: observations from the Mars Exploration Rover Spirit. Journal of Geophysical Research 111, E12, E12S09. DOI 10.1029/2006JE002743 Grieve RAF (2005). Economic natural resource deposits at terrestrial impact structures. In Mineral Deposits and Earth Evolution. (Eds I McDonald, AJ Boyce, IB Butler, RJ Herrington and DA Polya). Special Publication 248, pp. 1–29. The Geological Society, London. Grin E A and Cabrol NA (1997). Limnologic analysis of Gusev Crater paleolake, Mars. Icarus 130, 461–474. Grotzinger JP, Arvidson RE, Bell JF III, Calvin W, Clark BC, Fike DA, Golombek M, Greeley R, Haldemann A, Herkenhoff KE, Jollif BL, Knoll AH, Malin M, McLennan SM, Parker T, Soderblom L,
Sohl-Dickstein JN, Squyres SW, Tosca NJ and Watters WA (2005). Stratigraphy and sedimentology of a dry to wet eolian depositional system, Burns formation, Meridiani Planum, Mars. Earth and Planetary Science Letters 240, 11–72. Halevy I, Zuber MT and Schrag DP( 2007). A sulfur dioxide climate feedback on early Mars. Science 318, 1903–1907. Hartmann WK (1973). Ancient lunar mega-regolith andsubsurface structure. Icarus 18, 634–636. Hartmann WK (2003). A Traveler’s Guide to Mars. Workman Publishing, New York. Hartmann WK, Anguita J, de la Casa MA, Berman DC and Ryan EV (2001). Martian cratering 7: The role of impact gardening. Icarus 149, 37–53. Head J, Mustard JF, Kreslavsky KA, Millikan RE and Marchant DR (2006). Recent ice ages on Mars. Nature 426, 797–802. Heim JA, Vasconcelos PM, Shuster DL, Farley KA and Broadbent G (2006). Dating paleochannel iron ore by (U-Th)/He analysis of supergene goethite, Hamersley Province, Australia. Geology 34, 173–176. Hodges RR Jr (2002). Ice in the lunar polar regions revisited. Journal of Geophysical Research 107, E2,5011. DOI 10.1029/2000JE001491. Hollands CB, Nanson GC, Jones BMG, Bristow CS, Price DM and Pietsch TJ (2006). Aeolian-fluvial interaction; evidence for late Quaternary channel change and wind-rift linear dune formation in the northwestern Simpson Desert, Australia. Quaternary Science Reviews 25, 142–162. Horz F, Hartung JB and Gault DE (1971). Micrometeorite craters on lunar rock surfaces. Journal of Geophysical Research 76, 5770–5799. Hurowitz JA and McLennan SM (2007). A ~3.5 Ga record of water-limited, acidic weathering conditions on Mars. Earth and Planetary Science Letters 260, 432–443. Iriyama J and Honda M (1979). Gardening process of lunar surface layer inferred from the galactic cosmic-ray exposure ages of lunar samples. Bulletin of the Volcanological Society of Japan 24, 1–8. Jankowski DG and Squyres SW (1993). ‘Softened’ impact craters on Mars; implications for ground ice and the structure of the Martian megaregolith. Icarus 106, 365–379.
Extraterrestrial regolith
Kargel JS (2004). Mars – A Warmer, Wetter Planet. Springer Praxis Books, New York. Kargel JS, Komatsu G, Baker VR and Strom RG (1993). The volcanology of Venera and VEGA landing sites and the geochemistry of Venus. Icarus 103, 253–275. Lebreton J-P, Witasse O, Sollazzo C, Blancquaert T, Couzin P, Schipper A-M, Jones JB, Matson DL, Gurvits LI, Atkinson DH, Kazemineja B and PerezAyucar M. (2005). An overview of the descent and landing of the Huygens probe on Titan. Nature 438, 758–764. Leshin LA, Minitti M, Cosarinsky M, McAdam A and Niles P (2004). A view from the cheap seats: MER and its implications for future Mars exploration. Fall Meeting American Geophysical Union. Abstract #P24A-08. Lewis JS, Matthews MS and Guerrieri ML (Eds) (1993). Resources of Near-Earth Space. The University of Arizona Press, Tucson, Arizona. Lorenz RD, Wall S, Radebaugh J, Boubin G, Reffet E, Janssen M, Stofan E, Lopes R, Kirk R, Elachi C, Lunine J, Mitchell K, Paganelli F, Soderblom L, Wood C, Wye L, Zebker H, Anderson Y, Ostro S, Allison M, Boehmer R, Callahan P, Encrenaz P, Ori GG, Francescetti G, Gim Y, Hamilton G, Hensley S, Johnson W, Kelleher K, Muhleman D, Picardi G, Posa F, Roth L, Seu R, Shaffer S, Stiles B, Vetrell, S, Flamini E and West R (2006). The sand seas of Titan: Cassini RADAR observations of longitudinal dunes. Science 312, 724–727. Love SG and Brownlee DE (1993). A direct measurement of the terrestrial mass accretion rate of cosmic dust. Science 262, 550–553. McGill GE, Warner JL, Mailin M, Arvidson RE and Eliason E (1983). Topography, surface properties, and tectonic evolution. In Venus. (Eds DM Hunten, L Colin, TM Donahue and VI Moroz) pp. 69–130. The University of Arizona Press, Tucson. Mann GA, Clarke JDA and Gostin V (2004). Surveying the Mars analogue research sites in the central Australian deserts. Australian Geographical Studies 30,116–124. Milton DJ and Michel FC (1965). ‘Structure of a ray crater at Henbury, Northern Territory, Australia.’ Professional Paper 525-C p. C5-C11. US Geological Survey, Reston, Virginia.
Mitri G, Showman AP, Lunine JI and Lorenz RD (2007). Hydrocarbon lakes on Titan. Icarus 186, 385–394 Mitrofanov IG, Zuber MT, Litvak ML, Boynton WV, Smith DE, Drake D, Hamara D, Kozyrev AS, Sanin AB, Shinohara C, Saunders RS and Tretyakov V (2003). CO2 Snow depth and subsurface water-ice abundance in the Northern Hemisphere of Mars. Science 300, 2081–2084. Miyamoto M, Yano H, Scheeres DJ, Abe S, BarnouinJha O, Cheng AF, Demura H, Gaskell RW, Hirata N, Ishiguro M, Michikami T, Nakamura AM, Nakamura R, Saito J and Sasaki S (2007). Regolith Migration and Sorting on Asteroid Itokawa. Science 316, 1011–1014. Morris RV, Klingelhöfer G, Schröder C, Rodionov DS, Yen A, Ming DW, de Souza PA, Wdowiak T, Fleischer I, Gellert R, Bernhardt B, Bonnes U, Cohen BA, Evlanov EN, Foh J, Gütlich P, Kankeleit E, McCoy T, Mittlefehldt DW, Renz F, Schmidt ME, Zubkov B, Squyres SW and Arvidson RE (2006). Mössbauer mineralogy of rock, soil, and dust at Meridiani Planum, Mars: Opportunity’s journey across sulfate-rich outcrop, basaltic sand and dust, and hematite lag deposits. Journal of Geophysical Research 111, DOI: 10.1029/2006JE002791. Moroz VI (1983). Summary of preliminary results of the Venera 13 and 14 missions . In Venus. (Eds DM Hunten, L Colin, TM Donahue and VI Moroz) pp. 45–68. University of Arizona Press, Tucson. Nanson GC, Chen XY and Price DM (1995). Aeolian and fluvial evidence of changing climate and wind patterns during the past 100 ka in the western Simpson Desert, Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 113, 87–102. Nelson PA, Manga M, Bourke MC and Clarke JDA (2007). A model for mound spring formation and evolution. Abstracts of the 37th Lunar and Planetary Science Conference. Abstract #2111. Neukum G, Jaumann R, Hoffmann H, Hauber E, Head JW, Basilevsky AT, Ivanov BA, Werner SC, van Gasselt S, Murray JB, McCord T and The HRSC Co-Investigator Team (2004). Recent and episodic volcanic and glacial activity on Mars revealed by the high resolution stereo camera. Nature 432, 971–979.
405
406
Regolith Science
Nozette S and Lewis JS (1982). Venus – chemical weathering of igneous rocks and buffering of atmospheric composition. Science 216, 181–183. Oshigami S, Namiki1 N and Komatsu G (2007). A thermal erosion origin for venusian sinuous rilles. Abstracts of the 37th Lunar and Planetary Science Conference. Abstract #1185. Paganelli F, Elachi C, Lopes CM, West R, Stiles B, Janssen MA, Stofan ER, Wood CA, Lorenz RD, Lunine JL, Kirk RL, Roth LE, Wall SD and Soderblom LA (2005). Channels and fan-like features on Titan surface imaged by the Cassini radar. Abstracts of the 36th Lunar and Planetary Science. Abstract #2150. Pain CF and Ollier CD (1995). Inversion of relief – a component of landscape evolution. Geomorphology 12, 151–165. Pain CF, Clarke JDA and Thomas M (2007). Relief inversion on Mars. Icarus 190, 478-491. Perron JT, Mitrovica JX, Manga M, Matsuyama I and Richards MA (2007). Evidence for an ancient martian ocean in the topography of deformed shorelines. Nature 447, 840–843. Phillips RJ, Raubertas RF, Arvidson RE, Sarkar IC, Herrick RR, Izenberg N and Grimm RE (1992). Impact craters and Venus resurfacing history. Journal of Geophysical Research 97, 15,923–15,948. Pillans B (2004). Tektites as chronostratigraphic markers in Australian regolith. In Regolith 2004. (Ed. IC Roach) pp. 279–281. CRC LEME, Perth. Pillans B (2005). Geochronology of the Australian regolith. In Regolith Landscape Evolution Across Australia. (Eds RR Anand and P de Broekert) pp. 41–52. CRC LEME, Perth. Porco CC, Baker E, Barbara J, Beurle K, Brahic A, Burns JA, Charnoz S, Cooper N, Dawson DD, Del Genio AD, Denk T, Dones L, Dyudina U, Evans MW, Fussne, S, Giese B, Grazier K, Helfenstein P, Ingersoll AP, Jacobson RA, Johnson TV, McEwen A, Murray CD, Neukum G, Owen WM, Perry J, Roatsch T, Spitale J, Squyres S, Thomas P, Tiscareno M, Turtle EP, Vasavada AR, Veverka J, Wagner R, West R (2005). Imaging of Titan from the Cassini spacecraft. Nature 434, 159–168. Schaefer L and Fegley BJ (2004). Heavy metal frost on Venus. Icarus 168, 215–219.
Soderblom LA, Kirk RL, Lunine, JI, Anderson JA, Baines KH, Barnes JW, Barrett JM, Brown RH, Buratt BJ, Clark RN, Cruikshank DP, Elachi C, Janssen MA, Jaumann R, Karkosch E, Le Mouelic S, Lopes RM, Lorenz RD, McCordi TB, Nicholson PD, Radebaugh J, Rizk B, Sotin C, Stofan ER, Sucharski TL, Tomasko MG and Wall SD (2007a). Correlations between Cassini VIMS spectra and RADAR SAR images: implications for Titan’s surface composition and the character of the Huygens Probe Landing Site. Planetary and Space Science 55, 2025-2036. Soderblom LA, Tomasko MG, Archinal BA, Becker TL, Bushroe WL, Cook DA, Doose LR, Galuszka DM, Hare TM, Howington-Kraus E, Karkoschka E, Kirk RL, Lunine JI, McFarlane EA, Redding BL, Rizk B, Rosiek MR, See C and Smith PH (2007b). Topography and geomorphology of the Huygens landing site on Titan. Planetary and Space Science 55, 2015–2024. Squyres SW, Arvidson RE, Bollen D, Bell JF, Brückner J, Cabrol NA, Calvin WM, Carr MH, Christensen PR, Clark BC, Crumpler L, Des Marais DJ, d’Uston C, Economou T, Farmer J, Farrand WH, Folkner W, Gellert R, Glotch TD, Golombek M, Gorevan S, Grant JA, Greeley R, Grotzinger J, Herkenhoff KE, Hviid S, Johnson JR, Klingelhöfer G, Knoll AH, Landis G, Lemmon M, Li R, Madsen MB, Malin MC, McLennan SM, McSween HY, Ming DW, Moersch J, Morris RV, Parker T, Rice JW, Richter L, Rieder R, Schröder C, Sims M, Smith M, Smith P, Soderblom LA, Sullivan R, Tosca NJ, Wänke H, Wdowiak T, Wolff M and Yen A (2006). Overview of the Opportunity Mars Exploration Rover Mission to Meridiani Planum: Eagle Crater to Purgatory Ripple. Journal of Geophysical Research 111, DOI: 10.1029/2006JE002771. Steven NS and Atkinson FA (1972). The Undara lava tubes, north Queensland, Australia. In Proceedings of the International Symposium on Vulcanospeleology and its Extraterrestrial Applications Vol. 29, pp. 58–63. National Speleology Society, Seattle, Washington. Stofan ER, Elachi C, Lunine JI, Lorenz RD, Stiles B, Mitchell KL, Ostro S, Soderblom L, Wood C, Zebker H, Wall S, Janssen M, Kirk R, Lopes R,
Extraterrestrial regolith
Paganelli F, Radebaugh J, Wye L, Anderson Y, Allison M, Boehmer R, Callahan P, Encrenaz P, Flamini E, Francescetti G, Gim Y, Hamilton G, Hensley S, Johnson WTK, Kelleher K, Muhleman D, Paillou P, Picardi G, Posa F, Roth L, Seu R, Shaffer S, Vetrella S and West R (2007). The lakes of Titan. Nature 445, 61–64. Taylor SR (1975). Lunar Science: A Post Apollo View. Pergamon Press Inc, New York. Taylor SR (2001). Solar System Evolution. 2nd edn. Cambridge University Press, Cambridge, UK. Thomas M and Walter MR (2002). Application of hyperspectral infrared analysis of hydrothermal alteration on Earth and Mars. Astrobiology 2, 335–351. Thomas M, Clarke JDA and Pain CF (2005). Weathering, erosion and landscape processes on Mars identified from recent rover imagery, and possible earth analogues. Australian Journal of Earth Sciences 52, 365–378. Tilley D and Bevan AWR (1998). Prolonged weathering of iron and stony-iron meteorites. In Proceedings of Regolith 98: new approaches to an old continent. (Eds G Taylor and C Pain) pp. 77–88. CRC LEME, Perth. Tokano T, McKay CP, Neubauer FM, Atreya SK, Ferri F, Fulchignoni M and Niemann HB (2006). Methane drizzle on Titan. Nature 442, 432–334. Tomasko MG, Archinal B, Becker T, Bézard B, Bushroe M, Combes M, Cook D, Coustenis A, de Bergh C, Dafoe LE, Doose L, Douté S, Eibl A, Engel S, Gliem F, Grieger B, Holso K, Howington-Kraus E, Karkoschka E, Keller HU, Kirk R, Kramm R, Küppers M, Lanagan P, Lellouch E, Lemmon M, Lunine J, McFarlane E, Moores J, Prout GM, Rizk B, Rosiek M, Rueffer P, Schröder SE, Schmitt B, See C, Smith P, Soderblom L, Thomas N and West R (2005). Rain, winds and haze during the Huygens probe’s descent to Titan’s surface. Nature 438, 765–778. Veverka J, Robinson M, Thomas P, Murchie S, Bell JF III, Izenberg N, Chapman C, Harch A, Bell M, Carcich B, Cheng A, Clark B, Domingue D, Dunham D, Farquhar R, Gaffey MJ, Hawkins E, Joseph J, Kirk R, Li H, Lucey P, Malin M, Martin P, McFadden L, Merline JJ, Miller JK, Owen WM Jr,, Peterson C, Prockter L, Warren J, Wellnitz D, Williams
BG and Yeomans DK (2000). NEAR at Eros: imaging and spectral results. Science 289, 2088–2097. Veverka J, Farquhar B, Robinson MS, Thomas PC, Murchie SL, Harch A, Antreasian PG, Chesley SR, Miller JK, Owen WM Jr, Williams BG, Yeomans DK, Dunham D, Heyler G, Holdridge M, Nelson RL, Whittenburg KE, Ray JC, Carcich B, Cheng AF, Chapman C, Bell JF III, Bell M, Bussey B, Clark BC, Domingue D, Gaffey MJ, Hawkins E, Izenberg N, Joseph J, Kirk RL, Lucey PG, Malin MC, McFadden L, Merline WJ, Peterson CD, Prockter L, Warren J and Wellnitz D (2001). The landing of the NEAR-Shoemaker spacecraft on asteroid 433 Eros. Nature 413, 390–393. Waclawik V and Gostin V (2006). Significance of remnant gravel lags as landscape evolution indicators, Arkaroola Mars analogue region geology. In Mars Analogue Research. (Ed. JDA Clarke) Science and Technology Series 111, pp. 107–114. American Astronautical Society, Springfield, Virginia. Wang A, Haskin LA, Squyres SW, Jolliff BL, Crumpler L, Gellert R, Schröder C, Herkenhoff K, Hurowitz J, Tosca NJ, Farrand WH, Anderson R and Knudson A T (2006). Sulfate deposition in subsurface regolith in Gusev crater, Mars. Journal of Geophysical Research 111, DOI: 10.1029/2005JE002513. Weitz CM, Plaut JJ, Greeley R and Saunders RS (1994). Dunes and microdunes on Venus: why were so few found in the Magellan data? Icarus 112, 282–295. Yano H, Kubota T, Miyamoto H, Okada T, Scheeres D, Takagi Y, Yoshida K, Abe M, Abe S, Barnouin-Jha O, Fujiwara A, Hasegawa S, Hashimoto T, Ishiguro M, Kato M, Kawaguchi J, Mukai T, Saito J, Sasaki S, Yoshikawa M (2006). Touchdown of the Hayabusa spacecraft at the Muses Sea on Itokawa. Science 312, 1350–1353. Zarnecki JC, Leese MR, Hathi B, Ball AJ, Hagermann A, Towner MC, Lorenz RD, McDonnell JAM, Green SF, Patel MR, Ringrose T.J, Rosenberg PD, Atkinson KR, Paton M.D, Banaszkiewicz M, Clark BC, Ferri F, Fulchignoni M, Ghafoor NAL, Kargl G, Svedhem H, Delderfield J, Grande M, Parker DJ, Challenor PG and Geake JE (2005). A soft solid surface on Titan as revealed by the Huygens Surface Science Package. Nature 438, 792–795.
407
Appendix 1: Glossary of regolith terms Richard A Eggleton, Colin F Pain and Keith M Scott
INTRODUCTION This glossary is intended to provide definitions of words used in this book and other commonly used regolith terms. Where possible, definitions are taken from Eggleton (2001) and words defined within the AGI Glossary of Geology (Jackson 1997) are omitted. Generally, only one part of speech (for example, noun rather noun plus verb) is defined in the interests of brevity. Some regolith minerals are listed but the reader is directed to Nickel and Nichols (1991) for basic information on those not listed below. A ! HORIZON See soil profile diagram: Figure 6.3. abrasion The mechanical breaking of rocks or minerals by either friction against other rocks or by impact of other rock or mineral fragments; e.g. in streams and during aeolian transport. accretion The process by which an inorganic body increases in size by external addition, as by adhesion or precipitation (see also concretion). ACCUMULATION ZONE That part of a weathering profile that is characterised by the accumulation of some elements (commonly Al, Fe, Mn and Ni). ACID SULFATE SOIL A soil in which sulfuric acid produced by the weathering of sulfides affects soil characteristics (Sections 12.8.2; 7.7.1).
alcrete Duricrust cemented mainly by Al compounds (cf. bauxite). allite A weathering product with a high proportion of Al and Fe oxides. The word ‘ferrallite’ has the same meaning and may be preferred if the Fe minerals dominate the assemblage. Allitic or ferrallitic regolith may also contain some quartz and kaolinite. allocthonous Material formed or produced in a place different from where it now resides; e.g. aeolian material. Transported material. allophane Clay mineral, with composition in the range from Al2O3.SiO2.xH2O to Al2O3.2SiO2. xH2O (2<x<3). Commonly as hollow ~5 nm spheres in altered volcanic ash or soils derived therefrom (Section 4.5.5). alluvium Sediment deposited from rivers, streams and creeks. alunite Mineral, KAl 3 (SO4) 2 (OH) 6, found in acid argillic alteration, and in acid sulfate weathering in regolith of arid climates and in playa sediments. Also, any Al-rich member of the alunite supergroup. ALUNITE SUPERGROUP Minerals with the general formula, A B3 (XO4)2 (OH) 6 where A is a large ion such as K, Na, Ca, REE or Pb; B is generally Fe or Al; (XO4) is generally (SO4)2– or (PO4)3–. The main
410
Regolith Science
division is on occupancy of B sites – Fe-dominant = jarosite family; Al-dominant = alunite family. ALVEOLAR WEATHERING Honeycomb weathering. anabranch A channel divergent from a main stream and either rejoining downstream or remaining separate and diminishing in size until becoming indistinct. Anabranches are common on alluvial plains. See also anastomosing; cf. braid. anastomosing A branching and reconnecting netlike pattern similar to the veins of a leaf. The term is commonly used to describe such a network of drainage channels (see also anabranch). anatase Mineral, TiO2 – with a tetragonal structure. Most abundant polymorph of TiO2 in the regolith (Section 4.4.4), resulting from the weathering of Ti-bearing silicates, sphene and ilmenite. It is commonly found as very small (0.1 µm) crystals, and is a constituent of the fine-grained alteration assemblage known as ‘leucoxene’. Anatase has a creamcoloured appearance when it is concentrated, but mostly it is dispersed uniformly through silicateweathering products. It is a significant minor component of most silcretes and of QAZ cement. anglesite White mineral, PbSO4 – an oxidation product of galena (Table 4.2). ANOMALY A deviation from common or normal experience. For numerical data, a value that differs significantly (in a statistical sense) from a previously established threshold or defined population (Section 5.5.4). Anthroposols Soils resulting from human activities that have led to a profound modification, truncation or burial of the original soil horizons, or the creation of new soil parent materials by a variety of mechanical means. Where burial of a pre-existing soil is involved, the anthropic materials must be 0.3 m or more thick. Pedogenic features may be the result of in situ processes (usually the minimal development of an A1 horizon; sometimes the stronger development of typical soil horizons) or the result of pedogenic processes prior to modification or placement (i.e. the presence of identifiable pre-existing soil material). !P HORIZON That part of the A soil horizon disturbed by tillage (ploughing).
apedal Term applied to soils with no observable peds. AQUA REGIA Mixture of HNO3 and HCl, which dissolves most geochemical pathfinder elements, including Au. aquiclude A stratum or layer of rock that prevents water ingress. An aquiclude that is a boundary to an adjacent aquifer is called a ‘confining bed’ (this term is recommended). AQUIFER Permeable stratum carrying accessible water. aquitard A confining bed that retards, but does not prevent, the flow of water to or from an adjacent aquifer; a leaky confining bed. It does not readily yield water to wells or springs, but may serve as a storage unit for groundwater. arene (French) Facies observed at the bottom of a weathering profile developed on granites or gneisses that still contain 1–10 mm-sized grains of fresh primary minerals (feldspars and micas). The texture is coarse. An arene is equivalent to a coarse saprolite. See also grus. arenose Sandy; composed largely of sand-sized particles. argillaceous/argillic Containing significant clay; clay-rich. argillan A cutan composed dominantly of clay minerals; e.g. a clay skin. argillisation Conversion to clay; said of the alteration of aluminosilicates to clay minerals. ARGILLOFERRAN Cutan composed of clays and Fe oxides. artesian An adjective referring to groundwater confined under hydrostatic pressure. !3$® Analytical spectral device. A field-portable reflectance spectrometer operating over the 350–2500 nm wavelength interval. It is useful to rapidly identify mineralogical components in regolith samples, especially Fe oxides and layer silicates (Section 4.8). AZONAL SOILS Soils having no certain relationship with climatic zones; e.g. lithosols. AZURITE Bright blue mineral, Cu3 (CO3)2 (OH)2 – a weathering product of copper minerals, such as chalcopyrite. See also malachite.
