Tufas and Speleothems: Unravelling the Microbial and Physical Controls
The Geological Society of London Books Editorial Committee Chief Editor
BOB PANKHURST (UK) Society Books Editors
JOHN GREGORY (UK) JIM GRIFFITHS (UK) JOHN HOWE (UK) RICK LAW (USA) PHIL LEAT (UK) NICK ROBINS (UK) RANDELL STEPHENSON (UK) Society Books Advisors
MIKE BROWN (USA) ERIC BUFFETAUT (FRANCE ) JONATHAN CRAIG (ITALY ) RETO GIERE´ (GERMANY ) TOM MC CANN (GERMANY ) DOUG STEAD (CANADA ) MAARTEN DE WIT (SOUTH AFRICA )
Geological Society books refereeing procedures The Society makes every effort to ensure that the scientific and production quality of its books matches that of its journals. Since 1997, all book proposals have been refereed by specialist reviewers as well as by the Society’s Books Editorial Committee. If the referees identify weaknesses in the proposal, these must be addressed before the proposal is accepted. Once the book is accepted, the Society Book Editors ensure that the volume editors follow strict guidelines on refereeing and quality control. We insist that individual papers can only be accepted after satisfactory review by two independent referees. The questions on the review forms are similar to those for Journal of the Geological Society. The referees’ forms and comments must be available to the Society’s Book Editors on request. Although many of the books result from meetings, the editors are expected to commission papers that were not presented at the meeting to ensure that the book provides a balanced coverage of the subject. Being accepted for presentation at the meeting does not guarantee inclusion in the book. More information about submitting a proposal and producing a book for the Society can be found on its web site: www.geolsoc.org.uk.
It is recommended that reference to all or part of this book should be made in one of the following ways: PEDLEY , H. M. & ROGERSON , M. (eds) 2010. Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336. DITTRICH , M. & SIBLER , S. 2010. Calcium carbonate precipitation by cyanobacterial polysaccharides. In: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 51 –63.
GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 336
Tufas and Speleothems: Unravelling the Microbial and Physical Controls
EDITED BY
H. M. PEDLEY and M. ROGERSON University of Hull, UK
2010 Published by The Geological Society London
THE GEOLOGICAL SOCIETY The Geological Society of London (GSL) was founded in 1807. It is the oldest national geological society in the world and the largest in Europe. It was incorporated under Royal Charter in 1825 and is Registered Charity 210161. The Society is the UK national learned and professional society for geology with a worldwide Fellowship (FGS) of over 9000. The Society has the power to confer Chartered status on suitably qualified Fellows, and about 2000 of the Fellowship carry the title (CGeol). Chartered Geologists may also obtain the equivalent European title, European Geologist (EurGeol). One fifth of the Society’s fellowship resides outside the UK. To find out more about the Society, log on to www.geolsoc.org.uk. The Geological Society Publishing House (Bath, UK) produces the Society’s international journals and books, and acts as European distributor for selected publications of the American Association of Petroleum Geologists (AAPG), the Indonesian Petroleum Association (IPA), the Geological Society of America (GSA), the Society for Sedimentary Geology (SEPM) and the Geologists’ Association (GA). Joint marketing agreements ensure that GSL Fellows may purchase these societies’ publications at a discount. The Society’s online bookshop (accessible from www.geolsoc.org.uk) offers secure book purchasing with your credit or debit card. To find out about joining the Society and benefiting from substantial discounts on publications of GSL and other societies worldwide, consult www.geolsoc.org.uk, or contact the Fellowship Department at: The Geological Society, Burlington House, Piccadilly, London W1J 0BG: Tel. þ44 (0)20 7434 9944; Fax þ44 (0)20 7439 8975; E-mail:
[email protected]. For information about the Society’s meetings, consult Events on www.geolsoc.org.uk. To find out more about the Society’s Corporate Affiliates Scheme, write to
[email protected]. Published by The Geological Society from: The Geological Society Publishing House, Unit 7, Brassmill Enterprise Centre, Brassmill Lane, Bath BA1 3JN, UK (Orders: Tel. þ44 (0)1225 445046, Fax þ44 (0)1225 442836) Online bookshop: www.geolsoc.org.uk/bookshop The publishers make no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. # The Geological Society of London 2010. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of The Copyright Licensing Agency Ltd, Saffron House, 6 –10 Kirby Street, London EC1N 8TS, UK. Users registered with the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, USA: the item-fee code for this publication is 0305-8719/10/$15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 978-1-86239-301-1 Typeset by Techset Composition Ltd, Salisbury, UK Printed by MPG Books Ltd, Bodmin, UK Distributors North America For trade and institutional orders: The Geological Society, c/o AIDC, 82 Winter Sport Lane, Williston, VT 05495, USA Orders: Tel. þ1 800-972-9892 Fax þ1 802-864-7626 E-mail:
[email protected] For individual and corporate orders: AAPG Bookstore, PO Box 979, Tulsa, OK 74101-0979, USA Orders: Tel. þ1 918-584-2555 Fax þ1 918-560-2652 E-mail:
[email protected] Website: http://bookstore.aapg.org India Affiliated East-West Press Private Ltd, Marketing Division, G-1/16 Ansari Road, Darya Ganj, New Delhi 110 002, India Orders: Tel. þ91 11 2327-9113/2326-4180 Fax þ91 11 2326-0538 E-mail:
[email protected]
Contents PEDLEY , H. M. & ROGERSON , M. Introduction to tufas and speleothems
1
JONES , B. Microbes in caves: agents of calcite corrosion and precipitation
7
GONZA´ LEZ -MUN˜ OZ , M. T., RODRIGUEZ -NAVARRO , C., MARTI´ NEZ -RUIZ , F., ARIAS , J. M., MERROUN , M. L. & RODRIGUEZ -GALLEGO , M. Bacterial biomineralization: new insights from Myxococcus-induced mineral precipitation
31
DITTRICH , M. & SIBLER , S. Calcium carbonate precipitation by cyanobacterial polysaccharides
51
ROGERSON , M., PEDLEY , H. M. & MIDDLETON , R. Microbial influence on macroenvironment chemical conditions in alkaline (tufa) streams: perspectives from in vitro experiments
65
ARP , G., BISSETT , A., BRINKMANN , N., COUSIN , S., DE BEER , D., FRIEDL , T., MOHR , K. I., NEU , T. R., REIMER , A., SHIRAISHI , F., STACKEBRANDT , E. & ZIPPEL , B. Tufa-forming biofilms of German karstwater streams: microorganisms, exopolymers, hydrochemistry and calcification
83
ARENAS , C., OSA´ CAR , C., SANCHO , C., VA´ ZQUEZ -URBEZ , M., AUQUE´ , L. & PARDO , G. Seasonal record from recent fluvial tufa deposits (Monasterio de Piedra, NE Spain): sedimentological and stable isotope data
119
GRADZIN´ SKI , M. Factors controlling growth of modern tufa: results of a field experiment
143
PEDLEY , H. M. & ROGERSON , M. In vitro investigations of the impact of different temperature and flow velocity conditions on tufa microfabric
193
BENZERARA , K., MEIBOM , A., GAUTIER , Q., KAZ´ MIERCZAK , J., STOLARSKI , J., MENGUY , N. & BROWN , JR , G. E. Nanotextures of aragonite in stromatolites from the quasi-marine Satonda crater lake, Indonesia
211
BINDSCHEDLER , S., MILLIE` RE , L., CAILLEAU , G., JOB , D. & VERRECCHIA , E. P. Calcitic nanofibres in soils and caves: a putative fungal contribution to carbonatogenesis
225
PENTECOST , A. The fractionation of phosphorus in some modern and late-Holocene calcareous tufas in North Yorkshire, UK
239
¨ ZKUL , M., GO¨ KGO¨ Z , A. & HORVATINCˇ IC´ , N. Depositional properties and geochemistry of O Holocene perched springline tufa deposits and associated spring waters: a case study from the Denizli Province, Western Turkey
245
CAPEZZUOLI , E., GANDIN , A. & SANDRELLI , F. Calcareous tufa as indicators of climatic variability: a case study from southern Tuscany (Italy)
263
BALDINI , J. U. L. Cave atmosphere controls on stalagmite growth rate and palaeoclimate records
283
FAIRCHILD , I. J., SPO¨ TL , C., FRISIA , S., BORSATO , A., SUSINI , J., WYNN , P. M., CAUZID , J. & EIMF. Petrology and geochemistry of annually laminated stalagmites from an Alpine cave (Obir, Austria): seasonal cave physiology
295
vi
CONTENTS
MATTEY , D. P., FAIRCHILD , I. J., ATKINSON , T. C., LATIN , J.-P., AINSWORTH , M. & DURELL , R. Seasonal microclimate control of calcite fabrics, stable isotopes and trace elements in modern speleothem from St Michaels Cave, Gibraltar
323
HAMMER , Ø., DYSTHE , D. K. & JAMTVEIT , B. Travertine terracing: patterns and mechanisms
345
Index
357
Introduction to tufas and speleothems H. MARTYN PEDLEY* & MIKE ROGERSON Department of Geography, University of Hull, Cottingham Road, Hull, HU6 7RX, UK *Corresponding author (e-mail:
[email protected])
Ambient temperature freshwater carbonates precipitate as surface deposits within karstic stream, lake and swamp environments (tufas) and in subterranean situations (speleothems), where they line vadose caves and fracture systems. Although physico-chemical mineral precipitation contributes significantly to both kinds of deposit, there is a clear spatial association between the development of tufa deposits and the occurrence of microbial biofilms. This fact, and the recent discovery that the occurrence of certain heterotrophic bacteria promote precipitation onto the surface of stalactites (Cacchio et al. 2004), strongly implicates a degree of microbial influence in the calcite precipitation process, regardless of the environmental context. To add to the inherent complexity of these systems, there is considerable interplay between biological and physical processes to consider. Water velocity and turbulence will strongly affect biofilm colonization and may damage the community, thereby affecting carbonate precipitation rates, in addition to regulating important kinetic limitations on precipitation via modifications of the calcium ion delivery rate. Exchange of CO2 gas at the air– water interface is an important conditioner for precipitation in vadose systems but will also occur within surface systems as a consequence of photosynthesis. It is only by considering karst hydrological systems holistically that these processes can be untangled. Tufas and speleothems share the same soilderived meteoric water supply, represent zones of deposition of calcium ions chemically eroded from the same geological sources and produce laminated deposits which are superficially similar. In passing from cave environments via resurgences (Fig. 1) into surface waterways, individual packages of water pass down an interconnected hydrological system at any point in which the conditions necessary for calcite precipitation may be achieved. Within the deposits that this precipitation creates, it is apparent that there is a progressive gradation from massive, laminated speleothems fabrics into stromatolitic, biofilm dominated tufas fabrics. In fact, speleothems and tufa represent two end members within a continuum of freshwater carbonate reflecting different balances of physicochemically and biologically controlled precipitation.
On a regional scale the occurrences of tufas and speleothems are both controlled by water table fluctuations. Typically, tufa deposition is associated with predominantly high water tables and although tufas enjoy global distributions from the tropics to polar regions, they are most effective as bioconstructors where spring fed streams are not subjected to spate conditions. Similarly, tufa developments are severely impaired by fluctuating water tables associated with increasingly arid climatic cycles. Limitation on surface carbonate precipitation is consequently derived from the necessity for biofilm development combined with the equal necessity of adequate supply of dissolved calcium and carbon, which must be present at least in part as carbonate. The latter requirement of sufficient Ca2þ(aq) and CO322(aq) ionic activity demands that these ions are not lost from solution before resurgence, making it likely that tufas will develop best where caves are flooded, thereby minimizing the distribution of the subterranean vadose environment where speleothems develop most abundantly. Curiously, these elevated tables are frequently encouraged by the tufa growth itself as a consequence of the valley bottom ponding and back flooding caused by barrage development. Conversely, as lower water tables become established and the subterranean vadose environment becomes more important, speleothems will become established. Conceptually, the occurrence of abundant tufa or speleothem deposition simply reflects the position of a hydrochemical ‘knick-point’, which occurs when sufficient CO2 has been lost from solution for carbonate ions to become abundant, for example when the thermodynamic gradient promoting precipitation (Gibbs Free Energy) exceeds the barrier presented by the activation energy. This knick-point may occur either above or below ground depending on the height of the water table. As part of the same hydrochemical system, tufas and speleothems offer an inseparable duo when exploring the climatic archive, and will reflect the same processes within the catchment. Much palaeo-environmental information in tufas and speleothems can be extracted from geochemical time series created from these deposits. However, one of the greatest obstacles to collective use of these materials in ‘climate’ reconstruction is the
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 1–5. DOI: 10.1144/SP336.1 0305-8719/10/$15.00 # The Geological Society of London 2010.
2 M. PEDLEY & M. ROGERSON Fig. 1. A schematic diagram of terrestrial carbonate precipitation occurring within an interlinked system. This continuum exhibits dominantly physico-chemical processes at one extreme (speleothems, on the left of diagram, and travertines, on the right) and dominantly biological processes at the other (oncoid and lake carbonates). Tufa systems typically fall within the middle of this continuum, exhibiting both physical and microbial control. When calcite saturation no longer allows tufa framework development, oncoids typically are developed (top right, showing cross section of oncoid with stromatolitic structure as inset). All these deposits properly should be considered as part of the same depositional system within karst environments, with individual occurrences related to the position of the hydrochemical ‘knick-point’, that is, when precipitation activation energy is breached by local biological influences.
INTRODUCTION TO TUFAS AND SPELEOTHEMS
oversimplified understanding of the actual freshwater carbonate precipitation process and how it is controlled by the ever-changing environment. The traditional view that tufas and speleothems are formed entirely by physico-chemical precipitation processes provides the foundation for this research, but is unlikely to be sustainable indefinitely in the face of increasing evidence for biological influence on the precipitation process. This newly perceived consideration that both chemical and microbiological processes are involved in carbonate precipitation (in both marine systems as well as the freshwater environments considered in this volume) has resulted in the development of a new interface discipline which bridges both the physical and biological sciences. Its remit is to view and interpret the planet, in space and time, as an integrated system with life as an essential driver within all terrestrial and oceanic environmental systems. This interface discipline has variously been referred to as ‘Biogeology’, ‘Biogeoscience’ and ‘Geomicrobiology’ though it is most frequently referred to as ‘Geobiology’ (Noffke 2005), which is our preference. One of the paramount requirements for future freshwater carbonate research in this context is to build a new foundation for understanding carbonate precipitation processes based on a holistic approach deriving data from all relevant physical, chemical and biological processes acting within a system. Microbial processes fundamentally underpin many large scale carbonate depositional systems. Consequently, the present need within geosciences is for detailed assessments of the role of living microbial communities in modern environments and of how they act to modify their hydrochemical, geochemical and biochemical environments. Much current research focuses on exploring the role of microbial communities in present day sediment– water interface processes. Many of these processes are associated with carbonate precipitation, thereby providing a direct link into accumulation styles and deposition rates of ancient sediments. Equally importantly these microbial processes strongly influence diagenesis within sedimentary successions. Consequently, the new discipline of Geobiology has the potential to revolutionize the way we view geological processes at the local and regional scales (past, present and future).
Contributions in this volume This volume was conceived at a ‘Tufas and Speleothems’ workshop organized around ‘research in progress’ which was held in the Geography Department, University of Hull, UK in May 2008. The meeting was well attended by researchers from
3
France, Germany, Italy, Poland, Slovenia, Spain, Switzerland and the UK. It also attracted specialists from Israel, Turkey and Canada. The remit of the meeting was to bring tufa and speleothems researchers together, and especially sedimentologists and geochemists, in order to demonstrate that these increasingly separate research areas were essentially dealing with a single carbonate precipitate continuum. A one day fieldtrip to the Caerwys fluvial barrage tufas of North Wales provided further stimulus to informal discussion and gave an excellent opportunity for liaison between tufa and speleothems specialists. The articles herein reflect the work of 63 specialists (principally bacteriologists microbiologists, hydro- and geochemists and sedimentologists) based mainly in academic institutions. The remit of this volume is to develop a better understanding of the biological and chemical influences on carbonate precipitation associated principally with ambient temperature freshwater carbonates. Many processes are common to both surface and subterranean situations yet a research dichotomy has developed over the past 30 years. Tufa researchers have tended in the past to be carbonate sedimentologists or biostratigraphers interested in using facies models in order to characterise the flow dynamics and proxy-datable stratigraphy. In contrast, speleothems research has primarily been the domain of the geochemist. There is now a need to draw the two research themes together. The following thirteen chapters of this volume, dealing principally with carbonate precipitate associated microbial process and product, range from studies of individual reactions to overviews of entire biofilm communities involved with calcium carbonate precipitation. These are followed by four papers devoted to the hydrochemistry of cave waters, the geochemistry of speleothems and finally, the physical control on depositional morphology using travertine as a case study. The first paper, by Jones, takes on the thorny question of what effect microbial processes have on cave carbonates. The article is a comprehensive review of previous work to which is added a considerable body of work carried out in the Cayman Islands. The conclusion is that microbial processes are widespread both above and below ground. The article reaffirms the need to reconsider microbial effects when interpreting geochemical data derived from speleothems. Gonza´lez-Mun˜oz et al. show how a single bacterial genus induces a wide range of mineral precipitates. They conclude by suggesting that the mechanisms that control bacterial biomineralization are universal and are determined principally by the environment in which the bacteria live. Turning now to the specific situation of biomediated carbonate, Dittrich &
4
M. PEDLEY & M. ROGERSON
Sibler, present an experiment-based study of EPS associated with three strains of picocyanobacteria. The authors conclude that cyanobacterial extracellular polysaccharides have a strong potential to exchange protons with the environment. Additionally, the strains studied were capable of calcium carbonate precipitation. Rogerson et al. showed that the precipitation of calcium carbonate occurs preferentially under conditions of rising pH at the night– day transition by working with a flowing mesocosm laboratory experiment. The observations indicate that daylight length does not control overall precipitation rate and confirms suggestions that seasonal laminations require either strong variability in ambient physico-chemical activity or else an ecological/compositional change in the microbial community. However, Arp et al. present a very comprehensive study of natural stream-grown stromatolitic tufa together with details of the biofilm community which include cyanobacteria, a nonphototrophic prokaryote community and diatoms. This study concludes that annual laminae were the result of seasonal changes in temperature and light. Furthermore, the stable carbon isotope composition of the laminae was not affected by photosynthesis-induced microgradients but mirrored bulk water chemistry. Arenas et al. monitored physical, hydrochemical parameters and sedimentation rates over a 3.5-year period in an active stream. They demonstrate that sedimentation rates have a strong seasonal pattern with higher values in warm periods. Conversely, the sediment d18O composition shows a rhythmic variation, with higher values in cool periods. This was probably caused by fractionation due to seasonal temperature variations. The calculated temperatures, for a theoretical equilibrium precipitation, accord with the measured temperatures and display the same seasonal trend. In conclusion they suggest that: (a) tufa d18O values can be used to estimate relative palaeotemperature variations in fluvial carbonate deposits; and (b) laminated tufa sediments can be useful as high-resolution records of temperature change. Gradzin´ski monitored four active tufa field sites during a 14-month period and recorded the highest tufa growth rates in the fastest flow conditions. These precipitates are dominated by a crystalline texture or a highly encrusting mode. In contrast the areas of slow tufa growth are dominated by micritic textures. These observations are confirmed by Pedley & Rogerson from independent in vitro laboratory experiments. Furthermore, this study also demonstrates the physical structure of the freshwater biofilm. SEM studies reveal that heterotrophic bacteria are dominant in zones within the EPS where calcite precipitation occurs. The intimate relationships between nanofabrics and calcite precipitates are well demonstrated in this study. Benzerara
et al. provide an illuminating study, aided by microscopy and microspectroscopy, of the extent to which biological processes control stromatolite structure, and of the respective contributions of biological and abiotic processes in their formation. The chemical composition of the calcium carbonate precipitates is outlined and their developed within the stromatolite is shown. Some carbonate fabrics, however, continue to remain enigmatic and none more so than pedogenic calcitic nanofibres. Bindschedler et al. conclude that they are probably the product of the organic decay of fungal hyphae into cell wall microfibrils. These could then be replaced by calcite pseudomorphosis, perhaps also acting as templates for direct calcite precipitation. The article emphasises the importance of organic matter and fungi in carbonatogenesis generally. Gonza´lez-Mun˜oz et al. (earlier in the volume) indicate how bacteria are capable of precipitating a range of chemistries and Pentecost illustrates this with respect to the development of up to 19% phosphorous in the carbonate fraction of some active tufa sites. The percentage increase in some fossil examples to over 48% is attributed to further miner¨ zkul et al. alization of contained organic matter. O provide an exceptionally detailed account of a partly active perched springline tufa. The spring water hydrochemistry and trace element and stable isotopic composition of the tufa precipitates are well documented and provide a useful comparison between the composition of aqueous supply and tufa product. Radiocarbon dating of the older deposits demonstrates them to have developed between 2000–5800 yr BP. The final article within this section is by Capezzuoli et al. Here, the regional study of a Quaternary succession containing tufas demonstrates convincingly that tufa developments appear to be controlled by water availability rather than by ambient temperature. More precisely they consider that climate controls water table levels and indirectly the volume of water available to the system. Four depositional carbonate (tufa), events each separated by active down cutting events, are recognized in the Valdelsa succession. These depositional carbonate events are radiocarbon dated and correlate with the Last Glacial Interstadial through the Younger Dryas to the Atlantic ‘Optimum Climatic’, and Sub-Boreal episodes respectively. It would appear therefore that tufas are clear indicators of high palaeo-water tables. In developing the subterranean theme introduced by the opening article by Jones, Baldini explores the effects of soil derived atmospheric pCO2 within caves. He concludes that soil temperature is a major control on cave air pCO2 and that variations in pCO2 can trigger either precipitation of dissolution, thereby controlling stalagmite growth rates generally and holds the potential to skew the
INTRODUCTION TO TUFAS AND SPELEOTHEMS
seasonal geochemical proxy signals within them in favour of certain seasons. Fairchild et al. explore the seasonality which is encoded in speleothems. All samples show visible autumnal laminations which are associated with increases in trace element concentrations. Within these laminae synchrotron studies have resolved mm-scale Pb and Zn enrichments. The study is an excellent demonstration of how chemical variations within laminae express seasonal physiology in temperate caves. Mattey et al. monitored the water chemistry at three speleothems drip sites and demonstrate a strong link between local microclimate and the proxy–record. The data reveal a strong seasonal drip-water chemistry pattern with the calcite saturation being linked closely to regular seasonal variations in cave air pCO2. The relationships between stable isotope rations, Sr/Ca and speleothem laminae are consistent with a degassing – calcite precipitation process. In these modern speleothems rapid degassing controls the d13C of drip water and calcite whereas the slower rate of calcite precipitation caused seasonal Sr cyclicity. Nevertheless, caution is advised when linking paired speleothem fabrics to specific seasons because of the local processes operative within caves. Clearly, there is a dominantly physico-chemical control on the geochemical signals within speleothems but this need not be exclusive to subterranean sites. In fact, on a widespread macro-scale Hammer et al. presents a thoughtful study of the dominantly physico-chemical controls governing travertine terracing patterns. Such terracing occurs at a range of incremental scales though this study pays particular attention to micro-terrace development. The terrace morphologies are the product of interactions between any combination of water chemistry, precipitation kinetics, topography, hydrodynamics, carbon dioxide degassing, erosion, sedimentation and even biology. However, the
5
article provides some pointers to the likely process combinations controlling each development. Collectively, these research articles illustrate a wide range of modern calcium carbonate precipitation processes and related carbonate precipitates. All demonstrate that small scale variations in water chemistry correlate directly to changes in environmental conditions. Many of the articles confirm that such variations are faithfully recorded in the precipitated carbonate fabrics as proxypalaeoenvironmental indicators. Preservation quality appears excellent for the Holocene but has yet to be confirmed for older Quaternary and Neogene deposits. Research continues to seek new approaches (e.g. synchrotron analysis) which will reveal fine scale variations within the carbonate precipitation processes though much remains to be done. Nevertheless, it is already clear that depositional karst offers an unrivalled opportunity to fine-tune our understanding of Quaternary environmental change at the global scale. Speleothems provide the absolute chronometer of environmental change and monitor of hydrological regime within the aquifer whereas tufas provide a detailed view of karst surface processes and their influence on environment-driven ecological change. The next step in this Geobiology research domain is to resolve the dynamic choreography between these subaerial and subterranean karst players.
References C ACCHIO , P., C ONTENTO , R., E RCOLE , C., C APPUCCIO , G., M ARTINEZ , M. P. & L EPIDI , A. 2004. Involvement of microorganisms in the formation of carbonate speleothems in the Cervo Cave (L’Aquila-Italy). Geomicrobiology Journal, 21(8), 497–509. N OFFKE , T. 2005. Geobiology – a holistic scientific discipline. Palaeogeography, Palaeoclimatology, Palaeoecology, 219, 1– 3.
Microbes in caves: agents of calcite corrosion and precipitation BRIAN JONES Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Alberta, T6G 2E3, Canada (e-mail:
[email protected]) Abstract: Diverse biogenic and abiogenic processes produce calcite speleothems. From a biogenic perspective, cave microbes mediate a wide range of destructive and constructive processes that collectively influence the growth of calcite speleothems and their internal fabrics. Destructive processes include substrate breakdown by dissolution, boring and residue micrite production, whereas constructive processes include microbe calcification, trapping and binding of detrital particles to substrates, and microbial induced calcite precipitation. Biogenesis can be established from: (1) the presence of mineralized microbes; (2) fabrics, such as stromatolite-like structures, that can be attributed to microbial activity; and/or (3) geochemical proxies (carbon and oxygen isotopes, lipid biomarkers) considered indicative of microbe activity. Such criteria have, for example, been used to demonstrate microbial involvement in the formation of pool fingers, stalactites/stalagmites, cave pisoliths and moonmilk. Nevertheless, absolute proof of microbial biogenesis in calcitic speleothems is commonly difficult because taphonomic processes and/or diagenetic processes commonly mask evidence of microbial activity. The assumption that calcitic speleothems are abiogenic, which has been tacitly assumed in many studies, is dangerous as there is clear evidence that microbes thrive in most caves and can directly and indirectly influence calcite precipitation in many different ways.
With their wondrous and complex morphologies that develop in dark cave environs, speleothems have long been the subject of scientific attention and curiosity. The calcite crystals that form these speleothems have been the focus of many studies with emphasis commonly being placed on their crystallography and the factors that controlled their growth (White 1976; Kendall & Broughton 1978; Broughton 1983a, b, c). Such studies were commonly based on the tacit assumption that the calcite must be of abiogenic origin because it grew in the dark environs of caves. Such an assumption is false because microbes, including bacteria, fungi, actinomycetes, and Archaea, thrive in most caves despite the lack of light (e.g. Høeg 1946; Claus 1955; Mason-Williams 1967) – a fact that has been emphasized in many reviews (Barton et al. 2001; Northup & Lavoie 2001; Barton 2006; Barton & Jurado 2007; Barton & Northup 2007; Mulec 2008; Taborosˇi 2008). Given the presence of microbes in caves throughout the world, the role that they play in the growth and development of speleothems has become a topic of considerable interest. Caves can be divided into two parts: (1) the twilight zone, which is the transition between lightless cave interior and the outside world; and (2) the aphotic zone where there is no light. Cave walls in the twilight zone are commonly covered with a green, mucilaginous biofilm that is formed by microbes that thrive in the light that penetrates through the cave entrance. Although less obvious
in the aphotic zone, biofilms and their formative microbes are present on cave walls and the speleothems that adorn the cave. Calcite precipitation in caves has commonly been directly or indirectly linked to these microbial communities (e.g. Danielli & Edington 1983; Cox et al. 1989a, b; Polyak & Cokendolpher 1992; Cox et al. 1995; Cox et al. 2001; Melim et al. 2001). Potentially, microbes can mediate destructive (‘erosion’ of Cunningham et al. 1995) and constructive (‘microbe-assisted depositional forms’ of Cunningham et al. 1995) processes (Jones 2001). Although operational on a micro-scale, these processes have the potential of significantly impacting the growth and development of speleothems. Assessment of the role of microbes in caves lies at the interface between microbiology and geology: the former deals largely with determining the taxonomy of the microbes and their life modes, whereas the latter deals mostly with the destructive and construction geological processes mediated by the microbes. This paper approaches the problem from a geological perspective with emphasis being placed on the roles that microbes play in the growth, destruction, and development of calcitic speleothems. It is based on a comprehensive assessment of the literature and information obtained from speleothems that adorn various caves on the Cayman Islands (Fig. 1). Collectively, this information provides a critical assessment of the influences that microbes have on speleothem development and also highlights areas where our knowledge is in its infancy. Although considerable
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 7–30. DOI: 10.1144/SP336.2 0305-8719/10/$15.00 # The Geological Society of London 2010.
8 B. JONES Fig. 1. Location maps. (a) Location of Grand Cayman and Cayman Brac in Caribbean Sea. (b) Location of modern cave at Old Man Village and old, filled cave near Breakers. (c) Locations of Hospital and Bats caves (modern) on Cayman Brac.
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION
advances have been made over the last ten years, many problems and issues await resolution. It is a fertile ground for further research.
Microbes in caves Biotic composition and diversity Barton & Northup (2007) suggested that any discussion on microbes in caves should be split into two eras that they centered on the ‘Breakthroughs in Karst Geomicrobiology and Redox Geochemistry’ conference held in 1994 (Sasowsky & Palmer 1994). That date coincides roughly with the advent of new microbiological techniques, such as 16S rDNA gene sequencing, which changed the manner in which microbes were characterized and classified and showed that cave biotas were far more diverse than previously suspected. The presence of microbes in caves has long been recognized. Claus (1955), for example, noted that Scopoli (1772), Humboldt (1793) and Hoffman (1811) referred to fungi in caves, but unfortunately failed to include these citations in his reference list. Irrespective, many studies between 1900 and 1994 reported diverse microbial populations from caves throughout the world, including the following: † bacteria, fungal hyphae and cyanohycea from caves in Norway (Høeg 1946); † various species of bacteria from caves in south Wales (Mason-Williams 1959); † 69 species, varieties and forms of algae from Baradla cave in north Hungary (Claus 1955); † 93 species (including 58 cyanophytes) from caves in Baranya district; 90 species from Be´ke Cave, 41 species from Pa´lvo˜lgy Cave, and 13 species from Ko˜lyuk Cave, all being located in Hungary (Palik 1960); † bacteria, actinomycetes, and fungi from caves in France (Caumartin 1963); † 7 species of algae from Crystal Cave, Kentucky (Nagy 1965); † 27 taxa from Mammoth Cave, Kentucky (Jones 1965); and † 30 species of algae and 45 species of bacteria from caves in South Wales (Mason-Williams 1967). These examples amply demonstrate the diversity of microbes found in many different caves. Such diversity was emphasized by Draganov (1977) who compiled a list of 627 taxa of cave algae that had been recorded between the end of the 19th century and the time when he wrote his paper. This diverse biota included representatives of the Cyanophyta, Rhodophyta, Chlorophyta, Pyrrophyta and Euglenophyta.
9
Virtually all of the microbes listed in these papers were identified on the basis of culturing coupled with optical and electron microscopy. Culturing of microbes, however, commonly provides a very biased view of the total population. Amann et al. (1995) and Amann et al. (1996), for example, suggested that less than 1% of microbes in any given population are culturable by standard methods. Application of that ratio to the 627 taxa of microbes listed by Draganov (1977) implies that the microbial biotas found in caves may comprise at least 62 700 taxa! The difficulty of recognition using these methods is compounded by the fact that microbes are not that diverse in terms of their morphology as there are limits to the degree of morphological variance that can be expected among the morphologically-simple filamentous, coccoid, or bacilliform architectures of most microbes (e.g. Barton 2006). Over the last twenty years or so, methods for the identification and characterization of microbes have changed significantly. As a result, microbes are now typically identified by their chemotaxonomic characters, 16S rRNA gene sequencing, and polyphasic taxonomy. Such methods allow microbe identification without cultivation and places emphasis on the DNA of the microbes rather than their morphology and/or culture conditions. Application of these methods to microbial biotas collected from caves has revealed diverse, complex biotas, as is shown by the following examples: † 200 strains of actinomycetes associated with cave paintings in various crypts, caves, and tombs (Giacobini et al. 1987); † 350 taxa of actinomycetes identified from Altamira Cave, Spain (Groth & Saiz-Jimenez 1999); † Streptomyces (8 strains), Bacillus (4 strains), Nocardiopsis (3 strains) and various other taxa found in Grotto deo Cervi, Italy (Laiz et al. 2000); † 38 phylotypes related to Ortobacteria, Actinobacteria, and Cytophagales in Glenwood Caverns, Colorado (Barton et al. 2004); and † 12 bacterial divisions and subdivisions comprising 49 phylotypes identified in samples collected from Wind Cave in South Dakota, U.S.A. (Chelius & Moore 2004). These examples, based on chemotaxonomic techniques, further emphasize the tremendous diversity of microbes that can thrive in caves. Most new taxa of cave microbes are now defined solely on the basis of their chemotaxonomic attributes. Lee (2006), for example, defined a new species of actinomycetes from a cave on Jeju Island, Korea on the basis of its 16S rRNA signature but he described it as being ‘. . . a well-developed and branched substrate mycelium that fragments into rod-shaped
10
B. JONES
Fig. 2. SEM photomicrographs of substrates from the twilight zone of the Old Man Village cave (Grand Cayman). Figure 2a– f show exposed surfaces (i.e. wall of cave) whereas Figure 2g–l show features beneath exposed surfaces. (a) General view showing mucus (m) and bacteriform microbes on substrate. (b) Filamentous and bacteriform microbes covered calcite substrate. (c) Non-calcified microbes. (d) Bacteriform microbes associated with etched calcite substrate. (e) Filamentous microbe on surface formed of small, irregular-shaped calcite particles. (f ) Mucus, filaments, and calcite particles covering surface and openings to borings. (g) General view showing boundary between biofilm
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION
elements . . .’, failed to specify its size, and did not provide an illustration. Thus, the morphology of cave microbes, at least from a microbiological perspective, is now deemed of secondary importance in their description, characterization, and classification.
Migration and colonization As the microbial biotas of caves become better known and characterized it is also becoming apparent that similar taxa are present in caves that are geographically distant and isolated from each other. In other words, as asked by Jones (1965), ‘How did the algae get into the speleo-environment . . .’. As yet, the mechanism by which the same or related microbes colonize different, geographically disparate caves has not been resolved. Microbes are, by definition, very small (commonly ,1 mm long) and hence easy to transport. Thus, any assessment of the role that microbes may play in the formation of speleothems must first ascertain if the microbes are residents that thrive in the cave or simply ‘detritus’ that were transported into the cave. Claus (1955), for example, suggested that microbes may be carried into caves by streams, air current, animals (e.g. bats), and/or water that gradually seeps into the cave. Similarly, caution must also be exerted against contamination of samples during sampling and/or microbes that migrated into a sample, via cracks; long after the substrate was originally precipitated (Barton et al. 2001). Opening caves for public viewing of the prehistoric art work, for example, has commonly led to significant problems because this has disturbed atmospheric equilibrium in the caves as humidity, CO2 content of the air, and air temperature were modified and bacteria, fungi and algae were introduced (e.g. Lefevre & Laporte 1969). In many cases, microbes have been implicated in the deterioration of historically significant cave paintings (e.g. Giacobini et al. 1987; Monte & Ferrari 1993; Groth et al. 2001).
Microbial destruction of substrates Cave walls are commonly covered with biofilms (Fig. 2) that mediate substrate destruction through boring and dissolution (Can˜averas et al. 2001; Jones 2001). Such processes release Ca and CO3 back into the system, cause breakdown of the host substrate, and may generate residue micrite that can be transported to other parts of the cave (Jones & Kahle 1995).
11
Boring Boring microbes commonly infest and destroy calcite crystals that line near-surface cavities in the karst terrains of the Cayman Islands (Jones 1987). In contrast, Jones (1995) did not find any evidence of endolithic microbes in substrates collected from the twilight zones of various caves on the Cayman Islands. New samples collected from the twilight zone of the Old Man Village cave, however, did reveal the presence of boring microbes (Fig. 2f–l). That borings were found in these samples but not in previous samples probably reflects issues of sampling for features that are so variable even on a microscale. Borings in the calcitic substrates from the twilight zone of the Old Man Village cave, which are up to 10 mm in diameter, start beneath the biofilm (Fig. 2f) and penetrate to depths of 1 mm (Fig. 2g– j). Some borings contain calcified masses that may be preserved remnants of the formative microbe (Fig. 2j). Boring appears to have been mediated by etching as the walls of the borings are characterized by highly irregular microtopographies that display features indicative of dissolution (Fig. 2k, l). The lack of morphological features precludes definitive identification of the microbes that formed the borings.
Dissolution Dissolution of calcitic substrates, mediated by microbial biofilms, commonly produces, spiky calcite (Fig. 3a –c), irregular etching (Fig. 3d), blocky calcite (Fig. 3e, f ), and/or residual calcite (Fig. 3h, i). Such features are evident in natural samples (Jones 1987, 1989, 1995; Can˜averas et al. 2001) and have also been produced experimentally by allowing fungi to grow on crystals of Iceland Spar (Jones & Pemberton 1987a, b). Dissolution of this type, which commonly follows the crystallographic precepts of the substrate, is typically slow and probably reflects surface-reaction-controlled kinetics (cf. Berner 1978). Such etching is mediated by the biofilm and does not involve vast quantities of water. Residual calcite (¼ sparmicrite of Kahle 1977) is a common byproduct of substrate breakdown caused by boring and dissolution (Fig. 3h, i). Irregular etching of a substrate, for example, produces monticuli (small islands still connected to substrate by narrow neck) that will eventually break away from the substrate and form individual grains (Jones & Pemberton 1987b; Jones & Kahle 1995). Water
Fig. 2. (Continued) and calcite crystal (arrow). Note borings (below arrow) penetrating calcite crystal. (h, i) Borings in calcite crystal just below surface biofilm. (j) Enlarged view of borings in calcite crystal. (k) Borings extending from outer surface (upper part of image) into calcite crystals. Note etched surfaces on the borings. (l) Etched surface of boring.
12
B. JONES
Fig. 3. SEM photomicrographs showing etch features on substrates in twilight zone of caves. Figure 3a, b, c, g from Old Man Village Cave, Grand Cayman; Figure 3d from Bats Cave, Cayman Brac; Figure 3e, f, h, i from Hospital Cave, Cayman Brac. (a) Spikey calcite partly covered by mat formed of spores and filaments. (b, c) Spiky calcite beneath biofilm formed of mucus and rod-shaped microbes. (d) Irregularly etched surface covered by rod-shaped microbes. (e, f) Blocky calcite partly covered by microbial mat. (g) Biofilm on partly etched calcite crystal. (h, i) Residue micrite formed by microbial breakdown of calcitic substrate.
flowing over such substrates may transport these grains elsewhere.
Microbial construction of calcitic substrates Constructive processes include microbe calcification, trapping and binding by filamentous microbes, and/or mineral precipitation (Can˜averas et al. 2001; Jones 2001).
Microbe calcification When microbes are replaced and/or encrusted by calcite, they become part of the substrate (e.g. Jones & Motyka 1987; Arin˜o et al. 1997; Gradzinski 1999; Can˜averas et al. 2001; Gradzinski 2001; Baskar et al. 2007; Baskar et al. 2009). The type
of calcite crystals involved in such calcification is variable. Geitleria calcarea, a cyanobacterium that thrives under low light conditions (Friedmann 1979; Davis & Rands 1981; Abdelahad & Bazzichellia 1988) has been found in the twilight zone of caves in Israel, Romania, Yugoslavia, Florida, Cook Islands (Friedmann 1979), France (Bourrelly & Dupuy 1973; Coute´ 1982; Leclerq et al. 1983), Spain (Gracia Alonso 1974), Italy (Abdelahad & Bazzichellia 1988), the Bahamas (Davis & Rands 1981), and the Cayman Islands (Jones & Kahle 1986, 1993). Davis and Rands (1982) also reported Hapalosiphon intricatus with calcified sheaths from the humid, poorly illuminated parts of Orange Lake cave in Florida. Calcified Geitleria found on the walls of cavities in a breccia that fills a sinkhole on the southeast coast of Grand Cayman are encrusted with complex three-dimensional calcite dendrite crystals
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION
13
Fig. 4. SEM photomicrographs of Geitleria calcarea from walls of a poorly illuminated cavity in a limestone breccia that fills a sinkhole in the Bluff Formation, southeast Grand Cayman. (a) Individual specimen resting on biofilm that is formed of mucus and filamentous microbes. (b) Longitudinal section showing hollow core and wall of the filamentous microbes encrusted with dendritic calcite crystals. (c) Open lumen with encrusting dendritic calcite crystals. (d) Encrusting dendritic calcite crystals encasing filament. Black background introduced, using Photoshop #, in order to highlight encrusting crystals. (e) Exterior of microbe showing dendritic calcite crystals. (f ) Enlarged view of dendritic calcite crystals.
(Fig. 4). Being from the interior, poorly illuminated areas of the cavities, they are equivalent to forms found in the twilight zones of caves. The dendrite crystals, characterized by primary, secondary, and tertiary levels of branching form intricate crystal networks that completely envelope the microbes (Fig. 4e, f ). Other calcified microbes are
characterized by different crystal architectures. Some of the filamentous microbes on the cave walls in the twilight zone of the Old Man Village cave, for example, are encrusted with very small crystals that are difficult to characterize even at high magnifications on the SEM (Fig. 5). These crystals appear to be small prismatic forms that
14
B. JONES
Fig. 5. SEM photomicrographs of calcite precipitates associated with exposed surfaces (i.e. cave wall) in twilight zone of Old Man Village cave (Grand Cayman). (a) General view calcified filamentous microbes on a calcite substrate. (b) Enlarged view from Figure 5a showing calcified filament. (c) Enlarged view from Figure 5b showing calcite crystals forming wall of calcified filament. (d) General view of calcified filaments on calcite substrate. (e) Enlarged view from Figure 5d showing calcite crystals forming wall of calcite filament (f) Filamentous microbes associated with calcite crystals. (g, h) Partly calcified filaments on calcite crystal. (i) Enlarged view from Figure 5h showing calcite crystals
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION
have their long axes at 908 to the filament surface (Fig. 5b, c, e, h, i). Some filaments appear to be encased by a double layer of calcite crystals (Fig. 5b, c). Elsewhere, the same filaments are host to small crystal rosettes (Fig. 5j). Mucus associated with these filaments also seems to mediate the growth of calcite crystals (Fig. 5k, l). This style of calcification preserves the general form of the microbe but does not preserve many of the morphological features that might allow identification in terms of extant taxa. The factors that control the morphology of the calcite crystals involved in the calcification of microbes is open to debate. The idea that calcification style and crystal morphology may be taxonspecific (Desikachary 1959; Bourrelly 1970; Riding 1977; Krumbein 1979; Jones & Kahle 1986; Jones 1988; Merz 1992) arises because some microbes seem to be more susceptible to calcification than others, and different microbes commonly display different styles of calcification with crystals of different morphologies. Nevertheless, there is also the possibility that the style of calcification and/or morphology of the calcite crystals are controlled solely by environmental conditions such as saturation levels (Golubic 1973). The former implies that the microbes, through their metabolic activities, play a direct role in calcification, whereas the latter holds that the microbes played no role in calcite precipitation. This is difficult to assess as there is so little information that relates specific microbes to specific styles of calcification under specific environmental conditions.
15
filamentous microbes (Fig. 6d, e); (2) the presence of textures that are consistent with the notion that growth was achieved by the trapping and binding of sediment grains to the substrates; and (3) microbes forming a substantial part of the structures (Jones 2001). Nevertheless, it is virtually impossible to prove that the microbes played an active role in the growth and development of the laminated columns. Alternating light and dark lamina, calcite microrods, organic inclusions, microbialite-like structures found in stalactites from various caves in northern India have been attributed to the trapping and binding activity of microbes (Baskar et al. 2007; Baskar et al. 2009). The presence of mineralized filamentous microbes in some of the stromatolitelike structures from these stalactites added further support to this suggestion (Baskar et al. 2007). Stalactites form Old Man Village cave commonly contain small, irregular-shaped micritic columns that radiate from growth surfaces (Fig. 6f, g). The fact that enveloping calcite layers are deflected around such structures shows that they were growth structures that developed as projections from the exteriors of the stalactites (Fig. 6f). Although most of these structures are formed of dense accumulations of micrite, some stalactites have columns that are formed of diffuse micrite (Fig. 6g). Despite the lack of lamina in these ‘stromatolitic’ columns (Fig. 6a–c), it is still tempting to attribute their formation to biogenic processes. The fact that microbes have not yet been located in such structures, however, makes the case for biogenicity questionable. Thus, their origin and mode of growth must remain open to question.
Trapping and binding Filamentous microbes have the capacity to trap and bind detrital grains to a cave substrate (Cox et al. 1989a, b; James et al. 1994; Contos et al. 2001; Jones 2001). Can˜avas et al. (2001), for example, noted filamentous microbes that were actively trapping and binding micrite grains, animal fragments, and spores to cave substrates. The formation of laminated bulbous stromatolites is commonly attributed to the filamentous microbes that trap and bind sediment grains to a substrate. Thus, it is tempting to suggest that the outward expanding laminated columns found in some speleothems (Fig. 6a –c) originated through microbial activity as suggested by Jones & Motyka (1987). That microbes played an active role in the growth and development of the laminated columns can be inferred from: (1) the presence of
Mediation of CaCO3 precipitation Microbes, through their metabolic activities, can encourage CaCO3 precipitation by modifying the saturation index (SI) of the solution or by neutralizing kinetic factors that may inhibit precipitation (Bosak & Newman 2003, 2005; Barton & Northup 2007). The potential importance of biologically induced precipitation (Lowenstam 1981; Barton & Northup 2007) has been well illustrated by the following studies that used laboratory experiments to show that microbes can induce CaCO3 precipitation. † In laboratory experiments, Castanier et al. (1999) showed that the development of Bacillus cereus involved a phase of latency that was followed by a phase of exponential growth increase that eventually gave way to steady state bacterial
Fig. 5. (Continued) forming around filament. (j) Clusters of calcite crystals associated with filaments. (k) Surface of calcite crystal covered with mucus, filaments and small clusters of precipitated calcite. (l) Small calcite platelets on surface of calcite crystal – note consistent orientation of platelets.
16
B. JONES
Fig. 6. Internal features of some Cayman stalactites. (a) Transverse cross-section through stalactite from old cave near Breakers (Fig. 1) showing central ‘soda-straw’ (SS) surrounded by outward expanding laminated ‘stromatolitic’ columns encased by calcite. (b, c) Enlarged views of laminated ‘stromatolitic’ columns formed of micrite. (d, e) SEM photomicrographs of hollow filamentous microbes encrusted with ‘spiky’ calcite crystals, from ‘stromatolitic’ columns shown in Figure 6a. (f) Thin-section photomicrograph of showing irregular-shaped micritic columns (arrows) in stalactite from Old Man Village cave. Note deflected of outer lamina around the columns. (g) Thin-section photomicrograph showing small non-laminated columns formed of dark grey-black micrite in stalactite from Old Man Village cave. (h) Thin section photomicrograph showing diffuse column of micrite in stalactite from Old Man Village cave.
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION
growth. Their experiments showed that Cacarbonate precipitation took place during the phase of exponential growth. From these experiments, Castanier et al. (1999) developed the notion of the carbonatogenic yield, which is defined as the ratio between the weight of produced solid Ca-carbonate and the weight of organic matter input. † Rautaray et al. (2003) produced different types and different crystal forms of CaCO3 when they added Fusarium sp. (fungus) and Rhodococcus sp. (actinomycete) to an aqueous CaCl2 solution and incubated it at 27 8C. In these experiments, Ca came from the solution whereas the CO2 came from the microbes. Fusarium sp. yielded cruciform-shaped calcite crystals whereas Rhodococcus sp. produced rounded crystals of vaterite. Experiments by Ahmed et al. (2004) and Rautaray et al. (2004) produced similar results. Solutions inoculated with Fusarium oxysproum (fungi) produced CaCO3 crystallites that formed a circular superstructure composed of calcite with minor amounts of vaterite whereas Trichothecium sp. (fungi) produced plate-shaped crystals of calcite with preferential growth along the crystal edges (Ahmad et al. 2004). Similarly, solutions with Verticillium sp. (fungi) yielded circular-shaped nanocrystals (70–100 nm diameter) of vaterite whereas Thermomonos sp. (actinomycete) resulted in the growth of plate-shaped calcite and aragonite crystals, 20–30 nm thick, on the mycelia and extracellularly (Rautaray et al. 2004). Fourier transform infrared (FTIR) analyses showed that all of these precipitates had proteins incorporated in the crystal frameworks with the types of protein varying in accord with the mineral polymorph and crystal type. Indeed, all of these studies suggested that the proteins secreted by the microbes controlled the type of precipitate and crystal morphology. † Groth et al. (2001) tested the influence of various microbes (mainly actinomycetes), obtained from Grotta dei Cervi, Italy on mineral precipitation by culturing them on a variety of different media. These experiments showed that the actinomycetes produced different types of minerals (calcite, vaterite) and crystal forms that varied in accord with the carbon source and the salts that were present. † Using bacteria (mostly Bacillus and Arthrobacter) isolated from Stiffe cave, Italy, Ercole et al. (2001) and Cacchio et al. (2003) showed that many of them were capable of mediating CaCO3 precipitation. Although rhombohedral calcite crystals (Cacchio et al. 2003, figs 4–6) were the main precipitates, minor amounts of vaterite were also produced in some
17
experiments. Subsequently, Cacchio et al. (2004) conducted experiments using eleven strains of bacteria (Kocuria spp., Acinetobacter spp., Bacillus sp.) isolated from Cervo cave, Italy. Calcite produced in association with Kocuria sp. was in the form of spheres whereas Acinetobacter spp. produced crystal aggregates in globular forms. These experiments showed that all of the bacteria induced calcite precipitation with some also producing vaterite. † Hemispherical vaterite aggregates were produced in laboratory cultures of Acinetobacter spp. that Sanchez-Moral et al. (2003) prepared using microbial isolates from Altamira cave, Spain. These spherulites, 5 –20 mm in diameter, had a radial internal structure, and were in some cases, hollow. Crystals only formed when the Ca acetate/Mg acetate ratio was .1. Cultured Rhodococcus sp. produced spherulitic crystals formed of monohydrocalcite. Precipitation of metastable vaterite is favoured by various factors including the presence of organic substances such as amino acids or mucopolysaccharides (Manoli & Dalas 2000) and phosphorus-rich media (Katsifara & Spanos 1999). † Baskar et al. (2006) showed that Bacillus thuringiensis and B. pumilis, isolated from stalactites collected from a cave in Dehradin Valley, India, mediated the growth of calcite under laboratory conditions. By varying the temperatures of incubation, they showed that 25 8C was the optimum temperature for precipitation.
Microbial processes in the twilight zone The twilight zone, found around the entrance to the cave, is the transition between normal light conditions outside of the cave and the dark cave interiors (Cox 1977; Cox & Marchant 1977; Hoffman 1989). Geographical location and cave configuration influence the climatic conditions in the twilight zone with temperature, humidity, and light gradients being especially important (Hoffman 1989). That microbes thrive in the twilight zone is readily apparent from the green colour of the biofilms that coat the cave walls. The formative microbial communities of these biofilms change as light levels decrease towards the interior of the cave (Hoffman 1989; Rolda´n et al. 2004). Thus, green algae and cyanobacteria thrive on the well-illuminated substrates near the cave entrance (Dobat 1968) whereas the darker interior areas are characterized by fewer microbes (Hoffman 1989). Rolda`n et al. (2004) examined phototrophic biofilms from three limestone cavities in the karstic Garraf massif of Spain. Although not caves in the strictest sense, the trends they discovered in those
18
B. JONES
cavities are applicable to caves. They showed that systematic changes in the temperature, relative humidity, and light levels were reflected in the biofilms and thereby identified three zones: 1.
2.
3.
Entrance zone where the microclimate is strongly influence by the outside climate and light levels were high and microclimatic conditions fluctuated widely throughout the year. Algae, cyanobacteria, crustose lichens, and mosses generated the abundant mucilaginous biofilms found in this zone; Intermediate zone where light levels were low and microclimate conditions underwent only moderate oscillations. Although the algal biofilms are thinner and contain less mucilage than those in the entrance zone, the microbial communities are formed of cyanobacteria and green algae; and Deep zone where light levels ranged from very low to zero but microclimate variables were relatively stable. Only a few taxa are found in this zone, including Geitleria calcarea, Loriella osteophila, various bacteria, and filamentous fungi.
Besides controlling the species consortia, the light and relative humidity gradients also appear to influence the nature of the biofilm (Rolda´n et al. 2004). Thus, biofilms formed of mucilaginous and dark coloured cyanobacteria were gradually replaced by biofilms formed of calcified filaments with colourless sheaths as the light levels decreased and the relative humidity became more stable. It is important to note, however, that distance from the cave entrance may not be the only factor that controls the nature of the biofilm. The amount of light reaching a substrate, for example, also depends on the orientation of the cave wall relative to the cave entrance. Substrates facing the cave entrance will be under the influence of incoming light whereas substrates that face towards the cave interior will be in a shadow zone and hence receive less light. The type of biofilm that develops on a particular substrate may also be partly controlled by the water retention capabilities of the host rock (De Winder et al. 1989), the ability of a microbe to attach itself to a particular rock type (Guillitte & Dreesen 1995; Herna´ndez-Marine´ et al. 2001), and possibly, interactions between species (Costerton 2000). The biofilms found on cave walls in the twilight zone are important because they mediate a wide range of destructive and constructive geological processes that can have a significant impact on the host substrates (Jones 1995). Microbialites with stromatolite-like laminations formed of alternating CaCO3 (calcite and aragonite) and kerolite (Mg-rich silicate) found on cave walls in the twilight
zone of caves on the north coast of Kauai, Hawaii, are produced by biofilms (Le´veille´ et al. 2000).
Microbial processes in the aphotic zone The aphotic zone, defined simply as that part of the cave where there is no light, encompasses many different habitats, ranging from the phreatic realms in the cave pools, to flowing streams, to areas under the influence of dripping waters, to bare dry surfaces. The different processes that operate in these settings, be they biogenic or abiogenic in nature, operate in accord with local conditions. These contrasting settings and their associated microbial processes are herein illustrated by considering the growth and development of stalactites and stalagmites, pool fingers, cave pisoliths, and moonmilk.
Stalactites and stalagmites The growth of stalactites and stalagmites through the precipitation of calcite has commonly been viewed as an abiogenic process (e.g. Kendall & Broughton 1978; Broughton 1983a, b, c) even though the presence of microbes on stalactites and stalagmites has long been known (e.g. Palik 1960; Nagy 1965; Went 1969; Hasselbring et al. 1975). That microbial communities grow and thrive on stalactites and stalagmites in caves throughout the world has been amply demonstrated in many recent studies (Groth et al. 2001; Basker et al. 2005; Barton 2006; Baskar et al. 2006; Baskar et al. 2007; Baskar et al. 2009). Accordingly, it has been suggested that the microbes may play an important role in the growth of stalactites and/or stalagmites (Jones & Motyka 1987; Mulec et al. 2007). Lamina in some speleothems have been likened to microbialites/stromatolites because they: (1) contain mineralized microbes (Fig. 6d, e); (2) contain organic inclusions; (3) locally form outward expanding columns (Fig. 6a– c); and (4) include precipitates akin to those found with terrestrial microbialites (Jones 2001; Baskar et al. 2007; Baskar et al. 2009). Jones & Motyka (1987) suggested that microbes contributed to the growth of stalactites as they became calcified, provided micrite through the breakdown of calcified filaments, and trapped and bound sediment to the substrate on which they were growing.
Pool fingers Pool fingers, which are up to 2 cm in diameter and 50 cm long, were first identified from Lechuguilla Cave New Mexico, USA by Davies et al. (1990). Although no longer in water, Davies et al. (1990)
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION
considered them to be subaqueous pool structures because they: (1) hang down from shelfstone that formed around the edge of cave pools; (2) are the same colour as the shelfstone; and (3) are absent at levels above the shelfstone. Pool fingers from Hidden Cave, New Mexico, USA, which have a rough outer surface with knobs 10– 20 mm wide that extend outward for 1 –10 mm, are internally laminated with layers of dark micrite alternating with layers formed of clear dogtooth calcite crystals (Melim et al. 2001). Individual micrite laminae are variable in thickness, largely because local protrusions extend outward for 50– 150 mm (Melim et al. 2001, figs 6A– C). Needle fibre and dendrite calcite crystals commonly form between the micrite patches. Two types of calcified filaments were found in the micrite layers, but not in the layers formed of dogtooth spar calcite. These filaments c. 1 mm in diameter and 5–50 mm long, include forms that have a smooth exterior (Melim et al. 2001, figs 7A–B) and forms that display a crosshatch surface texture (Melim et al. 2001, figs 7C – D). The identity of these microbes and especially the one with the crosshatch surface texture remains open to debate (Melim et al. 2008). On the basis of: (1) internal fabrics that are similar to microbialites; (2) the presence of mineralized microbes; and (3) the consistent depletion of d13C in the micrite layers relative to the spar layers, Melim et al. (2001) argued that bacteria exerted a strong influence over the growth of the pool fingers and largely controlled their internal and external morphology.
Cave pisoliths These coated grains (also known as ‘cave pearls’, ‘oolites’), which are typically formed of a nucleus that is encased by concentrically laminated cortical lamina, have been reported from caves in Australia (Baker & Frostick 1947, 1951), Austria (Kirchmayer 1969, 1987), Belgium (Lie´geois 1956), the Cayman Islands (Jones & MacDonald 1989), Cuba (Gradzinski & Radomski 1967), Norway (Erdman 1902), Ireland (Coleman 1949), North America (Hess 1930; Stone 1932; Keller 1937; Pond 1945; Black 1952; Thrailkill 1963), Lebanon (Abdul-Nour 1991; Choppy 1991; Karkabi 1991; Nader 2007), and Poland (Barcyzk 1956; Gradzinski 1999). These coated grains have been divided into: (1) spherical to subspherical forms with a polished exterior and distinct, compact cortical lamina that typically grow in highly agitated splash pools; and (2) irregular shaped forms with a rough, unpolished surface and indistinct, porous lamina that form in pools where there is little agitation. The role that microbes may play in the growth and development of cave pisoliths is open to
19
debate. In many cases the possibility that microbes played a role in their formation was automatically excluded because it was assumed that algae could not grow in the absence of light (Donahue 1965, 1969; Gradzinski & Radomski 1967). In other situations, failure to detect microbes has been used as evidence for an abiogenic origin of the pisoliths (e.g. Nader 2007). Cave pisoliths from Old Man Village (Jones & MacDonald 1989), up to 8 cm long, have laminated cortices that are commonly characterized by laminated, outward expanding columns (Fig. 7). The presence of mineralized spores (Fig. 7a– h), mucus (Fig. 7h, i), and calcified filamentous microbes (Fig. 7i –l) in the cortices of these pisoliths raises the possibility that microbes influenced growth of the pisoliths by being direct contributors, trapping and binding of detrital grains brought into the spring pools, and possibly by indirectly mediating calcite precipitation. Gradzinski (2001, 2003), for example, suggested that cave pisoliths grow as calcite is precipitated due to the metabolic activity of bacteria that reside in the mucilaginous biofilm that coat the surfaces of the pisoliths.
Moonmilk Moonmilk, found in caves throughout the world, is a whitish, porous, plastic deposit formed of crystals and water (White 1976; Hill & Forti 1997). The crystals may be formed of many different minerals, including calcite, aragonite, monohydrocalcite, vaterite, huntite or gypsum (Onac & Ghergari 1993; Can˜averas et al. 1999; Borsato et al. 2000; Can˜averas et al. 2006; Martı´nez-Arkarazo et al. 2007). Calcitic moonmilk is typically formed of long, fiber crystals (Borsato et al. 2000; Can˜averas et al. 2006), including crystals known as lublinite (Krischtafowitsch 1906; Morozewicz 1907, 1911; Muegge 1914; Ulrich 1938; Kowalinski et al. 1972; Stoops 1976). Bernasconi (1981) and Jones & Kahle (1993), however, recommended abandonment of the term lublinite. Calcitic moonmilk has been attributed to both biogenic and abiogenic origins. The debate over the role of microbes in the formation of moonmilk has arisen because filamentous microbes and/or bacteria are commonly associated with these deposits (Mason-Williams 1959; Bertouille 1972; James et al. 1982; Dziadzio et al. 1993; Can˜averas et al. 1999; Mulec et al. 2002; Can˜averas et al. 2006). Some studies have suggested that the long fibre crystals form through the calcification of bacterial cells (Gradzinski et al. 1997), as coatings on filaments (Moore & Bukry 1968), or by dissolving the bedrock and thereby generating the CaCO3 needed for calcite precipitation (Ge`ze et al. 1956; Ge`ze & Pobe´guin 1956). In the absence of microbes,
20
B. JONES
Fig. 7. Microbes in pool fingers from Cottonwood Cave, New Mexico. (a) Group of pool fingers with palaeo water level indicated. Vertical height is c. 1 m. Photograph by Kenneth Ingram. (b, c) SEM photomicrographs filamentous microbes with distinctive reticulate morphology. Microbes revealed by etching of calcite from a pool finger. SEM images courtesy of Leslie Melim.
calcite moonmilk has been attributed to abiogenic processes (Shumenko & Olimpiev 1977; Harmon et al. 1983; Onac & Ghergari 1993; Hill & Forti 1997; Borsato et al. 2000). A number of issues hinder interpretation of the origin of moonmilk. The fact that the crystallographic form of fibrous calcite crystals may vary because of growth conditions and/or subsequent diagenetic modifications (Jones & Kahle 1993; Verrecchia & Verrecchia 1994; Can˜averas et al. 1999) is commonly disregarded in discussions concerning the origin of moonmilk. Thus, any information that may be elicited from the morphology or the crystals is ignored. Attributing the formation of fibrous calcite crystals to the direct calcification of filamentous microbes is difficult to support because: (1) such crystals are typically straight and very long whereas filamentous microbes are typically sinuous; (2) there are no examples of fibre crystals that have a central, longitudinal open lumen that runs the length of the crystal; and (3) there are no examples of fibre crystals that encompass calcified microbes in their crystalline structure. Nevertheless, Can˜averas et al. (2006) suggested that a two-stage process may be involved in the formation of moonmilk whereby: (1) precipitation of monocrystalline rods, preserving filament morphology, from fluids that had high
supersaturation levels due to microbial activity; and (2) epitaxial growth of the fibers as the cessation of microbial activity leads to decreased saturation levels and nucleation rates.