Appendix 1: Glossary of regolith terms
B " HORIZON See soil profile diagram: Figure 6.3. barite White mineral, BaSO4 (Table 4.2). It may form dense masses in regolith or occur dispersed. BAUXITE I Highly aluminous rock or regolith. Bauxite contains abundant Al hydroxides (gibbsite, less commonly boehmite or diaspore) and Alsubstituted Fe oxides or hydroxides, generally minor or negligible kaolin minerals, and may contain appreciable (20%) quartz. It is commonly a near-surface or surface regolith; often, but not essentially, has a texture of pisoliths or nodules, loose or cemented. It is interpreted to have formed by the intense weathering of any parent rock, but most typically of basalts, syenites, granites, arkoses, marls and shales in freely drained, strongly leached environments. Transport and deposition may form colluvial or alluvial bauxite. Named for the type district, Le Baux, France. (ii) Karst bauxite is a variant, which is formed in caves and hollows in limestones; the Fe and Al are presumably derived from common erosion products, transported by flowing water as denudation products and trapped in the caves. (iii) Economic term: the most common ore of Al. bedrock Solid rock that has not been affected by weathering. It is found at the surface of planetary bodies (usually Earth), or at some depth beneath the regolith. beidellite Clay mineral of the smectite group, approximately A0.3Al2[Si3.7Al0.3]O10.(OH)2.xH2O, where A is an exchangeable cation, such as K+, Na+ or 0.5Ca2+. Beidellite is a common early weathering product of feldspars. benthic Flora and fauna at the base of a water body. bentonite Rock, largely composed of smectite clay minerals, which is produced by the in situ alteration of volcanic tuff or ash. billabong Meandering stretch of river channel that has been abandoned on a floodplain. Similar to an OXBOW LAKE, which is only one meander long. BILLY Colloquial term for silcrete; also termed grey billy.
biomantle The upper portions of the regolith where the actions of biota are significant (Section 8.3.3; Figure 8.17). bioturbation The churning and stirring of sediment and regolith by animals and plants (Section 8.3). bisiallitisation The chemical change of primary silicates to 2:1 layer silicates, such as illite or smectite (Section 4.3). BLOCKY I Rock or regolith: structure resulting from three or more sets of intersecting joint planes; the enclosed mass is in situ, although the joints may leave a few millimetres space between the blocks; this space may be open or filled with secondary minerals; size of block >64 mm. (ii) Soil: accommodated, equant, structural elements of soil with a number of plane faces; reentrant angles between adjoining faces are virtually absent. boehmite Mineral, a-AlO(OH). Common mineral of bauxite, which is formed under condition of low water activity or as an alteration product of gibbsite. BOG IRON (i) A general term for a soft, spongy, and porous deposit of impure hydrous Fe oxides formed in bogs, marshes, swamps, peat mosses and shallow lakes by precipitation from Fe-bearing waters. The Fe oxides are impregnated with plant debris, clay and clastic material. Bog iron is especially abundant in the glaciated northern regions of North America and Europe. It may form a poorquality Fe ore. (ii) A term commonly applied to a loose, porous, earthy form of limonite occurring in wet ground. boinka A landform complex of groundwater discharge, which occupies shallow depressions in semi-arid regions of low relief. A boinka may include dunes, gypsum flats and salt lakes. bolus Ball of soil used to determine soil texture in the field (Section 12.3.3). borax White mineral, Na 2B4O7.10H2O – evaporite found in playa sediments. boxwork A fabric of angular cells in a highly ferruginous and/or siliceous weathered rock. The cells are commonly formed by the dissolution of minerals and are defined by a network of intersecting
411
412
Regolith Science
blades or plates of Fe oxides or silica that were deposited in cavities and original cleavages, grain boundaries and fracture planes of the dissolved mineral. Boxworks are typically associated with the weathering of sulphides and are found in some gossans and in caprocks over sulphide-bearing rocks; however, boxworks may form by the selective weathering of other minerals; e.g. carbonates and silicates such as olivine. braid In a river channel, to branch and rejoin repeatedly to form an intricate pattern or network of small interlacing stream channels within the broader channel (cf. anabranch). BREAKAWAY Scarp, usually capped by indurated and subindurated parts of the weathered mantle (also called a jumpup). brochantite Mineral, Cu4SO4 (OH) 6 (Table 4.2). Bright emerald to dark green oxidation product of Cu-sulfides, particularly under low pH arid conditions. C C3/C4 A division of plants according to their ability to fix CO2 as 3- or 4- C atom molecules. Trees and most plants are C3, but saltbush is C4. # HORIZON See soil profile diagram: Figure 6.3. CALCAREOUS SOIL Soil containing sufficient calcium carbonate or calcium-magnesium carbonate to effervesce visibly when treated with cold 0.1 M hydrochloric acid. calcite White mineral, CaCO3 – essential component of calcrete and common in regolith of more arid regions or over limestones. calcrete Used broadly to refer to regolith carbonate accumulations, forming more- or less-well-cemented aggregates composed largely of calcium carbonate, but not excluding dolomitic or magnesitic material. Although some regolith carbonates clearly cement fragmental regolith to form duricrusts, others may be pisolitic, nodular, pebbly, slabby or powdery. Calcrete is a convenient field term for all such carbonate accumulations responsive to the 0.1M HCl test, but more specific terms such as dolocrete or magcrete should be used if laboratory testing establishes such carbonate mineralogy.
caliche (i) Spanish–North American term used synonymously for calcrete. II A layer near the surface, more or less cemented by soluble minerals (such as calcite, dolomite, nitre or halite). It may occur as a soft thin soil horizon, as a hard thick bed just beneath the solum, or as a surface layer exposed by erosion. CAP ROCK The weathered product of any rock, but especially of disseminated sulphide mineralisation, which may be ferruginous (i.e. an ironstone) or siliceous (i.e. silcrete) forming a hard cap on top of a hill or upper slope. CAPTURED DRAINAGE Drainage capture, or piracy, occurs when drainage from one catchment is diverted into another catchment. Sharp bends in river direction and barbed drainage patterns can indicate river capture. The stream from which drainage is captured is said to be beheaded. CARAPACE AND NODULAIRE French term referring to the lower, poorly indurated horizon of the ferruginous zone of a laterite profile. Consists of Fe nodules and pisoliths in a weakly cemented matrix of kaolinite and Fe oxides. carnotite Yellow mineral, K2 (UO2)2 (VO4)2.3H2O – occurs in some groundwater calcretes, where it may reach economic amounts, and in weathered U-deposits. catena A sequence of soils from hill top to valley floor. The catena concept has evolved to mean a sequence of soils that are derived from the same or similar parent material but displaying variability that can be related to differences in the topography and drainage of their landscape position. CATION EXCHANGE CAPACITY#%# See Table 5.3). CAVERNOUS WEATHERING Chemical and mechanical weathering on a cliff face, in which grains and flakes or rock are loosened so as to enlarge hollows and recesses opened through a chemically hardened shell on the surface of the cliff face. It produces the tafoni in seaside cliffs. See also HONEYCOMB weathering and TAFONI CELLULAR DURICRUST See DURICRUST CELLULAR. CEMENTATION FRONT A transformation front resulting in cementation by components such as Al, Fe and Mn oxides, silica and Ca or Mg carbonates.
Appendix 1: Glossary of regolith terms
cemented Indurated; having a hard, brittle consistency because the constituent particles are held together by cementing substances such as humus, calcium carbonate, or Si, Fe and Al oxides. The hardness and brittleness persist even when wet. CENTRIFUGAL (In the context of regolith development) The migration of a transformation front or a transfer of material is centrifugal when it proceeds outwards from a void, a mottle, a nodule or a glaebule. The migration of the front can be centrifugal while the associated transfer of material is centripetal. centripetal (In the context of regolith development) The migration of a transformation front or a transfer of material is centripetal when it proceeds inwards toward a void, a mottle, a nodule or a glaebule. The migration of the front can be centripetal while the associated transfer of material is centrifugal. cerussite White mineral, PbCO3 – a weathering product of galena (Table 4.2). CHALCEDONY Mineral SiO2 – a cryptocrystalline variety of quartz. chalcophile Elements with a strong affinity for S (Section 5.2). Elements present in sulfide minerals and which can be used as geochemical pathfinders for sulfide mineralisation (Table 13.1). CHANNEL DEPOSIT Alluvium that is deposited in an alluvial channel. It is commonly coarser than surrounding deposits, and is found in both active and relict channels. It includes deposits in cut-off meanders, and point bar deposits. chelation Chelation, or complexing, is the holding of an ion – commonly a metal – by multidentate ligands. It is an important process in the aqueous chemistry of regolith and involves both mineral and organic materials. cheluviation Eluviation under the influence of chelates. Chromosols Soils with a clear or abrupt textural B horizon and in which the major part of the upper 0.2 m of the B2 horizon (or the major part of the entire B2 horizon if it is <0.2 m thick) is not sodic and not strongly acid. Soils with strongly subplastic upper B2 horizons are also included even if they are sodic.
chronosequence A sequence of soils developed under similar soil-forming conditions, but at different times. clathrate Compound containing molecules of a substance physically enclosed in the structure of another. CLAY MINERAL Phyllosilicate (layer silicate) mineral or other mineral that imparts plasticity to clay and which hardens upon drying or firing. CLAY SIZE A term used in the US and by the International Society of Soil Science for a rock or mineral particle in the soil, having a diameter of <0.002 mm (<2 µm). CLAY ZONE See PLASMIC ZONE, see also LATERITE PROFILE. CLAYPAN A term used in Australia for a shallow depression filled with clayey and silty sediment, and having a hard, sun-baked surface; a playa formed by deflation of alluvial topsoils in a desert, and in which water collects after rain. COLLAPSED SAPROLITE Residual regolith consisting of fragments having some of the lithic fabric of the underlying saprolite and bedrock. Individual fragments may have lost their original orientation, but megascopic features, such as quartz veins and lithological contacts, remain discernible. Collapsed saprolite results from selective volume loss (e.g. by leaching, eluviation) of saprolite and the settling of resistant portions. collophane Mineral – apatite in cryptocrystalline form. colluvium Heterogeneous materials of any particle size, which are generally composed of soil and/or rock fragments accumulated on the lower parts of slopes, transported there by gravity, soil creep, sheet-flow, rain-wash, mudflows or solifluction. concretion A hard, compact, rounded, normally sub-spherical (but may be any shape) mass or aggregate of mineral matter generally formed by precipitation from aqueous solution (commonly about a nucleus or centre, such as a leaf, shell, bone, or fossil) within a rock or regolith and generally of a composition widely different from that of the rock in which it is found and from which it is rather sharply separated (see also accretion).
413
414
Regolith Science
CONFINING BED Essentially impermeable stratum confining an aquifer. CONGRUENT DISSOLUTION Dissolution of a mineral without the formation of a different mineral; as the dissolution of halite or quartz (cf. INCONGRUENT dissolution). consolidated Said of regolith that has such firmness and coherence that a tool is needed to take a sample from outcrop or that remains coherent after extraction from a coring device. corestone A rounded, ellipsoidal or broadly rectangular block, composed of virtually fresh parent rock in saprock or saprolite; the residual relatively unweathered remnant of a joint block, originating from any massive type of parent rock – e.g. granite, gabbro, dolerite, basalt – but separated from it. cortex The outer, concentric shell or envelope of a particle, such as on a nodule or pisolith. A cortex may vary in thickness (cf. cutan). crackled A fabric of small to minute cracks in a discrete body such as a nodule; the cracks may be open or filled with other material. Such a fabric is generally developed by shrinkage during crystallisation or ageing of colloidal matter or gels (syneresis or septarian cracks). crandallite Mineral, CaAl3H(PO4)2 (OH) 6. Also a mineral group including florencite (having Ce and other rare earth elements in place of Ca) plumbogummite (Pb); gorceixite (Ba). These minerals are members of an isostructural series of regolith phosphates, which are ultimately derived from the weathering of apatite and monazite. CREEP DEPOSIT A thin layer of saprolite or earthy colluvium that moves very slowly down-slope. In some circumstances, it may be recognised by, for example, the bending and fragmentation of rock strata, dykes or quartz veins down slope, but in other cases creep can only be inferred. CRITICAL ZONE Recent term emphasising the importance of regolith in sustaining life on Earth. The critical zone is defined as the volume extending from the upper limit of vegetation down to the lower limit of groundwater (Brantley et al. 2007). CROUTE CALCAIRE French term for calcrete. CROWN SCARP The scarp left above a landslide.
CRYPTOCRYSTALLINE Composed of crystals below the resolution of optical microscopy. CRYPTOMELANE Black mineral, K 2Mn8O16 – isostructural with hollandite and coronadite. Found particularly in weathered manganese-rich rocks; e.g. at Broken Hill, NSW, Australia. (See also mangaNESE OXIDES WAD psilomelane). cuirasse French – armour – commonly translated in the context of a lateritic profile as ‘iron-crust’. The highly indurated upper facies or horizon of the ferruginous zone of a lateritic regolith, with a massive pisolitic, nodular or vesicular fabric. Synonymous with LATERITIC DURICRUST. cutan A skin, generally thin, on the natural surfaces in soil; i.e. on the walls of the voids, the surfaces of skeleton grains and aggregates (e.g. pisoliths) or associated structures (e.g. glaebules), or the boundaries of other associated structures. Cutans have a composition and/or fabric different from the objects they coat. Cutans may be argillans (clay), ferrans (Fe oxide), mangans (Mn oxide), calcans (or calcitans) (calcite). D $ HORIZON The term for a soil horizon below the C horizon, but unrelated to the C, B and A horizons above. An example would be where soil-forming processes have affected alluvium over granite and also affected the granite. The granite would show a D horizon; the A, B and C horizons being in the alluvium (Figure 6.3). DEEP LEAD A paleovalley fill generally investigated for potential placer deposits. DEFLATION Wind erosion. degradation (i) The wearing down or away – and the general lowering or reduction – of the Earth’s surface by the natural processes of weathering and erosion; e.g. the deepening by a stream of its channel. The term is sometimes used to include the process of transportation, and sometimes it is used synonymously with denudation, or used to signify the results of denudation. (ii) Less broadly, the vertical erosion or down-cutting performed by a stream in order to establish or maintain uniformity of grade or slope.
Appendix 1: Glossary of regolith terms
(iii) Change in a soil (decrease in exchangeable bases and destruction of layer-silicate clay) due to leaching. $%- Digital elevation model. A representation of elevation on a grid. DEM is relative; DTM (digital terrain model) is absolute. DENDRITIC DRAINAGE Integrated drainage pattern where small branch channels join – usually at acute angles – to feed a trunk channel. Dendritic drainage shows no preferred orientation, and is typical of areas where the underlying rock is more or less homogeneous. denudation The consequence of erosion; the wearing down of the landscape. DEPOSITIONAL REGIME A grouping of regolith mapping units in regolith-dominated terrain characterised by surficial deposits. These deposits may overlie lateritic residuum, saprolite or bedrock. See also 2%$ 3CHEME. DESERT ARMOUR A desert pavement whose surface of stony fragments protects the underlying finergrained material from further wind erosion. DESERT PAVEMENT The layer of gravel or stones left on the land surface in desert regions after the removal of the finer material by erosion. More commonly termed lag. DESERT VARNISH A glossy sheen or coating on outcrop, stones and gravel in arid regions. The coating is commonly ferruginous, but may also contain Mn and Si. diamicton Massive to poorly sorted sediment with a wide range of particle size; e.g. a mixture of glacially deposited gravel, sand and clay. dispersion (i) Geochemistry: Relocation of geochemical components by transport, either as particles or in solution. (ii) Soil: A phenomenon that may accompany slaking. Dispersion occurs when the soil microaggregates break down into individual clay particles, which drift out of the aggregates and cause the aggregate to have a halo of cloudiness. The more exchange sodium in the clay, the more likely it is that dispersion will occur. dolocrete As for calcrete, but with the carbonate mineral dominated by dolomite.
dolomite White mineral, CaMg(CO3)2 – found in many calcretes; essential component of dolocrete; commonly associated with the weathering of mafic rocks. drawdown (i) The difference between the height of the water table and that of the water in a well. (ii) Reduction of the pressure head as a result of the withdrawal of free water. DRIFT REGOLITH Any unconsolidated rock material, such as boulders, till, gravel, sand, loess or clay, transported from one place and deposited in another. DRYLAND SALINITY The development of salt in the regolith at places removed from the effects of evaporation of bodies of standing or running water. $4- Digital terrain model. A representation on a grid of terrain height above sea level. DTM is absolute; DEM is relative. $UPLEX SOIL Soil profiles with a texture contrast between the A and the B horizons. duricrust Regolith material indurated by a cement – or the cement only – occurring at or near the surface, or as a layer in the upper part of the regolith.. The cement may be, for example, siliceous (silcrete), ferruginous (ferricrete, lateritic duricrust), aluminous (alcrete), gypseous (gypcrete), manganiferous (manganocrete), calcareous (calcrete) dolomitic (dolocrete), salty (salcrete), or a combination of these. DURICRUST CELLULAR Duricrust characterised by irregular to rounded bladder, cell or bubble-shaped voids. May contain pisoliths and or nodules, and show development of a mottled fabric. DURICRUST FRAGMENTAL A duricrust that has a fragmental or blocky fabric in outcrop and/or hand specimens. The interstices between fragments are commonly occupied by a clayey, ferruginous or sandy matrix. The fabric of the parent bedrock may be preserved within the fragments. DURICRUST MASSIVE A duricrust having a homogeneous fabric at the hand-specimen scale. Rarely completely massive; usually contains minor amounts of vesicles and tubules, which may be filled or partly filled with clay and/or other sediment. May either be uniformly coloured or be multicoloured
415
416
Regolith Science
due to segregation of secondary minerals or to selective ferruginisation. Commonly underlies nodular duricrust. DURICRUST PACKED PISOLITIC Consists of pisoliths that are cemented as a framework-supported mass. The interstices between pisoliths are often cavities or are partly filled with clay or sandy clay. The degree of cementation is variable. DURICRUST PARTIALLY CEMENTED Duricrust with less than 70% cemented material, and generally with an open texture; also termed hardpan. dust Solid particles of varying character and size that are carried in suspension in the atmosphere. E EARTHY (i) Soil: Soil material that is coherent and characterised by the presence of pores, few if any peds, and a general floc condition throughout. Ultimate soil particles (sand grains, for example) are coated with oxides and/or clays and are arranged (clumped) around the pores. (ii) Mineralogy: Having the very dull lustre of a lump of earth. eluviation The generally downward movement of material in the regolith – largely from the soil A horizon to the B horizon – carried by water either in solution or suspension through cracks and channels. The term refers especially but not exclusively to the movement of colloids and clays, whereas the term leaching refers to the removal of soluble materials. eolian See aeolian. epilithic Growing on a rock surface. endolithic Growing within a rock (as some algae may do). EROSIONAL REGIME A grouping of regolith mapping units in partly eroded regolith-dominated terrain characterised by outcrop and subcrop of saprolite and/or bedrock. See also 2%$ 3CHEME. evaporite A precipitation of solutes on or near the land surface – typically as lacustrine sediments, or within the regolith; e.g. halite, gypsum. EXFOLIATION The shedding of thin sheets or shells from a weathering boulder or rock face. F FABRIC The physical nature of a regolith unit or component or rock according to the spatial arrange-
ment, orientation (or lack of it), and mutual relationships of the discrete elements of which it is composed, such as particles, crystals, cements and voids. FACET (i) A homogeneous element of a landscape, defined in terms of the slope, material and drainage conditions in which it is developed. The facet is the smallest unit of an ad hoc hierarchical system of land-surface classification, in which combinations of facets are termed landforms and combinations of landforms are termed land systems; e.g. a flat or a slope. (ii) A small, nearly planar surface produced on a rock fragment by abrasion, caused by wind or by the grinding action of a glacier. (iii) Any planar surface produced by erosion or faulting, and intersecting a general slope of the land; e.g. a triangular facet. FAN A gently sloping, fan-shaped mass of detritus forming a section of a very low cone, commonly at a place where there is a notable decrease in gradient. FANGLOMERATE A sedimentary rock consisting of slightly water-worn, heterogeneous fragments of all sizes – originally deposited in an alluvial fan and subsequently cemented into a firm rock – and characterised by a considerable persistence parallel to the depositional strike, but by a rapid down-dip thinning. FERRALL ITE A humid tropical soil, or in situ weathering product, which is formed by the leaching of silica and bases, and characterised by a large content of Fe or Al oxides, or both. Sometimes used synonymously for laterite, latosol, oxisol; not a recommended usage. FERRALLITIC See allite. Ferralsol Humid tropical soil characterised by a high Fe oxide content – formed by the leaching of silica and bases. Commonly (but incorrectly) used synonymously for laterite. FERRAN See cutan. FERRIARGILLAN A cutan of clay and Fe oxide minerals. FERRICRETE An indurated material formed by the in situ cementation of regolith by Fe oxides: mainly goethite and/or hematite. The fabric, mineralogy and composition of the cemented materials may reflect those of the parent (regolith) material. Some
Appendix 1: Glossary of regolith terms
authors restrict the term to the ferruginous horizon of lateritic regolith (and therefore synonymous with cuirasse, lateritic duricrust) but the more general definition is preferred. FERRIHYDRITE Mineral, approximate composition 5Fe2O3.9H2O. Ferrihydrite is the brown rusty scum visible at springs, where water seeps from cracks in rocks, or as an ‘oil slick’ on some swamp water. It is commonly associated with bacteria (Galionella and Lepthotrix: Section 7.4.1). It strongly adsorbs phosphate, silica, organic molecules, and heavy metals. Ferrihydrite generally transforms to goethite over a period of a few years (Section 4.4.3 and Table 12.1). FERROLITHIC Adjective applied to a lag derived from an original Fe-rich lithology. FERROLYSIS Originally used to refer to the gradual destruction of clays in a soil through repeated cycles of replacement of exchange ions by Fe2+ during reducing conditions, followed by oxidation of the Fe, consequent drop in pH, partial dissolution of the clays and the introduction of alumina in the exchange sites. More recently, it has been used to refer only to the decrease in pH caused by hydromorphic oxidation of Fe, according to the equation: Fe2+ + 3H2O = Fe(OH)3 + 3H+ + e –. FERRUGINOUS Pertaining to or containing Fe. Commonly used for regolith having obvious Fe oxides. FERRUGINOUS ZONE Highly weathered, upper part of a regolith profile composed principally of secondary Fe oxides (goethite, hematite and maghemite) and Al hydroxides (e.g. gibbsite). These minerals may incorporate clays and other secondary minerals and resistant primary minerals. The term may be used to encompass any or all of lateritic residuum, cuirasse, lateritic gravels, lateritic duricrust, carapace, plinthite or ferricrete. FIBRIC Peat–organic soil material with virtually all of the organic material remaining, which enables the identification of plant forms. FLOC Aggregate of clay-sized particles in a clay-water system. The product of flocculation. FLOOD PLAIN Alluvial plain characterised by frequently active aggradation by over-bank stream flow (i.e. by flooding more often than every 50 years) and erosion by channelled stream flow.