Effective role microbes in speleothem formation Calcitic speleothems, which are common in caves throughout the world, may form through abiogenic, biogenic, or a combination of abiogenic and biogenic processes. Many issues, however, conspire to make the assessment of biogenicity difficult, especially when the focus is on old speleothemic deposits. In lithified speleothems, for example, abiogenic mineral precipitates can easily be mistaken for mineralized microbes (e.g. Barton et al. 2001, table 1). Accordingly, Barton et al. (2001) suggested that putative fossil microbes should be assessed by the same criteria that Schopf (1999) validated fossil microbes from ancient strata, namely: (1) provenance; (2) age; (3) indigenousness; (4) syngenicity; and (5) biogenicity. The first four criteria are designed to identify contaminants that were introduced long after the deposit had formed, whereas the biogenicity assesses the biogenic origin. Ideally, biogenicity should only be accepted if the
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION
putative fossil: (1) is formed of organic matter or is clearly mineral-replaced; (2) is complex enough in cellular structure to rule out abiogenic origins; (3) is represented by numerous specimens; (4) is a member of a multicomponent assemblage; (5) exhibits a range of taphonomic degradation that is consistent with their mode of preservation; (6) exhibits morphological variability; (7) is found in a plausible living environment; (8) grew and reproduced by biological means of cell division; and (9) exhibits a biogenic isotopic signature (Schopf 1999; Barton et al. 2001). Although biogenicity should be supported by as many of these criteria as possible, preservation commonly dictates that many of these criteria cannot be verified. Establishing biogenicity depends on the preservation of the microbes and, in particular, how well the original features of the microbes are preserved. Clearly, the presence of mineralized microbes indicates that microbes were associated with the speleothems. The converse, however, does not necessarily indicate that microbes were absent for it could equally reflect a lack of preservation or failure to identify the microbes amid the calcite of the speleothem. Experimental work by Bartley (1996) demonstrated that the soft tissues of microbes will be rapidly lost to decay in a matter of days. Similarly, Le´veille´ et al. (2000) noted the paucity of preserved cells in mineralized microbialites found on cave walls in the twilight zone of caves on the north coast of Kauai, Hawaii. Observations such as these imply that the threedimensional preservation of some microbes through calcification must take place very quickly (Jones & Kahle 1986; Gradzinski 1999) as, for example, in the case of Geitleria (Fig. 4). Paradoxically, if microbes directly or indirectly encourage calcite precipitation they may be authoring their own demise as the precipitated calcite encases them (Barton et al. 2001; Barton & Northup 2007). These issues are not unique to caves, for the same debate arises with respect to the silicification of microbes in hot-spring systems (e.g. Jones et al. 1997), where some microbes remain viable even though encased with opal-A precipitates (Phoenix et al. 2000; Jones et al. 2001). In this case, it was suggested that the precipitated opal-A protected the microbe by shielding them from UV light (Phoenix et al. 2001). By analogy, this raises the possibility that early calcification of spelean microbes may be advantageous to them in some manner. The preservation potential of microbes also appears to be a function of the type of mineral being precipitated and the form of that precipitate. For microbes in hot spring systems, for example, it appears that the preservation potential of microbes is high if non-crystalline (¼ amorphous) precipitates are forming but low if crystalline
21
precipitates are forming (Jones & Renaut 2007). Similar processes may be operative in caves as microbes are usually more commonly detected in micrite lamina and rarely in spar calcite lamina. Fabrics and structures consistent with microbial activity such as stromatolite-like structures are commonly apparent in hand samples or thin sections (Figs 6a–c, f –h, 7a & 8a–b). In some cases, locating microbes is easy for they contrast readily with the surrounding substrates (Figs 2–5). In other cases, however, the search for mineralized microbes can become an exercise in frustration. In the cave pisoliths from Old Man Village cave, for example, the microbes are hidden and camouflaged in the micrite lamina. Thus, cursory inspection of the micrite lamina at low magnification (e.g. Fig. 8c, k) reveals the constituent grains but no indication of microbes. Careful inspection at high magnification, however, reveals some areas with numerous spores, c. 1 mm in diameter (Fig. 8d) and other areas with hints of mucus (Fig. 8i) and/or small, poorly preserved filaments (Fig. 8l). Elsewhere in these laminae there are spherical calcite bodies, c. 1 mm in diameter, that contrast sharply with the small calcite grains in the surrounding micrite (Fig. 8h). Although lacking obvious biogenic features, their overall similarity in size and shape to the spores found elsewhere in the lamina (Fig. 8e) suggests that they are calcified spores. The problem of locating and identifying microbes in micrite is not unique to cave deposits. Benzerara et al. (2006), for example, noted the difficulty of locating and identifying microbes amid finely crystalline aragonitic microbialites that are actively growing in Lake Van (Turkey). Although difficult to locate using conventional optical and scanning microscopy, Benzerara et al. (2006) showed that nanometerscale analyses using scanning transmission X-ray microscopy (STXM) combined with transmission electron microscopy (TEM) allowed the microbes to be located and thereby demonstrated the key role that they played in the development of the Lake Van microbialites. These examples, from the Cayman cave pisoliths and the Lake Van microbialites, clearly illustrate the problems that can arise in the search for evidence of microbes. Well-preserved calcified microbes, which should probably be considered the exception rather than the rule, invariably bring the desire to identify and name them in terms of extant taxa. As with mineralized microbes in hot-spring systems, this is fraught with problems (Jones et al. 2001, 2004). Although calcification leads to microbe preservation, the mineralizing processes commonly destroy or fail to preserve most of the features that are of taxonomic importance, in the same way that silicification does in hot spring settings (Jones et al. 2001, 2004). Mineralization can preserve
22
B. JONES
Fig. 8. Thin section (a & b) and SEM photomicrographs (c– l) of cave pisoliths from Old Man Village cave. (a, b) Thin section photomicrograph of transverse sections through two cave pisoliths showing internal laminations. Note ‘stromatolite-like’ columns in outer part of pisoliths shown in Figure 8b. Dark layers formed of micrite, light layers formed of spar calcite. (c) Peripheral micrite lamina (between white dots) in a pisolith that contains microbes and textures shown in Figure 8d– l. (d) Group of smooth actinomycetes spores, each with a central open pore. (e) Smooth actinomycetes amid micrite. Note thin wall of broken spore (upper part of image). (f ) Actinomycetes spore with reticulate coating and pores. (g) Two broken actinomycetes spores buried in micrite groundmass. (h) Spherical bodies
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION
morphological features but chemical signatures of the microbes are rarely, if ever, preserved. Thus, identification in terms of extant taxa depends, to a large extent, on the criteria first used to establish the taxonomy of those taxa. As chemotaxonomic characteristics are increasingly used to define new microbial taxa, so does the probability of associating a mineralized microbe with an extant microbe decrease. This is especially true in situations where modern microbes are defined with virtually no mention of their morphological attributes. Indeed, when dealing with mineralized microbes it is commonly difficult even to determine affinity to one of the major microbial groups. The notion that cave microbes can influence calcite precipitation arises, at least in part, from the fact that such microbes directly or indirectly induce calcite precipitation when cultured in the laboratory (e.g. Cacchio et al. 2003; Rautaray et al. 2003; Cacchio et al. 2004; Rautaray et al. 2004). Even this must be treated with some caution given the difficulty of exactly replicating cave conditions in a laboratory setting (Chalmin et al. 2008). Nevertheless, mineralized microbes found in calcite speleothems have commonly been implicated in their growth and development through a wide variety of destructive and constructive processes (e.g. Jones 2001; Melim et al. 2001; Northup & Lavoie 2001; Barton & Northup 2007). The mere presence of microbes, however, does not guarantee that they actually played any role in the formation of the surrounding calcite because they may simply have been buried in the calcite during precipitation (Polyak & Cokendolpher 1992; Forti 2001). Assessment of the exact role that the microbes played in the calcite accumulation/ precipitation may be virtually impossible to determine given the complexity of the systems involved. Even in controlled laboratory experiments it can be difficult to determine the exact reason why a particular CaCO3 polymorph was precipitated or why a particular crystal form resulted. Available evidence – be it from experiments, from microbiological studies, or from geological studies – clearly indicates that microbes may play an important role in the growth and development of calcitic speleothems. In most cases, however, it is difficult to know if microbes played an active role in calcite precipitation or were passive bystanders that suffered accidental burial in the calcite (Forti 2001). Interestingly, similar debates are ongoing with respect to the role(s) that microbes
23
play in the growth and development of opal-A precipitates that are accumulating on the discharge aprons of many springs (e.g. Konhauser et al. 2004). Absolute proof of the role that microbes play in the formation of calcitic speleothems may be impossible to attain given the microscale at which the processes operate. Despite the problems it should be possible to infer, with a reasonable level of confidence, that microbes were, in one way or another, instrumental in the growth and development of speleothems. If mineralized microbes are present, this can be based on documentation and recognition of microbes in numbers sufficient to have mediated any process that is attributed to them (Jones 2001; Melim et al. 2001), and recognition of structures and fabrics (e.g. stromatolites, dissolution features) that are consistent with known microbial activity (Jones 2001; Melim et al. 2001). In the absence of mineralized microbes, or if indirectly microbial influences are suspected, proxies for microbial activity must be obtained. Such proxies may include: (1) lighter d13C isotopic signatures that may result from microbial fractionation of C (Melim et al. 2001; Le´veille´ et al. 2007); (2) heavier d18O isotopic signatures that can be ascribed to bacterial activity (Gradzinski 2001, 2003); (3) biomarker signals related to microbially derived lipid groups (Blyth & Frisia 2008); or (4) the presence of proteins in the calcite that were inherited from the formative microbe (Rautaray et al. 2003; Ahmad et al. 2004; Rautaray et al. 2004).
Implications for palaeoclimate studies The calcite in stalagmites and stalactites is genetically linked to the groundwater from which it is precipitated and should therefore incorporate geochemical signals that can be related to the climate of the area. This premise has led to the development of various climatic proxies, including growth rates as determined from annual growth laminae (Genty & Quinif 1996; Qin et al. 1999; Proctor et al. 2000; Baker et al. 2008), stable isotopes (BarMattchews et al. 1999; Lauritzen & Lundberg 1999; McDermott et al. 1999), Sr isotopes (Banner et al. 1996; Goede et al. 1998; Verheyden et al. 2000), and trace elements (Roberts et al. 1998; Fairchild et al. 2001; Huang et al. 2001; Treble et al. 2003; Borsato et al. 2007). These studies tacitly assumed that the calcite crystals in the stalagmites
Fig. 8. (Continued) that are probably actinomycetes spores encased with calcite. Note calcified mucus in background (arrow). (i) Possible calcified mucus (arrows) coating surface of calcite crystals. (j) Calcified, partly collapsed filaments embedded in micrite matrix. (k) General view of micrite lamina with no obvious microbes. Square and white letter ‘l’ indicates position of Fig. 8l. (l) Poorly preserved collapsed filaments embedded in micrite groundmass.
24
B. JONES
and stalactites had been abiogenically precipitated for no consideration was given to the possible effects of microbial activity. Such an approach is curious given the overwhelming evidence for diversity microbial biotas that have been implicated in the broad range of processes that influence the corrosion and/or precipitation of speleothemic calcite. Trace elements concentrations in speleothemic calcite has been assessed using synchrontron radiation micro X-ray fluorescence (m–XRF), ion microprobe analyses (Borsato et al. 2007), excimer laser-ablation inductively coupled plasma mass spectrometry (Treble et al. 2003), and secondary ion mass spectrometry (SIMS) (Huang et al. 2001). All of these methods are capable of determining the concentrations of elements such as Y, Pb, Cu, Sr, Mg (e.g. Borsato et al. 2007), Si, Sr, Fe (e.g. Huang et al. 2001), and P, Na and H (e.g. Fairchild et al. 2001) in the 0–1000 ppm concentration range from spot analyses that are typically ,30 mm in diameter (Huang et al. 2001). The fact that fluctuations in the concentrations of these elements commonly follows a systematic pattern that seem to correlate with the laminations in the speleothems has led to the notion that such variations are proxies for seasonal climatic changes (e.g. Roberts et al. 1998; Fairchild et al. 2001; Huang et al. 2001; Borsato et al. 2007). Phosphor, for example, has been related to increased water infiltration that coincides with the first storm of the autumn (Fairchild et al. 2001; Huang et al. 2001; Treble et al. 2003; Treble et al. 2005; Borsato et al. 2007). Where P resides in the speleothems is unknown. Suggestions include: (1) individual P ions located in calcite defects (Fairchild et al. 2001; Huang et al. 2001; Mason et al. 2007); (2) various types of coexisting phosphate inclusions, including monetite (CaHPO4); and (3) various unidentified crystalline phases (Mason et al. 2007). The possibility that P and other trace elements may be related, in some manner, to microbes has been ignored. Jones (2009) recently demonstrated that much of the P in speleothems from Old Man Village cave on Grand Cayman was preferentially associated with the actinomycetids that colonized corrosion surfaces. Indeed, the actinomycetes had been mineralized with P. Thus, it could be argued that the P was there because of microbial mineralization, not because of a climatic event. Little attempt has been made to assess the affect that microbes may have on the geochemical attributes in speleothemic calcite that are used as climate proxies. This may prove difficult until methods for detecting microbes in these speleothems by direct observation (e.g. SEM images) or biomarker proxies (e.g. lipids, proteins) become well established. Scale becomes an issue because
most cave microbes are submicron in size and the amount of an element that can be fixed by an individual microbe is, as yet, unknown. With geochemical analyses seeking palaeoclimate signals, however, the trend is towards smaller spot sizes (1 mm in some cases) and ever-lower concentrations of an element (,100 ppm). Thus, such analyses are increasingly approaching the scale of the microbes, both in terms of the size of the sample being analyzed and the amount of an element being determined. Accordingly, there is the possibility that such geochemical analyses may be derived from mineralized microbe(s) rather than from the calcite itself is also increasing. Clearly the a priori assumption of abiogenicity that is inherent to many geochemical studies of speleothems must be treated with great caution for there is the very real possibility that such analyses may in fact reflect microbial mineralization processes. This is clearly illustrated in the case of the P in the Cayman speleothems that is directly associated with microbe mineralization (Jones 2009).
Conclusions With respect to microbes in caves, it is important to stress and highlight the following points. † Although most caves are home for rich, diverse microbial biotas, the diversity and nature of that biota is still poorly known. † Identification of mineralized (e.g. calcified) microbes in terms of extant taxa is difficult and depends largely on the criteria used to define the extant taxa. There is little chance of matching mineralized microbes with extant taxa defined solely by DNA characteristics. Comparisons based on morphological attributes are also difficult because mineralization commonly destroys or disguises the taxonomically important features. † Microbial boring and microbially mediated dissolution leads to substrate destruction and, in many cases, production of residual micrite. † Microbes mediate constructive processes through microbe calcification, trapping and binding of detrital grains to a substrate, and/or inducing calcite precipitation by modifying microenvironmental conditions. † Recognition of microbes in speleothems depends on the location of mineralized forms and/or the detection of biogenic proxies. † The mere presence of microbes in a speleothem does not guarantee that they played a formative role in the growth and development of that speleothem. † That microbes played a role in the growth and development of a speleothem can be inferred
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION
from the presence of mineralized microbes and the presence of fabrics indicative of microbial activity. Over the last 10–20 years, knowledge of spelean microbial communities has increased greatly as microbiologists have focused attention on these biotas. Despite this, much remains to be learnt about the roles that these microbes play in the growth and development of cave deposits and, in particular, calcite speleothems. I am indebted to the Natural Sciences and Engineering Research Council of Canada for funding of the research on which this paper is based (grant A6090); Captain Ned J. Millar for permission to collect samples from the Old Man Village cave that is on his property; Hendrik van Genderen, The Water Authority, Cayman Islands, for his assistance in the field; George Braybrook, University of Alberta, who took the SEM images used in this paper; Dr Martyn Pedley for encouraging me to write this paper, and Ian Fairchild and an anonymous reviewer who critically reviewed an earlier version of this paper.
References A BDELAHAD , N. & B AZZICHELLIA , G. 1988. Geitleria calcarea Friedmann, cyanophyce´e cavernicole nouvelle pour l’Italie. Nova Hedwigia, 46, 265–270. A BDUL -N OUR , H. 1991. Perles cubiques. Liban Souterrain. Bulletin du Groupe d’Etudes et de Recherches Souterraines du Liban, 3, 64– 65. A HMAD , A., R AUTARAY , D. & S ASTRY , M. 2004. Biogenic calcium carbonate: calcite crystals of variable morhology by the reaction of aqueous Ca2þ ions with fungi. Advanced Functional Materials, 14, 1075–1080. A MANN , R. I., L UDWIG , W. & S CHLEIFER , K. H. 1995. Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiological Reviews, 59, 143– 169. A MANN , R. I., S NAIDR , J., W AGNER , M., L UDWIG , W. & S CHLEIFER , K. H. 1996. In situ visualization of high genetic diversity in a natural community. Journal of Bacteriology, 178, 3496–3500. A RIN˜ O , X., H ERNA´ NDEZ -M ARINE´ , M. & S AIZ -J IMENEZ , C. 1997. Colonization of Roman tombs by calcifying cyanobacteria. Phycologia. 36, 366–373. B AKER , A., S MITH , C. L., J EX , C., F AIRCHILD , I. J., G ENTY , D. & F ULLER , L. 2008. Annually laminated speleothems: a review. International Journal of Speleology, 37, 193–206. B AKER , G. & F ROSTICK , A. C. 1947. Pisoliths and ooliths from some Australian caves and mines. Journal of Sedimentary Petrology, 17, 39– 67. B AKER , G. & F ROSTICK , A. C. 1951. Pisoliths, ooliths and calcareous growths in limestone caves at Port Campbell, Victoria, Australia. Journal of Sedimentary Petrology, 21, 85– 104. B ANNER , J. L., M USGROVE , M. L., A SMERON , Y., E DWARDS , R. L. & H OFF , J. A. 1996. High-resolution temporal record of Holocene ground-water
25
chemistry: tracing links between climate and hydrology. Geology, 24, 1049–1053. B AR -M ATTCHEWS , M., A YALON , A. & K AUFMAN , A. 1999. The eastern Mediterranean paleoclimate as a reflection of regional events: Soreq cave, Israel. Earth and Planetary Science Letters, 166, 85– 95. B ARCYZK , W. 1956. Pizolity jaskiniowe w jednej z grot w Wojcieszowie (On cave pisoliths from Wojcieszo´w, Polish Sudetin). Acta Geologica Polonica, 6, 327– 336. B ARTLEY , J. K. 1996. Actualistic taphonomy of cyanobacteria: implications for the Precambrian fossil record. Palaios, 11, 571– 586. B ARTON , H. A. 2006. Introduction to cave microbiology: a review for the non-specialist. Journal of Cave and Karst Studies, 68, 43–54. B ARTON , H. A. & J URADO , V. 2007. What’s up down there? Microbial diversity in caves. Microbe, 2, 132– 138. B ARTON , H. A. & N ORTHUP , D. E. 2007. Geomicobiology in cave environments: past, current and future perspectives. Journal of Cave and Karst Studies, 69, 163 –178. B ARTON , H. A., S PEAR , J. R. & P ACE , N. R. 2001. Microbial life in the underworld: biogenicity in secondary mineral formations. Geomicrobiological Journal, 18, 359– 368. B ARTON , H. A., T AYLOR , M. R. & P ACE , N. R. 2004. Molecular phylogenetic analysis of a bacterial community in an oligotrophic cave environment. Geomicrobiological Journal, 21, 11– 20. B ASKAR , S., B ASKAR , R. & K AUSHIK , A. 2007. Evidences for microbial involvement in the genesis of speleothem carbonates, Borra Caves, Visakhapatnam, India. Current Science, 92, 350– 355. B ASKAR , S., B ASKAR , R., L EE , N. & T HEOPHILUS , P. K. 2009. Speleothems from Mawsmai and Krem Phyllut caves, Meghalaya, India: some evidences on biogenic activities. Environmental Geology, 57, 1169–1186. B ASKAR , S., B ASKAR , R., M AUCLAIRE , L. & M C K ENZIE , J. A. 2006. Microbially induced calcite precipitation in culture experiments: possible origin for stalactites in Sahastradhara caves, Dehradun, India. Current Science, 90, 58–64. B ASKER , S., B ASKAR , R., M AUCLAIRE , L. & M C K ENZIE , J. A. 2005. Role of microbial community in stalactite formation, Sahastradhara caves, Dehradun, India. Current Science, 88, 1305–1308. B ENZERARA , K., M ENGUY , N., L O´ PEZ -G ARCIA , P., Y OON , T.-H., K AZMIERCZAK , J., T YLISZCZAK , T., G UYOT , F. & B ROWN , G. E. 2006. Nanoscale detection of organic signatures in carbonate microbialites. Proceedings of the National Academy of Sciences of the United States of America, 103, 9440– 9445. B ERNASCONI , R. 1981. Moonmilch: two questions of terminology. Proceedings of the 8th International Congress of Speleology, 113–116. B ERNER , R. A. 1978. Rate control on mineral dissolution under earth surface conditions. American Journal of Science, 278, 1235– 1252. B ERTOUILLE , H. 1972. Le monmilch, repe`re climatique? Annales de Spe´le´ologie, 27, 609– 623. B LACK , D. M. 1952. Cave pearls in Carlsbad Caverns. Scientific Monthly, 74, 206–210.
26
B. JONES
B LYTH , A. J. & F RISIA , S. 2008. Molecular evidence for bacterial mediation of calcite formation in cold high-altitude caves. Geomicrobiological Journal, 25, 101– 111. B ORSATO , A., F RISIA , S., F AIRCHILD , I. J., S OMOGYI , A. & S USINI , J. 2007. Trace element distribution in annual stalagmite laminae mapped by micrometerresolution X-ray fluorescence: implications for incorporation of environmentally significant species. Geochemica et Cosmochemica Acta, 71, 1494– 1512. B ORSATO , A., F RISIA , S., J ONES , B. & VAN DER B ORG , L. 2000. Calcite moonmilk: crystal morphology and environment of formation in caves in the Italian Alps. Journal of Sedimentary Research, 70, 1179– 1190. B OSAK , T. & N EWMAN , D. K. 2003. Microbial nucleation of CaCO3 in the Precambrian. Geology, 31, 577–580. B OSAK , T. & N EWMAN , D. K. 2005. Microbial kinetic controls on calcite morphology in supersaturated solutions. Journal of Sedimentary Research, 75, 190– 199. B OURRELLY , P. 1970. Les algues d’Eau Douce. III: Les Algures bleues et rouges, les Eugle´niens, Peridiniens et Cryptomonadines. N. Boube´e and Cie, Paris, France. B OURRELLY , P. & D UPUY , P. 1973. Quelques stations franc¸aises de Geitleria calcarea, Cyanophyce´e cavernicole. Revue Suisse d’Hydrologie, Schweizerische Zeitschrift fu¨r Hydrologie, 35, 136 –140. B ROUGHTON , P. L. 1983a. Environmental implications of competitive growth fabrics in stalactitic carbonate. International Journal of Speleology, 13, 31–41. B ROUGHTON , P. L. 1983b. Lattice deformation and curvature in stalactitic carbonate. International Journal of Speleology, 13, 19–30. B ROUGHTON , P. L. 1983c. Secondary origin of the radial fabric in stalactitic carbonate. International Journal of Speleology, 13, 43–66. C ACCHIO , P., C ONTENTO , R., E RCOLE , C., C APPUCCIO , G., M ARTINEZ , M. P. & L EPIDI , A. 2004. Involvement of microorganisms in the formation of carbonate speleothems in the Cervo Cave (L’Aquila-Italy). Geomicrobiological Journal, 21, 497–509. C ACCHIO , P., E RCOLE , C., C APPUCCIO , G. & L EPIDI , A. 2003. Calcium carbonate precipitation by bacterial strains isolated from a limestone cave and from a loamy soil. Geomicrobiological Journal, 20, 85– 98. C AN˜ AVERAS , J. C., C UEZVA , S., S ANCHEZ -M ORAL , S., L ARIO , J., L AIZ , L., G ONZA´ LEZ , J. M. & S AIZ J IMENEZ , C. 2006. On the origin of fiber calcite crystals in moonmilk deposits. Naturwissenschaften. 93, 27– 32. C AN˜ AVERAS , J. C., H OYOS , M. ET AL . 1999. Microbial communities associated with hydromagnesite and needle-fiber aragonite deposits in a karstic cave (Altamira, Northern Spain). Geomicrobiological Journal, 16, 9 –25. C AN˜ AVERAS , J. C., S ANCHEZ -M ORAL , S., S OLER , V. & S AIZ -J IMENEZ , C. 2001. Microorganisms and microbially induced fabrics in cave walls. Geomicrobiological Journal, 18, 223–240. C ASTANIER , S., L E M E´ TAYER -L EVREL , G. & P ERTHUISOT , J.-P. 1999. Ca-carbonates precipitation and limestone genesis – the microbiogeologist point of view. Sedimentary Geology, 126, 9 –23.
C AUMARTIN , V. 1963. Review of the microbiology of underground environments. National Speleological Society Bulletin, 25, 1 –14. C HALMIN , E., S ANSOT , E., O RIAL , G., B OUSTA , F. & R EICHE , I. 2008. Microanalysis and synthesis of calcite. Growth mechanisms on prehistoric paintings in the Large Cave, Arcy-sur-Cure (Yonne, France). X-Ray Spectrometry, 37, 424– 434. C HELIUS , M. K. & M OORE , J. C. 2004. Molecular phylogenetic analysis of archaea and bacteria in Wind Cave, South Dakota. Geomicrobiological Journal, 21, 123– 134. C HOPPY , J. 1991. Perles des cavernes Cubiques et Polyedriques. Al-Ouat’Ouate (Revue Libanaise de Spe´le´ologie et de Karstologie, 6, 43–47. C LAUS , G. 1955. Algae and their mode of life in the Baradla Cave at Aggtelek. Acta Academiae Scientiarum Hungaricae, 2, 1– 26. C OLEMAN , J. C. 1949. Irish cave pearls. Proceedings, University of Bristol Speleology Society, 6, 68– 71. C ONTOS , A. K., J AMES , J. M., H EYWOOD , B., P ITT , K. & R OGERS , P. 2001. Morphoanalysis of bacterially precipitated subaqueous calcium carbonate from Weebubbie Cave, Australia. Geomicrobiological Journal, 18, 331–343. C OSTERTON , J. W. 2000. Biofilms in the new millenium: musings from a peak in Xanadu. In: A LLISON , D. G., G ILBERT , P., L APPIN -S COTT , H. M. & W ILSON , M. (eds) Community Structure and Co-operation in Biofilms. Cambridge University Press, Cambridge, 330–344. C OUTE´ , A. 1982. Ultrastructure d’une Cyanophyce´e ae´rienne calcifie´e cavernicole: Geitleria calcarea Firedman. Hydrobiologia, 97, 255– 274. C OX , G. 1977. A ‘living fossil’ in the twilight zone: a cavewall bacterium of unique ultrastructure. In: F ORD , T. D. (ed.) Proceedings of the 7th International Speleological Congress, Sheffield. British Cave Research Association, Bridgewater, UK, 155–156. C OX , G., J AMES , J. M., L EGGETT , K. E. A. & O SBORNE , R. A. L. 1989a. Cyanobacterially deposited speleothems: subaerial stromatolites. Geomicrobiological Journal, 7, 245–252. C OX , G., J AMES , J. M., L EGGETT , K. E. A. & O SBORNE , R. A. L. 1989b. Stromatolitic crayfish-like stalagmites. Proceedings of the Bristol Speleological Society, 18, 339–358. C OX , G., J AMES , J. M., O SBORNE , R. A. L. & L EGGETT , K. E. A. 2001. Morphoanalysis of bacterially precipitated subaqueous calcium carbonate from Weebubbie Cave, Australia. Geomicrobiological Journal, 18, 331–343. C OX , G. & M ARCHANT , H. 1977. Photosynthesis in the deep twilight zone: micro-organisms with extreme structural adaptations to low light. In: F ORD , T. D. (ed.) Proceedings of the 7th International Speleological Congress, Sheffield. British Cave Research Association, Bridgewater, UK, 131– 133. C OX , G., S ALIH , A., J AMES , J. M. & A LLAWAY , B. 1995. Confocal microscopy of cyanobacteria in calcite speleothems. Zoological Studies, 34, 5– 6. C UNNINGHAM , K. I., N ORTHUP , D. E., P OLLASTRO , R. M., W RIGHT , W. G. & L A R OCK , E. J. 1995. Bacteria, fungi and biokarst in Lechuguilla Cave, Carlsbad Caverns National Park, New Mexico. Environmental Geology, 25, 2 –8.
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION D ANIELLI , H. M. C. & E DINGTON , M. A. 1983. Bacterial calcification in limestone caves. Geomicrobiology Journal, 3, 1 –16. D AVIES , D. G., P ALMER , M. V. & P ALMER , A. N. 1990. Extraordinary subaqueous speleothems in Lechuguilla Cave, New Mexico. National Speleological Society Bulletin, 52, 70–86. D AVIS , J. S. & R ANDS , D. G. 1981. The genus Geitleria (Cyanophyceae) in a Bahamian cave. Schweizerische Zeitschrift fu¨r Hydrologie (Swiss Journal of Hydrology), 43, 63–68. D AVIS , J. S. & R ANDS , D. G. 1982. Lime incrusting Hapalosiphon intricatus (Cyanophyceae) and phosphate availability in a Florida cave. Schweizerische Zeitschrift fu¨r Hydrologie, 44, 289–294. D E W INDER , B. D., M ATTHIJS , H. C. P. & M UR , L. R. 1989. The role of water retaining substrate on the photsynthetic response to three drought tolerant phototropic microorganisms isolated from a terrestrial habitat. Archives fu¨r Mikrobiologie, 152, 458– 462. D ESIKACHARY , T. V. 1959. Cyanophyta. Indian Council of Agriculture Research, New Delhi. D OBAT , K. 1968. Die Pflanzen- und tierwelt der Charlottenho¨le. Abhandlinger Karst-u Ho¨hlenkunde Reihe A, 3, 37– 50. D ONAHUE , J. 1965. Laboratory growth of pisolite grains. Journal of Sedimentary Petrology, 35, 251–256. D ONAHUE , J. 1969. Genesis of oolite and pisolite grains. Journal of Sedimentary Petrology, 39, 360–377. D RAGANOV , S. J. 1977. Taxonomic structure of cave algal flora. In: F ORD , T. D. (ed.) Proceedings of the 7th International Speleological Congress, Sheffield. British Cave Research Association, Bridgewater, UK, 155–156. D ZIADZIO , P., R OZNIAK , R. & S ZULC , J. 1993. Geneza polew naciekowych jaskin psiej i naciecowey W Tatrach zachodnich. Przeglad Geolgiczny, 41, 767–775. E RCOLE , C., C ACCHIO , P., C APPUCCIO , G. & L EPIDI , A. 2001. Deposition of calcium carbonate in karst caves: role of bacteria in Stiffe’s Cave. International Journal of Speleology, 30, 69– 79. E RDMAN , E. 1902. Stalagmit. och pisolitartade bildnigar I Hoganas Stenkolsgrufva, Skane. Geologiska Fo¨reningens i Stockholm Fo¨rhandlingar, 24, 501– 507. F AIRCHILD , I. J., B AKER , A., B ORSATO , A., F RISIA , S., H INTON , R. W., M C D ERMOTT , F. & T OOTH , A. F. 2001. Annual to sub-annual resolution of mulitple trace-element trends in speleothems. Journal of the Geological Society, London, 158, 831– 841. F ORTI , P. 2001. Biogenic spelethems: an overview. International Journal of Speleology, 30, 39–56. F RIEDMANN , E. I. 1979. The genus Geitleria (Cyanophyceae or cyanobacteria): distribution of G. calcarea and G. floridana n. sp. Plant Systematics and Evolution, 131, 169– 178. G ENTY , D. & Q UINIF , Y. 1996. Annually laminated sequences in the internal structure of some Belgian stalagmites – importance for paleoclimatology. Journal of Sedimentary Research, 66, 275 –288. G E` ZE , B., L AGRANGE , R. & P OBE´ GUIN , T. 1956. Sur la nature du reve`tment occasionel des parois ou du sol des grottes (mondmilk). Academie des Sciences [Paris], Comptes Rendus, 242, 144– 145.
27
G E` ZE , B. & P OBE´ GUIN , T. 1956. Contribution a l’etude des concretions carbonate`es. 2eme Congre`s Internationale de Speleologie. Actes, Bari, Italy, 396–414. G IACOBINI , C., DE C ICCO , M. A., T IGLIE , I. & A CCARDO , G. 1987. Actinomycetes and biodeterioration in the field of fine art. Biodeterioration, 7, 418– 423. G OEDE , A., M C C ULLOCH , M. T., M C D ERMOTT , F. & H AWKESWORTH , C. J. 1998. Aeolian contribution to strontium and strontium isotope variaiton in a Tasmanian speleothem. Chemical Geology, 149, 37–50. G OLUBIC , S. 1973. The relationship between blue-green algae and carbonate deposits. In: C ARR , N. B. & W HITTON , B. A. (eds) The Biology of the Blue-Green Algae. Botanical Monographs. Blackwell Scientific Publications, Oxford, 434–474. G RACIA A LONSO , C. A. 1974. Geitleria calcarea Friedmann neuva alga cavernicola para Espan˜a. Speleon, 21, 133– 136. G RADZINSKI , M. 1999. Role of micro-organisms in cave pearls formation. Journal of Conference Abstracts, 24, 924. G RADZINSKI , M. 2001. Role of bacteria in the growth of cave pearls. 13th International Congress of Speleology, 205 –281. G RADZINSKI , M. 2003. Bacterial influence on speleothem oxygen isotope composition: an example based on cave pisoids from Perlova´ cave (Slovakia). Geologica Carpathica, 54, 199–204. G RADZINSKI , M., S ZULC , J. & S MYK , B. 1997. Microbial agents of moonmilk calcification. In: J EANNIN , P. Y. (ed.) Symposium 7: Physical Speleology & Symposium 8: Karst geomorphology (Proceedings of the 12th international congress of speleology). International Union of Speleology, 275– 278. G RADZINSKI , R. & R ADOMSKI , A. 1967. Pisoliths from Cuban caves. Rocznik Polskie Towarzystwo Geologiske, 37, 243– 265. G ROTH , I. & S AIZ -J IMENEZ , C. 1999. Actinomycetes in hypogean environments. Geomicrobiological Journal, 16, 1–8. G ROTH , I., S CHUMANN , P., L AIZ , L., S ANCHEZ -M ORAL , S., C AN˜ AVERAS , J. C. & S AIZ -J IMENEZ , C. 2001. Geomicrobiological study of the Grotta dei Cervi, Porto Badisco, Italy. Geomicrobiological Journal, 18, 241 –258. G UILLITTE , O. & D REESEN , R. 1995. Laboratory chamber studies and petrographical analysis as bioreceptivity assessment tools of building materials. Science of the Total Environment, 167, 365– 374. H ARMON , R. S., A TKINSON , T. C. & A TKINSON , J. L. 1983. The mineralogy of Castleguard Cave, Columbia Icelfields, Alberta, Canada. Arctic and Alpine Research, 15, 503–506. H ASSELBRING , T. S., T ANSEY , M. R. & J ACK , M. A. 1975. Fingi associated with growing stalagmites. Mycologia, 67, 171–172. H ERNA´ NDEZ -M ARINE´ , M., R OLDA´ N M OLINA , M., C LAVERO , A., C ANALS , A. & A RIN˜ O , X. 2001. Phototrophic biofilm morphology in dim light. The case of the Puigmolto´ sinkhole. In: E LSTER , J., S ECKBACH , J., V INCENT , W. P. & L HOTSKY´ , O. (eds) Algae and Extreme Environments. Nova Hedwigia Beiheft, 237– 253.
28
B. JONES
H ESS , F .L. 1930. Oolites or cave pearls in the Carlsbad Caverns. Proceedings of the United States National Museum, 76, 1–5. H ILL , C. A. & F ORTI , P. 1997. Cave Minerals of the World. National Speleological Society, Huntsville, Alabama. H OFFMAN , L. 1989. Algae of terrestrial habitats. Botanical Review, 55, 77– 105. H UANG , Y., F AIRCHILD , I. J., B ORSATO , A., F RISIA , S., C ASSIDY , N. J., M C D ERMOTT , F. & H AWKESWORTH , C. J. 2001. Seasonal variations in Sr, Mg, and P in modern speleothems (Grotta di Ernesto, Italy). Chemical Geology, 175, 429 –448. H ØEG , O. A. 1946. Cyanophyceae and bacteria in calcareous sediments in the interior of limestone caves in Nord-Rana, Norway. Nyatt Magasine for Naturvidenskapene, 85, 99– 104. J AMES , J. M., J ENNINGS , J. N. & D YSON , H. G. 1982. Mineral decoration and weathering of the caves. In: S DYSON , H. G., E LLIS , R. & J AMES , J. M. (eds) Wombeyan Caves. Sydney Speleological Society Paper 8, 121–136. J AMES , J. M., P ATSALIDES , E. & C OX , G. 1994. Amino acid composition of straomatolitic stalagmites. Geomicrobiological Journal, 12, 183 –194. J ONES , B. 1987. The alteration of sparry calcite crystals in a vadose setting, Grand Cayman Island. Canadian Journal of Earth Sciences, 24, 2292–2304. J ONES , B. 1988. The influence of plants and microorganisms on diagenesis in caliche: example from the Pleistocene Ironshore Formation on Cayman Brac, British West Indies. Bulletin of Canadian Petroleum Geology, 36, 191– 201. J ONES , B. 1989. The role of micro-organisms in phytokarst development on dolostones and limestones, Grand Cayman, British West Indies. Canadian Journal of Earth Sciences, 26, 2204–2213. J ONES , B. 1995. Processes associated with microbial biofilms in the twilight zone of caves: examples from the Cayman Islands. Journal of Sedimentary Research, A65, 552–560. J ONES , B. 2001. Microbial activity in caves - A geological perspective. Geomicrobiology Journal, 18, 1– 13. J ONES , B. 2009. Mineralized actinomycetes and phosphates in speleothems from Grand Cayman, British West Indies. Sedimentary Geology, 219, 302– 317. J ONES , B. & K AHLE , C. F. 1986. Dendritic calcite crystals formed by calcification of algal filaments in a vadose environment. Journal of Sedimentary Petrology, 56, 217– 227. J ONES , B. & K AHLE , C. F. 1993. Morphology, relationship, and origin of fiber and dendrite calcite crystals. Journal of Sedimentary Petrology, 63, 1018–1031. J ONES , B. & K AHLE , C. F. 1995. Origin of endogenetic micrite in karst terrains: a case study from the Cayman Islands. Journal of Sedimentary Research, 65, 283–293. J ONES , B. & M AC D ONALD , R. W. 1989. Microorganisms and crystal fabrics in cave pisoliths from Grand Cayman, British West Indies. Journal of Sedimentary Petrology, 59, 387– 396. J ONES , B. & M OTYKA , A. 1987. Biogenic structures and micrite in stalactites from Grand Cayman Island, British West Indies. Canadian Journal of Earth Sciences, 24, 1402– 1411.
J ONES , B. & P EMBERTON , S. G. 1987a. Experimental formation of spiky calcite through organically mediated dissolution. Journal of Sedimentary Petrology, 57, 687–694. J ONES , B. & P EMBERTON , S. G. 1987b. The role of fungi in the diagenetic alteration of spar calcite. Canadian Journal of Earth Sciences, 24, 903–914. J ONES , B. & R ENAUT , R. W. 2007. Selective mineralization of microbes in Fe-rich precipitates (jarosite, hydrous ferric oxides) from acid hot springs in the Waiotapu geothermal area, North Island, New Zealand. Sedimentary Geology, 194, 77– 98. J ONES , B., R ENAUT , R. W. & R OSEN , M. R. 1997. Biogenicity of silica precipitation around geysers and hot-spring vents, North Island, New Zealand. Journal of Sedimentary Research, 67, 88–104. J ONES , B., R ENAUT , R. W. & R OSEN , M. R. 2001. Taphonomy of silicified filamentous microbes in modern geothermal sinters– implications for identification. Palaios, 16, 580–592. J ONES , B., R ENAUT , R. W. & R OSEN , M. R. 2004. Taxonomic fidelity of silicified microbes from hot spring systems in the Taupo Volcanic Zone, North Island, New Zealand. Transactions of the Royal Society of Edinburgh: Earth Sciences, 94, 475– 483. J ONES , H. J. 1965. Algological investigations in Mammoth Cave, Kentucky. Speleology, 1, 491–516. K AHLE , C. F. 1977. Origin of subaerial Holocene calcareous crusts: role of algae, fungi and sparmicritization. Sedimentology, 24, 413–435. K ARKABI , S. 1991. La perle de Caverne Hexagonale. Al-Ouat’Ouate (Revue Libanaise de Spe´le´ologie et de Karstologie, 6, 48– 53. K ATSIFARA , A. & S PANOS , N. 1999. effect of inorganic phosphate ions on spontaneous precipitation of vaterite and the transformtion of vaterite to calcite. Journal of Crystal Growth, 204, 183–190. K ELLER , W. D. 1937. “Cave Pearls” in a cave near Columbia, Missouri. Journal of Sedimentary Petrology, 7, 108– 109. K ENDALL , A. C. & B ROUGHTON , P. L. 1978. Origin of fabrics in speleothems composed of columnar calcite crystals. Journal of Sedimentary Petrology, 48, 519–538. K IRCHMAYER , M. 1969. Crystallizations and recrystallizations in cave pearls from mines. Østerr Akademie der Wissenschaften Mathem, 8, 233 –245. K IRCHMAYER , M. 1987. Universal stage grain fabrics statistics on dendrites of mine cave pearls. Carbonates and Evaporites, 2, 59–65. K ONHAUSER , K. O., J ONES , B., P HOENIX , V. R., F ERRIS , F. G. & R ENAUT , R. W. 2004. The microbial role in hot spring silicification. Ambio, 33, 546– 552. K OWALINSKI , S., P ONS , L. J. & S LAGER , S. 1972. Micromorphological comparison of three soils derived from loess in different climatic regions. Geoderma, 7, 141–158. K RISCHTAFOWITSCH , N. 1906. Bibliographische notix zu dem Artikel von Herrn L. L. Iwanoff “Ein wasserhaltiges calciumcarbonat sua den umgebungen von Nove-Alexandria, Gouv. Lublin”. Annuaire Ge´ologie et Mine´ralogie de la Russie, 8, 124–125. K RUMBEIN , W. E. 1979. Calcification by bacteria and algae. In: T RUDINGER , P. A. & S WAINE , D. J. (eds) Biogeochemical Cycling of Mineral-forming Elements.
MICROBES IN CAVES: AGENTS OF CALCITE CORROSION AND PRECIPITATION Studies in Environmental Science 3. Elsevier Scientific Publishing Company, Amsterdam, 47– 68. L AIZ , L., G ROTH , I., S CHUMANN , P., Z EZZA , F., F ELSKE , A., H ERMOSIN , B. & S AIZ -J IMENEZ , C. 2000. Microbiology of the stalactites from Grotta dei Cervi, Porto Badisco, Italy. International Microbiology, 3, 25–30. L AURITZEN , S.-E. & L UNDBERG , J. 1999. Calibration of the speleothem delta function: an absolute temperature record for the Holocene in northern Norway. The Holocene, 9, 659–669. L ECLERQ , J. C., C OUTE´ , A. & D UPUY , P. 1983. Le climat annuel de deux grottes et d’une e´glise du Poitou, ou` vivent des colonies pures d’algues sciaphiles. Cryptogamie, Algologie, 4, 1– 19. L EE , S. D. 2006. Nocardia jejuensis sp. nov., a novel actinomycete isolated from a natural cave on Jeju Island, Republic of Korea. International Journal of Systematic and Evolutionary Microbiology, 56, 559–562. L EFEVRE , M. & L APORTE , G. S. 1969. The “Maladie Verte” of Lascaux. Studies in Speleology, 2, 34–44. L E´ VEILLE´ , R. J., F YFE , W. S. & L ONGSTAFFE , F. J. 2000. Geomicrobiology of carbonate–silicate microbialites from Hawaiian basaltic sea caves. Chemical Geology, 169, 339– 355. L E´ VEILLE´ , R. J., L ONGSTAFFE , F. J. & F YFE , W. S. 2007. An isotopic and geochemical study of carbonate-clay mineralization in basaltic caves: abiotic versus microbial processes. Geobiology, 5, 235–249. L IE´ GEOIS , P. G. 1956. A propos des perles de cavernes et concre´tions analogues non encore de´crites. Annales de la Socie´te´ Ge´ologique de Belgique, 80, 165– 169. L OWENSTAM , H. A. 1981. Minerals formed by organisms. Science, 211, 1126–1131. M ANOLI , F. & D ALAS , E. 2000. Spontaneous precipitation of calcium carbonate in the presence of chondroitin sulfate. Journal of Crystal Growth, 217, 416–421. M ARTI´ NEZ -A RKARAZO , I., A NGULO , M., Z ULOAGA , O., U SOBIAGA , A. & M ADARIAGA , J. M. 2007. Spectroscopic characterization of moonmilk deposits in Pozalagua tourist cave (Karrantza, Basque Country, North of Spain). Spectrochimica Acta Part A, 68, 1058– 1064. M ASON , H. E., F RISIA , S., T ANG , Y., R EEEDER , R. J. & P HILLIPS , B. L. 2007. Phosphorus speciation in calcite speleothems determined from solid-state NMR spectroscopy. Earth and Planetary Science Letters, 254, 313– 322. M ASON -W ILLIAMS , A. 1959. The formation and deposition of moon-milk. Cave Research Group, Transactions, Great Britain, 10, 135– 138. M ASON -W ILLIAMS , M. A. 1967. Further investigations into bacterial and algal populations of caves in South Wales. Speleology, II, 389–395. M C D ERMOTT , F., F RISIA , S. ET AL . 1999. Holocene climate variability in Europe, evidence from d18O and textural variations in speleothems. Quaternary Science Reviews, 18, 1021–1038. M ELIM , L. A., N ORTHUP , D. E., S PILDE , M. N., J ONES , B., B OSTON , P. J. & B IXBY , R. J. 2008. Reticulated filaments in cave pool speleothems: microbe or mineral? Journal for Cave and Karst Studies, 70(3), 135–141.
29
M ELIM , L. A., S HINGLMAN , K. M., B OSTON , P. J., N ORTHUP , D. E., S PILDE , M. N. & Q UEEN , J. M. 2001. Evidence for microbial involvement in pool finger precipitation, Hidden Cave, New Mexico. Geomicrobiological Journal, 18, 311– 329. M ERZ , M. U. E. 1992. The biology of carbonate precipitation by Cyanobacteria. Facies, 26, 81– 101. M ONTE , M. & F ERRARI , R. 1993. Biodeterioration in subterranean environments. Aerologica, 9, 141–148. M OORE , G. W. & B UKRY , D. 1968. Electronmicrograph of moonmilk from caverns of Sonora, Texas. National Speleological Society News, 26, 122. M OROZEWICZ , J. 1907. Lublinit (Przyczynki do znajomosci weglanu wapniowego). Kosmos, 32, 487–495. ¨ ber lublinit, eine neue Varita¨t M OROZEWICZ , J. 1911. U des kalkspates. Centralblatt fur Mineralogie, Geologie und Pala¨ontologie, 229– 230. M UEGGE , O. 1914. Uver die Lublinit genannte angeblich neue modifikation des kohlensauren kalkes. Centralblatt fu¨r Mineralogie, Geologie und Pala¨ontologie, 673– 675. M ULEC , J. 2008. Microorganisms in hypogeon: examples from Slovenian karst caves. Acta Carsologica, 37, 153– 160. M ULEC , J., K OSI , G. & V RHOVSEK , D. 2007. Algae promote growth of stalagmites and stalactites in karst caves (Skocjanske jame, Slovenia). Carbonates and Evaporites, 22, 6 –9. M ULEC , J., Z ALAR , P., H AJNA , N. Z. & R UPNIK , M. 2002. Screening for cultural microorganisms from cave environments (Slovenia). Acta Carsologica, 8, 177– 187. N ADER , F. H. 2007. Petrographic and geochemical study on cave pearls from Kanaan Cave (Lebanon). International Journal of Speleology, 36, 39–50. N AGY , J. P. 1965. Preliminary notes on the algae of Crystal Cave, Kentucky. Speleology, 1, 479– 490. N ORTHUP , D. E. & L AVOIE , K. H. 2001. Geomicrobiology of caves: a review. Geomicrobiological Journal, 18, 199 –222. O NAC , B. P. & G HERGARI , L. 1993. Moonmilk mineralogy in some Romanian and Nowegian caves. Cave Science, 20, 107– 111. P ALIK , P. 1960. Study into the algal flora of caves. Hidrologiani Kozlony, 5, 417–422. P HOENIX , V. R., A DAMS , D. G. & K ONHAUSER , K. O. 2000. Cyanobacterial viability during hydrothermal biomineralisation. Chemical Geology, 169, 329– 338. P HOENIX , V. R., K ONHAUSER , K. O., A DAMS , D. G. & B OTTRELL , S. H. 2001. Role of biomineralization as an ultrviolet shield: implications for Archean life. Geology, 29, 823–826. P OLYAK , V. J. & C OKENDOLPHER , J. C. 1992. Recovery of microfossils from carbonate speleothems. National Speleological Society Bulletin, 54, 66– 68. P OND , A. W. 1945. Calcite oolites or “cave pearls” found the “Cave of the Mounds”. Journal of Sedimentary Petrology, 15, 55–58. P ROCTOR , C. J., B AKER , A., B ARNES , W. L. & G ILMOUR , M. A. 2000. A thousand year speleothem proxy record of North Atlantic climate from Scotland. Climate Dynamics, 16, 815– 820. Q IN , X., T AM , M., L IU , T., W ANG , X., L I , T. & L U , J. 1999. Spectral analysis of 1 100-year stalagmite
30
B. JONES
lamina-thickness record from Shihua Cavern, Beijing, China, and its climatic significance. The Holocene, 9, 689– 694. R AUTARAY , D., A HMAD , A. & S ASTRY , M. 2003. Biosynthesis of CaCO3 crystals of complex morphology using a fungus and an actinomycete. Journal of the American Chemical Society, 125, 14656– 14657. R AUTARAY , D., A HMAD , A. & S ASTRY , M. 2004. Biological synthesis of metal carbonate minerals using fungi and actinomycetes. Journal of Materials Chemistry, 14, 2333– 2340. R IDING , R. 1977. Calcified Plectonema (blue-green algae), a recent example of Girvanella from Aldabra Atoll. Palaeontology, 20, 33– 46. R OBERTS , M. S., S MART , P. L. & B AKER , A. 1998. Annual trace element variations in a Holocene speleothem. Earth and Planetary Science Letters, 154, 237– 246. R OLDA´ N , M., C LAVERO , E., C ANALS , T., G O´ MEZ B OLEA , A., A RIN˜ O , X. & H ERNA´ NDEZ -M ARINE´ , M. 2004. Distribution of phototrophic biofilms in cavities (Garraf, Spain). Nova Hedwigia, 78, 329 –351. S ANCHEZ -M ORAL , S., C AN˜ AVERAS , J. C., L AIZ , L., S AIZ -J IMENEZ , C., B EDOYA , J. & L UQUE , L. 2003. Biomediated precipitation of calcium carbonate metastable phases in hypogean environments: a short review. Geomicrobiological Journal, 20, 491– 500. S ASOWSKY , I. D. & P ALMER , M. V. 1994. Breakthroughs in karst geomicrobiology and redox geochemistry: Abstracts and field-trip guide for the symposium geld February 16 through 19, 1994, Colorado Springs, Colorado. Special Publication No. 1, 1 –111. Karst Waters Institute, Inc., Charles Town, West Virginia. S CHOPF , J. W. 1999. Fossils and pseudofossils: lessons from the hunt for early life on Earth. Space Studies Board, National Research Council, 88– 93. S HUMENKO , S. I. & O LIMPIEV , I. V. 1977. Rock milk from caves in Crimea and Abhasia. Lithology and Mineral Resources, 12, 240–243.
S TONE , R. W. 1932. Cave concretions. Pennsylvanian Academy of Science, 6, 106–109. S TOOPS , G. J. 1976. On the nature of lublinite from Hollanta (Turkey). American Mineralogist, 61, 172. T ABOROSˇ I , D. 2008. Biologically influenced carboante speleothems. In: H ARMON , R. S. & W ICKS , C. (eds) Perspectives on Karst Geomorphology, Hydrology, and Geochemistry – A Tribute to Derek C. Ford and William B. White. Geological Society of America Special Paper, 404, 307–317. T HRAILKILL , J. V. 1963. Moonmilk, cave pearls and pool accretions from Fulford Cave, Colorado. National Speleological Society Bulletin, 25, 88–89. T REBLE , P. C., C HAPPELL , J. & S HELLEY , J. M. G. 2005. Complex speleothem growth processes revealed by trace element mapping and scanning electron microscopy of annual layers. Geochimica et Cosmochimica Acta, 69, 4855–4863. T REBLE , P. C., S HELLEY , J. M. G. & C HAPPELL , J. 2003. Comparison of high resolution sub-annual records of trace elements in a modern (1911–1992) speleothem with instrumental climate data from southwest Australia. Earth and Planetary Science Letters, 216, 141–153. U LRICH , F. 1938. Houby jako rozrusovatele a tvurci nerosbi a hornin. Verda Prorodni, 19, 45– 50. V ERHEYDEN , S., K EPPENS , E., F AIRCHILD , I. J., M C D ERMOTT , F. & W EIS , D. 2000. Mg, Sr and Sr isotope geochemistry of a Belgian Holocene speleothem: implications for paleoclimate reconstructions. Chemical Geology, 169, 131–144. V ERRECCHIA , E. P. & V ERRECCHIA , K. E. 1994. Needlefiber calcite: a critical review and a proposed classification. Journal of Sedimentary Research, A64, 650–664. W ENT , F. W. 1969. Fungi associated with stalactite growth. Science, 166, 386–387. W HITE , W. B. 1976. Cave minerals and speleothems. In: F ORD , D. T. & C ULLINGFORD , C. H. (eds) The Science of Speleology. Academic Press, London, 267–327.
Bacterial biomineralization: new insights from Myxococcus-induced mineral precipitation ´ LEZ-MUN ˜ OZ1, CARLOS RODRIGUEZ-NAVARRO2*, MARIA TERESA GONZA 3 ´ FRANCISCA MARTINEZ-RUIZ , JOSE MARIA ARIAS1, MOHAMED L. MERROUN4 & MANUEL RODRIGUEZ-GALLEGO2 1
Departamento de Microbiologı´a, Universidad de Granada, Fuentenueva s/n, 18002, Granada, Spain
2
Departamento de Mineralogı´a y Petrologı´a, Universidad de Granada, Fuentenueva s/n, 18002, Granada, Spain
3
Instituto Andaluz de Ciencias de la Tierra, CSIC – Universidad de Granada, Fuentenueva s/n, 18002, Granada, Spain 4
Institute of Radiochemistry, Forschungszentrum Dresden-Rossendorf, D – 01314, Dresden, Germany; Present address: Departamento de Microbiologı´a, Universidad de Granada, Granada, Spain *Corresponding author (e-mail:
[email protected]) Abstract: Bacteria have contributed to the formation of minerals since the advent of life on Earth. Bacterial biomineralization plays a critical role on biogeochemical cycles and has important technological and environmental applications. Despite the numerous efforts to better understand how bacteria induce/mediate or control mineralization, our current knowledge is far from complete. Considering that the number of recent publications on bacterial biomineralization has been overwhelming, here we attempt to show the importance of bacteria– mineral interactions by focusing in a single bacterial genus, Myxococcus, which displays an unusual capacity of producing minerals of varying compositions and morphologies. First, an overview of the recent history of bacterial mineralization, the most common bacteriogenic minerals and current models on bacterial biomineralization is presented. Afterwards a description of myxobacteria is presented, followed by a section where Myxococcus-induced precipitation of a number of phosphates, carbonates, sulphates, chlorides, oxalates and silicates is described and discussed in lieu of the information presented in the first part. As concluding remarks, implications of bacterial mineralization and perspectives for future research are outlined. This review strives to show that the mechanisms which control bacterial biomineralization are not mineral- or bacterial-specific. On the contrary, they appear to be universal and depend on the environment in which bacteria dwell.
Microbes are considered the oldest living creatures on Earth (Schopf 1993). Although there is still controversy as to when, how and where they first appeared (between 3550–3800 Ma) (Mojzsis et al. 1996; Altermann & Kazmierczak 2003; Garcı´aRuiz et al. 2003), there is compelling evidence that microbes have been closely related to the solid phases they were forced to live with over the last c. 3.5 billion years: that is, minerals (Ehrlich 2002). That long-term relationship has been extraordinarily productive. It has contributed to the shaping of the Earth surface (and near-subsurface) sediments and rocks, and to the evolution of the chemistry and composition of the oceans (Banfield et al. 1998) and of the Earth atmosphere (Kasting & Siefert 2002; Wiechert 2002). It is believed for
instance that cyanobacteria have aided in the formation of mineralized stromatolites since the Precambrian (Arp et al. 2001; Bosak & Newman 2003), their photosynthetic activity thus contributing to atmosphere oxygen increase and CO2 budgeting by carbonate precipitation since c. 2.15 billion-years-ago (Buick 1992). Fe(II)-oxidizing bacteria may have been responsible for the formation of Precambrian iron formations (Konhauser et al. 2002) and may have had an important impact on ancient metal cycling. Microbes are assumed to have played a major role on evolution during snowball Earth episodes when it is suspected that only extremophiles could survive and repopulate the planet (Kerr 1997). Examples of the role of bacteria in Earth History are abundant since they have been
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 31– 50. DOI: 10.1144/SP336.3 0305-8719/10/$15.00 # The Geological Society of London 2010.
32
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
and are at present biotic fundamental components of natural biogeochemical cycles (Newman & Banfield 2002). In this context bacterially mediated biomineralization is crucial for the complex interactions of biological, chemical and physical processes. Discovery of the biosphere in deep subseafloor has shown that microbial communities are active in sediments of several million years of age and up to several hundred metres in depth. It has also been confirmed that the deep biosphere is significantly involved in the global cycling of elements and represents a major reservoir of organic carbon (e.g. Pedersen 2000). On the other hand, the technological and environmental applications of bacterial mineralization are now being demonstrated to be far reaching (McIntosh & Groat 1997). For instance, the use of biogenic minerals for the sequestration of radionuclides and heavy metals in contaminated sediments and groundwater offers considerable potential for environmental cleanup (Bhagat et al. 2004; Haferburg et al. 2008). Long-lived radionuclides neptunium and plutonium have been removed from contaminated solutions using uranyl phosphatecoated bacteria, with likely exchange of neptunyl and plutonyl ions for uranyl ions within the crystal lattice (Macaskie & Baskanova 1998). The efficiency of these microbiological-based strategies is enhanced by the high capacity of biogenic minerals to remove heavy metal if compared with those produced abiotically. For instance, biogenic Mn oxides have been shown to display a larger metal binding capacity than well-crystallized synthetic Mn oxides (Nelson et al. 1999). Bacterial mineralization has also been recently proposed as a method for the conservation of ornamental stone (Castanier et al. 2000; Rodriguez-Navarro et al. 2003; Webster & May 2006; Gonza´lez-Mun˜oz 2008). Due to the ubiquity and importance of bacteriamineral interactions, the number of publications on this topic has grown steadily over the last decades. Here, we will therefore focus on studies of biomineralization by a microorganism, Myxococcus. We have chosen to study these bacteria because they display an unusual capacity for producing mineral precipitates of varying compositions and morphologies. The following section presents an overview of the recent history of bacterial mineralization and briefly describes which minerals are most commonly precipitated in the presence of bacteria. Current models on bacterial precipitation of minerals are also described. The third section presents a description of myxobacteria. Myxococcus-induced precipitation of a number of phosphates, carbonates, sulphates, chlorides, oxalates, and silicates is presented and discussed in the fourth section. Finally, implications of bacterial mineralization and perspectives for future research are outlined.
The significance of bacterial mineralization Microorganisms have been recognized to influence the formation of a wide variety of minerals such as carbonates, oxides, sulphides, phosphates, sulphates, nitrates, halides and silicates (Ferris et al. 1986; Thompson & Ferris 1990; Schultze-Lam et al. 1996; Fortin et al. 1997; Po´sfai et al. 1998; Castanier et al. 1999; Labrenz et al. 2000). Although it is now beyond doubt that bacterial activity and mineral formation are closely related in a range of environments, the complex interplay between biotic and abiotic processes in mineral precipitation is only beginning to be appreciated (Siering 1998). Current knowledge on how bacteria induce or mediate biomineralization is far from complete and the underlying causes of bacterial mineralization are a matter of controversy. For instance, some researchers believe that there is a genetic control on bacterial mineral precipitation (Barabesi et al. 2007), while others claim that bacteria induce/mediate mineral precipitation simply as a result of their metabolic activity without any kind of direct genetic control (Knorre & Krumbein 2000; Rodriguez-Navarro et al. 2007). As we will see below, there is growing evidence showing that the second line of thought could be correct. The implications of microbial mineralization are far reaching. Microbial mineralization may help disclose how and when living organisms first appeared on Earth and how they evolved. This is particularly relevant for our growing understanding of the interplay between geochemical and biological processes. In addition to the ability of microbes to precipitate minerals within, on or around cells, they dramatically affect the speciation and distribution of ions through redox reactions, through release of organic and inorganic by-products and by directly or indirectly changing the rates or mechanisms of mineral weathering (Banfield et al. 1998). As a result, microbes are co-responsible for the co-evolution of Earth surface and near surface environments, including Earth climate and the influence of this climate on the development of higher life forms (Schwartzman & Volk 1989). The science of microbiologically controlled or mediated geological processes, known as geomicrobiology, has undergone substantial growth in recent years (Ehrlich 2002). Early reviews on microbial mineralization can be found in Ehrlich (2002), Banfield & Nealson (1997) and McIntosh & Groat (1997). Despite the amount of work done on the precipitation of a single mineral by a number of microbes, the production of many mineral phases by a single microorganism has not yet been fully explored. The latter is quite relevant if we are to understand the ultimate mechanisms responsible
BACTERIAL BIOMINERALIZATION
for bacterially mediated/induced mineralization. This review will attempt to show that, with the exception of bacterially controlled mineralization by magnetobacteria (Bazylinsky & Moskowitz 1997), the mechanisms which control bacterial biomineralization are not mineral- or bacteriumspecific, but rather appear to be universal: they seem to depend on the environment in which bacteria dwell and not on any specific type of bacteria. It is believed that this can be evidenced by evaluating how Myxococcus activity induces mineral precipitation.
Types of minerals produced by bacterially induced or mediated processes Several pioneering works on the role of bacteria in mineral precipitation were published during the 19th century (see Ehrlich 2002 and references therein). However, it was not until the early 20th century that the study of bacterial mineralization gained momentum. Drew (1914) demonstrated that bacteria isolated from natural marine waters were able to precipitate calcium carbonate. This type of research prompted the systematic study of possible associations between bacteria and mineral precipitation. Since then, it has been shown that different kinds of bacteria produce mineral precipitates in both laboratory and natural environments (Banfield & Nealson 1997; McIntosh & Groat 1997; Ehrlich 2002). A substantial amount of research has focused on the precipitation of carbonate minerals (Boquet et al. 1973; Buczynski & Chafetz 1991; Knorre & Krumbein 2000; Rodriguez-Navarro et al. 2003; Ben Chekroun et al. 2004). Bacterial precipitation of calcium carbonate polymorphs (calcite, aragonite and vaterite) (Ben Chekroun et al. 2004; RodriguezNavarro et al. 2007) has become an attractive research topic because it has important implications in past and present formation of carbonate rocks and sediments (Krumbein 1979; Buczynski & Chafetz 1991; Braissant et al. 2003; Dupraz & Visscher 2005). Besides its geological significance, bacterial carbonate precipitation in terrestrial environments appears to be crucial for atmospheric CO2 budgeting (Braissant et al. 2002). While early work (until the 1980s) examined the ability of bacteria to precipitate calcium carbonates in marine environments, precipitation of carbonates in other environments such as lakes, travertines, caves, soils and monuments, has become the subject of much research in recent decades (Boquet et al. 1973; Chafetz & Folk 1984; Folk 1993; Urzi et al. 1999; Blyth & Frisia 2008). The study of dolomite [CaMg(CO3)2] precipitation in the presence of bacteria (Vasconcelos et al. 1995) is helping to solve the long standing
33
controversy on the origins of natural dolomite (the so-called ‘dolomite problem’). The suggestion that carbonates in Martian meteorite ALH84001 could be bacterial in origin (McKay et al. 1996) has prompted much debate. These Martian carbonate microfossils were assumed to be precipitated by dwarf bacteria, the so-called ‘nanobacteria’ (Folk 1993), the existence of which has been strongly disputed (e.g. Kirkland et al. 1999; Southan & Donald 1999). For instance, Aloisi et al. (2006) have shown that such calcified ‘nanobacteria’ are in fact nanoglobules originated from bacterial cell surface that act as calcium carbonate nucleation sites once released into the culture medium. In general, recognition of microbial carbonates in nature is controversial (Riding 2000). Iron oxides and hydroxides have been observed to precipitate within, on and outside microbial cells (Konhauser 1998). Magnetite (Fe3O4) appeared to form the striking precipitates lined in magnetosomes of magnetotactic bacteria (Bazylinski & Moskowitz 1997). Komeili et al. (2004) have shown that a magnetosome-associated protein, MamA, is required for the formation of functional magnetosomes and the growth of magnetite in magnetotactic bacteria. These works proved the intimate relationship between cell biology and bacterially controlled mineralization. Iron oxyhydroxides such as ferrihydrite (Fe2O3.0.5H2O) precipitate on bacterial cells (Casanova et al. 1999; Banfield et al. 2000). Chan et al. (2004) reported bacterial precipitation of akaganeite (FeOOH) pseudo-single crystals with unusual 1000:1 aspect ratios. They concluded that the bacterial cells extruded the polysaccharide strands to localize FeOOH precipitation in proximity to the cell membrane. Such work has opened new ways for engineering novel materials. Ferris et al. (1988) concluded that iron binding to and silica precipitation on bacterial cells was an important contributing factor to the fossilization of microbes. It should be indicated that the Fe(II) –Fe(III) redox cycle represents a major energy flux at the Earth surface (Lower et al. 2001). Schewanella is known to be an active element in this cycle by contributing to Fe reduction and the precipitation of iron oxides (Lower et al. 2001). Bacterial mineralization of other metal oxides such as Mn, has also been reported (Fortin et al. 1998; Northup & Lavoie 2001). For instance, several studies have proposed microbial participation in the formation of cave manganese deposits that include pyrolusite (MnO2), romanechite (Ba0.7Mn4.8Si0.1O10.1.2H2O), todoro3þ . kite (Na0.2Ca0.05K0.02Mn4þ 4 Mn2 O12 3H2O), and bir4þ messite (Na0.3Ca0.1K0.1Mn Mn3þO4.1.5H2O) (see review by Northup & Lavoie 2001). Bacterially induced or mediated precipitation of iron and manganese oxides is commonly associated with Siderocapsa, Gallionella, Leptothrix,
34
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
Sphaerotilus, Crenothrix and Clonothrix (Little et al. 1997). Environmental scanning electron microscopy (ESEM) has enabled the study of iron oxides and hydroxides production by Gallionella, in situ, at high magnification (Hallberg & Ferris 2004). This work demonstrates the enormous potential of such a tool for studying bacterial biomineralization processes. It has recently been found that Pyrolobus fumarii, an Fe(III)-reducing bacterium which is responsible for magnetite precipitation in hydrothermal vents, is able to survive at temperatures of around 120 8C (Kashefi & Lovley 2003). This study extends the upper temperature limit for life. Note that mineralization induced by extremophiles, that is, microbes that live in extreme environments (e.g. with very high or very low pH, high pressure, high temperature, high radiation, high salinity), may help us to understand how and where life appeared on Earth, or if life is possible on other planets (Rothschild & Mancinelli 2001). Iron sulphides such as pyrrhotite (Fe12xS), greigite (Fe3S4) and mackinawite (Fe9S8) also form in magnetosomes of magnetotactic bacteria (see review by Bazylinski & Moskowitz 1997, and references therein). Extracellular precipitation of iron and nickel sulphides has also been reported (Ferris et al. 1987). A striking case of bacterial sulphide mineralization is the occurrence of framboidal pyrite (FeS2) in antique books (Garcı´a-Guinea et al. 1997). It has been observed that sulphate-reducing bacteria can cause sphalerite (ZnS) deposition in nature (Labrenz et al. 2000). Sillitoe et al. (1996) reported chalcocite (Cu2S) enrichment during weathering. This study showed that chalcocite precipitated following Cu-binding by bacteria, thus resulting in a sevenfold enrichment of copper content in sulphide deposits. This work is a clear example of the important economic implications of bacterial mineralization. Microbially induced precipitation of sulphides around oceanic hydrothermal vents has recently drawn much attention, since this process could be evidence of the most primitive life on Earth (Banfield et al. 1998). These environments could have provided compounds such as (Ni,Fe)S which might be crucial for the origins of life (Huber & Wa¨chtersha¨user 1998). Examples of sulphate-reducing bacteria responsible for sulphide mineralization are Desulfovibrio and Desulfotomaculum (Gould et al. 1997). For additional information, see reviews by Gould et al. (1997) and Nordstrom & Southam (1997) on microbiological formation of sulphides. Bacterial activity may result in the formation of silicate phases (Ferris et al. 1986; Urrutia & Beveridge 1994). Fe– Al silicate, assumed to be the clay mineral chamosite [(Fe,Mg)5Al(Si3Al) O10(OH,H)8], was found to precipitate in the presence of bacteria (Ferris et al. 1987). Phoenix et al.