FOOTSLOPE A general term for the lower, generally concave, part of a hillslope. FRAGIPAN An earthy pan, which is generally loamy. A dry fragment of fragipan slakes in water, whereas a wet fragment does not slake in water, but has moderate or weak brittleness. Fragipans are more stable on exposure than overlying or underlying horizons. FRAGMENTAL DURICRUST See DURICRUST FRAGMENTAL. FULVIC ACIDS High molecular weight organic acids formed by the decomposition of dead plant matter (humus). Fulvic acids have H/C>1 implying a lesser aromatic component than humic acids. G garnierite Green nickel layer silicate of unspecified structure – occurring in weathered nickel-rich ultramafic rocks. gastrolith A rounded stone in or from the gut of an animal. GEOCHEMICAL PATHFINDER Element closely associated with the target mineralisation and which gives an indication of derivation from weathered mineralisation. geode A hollow, or partly hollow, globular or subspherical body – generally over 20 mm in diameter and with a relatively smooth outer surface; the void is lined with crystals, commonly quartz or calcite, but also various sulfides, such as barite, or gibbsite as rods, threads or irregularly shaped plates in geodes in bauxite. The central part of bauxitic geodes may be filled with soft kaolinitic clay or clayey gibbsite. A geode is separable (by weathering) from the rock in which it occurred as a discrete nodule or concretion. GEOMORPHIC PROCESSES Geomorphic processes are those which form or modify landforms. They can refer to either present or past activity. gibber An Australian term for a pebble or boulder, especially one of the wind-polished or windsculpted stones that compose the desert pavement, or the lag gravels of an arid region. gibbsite White mineral, Al(OH)3. Major mineral of bauxite, which is formed by the weathering of kaolinite or of primary Al-rich minerals (Section 4.4.2). Occurs in pisoliths and as earthy deposits. gilgai Surface micro-relief (up to a few metres) consisting of mounds and depressions of varying sizes
417
418
Regolith Science
and spatial distribution, sometimes separated by a subplanar or slightly undulating surface. Gilgai is associated with soils containing shrink–swell clays. ')3 Geographical information system. A data-handling and analysis system based on sets of spatial attributes The data sets are map oriented when they consist of qualitative attributes of an area recorded as lines, points and areas (in vector format), or image oriented, when the data are quantitative attributes referring to cells in a rectangular grid (usually in raster format). glaebule A three-dimensional compound unit within the matrix of a soil material, or occurring as a discrete physical fabric element – generally approximately equant or prolate in shape, and with a sharp boundary. GLEY Greyish, bluish or greenish coloured soil commonly produced under poor drainage; indicative of chemically reducing conditions. goethite Yellow-brown mineral, a-FeO(OH), with common substituents Al, Mn, Ni; polymorphous with lepidocrocite. The most common Fe oxide mineral in the regolith; common as cutans on ferruginous nodules. Very high surface area, resulting in high sorptive capacity for cations (e.g. Cu, Pb and Zn) and anions, especially phosphate (Appendix 2). Thus, it can be useful as a geochemical sampling medium in mineral exploration (Chapter 13). Substitution of Al for Fe reaches 30 mole % in goethites formed in hydromorphic regolith (Section 4.4.3; Figure 12.2; Table 12.1). gorceixite See crandallite. gossan The weathered expression of rocks that contained substantial sulfide mineralisation. Gossans derived from Fe-bearing sulfide assemblages typically consist largely of Fe oxides and are a form of ironstone, whereas gossans formed from the weathering of Fe-poor sulfides (e.g. carbonate-hosted Pb-Zn deposits) give Fe-poor gossans; such gossans may be siliceous or have a high manganese content. Gossans commonly exhibit a boxwork fabric derived from that of their sulfide precursors. The term gossan has no economic connotation (Section 13.4). GOSSAN FALSE A discrete ferruginous outcrop (ironstone) with a fabric and/or composition suggestive
of a gossan, but not developed over sulphides (see pseudo-gossan). 'RADATIONAL SOILS Soils with a profile dominated by the C horizon and showing gradual or diffuse boundaries on passing from the B into the A horizon. granular Said of the texture of a rock that consists of mineral grains of approximately equal size, generally granule-sized. GRANULE A rock fragment having a diameter in the range of 2–4 mm and being somewhat rounded or otherwise modified by abrasion in the course of transport. GRAVITY SLOPE The slope at the surface of debris eroded from a cliff or rock outcrop where the debris has fallen and tumbled to its resting place. The slope of the surface is influenced by the sizes of the particles; steeper slopes are associated with coarser debris. 'REAT SOIL GROUP A soil classification category in which soils are classified according to their mode of formation as reflected in major morphological characteristics and profile form. GREYBILLY GREY BILLY Colloquial term for silcrete. GROUNDWATER FERRICRETE Iron-oxide accumulation and induration developed at the surface of the groundwater. There may be a relation between its thickness and fluctuation in the level of the water table. Sometimes called ‘groundwater laterite’. The terms are based on poorly understood genetic processes and therefore not recommended for field description. groundwater Loosely, all sub-surface water. Formally, sub-surface water in the saturated zone; i.e. below the water table. See also VADOSE PHREATIC groutite Black mineral, a-MnO(OH). GROVEnINTERGROVE PATTERN A striking vegetation pattern of alternating lines of woodland about 4 m high and strips of bare or sparsely vegetated ground, found in semi-arid rangeland areas in West Africa, North America and Australia (where it occurs in mulga woodlands). In West Africa it occurs on shallow gravelly soils of very low slopes (<1%) on plateau lands formed by a thick lateritic curaisse of Tertiary age. Also termed ‘brousse tigree’. grus Fragmented disintegration product of largely unweathered coarse-grained igneous rocks, espe-
Appendix 1: Glossary of regolith terms
cially granite. Commonly applied to surface products, but also to a porous horizon ranging from a few centimetres to 10 m or more thick at the base of saprolite. In the French literature, ‘grus’ or ‘arene’ is used to designate such horizons over any lithology. Grus is saprolite: distinctive in that it is friable, gravelly or sandy, rather than compact. GYPCRETE Duricrust cemented mainly by gypsum. GYPSUM White mineral, CaSO4.2H2O, which is bladed, fibrous or powdery. Very common evaporite mineral that is found across arid Australia in saline lakes and soils. Important as an industrial mineral and soil modifier. H hematite Mineral, a-Fe2O3, with hexagonal closepacked structure. Very common in regolith of warm or arid regions. Colour black or blue-black where massive; red where fine-grained and dispersed. Its intense colour may mask the presence of goethite. Hematite has similar adsorption properties to goethite and can also be responsible for the fixation of P, Cu, As and other elements (Appendix 2; Section 4.4.3 and Table 12.1). hemic Peat–organic soil material intermediate in degree of decomposition between the less decomposed fibric and the more decomposed sapric material. halite White mineral, NaCl, common salt. Very common evaporite mineral, which is found across arid Australia in saline lakes, playas and regolith. HALLOYSITE Clay mineral, Al2Si2O5 (OH)4.H2O, which is similar to kaolinite in chemistry, properties and occurrence, but with cylindrical or spherical morphology. hardpan A relatively hard, impervious layer in the regolith lying at or near the surface. It offers great resistance to digging or drilling, and hampers root penetration and downward movement of water. Its hardness does not change appreciably with changes in moisture content, and it does not slake or become plastic when mixed with water; it can be shattered mechanically or by explosives. It is produced as a result of cementation of soil particles by precipitation of relatively insoluble materials – most commonly silica, with some Fe oxide, calcium carbonate, and organic matter. Red-brown
hardpans, such as the Wiluna Hardpan (WA), consist of a variety of transported or residual host materials, including soil, colluvium, pisolitic horizons and brecciated saprolite, set in a porous, redbrown, earthy matrix, cemented by silica (generally hyalite), clay, and Fe oxides. This material has a coarsely laminated appearance and commonly has Mn oxides on partings. hisingerite Clay mineral – brown, vitreous, almost amorphous, Fe3+ – end member of the kaolin group: Fe2Si2O5(OH)4. Low temperature alteration product of pyrite, olivine and other Fe-rich minerals. HONEYCOMB WEATHERING A type of chemical weathering in which innumerable pits are produced on a rock exposure. The pitted surface resembles an enlarged honeycomb and is characteristic of finely granular rocks, such as tuffs and sandstones, in an arid region. HORIZON (i) Soils: A layer of soil approximately parallel to the land surface, with morphological properties different from layers below and/or above it. (ii) Regolith: A layer within the regolith, approximately parallel to the land surface, with field observable properties different from layers below and/or above it. HUMIC ACIDS High molecular weight organic acids formed by the decomposition of dead plant matter (humus). Humic acids have H/C=1, implying an higher aromatic component than fulvic acids. humus The decomposed organic material in soil. Humus gives surface soil horizons their dark colours. HYALITE Mineral – a glassy form of opaline silica. HYDROLYSIS Chemical reaction between a solid and water. Hydrolysis changes both the solid and the water; e.g. the hydrolysis of anorthite may yield kaolinite and release lime. CaAl2Si2O8 + 3H2O = Al2Si2O5 (OH)4 + Ca(OH)2 HYDROMORPHIC Formed under conditions of water saturation. HYDROTHERMAL ALTERATION Alteration produced by chemical changes in rock materials caused by hot water and steam rising through country rock. This is not weathering, but produces very similar effects. The best field distinction between clay bodies formed by weathering and hydrothermal alteration is that weathering decreases with depth,
419
420
Regolith Science
and hydrothermal alteration commonly increases with depth. HYPERSPECTRAL Reflectance spectrometry over the 350–2500 nm wavelength interval, which is used to identify mineralogical components in regolith samples, especially Fe oxides and layer silicates. I iddingsite Petrographic term for a pseudomorphic alteration product of olivine; less commonly of pyroxene. Composed of goethite and minor smectite; red-brown, cryptocrystalline; commonly as rims or crack fillings or as complete replacement of the primary mineral. illite White clay mineral – approximately K0.75 (Al1.75Mg 0.25)(Si3.5Al0.5)O10 (OH)2 . Regolith illite forms chiefly by the weathering of muscovite and feldspar, but the mineral is better known as a diagenetic mineral in buried marine sediments. illuviation The process of deposition or accumulation of soil material that has moved from one horizon to another within the soil – generally from an upper to a lower horizon within the profile, but also laterally within a toposequence. Refers particularly to the transportation of material in suspension, especially colloidal particles. Can also apply to the transport of soluble material. in situ In its original place, e.g. in situ regolith INCONGRUENT SOLUTION Dissolution of a solid accompanied by reaction with the liquid so that ions in solution and a new solid are produced. e.g. anorthite ® kaolinite: CaAl2Si2O8 + 3H2O = Al2Si2O5 (OH)4 + Ca(OH)2 solution. INDURATED MATERIAL Regolith material that has been hardened and/or cemented. Indurated material can be further described by a prefix according to the dominant indurating material – bauxitic, calcareous, clay, ferruginous, gypsiferous, siliceous or humic. See also duricrust. induration The hardening of a rock, rock material or regolith by the action of heat, pressure, or the introduction of some cementing material not commonly contained in the original mass: especially the process by which relatively consolidated rock is made harder or more compact; e.g. the development of a hardpan or duricrust.
inselberg Range, ridge, or isolated hill that stands abruptly from the surrounding plains, like an island from the sea. It is characterised by steep slopes that meet the adjacent plain in a sharp, almost angular, junction; e.g. Uluru (Ayres Rock), Kata Tjuta (The Olgas). INSOLATION WEATHERING Insolation weathering occurs when varying insolation induces temperature changes, which cause expansion and contraction of rocks. Repeated temperature changes or rapid change causes rocks to fracture. Insolation weathering is recognised where rocks at the surface consist of interlocking angular fragments. INTERFLUVE The area between rivers; especially the relatively undissected upland or ridge between two adjacent drainage basins. INTERRUPTED DRAINAGE Drainage where the channel segments are short and unconnected. Typically this occurs in karst landforms, and in areas where the drainage pattern has not been fully integrated. Some parts of the arid centre of Australia show this pattern because of the lack of sufficient precipitation and disruption of drainage lines by wind blown materials. INTRAZONAL SOILS Soils with more or less well-developed characteristics that reflect the influence of some local factor other than climate, such as parent material, hydrology and relief. ION EXCHANGE CAPACITY The total amount of particular material’s exchangeable ions. It is commonly determined for soil or clay (Table 5.3). ION EXCHANGE The replacement of a weakly bonded ion of any solid (commonly a surface or interlayer cation of a clay or the alkaline cation of a zeolite) by an ion from solution. IRON SEGREGATIONS Dark, non-magnetic and goethiterich Fe enrichments within ferruginous saprolite or the upper saprolite – occurring as pods, lenses and large slabs. They lack cutans and range in size from 0.1 m to 25 m. ironstone Highly ferruginous weathered material consisting mainly of Fe oxides (Section 13.4; Figure 13.3). Examples are: I a part of a laterite profile, essentially conformable with the land surface; i.e. lateritic ironstone or duricrust;
Appendix 1: Glossary of regolith terms
(ii) essentially linear outcrop following an underlying geological unit or structure (i.e., a stratigraphic ironstone); (iii) ferruginous gravels composed of a majority of ferruginous grains, which may include hematitic and or goethitic pisoliths or nodules of irregular shape such as are released by erosion of the ferruginous and mottled zones of a weathering profile; (iv) gossans. IRONSTONE LATERITIC See CUIRASSE DURICRUST. IRONSTONE LEAKAGE Ironstone that forms where dissolved Fe precipitates. The Fe – originally derived from the weathering of Fe-bearing minerals – is precipitated at depth along joints, lithological contacts and faults, and at the surface in seepage areas and in drainages (Figure 13.3). isalterite French term, synonymous with saprolite. J jarosite–natrojarosite Mineral, KFe3 (SO4)2 (OH) 6 – NaFe3 (SO4)2 (OH) 6. Members of the alunite family (see ALUNITE SUPERGROUP). Formed in a reaction between sulfuric acid formed by pyrite oxidation and surrounding silicates. These minerals are common in regolith where pyrite is weathering, and are particularly so in acid sulfate soils and mine dumps. jumpup See BREAKAWAY. K kandite Mineral group term for the varieties of Al2Si2O5 (OH)4, comprising kaolinite, dickite, nacrite and halloysite (synonymous with kaolin: Section 4.3.5). The word has not been widely used in the mineralogical literature. KANKAR KUNKAR Term for calcrete. kaolin (i) Al2Si2O5 (OH)4-minerals. (ii) An unconsolidated rock in which the kaolin (kandite) minerals represent *80% of the minerals. It is generally a soft, fine, white, earthy, nonplastic material. kaolinite Clay mineral, Al2Si2O5 (OH)4. Major component of regolith, particularly in the plasmic zone of weathering profiles (Section 6.2). Formed by the weathering of aluminosilicate minerals, primarily plagioclase and muscovite (Section 4.3.5). Important as an industrial mineral.
KARST BAUXITE Bauxite occurring in limestone karst. karst Landform pattern, most commonly on limestone, of unspecified relief and slope (for specification use terms such as karst rolling hills) – typically with fixed, deep erosional stream channels forming a non-directional disintegrated tributary pattern and many closed depressions without stream channels. It is eroded by continuously active solution and rarely active collapse, with the products being removed through underground channels. knickpoint Any interruption or break of slope in the longitudinal profile of a stream or of its valley, especially a point of abrupt change or inflection, resulting from rejuvenation, glacial erosion, or the outcropping of a resistant bed. L LACUSTRINE PLAIN Level landform pattern with extremely low relief, which was formerly occupied by a lake, or lakes, but now partly or completely dry. lag A general term for a surface accumulation of materials of diverse origin, such as regolith, rock, and mineral particles, with most being in the granule to cobble range (2 to 256 mm); resulting from removal of finer material by pluvial and aeolian processes, or by matrix removal as a result of differential weathering. The type of lag present in an area is partly a function of the local regolith, landform and bedrock. LAMINATED CALCRETE Hard, finely laminated sheet consisting of layers of authigenic calcite or dolomite. The laminar zone may display wavy bands. Rarely exceeds 1 m in thickness. LAND SYSTEM A repeated association or pattern of landforms characteristic of a terrain. LANDFORM Any physical, recognisable aspect or feature of a planet’s surface, having a characteristic shape, and produced by natural processes. LANDFORM ELEMENT Area in the landscape of the order of 40 m or more across, such as a footslope, river flat or cliff face. See LANDFORM PATTERN. LANDFORM PATTERN Area in the landscape more than 600 m across – made up of landform elements. landscape All the natural features of land or territory encompassed in a single view, such as fields, hills, forests and water, which distinguish one part of a planet’s surface from another part.
421
422
Regolith Science
laterite Generally, regolith exhibiting the characters of all, or at least the upper part, of a laterite profile. Laterite, or lateritic regolith – commonly has a hard, more or less prominent, ferruginous surface expression, with some degree of chemical and mineralogical differentiation below, characterised by varying colour reflecting varying Fe and silicate distribution. Laterite is the product of weathering. LATERITE PROFILE A vertical sequence of regolith facies showing (or reasonably interpreted to have, if hidden) some or all of the following, from the bottom up: bedrock, saprock, saprolite, plasmic zone, mottled zone or ferruginous saprolite and lateritic residuum (lateritic duricrust, lateritic gravel). LATERITIC RESIDUUM A collective term for the ferruginous part of a laterite profile, which is composed dominantly of oxides of Fe or Al (goethite, hematite, maghemite, or gibbsite, boehmite) with or without quartz. The term includes both fragments and duricrust developed essentially by residual processes and, therefore, has a broad genetic and/ or compositional relationship with the substrate. Where a duricrust is present, the fragments commonly overlie the duricrust. Note: These materials have formed by both vertical and some minor (of the order of 5–50 metres) lateral movement of clasts and are therefore better regarded as residual rather than in situ. LATERITIC SOIL A term used by soil scientists for a suborder of zonal soils (great soil group) formed in warm, temperate and tropical regions. leaching The removal of soluble materials from one zone in soil to another via water movement in the profile. See also eluviation. LEAKAGE IRONSTONE See IRONSTONE LEAKAGE. lepidocrocite Orange mineral, g-FeO(OH). Polymorph with the oxygens in approximate cubic close packing (Section 4.4.3). Lepidocrocite is a relatively uncommon mineral – forming in preference to goethite as a direct oxidation product of ferrous Fe and in preference to ferrihydrite if oxidation is slow. It also seems to be precipitated rather than goethite in the presence of Cl–. leucoxene Petrographic term for cream-coloured, high brightness alteration product of titanium-
bearing mafic minerals such as ilmenite. Leucoxene is a mixture of minerals, including rutile, pseudorutile (Fe2Ti3O9), anatase, Fe oxides and clay minerals. LIESEGANG RINGS (weathering) Nested rings or bands of yellow/brown/red colour in weathered rocks – generally caused by the precipitation of Fe oxides from solution. limonite Earthy, ochreous mineral aggregate (rather than a mineral). It is largely composed of goethite, with variable amounts of lepidocrocite, hematite, clay silicates and other fine-grained minerals. lithomarge (French) Compact, massive, generally kaolinitic clay. It has been applied in the French and Indian literature to clay-rich zones of the regolith, particularly in the upper saprolite. lithophile Elements that generally concentrate in rock-forming minerals (e.g. Na, K, Si, Al, Mg and Ca (Section 5.2). lithorelic A weathered fragment of rock. lithorelict An unweathered fragment of rock in an assemblage of secondary minerals. Lithosol Skeletal soil. litter Freshly fallen plant (or animal) matter on the ground surface. LITTER DAM Litter that has moved across the ground surface until impeded by other litter or the decreased ability of water to transport it (Section 8.3.2). Litter dams may be useful in determining slopes on very flat ground. lixiviation Mechanism of subtraction of material by congruent dissolution. ,-7/! Low molecular weight organic acids. These typically contain 1–6 carbon atoms and 1–3 carboxyl groups. Examples are ascorbic, aspartic, hydroxybenzoic, tannic, malic, oxalic and citric acids – with the last three being di-carboxylic LMWOAs. loam Soil with 10–30% clay and approximately equal amounts of sand and silt (Section 12.3.3; Figure 12.3). loess Material transported and deposited by wind and consisting of predominantly silt-sized particles. M maghemite Black mineral, g-Fe2O3 – spinel. Maghemite can form by the oxidation of magnet-
Appendix 1: Glossary of regolith terms
ite, and by the dehydration of goethite or lepidocrocite during fires. Lepidocrocite can transform easily to maghemite, as both have cubic closepacked structures. Maghemite is strongly magnetic (Section 4.4.3; Table 12.1). magnesite White mineral, MgCO3 – found particularly in veins, sheets, nodules or bodies up to the size of a cauliflower in regolith over ultramafic rocks. malachite Bright green mineral, Cu2CO3 (OH)2 – weathering product of copper minerals such as chalcopyrite (Table 4.2). See also AZURITE. mangan (i) Prefix denoting manganese (ii) A cutan consisting of manganese oxides or hydroxides. MANGANESE OXIDES Minerals common in regolith – particularly in weathering profiles of ultramafic rocks, or manganese-bearing rocks and base metal deposits; e.g. Broken Hill, New South Wales, Groote Eylandt, Northern Territory. Typically, a variety of species occur together in black, very finegrained aggregates referred to as ‘wad’ if soft, or ‘psilomelane’ if hard. Minerals include: asbolane (Co-Ni-Mn oxyhydroxide) birnessite ((Na, K)4Mn14O27.9H2O) cryptomelane–coronadite–hollandite group (K, Pb, Ba)2-1Mn8O16) chalcophanite (ZnMn3O7.3H2O) lithiophorite ((Al, Li)MnO2 (OH)2) groutite (a-MnO(OH)) manganite (g-MnO(OH)) nsutite (Mn4+1-x Mn2+xO2-2x(OH)2x pyrolusite (b-MnO2) romanchèite (Ba,H2O)2 (Mn4+,Mn3+)5O10) todorokite ((Na,Ca,K) 0.3-0.7(Mn,Mg) 6O12. nH2O) vernadite (d-MnO2). Besides elements listed as essential components, significant abundances of geochemical pathfinder elements may also be present (Appendix 2). manganite Black mineral – g-MnO(OH). manganocrete Duricrust cemented mainly by manganese oxides. massive Homogeneous, without visible internal fabric. megamottles Mottles greater than 200 mm in mean diameter.