(2001) have described a remarkable case of silica precipitation on bacterial cells. The authors observed that an Fe-rich silica crust formed around cyanobacteria. They suggested that this mineral shell could act as a shield against UV radiation. However, these bacteria face a dilemma, namely, how to escape from the silica coating when it surrounds the entire microorganism. Cyanobacteria have solved this problem by restricting silica deposition on both ends, thus allowing the cell to exit the shell following cellular division. This strategy could have allowed these bacteria to endure the high UV irradiation conditions of early Earth. Sulphates were demonstrated to precipitate on bacterial cells, both in nature and in the laboratory (Thompson & Ferris 1990). Precipitation of common sulphates such as gypsum (CaSO4.H2O) (Thompson & Ferris 1990) and barite (BaSO4) (Gonza´lez-Mun˜oz et al. 2003) has been reported. Uncommon sulphates such as jarosite [KFe3(SO4)2 (OH)6] and schwertmannite [Fe8O8(OH)6SO4] precipitated in spring waters in the presence of bacteria (Kawano & Tomita 2001). Jarosite presence in Martian Gusev Crater soils has been reported (Morris et al. 2004), which raises the question: Did bacteria play a role in its formation? Bacterial oxidation of sulphides results in the precipitation of sulphates, which causes soil and shale heave and significant geotechnical problems (Gillot et al. 1974). Acidothiobacillus ferrooxidans transform pyrite into sulphates. Following pyrite oxidation, jarosite and gypsum precipitate within shales when potassium and calcium ions are available (Quigley et al. 1973). Acidic mine drainage has been also directly connected with iron sulphide oxidation by microorganisms (Davis 1997). Nitrates such as nitrocalcite [Ca(NO3)2.4(H2O)], also known as ‘saltpetre’, are commonly found as in cave deposits and in soils (Northup & Lavoie 2001). The formation of calcium nitrate has been associated to the activity of nitrifying bacteria such as Nitrosomonas and Nitrobacter. They convert ammonium ions into nitrite and then nitrate (Focht & Verstraete 1977), which in the presence of Ca ions leads to the precipitation of nitrocalcite. Phosphates, particularly those including Ca and Mg, such as apatite [Ca5(PO4)3(F,Cl,OH)] and struvite (NH4MgPO4.6H2O) have been shown to grow on the outer membrane of Gram-negative bacteria. Phosphate precipitation induced by bacteria is particularly relevant because of its health-related implications. For instance, formation of human kidney stones has been related to bacterial infection. The calcium phosphates (apatite and hydroxylapatite) which form the calculi systematically enclose bacterial bodies. This observation suggests that bacterial infection is directly related to the formation of these pathological concretions (Hess et al. 1994;
BACTERIAL BIOMINERALIZATION
Kajander & C¸iftcioglu 1998). However, the possible role of bacteria, as well as ‘nanobacteria’, in the development of such pathologies is a matter of strong controversy (Cisar et al. 2000; Aloisi 2008). Chlorides such as halite (NaCl) form in natural habitats in the presence of halophilic bacteria (Castanier et al. 1999). Despite the previous example, works linking halide precipitation with microbial activity are inconclusive. There is, however, evidence for the development of bacteria and archaea in hypersaline environment associated to halite precipitation. For instance, Vreeland et al. (2000) report the presence of viable bacteria included in 250 million-year-old halite crystals, although it is difficult to link past bacterial activity and the formation of such crystals. On the other hand, cyanobacteria have been found in halite rocks in the hyperarid core region of the Atacama Desert where no other life form has been detected (Wierzchos et al. 2006), while halite biomineralization by halophilic archaea has been suggested as a means for these microbes preservation (Adamski et al. 2006). As a result, much attention is paid to the study of brines and evaporites as analogs for microbial life in salt-rich deposits on Mars (Rothschild 1990; Mancinelli et al. 2004). Interestingly, halophylic bacteria appear to be highly versatile from a biomineralization perspective: they have been shown to induce the precipitation of a range of minerals including carbonates such as calcite, magnesium calcite, aragonite, hydromagnesite [Mg5(CO3)4(OH)2.2H2O] and monohydrocalcite (CaCO3.H2O), as well as phosphates such as struvite (Rivadeneyra et al. 2006). Economic deposits of native sulphur have been associated to bacterial activity (e.g. Sebastian-Pardo et al. 1983). In this respect, the bacterial origin of sulphur deposits in several caves was corroborated by 34S values showing enrichment in light sulphur isotopes (Northup & Lavoie 2001). Other native metals such as gold have also been found to precipitate in the presence of bacteria (Ferris 1997). Karthikeyan & Beveridge (2002) reported toxic soluble gold precipitation by Pseudomonas aeruginosa.
How does bacterial mineralization take place? Current models Biogenic mineral formation is thought to occur through either biologically controlled or induced processes (Lowenstam & Weiner 1989). Biologically controlled mineralization occurs in an isolated compartment within a living organism (e.g. magnetosomes in magnetotactic bacteria). Minerals formed in this way display a highly ordered structure and the organisms can control size, texture and
35
orientation of the precipitates (Mann 2001). This biomineralization process is typical of metazoan (e.g. shells of bivalves), but quite uncommon in bacteria. On the other hand, biologically induced (or mediated) mineralization is the result of microbial metabolism, but the microorganisms do not directly control how and where the precipitates form. Two stages can be identified in the biomineralization process induced or mediated by bacteria. A first stage includes the active modification of the physical-chemistry in the interior or surroundings of the bacteria (i.e. pH, Eh, ion concentration, PCO2). These changes ultimately lead to an increase in ion concentration (i.e. supersaturation) which is a prerequisite for mineral precipitation. The bacteria can directly contribute to ion concentration increase by producing inorganic metabolic by-products. Otherwise, bacteria can indirectly contribute to mineral precipitation by gathering and concentrating different ions both within and on/around the cells (Fortin et al. 1997). Nucleation of a mineral phase occurs in a second stage. Nucleation is a crucial moment for mineral precipitation. It can occur either homogeneously or heterogeneously (Mullin 1992). Homogeneous nucleation requires a significantly high supersaturation. The activation energy (i.e. supersaturation) can be drastically reduced by the presence of a foreign surface on which heterogeneous nucleation occurs. Heterogeneous nucleation, the most plausible nucleation process in nature, is promoted by the coupling between functional (macro)molecules in the bacterial cell wall and the new mineral phase. Heterogeneous nucleation may also occur on bacterial exopolymeric substances (EPS) (Braissant et al. 2003, 2007; Dupraz & Visscher 2005). This complex process is explained by the so-called theory of template-directed (or organic-matrix mediated) biomineralization (Mann 2001). This theory states that organic molecules can (self)assemble into a template so that there can be electrostatic (ionotropic effect), geometric, or stereochemical affinity/matching between the template and the inorganic precipitate (biomineral).
Organic macromolecules and biomineralization Mineralization concepts based on inorganic solid phase precipitation aided by organic macromolecules have been developed over the last few decades (Mann 2001). These concepts have been applied to the understanding of how biomineralization occurs (Lowenstam & Weiner 1989; Mann 2001; Rodriguez-Navarro et al. 2007). Organic macromolecules are normally acidic polyanionic polymers (e.g. proteins, glycoproteins, proteoglycans) which include carboxylic or
36
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
phosphatic functional groups. These charged groups play a key role in the heterogeneous nucleation of a new solid phase, as demonstrated by in vitro studies of calcium carbonate precipitation in the presence of amino acids (Jimenez-Lopez et al. 2003). Nonetheless, organic macromolecules play a dual role in the nucleation process. They can act as inhibitors or as promoters for mineral crystallization (Rodriguez-Navarro et al. 2007). Typically, crystallization inhibition occurs due to binding of counter-ions (i.e. metal ions), thus reducing the effective ion concentration in solution (i.e. reducing supersaturation). This reportedly occurs in bacterial biofilms due to the presence of EPS (Decho 2000; Arp et al. 2001; Dupraz & Visscher 2005). EPS displays a random distribution of charged functional groups which bind metals in a disordered array. The disordered arrangement of bonded ions prevents crystal nucleation (Arp et al. 2001). On the other hand, if a periodic, ordered arrangement of the functional groups in the organic matrix occurs, for instance on the bacterial cell membrane (SchultzeLam et al. 1996), or on altered EPS subjected to reorganization of acidic sites (Duprazz & Visscher 2005), stereochemical coupling between the organic substrate and the newly-formed solid phase can occur. The latter effect promotes heterogeneous nucleation of a crystalline phase. Thus a biotically induced or mediated mineral is formed on the cell surface.
Metabolic pathways leading to mineral precipitation The main metabolic activities resulting in mineral precipitation are: (a) oxygenic photosynthesis (e.g. cyanobacteria); (b) ammonification by protein degradation (e.g. soil bacteria such as Myxococcus or Bacillus); (c) ammonia oxidation (e.g. nitrifying bacteria); (d) iron (or manganese) oxidation (e.g. iron-oxidizing bacteria such as Gallionella ferruginea); (e) sulphur oxidation by sulphur oxidizing bacteria; and (f) sulphur reduction by sulphate reducing bacteria (for details see Nealson & Stahl 1997). Detailed description of these metabolic pathways falls outside the scope of this review. However, a representative example of the complex reactions leading to mineral precipitation is described below. Thompson & Ferris (1990) proposed the following metabolic pathway for calcium carbonate precipitation by the cyanobacterium Synechococcus. The bacterium converts intracellular HCO2 3 photosynthetically into reduced carbon (CH2O): HCO 3 þ H2 O ! (CH2 O) þ O2 þ OH
(1)
Intracellular OH2 ions are then exchanged for extracellular bicarbonate ions across the cell membrane.
The resulting alkalinization around the bacteria cells induces CO22 3 generation: 2 HCO 3 þ OH ! CO3 þ H2 O
(2)
In the presence of Ca ions, CaCO3 precipitates on the cell surface according to the reaction: Ca2þ þ CO2 3 ! CaCO3#
(3)
In the presence of SO22 4 ions, gypsum nucleation on the cell surface can also occur. Additional details regarding different bacterial metabolic pathways resulting in supersaturation with respect to a particular mineral phases can be found in Ehrlich (2002).
Myxobacteria Myxobacteria are Gram-negative heterotrophic bacteria whose cells are long rods with lengths of 3–12 mm and diameters of 0.7–1.2 mm (Fig. 1a). On solid surfaces they display gliding motility, which is necessary for their swarming and fruiting body (Fig. 1b, c) development (Kaiser 2003). Although phylogenetically they belong to the delta subdivision of the Proteobacteria in the order Myxobacterales (Shimkets & Woese 1992), they do not share many of the phenotypic features shown by the other members of this subdivision, such as Bdellovibrio or Desulfovibrio. The characteristics that define the order Myxobacterales are thoroughly described by Shimkets et al. (2005). Myxobacteria are very common in nature, particularly in topsoil rich in organic material. Within a pH range of between 5 and 8, and in aerated surface layers, all soils appear to contain some myxobacteria (Dawid 2000). They are also able to colonize other habitats, especially those rich in microbial life, such as plant rhizosphere, dung, decaying plant material, and bark from both living and dead trees (Reichenbach 1993). Myxobacteria are easily washed from soil into water where they can survive and multiply, consequently they are common in fresh water (Hook 1977), and in marine environments (Iizuka et al. 2003). Myxobacteria produce a large variety of active compounds, such as enzymes and antibiotics, which have great biotechnological applications (Reichenbach & Ho¨fle 1999; Ho¨fle & Reichenbach 2005). For example, myxobacteria are known to produce antibacterial, antifungal, antiviral, and, to a lesser extent, insecticidal compounds. Myxobacteria form large communities known as swarms that feed on a variety of macromolecules (i.e. proteins, starch, lipids, nucleic acids and even cellulose) because they produce numerous extracellular hydrolytic enzymes. They can also degrade
BACTERIAL BIOMINERALIZATION
37
Fig. 1. M. xanthus rods, myxospores and fruiting body: (a) vegetative cells as observed using a scanning electron microscope (SEM); (b) transmission electron microscope (TEM) photomicrograph of a fruiting-body myxospore; (c) SEM photomicrograph of the fruiting body.
entire cells of a wide range of microorganisms, including Gram-negative bacteria, yeasts, and cells of different species of myxobacteria. The products of all these enzymes are used as nutrients by the cells of the swarm. They have developed different adaptive mechanisms to survive and compete in their natural habitats. As the best-characterized myxobacterium is Myxococcus xanthus, most of this review will deal with this species. For other details on the biology of myxobacteria, the reader is referred to several books and reviews (Dworkin & Kaiser 1993; Dworkin 1996; Shimkets 1999; Dawid 2000; Whitworth 2008). When myxobacterial cells are living in a medium with nutrients they follow a vegetative life cycle during which each cell elongates and then divides, by binary fission, into two daughter cells. However, when nutrients are depleted, a developmental cycle is initiated. Although starvation triggers development, several other circumstances are required for the cycle to be completed, such as high cell density and a solid surface. When cell density is high enough, cells are motile, and nutrients are depleted, cells inside a swarm start to move to certain points, where they aggregate and originate multicellular structures known as fruiting bodies. Inside the fruiting bodies, cells differentiate and originate myxospores (Fig. 1b), which are dormant and resistant to several stress conditions such as desiccation, UVirradiation and sonication. Myxospores are able to germinate and resume a new vegetative cycle when nutrients are present again. Myxobacteria are also of particular interest because they can be used as a model to study various biological processes such as morphogenesis and differentiation. These processes have turned out to be very complex, involving large numbers of proteins. A significant step towards the better understanding of such processes has taken place recently by the completion of the genome sequence
of M. xanthus DK1622 (Goldman et al. 2006). Thus, myxobacteria represent one of the most interesting groups among prokaryotes, in both basic and applied microbiology.
Myxococcus-induced biomineralization Several species within the Myxococcus genus have been demonstrated to be highly versatile for biomineralization experiments. Depending on the culture media composition and type (solid or liquid), these bacteria are able to precipitate a wide range of minerals that include: phosphates, carbonates, sulphates, chlorides, oxalates and silicates.
Production of phosphates Struvite is one variety of insoluble phosphate that has received special attention from microbiologists. This is due to the fact that many different types of bacteria produce this mineral under laboratory conditions and that a relationship has been established between kidney calculi and urinary infections (Samuell & Kasidas 1995). Struvite production by myxobacteria (Fig. 2) has been reported using Myxococcus coralloides (Gonza´lez-Mun˜oz et al. 1993) and M. xanthus (Ben Omar et al. 1995; Da Silva et al. 2000). At first sight, the production of struvite by M. xanthus and M. coralloides, like in other cases, seems to be a consequence of their metabolism. This provides the physical–chemical conditions for its crystallization to occur: the pH increases [alkaline pH is necessary for the production of this mineral (Pe´rez-Garcı´a et al. 1989)], and NHþ 4, a by-product of the bacteria nitrogen metabolism, is supplied. However, Gonza´lez-Mun˜oz et al. (1993) and Ben Omar et al. (1994) found that pH increase and NHþ 4 production are not sufficient conditions for the formation of struvite: the physical presence
38
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
Fig. 2. SEM photomicrographs of struvite crystals precipitated in the presence of M. xanthus: (a) twined; and (b) bipyramidal struvite crystals.
of the bacteria is also necessary. The production of struvite using both living and dead (entire or disrupted) cells, as well as with and without EPS, indicate that M. xanthus cells may act as, or supply, heterogeneous nuclei for struvite crystallization when suitable media and bacterium culture ages are chosen (Ben Omar et al. 1995). Working with different strains of Pseudomonas and Azotobacter, Rivadeneyra et al. (1992) also found that, depending on culture age, heat-killed cells trigger struvite formation. Dead cells and cell debris may contribute to the formation of struvite deposits in nature. The crystallization processes of very different genera of bacteria previously reported (Myxococcus, Pseudomonas and Azotobacter) could indicate that this contribution is a widespread phenomenon. In the case of myxobacteria in general, their social behaviour,
morphogenesis and differentiation involve a phase in which 80 –90% of the cells undergo lysis (Dworkin & Kaiser 1993). Lysis contributes to the production of heterogeneous nuclei for crystallization. Since myxobacteria are abundant in organically enriched soils, the hypothesis that they participate in struvite precipitation in nature (Ben Omar et al. 1995) would appear to deserve more attention. Another interesting finding was that M. xanthus membranes (total membrane fraction: cytoplasmic and outer membranes) supply heterogeneous nuclei in the production of struvite (Gonza´lezMun˜oz et al. 1996). While some authors have found that each bacterial species displays a very narrow range of struvite crystalline habits (Pe´rez-Garcı´a et al. 1989), others have found that struvite production by M. xanthus and M. coralloides evidence a large number of diverse crystal morphologies (Fig. 2) (Ben Omar et al. 1996). Other phosphates produced by myxobacteria, newberyite (MgHPO4.3H2O) and schertelite [(NH4)2MgH2(PO4)2.4H2O], were reported by Gonza´lez-Mun˜oz et al. (1994). These minerals, which have been considered syngenetic with struvite (Dana 1966), were produced by M. coralloides D as minor mineral phases when struvite was found in certain liquid cultures under static conditions (Gonza´lez-Mun˜oz et al. 1994). The production of struvite syngenetic phosphates by a myxobacterium is of relevance: before the latter publication it had not been reported that bacteria could produce such minerals, and also because newberyite is associated with struvite in kidney calculi. Additionally M. xanthus shows a significant capacity for biosorption of heavy metals and lanthanides (Gonza´lez-Mun˜oz et al. 1997; Merroun et al. 1998, 2001, 2003). The metals biosorbed frequently appear as metal-phosphate (Merroun et al. 2003). These metals are linked mainly to extracellular polysaccharide and to cell wall (Fig. 3a). However, in some cases, they appear as intracellular polyphosphate granules (Fig. 3b). In this latter case highly crystalline phases are commonly observed, as shown by selected area electron diffraction (SAED) pattern of barium phosphate nodules formed within M. xanthus cells (inset in Fig. 3b). Extended X-ray absorption fine structure spectroscopy (EXAFS) analysis has shown that at pH 4.5, M. xanthus cells are able to precipitate uranium phosphate, a mineral phase belonging to the meta-autunite group (Jroundi et al. 2007). U-phosphate precipitates were localized mainly within EPS and on the cell surface, although some poorly-crystalline intracellular precipitates were also observed (Fig. 3c). The biomineralization of U(VI) may be associated with the activity of indigenous acidic phosphatase. The activity of this enzyme has been shown to play
BACTERIAL BIOMINERALIZATION
39
Fig. 3. TEM images of: (a) La-rich phosphates precipitated on M. xanthus cell wall and within the EPS (reprinted from Merroun et al. (2003), with permission from Elsevier); (b) Ba phosphate spherulites (arrows) formed within M. xanthus cells. The SAED pattern (insets) shows the crystalline nature of the precipitates; (c) U-rich phosphates precipitated on and within (arrow) M. xanthus cells and within EPS. The SAED pattern of intracellular precipitates (arrow) in insets show diffuse spots and rings confirming their poorly crystalline nature (reprinted from Gonza´lez-Mun˜oz et al. (1997), with permission from Elsevier).
a key role in the biomineralization of uranium phosphate by several bacterial strains including those isolated from uranium contaminated sites (Martinez et al. 2007; Merroun & Selenska-Pobell 2008) as well as by laboratory strains such as Citrobacter sp. (Jeong et al. 1997). The uranium phosphate precipitation by the cells of M. xanthus is affected by the chemistry of the metal solution (e.g. pH). At pH 2 no U-precipitation was observed, this radionuclide being bound to the cell surface organic phosphate groups. Similar pH-dependent U/bacteria interactions have been found in the case of other Gram-positive or Gram-negative bacterial strains, such as Bacillus sphaericus JG-7B, and Sphingomonas sp. SW366-3 (Merroun et al. 2006; Nedelkova et al. 2007). The fact that M. xanthus precipitates uranium as a mineral phase indicate that this bacterium may be implicated in the transport and mobility of U in the environment and can be used for bioremediation purposes in sites where bacterial populations belonging to the Myxobacteria group are predominant (e.g. uranium contaminated sediments at the U.S. DOE Field Research Center (FRC) near Oak Ridge, TN; Petrie et al. 2003).
Rodriguez-Navarro et al. 2003, 2007). Passive (or indirect) precipitation of calcite was reported using M. xanthus cellular membranes (Gonza´lez-Mun˜oz et al. 1996), and UV-killed bacterial cells (Ben Chekroun et al. 2004). Only recently has thoroughly documented research been carried out on the production of calcite and vaterite (a metastable calcium carbonate polymorph) by myxobacteria (Rodriguez-Navarro et al. 2003, 2007; Ben Chekroun et al. 2004). The aim of the study by Rodriguez-Navarro et al. (2003) was to assess the consolidation and protection of calcarenite (an ornamental stone) by Myxobacterium calcium-carbonate production. M. xanthus was cultivated in liquid media in the presence of calcarenite. Variations in the phosphate concentrations of the culture medium, which lead to changes in local pH and bacterial productivity, were shown to affect the precipitation of CaCO3 polymorphs (vaterite or calcite). M. xanthus metabolic activity involves production of NH3 by means of oxidative deamination of amino acids. NH3 creates an alkaline microenvironment around the cell according to the reaction, NH3(aq) þ H2 O , NHþ 4 þ OH
Production of carbonates Myxobacteria may play both an active (or direct) and a passive (or indirect) role in carbonate precipitation. Active calcium carbonate production by myxobacteria, as a consequence of its metabolism, was reported using living bacterial cultures of several species of Myxococcus genus (Ruiz et al. 1988; Gonza´lez-Mun˜oz et al. 2000;
(4)
Metabolic activity and the resulting rise in local pH may lead to a high local supersaturation with respect to either vaterite or calcite. CO2 is also produced by the bacterium. CO2 dissolves and trans22 forms into either HCO2 3 or CO3 depending on the pH. Once sufficient supersaturation is reached, calcium carbonate formation by heterogeneous nucleation occurs on the bacterial cell walls, as
40
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
can be observed in the SEM photomicrographs showing bacterial imprints within rhombohedral calcite crystals (Fig. 4). In this study vaterite was detected surrounding calcified bacterial cells (Fig. 5). Vaterite is a metastable polymorph of CaCO3 and is rare in natural environments. It is unstable and rapidly transforms into calcite (or aragonite) at room temperature in an aqueous solution. However, it commonly forms in synthetic processes where organics are present and has been reported to develop in the presence of microorganisms in nature (see Rodriguez-Navarro et al. 2007, and references therein). The presence of microbial films and the apparent affinity of the bacterial cell walls for vaterite appear to cause vaterite precipitation in a manner similar to that described by Mann et al. (1988). One complementary explanation for vaterite development in the presence of M. xanthus is suggested
by Ostwald’s step rule (Jimenez-Lo´pez et al. 2001). According to this rule, stable phase formation is sometimes preceded by metastable phases which are normally favored under non-equilibrium conditions (i.e. high supersaturation). Vaterite could therefore be a precursor of calcite, forming in localized areas in which supersaturation and pH rapidly rise as a consequence of local intense metabolic bacterial activity (Ben Chekroun et al. 2004). Local increases in metabolic activity are consistent with localized vaterite formation at high supersaturation and also with the euhedral calcite rhombohedra formed nearby at a lower supersaturation. A high supersaturation seems to be a prerequisite for vaterite formation in the laboratory. This is consistent with the observations of Rodriguez-Navarro et al. (2003, 2007) regarding vaterite habit, size and crystal density. The large number of tiny acicular
Fig. 4. SEM photomicrographs of: (a) calcite rhombohedra with bacterial imprints (bc) developed on calcarenite stone following a bioconsolidation treatment with M. xanthus; (b) detail of bacterial calcite rhombohedra.
Fig. 5. SEM photomicrographs of: (a) aggregate of vaterite spherulites formed in the presence of M. xanthus; (b) detail of a bacterial vaterite spherulite.
BACTERIAL BIOMINERALIZATION
vaterite crystals, as well as the presence of vaterite spherical aggregates (Fig. 5), is in agreement with their formation under high supersaturated conditions, while large, rhombohedral calcite crystals are formed at relatively low supersaturation. Mullin (1992) previously described the details of supersaturation influence on nucleation density, crystal size and crystal morphology. Rodriguez-Navarro et al. (2007) found precipitation of vaterite spherulites in the microenvironment around M. xanthus cells, and directly onto the surface of bacterial cells. In the latter case, fossilization of bacteria occurred. Vaterite crystals formed by aggregation of oriented nanocrystals with c-axes normal to the bacterial cellwall, or to the core of the spherulite when bacteria were not encapsulated. The preferred orientation of vaterite c-axes appeared to be determined by electrostatic affinity (ionotropic effect) between vaterite crystal (0001) planes and the negatively-charged functional groups of organic molecules on the bacterium cell-wall or on EPS. According to analyses of the changes in the culture medium chemistry as well as high resolution transmission electron microscopy (HRTEM) observations, polymorph selection was associated to both physicochemical (kinetic) factors (high supersaturation) and to stabilization by organics, both connected with bacterial activity. Precipitation of Mg-calcite induced by M. xanthus was reported by Gonza´lez-Mun˜oz et al. (2000). This work describes the results of experiments designed to study the crystallization of CaCO3 by a biogenic process in the presence of Mg2þ. For this purpose, a culture of M. xanthus was developed in solidified agar –agar nutritive solution. This work makes some interesting observations, regarding both the location of crystals in M. xanthus colonies and the diversity of the crystal morphologies. As a consequence of M. xanthus metabolic activity, the various metabolites produced (CO2 and NH3) have to diffuse from the zones of higher density of microbial growth to the rest of the culture medium. Ammonia production also led to rises in pH. The consumption of nutrients from the culture media brought about the diffusion of different ions in the opposite direction. Counter-diffusion of microbial metabolites and culture ions produces spatio-temporal concentration gradients which lead to supersaturation conditions. Such conditions permit the onset of precipitation at certain moments and at certain points. As a consequence, the crystal precipitation zone is a band located on the periphery of the M. xanthus colonies with a width of c. 1 mm. Appropriate supersaturation and pH conditions occur in this zone. In addition, the pattern in which counterdiffusion evolves influences the development of different crystal morphologies produced by Mg-calcite (Fig. 6a). On the other hand, a striking relationship was found between
41
Fig. 6. SEM photomicrographs of: (a) dumbbells and spheres of magnesium calcite crystal aggregates; and (b) monohydrocalcite dipyramidal crystal formed in the presence of M. xanthus.
viscosity of the medium and the type of precipitate. The precipitation of the Mg-calcite invariably took place in gelled media. The viscosity of the medium, and hence the rate of ion diffusion and precipitation, seem to be the overriding control. These results are consistent with the observations of Buczynski & Chafetz (1991) showing bacterially induced precipitation of calcium carbonate in the gel-like slime (biofilm) present in stromatolites. On the other hand, Mg probably plays a key role in the development of the morphologies of the precipitates since these morphologies had never been observed with M. xanthus in the absence of Mg. In general, when a particular form of a crystal or a crystal bundle is found in a culture, most of the morphologies are similar (Buczynski & Chafetz 1991). In contrast, it is noteworthy that M. xanthus gives rise, both simultaneously and in the same proportion, to a diversity
42
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
of morphologies (Fig. 6a). All are Mg-calcite of almost similar low Mg content. Morphologies similar to those produced by M. xanthus have been described by various authors studying biotic and abiotic, laboratory as well as natural systems (e.g. Krumbein 1979; Buczynski & Chafetz 1991; Ferna´ndez-Diaz et al. 1996). Precipitation of monohydrocalcite in the presence of M. xanthus has been reported too (Ben Chekroun 2000). Euhedral hydrocalcite crystals (Fig. 6b) formed, most probably at a high supersaturation, as precursors of more stable calcite (Jimenez-Lopez et al. 2001). When appropriate supersaturated solutions are employed, calcium carbonate has also been found to precipitate on M. xanthus cellular membranes under abiotic conditions, a precipitation which can be considered passive or indirect (Gonza´lez-Mun˜oz et al. 1996). The precipitation of these calcite crystals probably takes place on the negatively charged points of the external side of the cellular structures. De´farge et al. (1996) found that microscopic three-dimensional organic networks inherited from sheaths of dead cyanobacteria acted as a matrix for calcification. They reported that crystal nucleation began at acidic sites which are capable of binding a wide range of cations. It has been suggested that specific attributes of certain bacteria induce and affect calcium carbonate formation (Hammes et al. 2003). Precipitation occurs preferentially on macromulecules such as lipid bilayers of vesicles and glycoproteins and proteoglycans that are constituents of bacterial cell membranes. Such organics act as a nucleation template for calcium carbonate. The nature of such an organic matrix may determine which ion is preferentially adsorbed and, consequently, which mineral phase is formed. Thus, biomineralization could be considered strain-specific. For instance, bacteria that preferentially adsorb Mg2þ on their membranes induce dolomite formation, whereas calcite precipitation is induced by preferential adsorption of Ca2þ (Van Lith et al. 2003). However, M. xanthus appears to challenge the aforementioned hypothesis of strain-specific biomineralization since it is able to induce precipitation of carbonates with contrasting structure and composition.
observed barite precipitation in laboratory experiments using M. xanthus (Fig. 7). Barite production started with a phase dominated by P and Ba that evolved to well-crystallized barite crystals (Fig. 7a). The initial poorly crystalline P-rich precursor phase (see the diffuse rings in the SAED pattern in inset of Fig. 7b) suggested that phosphoryl and carboxyl groups in the structural polymers of the cell wall outer membrane may be sorbent constituents which play an important role in the precipitation process. Deprotonation of these groups provided discrete complexation sites for
Production of sulphates Myxococcus also mediates the precipitation of barite (Gonza´lez-Mun˜oz et al. 2003) and taylorite [(K,NH4)2SO4] (Gonza´lez-Mun˜oz et al. 1994). Barite dissolution by sulphate reducing bacteria has been proposed by several authors (e.g. Phillips et al. 2001), but no bacterial contribution to barite precipitation was considered until the finding of Gonza´lez-Mun˜oz et al. (2003). These authors
Fig. 7. M. xanthus induced barite precipitation: (a) SEM photomicrographs of barite aggregates; and (b) TEM image on a barite precursor aggregate showing diffuse rings in the SAED pattern (inset) thus confirming its poorly-crystalline nature (reprinted from Gonza´lezMun˜oz et al. (2003), with permission from the American Society for Microbiology).
BACTERIAL BIOMINERALIZATION
metals in solution. Sorption is enhanced as pH increases and as surface groups deprotonate. Sorption of large cations such as Ba2þ is particularly favoured by PO32 4 ligands and this process enables the formation of large coordination polyhedra (a coordination of 10 or higher). As the SO22 4 content of the culture medium increases (likely due to degradation and oxidation of amino acids), the ions are captured by the Ba ions, thus giving rise to a barite growth nucleus. Because mineral production only occurred in the living bacterial colonies, favourable conditions for crystallization relate to bacterium presence and metabolism. Although these results cannot be directly extrapolated to natural environments, this represents a significant step in our understanding of the biogeochemical cycle of Ba, which has attracted considerable research due to its close relationship with marine biological productivity (e.g. Paytan et al. 1996). Taylorite, a mineral that frequently appears associated with struvite in guano deposits (Dana 1966), was produced by M. corolloides D (Gonza´lezMun˜oz et al. 1994), as in nature, when this bacterium produces struvite in certain conditions. This work was the first to report bacterial production of this mineral. Although myxobacteria have not to date been isolated from guano deposits, they are commonly found in soils rich in decomposing organic material, and in fact, dungs of various animals are some of the preferred isolation sources (Dworkin & Kaiser 1993). On the other hand, heterotrophic, ammonifying microorganisms are present in pigeon dung (Strzelczyk 1981). Taylorite production by M. corolloides may therefore provide a basis from which a model for the biogenesis of this mineral can be developed.