MICROAGGREGATION FRONT A transformation (front) associated with a change to a microaggregated fabric in friable soil. MICRORELIEF Small-scale, local differences in topography, including mounds, swales or pits that are only a metre or so across with elevation differences of up to 2 metres. microtubular A void consisting of sinuous tubes with a diameter of <0.5 mm – occurring in otherwise more or less massive and fine-grained rock or soil. minimottle Mottle <10 mm in mean diameter. monosiallitisation The chemical change of silicates to 1:1 layer silicates such as kaolinite, halloysite (Section 4.3). montmorillonite White clay mineral of the smectite group, approximately A0.7(Al1.3Mg0.7)[Si4] O10 (OH).2H2O, where A is an exchangeable cation, such as K+, Na+, or 0.5Ca2+. Montmorillonite is less common the than the other aluminous smectite, beidellite. (Formerly used also for the smectite group.) mottle Segregation of subdominant colour different from the surrounding region’s colour. In regolith, mottles may have sharp, distinct or diffuse boundaries. They typically range in size from 10–100 mm, but may reach several metres in size. Larger mottles (>200 mm) have been termed megamottles, and those <10 mm minimottles. MOTTLED ZONE Part of a weathering profile showing mottles. Where present in a laterite profile, it underlies the lateritic residuum and is commonly above the plasmic zone (Figure 6.1). mulch Loose material on the soil surface – generally decaying plant material or small pieces of soil. See also SELF MULCHING -UNSELL COLOUR SYSTEM A colour designation system specifying the relative degrees of the three simple variables of colour: hue, value, and chroma (Section 12.3.1; Figure 12.1). MYCORRHIZAE The fungi which have a symbiotic association with the roots of a plant. N NEOFORMED Newly formed; authigenic. neotectonics The study of the structures and structural history of the Earth’s crust, after the Miocene
423
424
Regolith Science
and during the later Tertiary and the Quaternary. Most neotectonic features are produced by vertical movements. nepouite Green mineral – nickel equivalent of the serpentine mineral, lizardite. nitre White evaporitic mineral, KNO3 – found in very arid climates. nodule (regolith) A lump of regolith – generally pebble-sized, not rock, different from its immediate surrounds. Nodules are distinguished from other similar-sized bodies by having the following characters: they may have an irregular but commonly smooth shape and, generally, although not necessarily, have a contrasting cortex, but are more uniform inside, except they may have either a nucleus or a central cavity. As the sphericity of the smaller nodules in a weathering profile increases, the terms nodule and pisolith merge. nontronite Green clay mineral of the smectite group, approximately A0.7(Fe23+)[Si3.3Al0.7]O10. (OH)2.4H2O, where A is an exchangeable cation, such as K+, Na+, or 0.5Ca2+. Nontronite is a common weathering product of Fe-bearing silicates, and is the main colouring mineral in ‘altération pistaché’. nucleus A central part; e.g. the central part of a pisolith or nodule; it may be of similar or contrasting shape, chemistry and mineralogical composition to the cortex. O ochre A strongly coloured, generally earthy, regolith material – typically red, yellow, orange or brown, and composed of clays and Fe oxide minerals. ooid An individual spherite of an oolitic rock; an oolith. The term has been used in preference to ‘oolith’ to avoid confusion with ‘oolite’. oolite Rock composed of ooliths. oolith A small (0.25 to 2 mm), round (ovate, spherical, or oblate elipsoidal) body in the regolith or in a sedimentary rock, in aggregation resembling the roe of fish. Regolith ooliths may be composed of gibbsite, hematite or goethite, or combinations of these minerals. Sedimentary ooliths are generally formed of calcium carbonate (but may be of dolomite, silica, Fe oxide, pyrite, or other minerals). An
oolith typically has successive concentric layers commonly around a nucleus (such as a shell fragment, an algal pellet or a quartz-sand grain). opal Silica mineral, SiO2.nH2O – amorphous to poorly crystalline. Electron microscopy shows a microstructure of packed silica spheres. Precious opal displays an iridescent play of colour; common opal is varicoloured, vitreous, glassy (hyalite) or translucent. Commonly formed from solution and re-precipitation of silica in regolith; in Australia, notably within late Cretaceous and Cainozoic deeply weathered profiles which developed in the Eromanga Basin sequence of South Australia, Queensland and New South Wales, in red-brown hardpans. Also found as late stage deuteric alteration within volcanic rocks (Section 4.4.1). /RGANIC SOIL Soil with a profile dominated by plant remains – the organic fraction – such that the surface 30 cm contains 20% or more organic matter if the clay content of the fine earth is 15% or lower, or 30% or more organic matter if the clay content of the fine earth is higher than 15%. The organic matter should equal, or exceed, the stated amounts at all depths down to, and including, the 30 cm. OVER BANK DEPOSIT Alluvium deposited outside an alluvial channel from flowing water that has overflowed from the channel. It includes levees and back swamp deposits. overburden (i) Economic geology: Barren rock material – generally unconsolidated – overlying a mineral deposit, which must be removed prior to mining. (ii) Sedimentary/regolith: The unconsolidated material overlying bedrock – either transported or formed in place. OXBOW LAKE See billabong. OXYHYDROXIDE Minerals in which the anions are both O and (OH): most commonly goethite, lepidocrocite (FeO(OH)), boehmite, diaspore (AlO(OH)), and groutite (MnO(OH)). (Generally included in terms ‘Fe oxides’ etc. in this book.) P paleosol Soil formed under environmental conditions different from those of the present. May be buried.
Appendix 1: Glossary of regolith terms
pallid In the context of a weathering profile: very pale to white, lacking pigmentation (cf. bleached). PALLID ZONE A zone or portion of a weathering profile that lacks colour. Its position in the profile and its mineralogy may vary, but the zone is generally dominated by kaolin ± quartz. In a laterite profile, the pale-coloured region below the mottled zone, incorporating parts of the plasmic zone and or saprolite (Figure 6.1), has been referred to as the pallid zone. PALYGORSKITE White fibrous mineral, approximately (Mg,Al)5Si8O20 (OH)2.4H2O – generally found in arid, evaporitic environments. pan An indurated or cemented soil horizon. See also hardpan. parna Silt- and sand-sized aggregates of mixtures of silts and clays. The aggregates may be bonded internally by clays, Fe oxides, carbonates, gypsum or other salts. See also loess. PATTERNED GROUND Land surface with distinctive arrangement of stones or microtopography. Patterned ground can be due to the effects of ground freezing and seasonal frost; characteristic of periglacial environments. Such patterned ground includes stone stripes, sorted circles and tundra polygons. ped A unit of soil structure – such as an aggregate, crumb, prism, block or granule – formed by natural processes (in contrast with a clod, which is formed anthropogenically). pediment Gently inclined to level landform pattern of extremely low relief, underlain by bedrock, characteristically lying down-slope from adjacent hills with markedly steeper slopes. pediplain Level to very gently inclined plain with extremely low relief and no stream channels – formed by the coalescence of pediments. pedogenesis Soil formation. pedolith Upper part of the regolith, above the pedoplasmation front, which has been subjected to soilforming processes resulting in the loss of the fabric of the parent material and the development of new fabrics (Figure 6.1). PEDOPLASMATION FRONT Transformation front at which the lithic fabric is destroyed, although com-
monly with little chemical reworking. It forms the boundary between the saprolith and pedolith in deeply weathered profiles (Figure 6.1). pedoturbation Mixing of soil components by biological activity (bioturbation: Section 8.3) and by shrink–swell processes. peneplain Level to gently undulating landform pattern with extremely low relief – formed from slope decline by the processes of long-continued subaerial erosion. periglacial An environment in which frost action is an important factor. PERMAFROST The state of natural materials that exist at below freezing temperatures for a relatively long period of time (*2 years). phreatic Of sub-surface water, that which is below the water table. Formally equivalent to groundwater. PHYSICAL WEATHERING Weathering without chemical change. The break-up of rocks, rock fragments and minerals by physical force, such as gravity, heating/ cooling and crystal growth. PHYTOLITHS Microscopic crystalline forms made up of minerals within various parts of plants. 0)-!® Portable infra-red mineral analyser. A fieldportable reflectance spectrometer operating over the 1000–2500 nm wavelength interval. It is useful to rapidly identify layer silicates in regolith samples (Section 4.8). pisoidal Texture of a rock composed of pisoliths. pisolite A sedimentary rock made up chiefly of pisoliths cemented together. pisolith A spherical or ellipsoidal body resembling a pea in shape and limited in size to between 2 mm and about 64 mm in diameter. It may have a concentric internal structure, but concentric lamination is not diagnostic; however, most pisoliths have an outer cortex or skin. placer A mineral deposit formed by the accumulation of weathering resistant minerals, commonly in alluvium or on a shore. Most placer deposits are of dense, durable minerals such as diamond or cassiterite. plasma All the material of regolith – inorganic and organic, crystalline and amorphous – of clay size (<0.002 mm).
425
426
Regolith Science
PLASMIC HORIZON Plasmic zone with essentially horizontal boundaries. PLASMIC ZONE Mesoscopically homogeneous part of a weathering profile, made up of clay or silty clay, which has neither the lithic fabric of saprolith, nor the significant development of secondary entities such as nodules or pisoliths (Figure 6.1). Major structural features, such as quartz veins and lithological contacts, may be preserved – possibly with change in orientation. PLAYA Vegetation-free, flat area at the lowest part of an undrained desert basin, underlain by stratified clay, silt or sand, and commonly by soluble salts, dry most of the time. plinthite Iron-rich, humus-poor mixture of clay and quartz, commonly occurring as mottles, firm but uncemented when moist, but hardening irreversibly into an ironstone hardpan or irregular aggregates on repeated wetting and drying. Used synonymously for lateritic duricrust or cuirasse by some authors. plumbogummite Mineral, PbAl3H(PO4)2 (OH) 6. Member of the alunite supergroup. Commonly formed as weathering product of galena (Table 4.2), with the phosphate commonly derived from apatite. See crandallite. plumbojarosite Mineral, Pb0.5Fe3(SO4)2(OH) 6. Member of the alunite supergroup. Commonly formed as weathering product of galena (Table 4.2). PODZOLIC Group of acid soils with a strongly differentiated profile with a bleached sub-surface horizon overlying an horizon that is rich in sesquioxides relative to those lying above and below it. POINT OF ZERO CHARGE pzc. The pH at which the number of protonated and de-protonated surface sites on a mineral surface are equal (Section 5.3.7; Table 5.3). POLYGON In regolith mapping, it is the term used for an area on a map bounded by a line. porcellanite A hard, dense, siliceous rock having the texture, dull lustre, hardness, fracture or general appearance of unglazed porcelain; it is less hard, dense and vitreous than chert. The term has been used for: (i) various kinds of rocks, such as an impure chert, in part argillaceous and in part calcareous, or more rarely, sideritic
(ii) an indurated or baked clay or shale with a dull, light-coloured, cherty appearance, often found in the roof or floor of a burned-out coal seam (iii) a fine-grained, acidic tuff compacted by secondary silica, or a silica-replaced silt- or clay-stone. In Australia, the term is commonly applied to silicified clay- or mud-stone. POROSITY The amount of pore space present, expressed as a percentage of the total volume of the material. protolith General term for parent material from which the regolith has formed. pseudogossan See GOSSAN FALSE. pseudomorph A mineral or mineral aggregate that has the outward shape of a former mineral species or aggregate that has been replaced through alteration; e.g. goethite after pyrite. psilomelane Field term for hard aggregates of manganese oxides. Formerly applied to the mineral now known as ‘romanchèite’. See also MANGANESE OXIDES WAD Q 1!: CEMENT Acronym for quartz-anatase-zircon cement: the cement of silcretes on some granites and quartzites. R 2 HORIZON The continuous masses (not boulders) of moderately strong to very strong rock, such as saprolite and saprolith, at the base of a soil profile (Section 6.2; Figure 6.3). R horizons may have cracks, but these are few enough and/or fine enough that few roots penetrate and there is no significant displacement of rock. It is usually too strong to dig with hand tools, even when moist. 2%$ SCHEME Acronym for relict, erosional and depositional. A means of interpreting factual regolith maps of deeply weathered terrain, which was initially developed for application to geochemical exploration on the Yilgarn Craton, Australia (Anand et al. 1993). It is based on the concept of a landscape that was characterised by an extensive blanket of lateritic residuum and that has been modified by erosion and deposition. This concept is somewhat of an over-simplification, because the lateritic residuum probably did not form a widespread, continuous
Appendix 1: Glossary of regolith terms
unit on a peneplained surface, but a discontinuous cover on a broadly undulating plateau. Nonetheless the scheme provides a practical guide for geochemical sampling and interpretation and has application in equivalent terrains elsewhere. redox Abbreviation for reduction–oxidation. The term is also used to refer to the oxidation state (Eh or pe) of material, as in ‘the mineral assemblage allows the redox conditions to be estimated’ or to describe a reduction–oxidation interface; i.e. ‘redox boundary’ (Section 5.3.1; Figure 5.6). reductomorphic Formed under chemically reducing conditions, such as generally exist below the water table or in gleyed soils. regolith (i) The entire unconsolidated or secondarily recemented cover that overlies more coherent bedrock, that has been formed by weathering, erosion, transport and/or deposition of the older material. The regolith thus includes fractured and weathered basement rocks, saprolites, soils, organic accumulations, glacial deposits, colluvium, alluvium, evaporitic sediments, aeolian deposits and ground water (Section 1.1). II Everything from fresh rock to fresh air. REGOLITHnLANDFORM REGIMES Broad genetic groupings of some landforms and their associated regolith. They may form the basis of regolith–landform models, particularly for weathered terrain. In these models, the development of an extensive deeply weathered mantle is proposed as the first stage, and this is subsequently modified by erosion and deposition. In broad terms, three major regimes are perceived as being widely applicable in lateritic terrain: namely, relict, erosional and depositional. See also 2%$ SCHEME. REGOLITHnLANDFORM UNIT (RLU). A land area characterised by similar landform and regolith attributes; it refers to an area of land of any size that can be isolated at the scale of mapping. REGOLITH STRATIGRAPHY The organisation of regolith units in space and time. REGOLITH TERRAIN UNIT 245 Old name for regolith landform unit. REGOLITH UNIT A subdivision of the regolith, which is generally mappable and has visibly distinguishable boundaries, unless defined outside the visible spectrum using remotely sensed electromagnetic or
chemical data. The term may be used for zones or horizons of weathering profiles, such as soil, duricrust, gravel, mottled regolith and saprolite, or mappable entities with boundaries associated with a change in landform. rejuvenated Said of a stream that has had its erosive power increased. RELICT REGIME A grouping of regolith mapping units in regolith-dominated terrain that are characterised by the occurrence of lateritic residuum at or close to the surface. (The term ‘relict’ implies that these units represent an ancient weathered surface, but the grouping is based only on factual observation). Is also called residual regime, but relict is preferred. See 2%$ SCHEME. residual Left in its original place. Residual regolith results from the weathering of rock without significant lateral movement of the solid weathered products. RESIDUAL REGIME See RELICT REGIME. residuum That which is left after weathering; e.g. lateritic residuum. resistate Mineral more resistant to weathering than most; e.g. cassiterite, ilmenite, spinel, zircon and chromite. These minerals may not be completely dissolved during acid digests (Appendix 2). REVERSED DRAINAGE The reversal of flow in a river channel; can occur as a result of tectonic tilting – causing lowering of the headwaters of a river system. Sometimes river capture can lead to reversed drainage. reworked Geological materials that have been displaced from their place of origin and incorporated in a still recognisable form in a younger formation. RHIZOCONCRETION A root-shaped, solid or hollow concretion – cylindrical or conical in shape – generally branching. RHIZOMORPH see RHIZOCONCRETION. RHIZOSPHERE The soil in the immediate vicinity of the plant roots in which the abundance or composition of the microbial population is affected by the presence of roots. rill A channel <0.3 m deep resulting from erosion. Shallow gully. rises Landform pattern of very low relief (9–30 m) and very gentle to steep slopes.
427
428
Regolith Science
RLU Regolith landform unit. ROOT CHANNEL A tubular void of up to a few centimetres diameter and up to several metres long, generally more-or-less vertical and fairly straight; remnants of the original root may still be present, or the cavity may be filled with regolith matter such as grains, small pisoliths or clay. 245 Regolith terrain unit. S salcrete Duricrust cemented mainly by halite. SALINITY Accumulation of salts at the surface or within the near-surface soil. This can arise from a number of causes, ranging from a rise in water table levels in irrigated areas to emergence of subsurface water in lower footslope areas. Also called dryland salinity or groundwater-associated salinity (Section 12.5). saponite Mineral of the smectite group – A0.7Mg3[Si3.3Al0.7]O10.(OH)2.4H2O, where A is an exchangeable cation, such as K+, Na+, or 0.5Ca2+. Alteration product of magnesium silicates (Section 4.3.1). sapric Peat–organic soil material in which virtually all of the organic material is decomposed, which does not enable identification of plant forms. saprock Compact, slightly weathered rock with low porosity; defined as having less than 20% of weatherable minerals altered but generally requiring a hammer blow to break. Weathering effects are present mainly at the micro-sites of contacts between minerals and intra-mineral fissures, along shears and fractures through the rock as a whole, or affecting only a few individual mineral grains or mineral species (Section 6.2; Figure 6.1). saprolite Weathered bedrock in which the fabric of the parent rock – originally expressed by the arrangement of the primary mineral constituents of the rock (e.g. crystal or grains) – is retained. Compared with saprock, saprolite has more than 20% of weatherable minerals altered (Section 6.2; Figure 6.1). Saprolite is commonly the material referred to as the C horizon in pedology (Figure 6.3). saprolith The saprolith is the (generally lower) part of the regolith that has retained the fabric of
the parent rock. That is, saprock plus saprolite (Figure 6.1). scald Flat area, bare of vegetation, from which soil has been eroded or excavated by surface wash or wind. SELF MULCHING The condition of well-aggregated soil in which the surface layer forms a shallow mulch of soil aggregates when dry. Aggregation is maintained largely as a response of the clay minerals to the natural processes of wetting and drying. Such soils typically have moderate to high clay contents and marked shrink–swell potential. Any tendency to crust and seal under the impact of rain is counteracted by shrinkage and cracking, thus producing a mulch effect as the soil dries out. Tillage when wet may appear to destroy the surface mulch but it will re-form upon drying. Black earths are commonly self-mulching. sepic A soil fabric visible in a petrographic microscope characterised by the presence of patches and/ or zones with a striated interference pattern. sepiolite Fibrous white mineral, approximately Mg8Si12O30 (OH)4.4H2O. Generally found in arid, evaporitic environments. See also PALYGORSKITE sesquioxide Mineral such as Al2O3 or Fe2O3 or Mn2O3. SHEET EROSION Erosion of material by sheet flow. It commonly removes fine material – leaving coarser material behind as a lag deposit. Also called hillwash, sheet-wash, surface wash, slope wash or rain wash. siderophile Elements with an affinity for Fe (Section 5.2). In the regolith this is reflected by incorporation into Fe oxides (Section 5.4.3; Appendix 2). silcrete Strongly silicified, indurated regolith, generally of low permeability and commonly having a conchoidal fracture with a vitreous lustre. Silcrete appears to represent the complete or near-complete silicification of precursor regolith by the infilling of available voids, including fractures. Most are dense and massive, but some may be cellular, with boxwork fabrics. The fabric, mineralogy and composition of silcretes may reflect those of the parent (regolith) material and hence, if residual, the underlying lithology. Thus, silcretes over granites and sandstones have a floating or terrazzo fabric
Appendix 1: Glossary of regolith terms
and tend to be enriched in Ti and Zr (see also 1!: cement); silcretes with lithic fabrics (e.g. on dunites) are silicified saprolites with initial constituents diluted or replaced by silica. silica SiO2. In the mineral processing context, two varieties are commonly recognised: (i) reactive silica, which is the silica present in the clay minerals – in particular kaolinite and halloysite – or in a very fine-grained state, which is soluble in hot sodium hydroxide solutions. (ii) free silica = quartz. SILICEOUS INDURATION Induration leading to either the absolute or relative accumulation of silica as a cementing agent. SILICIFICATION The introduction of, or replacement by, silica, generally resulting in the formation of fine-grained quartz, chalcedony, or opal, which may both fill pores and replace existing minerals. sillite An accumulation of silica as quartz. SKELETAL SOIL See lithosols. slaking Breakdown of soil aggregates when submerged – caused by the release of entrapped air and/or the swelling of expandable clay minerals. smectite A group name for layer silicates – formerly known as montmorillonite. Smectites readily take water between their structural layers in the form of molecules bound to exchangeable Ca, K or Mg (or, less commonly, Na). The ‘shrink–swell’ character of smectites gives soils much of their mobility and cracking character. Smectites typically form the finest particles in a soil. Their small size also gives them a high edge exchange capacity (Section 4.3.1; Table 5.3). A small amount of smectite in a soil therefore has a considerable effect on the soil’s properties. See also BEIDELLITE MONTMORILLONITE nontronite and saponite. smithsonite White mineral, ZnCO3 – weathering product of sphalerite (Table 4.2). SODIC SOIL A soil with so much adsorbed Na+ present that it affects the soil structure (Section 12.7.1). soil The unconsolidated mineral matter on or near the surface of the Earth that has been subjected to, and influenced by, genetic and environmental factors such as climate (including moisture and temperature effects), macro- and micro-organisms and topography – all acting to produce a product
(soil) that differs from the material from which it is derived in many physical, chemical, biological and morphological properties, and characteristics (Section 6.2; Figure 6.3). SOIL TEXTURE FIELD A measure of the behaviour of a small handful of soil when moistened and kneaded into a ball and then pressed out between thumb and forefinger. (Section 12.3.3: Table 12.2). See also TEXTURE TRIANGLE SOILS DIFFERENTIATED Soils showing development of pedogenic horizons, such as A and B horizons (Figure 6.3). SOLIFLUXION The slow flow of water-logged soil down-slope – associated with alternating freezing and thawing. solonised Said of a soil profile in which the pH varies from slightly acid in the upper horizon to slightly alkaline in the lower. Carbonate nodules may occur in the deeper parts of the profile. solum The upper part of the regolith profile; the A and B soil horizons (Figure 6.3). speleothem A secondary mineral deposit that is formed in a cave by the action of water. SPHEROIDAL WEATHERING Rounding of a pebble, boulder or tor by preferential weathering of protuberances and corners, and by exfoliation of concentric shells. SPONGY Said of a rock with numerous, more-or-less evenly distributed voids – interconnected or not – generally ranging in size from 2–16 mm. STAGNANT ALLUVIAL PLAIN Alluvial plain on which both erosion and aggradation are essentially inactive. STONE LINE Layer in the regolith composed of gravelsize angular to sub-rounded fragments of weathering-resistant rock – commonly quartz – and normally occurring at a depth between 0.3 and several metres below a gently sloping ground surface. Stone lines have been variously interpreted as marking an unconformity – as the base of bioturbation (Figure 8.17) – or as the boundary between in situ saprolite below and a mobile layer above. STRIPPED SLOPES Erosional slopes flanking hills. Stripped slopes can have much rock exposed and/ or a shallow mantle of stony colluvium. STRUCTURED SAPROLITE See saprock.
429
430
Regolith Science
subrosion Removal of regolith material in solution or as solids by flow of groundwater along sub-surface channels and pipes. Such removal may cause collapse and settling of overlying regolith. SUB SURFACE SOLUTIONPIPING The movement of materials by sub-surface water flow – both in solution and in suspension – to result in changes to the shape of the land surface. Such removal can lead to both circular and linear depressions; the latter are percolines. Sub-surface removal of weathered materials can lead to the development of tunnels or pipes, which may collapse to form shafts. Such processes can occur in most regolith materials and rocks, and the phenomenon reaches its best development in limestone karst areas. SULFIDIC (soils): Soil containing sulfides (generally pyrite) which are susceptible to oxidation and the formation of acid sulfate soils (Section 12.8.2). SULFURIC (soils): Soil with a pH<4. An acid sulfate soil (Section 12.8.2). supergene Said of a mineral deposit or enrichment formed near the surface – commonly by descending solutions; also, said of those solutions and of that environment. SUPERIMPOSED DRAINAGE Drainage is interpreted to be superimposed when a stream’s course is unrelated to the geologic structures and underlying rocks over which it flows. Such an interpretation presumes that the drainage developed on a previous land surface at a higher level, and that differential erosion has subsequently emphasised the geologic structures. SURFACE WASH Sheet erosion. swale Linear, level-floored open depression excavated by wind, or left relict between ridges built up by wind or waves, or built up to a lesser height than them. 37)2 Short wavelength infra-red. SYNERESIS The spontaneous separation or throwing off of a liquid from or by a gel or flocculated colloidal suspension during aging, resulting in shrinkage and in the formation of cracks, pits, mounds, cones or craters. T TAFONE TAFONI A form of cavernous weathering chiefly found in igneous rocks – commonly in coastal or arid environments. The holes or recesses may reach
a diameter of a metre or more. See also HONEYCOMB weathering and CAVERNOUS WEATHERING. talc White mineral, Mg3Si4O10 (OH)2 . A common secondary mineral derived by hydration of nonaluminous magnesium silicates (olivine, enstatite, tremolite) in mafic and ultramafic igneous rocks. talus Fragments of rock and soil material accumulated by gravity at the foot of cliffs or steep slopes. Scree. TAXONOMY The classification of living organisms. termitaria The nests of a termite colony. TERRA ROSSA SOILS Deep-red, residual soil derived from impurities when carbonates weather (Section 6.3.5). terracette Small parallel terraces on a steep hillside – usually caused by grazing animals such as sheep. texture (regolith) The physical nature of a regolith unit or component according to the proportions of different size fractions (e.g. sand, silt or clay). For physical characteristics related to the spatial arrangement of the constituents at the microscopic and mesoscopic scale, the term ‘fabric’ is preferred. TEXTURE TRIANGLE A classification of soil according to the relative proportions of clay, silt and sand (Section 12.3.3; Figure 12.3). See SOIL TEXTURE FIELD thalweg The line of continuous maximum descent from any point on a land surface; e.g. the line of greatest slope along a valley floor, or the line crossing all contour lines at right angles, or the line connecting the lowest points along the bed of a stream. tholin Dark-red organic polymers formed by solar irradiation of simple organic compounds under reducing conditions (Section 14.5.2). Not present on Earth due to its oxidising atmosphere. tigree See GROVEnINTERGROVE PATTERN titania TiO2. toeslope The lower, gentle slope of a hillside – lying at the foot of an escarpment or steep rock face and generally covered by an accumulation of talus; it is less steep than the slope element above and commonly consists of alluvial fans or pediments. toposequence A sequence of soils whose variation is related to their topographic position. TRANSFORMATION FRONT The boundary of a change (or transformation) in the composition or other property of the regolith.
Appendix 1: Glossary of regolith terms
TRANSPORTED OVERBURDEN A term referring to material of exotic or redistributed origin such as alluvium, colluvium and aeolian material that blankets fresh or weathered bedrock. It may be friable or partially or wholly consolidated, or cemented. TREE THROW Disturbance of the regolith caused by the uprooting of a tree. TRELLIS DRAINAGE A drainage pattern where secondary channels flow at right angles to the main channel. The secondary channels are, in turn, joined at right angles by small tributaries flowing parallel to the main channel. This pattern is common in well-bedded rocks – commonly with scarp and dip slopes. Small tributaries on the scarp slopes are short and steep, whereas those on the dip slopes are longer and more gently sloping. tubule Tube with diameter <2 mm. U unconsolidated Primary property of looseness of regolith constituents, that allow pieces of regolith to be crumbled or deformed with the fingers. 54- Universal Transverse Mercator grid; present on most topographic maps and used for quantitative description of locations. In this system the world is divided into a series of zones and points are located according to grid co-ordinates (eastings and northings) in each zone. V vadose Of sub-surface water, that which is above the water table. Water below the water table is formally defined as groundwater. VENTIFACT Stone faceted or polished by wind action. vermicular See VERMIFORM vermiculite Mineral structurally mid-way between biotite and chlorite. It has a trioctahedral 2:1 layer, and a hydrous interlayer bearing Mg or Al. Vermiculate has a high layer charge and a high cation exchange capacity (Chapter 4.3.3). It is the first weathering product of biotite in granite (Eqn 4.2), and also forms from pyroxene or amphibole during weathering of mafic rocks. VERMIFORM Having the form of a worm. In the regolith, a fabric consisting of tubes, pipes or wormshaped voids which may be filled or partly filled with, for example, clays, sandy sediments or Fe oxides. Synonym ‘vermicular’.
VNIR Visible near infra-red. W wad Mineral aggregate; hand specimen term for soft, black, earthy Mn oxides, typically including pyrolusite, romanchèite, manganite, nsutite, todorokite. Largely equivalent to psilomelane. Occurs with other Mn oxides in the weathering profiles of ultramafic rocks, or Mn-bearing rocks and base metal deposits; e.g. Broken Hill, NSW. See also MANGANESE OXIDES. WATER TABLE The surface to which water rises in an open well or piezometer. weathered Having experienced weathering. In the strict sense, the word may properly only be applied to a rock or mineral where there remain recognisable aspects of the parent material, such as fabric or mineral pseudomorphs. Applied to an outcrop or profile, the term may refer to the extent to which it has undergone chemical and mineralogical change by weathering. weathering Weathering refers to any process that – through the influence of gravity, the atmosphere, hydrosphere and/or biosphere at ambient temperature and atmospheric pressure – modifies rocks, either physically, or chemically (Chapter 1). WEATHERING DEEP Term properly applied to the depth of weathering (as measured in metres), not to the degree of weathering. Deep weathering is a relative term, which is applicable to a profile weathered to a greater depth than the average for the region. WEATHERING FRONT The boundary between unweathered and weathered rock (Figure 6.3). willemseite Light green, Ni analogue of talc, (Ni,Mg)3Si4O10 (OH)2. WUSTENQUARTZ Red (Fe-stained) quartz. Z ZONAL SOILS Soils whose characteristics are allegedly dominated by climate. See also INTRAZONAL SOILS ZONE In the regolith context, a part of the regolith having a distinctive character, differing from parts adjacent, as in ‘mottled zone’ (Figure 6.1).
REFERENCES Anand RR, Churchward HM, Smith RE, Smith K, Gozzard JR, Craig MA and Munday TJ (1993).
431
432
Regolith Science
‘Classification and atlas of regolith-landform mapping units: Exploration perspectives for the Yilgarn Craton’. CSIRO Exploration Geoscience /AMIRA Report 440R. (Reissued as Open File Report 2 CRC LEME, Perth, 1998.) Brantley SL, Goldhaber MB and Ragnarsdottir KV (2007). Crossing disciplines and scales to understand the Critical Zone. Elements 3, 307–314.
Eggleton RA (Ed.) (2001). The Regolith Glossary: Surficial Geology, Soils and Landscape. CRC LEME, Canberra and Perth. Jackson JA (1997). Glossary of Geology. 4th edn. American Geological Institute, Alexandria, Virginia. Nickel EH and Nichols MC (1991. Mineral Reference Manual. Van Nostrand Reinhold, New York.
Appendix 2: Regolith geochemistry of elements Keith M Scott
The basic features of elements, with relevance to their likely host minerals in fresh rocks and in regolith samples, are listed below on an elemental group basis (see Figure 5.1 for the complete Periodic Table). Much of the data on abundances of elements in primary minerals is compiled from Wedepohl (1978) – maximum values are quoted. Where possible, ionic radii data are for the commonly occurring ions in octahedral coordination from Lide (1997), otherwise older data from Weast (1989) are quoted in parentheses. Octahedral coordination is used because that is the most common mode of occurrence in regolith materials. (Ions in tetrahedral coordination have smaller radii – the exception being Fe2+ with radii
octahedral/tetrahedral of 0.61 versus 0.63 Å. Larger ions such as K+, Pb2+ and REE may also be present in 12-fold coordination in regolith minerals.) Elemental abundances in regolith materials are based largely on the author’s experience, with a bias toward mineral exploration environments (extensive data for environmental studies is provided by Reimann and de Caritat 1998). Analytical methods quoted are those which give the low detection limits and which are commonly available – usually based on an HClHClO4-HNO3-HF digestion and analysis by inductively coupled plasma mass spectrometry (ICPMS) or inductively coupled plasma emission spectrometry (ICPES) (e.g. ALS 2008; Genalysis 2008).
434
Regolith Science
GROUP 1 ELEMENTS (ALKALIS) Analytical method (detection limit)
Element
Atomic number
Charge
Ionic radius (Å )
Common host primary minerals
Common occurrence in regolith
Li
3
+1
0.76
Pegmatite minerals (Li- micas, tourmaline)
Water, adsorbed onto clay
ICPMS (0.1 ppm)
Na
11
+1
1.02
Feldspars, amphiboles, micas
Water, alunite– jarosite
ICP (20 ppm)
K
19
+1
1.38
Feldspars, amphiboles, micas, alunite
Water, alunite– jarosite
ICP (20 ppm)
Rb
37
+1
1.52
Pegmatitic minerals, K-feldspar (900 ppm) and micas (2500 ppm)
Water
ICPMS (0.05 ppm)
Cs
55
+1
1.67
Pegmatitic minerals, K-feldspar (100 ppm) and micas (3400 ppm)
Water, adsorbed onto clay
ICPMS (0.05 ppm)
Fr
87
+1
1.80
Only occurs as the short-lived radioactive isotope
During weathering, the alkalis are released from their primary minerals to form monovalent ions and dispersed by groundwaters or incorporated into/onto neoformed clay minerals. The larger alkali ions have greater exchange capacity than the smaller ions (Figure 5.12) 3ODIUM tends to be more mobile than K during weathering (Figure 5.10), so that the ratio Na/K generally decreases as weathering proceeds to form soils. The K incorporated into muscovite is generally retained better than that in biotite or feldspar (Figure 5.3). Na and K may be stabilised in weathering profiles by their incorporation into alunite–jarosite min-
erals, which may also stabilise usually mobile pathfinder elements. Potassium is essential for all organisms. Rubidium is preferentially concentrated into residual liquids relative to K during crystallisation processes (K/Rb <150 may imply a pegmatite or an highly differentiated granite). Thus Rb tends to be concentrated in biotite and muscovite rather than feldspar. The greater stability of muscovite during weathering (Figure 5.3) can mean that K/Rb ratios, preserved in muscovite in regolith material, may reflect the original mica paragenesis (e.g. pegmatitic or from an highly differentiated granite).
Appendix 2: Regolith geochemistry of elements
GROUP 2 ELEMENTS (ALKALINE EARTHS) Common occurrence in regolith
Analytical method (detection limit)
Pegmatite minerals (e.g. beryl)
Adsorbed onto clay
ICPMS (0.05 ppm)
0.72
Olivine, amphiboles, micas, chlorite, carbonates
Carbonates, smectites, vermiculite, water
ICP (20 ppm)
+2
1.00
Feldspars, amphiboles, micas, apatite, sphene, carbonates, gypsum, anhydrite
Water, carbonates, sulfates, smectites
ICP (10 ppm)
38
+2
1.18
Feldspars (5000 ppm), micas (400 ppm), apatite (1%), carbonates (6000 ppm)
Water, carbonates, sulfates
ICPMS (0.05 ppm)
Ba
56
+2
1.35
K-feldspar (5%) and micas(1%), barite
Water, barite, hollandite
ICPMS(0.1 ppm)* XRF(10 ppm)
Ra
88
+2
(1.43)
Occurs as a short-lived radioactive isotope, associated with Th- and U-bearing minerals
Water, barite, Fe-Mn oxides, organics
Charge
Ionic radius (Å )
Common host primary minerals
4
+2
0.45
Mg
12
+2
Ca
20
Sr
Element
Atomic number
Be
* Ba determined by ICPMS may be low due to incomplete dissolution of barite and/or precipitation of neoformed barite in the presence of sulfate in solution.
Alkaline earth ions are absorbed into/onto clay minerals better than the alkali ions (Figure 5.12). Mg and Ca are strongly depleted during weathering, but they may be introduced/retained in secondary carbonate minerals (‘calcrete’: Sections 4.5.2, 5.4.3 and 13.7). Because of the substantial size difference between Mg and Ca (0.72 versus 1.00 Å), Mg is not readily incorporated into calcite and increasing Mg/ Ca with depth in a profile could reflect calcite and dolomite rather than increased Mg in the calcite (Khider and McQueen 2006). Thus, the mineralogy of the calcrete should be checked; e.g. by X-ray diffraction. 3TRONTIUM is intermediate in size between Ca and K (1.18 versus 1.00 and 1.38 Å) and substitutes into both Ca and K minerals, but does not strongly enter micas (Taylor 1966). As seen above, Rb follows K strongly into micas, so Rb/Sr ratios might be expected to vary considerably within a regolith profile and probably cannot be used in a general predictive manner unless the mineralogy is also known. Significant Sr may be present in secondary barite, carbonates and alunite–jarosites, as well as within residual
barite. Up to 7.6% Sr occurs in the secondary barite immediately above the supergene sulfide zone at Elura, NSW (Scott and Taylor 1987b) and up to 3200 ppm has been found in cryptomelane from Mt Magnet, WA (Scott 1990a). Lintern et al. (2006) report up to 1% Sr in calcrete, but probably as celestite (SrSO4) rather than in the calcite structure. Up to 15% Sr has also been reported in calcrete samples (H Waldron pers. comm. 2008). Radiogenic 87Sr is formed by the decay of 87Rb. Thus Sr isotopic ratios (87Sr/86Sr) are widely used in dating and petrogenetic studies in fresh rocks. However, Sr isotopic ratios can also be useful in regolith studies; e.g. in identifying sources of Ca present in calcrete because of the differences between oceanic and bedrock Sr isotopic ratios (Dart et al. 2005, 2007; Lintern et al. 2006). Barium may be present in percentage amounts in members of the coronadite-cryptomelane-hollandite series, but only minor amounts in lithiophorite from the same sample (e.g. ~7.8% Ba in hollandite, but only 0.2% Ba in coexisting lithiophorite at New Cobar, NSW: Scott and McQueen 2001).
435
436
Regolith Science
Secondary barite may form, and be concentrated in, the supergene oxidate zones of weathering sulfide deposits even when the Ba content of the primary ore zone is low (e.g. Scott et al. 2001). It may also be formed in regional calcrete zones (e.g. McQueen 2006). Ba/Sr and Ba/Rb ratios in feldspars reflect fractionation but the variable weathering of their host minerals probably precludes their use in regolith samples.
Radium is separated from other radioactive isotopes during weathering due to its alkaline earth features (greater solubility). It is particularly mobile in highly saline groundwaters (Dickson 1990) and Ra may be concentrated on the margins of salt lakes and near sandstone escarpments in swamps with only low accompanying U content (Dickson et al. 1987).
GROUP 3 ELEMENTS Common occurrence in regolith
Analytical method (detection limit)
Charge
Ionic radius (Å )
5
+1 +3
(0.35) (0.23)
Tourmaline, micas (2000 ppm)
Tourmaline. Soluble borates
ICP (10 ppm) (special digest)
Al
13
+3
0.54
Feldspars, amphiboles, garnets, micas, chlorites, Al2SiO5 polymorphs
Muscovite, kaolinite, Fe/Mn oxides, alunite– jarosites. Bauxite. Soluble Al3+
ICP (20 ppm)
Ga
31
+3
0.62
Micas, tourmaline (300 ppm)1, magnetite, sphalerite (1000 ppm)
Tourmaline, bauxite, alunite– jarosite, Fe oxides
ICPMS (0.05 ppm)
In
49
+3
0.80
Tourmaline (1%), micas (1800 ppm), sphalerite (4700 ppm), chalcopyrite/ stannite (1500 ppm)
Soluble. Precipitated with Fe/Mn oxides
ICPMS (0.005 ppm)
Tl
81
+1 +3
1.50 0.89
Micas (400 ppm) and K-feldspar (600 ppm), especially in pegmatites. Pyrite (210 ppm), marcasite (9000 ppm), sphalerite (1000 ppm), galena (5000 ppm)
Fe/Mn oxides
ICPMS (0.02 ppm)
Element
Atomic number
B
1
Common host primary minerals
Data from Telfer, WA (Scott et al. 1993)
Boron in tourmaline is retained during weathering, but B in other primary minerals may be mobilised and fixed by adsorption onto neoformed Fe/Al oxides. Authigenic clay minerals incorporate B, with more being present in marine than freshwater shales (Wedepohl 1978). It is toxic to plants at high levels. Aluminium (Al3+) is soluble at pH<4 and pH>9, whereas Fe3+ is only soluble under acid conditions (Figure 5.7); thus, when an acid weathering solution is
neutralised, Fe(OH)3 is precipitated and Al3+ remains in solution – i.e. two major components of regolith materials may be separated and lead to Al concentration in bauxite (Section 6.3.1). The Al distribution in soils is strongly affected by plant roots even though it is a non-nutrient (Section 8.3.1). 'ALLIUM (Ga3+) substitutes for Al3+ and Fe3+ in the regolith because of their similar sizes (0.62 versus 0.51 and 0.64 Å). Thus some of the high Ga content in
Appendix 2: Regolith geochemistry of elements
sphalerite may be retained in Fe and Al oxides (or alunite–jarosite minerals) when sphalerite weathers. Indium (In3+) from sphalerite and chalcopyrite/ stannite is generally dispersed but, with the low limits of detection now possible with ICPMS, small amounts precipitated with Fe-Mn oxides may be detected (N Radford pers. comm. 2000).
4HALLIUM in white micas may be retained during weathering but Tl, freed when sulfide hosts weather, is generally lost except for that retained in Fe-Mn oxides. Crytomelane in the deeply weathered Conta História Mn deposit, Minas Gerais, Brazil contains 19 ppm Tl (Cabral et al. 2002)
GROUP 4 ELEMENTS Common occurrence in regolith
Analytical method (detection limit)
Carbonates, graphite, coal/lignite
Carbonates, graphite, coal/ lignite, water, atmosphere
Leco furnace (100 ppm)
0.40
Quartz, feldspars, amphiboles, pyroxenes, garnets, phyllosilicates, Al2SiO5 polymorphs
Quartz, muscovite, kaolinite, Soluble Si4+, Secondary silica
Fusion ICP (100 ppm)
+2 +4
0.73 0.53
Topaz (700 ppm), garnet (100 ppm), sphalerite (2000 ppm)
Soluble during weathering. Fixed by clays
ICPMS (0.1 ppm)
50
+4
0.69
Cassiterite, stannite, sphene (9000 ppm), micas (3500 ppm)
cassiterite
ICPMS (0.1 ppm)*
82
+2 +4
1.19 0.78
Galena, K-feldspar (1%), micas (100 ppm)
Fe-Mn oxides, alunite–jarosites
ICPMS (0.5 ppm)
Charge
Ionic radius (Å )
Common host primary minerals
6
+4
0.16
Si
14
+4
Ge
32
Sn Pb
Element
Atomic number
C
* Sn as cassiterite may not be completely digested.
The carbon contents of shales associated with mineralisation may be high (e.g. nearly 5% C in black shale at McArthur River, NT; Corbett et al. 1975), but usually black shales contain <1% C. During weathering the C is oxidised to leave bleached white saprolite (Section 6.3.4). Lignite is commonly found in paleochannels and the reducing environment may facilitate retention of base metal – commonly as secondary sulfides and U minerals; e.g. at Mulga Rock; Douglas et al. 2005). Significant H, N and O may also be present in low-rank organic matter (e.g. Saxby 1976). 14C isotopes are commonly used to date groundwaters (Section 10.9).
The silicon distribution in soils is strongly affected by plant roots (Section 8.3.1). Germanium from sphalerite is lost during weathering, although Ge 4+ can readily enter Fe3+ sites (ionic radius 0.53 versus 0.55 Å) in Fe oxides and clays. Ge may be present in hot springs but >10 ppm Ge is toxic to plants. 4IN is resist to weathering when present as cassiterite (this Sn may not be fully dissolved during dissolution for analysis), but Sn in less resistant minerals may be dissolved although it is generally re-precipitated nearby. Lead (Pb2+) is similar in size to K, Sr, Ba and Ca (1.19 versus 1.38, 1.18, 1.35, 1.00 Å) and so it may go
437
438
Regolith Science
into K and Ca-rich minerals; conversely these elements may be associated with Pb in secondary minerals (e.g. Ca in pyromorphite at Pegmont, Qld; Scott and Taylor 1987a). The Pb present in weathered rocks is generally present in Mn oxides (~29% Pb in coronadite but <0.1% in lithiophorite; Scott 1986), Fe oxides (up to 4.2% and 6.4% Pb in goethite and hematite, respec-
tively: Scott 1986), alunite–jarosite minerals (up to 36% in plumbogummite). Up to 6% Pb is also present in secondary barite from just above the supergene sulfide zone at Elura, NSW (Scott and Taylor 1987b). Pb isotopes have been used to determine whether anomalous Pb in weathered samples is likely to be ore-related (e.g. Gulson 1984; Vaasjoki and Gulson 1985).