Other minerals produced by M. xanthus M. xanthus can produce other types of minerals depending on the culture medium composition and/or culture conditions. Among the phosphates: apatite, hydroxylapatite and natrophosphate [Na7(PO4)2F.19(H2O)] have been obtained in small amounts under different conditions (Ben Chekroun 2000; Ferna´ndez-Luque 2002). Production of a chloride, chlorargyrite (AgCl), has been reported by Merroun et al. (2001) in silver biosorption experiments in which non-proliferative M. xanthus biomass was used. The silver was located as electron-dense deposits in the EPS, on the cell wall and within the cytoplasm. Also, the production of the oxalate weddellite (CaC2O4.2H2O) occurs (Fig. 8a) when culture medium contains calciumacetate (Ben Chekroun 2000; Ferna´ndez-Luque 2002) or in the heavy metal presence. Small amounts of weddellite have been observed to develop during M. xanthus-based consolidation of ornamental
43
Fig. 8. SEM photomicrographs of weddellite aggregates: (a) fibrous-radiate crystal aggregate formed in solid medium inoculated with M. xanthus; and (b) prismatic-shaped aggregate formed on calcarenite stone following a M. xanthus conservation treatment.
carbonate stone (Fig. 8b). Finally, data exist regarding the production of silica minerals; for instance, opal CT has been found using a culture medium with silica (Ben Chekroun 2000).
Concluding remarks Bacteria are able to precipitate a large number of minerals, playing a direct and/or an indirect role in this process. Their role is direct when metabolic activity leads to supersaturation with respect to a particular mineral phase. Bacterial dead cells and their membranes, as well as EPS, play an indirect role when acting as substrates for heterogeneous nucleation. A single bacterial genus, Myxococcus, can precipitate phosphates, carbonates, sulphates, silicates,
44
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
oxalates and chlorides. This clearly evidences that the mechanisms which control bacterial biomineralization are not mineral- or bacterium-specific. On the contrary, they appear to be universal and depend on the environment in which bacteria dwell. This is in agreement with studies on the biomineralization capacity of other bacteria such as Schewanella sp. that is able to induce the precipitation of different iron oxi-hydroxides and ferric phosphates (Hyacinthe et al. 2008) as well as arsenic sulphides (Lee et al. 2007), Synecococcus which is able to precipitate gypsum, calcite and magnesite (Thompson & Ferris 1990), or Halomonas sp. that is able to induce the precipitation of calcite, Mg-calcite, aragonite, struvite, monohydrocalcite and hydromagnesite (Rivadeneyra et al. 2006). Regarding future research trends on bacterially induced or mediated biomineralization, a number of important minerals and poorly understood biomineralization processes should be studied in detail. For instance, bacteria play a crucial role determining rates, pathways and end products of the formation and degradation of rocks, minerals and organic matter. Consideration of the role of bacterial biomineralization is thus required to gain a deeper understanding of global biogeochemical cycles and atmosphere-lithosphere-biosphere exchanges. Significant progress has been made concerning well known element cycles, such as Fe, N and S. However, other cycles (e.g. the Ba cycle) are still poorly understood and may also rely on bacterial activity. The formation of quartz in sedimentary environments has puzzled the scientific community for decades. Bacterial activity could help explain low temperature quartz formation. Experiments on quartz precipitation in the presence of bacteria would clarify this long standing geological controversy. Microbial growth in the deep biosphere may have significant geobiochemical implications since a tenth of the Earth biomass lives in deep sea-floor layers. Future research should focus on those biomineralization processes which may be associated to this hidden microbial world. The study of bacteria which are adapted to extreme environments could offer new prospects for bioremediation and pollution control. Mineralization experiments using extremophiles might also further our understanding of geobiochemical processes assumed to exist on early Earth, and elsewhere. This research may also allow us to more precisely establish the microbiological links among most natural chemical and geological processes, as well as to define constraints for life development. For instance, halophilic bacteria and Archaea induced mineralization in hypersaline, low T environments could help disclose if microbial
life is possible on Mars. Furthermore, precipitation of unusual Fe and Mg carbonates by bacteria, with a spherulitic morphology similar to that observed in bacterial vaterite, could help disclose whether or not Martian meteorite spherulitic carbonates are, in fact, bacteriogenic. Detailed studies should be conducted regarding the so-called nanobacteria and their putative role in mineralization in natural environments, human pathological concretions and on other planets. These studies will help corroborate the existence of such bacteria, and if so, how they contribute, for instance, to the mineralization of phosphates and carbonates in kidney stones. A critical, but poorly understood, aspect of bacterial mineralization is the role of organics (in bacterial cells and sheaths, as well as EPS) on the nucleation of different solid phases. Although there is some evidence pointing to a possible template effect of organics in bacterial mineralization, detailed studies of the organic-mineral interface at the nanoscale should be performed. The combined use of advanced techniques such as high resolution transmission electron microscopy and near edge X-ray absorption fine structure spectroscopy (Benzerara et al. 2006), as wells X-ray and/or neutron diffraction (e.g. using synchrotron facilities for in-situ analysis of the early stages of mineralization) could enable a better understanding of the role of organics in bacterial mineralization and define criteria for recognition of microbial biosignatures. Another important issue, not directly dealt with here, is how bacterial minerals grow. Banfield et al. (2000) have shown that a self-assembled coarsening mechanism is responsible for the growth of bacterial oxyhydroxides. Similar results have been observed by Rodriguez-Navarro et al. (2007) in the case of bacterial vaterite. During the oriented aggregation of nanoparticles precipitated in the presence of bacteria, organics produced by the bacteria (e.g. EPS) can be trapped within the aggregate, thus offering a way to identify bacterial biominerals. Research should be carried out to determine whether or not this process is universal. The relationship between bacterial infection in humans and the formation of pathological precipitates will also deserve further research. A causal relationship has been found between bacteria and phosphate mineralization in humans. However, it has been suggested that abiotic processes could lead to similar phosphate precipitation in human pathological concretions. Nonetheless, pathological concretions include other minerals such as vaterite, typically associated to biogenic processes. Therefore, research should also focus on vaterite formation in the presence of bacteria. Bacteria (or other dead cells, like osteoblasts) may play a purely passive role in the process (i.e. providing heterogeneous
BACTERIAL BIOMINERALIZATION
nuclei), or their metabolic activity may be crucial for vaterite mineralization. In the latter case, antibiotics could help prevent these health problems. In summary, our understanding of bacterial mineralization has improved substantially in recent decades. However, much more remains to be done if we are to determine with greater precision how and where bacterial mineralization takes place. Finally, this research will help define the implications of bacterial mineralization in a variety of chemical, biological and geological processes. This work was supported by the Spanish Direccio´n General de Investigacio´n (DGI contracts MAT2006-05411, MAT2009-11332, CGL2006-13327-CO4-04, MARM200800050084447, CGL2009-07603 and CGL2007-61489/ BOS), by the Junta de Andalucı´a (contract RNM3943) and by Research Groups NMR 179, FQM 195, RNM 5212 and BIO 103 of the Junta de Andalucı´a. Editing of the English was done by Marco Bettini.
References A DAMSKI , J. C., R OBERTS , J. A. & G OLDSTEIN , R. H. 2006. Entrapment of bacteria in fluid inclusions in laboratory-grown halite. Astrobiology, 6, 552–562. A LOISI , G., G LOTER , A., K RU¨ GER , M., W ALLMANN , K., G UYOT , F. & Z UDDAS , P. 2006. Nucleation of calcium carbonate on bacterial nanoglobules. Geology, 34, 1017–1020. A LOISI , G. 2008. An alternative origin for nanobacteria in kidney stones. Bioscience Hypotheses, 1, 138– 141. A LTERMANN , W. & K AZMEIRCZAK , J. 2003. Archean microfossils: a reappraisal of early life on Earth. Research in Microbiology, 154, 611– 617. A RP , G., R EIMER , A. & R EITNER , J. 2001. Photosynthesis-induced biofilm calcification and calcium concentrations in Phanerozoic oceans. Science, 292, 1701–1704. B ANFIELD , J. F. & N EALSON , K. H. (eds) 1997. Geomicrobiology: Interactions between Microbes and Minerals. Mineralogical Society of America, Reviews in Mineralogy, Washington, DC, 35. B ANFIELD , J. F., W ELCH , S. A. & E DWARDS , K. J. 1998. Microbes as geochemical agents. Newsletter of the Goechemical Society, 96, 11–17. B ANFIELD , J. F., W ELCH , S. A., Z HANG , H., E BERT , T. T. & P ENN , R. L. 2000. Aggregation-based growth and microstructure development in natural iron oxyhydroxide biomineralization products. Science, 289, 751–754. B ARABESI , C., G ALIZZI , A., M ASTROMEI , G., R OSSI , M., T AMBURINI , E. & P ERITO , B. 2007. Bacillus subtilis gene cluster involved in calcium carbonate biomineralization. Journal of Bacteriology, 189, 228– 235. B AZYLINSKI , D. A. & M OSKOWITZ , B. M. 1997. Microbial biomineralization of magnetic iron minerals: microbiology, magnetism and environmental significance. In: B ANFIELD , J. F. & N EALSON , K. H. (eds) Geomicrobiology: Interactions between Microbes and Minerals. Mineralogical Society of
45
America, Reviews in Mineralogy, Washington, DC, 35, 181–223. B EN C HEKROUN , K. 2000. Produccio´n de diversos minerales por nu´cleos heteroge´neos de cristalizacio´n aportados por Myxococcus xanthus y comparacio´n con otros microorganismos. PhD thesis, Universidad de Granada, Spain. B EN C HEKROUN , K., R ODRIGUEZ -N AVARRO , C., G ONZA´ LEZ -M UN˜ OZ , M. T., A RIAS , J. M., C ULTRONE , G. & R ODRIGUEZ -G ALLEGO , M. 2004. Precipitation and growth morphology of calcium carbonate induced by Myxococcus xanthus: implications for recognition of bacterial carbonates. Journal of Sedimentary Research, 74, 124–132. B EN O MAR , N., E NTRENA , M., G ONZA´ LEZ -M UN˜ OZ , M. T., A RIAS , J. M. & H UERTAS , F. 1994. The effects of pH and phosphate on the production of struvite by Myxococcus xanthus. Geomicrobiology Journal, 12, 81–90. B EN O MAR , N., M ARTI´ NEZ -C AN˜ AMERO , M., G ONZA´ LEZ -M UN˜ OZ , M. T., A RIAS , J. M. & H UERTAS , F. 1995. Myxococcus xanthus’ killed cells as inducers of struvite crystallization. Its possible role in the biomineralization processes. Chemosphere, 30, 2387–2396. B EN O MAR , N., E NTRENA , M., G ONZA´ LEZ – M UN˜ OZ , M. T., A RIAS , J. M., H UERTAS , F. & R ODRI´ GUEZ G ALLEGO , M. 1996. Morphological diversity of struvite crystals produced by Myxococcus xanthus and Myxococcus coralloides. Toxicology and Environmental Chemistry, 53, 209– 217. B ENZERARA , K., M ENGUY , N., L O´ PEZ -G ARCI´ A , P., Y OON , T. H., K AZMIERCZAK , J., T YLISCZAK , T., G UYOT , F. & B ROWN JR , G. E. 2006. Nanoscale detection of organic signatures in carbonate microbialites. Proceedings of the National Academy of Sciences USA, 103, 9440–9445. B HAGAT , M., B URGESS , J. E., P AULA , A., A NTUNES , M., W HITELEY , C. G. & D UNCAN , J. R. 2004. Precipitation of mixed metal residues from wastewater utilising biogenic sulphide. Minerals Engineering, 17, 925 –932. B LYTH , A. J. & F RISIA , S. 2008. Molecular evidence for bacterial mediation of calcite formation in cold high-altitude caves. Geomicrobiology Journal, 25, 101– 111. B OQUET , E., B ORONAT , A. & R AMOS -C ORMENZANA , A. 1973. Production of calcite (calcium carbonate) crystals by soil bacteria is a common phenomenon. Nature, 246, 527–529. B OSAK , T. & N EWMAN , D. K. 2003. Microbial nucleation of calcium carbonate in the Precambrian. Geology, 31, 577 –580. B RAISSANT , O., V ERRECCHIA , E. P. & A RAGNO , M. 2002. Is the contribution of bacteria to terrestrial carbon budget greatly underestimated? Naturwissenschaften, 89, 366– 370. B RAISSANT , O., C AILLEAU , G., D UPRAZ , C. & V ERRECCHIA , E. P. 2003. Bacterially induced mineralization of calcium carbonate in terrestrial environments: the role of exopolysaccharides and amino acids. Journal of Sedimentary Research, 73, 485– 490. B RAISSANT , O., D ECHO , A. W., D UPRAZ , C., G LUNK , C., P RZEKOP , K. M. & V ISSCHER , P. T. 2007.
46
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
Exopolymeric substances of sulfate-reducing bacteria: Interactions with calcium at alkaline pH and implication for formation of carbonate minerals. Geobiology, 5, 401–411. B UCZYNSKI , C. & C HAFETZ , H. S. 1991. Habit of bacterially induced precipitates of calcium carbonate and the influence of medium viscosity on mineralogy. Journal of Sedimentary Petrology, 61, 226 –233. B UICK , R. 1992. The antiquity of oxygenic photosynthesis: evidence from stromatolites in sulfate-deficient Archaean lakes. Science, 25, 74– 77. C ASANOVA , J., B ODE´ NAN , F., N E´ GREL , P. & A ZAROUAL , M. 1999. Microbial control on the precipitation of modern ferrihydrite and carbonate deposits from the Ce´zallier hydrothermal spring (Massif Central, France). Sedimentary Geology, 126, 125– 145. C ASTANIER , S., P ERTHUISOT , J. P., M ATRAT , M. & M ORVAN , J. Y. 1999. The salt ooids of Berre salt works (Bouches du Rhoˆne, France): the role of bacteria in the salt crystallisation. Sedimentary Geology, 125, 9 –21. C ASTANIER , S., L E M E´ TAYER -L EVREL , G., O RIAL , G., L OUBIE` RE , J. F. & P ERTHUISOT , J. P. 2000. Bacterial carbonatogenesis and applications to preservation and restoration of historic property. In: C IFERRI , O., T IANO , P. & M ASTROMEI , G. (eds) Of Microbes and Art: The Role of Microbial Communities in the Degradation and Protection of Cultural Heritage. Plenum, New York, 201– 216. C HAFETZ , H. S. & F OLK , R. L. 1984. Travertines: depositional morphology and the bacterially constructed constituents. Journal of Sedimentary Petrology, 54, 289– 316. C HAN , C. S., DE S TASIO , G., W ELCH , S. A., G IRASOLE , M., F RAZER , B. H., N ESTEROVA , M. V., F AKRA , S. & B ANFIELD , J. F. 2004. Microbial polysaccharides template assembly of nanocrystal fibers. Science, 303, 1656– 1658. C ISAR , J. O., X U , D.-Q., T HOMPSON , J., S WAIM , W., H U , L. & K OPECKO , D. J. 2000. An alternative interpretation of nanobacteria-induced biomineralization. Proceedings of the National Academy of Sciences USA, 97, 11 511– 11 515. D ANA , E. S. 1966. Dana’s textbook in mineralogy with an extended treatise on crystallography and physical mineralogy. In: F ORD , W. E. (ed.) Dana’s Mineralogy. 4th edn. John Wiley and Sons, New York. D A S ILVA , S., B ERNET , N., D ELGENE` S , J. P. & M OLETTA , R. 2000. Effect of culture conditions on the formation of struvite by Myxococcus xanthus. Chemosphere, 40, 1289– 1296. D AVIS , B. S. 1997. Geomicrobiology of the oxic zone of sulfide mine tailings. In: M C I NTOSH , J. M. & G ROAT , L. A. (eds) Biological –Mineralogical Interactions. Mineralogical Association of Canada, Short Course Series, Ottawa, Ontario, 25, 93–112. D AWID , W. 2000. Biology and global distribution of myxobacteria in soils. FEMS Microbiology Review, 24, 403–427. D ECHO , A. W. 2000. Microbial biofilms in intertidal systems: and overview. Continental Shelf Research, 20, 1257– 1273. D E´ FARGE , C., T RICHET , J., J AUNET , A. M., R OBERT , M., T RIBBLE , J. & S ANSONE , F. J. 1996. Texture of
microbial sediments revealed by cryo-scanning electron microscopy. Journal of Sedimentary Research, 66, 935– 947. D REW , G. H. 1914. On the precipitation of calcium carbonate in the sea by marine denitrifying bacteria. Carnegie Institution Washington Publications, 1825, 9– 45. D UPRAZ , C. & V ISSCHER , P. T. 2005. Microbial lithification in marine stromatolites and hypersaline mats. TRENDS in Microbiology, 13, 429– 438. D WORKIN , M. 1996. Recent advances in the social and developmental biology of the myxobacteria. Microbiology Reviews, 60, 70–102. D WORKIN , M. & K AISER , D. (eds) 1993. Myxobacteria II. American Society for Microbiology, Washington DC. E HRLICH , H. L. 2002. Geomicrobiology, 4th edn. Marcel Dekker, New York. F ERNA´ NDEZ -D ´I AZ , L., P UTNIS , A., P RIETO , M. & P UTNIS , C. V. 1996. The role of magnesium in the crystallization of calcite and aragonite in porous medium. Journal of Sedimentary Research, 66, 482–491. F ERNA´ NDEZ -L UQUE , B. 2002. Estudio fenotı´pico del mutante DZF1Doar de Myxococcus xanthus y de su capacidad biomineralizadora. PhD thesis. Universidad de Granada, Spain. F ERRIS , F. G. 1997. Formation of authigenic minerals by bacteria. In: M C I NTOSH , J. M. & G ROAT , L. A. (eds) Biological –Mineralogical Interactions. Mineralogical Association of Canada. Short Course Series, Ottawa, Ontario, 25, 187–208. F ERRIS , F. G., B EVERIDGE , T. J. & F YFE , W. S. 1986. Iron-silica crystallite nucleation by bacteria in a geothermal sediment. Nature, 320, 609– 611. F ERRIS , F. G., F YFE , W. S. & B EVERIDGE , T. J. 1987. Bacteria as nucleation sites for authigenic minerals in a metal-contaminated lake sediment. Chemical Geology, 63, 225 –232. F ERRIS , F. G., F YFE , W. S. & B EVERIDGE , T. J. 1988. Metallic ion binding by Bacillus subtilis: implications for the fossilization of microorganisms. Geology, 16, 149–152. F OCHT , D. D. & V ERSTRAETE , W. 1977. Biochemical ecology of nitrification and denitrification. Advances in Microbial Ecology, 1, 135–214. F OLK , R. L. 1993. SEM imaging of bacteria and nanobacteria in carbonate sediments and rocks. Journal of Sedimentary Petrology, 63, 990–999. F ORTIN , D., F ERRIS , F. G. & B EVERIDGE , T. J. 1997. Surface–mediated mineral development by bacteria. In: B ANFIELD , J. F. & N EALSON , K. H. (eds) Geomicrobiology: Interactions between Microbes and Minerals. Mineralogical Society of America, Reviews in Mineralogy, Washington, DC, 35, 161– 180. F ORTIN , D., F ERRIS , F. G. & S COTT , S. D. 1998. Formation of Fe-silicates and Fe-oxides on bacterial surfaces in samples collected near hydrothermal vents on the Southern Explorer Ridge in the northeast Pacific Ocean. American Mineralogist, 83, 1399– 1408. G ARCI´ A -G UINEA , J., M ARTINEZ -F RIAS , J., G ONZALEZ M ARTIN , R. & Z AMORA , L. 1997. Framboidal pyrites in antique books. Nature, 388, 631. G ARCI´ A -R UIZ , J. M., H YDE , S. T., C ARNERUP , A. M., C HRISTY , A. G., V AN K RANENDONK , M. J. &
BACTERIAL BIOMINERALIZATION W ELHAM , N. J. 2003. Self-assembled silica-carbonate structures and detection of ancient microfossils. Science, 302, 1194–1197. G ILLOT , J. E., P ENNER , E. & E DEN , W. J. 1974. Microstructure of Billing shale biochemical alteration products, Ottawa, Canada. Canadian Geotechnical Journal, 11, 482–489. G OLDMAN , B. S., N IERMAN , W. C. ET AL . 2006. Evolution of sensory complexity recorded in a myxobacterial genome. Proceeding of the National Academy of Sciences, USA, 103, 15 200–15 205. G ONZA´ LEZ -M UN˜ OZ , M. T. 2008. Bacterial biomineralization applied to the protection-consolidation of ornamental stone: current development and perspectives. Coalition, 15, 12–18. G ONZA´ LEZ -M UN˜ OZ , M. T., A RIAS , J. M., M ONTOYA , E. & R ODRI´ GUEZ -G ALLEGO , M. 1993. Struvite production by Myxococcus coralloides. Chemosphere, 26, 1881– 1887. G ONZA´ LEZ -M UN˜ OZ , M. T., E NTRENA , M., B EN O MAR , N., H UERTAS , F. & A RIAS , J. M. 1994. Produccio´n of syngenetic minerals with struvite by Myxococcus coralloides D. Geomicrobiology Journal, 26, 279–283. G ONZA´ LEZ -M UN˜ OZ , M. T., B EN O MAR , N., M ARTI´ NEZ C AN˜ AMERO , M., R ODRI´ GUEZ -G ALLEGO , M., L O´ PEZ G ALINDO , A. & A RIAS , J. M. 1996. Struvite and calcite crystallization induced by cellular membranes of Myxococcus xanthus. Journal of Crystal Growth, 163, 434– 439. G ONZA´ LEZ -M UN˜ OZ , M. T., M ERROUN , M. L., B EN O MAR , N. & A RIAS , J. M. 1997. Biosorption of uranium by Myxococcus xanthus. International Biodeterioration and Biodegradation, 40, 107– 114. G ONZA´ LEZ -M UN˜ OZ , M. T., B EN C HEKROUN , K., B EN A BOUD , A., A RIAS , J. M. & R ODRI´ GUEZ -G ALLEGO , M. 2000. Bacterially induced Mg-calcite formation: Role of Mg2þ in development of crystal morphology. Journal of Sedimentary Petrology, 70, 559–564. G ONZA´ LEZ -M UN˜ OZ , M. T., F ERNA´ NDEZ -L UQUE , B. ET AL . 2003. Precipitation of barite by Myxococcus xanthus: possible implications for the biogeochemical cycle of barium. Applied and Environmental Microbiology, 69, 5722– 5725. G OULD , W. D., F RANCIS , M., B LOWES , D. W. & K ROUSE , H. R. 1997. Biomineralization: microbiological formation of sulfide minerals. In: M C I NTOSH , J. M. & G ROAT , L. A. (eds) Biological –Mineralogical Interactions. Mineralogical Association of Canada, Short Course Series, Ottawa, Ontario, 25, 169–186. H AFERBURG , G., K LOESS , G., S CHMITZ , W. & K OTHE , E. 2008. ‘Ni-struvite’ – A new biomineral formed by a nickel resistant Streptomyces acidiscabies. Chemosphere, 72, 517–523. H ALLBERG , R. & F ERRIS , F. G. 2004. Biomineralization by Gallionella. Geomicrobiology Journal, 21, 325– 330. H AMMES , F., B OON , N., DE V ILLIERS , J., V ERSTRAETE , W. & S ICILIANO , S. D. 2003. Strain-specific ureolytic microbial calcium carbonate precipitation. Applied and Environmental Microbiology, 69, 4901–4909. H ESS , B., M ETZGER , R. M., A CKERMANN , D., M ONTANDON , A. & J AEGER , P. 1994. Infection-induced stone formation in renal allograft. American Journal of Kidney Disease, 24, 868–872.
47
H O¨ FLE , G. & R EICHENBACH , H. 2005. Epothilone, a myxobacterial metbabolite with promising antitumor activity. In: C RAGG , G. M., K INGSTON , D. G. & N EWMAN , D. (eds) Anticancer Agents from Natural Products. Taylor & Francis, Boca Raton, FL, 413– 450. H OOK , L. A. 1977. Distribution of myxobacteria in aquatic habitats in an alkaline bog. Applied and Enviromental Microbiology, 34, 333–335. H UBER , C. & W A¨ CHTERSHA¨ USER , G. 1998. Peptides by activation of amino acids with CO on (Ni,Fe)S surfaces: implications for the origin of life. Science, 281, 670–672. H YACINTHE , C., B ONNEVILLE , S. & V AN C APPELLEN , P. 2008. Effect of sorbed Fe(II) on the reduction kinetics of 6-line ferrihydrite and amorphous ferric phosphate by Shewanella putrefaciens. Geomicrobiology Journal, 25, 181– 192. I IZUKA , T., J OJIMA , Y., F UDOU , R., H IRAISHI , A., A HN , J. W. & Y AMANAKA , S. 2003. Plesiocystis pacifica gen. nov., sp. nov., a marine myxobacterium that contains dihydrogenated menaquinone, isolated from the Pacific coasts of Japan. International Journal of Systematic Evolutional Microbiology, 50, 205–209. J EONG , B. C., H AWES , C., B ONTHRONE , K. M. & M ACASKIE , L. E. 1997. Localization of enzymatically enhanced heavy metal accumulation Citrobacter sp. and metal accumulation in vitro by liposomes containing entrapped enzyme. Microbiology, 43, 2497– 2507. J IMENEZ -L O´ PEZ , C., C ABALLERO , E., H UERTAS , F. J. & R OMANEK , C. S. 2001. Chemical, mineralogical and isotope behavior, and phase transformation during the precipitation of calcium carbonate minerals from intermediate ionic solutions at 25 8C. Geochimica et Cosmochimica Acta, 65, 3219–3231. J IMENEZ -L O´ PEZ , C., R ODRIGUEZ -N AVARRO , A., D OMINGUEZ -V ERA , J. M. & G ARCIA -R UIZ , J. M. 2003. Influence of lysozyme on the precipitation of calcium carbonate: a kinetic and morphologic study. Geochimica et Cosmochimica Acta, 67, 1667–1676. J ROUNDI , F., M ERROUN , M. L., A RIAS , J. M., R OSS´ LEZ BERG , A., S ELENSKA -P OBELL , S. & G ONZA M UN˜ OZ , M. T. 2007. Spectroscopic and microscopic characterization of uranium biomineralization in Myxococcus xanthus. Geomicrobiology Journal, 24, 441– 449. K AISER , D. 2003. Coupling cell movement to multicellular development in myxobacteria. Nature Reviews Microbiology, 1, 45– 54. ¸ IFTCIOGLU , N. 1998. Nanobacteria: K AJANDER , E. O. & C An alternative mechanism for pathogenic intra- and extracellular calcification and stone formation. Proceedings of the National Academy of Sciences USA, 95, 8274–8279. K ARTHIKEYAN , S. & B EVERIDGE , T. J. 2002. Pseudomonas aeruginosa biofilms react with and precipitate toxic soluble gold. Environmental Microbiology, 4, 667– 675. K ASHEFI , K. & L OVLEY , D. R. 2003. Extending the upper temperature limit for life. Science, 301, 934. K ASTING , J. F. & S IEFERT , J. L. 2002. Life and the evolution of Earth’s atmosphere. Science, 296, 1066– 1068. K AWANO , M. & T OMITA , K. 2001. Geochemical modelling of bacterially induced mineralization of
48
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
schwertmannite and jarosite in sulfuric acid spring water. American Mineralogist, 96, 1156–1165. K ERR , R. A. 1997. Life goes to extremes in the deep Earth – and elsewhere? Science, 276, 703–704. K IRKLAND , B. L., L YNCH , F. L., R AHNIS , M. A., F OLK , R. L., M OLINEUX , I. J. & M C L EAN , R. J. C. 1999. Alternative origins for nannobacteria-like objects in calcite. Geology, 27, 347– 350. K NORRE , H. & K RUMBEIN , W. E. 2000. Bacterial calcification. In: R IDING , R. E. & A WRAMIK , S. M. (eds) Microbial Sediments. Springer-Verlag, Berlin, 25–31. K OMEILI , A., V ALI , H., B EVERIDGE , T. J. & N EWMAN , D. K. 2004. Magnetosome vesicles are present before magnetite formation, and MamA is required for their activation. Proceedings of the National Academy of Sciences USA, 101, 3839–3844. K ONHAUSER , K. O. 1998. Diversity of bacterial iron mineralization. Earth Science Reviews, 43, 91–121. K ONHAUSER , K. O., H AMADE , T., R AISWELL , R., M ORRIS , R. C., F ERRIS , F. G., S OTHAM , G. & C ANFIELD , D. E. 2002. Could bacteria have formed the Precambrian banded iron formations? Geology, 30, 1079– 1082. K RUMBEIN , W. E. 1979. Photolithotrophic and chemoorganotrophic activity of bacteria and algae as related to beachrock formation and degradation (Gulf of Aqaba, Sinai). Geomicrobiology Journal, 1, 139– 203. L ABRENZ , M., D RUSCHEL , G. K. ET AL . 2000. Formation of sphalerite (ZnS) deposits in natural biofilms of sulfate-reducing bacteria. Science, 290, 1744– 1747. L EE , J. H., K IM , M. G. ET AL . 2007. Biogenic formation of photoactive arsenic– sulfide nanotubes by Schewanella sp. strain HN-41. Proceedings of the National Academy of Sciences USA, 104, 20 410–20 415. L ITTLE , B. J., W AGNER , P. A. & L EWANDOWSKI , Z. 1997. Spatial relationships between bacteria and mineral surfaces. In: B ANFIELD , J. F. & N EALSON , K. H. (eds) Geomicrobiology: Interactions between Microbes and Minerals. Mineralogical Society of America, Reviews in Mineralogy, Washington, DC, 35, 123–159. L OWENSTAM , H. A. & W EINER , S. 1989. On Biomineralization. Oxford University Press, New York. L OWER , S. K., H OCHELLA , M. F. & B EVERIDGE , T. J. 2001. Bacterial recognition of mineral surfaces: nanoscale interactions between Shewanella and a-FeOOH. Science, 292, 1360– 1363. M ACASKIE , L. E. & B ASNAKOVA , G. 1998. Microbiallyenhanced chemisorption of heavy metals (MECHM): a method for the bioremediation of solutions containing long-lived isotopes of neptunium and plutonium. Environmental Science and Technology, 32, 184–187. M ANCINELLI , R. L., F AHLEN , T. F., L ANDHEIM , R. & K LOVSTAD , M. R. 2004. Brines and evaporites: analogs for Martian life. Advances in Space Research, 33, 1244– 1246. M ANN , S. 2001. Biomineralization: principles and concepts in bioinorganic materials chemistry. Oxford University Press, Oxford. M ANN , S., H EYWOOD , B. R., R AJAM , S. & B IRCHALL , J. D. 1988. Controlled crystallization of CaCO3 under stearic acid monolayers. Nature, 334, 692–695.
M ARTINEZ , R. J. M., B EAZLEY , J., T EILLEFERT , M., A RAKAKI , A. K., S KOLNICK , J. & S OBECKY , P. A. 2007. Aerobic uranium (VI) bioprecipitation by metal-resistant bacteria isolated from radionuclideand metal-contaminated subsurface soils. Environmental Microbiology, 9, 3122– 3133. M C I NTOSH , J. M. & G ROAT , L. A. (eds) 1997. Biological– Mineralogical Interactions. Mineralogical Association of Canada, Short Course Series, Ottawa, Ontario, 25. M C K AY , D. S., G IBSON , E. K. ET AL . 1996. Search for past life on Mars: Possible relic biogenic activity in Martian meteorite ALH84001. Science, 273, 924–930. M ERROUN , M. L., B EN O MAR , N., G ONZA´ LEZ -M UN˜ OZ , M. T. & A RIAS , J. M. 1998. Myxococcus xanthus biomass as biosorbent for lead. Journal of Applied Microbiology, 84, 63– 67. M ERROUN , M. L., B EN O MAR , N., A LONSO , E., A RIAS , J. M. & G ONZA´ LEZ -M UN˜ OZ , M. T. 2001. Silver sorption to Myxococcus xanthus biomass. Geomicrobiology Journal, 18, 183– 192. M ERROUN , M. L., B EN C HEKROUN , K., A RIAS , J. M. & G ONZA´ LEZ -M UN˜ OZ , M. T. 2003. Lanthanum fixation by Myxococcus xanthus: cellular location and extracellular polysaccharide observation. Chemosphere, 52, 113– 120. M ERROUN , M. L., N EDELKOVA , M., R OSSBERG , A., H ENNIG , C. & S ELENSKA -P OBELL , S. 2006. Interaction mechanisms of uranium with bacterial strains isolated from extreme habitats. Radiochimica Acta, 94, 723– 729. M ERROUN , M. L. & S ELENSKA -P OBELL , S. 2008. Bacterial interactions with uranium: and environmental perspective. Journal of Contaminant Hydrology, 102, 285–295. M OJZSIS , S. J., A RRHENIUS , G., M C K EEGAN , K. D., H ARRISON , T. M., N UTMAN , A. P. & F RIEND , C. R. L. 1996. Evidence for life on Earth before 3,800 million years ago. Nature, 384, 55–59. M ORRIS , R. V., K LINGELHO¨ FER , G. ET AL . 2004. Mineralogy at Gusev Crater from the Mo¨ssbauer spectrometer on the Spirit rover. Science, 305, 833–836. M ULLIN , J. W. 1992. Crystallization. 3rd edn. Butterworth, London. N EALSON , K. H. & S TAHL , D. H. 1997. Microorganisms and biogoechemical cycles: what can we learn from stratified communities? In: B ANFIELD , J. F. & N EALSON , K. H. (eds) Geomicrobiology: Interactions between Microbes and Minerals. Mineralogical Society of America, Reviews in Mineralogy, Washington, DC, 35, 5 –34. N EDELKOVA , M., M ERROUN , M. L., R OSSBERG , A., H ENNIG , C. & S ELENSKA -P OBELL , S. 2007. Microbacterium isolates from the vicinity of a radioactive waste depository and their interactions with uranium. FEMS Microbiology Ecology, 59, 694– 705. N ELSON , Y. M., L ION , L. W., G HIORSE , W. C. & S HULER , M. L. 1999. Production of biogenic Mn oxides by Leptothrix discophora SS-1 in a chemically defined growth medium and evaluation of their Pb adsorption characteristics. Applied and Environmental Microbiology, 65, 175–180. N EWMAN , D. K. & B ANFIELD , J. F. 2002. Geomicrobiology: how molecular-scale interactions underpin biogeochemical systems. Science, 296, 1071– 1077.