GROUP 5 ELEMENTS
Charge
Ionic radius (Å )
7
+3 +5
P
15
As
Analytical method (detection limit)
Common host primary minerals
Common occurrence in regolith
0.16 0.13
Ammonium micas and feldspars
Nitrates, organic matter
+5
0.38
Apatite, REE phosphates (e.g. monazite, xenotime)
Soluble during acidic weathering, precipitated under alkaline conditions. Secondary phosphates, Fe oxides, alunite– jarosites
ICP (10 ppm)
33
+3 +5
0.58 0.46
Arsenopyrite, tetrahedritetennantite, sulfosalts, pyrite (8000 ppm)1
Soluble in water. Fe oxides, alunite– jarosites
ICPMS (0.2 ppm)
Sb
51
+3 +5
0.76 0.60
Tetrahedrite–tennantite, Galena (1%), sphalerite (5000 ppm), pyrite (6000 ppm)1, arsenopyrite (3%)1, rutile (6.6%)2
Soluble in water. Fe oxides, alunite– jarosites, tripuhyite
ICPMS (0.05 ppm)
Bi
83
+3 +5
1.03 0.76
Bismuthinite, galena (4%) sphalerite (4000 ppm), chalcopyrite (2000 ppm), pyrite (1000 ppm)
Fe oxides
ICPMS (0.01 ppm)
Element
Atomic number
N
1 Data 2 Data
from Elura, NSW (Scott 1991) from Hemlo, Ontario (Urban et al. 1992)
Nitrogen in primary minerals is found mainly as NH4+ (ionic radius =1.43 Å). Thus it fits readily into K+ sites (radius =1.38 Å) and might be expected to be retained in the more stable white micas, but not feldspars, during weathering (cf. Figure 5.3). N in surficial soils is mainly present within organic matter. The amount of N in a rock may be directly related to its shale content. Nitrates are generally only common where biological activity is absent and there is a lack of leaching by
groundwaters allowing nitrates to accumulate (e.g. Attacama Desert). Phosphorus in regolith materials is mainly in the Fe oxides (up to several % P2O5), with goethite generally a better host than hematite (Scott 1986). P is essential for growth and reproduction of plants and animals (Section 8.3.1). Arsenic (As3+) readily substitutes for Fe3+ (ionic radii: 0.58 versus 0.55 Å). In the +5 OS, it occurs as arsenate and behaves similarly to phosphate and may
Appendix 2: Regolith geochemistry of elements
be present in alunite–jarosite minerals as well as adsorbed onto/incorporated into Fe oxides, with up to 5.7% and 3.3% As, respectively, reported in goethite and hematite at Elura, NSW (Scott and Taylor 1989). Because the As association with Fe is very strong in regolith materials, care must be exercised to distinguish between normal and anomalous As accumulation in Fe oxides (e.g. almost 200 ppm As can be present in unmineralised Fe oxides: McQueen et al. 2004). !NTIMONY behaves similarly to As and Bi. Sb2O5 is more soluble than Sb2O3 (8.77 mg/100 ml versus 1.31 mg/100 ml). Sb mainly occurs in the +5 OS in weath-
ering profiles within Fe oxides (up to 1.0 and 2.1% in goethite and hematite at Elura, NSW: Scott and Taylor 1987b; 1989) and alunite–jarosite minerals (up to 1.3%; Scott and Taylor 1987b). It is also retained in residual rutile (e.g. Big Bell; Scott and Radford 2007) Bismuth is geochemically associated with As, Sb and Pb. It may show coupled substitution in galena Ag+ + Bi3+ = 2Pb2+ (e.g. Scott 1991). Bi is freed when galena weathers and it may be incorporated into alunite–jarosites, hematite and goethite (up to 9700 ppm in hematite at Mt Leyshon, Qld: Scott 1987b; 700 ppm is found in coexisting goethite and hematite from New Cobar, NSW; Scott and McQueen 2000).
GROUP 6 ELEMENTS Analytical method (detection limit)
Element
Atomic number
Charge
Ionic radius (Å )
O
8
–2
1.40
Silicates, oxide minerals
Quartz, clay minerals, Fe/Mn/Al oxides
S
16
–2 +4 +6
1.84 0.37 0.29
Native S. Sulfates (e.g. gypsum, anhydrite, alunite, barite). Sulfides
Soluble during acidic weathering (except in presence of alkaline earths). Fe oxides, alunite–jarosites
Leco furnace (50 ppm) ICP (10 ppm)
Se
34
+4 +6
0.50 0.42
Sulfides, galena (20%), pyrite (3%) molybdenum and chalcopyrite (1000 ppm) Native S (5%)
Rapidly oxidised. Incorporated into Fe oxides, alunite– jarosites
ICPMS (1 ppm)
Te
52
+4 +6
0.97 0.56
Au/Ag tellurides
Au/Ag tellurides stable during oxidation. Tellurites
ICPMS (0.05 ppm)
Po
84
+4
0.97
Decay product of U and Th present in U minerals
U minerals
Common host primary minerals
Common occurrence in regolith
/XYGEN is the most abundant element in the Earth’s crust. Depletion of gaseous O2 in soil gases above weathering buried massive sulfide deposits has been used as an exploration method (McCarthy et al. 1986: see also Section 5.3.6). Oxygen isotopes can be used to determine the latitude at the time of formation: dO18 becomes more negative at higher latitudes. dO18 may also decrease toward mineralisation (e.g. Klein and Criss 1988). It is elevated in sinters and relative to higher temperature
quartz in epithermal Au districts (Ewers 1991). This variation can potentially be seen in more residual minerals. 3ULFUR isotopic ratios of sulfides are retained when the sulfide oxidises. Thus the source of the S can sometimes be reflected by the oxidation product (see Taylor et al. 1984). 3ELENIUM commonly substitutes for S in volcanogenic deposits but Se is more strongly fixed onto Fe oxides than S. Thus, oxidising waters are generally
439
440
Regolith Science
low in Se, and Se might be expected to be present in ferruginous samples derived by the weathering of volcanogenic sulfide mineralisation (e.g. Scott et al. 2001). Se may be concentrated in goethite relative to hematite (1200 versus 600 ppm, respectively, in a gossan profile at New Cobar, NSW; Scott and McQueen 2000). In gossanous material from Lewis Ponds, NSW, up to 2100 ppm Se is present in goethite (Scott, unpublished data, 1999).
Se is essential for many organisms but at too high levels it becomes toxic for living matter. 4ELLURIUM present as precious metal tellurides, is stable during weathering and these minerals can be found in placer deposits. Te would be expected to be very low in soils. Polonium is radioactive, with Po210 being the most stable isotope: having an half life of 140 days.
GROUP 7 ELEMENTS (HALIDES)
Charge
Ionic radius (Å )
1
–1 OH –
F
9
Cl
Analytical method (detection limit)
Common host primary minerals
Common occurrence in regolith
(1.54) 1.37
Hydrated silicates, water
Clay minerals, Fe/Al oxides, organic matter
–1
1.33
Topaz, fluorite, apatite (3%), amphibole (3%), mica (8%), tourmaline (1%)
Soluble during weathering. Topaz, white mica, F-apatite, tourmaline
Fusion/ Specific Ion Electrode (20 ppm)
17
–1
1.81
Halite, micas (1%), scapolite (3%), apatite (3%), amphiboles (1%)
Highly soluble. Halite
Acid Digestion/ titration (10 ppm)
Br
35
–1
1.96
Minor component in evaporite minerals. Less than 10 ppm in rock-forming minerals
Soluble
Neutron Activation (1 ppm)
I
53
–1
2.20
Less than 1 ppm in rockforming minerals
Soluble. Iodides (e.g. AgI)
Neutron Activation (5 ppm)
At
85
+7
(0.62)
Short-lived radio isotope
Element
Atomic number
H
Fluoride readily substitutes for OH (radii = 1.33 versus 1.37 Å). Fluoroapatite is the least soluble of the apatite minerals. Shales can contain up to 1000 ppm F. The F present in white mica and topaz, which are associated with some tin deposits, may be retained during weathering, so that F anomalous zones can be defined using partially weathered material (e.g. Scott and Rampe 1984). Chlorine doesn’t generally substitute for F because of its size but it can fit into OH sites slightly more easily (radii Cl: F: OH= 1.81: 1.33: 1.37 Å).
Bromine may be more concentrated relative to Cl in volcanic rocks. Br in soils correlates with organic content. Br in Australian vegetation tends to be higher than that from other countries (H Waldron pers. comm. 2008). Iodine in soils correlates with organic content. Iodine is commonly more concentrated in soils relative to parent rocks – possibly due to plant or atmospheric contributions.
Appendix 2: Regolith geochemistry of elements
GROUP 8 ELEMENTS Ionic radius (Å )
Common host primary minerals
Common occurrence in regolith
Ferromagnesian minerals, beryl
Atmosphere
Open spaces in minerals
Atmosphere
Element
Atomic number
He
2
Ne
10
+1
(1.12)
Ar
18
+1
(1.54)
Kr
36
Mineral springs, atmosphere
Xe
54
Mineral springs, atmosphere
Rn
86
Charge
Analytical method (detection limit)
Mineral springs, atmosphere
Radioactive (half life = 4 days)
Helium in soil and overburden gas has been used as a pathfinder for U/Th mineralisation but such surveys produce only weak anomalies at best. However, reproducible Ra anomalies can be detected in gas at depths of 1 m (Butt and Gole 1985). He isotopes can be used to date regolith materials, including groundwaters (Figure 10.21).
Argon isotopes have been used to date regolith materials, especially alunite–jarosites, K-bearing Mn oxides and groundwaters (Vasconcelos 1999; Section 2.3.2; Figure 10.21). Radon concentrates in petroleum from surrounding rocks. Ra in overburden gas may be a pathfinder for U mineralisation. (See He above.)
441
442
Regolith Science
FIRST TRANSITION SERIES
Element
Atomic number
Charge
Ionic radius (Å )
Sc
21
+3
0.75
Ti
22
V
23
+2 +3 +4 +2 +3 +4 +5
0.86 0.67 0.61 0.79 0.64 0.58 0.54
Cr
24
+2 +3 +4 +6
0.73 0.62 0.55 0.44
Mn
25
+2 +3 +4
0.83 0.58 0.53
Fe
26
+2 +3
0.61 0.55
Co
27
+2 +3
0.65 0.55
Ni
28
+2 +3
0.69 0.56
Cu
29
+1 +2
0.77 0.73
Zn
30
+2
0.74
Common host primary minerals Beryl (1%), xenotime (1%), columbite (4%) Ti oxides, spinels, sphene, ferromagnesian minerals e.g. phlogopite (5%) Roscoelite, magnetite (1.4%),1 hematite (1600 ppm), pyroxene (300 ppm), amphibole (300 ppm), biotite (400 ppm), tourmaline (1%),2 rutile (3.5%)3 Fuchsite, spinels, garnet (5%), pyroxene (1.5%), amphibole (8000 ppm), phlogopite (5.6%), tourmaline (8.9%),2 Al2SiO5 polymorphs e.g. kyanite (8.8%), rutile (4.3%),2 ruby (6000 ppm), emerald (1000 ppm) Ferromagnesian minerals e.g. hornblende (1.4%) and garnet 6.7%),4 epidote, Ti-mte (3.8%), ilmenite (23%), carbonates Ferromagnesian minerals, magnetite/ hematite, ilmenite, carbonates, sulfides (esp pyrite, pyrrhotite) Ferromagnesian minerals e.g. liebenbergite (1.4%),5 spinels (8500 ppm),6 sulfides/sulfosalts Ferromagnesian minerals, especially liebenbergite, spinels (1000 ppm), sulfides (esp pentlandite, violarite, pyrrhotite) Ferromagnesian minerals, e.g. biotite (1.2%), chlorite (2.3%),7 muscovite (1.1%),8 sulfides (esp chalcopyrite) Ferromagnesian minerals, e.g. garnet (5300 ppm) and biotite (5000 ppm), gahnite, magnetite (4000 ppm), chromite (5000 ppm), staurolite (7%) 6, sulfides (sphalerite), siderite (11%) 9
*Ti and Cr determined by ICP may be low due to incomplete dissolution of Ti oxides and spinels. 1 Data from Windimurra, WA (Bolton and Alexander 2005) 2 Data from Telfer, WA (Scott et al. 1993) 3 Data from Hemlo, Ontario (Urban et al. 1992) 4 Data from Big Bell, WA (K.M. Scott, unpublished data, 1994) 5 Data from Barberton, Transvaal (De Waal and Calk 1973)
Common occurrence in regolith Fe/Al oxides Ti oxides, spinels
Analytical method (detection limit) ICPMS (0.1 ppm) ICP (5 ppm)* XRF (10 ppm)
Clay minerals, Fe/Mn oxides. Vanadate minerals (e.g. with U)
ICP (1 ppm)
Spinels (e.g. chromite). Soluble Cr incorporated into Fe/Mn oxides
ICP (1 ppm)* XRF (5 ppm)
Mn oxides. Incorporated into secondary carbonates
ICP (1 ppm)
Soluble in water. Forms hematite/goethite
ICP (0.01%)
Soluble. Incorporated into Fe/Mn oxides
ICPMS (0.1 ppm)
Soluble, solubility controlled by phosphate, carbonate and hydroxide. Incorporated into Fe/Mn oxides Soluble. Incorporated into Fe/Mn oxides and alunite–jarosite minerals
ICP (1 ppm)
Soluble. Incorporated into Fe/Mn oxides, alunite–jarosite minerals, smectites, carbonates
ICP (1 ppm)
ICP (1 ppm)
6 Data from Bleikvassli, Norway and Mineral Hill, Maryland (Spry and Scott 1986). 7 Data from Mount Dore, Qld (Scott 1988) 8 Data from Sar Cheshmeh, Iran (Henley 1970) 9 Data from Elura, NSW (Scott, unpublished)
Appendix 2: Regolith geochemistry of elements
3CANDIUM substitutes for Fe3+ and Al3+ in regolith materials (radii = 0.75 versus 0.55 and 0.54 Å). 4ITANIUM (Ti 4+) substitutes for Fe3+ and Al3+ in regolith materials (radii = 0.61 versus 0.55 and 0.54 Å). Ti oxides (rutile, brookite and anatase) are stable during weathering and such minerals commonly contain substantial impurities such as Sc, V, Nb, Ta, Cr and W. See second transition series (below) for Ti/Zr ratios. Vanadium (V3+) substitutes for Fe3+ and Mn4+ (radii = 0.64 versus 0.55 and 0.53 Å, respectively). Up to 6000 ppm V has been found in Mn oxide from the Scuddles Cu-Zn deposit, WA (M le Gleuher pers. comm. 2007). Vanadium is associated with U in red bed U deposits of the Colorado Plateau and in carnotite (K 2 (UO2) V2O8.3H2O) in groundwater calcrete precipitated at or below the water table in sediments within broad valleys (Butt et al. 1977, 2005). Chromium (Cr3+) follows Al3+ (radii = 0.62 versus 0.54 Å). Cr imparts the colour to corundum to turn it into ruby and beryl to turn it into emerald. During weathering, the Cr firmly bound in spinel structures is stable, but that in silicates may be freed. If it remains in the +3 OS, it is generally insoluble and up to 6500 and 3800 ppm Cr may be incorporated into goethite and hematite in weathered mafic/ultramafic rocks at Mt Magnet WA (Scott 1990b) and 0.9% Cr into smectite at the Perseverance Ni deposit, WA (Nickel et al. 1977). Kaolinite from the former location only contains up to 600 ppm Cr (Scott 1990 b).When oxidised to the +6 OS, Cr may be more mobile and be present in water to levels above potable standards (Gray 2003). However, as noted in Section 10.6.2, Cr6+ is readily adsorbed by organic matter. -ANGANESE generally occurs as Mn2+ in primary minerals (the exception being Mn3+ in epidote (piemontite)). It forms Mn oxide minerals (with a +4 oxidation state) during weathering and these may incorporate percentage amounts of elements such as Co, Ni, Zn, Pb, Cu, Zn, Ba, K and Li (Appendix 1). Iron occurs in the +3 OS in the regolith. Formation of hematite and goethite allows incorporation of elements, which may reflect the presence of former sulfide mineralisation. Cobalt AND .ICKEL follow Fe and Mg (radii of divalent ions = 0.65 and 0.69 versus 0.61 and 0.72 Å, respectively). Non-sulfide Ni may be particularly high in ultramafic rocks and such Ni is further enriched
during weathering. The Co in such deposits may also be substantial. The Co and Ni present in weathered rocks is generally present in Mn oxides (up to 12% Co and 8.6% Ni in lithiophorite–asbolan: Wells and Scott 2001), Fe oxides (up to 1.0% Co and 3.8% Ni in goethite; Scott 1994; Nickel et al. 1977) and smectite (up to 500 ppm Co and 2.9% Ni; Nickel et al. 1977). The Co and Ni contents of kaolinite are generally low. Thus they are <200 ppm even when the precursor chlorite contains up to 6000 ppm Co and 600 ppm Ni at Mount Dore, Qld (Scott 1988). Vermiculite contains up to 7400 ppm V at Mt Percy, WA (M le Gleuher pers. comm. 2007). Copper AND :INC follow Fe and Mg in primary rocks (radii of divalent ions = 0.73 and 0.74 versus 0.61 and 0.72 Å, respectively). Weathering of sulfide mineralisation generally leaches these elements, but they may be stabilised under alkaline conditions (e.g. by carbonates). Supergene enrichment of Cu is common at the base of oxidation of sulfide deposits (even when the Cu content of the primary ore is not substantial). The Cu and Zn present in weathered rocks is generally present in Mn oxides (up to 20% Zn in chalcophanite and 0.5% Cu in lithiophorite; Scott 1986; 1987d), Fe oxides (up to 2.3% Cu and 5.3% Zn in goethite and 0.7% Cu and 3.2% Zn in hematite: Scott and McQueen 2001; Scott 1986), alunite–jarosite minerals (up to 9.3% Cu in beaverite and 7.6% Zn in zincian alunite; Taguchi et al. 1972; Scott 1987c) and smectite (up to 2700 ppm Cu and 4.7% Zn; Nickel et al. 1977; Scott and Taylor 1987a), but not kaolinite (generally <100 ppm; Wedepohl 1978) although 4.5% Cu (and 4.7% Ni) is reported in kaolinitic clay at the Perseverance Ni deposit, WA (Nickel et al. 1977). Vermiculite incorporates up to 1.0% Cu at Boddington, WA (le Gleuher et al. 2005). Copper and Zn contents of kaolinite are at least an order of magnitude lower than in the Fe oxides at the Mount Dore Cu-Ag deposit, Qld (Scott 1988).The presence of other elements within the goethite structure may affect abundances of the base metals, but no direct relationship appears to be present for phosphate/Cu and Zn interaction (Scott 1986), the incorporation of several percent Al into the goethite structure appears to facilitate the incorporation of Cu and Zn (e.g. at New Cobar; Scott and McQueen 2001). Zinc ~2.4% is reported in secondary siderite at the Pegmont Pb-Zn deposit, Qld (Scott and Taylor 1987a).
443
444
Regolith Science
SECOND TRANSITION SERIES Analytical method (detection limit)
Element
Atomic number
Charge
Ionic radius (Å )
Y
39
+3
0.90
Xenotime, apatite (1400 ppm)
Xenotime. Soluble as carbonate, precipitated with Mn oxides
ICPMS (0.05 ppm)
Zr
40
+4
0.72
Zircon, xenotime (4%), aegirine (1.4%)
Zircon, xenotime
ICPMS (0.1 ppm)* XRF (5 ppm)
Nb
41
+3 +4 +5
0.72 0.68 0.64
Ilmenorutile, ilmenite (1000 ppm) 1, sphene (6100 ppm), rutile (4.2%)2, Ti-mte (3500 ppm), cassiterite (1.3%)
Ilmenorutile, rutile cassiterite, zircon
ICPMS (0.05 ppm)* XRF (5 ppm)
Mo
42
+3 +4 +5 +6
0.69 0.65 0.61 0.59
Molybdenite, pyrite (2300 ppm), chalcopyrite (1500 ppm)
Oxidised to +6 oxidation state and incorporated into Fe/ Mn/Al oxides. Wulfenite in gossans
ICPMS (0.05 ppm)
Tc
43
+4
0.65
Does not naturally exist
Ru
44
+3 +4 +5
0.68 0.62 0.57
Native element/alloy, chromite (0.5 ppm)
Native element/alloy is stable
Fire assay/ ICPMS (0.002 ppm)
Rh
45
+3
0.68
Native element/alloy, chromite (0.5 ppm)
Native element/alloy is stable
Fire assay/ ICPMS (0.002 ppm)
Pd
46
+2 +3 +4
0.86 0.76 0.62
Native element/alloy, zircon (5 ppm)
Native element/alloy is stable
Fire assay/ ICPMS (0.001 ppm)
Ag
47
+1 +2
1.15 0.94
Sulfides (especially tetrahedrite/ tennantite,galena and chalcocite), sulfosalts tellurides, gold alloys
More soluble than Au in saline solutions. Forms native Ag, chlororargyrite and alunite–jarosites
Aqua regia/ ICPMS (0.01 ppm) ICPMS (0.1 ppm)
Cd
48
+2
0.95
Sphalerite (6%)3
Freed when sphalerite weathers, but may be stabilised in carbonates
ICPMS (0.02 ppm)
Common host primary minerals
Common occurrence in regolith
*Zr and Nb determined by ICPMS may be low due to incomplete dissolution of its host minerals (e.g. zircon and Ti oxides). 1 Data from Telfer, WA (Scott et al. 1993) 2 Data from Big Bell, WA (Scott and Radford 1995) 3 Data from Miclere, Queensland (K.M. Scott, unpublished data, 1994)
Appendix 2: Regolith geochemistry of elements
Yttrium is generally considered with the lanthanides, and consideration of Y contents relative to the lanthanides may resolve some parental questions (see lanthanide series below). During weathering, the heavier lanthanides are more soluble than the lighter ones (and Y). Zirconium is mainly present in zircon and is generally residually concentrated during weathering. Thus Zr is generally regarded as immobile during weathering; however, the presence of corroded zircons in hydrothermally altered samples suggests that this may not always be true. Ti/Zr ratios in weathered rocks can be used as an indicator of the original rock type (e.g. Hallberg 1984: see also Section 6.6). However, accurate determinations of Ti and Zr are required – determining these elements by any method requiring dissolution will be unreliable due to incomplete dissolution of the Ti oxides and zircon. Niobium may be residually concentrated as host ilmenorutile; rutile, cassiterite and zircon accumulate during weathering, but sphene is less stable during weathering. Both Nb and Ta tend to enter Ti sites (radii Nb5+:Ta5+:Ti4+ =0.64:0.64:0.61 Å) thus Ta also tends to be concentrated in these minerals. However, Nb is more soluble than Ta so that, for the same rock type, Nb/Ta tends to be higher in the resultant soils in arid areas than in more humid areas. -OLYBDENUM is present as molybdenates in weathered material and may be particularly enriched in
soils associated with porphyry Cu deposits and granite- associated deposits. Up to 1600 and 640 ppm Mo occurs in Mn oxides and goethite, respectively, at Mount Dore, Qld (Scott 1988). 4ECHNETIUM (Tc99) may exist as a natural fission product of U. 2UTHENIUM RHENIUM AND PALLADIUM are part of the platinoid group (along with Os, Ir and Pt) which tend to be enriched in the minerals of ultramafic rocks. 3ILVER is commonly present in gold alloys, but its greater solubility in the presence of Cl, can lead to the separation from Au in saline weathering environments and the formation of secondary Au-rich, Ag-poor rims about more Ag-rich cores in gold grains (Freyssinet and Butt 1988). Up to 600 ppm and 1600 ppm Ag may occur in coexisting goethite and alunite–jarosite minerals at Elura, NSW (Scott and Taylor 1989), although at the Mt Leyshon Au deposit, Qld, up to 1500 ppm Ag may occur in colloform goethite whereas alunite contains only 25 ppm Ag (Scott 1987b). Cadmium follows Ca (radii =0.95 versus 1.00 Å). It concentrates in sphalerite and is dispersed during the acidic conditions of sulfide weathering but Cd and Zn may be retained under alkaline weathering conditions (e.g. at Dugald River; Taylor and Scott 1983). The relative dispersion of Zn and Cd is not well documented.
445
446
Regolith Science
THIRD TRANSITION SERIES Analytical method (detection limit)
Element
Atomic number
Charge
Ionic radius (Å )
La
57
+3
1.03
Monazite (16%)1
Monazite
ICPMS (0.01 ppm)
Hf
72
+4
0.71
Hafnon-zircon
Zircon, Hf-rich zircon more susceptible to breakdown during weathering
ICPMS (0.01 ppm)*
Ta
73
+3 +4 +5
0.72 0.68 0.64
Tantalite, ilmenite/pseudorutile (1300 ppm),1 cassiterite (6.9%), sphene (600 ppm), Ti mte (340 ppm)
Tantalite, rutile
ICPMS (0.01 ppm)*
W
74
+4 +5 +6
0.66 0.62 0.60
Wolfamite, scheelite, rutile (9.5%)2
Wolfamite, rutile. Scheelite weathers readily to W-rich Fe oxides
ICPMS (0.1 ppm)*
Re
75
+4 +5 +6 +7
0.63 0.58 0.55 0.53
Molybdenite, pyrrhotite/ pentlandite
Follows Mo during oxidation
ICPMS (0.002 ppm)
Os
76
+4 +5 +6
0.63 0.58 0.55
Chromite, pyrrhotite/pentlandite
Chromite
Fire assay/ ICPMS (0.002 ppm)
Ir
77
+3 +4 +5
0.68 0.63 0.57
Chromite (5 ppm), pyrrhotite/ pentlandite
Chromite
Fire assay/ ICPMS (0.002 ppm)
Pt
78
+2 +4
0.80 0.63
Chromite (1 ppm), pyrrhotite/ pentlandite
Chromite
Fire assay/ ICPMS (0.0005 ppm)
Au
79
+1 +3
1.37 0.85
Native Au, electrum
Native Au, electrum
Fire assay/ ICPMS (0.001 ppm)
Hg
80
+1 +2
1.19 1.02
Cinnabar, amalgam, sphalerite
Cinnabar, amalgam, calomel
Aqua regia/ ICPMS (0.01 ppm)
Common host primary minerals
Common occurrence in regolith
*Hf, Ta and W determined by ICPMS may be low due to incomplete dissolution of their host minerals (e.g. zircon, Ti oxides and wolframite). 1 Data from Telfer, WA (Scott et al. 1993) 2 Data from Kalgoorlie, WA (Scott and Radford 1995)
Appendix 2: Regolith geochemistry of elements
(AFNIUM precipitates at higher pH than Zr and hence may be separated from it. However, usually neither element is highly soluble. Both are found in Mn oxide nodules. Hf/Zr ratios in zircons can sometimes be used to determine the parentage of regolith material (Figure 6.15). 4UNGSTEN may be incorporated into Mn oxides adjacent to hot springs and W contents have been suggested (Wedepohl 1978) as a discriminant between hypogene and supergene Mn oxides. Indeed, up to 400 ppm W is present in colloform cryptomelane from gossan at the Bluegum epithermal Au deposit, Qld (Scott, unpublished data, 1991). Although, up to 1200 ppm W is present in goethite (at Panglo, WA; Scott 1990a), 300 ppm W has been observed in coexisting hematite and goethite (at depth in the weather-
ing profile at New Cobar, NSW; Scott and McQueen 2000), suggesting that there is no marked preference for W between the two Fe oxide types. Pt/Pd increases during weathering under arid saline conditions; i.e. Pd may be separated from Pt during weathering. Gold may be separated from Ag during weathering under saline weathering conditions (see Ag above). -ERCURY in amalgams stabilises Ag (e.g. Elura; Scott 1987a: Chapman and Scott 2005). The sulfide, cinnabar, is stable to weathering, even when other sulfides have weathered (Wedepohl, 1978). Hg (5 ppm) may be retained in cryptomelane in the deeply weathered Conta História Mn deposit, Minas Gerais, Brazil (Cabral et al. 2002).
447
448
Regolith Science
LANTHANIDE SERIES
Charge
Ionic radius (Å )
58
+3 +4
Pr
59
Nd
Analytical method (detection limit)
Common host primary minerals
Common occurrence in regolith
1.01 0.87
Monazite, basnaesite, davidite
Florencite. Incorporated into Mn oxides
+3 +4
0.99 0.85
Monazite (4.1%)1
60
+3
0.98
Monazite (9%),1 xenotime (1200 ppm)1
Pm
61
+3
0.97
Sm
62
+2 +3
1.19 0.96
Eu
63
+2 +3
1.17 0.95
Gd
64
+3
0.94
Tb
65
+3 +4
0.92 0.76
Dy
66
+2 +3
1.07 0.91
Ho
67
+3
(0.89)
Er
68
+3
0.89
Tm
69
+2 +3
1.01 0.88
Fusion/ICPMS (0.1 ppm)
Yb
70
+2
1.02
Fusion/ICPMS (0.1 ppm)
Lu
71
+3
0.86
Fusion/ICPMS (0.05 ppm)
Element
Atomic number
Ce
1 Data
Monazite (2.4%)1, xenotime (4000 ppm)
ICPMS (0.01 ppm) Fusion/ICPMS (0.1 ppm) Fusion/ICPMS (0.05 ppm)
Florencite
Fusion/ICPMS (0.1 ppm)
Fusion/ICPMS (0.1 ppm) Fusion/ICPMS (0.1 ppm)
Xenotime
Fusion/ICPMS (0.1 ppm) Fusion/ICPMS (0.05 ppm)
Xenotime
Fusion/ICPMS (0.1 ppm) Fusion/ICPMS (0.1 ppm)
Xenotime
Fusion/ICPMS (0.1 ppm)
from Telfer, WA (Scott et al. 1993)
The lanthanide series elements (plus La and Y) are generally determined as a group and normalised relative to chondrite rare element abundances to determine rare element fractionation in rocks. Abundances of heavy and light REE and Ce/Nd ratios may reflect the degree of weathering (Section 6.8; Figures 6.20; 6.21).
#ERIUM can be oxidised to +4 OS and hence separated from other rare earth elements, which generally occur as +3 OS only during weathering (Section 6.8).
Appendix 2: Regolith geochemistry of elements
ACTINIDE SERIES
Element
Atomic number
Charge
Ionic radius (Å )
Th
90
+4
0.94
Pa
91
+3 +4 +5
1.04 0.90 0.78
U
92
+3 +4 +5 +6
1.03 0.89 0.76 0.73
Np–Lw
93–103
1
Analytical method (detection limit)
Common host primary minerals
Common occurrence in regolith
Thorite, allanite, monazite (900 ppm),1 sphene (500 ppm), epidote (500 ppm), zircon (600 ppm), xenotime (500 ppm),1 apatite (250 ppm)
Immobile, residually concentrated in soils. Fe oxides
ICPMS (0.01 ppm)
uraninite, zircon (6000 ppm), monazite (3000 ppm) allanite (1000 ppm), epidote (200 ppm), xenotime (1600 ppm),1 apatite (100 ppm)
+4 OS immobile but +6 OS is soluble. Fe oxides
ICPMS (0.01 ppm)
Not stable naturally
Data from Telfer, WA (Scott et al. 1993)
4HORIUM in Al goethite (from ferruginous material in sandstone near Sydney, NSW) may reach 1500 ppm (Scott, unpublished data, 1990). However, up to 480 ppm Th is present in hematite and maghemite from ferruginous material from Wagga Tank, NSW (Scott, unpublished data, 1989). Such results explain the common occurrence of elevated Th response mapping drainage channels during aerial gamma-ray surveys (Dickson and Scott 1997). Uranium up to 550 ppm and 250 ppm occurs in colloform and residual goethite from a gossan at Arrawa, NSW (Dickson and Scott 1990). However, up to 300 and 160 ppm U occurs in hematite and goethite (respectively) in subcropping gossanous material at Wagga Tank, NSW (Scott, unpublished data, 1989), suggesting that hematite may be the better host. Th/U ratios increase with increasing differentiation in igneous rocks, but the ratio is commonly ~3–5. Both elements are commonly enriched in soils relative to their parental rocks (Dickson and Scott 1997).
REFERENCES ALS (Australian Laboratory Services) (2008). Web site. <www.alschemex.com>
Bolton D and Alexander J (2005). Windimurra vanadium deposits, Murchison Region, WA. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 121–123. CRC LEME, Perth. Butt CRM and Gole MJ (1985). Helium in soil and overburden gas as an exploration pathfinder – an assessment. Journal of Geochemical Exploration 24, 141–173. Butt CRM, Horwitz RC and Mann AW (1977). ‘Uranium occurrences in calcrete and associated sediments in Western Australia.’ Report FP 16. CSIRO Division of Mineralogy, Floreat Park, WA. Butt CRM, Scott KM, Cornelius M and Robertson IDM (2005). Sample media. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 53–79. CRC LEME, Perth. Cabral AR, Lehmann B, Sattler CD, Pires FRM and Kaneko K (2002). Hg-Tl-bearing manganese oxide from Conta Historia manganese deposit, Quadrilátero Ferrifero, Minas Gerais, Brazil. Transactions of the Institution of Mining and Metallurgy (Section B: Applied Earth Sciences) 111, B123–127. Chapman J and Scott K (2005). Supergene minerals from the oxidised zone of the Elura (Endeavor)
449
450
Regolith Science
lead-zinc-silver deposit. Australian Journal of Mineralogy 11, 83–90. Corbett JA, Lambert IB and Scott KM (1975). ‘Results from analyses of rocks from the McArthur River area, Northern Territory.’ Technical Communication 57. CSIRO Minerals Research Laboratories. North Ryde, New South Wales. Dart RC, Barovitch KM and Chittleborough D (2005). Pedogenic carbonates, strontium isotopes and the relationship with Australian dust processes. In Regolith 2005-Ten Years of CRC LEME. (Ed. IC Roach) pp. 64–66. CRC LEME, Perth. Dart RC, Barovitch KM, Chittleborough D and Hill SM (2007). Calcium in regolith carbonates of central and southern Australia: Its source and implications for global carbon cycle. Palaeogeography, Palaeoclimatology, Palaeoecology 249, 322–334. De Waal SA and Calk LC (1973). Nickel minerals from Barberton, South Africa. VI. Liebenbergite, a nickel olivine. American Mineralogist 58, 733–735. Dickson BL (1990). ‘Radium in groundwaters. Chapter 4-2- The environmental behaviour of radium.’ Technical Report Series 310, pp. 335–372. IAEA, Vienna. Dickson BL and Scott KM (1990). ‘An unusual K+Th anomaly within an airborne radiometric survey north of Cobar, NSW.’ Restricted Report 193R. CSIRO Division of Exploration Geoscience, Perth. Dickson BL and Scott KM (1997). Interpretation of aerial gamma ray surveys – adding the geochemical factors. AGSO Journal of Australian Geology and Geophysics 17, 187–200. Dickson BL, Giblin AM and Snelling AA (1987). The source of radium in anomalous accumulations near sandstone escarpments, Australia. Applied Geochemistry 2, 385–398. Douglas GB, Butt CRM and Gray DJ (2005). Mulga Rock uranium and multielement deposits, Officer Basin, WA. In Regolith Expression of Australian Ore Systems. (Eds CRM Butt, IDM Robertson, KM Scott and M Cornelius) pp. 415–417. CRC LEME, Perth. Ewers GR (1991). Oxygen isotopes and the recognition of siliceous sinters in epithermal ore deposits. Economic Geology 86, 173–178. Freyssinet P and Butt CRM (1988). ‘Morphology and geochemistry of gold in a laterite profile, Reedy
Mine, Western Australia.’ CSIRO Division of Exploration Geoscience Restricted Report MG 58R. (Reissued as Open File Report 4, 1998. CRC LEME, Perth). Genalysis (2008). Web site. <www.genalysis.com.au> Gray DJ (2003). Naturally occurring Cr6+ in shallow groundwaters of the Yilgarn Craton, Western Australia. Geochemistry: Exploration. Environment, Analysis 3, 359–368. Gulson BL (1984). Assessment of massive sulphide base metal targets using lead isotopes in soils. Journal of Geochemical Exploration 22, 291–313. Hallberg JA (1984). A geochemical aid to igneous rock identification in deeply weathered terrain. Journal of Geochemical Exploration 20, 1–8. Henley KJ (1970). Cuperiferous sericite from the Sar Cheshmeh porphyry copper ore, Kerman Province, Iran. Mineralogical Magazine 37, 945–947. Khider K and McQueen KG (2006). Gold dispersion in calcrete-bearing regolith of the Girlambone Region, western NSW. In Regolith 2006. (Eds RW Fitzpatrick and P Shand) pp. 190–195. CRC LEME, Perth. Klein TL and Criss RE (1988). An oxygen isotope and geochemical study of meteoric-hydrothermal systems at Pilot Mountain and selected other localities, Carolina Slate Belt. Economic Geology 83, 801–821. Lide DR (Ed.) (1997). CRC Handbook of Chemistry and Physics. 78th edn). pp. 12-14–12-16. CRC Press, Boca Raton, Florida. le Gleuher M, Anand R, Eggleton RA and Radford N (2005). Mineral hosts for gold and trace metals in regolith. In 22nd IGES: From Tropics to Tundra. 19–23 September, Perth. pp. 67–68. Promaco Conventions Pty Ltd, Perth. Lintern MJ, Sheard MJ and Chivas AR (2006). The source of pedogenic carbonate associated with goldcalcrete anomalies in the western Gawler Craton, South Australia. Chemical Geology 235, 299–324. McCarthy JH, Lambe RN, and Dietrich JA (1986). A case study of soil gases as an exploration guide in glaciated terrain – Crandon Massive Sulfide Deposit, Wisconsin. Economic Geology 81, 408–420. McQueen KG (2006). ‘Calcrete geochemistry in the Cobar-Girilambone region, New South Wales.’ Open File Report 200. CRC LEME, Perth.
Appendix 2: Regolith geochemistry of elements
McQueen KG, Munro DC, Gray D and Le Gleuher M (2004). Weathering-controlled fractionation of ore and pathfinder elements Part II: The lag story unfolds. In Regolith 2004. (Ed. IC Roach) pp. 241– 246. CRC LEME, Perth. Nickel EH, Allchurch PD, Mason MG and Wilmshurst JR (1977). Supergene alteration at the Perseverance nickel deposit, Agnew, Western Australia. Economic Geology 72, 184–203. Reimann C and de Caritat P (1998). Chemical Elements in the Environment: Factsheets for the Geochemist and Environmental Scientist. Springer-Verlag, Berlin. Saxby JD (1976). The significance of organic matter in ore genesis. In Handbook of Stratabound and Stratiform Ore Deposits. Vol 2. Geochemical Studies. (Ed. KH Wolf) pp.111–133. Elsevier, Amsterdam. Scott KM (1986). Elemental partitioning into Mnand Fe-oxides derived from dolomitic shale-hosted Pb-Zn deposits, northwestern Queensland, Australia. Chemical Geology 57, 395–414. Scott KM (1987a). ‘Precious metal distribution in weathered Elura Zn-Pb- Ag ore.’ Restricted Investigation Report 1688R. CSIRO Institute of Energy and Earth Resources, Melbourne. Scott KM (1987b). ‘Electron microprobe investigations of outcrop at the Mt Leyshon Gold Deposit, N.E.Qld.’ Restricted Investigation Report 1711R. CSIRO Division of Mineral Physics and Mineralogy, North Ryde, NSW. Scott KM (1987c). Solid solution in, and classification of, gossan-derived members of the alunite-jarosite family, northwest Queensland, Australia. American Mineralogist 72, 178–187. Scott KM (1987d). Significance of a lithiophorite interface between cryptomelane and florencite. American Mineralogist 72, 429–432. Scott KM (1988). ‘Mineralogical studies of primary and secondary minerals and their constraints upon the genesis of the Mount Dore Cu-Ag Deposit, N.W. Queensland.’ Restricted Investigation Report 1763R. CSIRO Division of Exploration Geoscience, Perth. Scott KM (1990a). ‘The mineralogical and geochemical effects of weathering on volcanics from the Panglo Deposit, Eastern Goldfields, WA.’ CSIRO Division of Exploration Geoscience Restricted
Report 143R. (Reissued as Open File Report 24, 1998. CRC LEME, Perth.) Scott KM (1990b). ‘Electron microprobe studies of minerals from weathered profiles, Parkinson Pit and environs, Mt Magnet, WA.’ CSIRO Division of Exploration Geoscience Restricted Report 147R. (Reissued as Open File Report 25, 1998. CRC LEME, Perth.) Scott KM (1991). ‘Trace and minor element contents of sulfides, 4 and 5 Levels, Elura Mine.1. Electron microprobe studies.’ Restricted Report 224R. CSIRO Division of Exploration Geoscience, Perth. Scott KM (1994). ‘Alteration features in gangue minerals as an aid to exploration in the Copper Canyon-Greenmount area of N.W. Queensland.’ Report 54R. CSIRO Exploration and Mining, Perth. Scott KM and McQueen KG (2000). ‘The mineralogy and geochemistry of the New Cobar Au-Cu mineralisation in the regolith and exploration implications for the Cobar District, western NSW.’ CRC LEME Restricted Report 137R/ CSIRO Exploration and Mining Report 728R. CRC LEME, Perth/ CSIRO Exploration and Mining, Perth. Scott KM and McQueen KG (2001). ‘Mineralogy and geochemistry of the New Cobar mineralisation in the oxidate and supergene zones, Cobar District, western NSW.’ CRC LEME Restricted Report 167R/ CSIRO Exploration and Mining Report 845R. (Reissued as Open File Report 213, 2008. CRC LEME, Perth.) Scott KM and Radford NW (1995). ‘Rutile compositions in the Big Bell and Kalgoorlie Gold Deposits, W.A.: Occurrences of some unusual V-rich rutiles’. Report 131R. CSIRO Exploration and Mining, Perth. Scott KM and Radford NW (2007). Rutile compositions at the Big Bell Au deposit as a guide for exploration. Geochemistry: Exploration. Environment, Analysis , 353–361. Scott KM and Rampe M (1984). Integrated mineralogical and geochemical exploration for tin in the Bygoo region of the Ardlethan Tin Field, southern NSW, Australia. Journal of Geochemical Exploration 20, 337–354.
451
452
Regolith Science
Scott KM and Taylor GF (1987a). The oxidized profile of BIF-associated Pb-Zn mineralization: Pegmont, northwest Queensland, Australia. Journal of Geochemical Exploration 27, 103–124. Scott KM and Taylor GF (1987b). ‘Weathering of siliceous ore and wall rocks on 1 Drill Level, Elura ZnPb-Ag deposit.’ Restricted Investigation Report 1715R. CSIRO Division of Mineral Physics and Mineralogy, North Ryde, NSW. Scott KM and Taylor GF (1989). ‘Weathering of pyritic and pyrrhotitic ore on 1 Drill Level and above, Elura Zn-Pb-Ag orebody, NSW.’ Restricted Report 55R. CSIRO Division of Exploration Geoscience, Perth. Scott KM, Ashley PM and Lawie DC (2001). The geochemistry, mineralogy and maturity of gossans derived from volcanogenic Zn-Pb-Cu deposits of the eastern Lachlan Fold Belt, NSW, Australia. Journal of Geochemical Exploration 72, 169–191. Scott KM, Ramsden AR and French DH (1993). ‘Preliminary results from rutile and other resistate mineral investigations at Telfer, Western Australia.’ Restricted Report 379R. CSIRO Division of Exploration Geoscience, Perth. Spry PG and Scott SD (1986). Zincian spinel and staurolite as guides to ore in the Appalachians and Scandinavian Caledonides. Canadian Mineralogist 24, 147–163. Taguchi Y, Kizawa Y and Okada N (1972). On beaverite from the Osarizawa mine. Journal of the Mineralogical Society of Japan 10, 313–325. Taylor GF and Scott KM (1983). Weathering of the zinc-lead lode, Dugald River, northwest Queens-
land.II. Surface mineralogy and geochemistry. Journal of Geochemical Exploration 18, 111–130. Taylor GF, Wilmshurst JR, Togashi Y and Andrew AS (1984). Geochemical and mineralogical haloes about the Elura Zn-Pb-Ag orebody, western New South Wales. Journal of Geochemical Exploration 22, 265–290. Taylor SR (1966). The application of trace element data to problems in petrology. Physics and Chemistry of the Earth 6, 133–213. Urban AJ, Hoskins BF and Grey IE (1992). Characterization of V-Sb-W-bearing rutile from the Hemlo Gold Deposit, Ontario. Canadian Mineralogist 30, 319–326. Vaasjoki M and Gulson BL (1985). Evaluation of drilling priorities using lead isotopes at the Lady Loretta lead-zinc-silver deposit, Australia. Journal of Geochemical Exploration 24, 305–316. Vasconcelos PM (1999). K-Ar and 40Ar/39Ar geochronology of weathering processes. Annual Reviews of Earth and Planetary Sciences 27, 183–229. Weast RC (Ed.) (1989). CRC Handbook of Chemistry and Physics. 70th edn., p. F-187, CRC Press, Boca Raton, Florida. Wedepohl KH (Ed.) (1978). Handbook of Geochemistry. Springer Verlag, Berlin. Wells MA and Scott KM (2001). ‘Mineralogy of the Mn-Ni-Co association in the Syerston Ni-Co-Pt deposit, New South Wales.’ CRC LEME Restricted Report 165R/ CSIRO Exploration and Mining Report 819R. CRC LEME, Perth/ CSIRO Exploration and Mining, Perth.