BACTERIAL BIOMINERALIZATION N ORDSTROM , D. K. & S OUTHAM , G. 1997. Geomicrobiology of sulfide mineral interactions. In: B ANFIELD , J. F. & N EALSON , K. H. (eds) Geomicrobiology: Interactions between Microbes and Minerals. Mineralogical Society of America. Reviews in Mineralogy, Washington, DC, 35, 361– 390. N ORTHUP , D. E. & L AVOIE , K. H. 2001. Geomicrobiology of caves: a review. Geomicrobiology Journal, 18, 199–222. P AYTAN , A., K ASTNER , M. & C HAVEZ , F. 1996. Glacial to interglacial fluctuations in productivity in the equatorial Pacific as indicated by marine barite. Science, 274, 1355–1357. P EDERSEN , K. 2000. Exploration of deep intraterrestrial microbial life: current perspectives. FEMS Microbiology Letters, 185, 9 –16. P E´ REZ -G ARCI´ A , I., R IVADENEYRA , M. A. & R AMOS C ORMENZANA , A. 1989. The influence of pH on struvite formation by bacteria. Chemosphere, 18, 1633–1638. P ETRIE , L., N ORTH , N. N., D OLLHOPF , S. L., B ALKWILL , D. L. & K OSTKA , J. E. 2003. Enumeration and characterization of iron(III)-reducing microbial communities from acidic subsurface sediments contaminated with uranium(VI). Applied and Environmental Microbiology, 69, 7467–7479. P HILLIPS , E. J. P., L ANDA , E. R. & K RAEMER , T. 2001. Sulfate-reducing bacteria release barium and radium from naturally occurring radioactive material in oil-field barite. Geomicrobiology Journal, 18, 167–182. P HOENIX , V. R., K ONHAUSER , K. O., A DAMS , D. G. & B OTTRELL , S. H. 2001. Role of biomineralization as an ultraviolet shield: implications for Archaean life. Geology, 29, 823– 826. P O´ SFAI , M., B USECK , P. R., B AZYLINSKI , D. A. & F RANKEL , R. B. 1998. Reaction sequence of iron sulfide minerals in bacteria and their use as biomarkers. Science, 280, 880– 883. Q UIGLEY , R. M., Z AJIC , J. E., M C K YES , E. & Y ONG , R. N. 1973. Oxidation and heave of black shale. Journal of Soil Mechanics Foundation Division, 99, 417–421. R EICHENBACH , H. 1993. Biology of the myxobacteria: ecology and taxonomy. In: D WORKIN , M. & K AISER , D. (eds) Myxobacteria II. American Society for Microbiology, Washington, DC, 13– 62. R EICHENBACH , H. & H O¨ FLE , G. 1999. Myxobacteria as producers of secondary metabolites. In: G RABLEY , S. & T HIERICKE , R. (eds) Drug Discovery from Nature. Springer Verlag, Berlin, 149– 179. R IDING , R. 2000. Microbial carbonates: the geological record of calcified bacterial–algal mats and biofilms. Sedimentology, 47 (Suppl. 1), 179–214. R IVADENEYRA , M. A., P E´ REZ -G ARCI´ A , I. & R AMOS C ORMENZANA , A. 1992. Influence of ammonium ion on bacterial struvite production. Geomicrobiology Journal, 10, 125–137. R IVADENEYRA , M. A., D ELGADO , R., P A´ RRAGA , J., R AMOS -C ORMENZANA , A. & D ELGADO , G. 2006. Precipitation of minerals by 22 species of moderately halophilic bacteria in artificial marine salts media: Influence of salt concentration. Folia Microbiologica, 51, 445–453.
49
R ODRIGUEZ -N AVARRO , C., R ODRIGUEZ -G ALLEGO , M., B EN C HEKROUN , K. & G ONZALEZ -M UN˜ OZ , M. T. 2003. Conservation of ornamental stone by Myxococcus xanthus-induced carbonate biomineralization. Applied and Environmental Microbiology, 69, 2182– 2193. R ODRIGUEZ -N AVARRO , C., J IMENEZ -L OPEZ , C., R ODRIGUEZ -N AVARRO , A., G ONZALEZ -M UN˜ OZ , M. T. & R ODRIGUEZ -G ALLEGO , M. 2007. Bacterially mediated mineralization of vaterite. Geochimica et Cosmochimica Acta, 71, 1197–1213. R OTHSCHILD , L. J. 1990. Earth analogs for Martian life: Microbes in evaporites, a new model system for life on Mars. Icarus, 88, 246–260. R OTHSCHILD , L. J. & M ANCINELLI , R. L. 2001. Life in extreme environments. Science, 409, 1092–1101. R UIZ , C., M ONTEOLIVA -S A´ NCHEZ , M., H UERTAS , F. & R AMOS -C ORMENZANA , A. 1988. Calcium carbonate precipitation by several Myxococcus. Chemosphere, 17, 835 –838. S AMUELL , C. T. & K ASIDAS , G. P. 1995. Biochemical investigations in renal stone formers. Annals Clinical Biochemistry, 32, 112 –122. S CHOPF , J. W. 1993. Microfossils of the Early Archean Apex chert: new evidence of the antiquity of life. Science, 260, 640–646. S CHULTZE -L AM , S., F ORTIN , D., D AVIS , B. S. & B EVERIDGE , T. J. 1996. Mineralization of bacterial surfaces. Chemical Geology, 132, 171–181. S CHWARTZMAN , D. W. & V OLK , T. 1989. Biotic enhancement of weathering and the habitability of Earth. Nature, 340, 457–460. S EBASTIAN -P ARDO , E., G ARCI´ A -C ERVIGO´ N , A. & R ODRIGUEZ -G ALLEGO , M. 1983. El yacimiento de azufre nativo de Benamaurel (Granada): genesis y mineralogia de las arcillas asociadas. Revista de la Academia de Ciencias de Granada, 2, 73– 91. S HIMKETS , L. J. 1999. Intercellular signalling during fruiting-body development of Myxococcus xanthus. Annual Review of Microbiology, 53, 525 –549. S HIMKETS , L. J. & W OESE , C. R. 1992. A phylogenetic analysis of the myxobacteria: basis for their classification. Proceedings of the National Academy of Sciences USA, 89, 9459– 9463. S HIMKETS , L. J., R EINCHENBACH , H. & D WORKIN , M. 2005. The Myxobacteria. In: D WORKIN , M. (ed.) The Prokaryotes. 3rd ed. Springer, New York, 31–115. S IERING , P. L. 1998. The double helix meets the crystal lattice: the power and pitfalls of nucleic acid approaches for biomineralogical investigation. American Mineralogist, 83, 1593– 1607. S ILLITOE , R. H., F OLK , R. L. & S ARIC , N. 1996. Bacteria as mediators of copper sulfide enrichment during weathering. Science, 272, 1153–1155. S OUTHAM , G. & D ONALD , R. 1999. A structural comparison of bacterial microfossils vs. ‘nanobacteria’ and nanofossils. Earth Science Reviews, 48, 251–264. S TRZELCZYK , A. B. 1981. Stone. In: R OSE , A. H. (ed.) Microbial Biodeterioration. Academic Press, London, 61–80. T HOMPSON , J. B. & F ERRIS , F. G. 1990. Cyanobacterial precipitation of gypsum, calcite, and magnesite from natural alkaline lake water. Geology, 18, 995 –998.
50
´ LEZ-MUN ˜ OZ ET AL. M. T. GONZA
U RRUTIA , M. M. & B EVERIDGE , T. J. 1994. Formation of fine-grained metal and silicate precipitates on a bacterial surface (Bacillus subtilis). Chemical Geology, 116, 261– 280. U RZI , C., G ARCIA -V ALLES , M., V ENDRELL , M. & P ERNICE , A. 1999. Biomineralization processes on rock and monument surfaces observed in field and laboratory conditions. Geomicrobiology Journal, 16, 39–54. V AN L ITH , Y., W ARTHMANN , R., V ASCONCELOS , C. & M C K ENZIE , J. A. 2003. Microbial fossilization in carbonate sediments: a result of the bacterial surface involvement in dolomite precipitation. Sedimentology, 50, 237–245. V ASCONCELOS , C., M C K ENZIE , J. A., B ERNASCONI , S., G RUJIC , D. & T IEN , A. J. 1995. Microbial mediation as a possible mechanism for natural
dolomite formation at low temperatures. Nature, 377, 220–222. V REELAND , R. H., R OSENZWEIG , W. D. & P OWERS , D. W. 2000. Isolation of a 250 million-year-old halotolerant bacterium from a primary salt crystal. Nature, 407, 897 –900. W EBSTER , A. & M AY , E. 2006. Bioremediation of weathered-building stone surfaces. TRENDS in Biotechnology, 24, 255–260. W HITWORTH , D. E. (ed.) 2008. Myxobacteria: Multicellularity and Differentiation. American Society for Microbiology, ASM Press, Washington, DC. W IECHERT , U. H. 2002. Earth’s early atmosphere. Science, 298, 2341–2342. W IERZCHOS , J., A SCASO , C. & M C K AY , C. P. 2006. Endolithic cyanobacteria in halite rocks from the hyperarid core of the Atacama Desert. Astrobiology, 6, 415–422.
Calcium carbonate precipitation by cyanobacterial polysaccharides M. DITTRICH1* & S. SIBLER1 1
Swiss Federal Institute for Environmental Science and Technology, EAWAG and Swiss Federal Institute of Technology, ETH, Limnological Research Center, Seestrasse 79, 6047 Kastanienbaum, Switzerland *Present address: University of Toronto, Department of Physical and Environmental Sciences, Toronto, Scarborough, Canada (e-mail:
[email protected]) Abstract: Cyanobacteria have been recognized as key players in the precipitation of calcium carbonate in marine and freshwater systems. These bacteria increase pH, (as a result of photosynthetic activity) and also produce extracellular polysaccharides, which act as binding sites for Ca2þ and CO22 3 . Both processes influence the morphology and the mineralogy of the carbonate minerals. In order to clarify the role of polysaccharides of picocyanobacteria upon calcium carbonate precipitation, both their buffering capacity and ability to induce precipitation need to be investigated. In this experimental study, we characterized the polysaccharides of three unicellular autotrophic picocyanobacterial Synechococcus-type strains by potentiometric titration and infrared spectroscopy. Potentiometric titrations were conducted to determine the total buffering capacity. The nature and concentration of active sites of the polysaccharides was clarified with the aid of potentiometric titration and spectral analysis of an aqueous cellular suspension. Precipitation experiments with polysaccharides of different strains allowed an estimation of their potential to precipitate calcium carbonate. The results presented here indicate that polysaccharides from cyanobacteria have a strong potential to exchange protons with their surrounding environment. Precipitation experiments demonstrated that extracellular polysaccharides of all the strains studied able to precipitate calcium carbonate.
Extracellular polymeric substances of microbial origin are an important class of polymeric materials that have been involved in different processes including biofilm development and mineral precipitation (Decho 1990; Riding 2000; Decho et al. 2005; Shiraishi et al. 2008). Substantial work has been carried out over the last decade on the properties and distribution of microbial extracellular polymeric substances (Decho 2000). Mechanisms with which they regulate various processes, including particle formation, sedimentation, organic carbon mineralization, and cycling of dissolved metals have been investigated and adopted in sediment ecology and biochemistry (see Bhaskar & Bhosle 2005 for review). Extracellular polymeric substances (EPS) consist of various organic substances, mostly of extracellular polysaccharides, but also of uronic acids, proteins, nucleic acids and lipids (Nichols & Nichols 2008). Extracellular polysaccharides produced by microorganisms can be tightly bound (cell attached or capsular) or loosely adhered (slime type, free ore released) to cells, or exist as free dissolved matter (Nielsen & Jahn 1999; Bhaskar & Bhosle 2005). From a physicochemical viewpoint, an EPS-covering on a cell surface is regarded as polyelectrolytes adsorbed onto a colloidal particle. The molecular masses of EPS range from a few thousand
to several million Daltons and comprise various functional groups including carboxyl, amino and phosphate (Wingender et al. 1999). Microbial extracellular polymeric substances strongly impacted both mineralogy and morphology of precipitated calcium carbonate crystals in the laboratory experiments and in stromatolites and soils in the natural environment (Kawaguchi & Decho 2002; Braissant et al. 2003). Microbial free ore released extracellular polymeric substances are also known to impact CaCO3 precipitation with their decomposition, organo-mineralization (Trichet & Defarge 1995) and their ability to bind cations (Dupraz & Visscher 2005). Following organomineralization processes, extracellular polymeric substances is re-structured in a way that new binding sites will be created, these binding sites are templates for CaCO3-formation will be created (Trichet & Defarge 1995). Templates are most probable the reorganized acid binding sites, which enable CaCO3 to precipitate (Dupraz & Visscher 2005). Microbial EPS can promote CaCO3 precipitation via the binding of Ca2þ to negatively charged functional groups of the EPS. This can also reduce the activation energy barrier that normally retards spontaneous nucleation (through the uptake and retention of metal ions), thereby reducing the critical saturation state at which precipitation can begin
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 51– 63. DOI: 10.1144/SP336.4 0305-8719/10/$15.00 # The Geological Society of London 2010.
52
M. DITTRICH & S. SIBLER
(Schultze-Lam et al. 1996; Ferris & Lowson 1997). Once all EPS binding places are occupied, the solution saturation state increases if there is a continued rise in local concentration of dissolved Ca2þ and HCO2 3 (Arp et al. 2003). The decomposition of 2þ ions into EPS causes a release of HCO2 3 and Ca the local environment, thus increasing the saturation state regarding relation to calcium carbonate and promotes precipitation. Continued precipitation will reduce the saturation state as Ca2þ and HCO2 3 are removed from the solution (Arp et al. 2003). Uncharacterized EPS produced by Desulfovibrio desulfuricans G20, (a strain of sulphate reducing bacteria, SRB), altered the CaCO3 mineral morphology (Bosak & Newman 2005). The influence of EPS on calcium carbonate precipitation is supposed to be based on their calcium binding capacity. The EPS of three different SRB strains have recently been characterized and their calcium binding capacity has been estimated (Braissant et al. 2007). Cyanobacteria have been observed to precipitate CaCO3 in a range of environments (Dittrich et al. 2004; Lee et al. 2004). As shown by many researchers, different cyanobacterial species exhibit different calcification fabrics (e.g. Pentecost 1991; Merz 1992). Furthermore, cyanobacteria have been known as potential EPS producers for a long time (De Philippis et al. 1991). This has highlighted the potential of cyanobacterial EPS from strains such as Cyanospira capsulata and Aphanothece halophytica GR02 for biotechnological applications (see for a review (De Philippis et al. 2001). It is assumed that acidic EPS probably play an important role in crystal nucleation, although the effect of Ca2þ-binding by acidic EPS on sustaining CaCO3 precipitation is minor in freshwater biofilms (Shiraishi et al. 2008). Therefore, EPS can more than likely influence the formation of tufa fabrics by providing nucleation sites, as can the cell surfaces of heterotrophic bacteria (e.g. Ferris & Beveridge 1984; Bosak & Newman 2003). Picocyanobacteria are small unicellular cyanobacteria with a cell diameter of 0.2 to 2 mm, commonly found in soils and freshwater. They contribute significantly to the overall primary production in ecosystems of all climatic zones (Agawin et al. 2000; Stockner et al. 2000; Bell & Kalff 2001). Picocyanobacteria have also been observed in mats, biofilms in hot springs, as well as in hypersaline ponds (Ferris et al. 1996; GarciaPichel et al. 1998; Ward et al. 1998; Miller & Castenholz 2000). Robbins & Blackwelder hypothesized that calcium carbonate crystals can be nucleated on both the organics and cell membranes of picoplankton cells (Robbins & Blackwelder 1992). Interestingly, picocyanobacteria from both the pelagic and biofilms in the euphotic zone of
temperate-zone lakes belong to the same evolutionary lineage of cyanobacteria (Becker et al. 2004). Knowledge about EPS compositions of cyanobacteria is crucial in order to understand biofilm formations, cell attachment to surfaces and cell– mineral interactions (de Winder et al. 1999; Hirst et al. 2003). Until now, the functional groups of extracellular polysaccharides of picocyanobacteria of Synechococcus-type have not been investigated in that respect. Cyanobacterial extracellular polymers are characterized by a presence of different proteins, uronic acids, pyruvic acid, and sulphate groups (Parikh & Madamwar 2006). The total buffering capacity plays an extremely important role in this respect as it reflects the binding capacity of polymers. Previous work has shown that EPS in cyanobacterial mats probably plays an important role in carbonate nucleation (Shiraishi et al. 2008). This important geochemical attribute of cyanobacteria has not been assessed in cyanobacterial cultures obtained from freshwater. Despite EPS ubiquitous distribution, there is still a great lack of knowledge concerning the diversity of extracellular polysaccharides of different picocyanobacterial strains and about those EPS components that may be responsible for calcium carbonate precipitation. The aim of this study is three-fold: to determine the total buffering capacity of the extracellular polysaccharides of three different strains of picocyanobacteria using potentiometric acid-base titrations; to characterize the functional groups by infrared spectroscopy; and to investigate their potential to precipitate calcium carbonate using batch precipitation experiments.
The isolation of extracellular polysaccharides PCC 7942, Syn. Green and Syn. Red picocyanobacteria Synechococcus-type strains were used in all experiments presented here. The PCC 7942 strain was obtained from the Pasteur Institute in Paris, France. The Syn. Green and Red strains were isolated from the water column of two stratified lakes: the Plo¨ner See and Lago Maggiore (courtesy of C. Callieri). Cells were grown as a batch culture using modified Z/10 medium, under a 14 h/10 h light/dark condition, with a light intensity of c. 10 mE m22 s21 (Dittrich & Sibler 2005). Different growth conditions and physical parameters are known to affect the production and properties of extracellular polymeric substances in algae and cyanobacteria (De Philippis et al. 1991). In order to generate reproducible experimental results that reflect the environmental conditions in biofilms, cyanobacterial cells in the stationary growth phase were used for the polysaccharides isolations.
CARBONATE, CYANOBACTERIA AND POLYSACCHARIDES
The cultures were harvested by centrifugation at 7000 rpm for 10 min at 20 8C, washed with 0.001 M EDTA and three times washed with 0.1 M NaNO3. The cells were centrifuged under the same conditions listed above. The cells were finally batched and re-suspended in the 0.1 M NaNO3 to a minimum of concentration of around 0.2 g L21 of bacteria. Our preliminary study showed that this protocol is the most efficient in terms of EPS quantities. The polysaccharides present in the cultures were extracted with phenol according to the conventional procedures for extracting bacterial polysaccharides. These samples, consisting of cells, were placed in a 2.0 mL Eppendorf-cap and centrifuged for one minute at 13000 rpm. After discarding the supernatant, another 2 mL of the sample was added into Eppendorf-caps and again centrifuged. In order to elute salts, the pellets were washed with Phosphate-Buffer (pH 7), centrifuged again and the supernatant was discarded. To induce separation, the pellet was thoroughly mixed with 0.5 mL Phosphate-Buffer and 0.5 mL Phenol (80%) and incubated in heated water bath for 20 minutes (Blaschek 1991). After every 5 minutes of heating the samples were thoroughly mixed and put on ice for one minute. After 20 minutes of heat exposure the samples were stored on ice for 5 minutes and centrifuged for 10 minutes at 5000 rpm. The supernatant was carefully transferred into a dialysis membrane (Sigma) and closed with labelled clamps. The membrane was stored overnight in nanopure water at þ4 8C. The next day the samples were transferred into 1.5 mL Eppendorf caps and dried in a freeze dryer overnight (Blaschek 1991). Finally, the dried samples were dissolved in 50–250 mL of sterile nanopure water and stored at 220 8C in 12 mL Eppendorf tubes.
Potentiometric titration and data analysis Titration solutions The solutions were degassed with N2 for 20 minutes to dissipate O2 and CO2. The concentration of NaNO3 used in this experiment was 0.1 M. The NaOH solution was prepared according to the following method: c. 0.1 M of solution was prepared from NaOH using degassed 18 MV water. The exact NaOH concentration was determined prior to the titration experiment with a relative standard deviation of 1% (Dittrich & Sibler 2005).
Titrator setting Deprotonation constants and surface site concentrations were determined from acid –base titrations of extracted polysaccharides from the three bacterial
53
strains in a background electrolyte of 0.01 M NaNO3. All titrations were performed in a glass vessel with a lid as part of a Metrohm GP 736 Titrino unit interfaced by Titrino software TITRINET to a personal computer. Two separate buret exchange units (20 mL and 10 mL) were used, one for the acid and one for the base. We also used a Metrohm titrator vessel lid. The temperature was recorded with a temperature sensor; the error of the temperature probe was 0.1 8C. The pH electrode was three-point calibrated with buffers (pH 4, 7 and 10) before each experiment, and the slope was constant at 99% of the Nernst value. The Titrino unit was programmed with a dynamic mode (DYN) for the titration, which adds the variable amount of titrant according to the pH changes: the smaller volume of titrant was added at the slope of the pH curve. The successive titrant additions were only made when the signal drift reached 10 mV min21.
Potentiometric titration The titrator electrode was first calibrated as mentioned above. The optical density of the bacterial suspension in the NaNO3 electrolyte solution was measured prior to each extraction and titration run. In order to determine the concentration of bacteria (mg L21) and the bacterial cell numbers (cells L21), the measured absorbance was compared to a prepared calibration curve. The dry weight of bacteria was defined by drying at 65 8C until a constant weight was attained. The absorbance used is ranged between 0.41 –0.81, which corresponds to 0.063– 0.122 g of bacteria L21. A known amount of suspension, c. 50 mL, was then transferred to the titration vessel, which was immediately attached to the lid setup connected with the N2 gas line. A magnetic stir bar was also added to the vessel. The whole system was then degassed for 30 –40 min to exclude atmospheric CO2. Following the degassing procedure a positive pressure of N2 was maintained by allowing a gentle flow of N2 into the headspace during the titration. The EPS aliquot was then titrated quickly to pH ¼ 2.9 with 0.1 M HCl. The buret exchange unit was subsequently changed and the titration of the aliquot (with NaOH up to pH ¼ 10) began. The total time for each titration was c. 40 min. Some titrations were reversed by conducting an acid titration, immediately after the base titration. The results of reversed titrations were not significantly different from the forward titrations suggesting reversibility of the proton adsorption – desorption reactions. The titration data were analyzed using the linear programming method (LPM) or the so-called pKa spectrum method as proposed by Brassard and
54
M. DITTRICH & S. SIBLER
others (1990; Sokolov et al. 2001). Proton dissociation from cell surface ligands can be described by the following equation: 0
þ
HL , H þ L
[H þ ][L ] [HL0 ]
(2)
where Ka is the dissolution constant for HL 0. The useful transformation of the acid base raw data for the j’th addition of acid or base is the charge balance expression (Brassard et al. 1990): bmeas, j ¼ CBj CAj þ [H þ ] j [OH ] j
n X i¼1
Kai LTi þS Kai þ [H þ ] j
Kai Kai þ [H þ ]j
(5)
The n 1 vector contains the ligand concentrations for each of the m sites and the m 1 vector b contains the measured charge excess CBj CAj þ [H þ ]j [OH ]j .
Infrared spectroscopy Pellets for infrared analysis were obtained by carefully grinding a mixture of 1–2 mg of polymeric substancies with 300 mg of dry KBr and then pressing them into in a 16-mm diameter mold. The pellet technique was used because most bacterial polymers were poor water-soluble and films could not be prepared. Fourier transform infrared spectroscopy (FTIR) spectra were recorded on a Perkin Elmer instrument SPECTRUM (PE-IR) with a resolution of 1 cm21. Spectra were run in the region 400–4500 cm21. No smoothing was performed.
(3)
where CBj and CAj correspond to concentrations of base and acid for the j’th addition of titrant, [H þ]j and [OH 2]j are obtained from the measured proton concentration. As described previously, the charge excess bmeas,i can be calculated as a function of measured [H þ] and adjustable (Ka and L T) speciation parameters (Martinez et al. 2002) as:
bcalc, j ¼
aij ¼
(1)
where L 2 is the deprotonated binding site with a negative charge and H þ is a proton in solution, whose activity in the bulk solution was measured with a pH electrode. The concentration of protonated and deprotonated surface sites can be quantified with the corresponding mass action equation: Ka ¼
sites and m additions of titrant is:
(4)
where S is a constant term analogous to the acid neutralizing capacity or the initial protonation state of the surface (Brassard et al. 1990; Cernik et al. 1995). The surface sites are considered as a sum of n monoprotonic ligands [L 2] with dissociation constants Kai and total concentrations L Ti . In practice, S allows a modeling positive charge on the surface. Equation (3) could be solved by fixing the pKa values as a grid from a minimum to maximum value at fixed step sizes (Cox et al. 1999). The ligand concentration associated with each pKa value is assigned a positive value where zero is a possible result; the result is the so-called pKa spectrum. The pKa spectrum approach is used here to determine the best fit of Kai/L Ti pairs, with pKa values fixed as a grid from 4 to 10 at fixed step sizes (0.2). Once the pKa values are selected, the matrix version of equation (3) is set up as Ax ¼ b. The entry aij in the m n matrix A for n proposed
Calcium carbonate precipitation experiments on agarose beads Agarose beads were prepared by modified method which is described in details in Strathmann et al. (2000). Briefly, two solutions were prepared: for the first solution 8 mL of Span85 was added into 200 mL of 60 8C cyclohexane. For the second solution, agarose powder was added to 98 8C nanopure water (1–5%, w/v) under constant stirring. When completely dissolved, the agarose solution was cooled down to 60 8C and then emulsified in the cyclohexane/Span85 solution (60 8C) at a stirring speed of 500 min21. After 10 minutes of stirring the water/oil emulsion was cooled down to 20 8C without stirring. The supernatant was decanted and the remaining beads washed 4 times with nanopure water. The beads can be stored in 50 mL Eppendorf tubes at 280 8C.
Calcium carbonate precipitation experiments After slowly defrosting the 12 mL Eppendorf-tubes containing the polysaccharides samples, a spatula tip of agarose beads was added to each sample. Then 37% formaldehyde was added to make 4% solution and the tubes were stored horizontally for 3 hours at room temperature to allow the polysaccharides to attach onto the beads. Finally agarose beads with attached polysaccharides were washed with phosphate buffer. Subsequently, beads were transferred into the prepared and labelled vials which contained 1– 2 mL of the mixture containing
CARBONATE, CYANOBACTERIA AND POLYSACCHARIDES
10 mL CaCl2 and 1.5 mL NaHCO3 solution. Blank experiments without polysaccharides were prepared by adding agarose beads into the mixture containing 10 mL CaCl2 and 1.5 mL NaHCO3 solution. Initial saturation index in respect to calcium carbonate (SI) is 1.96. Each day for 5 days, vials were gently agitated and samples were extracted with a sterile syringe and filtered through 0.2 mm polycarbonate filters, washed with NH3 solution (pH ¼ 8), air-dried and filters were stored in a desiccator. Dry material was then deposited onto the SEM stub with carbon tabs. The carbon layer underneath the particles allowed us to analyse the uncoated specimens. The morphology of the precipitates was characterized by scanning electron microscopy (SEM, Philips XL30, LaB6 filament) and the elementary composition of the crystals was determined qualitatively with an EDAX EDS detector.
Extraction protocol The content and composition of EPS require elucidation to clarify their role in various geochemical processes. However, the first step in the studies, the extraction protocols of extracellular polymers is a matter of debates. Comte et al. (2006a) noted that applied chemical reagents could contaminate collected EPS. Further study by Comte et al. (2007) revealed that applied chemical reactants could affect the high-pressure size exclusion chromatography fingerprint of EPS whereas physical extraction methods only affect corresponding molecular weight distributions. Additionally, the authors noted that physical means (such as centrifugation) were either inefficient for extraction or could induce significant cell lysis (e.g. heating) and contaminate the EPS. Recently, extracellular polymeric substances were extracted from aerobic granules using seven extraction methods (Adav & Lee 2008). Aerobic granules are compact bioaggregates with a compact interior structure. Ultrasound followed by the chemical reagents formamide and NaOH outperformed other methods in extracting EPS from aerobic granules of compact interior. The collected EPS revealed no contamination by intracellular substances and consisted mainly of proteins, polysaccharides, humic substances and lipids. We just started the work on the role of extracellular polymeric substances of picocyanobacteria strains in geochemical processes. The work is now in progress to compare this other extraction protocol in terms of quantities and qualities of extracellular polymeric substances (Comte et al. 2006b). More research is needed to determine the most effective protocol for extracellular polymers collections.
55
Functional groups revealed by infrared spectroscopy FTIR spectra (Fig. 4) revealed extensive homology between the samples and indicated the presence of the same functional groups mentioned in previous studies (Comte et al. 2006c; Beech & Tapper 1999). Absorption bands have been assigned to the different functional groups of the skeleton, that is, ether, carboxylic, carboxylate or sulphate groups. All of the samples analyzed in this study were characterized by a broad band above 3000 cm21 and intense absorptions of around 1650 and 1050 cm21 (Fig. 4). Characteristic absorption peaks of around 3500 –3200 cm21 reflect the stretching of the N –H bond of amino groups present in the polymers. This N –H stretching peak lies in a spectrum region occupied by a broad and strong band (3500–3000 cm21), which may be due to hydroxyl groups that are hydrogen bonded to various degrees. The weak peak at 2850 cm21 indicates the presence of saturated carbohydrates in samples of PCC strains. The C –H stretching bands between 2800 –3000 cm21 were poorly resolved and their intensities were weak. The corresponding CH2 deformation modes were located in the region 1430–1400 cm21. Protein related bands, the nCvO of amide I was present at 1650 cm21 and the region in the spectrum of polysaccharides. However, the presence of N – acetyl groups may also be manifested by the absorbance band in this range (Beech & Tapper 1999). Vibrations due to the carbohydrate backbone were common in all spectra. Strong complex absorptions, centred between 1060 –1080 cm21 for the exopolymers, are ascribed to complex vibrations of the carbohydrate skeleton, ring structures, including bending, stretching and coupling between these modes (Beech & Tapper 1999). The absorption peaks between the 1000–1200 cm21 regions ascertained the presence of gluuronic and mannuronic acids, the main carboxylic building blocks of alginate (Kazy et al. 2002). Absorption bands at a region of 1350 cm21, assigned to the nCZO of carboxylic acids, suggested that the exopolymer were acidic. The significant differences between the spectra of PCC 7942 and Syn Red are observed in the sugar/sugar phosphate region at around 950 cm21. The complex absorptions at the c. 2920 cm21 region are ascribed to the asymmetric stretching of nCZH bond of ZCH2 groups combined with that of the CH3 groups. The corresponding symmetric stretching of the same bond was found at the c. 2850 cm21 region (Beech & Tapper 1999). The observed peaks in the spectra at the 1400 to 1450 cm21 region are characteristic for the presence of carboxyl groups (Kazy et al. 2008).
56
M. DITTRICH & S. SIBLER
Fig. 1. (a) Charge excess (mM g21 of bacteria) measured by potentiometric titration for EPS from PCC 7942. Also shown are results of linear programming as lines. (b) pKa spectra determined by linear programming analysis for each of the titration curves shown in Figure 1a. The symbols of bars are correspondent to the titration curves in Figure 1a. The position of the bar reflects the pKa value and the height of the bar reflects the concentration of a binding site.
Lijour et al. (1994) reveal sulphate content of EPS via quantification of the FTIR spectra peaks at 1260–1230 cm21. These peaks usually form a slight doublet at 1250 cm21 corresponding to the OvSvO antisymmetric stretching vibrations. In our study we observed peaks at 1250 cm21 for both cyanobacterial strains. Many studies have related the structure of carrageenans to a set of infrared bands between 1000–800 cm21. Some represented pseudo-symmetric CZOZS stretching vibrations and gave structural information on the location of sulphate groups on the saccharide units. Their intensities have been measured by Rochas et al. (1986). The degree of sulphation of the polymers was perfomed using the ratio of the absorbance band at 1250–1050 cm21 with a precision of 4% (Lijour et al. 1994). Also Comte et al. 2006c
highlighted the range at ,1000 cm21 as fingerprint zone of phosphate or sulphur functional groups (Comte et al. 2006c, page 819, table 2).