Index See also Appendix 1: Glossary of regolith terms and Appendix 2: Regolith geochemistry of elements
absorption, precipitation and ionic exchange 87–9 acid sulfate soils (ASS) 5, 59, 61, 134, 136, 139, 147, 307–8, 310, 322–8, 330 sulfur biogeochemistry in 144–6 acidity, soil 321–2 aeolion 282, 295 airborne electromagnetics (AEM) 235–7, 290–1, 333 alkaline earths 435–7 alkalis 434 allophane 63–4 Al-oxides 58–9 aluminium 59, 111, 113, 436 mobility 92, 184–5 aluminium-Fe oxyhydroxides 64 alunite 19–21, 61–3, 67, 93, 95 amorphous minerals 63–4 amphiboles 46–7 anatase 60–1 animals redistribution of regolith materials 185–7 see also fauna antimony 439 apatite fission track dating 12–13 argon 441 40Ar/39Ar dating 19–21 aridity, brief history 22–3 arsenates 62 arsenic 438–9 asteroids 394–6 bacterial action 81–2 bacterially induced mineralisation (BIM) 141 barium 435–6 base metals 347–350, 354 basins, sediment accumulation 10–11 bassanite 61 biochemical uptake and redistribution of regolith materials by plants 182–5
biofilms 196–7 biogenic dispersion 84, 86–7 biogenic processes 81–2 biogenic weathering 196–7 biogeochemical element cycles 128–36 carbon cycle 129–33 nitrogen cycle 133–4 sulfur cycle 134–6 biogeochemical cycling 182–3 biogeochemistry 151, 185, 204, 355–6 rhizosphere 180–1 sulfur, in acid sulfate soils 144–6 biogeomorphology 202–8 biointrusion 182–3 biomineralisation 141, 196 bioprospecting 182–3 bio-signatures 143–4 biota 4, 175–209 and regolith physical characteristics 198–202 role of, in regolith development 41–2 biotic crusts 196–7 biotic effects and processes at local or meso scales 181–202 biotic effects at broad spatial and long timescales 202–8 biotic paleoforms in regolith 181 biotic processes on regolith surfaces 196–8 biotic surface conditions 199 biotite 32, 51, 65, 77, 110–1 biotransfer 187–94 bioturbation 187–94 consequences within the regolith 194–6 invertebrate fauna 191–4 vertebrate fauna 190–1 boehmite 58–9, 64, 110, 350 borates 63 borehole geophysics 230–1, 239 boron 436 Bowen’s reaction series 76–7
454
Regolith Science
bromide 440 bryophytes 198 cadmium 445 calcrete (regolith carbonate) 36–7, 39, 63, 94, 112, 114, 122, 144, 147, 239, 297, 314, 359, 362, 366, 435–6 Au- anomalous 146–9, 357 groundwater 35 pedogenic 352–4 pedogenic, as a sampling medium 353–4 sampling procedures 354 Cambrian regolith 15 Canning Basin 2, 11, 16 carbon 437 carbon cycle 129–33 carbonate 5, 62–3, 141 see also calcrete (regolith carbonate) carbonate equilibrium 267–8 carbon-rich rocks, weathering 115 Carnarvon Basin 2 catena 284 cation exchange capacity (CEC) 54–8, 66, 88–9, 226 Cenozoic 2, 18–23 lava flows as dated reference surfaces 11 China 115–16, 326 chlorine 440 chlorite 47–50, 55 chromium 443 clay layer silicates 50–8 clay minerals associations with 95 strategy for quantifying 65–7 clays, interstratified 57–8 climate 2–3 climatic change global 18–23 weathering, effects of long-term 118–19 cobalt 443 Cobar region 121–4, 356 copper 443 cosmogenic isotopes 11–12 Darcy’s Law 252–6 databases 292–4 denudation rates, measuring long-term 10–13
deposition, effects on weatering profiles 115–16 descriptive 3D whole-of-landscape process models 333 descriptive soil–regolith models 329–30 detrital sedimentary rocks, weathering 113–15 diaspore 58 direct mineral precipitation 141–2 disordered clay 51, 67 dryland salinity 207–8 Eh-pH 82–3, 269–70 electrical conductivity 226–7 electrical geophysical surveying 229 electrical resistivity 220–1, 229–33 electrochemical dispersion 85–6 electromagnetic frequency and time domain 230–1 electromagnetic (EM) surveying techniques 235–7 element associations, regolith-related 92–5 element behaviour, fundamental controls 73–6 element dispersion 84–7 element dispersion-retention 76–91 absorption, precipitation and ionic exchange 87–9 basic processes 76–81 biogenic processes and bacterial action 81–2 Eh-pH 82–3 metal complexing 83 sulfide weathering 89–91 trace element behaviour during weathering 82 element variations during weathering 91–2 elements, regolith geochemistry of 433–49 actinide series 449 first transition series 442–3 second transition series 444–5 third transition series 446–7 group 1 elements (alkalis) 434 group 2 elements (alkaline earths) 435 group 3 elements 436–7 group 4 elements 437–8 group 5 elements 438–9 group 6 elements 439–40 group 7 elements (halides) 440–1 group 8 elements 441 lanthanide series 448 environmental applications of regolith maps 299–300 environmental isotopic tracers 273–4 Eromanga Basin 2, 10–11, 18, 39
Index
erosion effects on weathering profiles 115–16 and vegetation 202–4 Eucla Basin 2, 11, 21–2 evaporative zones, associations with 94 explanatory soil–regolith models 330–1 extraterrestrial regolith 377–402 asteroids 394–6 mapping 396–7 Mars 380–7, 400 Moon 377–80, 400 regolith processes on rocky and icy bodies 397–8 solar system processes 400–1 Titan 391–4, 400 Venus 387–91 fauna bioturbation by invertebrate 191–4 bioturbation by vertebrate 190–1 Fe-oxides 59–60 associated trace elements 93–4, 435–49 see also ferrihydrite; goethite; hematite; lepidocrocite; maghemite; magnetite feldspars 32, 46, 51, 77, 110–12 feldspathoids 46 ferricretes 18–19, 35–9, 40–1, 224, 239, 282, 347, 358, 359 ferrihydrite 59, 88, 310 ferruginous weathering products 19 fire effects on regolith 206–7 fluoride 440 gallium 436–7 garnet 46 Gawler Craton 22, 148, 297, 353–4, 357, 359, 362 geobotany 182–3, 355 geochemical anomalies 95–101 space, time and source aspects 100–1 geochemical data multivariate statistical analysis 98 univariate statistical methods 96–8 geochemical dispersion under different weathering regimes 119–20 geochemical exploration, regolith sampling 341–70 gossans and ironstones 347–50 groundwater 364–7
indurated materials 358–9 interface sampling 359–60 introduction 341–2 lag 352 lateritic residuum 350–2 pedogenic calcrete 352–4 sample preparation and contamination 367–70 soil 342–5 stream sediments 363–4 transported overburden 360–3 vegetation 355–7 weathered bedrock (saprolite) 345–7 geochemical reactions in groundwater 266–71 geochemistry of elements see elements, regolith geochemistry of geological cycle and weathering 3 geology maps, surficial 286 see also regolith geology geomicrobial cycling of Au 146–9 geomicrobial reactions 175–7 geomicrobiology of the regolith 4, 127–51 biogeochemical element cycles 128–36 case studies 144–9 conculsions and applications 151 introduction 127 isotopic bio-signatures 143–4 microbial mineral formation 140–3 microorganisms 127–8, 136–40 molecular tools and techniques 149–51 weathering 136–40 geomorphic mapping 284 geomorphology, importance of 287–8 geophysical abbreviations 221 geophysical survey planning 240–1 geophysical techniques 219–20 geophysical technologies, principles and applications 228–40 germanium 437 gibbsite 58–9, 65, 77, 88, 92, 110–11, 350 Gippsland Basin 2, 17 goethite 59–60, 68,77, 82, 88, 92–4, 107, 110–16, 141, 224, 309, 350, 438–40, 443 gold (Au) 5, 14, 19, 64, 74, 83, 85, 91, 93, 95–6, 107, 143–4, 184, 224, 232, 297–8, 331, 341, 343–4, 349, 352–4, 357, 359, 362, 364–6, 370, 447
455
456
Regolith Science
geomicrobial cycling of 146–9 microbially induced precipitation of secondary 142–3 Goldschmidt’s classification 74–6 gossans 62, 347–50 granitoids, weathering 31–33, 110 gravity 223 gravity surveying 228, 230–1 ground acoustic penetration (GAP) 221, 228, 240 ground magnetic surveys 229 ground-penetrating radar (GPR) 227, 230–1, 237, 291 groundwater calcrete 353 chemical components 260 flow systems 258–9 and geochemical exploration 364–7 geochemical reactions 266–71 graphical display 260–1 and regolith 5, 256 as a sample medium 365–6 sampling 262–4 sampling procedures 366–7 sources of 257–8 storage 258 groundwater-associated salinity (GAS) 315, 318 gypsum 45, 61–2, 77, 80, 94, 265, 295, 310, 314, 330, 332, 354 hafnium 447 halide 62–3 halloysite 47–50 hardpan 35, 58, 224, 284, 358–9 helium 441 hematite 60, 68, 88, 92–3, 108, 309–10 hisingerite 51, 64 hydrobiogeochemical cycles 204–5 hydrochemical processes in the regolith 265–6 hydrochronology 274–5 hydrogeochemistry master variables 261–2 of the regolith 259–65 hydrogeological considerations 262–4 properties of the regolith 256–9 hydromorphic dispersion 84–5
hydroxides 58–61 hyperspectral techniques in regolith studies 67–9, 290 illite 55–7 imogolite 64 indium 437 indurated materials and geochemical exploration 358 interface sampling 359–60 iodine 440 ion exchange reactions 270–1 ionic radii 75–6, 84–5, 89, 415–28 iron 443 iron mobility 184–5 ironstones 347–50 irrigation-associated salinity (IAS) 315–18 isotopic bio-signatures 143–4 isotopic ratios 8, 273–5, 349–50, 435, 437–9 jarosite–natrojarosite 61, 324–5, 330, 332 kaolin 57 kaolinite 47–50, 67–8, 110 disorded 67, 362 karst 115 Lachlan Fold Belt 161, 232, 355, 357, 366 lag 352 land management planning, pictorial manuals 333–4 land systems mapping 285–6 landform models, regolith 286 landforms Australian 7–10 Cenozoic 18–23 continuous exposure or burial and exhumation 14–15 mapping principles 281–2 and Precambrian regolith 13–14 survival of ancient 23 landscape 4 evolutionary processes 331–3 field indicators and soils 328–9 history and Mesozoic continental breakup 16–18
Index
Permo-Carboniferous inheritance 15–16 and regolith 31–9 landscape-based mapping 282–6 lateritic residuum 350–2 lead 437–8 lepidocrocite 60, 93, 310 lichens 198 litter fall/dams and redistribution of regolith materials 186–7 low molecular weight organic acids (LMWOAs) 137–9, 177–9, 264 mafic rocks, weathering 110–12 maghemite 60, 64, 94, 207, 220, 222–4, 227, 230, 290, 310, 350, 352 magnetics 220, 223, 227–9, 230–1, 239, 242, 297–8, 333 magnetic resonance sounding (MRS) 240 magnetic surveying 228–9 magnetite 60, 141, 223, 225, 229 magnetometric resistivity (MMR) 233 manganese 443 mobility 184–5 mapping 227 compilation and interpolation 292 data collection 291–2 extraterrestrial regolith 396–7 geomorphic 284 implications for 227 information, presentation of 292–9 land systems 285–6 landform principles 281–2 landscape-based 282–6 methods 291–2 modellers 300 new directions 297–9 regolith maps 294–7, 299–300 regolith-landform 281, 286–99 scale 288–90 soil 284–5, 298 surficial geology maps 286 maps environmental applications 299–300 interpretive 296–7 mineral exploration 299 regolith 294–7, 299–300
standard 294–6 three-dimensional regolith 299 user groups 299–300 Mars 380–7, 399–400 current surface processes 384–7 regolith architecture 382–4 rover observations 387 surface chronology 380–2 surface features 382 mechanical dispersion 87 Menzies Line 335 mercury 447 Mesozoic continental breakup 16–18 weathering profiles 17 metal complexing 83 micas 47–50 microbial mineral formation 140 microbially induced precipitation of secondary Au 142–3 microorgamisms 127–8, 136–40, 175–81 and weathering 175–7 microtremor array method (MTM) 239 mineral dissolution, effect of organic ligands 139 mineral exploration industry 299 mineral weathering 64–5, 136–40 Mn minerals, associations with secondary 94 Mn-oxides 61 molecular microbial tools and techniques 149–51 molybdenum 445 montmorillonite 49, 51, 53–5, 68, 89 Moon, the 377–80, 399, 400 architecture of the lunar surface 378–9 chronology of the lunar surface 378 composition of the lunar regolith 379–80 polar ice deposits 380 Mt Isa region 349–50, 355 Murray Basin 2, 11, 39, 366 muscovite 55–7 natural resource management 275–6 nickel 443 niobium 445 nitrates 63 nitrogen 438 nitrogen cycle 133–4
457
458
Regolith Science
non-groundwater-associated salinity (NAS) 315, 318–20 olivine 46, 65, 77, 386 organic acids 177–8 orthosilicates 46 overburden, transported 360–3 oxides 58–61 oxygen 439 palladium 445 pedogenic calcrete 352–4 Perth Basin 2, 10–11, 14, 19, 239 petrophysical properties of regolith materials 220–8 phosphates 62–3, 141 minerals, associations with 95 phosphorus 438 Piper diagram 263 plants biochemical uptake and redistribution of regolith materials by 182–5 growth of roots 186, 200–2 hydraulic lift by vegetation 200 redistribution of regolith materials 185–7 surficial root mats 197–8 see also vegetation point of zero change (pzc) 87–88, 137, 271 polonium 440 potassium 434 Precambrian regolith 13–14, 15 predictive soil–regolith models 331–3 Pt/Pd 447 pyroxenes 46–7 qanats 275, 276 quartz 46 radiometrics 227, 230–1, 238–9, 290, 333 radium 436 radon 442 rainfall inputs 265 redox conditions 129, 137–8, 270 redox reactions 79, 80–1, 90, 262, 268–70 regolith in Australia 7–10, 23
biotic paleoforms 181 and biota 4, 175–209 Cambrian 15 characteristics 223–7, 281–2 defined 1–2 development and role of biota 41–2 and dryland salinity 207–8 as a dynamic medium 271–3 economic deposits within 5–6 fire effects 206–7 and geomicrobiology 4 and groundwater 256 information, presentation of 292–9 and landscape 31–9 mapping 227 maps 294–7, 299–300 physical characteristics 198–202 Precambrian 13–15 processes on rocky and icy bodies 397–8 surfaces, biotic processes on 196–8 structure of 105–9 terminology 6 vegetation effects 206–7 see also extraterrestrial regolith regolith and water 4–5, 207–8, 251–76 environmental isotopic tracers 273–4 geochemical reactions in groundwater 266–71 groundwater 256 hydrochemical processes 265–6 hydrochronology 274–5 hydrogeochemistry 259–65 hydrogeological properties 256–9 introduction 251 water and natural resource management 275–6 water cycle 251–6 regolith carbonate see calcrete (regolith carbonate) regolith description and mapping 281–300 characteristics 281–2 geomorphic mapping 284 implications for mapping 227 information presentation 292–9 introduction 281 land systems mapping 285–6 landform mapping principles 281–2 landscape-based mapping 282–6 modellers 300
Index
regolith-landform mapping 286–92 regolith-landform models 286, 300 soil mapping 284–5 user groups 299–300 see also mapping; regolith–landform mapping regolith geochemistry 73–101 associations in evaporative zones 94 associations with alunite supergroup minerals 95 associations with clay minerals 95 associations with Fe oxides 93–4 associations with phosphate minerals 95 associations with resistate minerals 95 associations with secondary Mn minerals 94 chemistry of weathering and element dispersion/ retention 76–91 element distributions and weathering related fractionation 91–5 fundamental controls on element behaviour 73–6 geochemical anomalies 95–101 physical processes affecting 91 regolith-related element associations 92–5 see also elements, regolith geochemistry of regolith geology correlation and landscape evolution 40–1 landscapes 39 role of biota in regolith development 41–2 weathering 39–40 regolith geophysics 219–42 developments and implications 241–2 geophysical survey planning and design 240–1 geophysical techniques 219–20 geophysical technologies, principles and applications 228–40 introduction 219 petrophysical properties 220–8 regolith materials biochemical uptake and redistribution by plants 182–5 characterising and identifying 121–2 petrophysical properties 220–8 redistribution by plants and animals 185–7 regolith minerology clay minerals 65–7 hyperspectral techniques 67–9 mineral groups 45
mineral weathering 64–5 other minerals 61–4 oxides and hydroxides 58–61 silicates, clay layer 50–8 silicates, rock-forming 46–50 regolith sampling for geochemical exploration see geochemical exploration, regolith sampling regolith–landform mapping 281, 286–300 compilation and interpolation 292 database 292–7 data collection 291–2 general concepts 286–7 ground geophysics 291 image data 290–1 importance of geomorphology 287–8 mapping methods 291–2 maps 294–7, 299–300 modellers 300 new directions 297–9 scale 288–90 standard terminology 293–4 user groups 299–300 regolith–landform mapping (RTMAP) 292–7 regolith–landform models 286, 300 regolith-related element associations 92–5 resistate minerals 64, 95 resistivity 229–33 resonance acoustic profiling (RAP) 221, 240 rhenium 445 rhizosphere 177, 178–81 rock types 3–4 discriminating parent 120–1 weathering profiles on common 109–15 rock-forming silicates 46–50 rocky and icy bodies 397–8 root growth 200–2 and redistribution of regolith materials 186 rubidium 434 ruthenium 445 saline soils 314–18 drainage and disturbance 318 groundwater-associated salinity (GAS) 315, 318 irrigation-associated salinity (IAS) 315–18 non-groundwater-associated salinity (NAS) 315, 318–20
459
460
Regolith Science
sample contamination 367–70 scandium 443 schwertmannite 310, 324, 330, 332 sediment accumulation in basins 10–11 seismic surveys 223–6, 230–1, 237–8 seismoelectric surveys 239–40 selenium 439–40 sepiolite–palygorskite 58 silica minerals 58 silica mobility 184–5 silicates clay layer 50–8 framework 46 layer 47–50 rock-forming 46–50 secondary 62–4 silicon 437 silcrete 23, 36–9, 61, 95, 107, 354, 358–9 ages and origins in Australia 21 silver 445 smectite 47–50, 53–5 sodicity, soil 320–2 sodium 434 soil–regolith processes 307–8 soil–regolith toposequence models 329–33 soils acid sulfate 322–8 acidity 321–2 colour 308–11 consistence 311 and geochemical exploration 342–5 groundwater-associated salinity (GAS) 315, 318 introduction 307 irrigation-associated salinity (IAS) 315–18 and landscape field indicators 328–9 mapping 284–5, 298 non-groundwater-associated salinity (NAS) 315, 318–20 processes 265–6 profile 108, 145, 173, 185, 208, 285, 293, 309, 318–20, 324–5, 353 saline 314–18 samples, making comparisons between 308–14 sampling procedures 342–5 segregations and course fragments 314 sodicity 320–1
structure 314 sub-aqueous 322 texture 311–14 solar system processes in terrestrial regolith 400–1 sorption reactions 271 species specific differentiation 183–4 stream sediments and geochemical exploration 363–4 as a sample medium 363–4 sampling procedures 364 stone lines 189 strontium 435 sub-audio magnetics (SAM) 229, 233–5 sub-aqueous soils 322 sulfates 61–3 sulfide minerals, oxidative breakdown 138–9 sulfide weathering 89–91 sulfur 439 sulfur biogeochemistry in acid sulfate soils 144–6 sulfur cycle 134–6 Surat Basin 2 surface lag 95, 105, 108–9, 352 surface nuclear magnetic resonance (SNMR) 240 surficial geology maps 286 surficial root mats 197–8 technetium 445 tellurium 440 terminology 6 terra rossa 115 tin 437 Titan 391–4, 399, 400 remote sensing data 392–3 surface data 393–3 titanium 443 trace element behaviour during weathering 82 trace element mobilisation 139–40 trace metal solubilisation 136–40 tree fall and redistribution of regolith materials 186–7 tungsten 447 ultramafic rocks, weathering 112–13 uranium 94, 116–18, 353, 365–6, 449 (U-Th/He) thermochronology 12–13
Index
vanadium 443 vegetation bands 206 biogeochemistry 355–6 dryland salinity 207–8 effects on regolith 206–7 erosion and weathering controls 202–4 geobotany 355 and geochemical exploration 355–7 hydraulic lift by 200 as a sample medium 356–7 sampling procedures 357 water table control 207–8 see also plants Venus 387–91, 399 surface imagery 391 surface processes 388–91 vermiculite 49–51, 55–7, 66–7, 177 vertical electric sounding (VES) 229 water cycle 251–6 and natural resource management 275–6 see also groundwater; regolith and water water table 3, 5, 21, 32, 39–40, 42, 59, 85, 109, 119–20, 145–6, 201, 253–5, 256–8, 266, 272, 276, 315, 317–21, 328, 331, 347–8, 353, 364, 367 table control 207–8 weathered bedrock 21, 225, 282, 284, 286, 295, 342, 345–7, 356 weathering biogenic 196–7 carbon-rich rocks 115 climate change, effects of long-term 118–19 degree of 122–4 in the geological cycle 3 granitoids 110 history 122–4 mafic rocks 110–12 major element variations 91–2 and microorganisms 175–7 mineral 64–5, 136–40 process 2–3 products, ferruginous 19 profiles, effects of erosion and deposition 115–16
profiles on common rock types 109–15 regimes 116–20 regimes, geochemical dispersion under different 119–20 regolith geology 39–40 related fractionation, element distributions and 91–5 sedimentary rocks 113–15 sulfide, associated element dispersion and 89–91 trace element behaviour 82 ultramafic rocks 112–13 and vegetation 202–4 weathering, element dispersion/retention and chemistry of 76–91 adsorption, precipitation and ionic exchange 87–8 basic processes 76–81 biogenic dispersion 86–7 biogenic processes and bacterial action 81–2 cation exchange capacity 88–9 co-precipitation 89 dissolution/precipitation reactions 79 Eh-pH 82–3 electrochemical dispersion 85–6 element dispersion 84–7 hydration-dehydration and transformation processes 80 hydrolysis reactions 79–80 hydromorphic dispersion 84–5 mechanical dispersion 87 metal complexing 83–4 reduction–oxidation (redox) reactions 80–1 sulfide weathering and associated element dispersion 89–91 trace element behaviour 82 Western Australian Shield 14, 19 Yilgarn Craton 10–11, 14, 22, 36, 63, 111–16, 118–21, 159–60, 162–3, 168–9, 222, 227, 284, 297–8, 344–5, 351–4, 356–7, 359–60, 362, 365 yttrium 445 zeolites 46 zinc 443 zirconium 445 zoogeomorphology 202–8
461