Analysis of the total buffering capacity The total buffering capacity, or charge excess, and pKa spectra for the EPS derived from PCC 7942, Syn. Red and Syn. Green, following the transformation to the charge balance expression (3) of replicate titrations, are shown in Figures 1 –3. Consistent trends are observed for the titration curves in each set. It can be seen that the data is distributed in approximately equal pH steps, as required by the LPM, so that no sites are assigned preferential weight in the fitting procedure (Brassard et al. 1990).
Fig. 2. (a) Charge excess (mM g21 of bacteria) measured by potentiometric titration for EPS from Syn. Red. Also shown are results of linear programming as lines. (b) pKa spectra determined by linear programming analysis for each of the titration curves shown in Figure 2a. The symbols of bars are correspondent to the titration curves in Figure 2a. The position of the bar reflects the pKa value and the height of the bar reflects the concentration of a binding site.
CARBONATE, CYANOBACTERIA AND POLYSACCHARIDES
57
Fig. 3. (a) Charge excess (mM g21 of bacteria) measured by potentiometric titration for EPS from Syn. Green, also shown are results of linear programming as lines. (b) pKa spectra determined by linear programming analysis for each of the titration curves shown in Figure 3a. The symbols of bars are correspondent to the titration curves in Figure 3a. The position of the bar reflects the pKa value and the height of the bar reflects the concentration of a binding site.
The titration curves, which have approximately the same shape for all three strains, showed that EPS influences the buffering capacity of the electrolyte. The functional groups are de-protonating due to the addition of the base. The reproducibility of the buffering capacity of three strains is variable. The data for Syn. Green exhibit an excellent repeatability (Fig. 3a), whilst data for PCC 7942 (Fig. 1a) and Syn. Red (Fig. 2a) have a rather poor reproducibility. The variation of the buffer capacity can be caused, on one hand, by variations in batch cultures at different times and, on the other hand, by the
impact of the extractive procedure on polysaccharides. Polysaccharides were extracted from the strains’ batch cultures. The batch cultures represent a mixture of cells and the production of different strains may vary (Mata et al. 2008). Furthermore, the polymers substances of three strains have slightly different compositions, as we already observed different surface properties of the investigated strains by infrared spectroscopy (Dittrich & Sibler 2005). For PCC 7942, the site identified within the pKa range 3–4.6 is likely to correspond to a carboxylic group (Cox et al. 1999; Fowle & Fein
Fig. 4. Reflectance-absorbance FTIR spectra of extracellular polymers produced by cyanobacteria. The spectra have been vertically displaced for the sake of clarity. AU means absorbance units.
58
M. DITTRICH & S. SIBLER
2000) and had an average concentration of 0.95– 0.35 mM g21 of bacteria. The carboxyl group is a basic component of several extracellular polymers constituents (Wingender et al. 1999). For comparison, the carboxyl sites on the bacterial cell surface were reported to have pKa values of 5.17 and 5.25 (Cox et al. 1999), 4.85 + 0.31 or 4.98 + 0.16 (Dittrich & Sibler 2005) and the humic acid to have a pKa value of 4.2 (Brassard et al. 1990). The pH range of the 5.6 –6.0 sites may be attributed to carboxyl or phosphoric groups (Comte et al. 2006c). The average concentration of this site was 0.64 mM g21 of bacteria. The site at pKa 6.8 may be attributed to the phosphoric sites, e.g. triprotoic phosphoric acid has the second dissociation constant pKa value of 6.82. Phosphate groups can exist in several different forms: inorganic forms of phosphate such as orthophosphate and its oligomers, and organic species in the form of phosphate monoand diesters. The concentration of this site was 0.41 mM g21 bacteria for PCC 7942. A site at pKa 7.8–8.2, which was likely attributed by the sulphydryl groups, was identified (Hornback 1998). The sites at pKa 9.4 and 10 had a concentration of 0.09 and 0.71 mM g21 bacteria for PCC 7942 was likely attributed by phenolic or amine (Cox et al. 1999). These compounds are abundant in humic substances and amino sugars. Among the five binding sites for polysaccharides of PCC 7942, the carboxylic site has the highest concentration of 1.23 mM g21 bacteria. Liu & Fang (2002) studied the electrostatic characteristic of binding sites of EPS and found a pKa of 6.0 for carboxylic/phosphoric groups, a pKa of 7.0–7.4 for phosphoric groups, a pKa of 9.4–9.8 for amine/ phenolic groups. The corresponding results for extracellular polysaccharides extracted from Syn. Red illustrated in
Figure 2a–b. The five sites found were comparable to those found in the samples of PCC 7942 strain. The concentrations of binding sites at pKa ¼ 3– 4.2 and 10 were highest. The results of the potentiometric titrations for Syn. Green are very similar to those for strain PCC 7942 and Syn. Red (Fig. 3a). Table 1 summarized the modelled concentrations of binding sites in polysaccharides extracted from three strains. PCC 7942 and Syn. Green have the same total excess charge whereas it was five times higher for Syn. Red. It was c. 3.7 mM g21 bacteria for PCC 7942, 15.0 mM g21 bacteria for Syn. Red and 4.1 mM g21 of bacteria for Syn. Green. For comparison, the sheath of gram-negative Calotrix has 0.18 mM and Calotrix cells have 1.46 mM g21 of bacteria (Phoenix et al. 2002). Borrok et al. (2005) compared the buffering capacity of 36 different bacterial species, and it was found to be around 3.2 mM g21 of dry bacteria or 0.32 mM g21 of wet bacteria. The total concentrations of the electrostatic binding sites found in this study were comparable to those reported for natural organic matters (4–24 mM g21) using the same method (Bird & Wyman 2003). In this study, titration was carried out starting at pH ¼ 3. The pH in natural water normally ranges between 6 and 9. At pH 6 only binding sites at pKa ¼ 4.4–4.8 carry negative charge. Therefore, the charges carried at pH 6 in the three strains were 21 meq g21, 25 meq g21 and 2 1 meq g21 assuming the absence of sites carrying positive charge. These values are lower than 215.7 to 21.2 meq g 21 – EPS measured for three anaerobic sludges (Jia et al. 1996). This could be due to the negligence of those sites at pKa below 3 in this study. The deprotonation constants obtained here (see Table 1) represent functional groups similar to
Table 1. Deprotonation constants and concentrations of corresponding binding sites in EPS extracted from three strains Functional groups
EPS of picocyanobacteria PCC 7942
Carboxylic Carboxylic – phosphoric Phosphoric Sulfhydryl Amin – phenolic Hydroxyl Total
Syn. Red
Syn. Green
pKa
LT (mM g21 bact.)
pKa
LT (mM g21 bact.)
pKa
LT (mM g21 bact.)
3/3.8–4.6 5.6–6.0
0.98/0.35 0.64
3/4.2 5.2–6.4
2.94 2.65
3/4.4– 5.4 6
0.86/0.38 0.52
6.8 7.8–8.2 9.4
0.41 0.48 0.09
6.8–7.4 8.2 –
2.76 1.43 –
6.8 8 9.2
0.82 0.43 0.21
10
0.71 3.66
10
3.06 14.97
10
0.93 4.15
CARBONATE, CYANOBACTERIA AND POLYSACCHARIDES
those obtained by previous studies for both gramnegative and -positive bacteria and their polymeric substancies, see Tourney et al. 2008 for a review. The results presented here indicate that polysaccharides from cyanobacteria have a strong potential to exchange protons with their surrounding environment. The amount of polysaccharides produced in cultures of the three strains tested shows that this component cannot be neglected. Cyanobacteria are often habited calcified mats with extreme daily fluctuations in geochemical conditions, for example, typical variations in pH from 8 to 9 during day-night time (Shiraishi et al. 2008). Under such circumstances, sulphydric and amine groups will periodically change their protonation states, releasing protons, when the pH increases, and binding protons, when pH decreases. Therefore, the functional groups with pKa values from 7 –9 will contribute to the buffer capacity and also the alkalinity balance, and therefore, influence the saturation index of carbonate. It is interesting to note that our titration and FTIR data suggest the presence of the sulphur-containing
59
groups. The degradation products of these groups may act as the energy and carbon sources for anaerobic heterotrophs (Lovley & Coates 2000). The intimate coupling of C- and S-cycles in the mat through metabolic activity of cyanobacteria and SRB has been suggested to result in the biogenic production of the sulphur compounds that represents an important source of volatile compounds typically emitted from mats and greatly impact the Earth’s atmosphere (Visscher et al. 2003). Our study showed one possible link between cyanobacteria and SRB through the degradation of cyanobacterial polysaccharides under the fluctuating geochemical conditions in mats.
CaCO3 precipitation by polysaccharides CaCO3 minerals were present on the surface of polysaccharides-coated agarose beads after 5 days of incubation (Fig. 5a, b, c). In controls lacking EPS, CaCO3 minerals were not observed (Fig. 6a, b). The presence of calcium carbonate in the
Fig. 5. Scanning electron microscopy images of coated agarose beads after CaCO3 precipitation experiments. (a) Spherical bead with small rhombohedral precipitates on surfaces. (b) Close-up of rhombohedral precipitates. (c) EDX spectrum of precipitates which is typical for CaCO3.
60
M. DITTRICH & S. SIBLER
Fig. 6. Scanning electron microscopy images of agarose beads without additions of extracellular polysaccharides (blank experiments) after CaCO3 precipitation experiments. (a) Spherical bead after CaCO3 experiments without EPS. (b) Close-up of spherical bead.
precipitates was confirmed using X-ray diffraction analyses by electron microscopy. The results of this study showed that extracellular polysaccharides of picocyanobacteria induced the precipitation of calcium carbonate. All extracellular polysaccharides have a buffering capacity at pH values from 3–4, in acidic range. Therefore, acidic polysaccharides are responsible for calcium carbonate precipitation in our experiments; they comprise L-glutamic and L-aspatic acids which were shown to be able to nucleate calcium carbonate (Braissant et al. 2003). Indeed, L-glutamic acids pKa’s ¼ 2.23, 4.25, 9.67, Lglutamine pKa’s ¼ 2.23, 4.42 and 9.95, and L-aspatic acid have pKa’s ¼ 1.99, 3.9 and 10.02 (Liu & Fang 2002). Stereo-chemical structure in extracellular polysaccharides, which is a result of attaching to solid surfaces, has been suggested to be an important factor in calcium carbonate polymorphisms. In our study, stereo-structures of polymeric substances were controlled through the attachment of agarose beads. As it can be seen from our data and previous studies (Kawaguchi & Decho 2002), calcium carbonate nucleation is induced by polysaccharides. The polysaccharides of three cyanobacterial strains have similar binding sites as we observed by the titration experiments and infrared spectra. In cyanobacterial mats, EPS was shown to affect the precipitation and dissolution of CaCO3 in different way, even in opposite directions (Dupraz & Visscher 2005). In cyanobacterial mats, it is a matter of debate, whether the saturation index of carbonate is a result of physical (e.g. CO2 degassing) or photosynthetic activity (Shiraishi et al. 2008). There were suggestions made that the photosynthetic activity is the key factor for promoting carbonate precipitation and EPS was quantitatively
of minor importance with regard to maintaining CaCO3 precipitation in calcifying biofilms. Our studies demonstrated that calcium carbonate precipitates in the presence of cyanobacterial polysaccharides, without the photosynthetic activity. The mechanism behind it shall need to be investigated in future studies. It is possible that the binding calcium or carbonates on extracellular polymers creates templates for crystal nucleation. However, this hypothesis is needed to be examined as, for example, Shiraishi et al. (2008) showed that the EPS-binding Ca plays a minor part on Ca flux. Cycling of EPS has been shown to be rapid under oxic and anoxic conditions (Decho et al. 2005). It was also demonstrated that the EPS pools of stromatolites are secreted largely by cyanobacteria (Kawaguchi et al. 2003). During anoxic conditions EPS is partly decomposed inducing the decrease of saturation index and dissolution of calcium carbonate. Our experiments demonstrated that isolated polymeric substances from cyanobacteria have a remarkable buffering capacity and are able to induced calcium carbonate formation.
Conclusions In this study, the functional groups of extracellular polysaccharides of three picocyanobacteria strains from hardwater lakes were experimentally examined by potentiomentric titrations and infrared spectroscopy. The results demonstrated that their deprotonation constants are very similar. Modelling and FTIR results are consistent with the presence of five to six distinct surface sites, corresponding to carboxyl, phosphoric, sulphydryl, amine/phenol, and hydroxyl groups, with a total concentration of 3.66 –14.97 mM g21 of bacteria. The carboxyl
CARBONATE, CYANOBACTERIA AND POLYSACCHARIDES
group (pKa ¼ 3) and carboxyl-phosphoric groups (pKa ¼ 3.8–4.6) dominate in all strains with 30– 37% and 13– 17% respectively, closely followed by the hydroxyl groups (pKa ¼ 10) which represent 20 –22%. The small fraction was provided by phosphoric groups (pKa ¼ 6.8–7.4) which was similar in all strains with c. 10%. The extracellular polysaccharides of picocyanobacteria are negatively charged at a pH range between 6–7, which is typical for natural surface water. Calcium cations can therefore be easily attracted. On one hand, this reaction is important for the nucleation of calcium carbonate; on the other hand, calcium removal from solution leads to reduction of saturation in respect to calcium carbonate and therefore, inhibits the precipitation. CaCO3 precipitation experiments clearly demonstrated that extracellular polysaccharides of three picocyanobacterial strains have a potential to precipitate calcium carbonate. Due to the dominance of carboxyl groups, extracellular polysaccharides of picocyanobacteria can play an important role in metal cycling in aquatic systems. For that reason, picocyanobacteria have a general potential for applications such as reducing metal concentrations at polluted sites. However, reaping this potential requires further investigation of the EPS characteristics and the development of techniques to effectively cultivate picocyanobacteria.
References A DAV , S. S. & L EE , D.-J. 2008. Extraction of extracellular polymeric substances from aerobic granule with compact interior structure. Journal of Hazardous Materials, 154, 1120– 1126. A GAWIN , N. S. R., D UARTE , C. M. & A GUSTI , S. 2000. Nutrient and temperature control of the contribution of picoplankton to phytoplankton biomass and production. Limnology and Oceanography, 3, 591–600. A RP , G., R EIMER , A. & R EITNER , J. 2003. Microbialite formation in seawater of increased alkalinity, satonda crater lake, indonesia. Journal of Sedimentary Research, 73, 105– 127. B ECKER , S., S INGH , A. K., P OSTIUS , C., B OGER , P. & E RNST , A. 2004. Genetic diversity and distribution of periphytic Synechococcus spp. in Biofilms and Picoplankton of Lake Constance. FEMS Microbiology Ecology, 49, 181–190. B EECH , I. B. & T APPER , R. C. 1999. Exopolymers of sulphate-reducing bacteria. In: W INGENDER , J., N EU , T. R. & F LEMMING , H.-C. (ed.) Microbial extracellular polymeric substances: Characterization, structure, and function. Spinger, Berlin, Heidelberg, New York, 119–127. B ELL , T. & K ALFF , J. 2001. The contribution of the picophytoplankton in marine and freshwater systems of different trophic status and depth. Limnology and Oceanography, 46, 1243–1248.
61
B HASKAR , P. V. & B HOSLE , N. B. 2005. Microbial extracellular polymeric substances in marine biogeochemical processes. Current Science, 88, 45–53. B IRD , C. & W YMAN , M. 2003. Nitrate/nitrite assimilation system of the marine picoplanktonic cyanobacterium synechococcus sp strain wh 8103: effect of nitrogen source and availability on gene expression. Applied and Environmental Microbiology, 69, 7009– 7018. B LASCHEK , W. 1991. Isolierung und analytik von polysacchariden. In: F RANZ , G. (ed.) Polysaccharide. Springer-Verlag, New York. B ORROK , D., T URNER , B. F. & F EIN , A. B. 2005. A universal surface complexation framework for modeling proton binding onto bacterial surfaces in geologic settings. American Journal of Science, 305, 826–853. B OSAK , T. & N EWMAN , D. K. 2003. Microbial nucleation of calcium carbonate in the precambrian. Geology, 31, 577– 580. B OSAK , T. & N EWMAN , D. K. 2005. Microbial kinetic controls on calcite morphology in supersaturated solutions. Journal of Sedimentary Research, 75, 190– 199. B RAISSANT , O., C AILLEAU , G., D UPRAZ , C. & V ERRECCHIA , A. P. 2003. Bacterially induced mineralization of calcium carbonate in terrestrial environments: The role of exopolysaccharides and amino acids. Journal of Sedimentary Research, 73, 485– 490. B RAISSANT , O., D ECHO , A. W., D UPRAZ , C., G LUNK , C., P RZEKOP , K. M. & V ISSCHER , P. T. 2007. Exopolymeric substances of sulfate-reducing bacteria: interactions with calcium at alkaline pH and implication for formation of carbonate minerals. Geobiology, 5, 401– 411. B RASSARD , P., K RAMER , J. R. & C OLLINS , P. V. 1990. Binding site analysis using linear programming. Environmental Science & Technology, 24, 195–201. C ERNIK , M., B ORKOVEC , M. & W ESTALL , J. C. 1995. Regularized least-squares methods for the calculation of discrete and continuous affinity distributions for heterogeneuos sorbents. Environmental Science & Technology, 29, 413– 425. C OMTE , S., G UIBAUD , G. & B AUDU , M. 2006a. Relations between extraction protocols for activated sludge extracellular polymeric substances (eps) and complexation properties of pb and cd with eps part II. Consequences of eps extraction methods on Pb2þ and Cd2þ complexation. Enzyme and Microbial Technology, 38, 246–252. C OMTE , S., G UIBAUD , G. & B AUDU , M. 2006b. Relations between extraction protocols for activated sludge extracellular polymeric substances (eps) and eps complexation properties part I. Comparison of the efficiency of eight eps extraction methods. Enzyme and Microbial Technology, 38, 237– 245. C OMTE , S., G ULBAUD , G. & B AUDU , M. 2006c. Biosorption properties of extracellular polymeric substances (eps) resulting from activated sludge according to their type: Soluble or bound. Process Biochemistry, 41, 815 –823. C OMTE , S., G UIBAUD , G. & B AUDU , M. 2007. Effect of extraction method on eps from activated sludge: an hpsec investigation. Journal of Hazardous Materials, 140, 129–137.
62
M. DITTRICH & S. SIBLER
C OX , J. S., S MITH , D. S., W ARREN , L. A. & F ERRIS , F. G. 1999. Characterizing heterogeneous bacterial surface functional groups using discrete affinity spectra for proton binding. Environmental Science and Technology, 33, 4514– 5421. D E P HILIPPIS , R., S ILI , C., P APERI , R. & V INCENZINI , M. 2001. Exopolysaccharide-producing cyanobacteria and their possible exploitation: a review. Journal of Applied Phycology, 13, 293–299. DE W INDER , B., S TAATS , N., S TAL , L. J. & P ATERSON , D. M. 1999. Carbohydrate secretion by phototrophic communities in tidal sediments. Journal of Sea Research, 42, 131–146. D ECHO , A. W. 1990. Microbial exopolymer secretions in ocean environments: their role(s) in food webs and marine processes. Oceanography and Marine Biology: An Annual Review, 28, 73–153. D ECHO , A. W. 2000. Microbial biofilms in intertidal systems: An overview. Continental Shelf Research, 20, 1257– 1273. D ECHO , A. W., V ISSCHER , P. T. & R EID , R. P. 2005. Production and cycling of natural microbial exopolymers (eps) within a marine stromatolite. Palaeogeography Palaeoclimatology Palaeoecology, 219, 71– 86. D EPHILIPPIS , R., S ILI , C., T ASSINATO , G., V INCENZINI , M. & M ATERASSI , R. 1991. Effects of growth conditions on exopolysaccharide production by cyanospira-capsulata. Bioresource Technology, 38, 101– 104. D ITTRICH , M., K URZ , P. & W EHRLI , B. 2004. The role of autotrophic picocyanobacteria in calcite precipitation in an oligotrophic lake. Geomicrobiology Journal, 21, 45– 53. D ITTRICH , M. & S IBLER , S. 2005. Cell surface groups of two picocyanobacteria strains studied by zeta potential investigations, potentiometric titration, and infrared spectroscopy. Journal of Colloid and Interface Science, 286, 487–495. D UPRAZ , C. & V ISSCHER , P. T. 2005. Microbial lithification in marine stromatolites and hypersaline mats. Trends in Microbiology, 13, 429–438. F ERRIS , F. G. & B EVERIDGE , T. J. 1984. Binding of a paramagnetic metal cation to Escherichia coli k– 12 outer membrane vesicles. FEMS Microbiology Letters, 24, 43– 46. F ERRIS , F. G. & L OWSON , E. A. 1997. Ultrastructure and geochemistry of endolithic microorganisms in limestone of the Niagara Escarpment. Canadian Journal of Microbiology, 43, 211 –219. F ERRIS , M. J., R UFF -R OBERTS , A. L., K OPCZYNSKI , E. D., B ATESON , M. M. & W ARD , D. M. 1996. Enrichment culture and microscopy conceal diverse thermophilic synechococcus populations in a single hot spring microbial mat habitat. Applied and Environmental Microbiology, 62, 1045–1050. F OWLE , D. A. & F EIN , J. B. 2000. Experimental measurements of the reversibility of metal-bacteria adsorption reactions. Chemical Geology, 168, 27–36. G ARCIA -P ICHEL , F., N UBEL , U. & M UYZER , G. 1998. The phylogeny of unicellular, extremely halotolerant cyanobacteria. Archives of Microbiology, 169, 469– 482. H IRST , C. N., C YR , H. & J ORDAN , I. A. 2003. Distribution of exopolymeric substances in the littoral
sediments of an oligotrophic lake. Microbial Ecology, 46, 22–32. H ORNBACK , J. 1998. Organic chemistry. Brooks Cole Publishing Company, Pacific Grove, CA. J IA , X. S., F URUMAI , H. & F ANG , H. H. P. 1996. Extracellular polymers of hydrogen-utilizing methanogenic and sulfate-reducing sludges. Water Research, 30, 1439– 1444. K AWAGUCHI , T. & D ECHO , A. W. 2002. A laboratory investigation of cyanobacterial extracellular polymeric secretions (eps) in influencing CaCO3 polymorphism. Journal of Crystal Growth, 240, 230–235. K AWAGUCHI , T., S AYEGH , H. A. & D ECHO , A. W. 2003. Development of an indirect competitive enzymelinked immunosorbent assay to detect extracellular polymeric substances (eps) secreted by the marine stromatolite-forming cyanobacteria, Schizothrix sp. Journal of Immunoassay & Immunochemistry, 24, 29–39. K AZY , S. K., S AR , P. & D’S OUZA , S. F. 2008. Studies on uranium removal by the extracellular polysaccharide of a Pseudomonas aeruginosa strain. Bioremediation Journal, 12, 47–57. K AZY , S. K., S AR , P., S EN , A. K., S INGH , S. P. & D’S OUZA , S. F. 2002. Extracellular polysaccharides of a copper-sensitive and a copper-resistant Pseudomonas aeruginosa strain: synthesis, chemical nature and copper binding. World Journal of Microbiology & Biotechnology, 18, 583– 588. L EE , B. D., A PEL , W. A. & W ALTON , M. R. 2004. Screening of cyanobacterial species for calcification. Biotechnology Progress, 20, 1345–1351. L IJOUR , Y., G ENTRIC , E., D ESLANDES , E. & G UEZENNEC , J. 1994. Estimation of the sulfate content of hydrothermal vent bacterial polysaccharides by fourier-transform infrared-spectroscopy. Analytical Biochemistry, 220, 244 –248. L IU , H. & F ANG , H. H. P. 2002. Characterization of electrostatic binding sites of extracellular polymers by linear programming analysis of titration data. Biotechnology and Bioengineering, 80, 806– 811. L OVLEY , D. R. & C OATES , J. D. 2000. Novel forms of anaerobic respiration of environmental relevance. Current Opinion in Microbiology, 3, 252–256. M ARTINEZ , R. E., S MITH , D. S., K ULCZYCKI , E. & F ERRIS , F. G. 2002. Determination of intrinsic bacterial surface acidity constants using a donnan shell model and a continuous pka distribution method. Journal of Colloid and Interface Science, 253, 130–139. M ATA , J. A., B EJAR , V., B RESSOLLIER , P., T ALLON , R., U RDACI , M. C., Q UESADA , E. & L LAMAS , I. 2008. Characterization of exopolysaccharides produced by three moderately halophilic bacteria belonging to the family alteromonadaceae. Journal of Applied Microbiology, 105, 521–528. M ERZ , M. 1992. The biology of carbonate precipitation by cyanobacteria. Facies, 26, 81– 102. M ILLER , S. R. & C ASTENHOLZ , R. W. 2000. Evolution of thermotolerance in hot spring cyanobacteria of the genus Synechococcus. Applied and Environmental Microbiology, 66, 4222–4229. N ICHOLS , P. D. & N ICHOLS , C. A. M. 2008. Microbial signature lipid profiling and exopolysaccharides: Experiences initiated with Professor David C White
CARBONATE, CYANOBACTERIA AND POLYSACCHARIDES and transported to Tasmania, Australia. Journal of Microbiological Methods, 74, 33– 46. N IELSEN , P. H. & J AHN , A. 1999. Extraction of eps. In: W INGENDER , J., N EU , T. R. & F LEMMING , H.-C. (eds) Microbial extracellular polymeric substances: Characterization, structure, and function. Springer, Berlin Heidelberg, 49–72. P ARIKH , A. & M ADAMWAR , D. 2006. Partial characterization of extracellular polysaccharides from cyanobacteria. Bioresource Technology, 97, 1822–1827. P ENTECOST , A. 1991. Calcification processes in algae and cyanobacteria. In: R IDING , R. (ed.) Calcareous Algae and Stromatolites. Springer, Berlin, 3– 20. P HOENIX , V., M ARTINEZ , R. E., K ONHAUSER , K. O. & F ERRIS , F. G. 2002. Characterization and implication of the cell surface reactivity of Calothrix sp. Strain KC97. Applied and Environmental Microbiology, 68, 4827–4834. R IDING , R. 2000. Microbial carbonates: The geological record of calcified bacterial–algal mats and biofilms. Sedimentology, 47, 179– 214. R OBBINS , L. L. & B LACKWELDER , P. L. 1992. Biochemical and ultrastructural evidence for the origin of whiting: A biologically induced calcium carbonate precipitation mechanism. Geology, 20, 464–468. R OCHAS , C., L AHAYE , M. & Y APHE , W. 1986. A new procedure for determining the heterogeneity of agar polymers in the cell-walls of Gracilaria spp (Gracilariaceae, Rhodophyta). Canadian Journal of Botany, 64(3), 579– 585. S CHULTZE -L AM , S., F ORTIN , D., D AVIS , B. S. & B EVERIDGE , T. J. 1996. Mineralization of bacterial surfaces. Chemical Geology, 132, 171– 181. S HIRAISHI , F., B ISSETT , A., DE B EER , D., R EIMER , A. & A RP , G. 2008. Photosynthesis, respiration and exopolymer calcium-binding in biofilm calcification
63
(Westerhfer and Deinschwanger Creek, Germany). Geomicrobiology Journal, 25, 83–94. S OKOLOV , I., S MITH , D. S., H ENDERSON , G. S., G ORBY , Y. A. & F ERRIS , F. G. 2001. Cell surface electrochemical heterogeneity of the Fe(iii)-reducing bacteria Shewanella putrefaciens. Environmental Science & Technology, 35, 341–347. S TOCKNER , J., C ALLIERI , C. & C RONBERG , G. 2000. Picoplankton and other non-bloom-forming cyanobacteria in lakes. In: W HITTON , B. A. & P OTTS , M. (eds) The Ecology of Cyanobacteria. Kluwer Academic Publishers, The Netherlands, 195– 231. S TRATHMANN , M., G RIEBE , T. & F LEMMING , H.-C. 2000. Artificial biofilm model - a useful tool for biofilm research. Applied Microbiology and Biotechnology, 54, 231– 237. T OURNEY , J., N GWENYA , B. T., M OSSELMANS , J. W. F., T ETLEY , L. & C OWIE , G. L. 2008. The effect of extracellular polymers (eps) on the proton adsorption characteristics of the thermophile Bacillus licheniformis S-86. Chemical Geology, 247, 1–15. T RICHET , J. & D EFARGE , C. 1995. Non-biologically supported organomineralization. Bulletin de l’Institute oceanographiqur, Monaco, 14, 203–236. V ISSCHER , P. T., B AUMGARTNER , L. K. ET AL . 2003. Dimethyl sulphide and methanethiol formation in microbial mats: Potential pathways for biogenic signatures. Environmental Microbiology, 5, 296– 308. W ARD , D. M., F ERRIS , M. J., N OLD , S. C. & B ATESON , M. M. 1998. A natural view of microbial biodiversity within hot spring cyanobacterial mat communities. Microbiology and Molecular Biology Reviews, 62, 1353– 1370. W INGENDER , J., N EU , T. R. & F LEMMING , H.-C. 1999. Microbial extracellular polymeric substances: Characterization, structure, and function. Springer, Berlin, Heidelberg, New York.
Microbial influence on macroenvironment chemical conditions in alkaline (tufa) streams: perspectives from in vitro experiments M. ROGERSON*, H. M. PEDLEY & R. MIDDLETON Department of Geography, University of Hull, Cottingham Road, Hull, HU6 7RX, UK *Corresponding author (e-mail:
[email protected]) Abstract: Tufas represent a palaeoclimatic archive of potentially global significance. However, uncertainty remains over the exact process of calcite precipitation from these systems, inhibiting our ability to decipher the precise meaning of geochemical records. For example, field studies of alkaline stream systems are unable to disentangle the influence of temperature and photosynthesis on ambient hydrochemistry on diurnal and annual timescales. This report describes a series of flume experiments in which temperature and light conditions are manipulated separately. These experiments reveal that precipitation of calcite occurs preferentially under conditions of rising pH, and consequently at the night– day transition. The amplitude of diurnal changes is regulated by the buffering capacity (i.e. alkalinity) of the ambient water and by the daytime balance of photosynthesis and respiration. Respiration is shown to be strongly affected by temperature, whereas photosynthesis is found to be limited by nutrient and/or DIC availability making temperature impacts minor. Consequently, macroenvironment pH during both day and night-time tend to be higher under lower temperatures, in contrast to expectation. These observations may have potential implications for the isotopic geochemistry of tufa carbonate, promoting slightly lower d18O, due to the carbonate ion effect, and more significantly negative d13C, due to incorporation of respired CO2 accumulated during the night. The observation that long periods of daylight are not necessarily needed for photosynthetically induced precipitation to occur confirm previous arguments that seasonal lamination requires either strong variability in ambient physicochemical activity or an ecological change in the microbial assemblage, and cannot be ascribed to reduced temperature and light intensity.
The freshwater systems from which tufa precipitates are subject to a wide range of physical and biological influences that alter the state of ambient (macroenvironment) water chemistry (Pentecost 1992; Drysdale et al. 1997, 2003; Kawai et al. 2006; Shiraishi et al. 2008a, b). In the field, these influences can be difficult to disentangle, leading to ambiguity in our understanding of whether physicochemical or biological processes are controlling temporal or spatial variability (Kawai et al. 2006). In this regard, several investigations in which single sites have been monitored over diurnal cycles are of particular importance (Spiro & Pentecost 1991; Bayari & Kurttas 1995; Drysdale et al. 2003; Liu et al. 2008; Takashima & Kano 2008), and these studies collectively highlight the role played by the relative influences of photosynthesis and respiration. Despite some agreement over the importance of photosynthesis in these studies, the major phototrophic agent seems to be regionally variable with cyanobacteria dominant in the UK (Spiro & Pentecost 1991), algae in Turkey (Bayari & Kurttas 1995) and a combination of micro- and macrophytes in China (Liu et al. 2008) identified as dominant. It should also be noted that biological control of diurnal chemistry changes is not universally agreed
upon, as studies of hydrochemistry in an Australian creek are argued to reflect air temperature as the major influence on calcite supersaturation (Drysdale et al. 2003). However, as light intensity will closely correlate with water temperature on diurnal timescales, the impact of biotic and abiotic influences in isolation remain poorly constrained (Liu et al. 2008). As with diurnal hydrochemistry changes, annual lamination in tufa stromatolites have frequently been linked to biological influences via diurnal or seasonal cycles in photosynthesis (Irion & Mu¨ller 1968; Hevesi 1970; Pentecost 1987), biofilm ecology (Arp et al. 2001) or both (Janssen et al. 1999). Though a significant literature has developed concerning laminated tufas, which present a very exciting potential archive of sub-annual climate (Andrews & Brasier 2005), few studies have been performed in which tufa systems have been monitored over several years, so the forcing mechanisms behind lamina formation remain largely the subject of conjecture. One study (Kano et al. 2003) revealed that at sites proximal to carbonate springs, hydrochemical processes within the aquifer provide the dominant control on precipitation rate and suggests little biotic influence. However, this gives little insight into more distal locations where biological
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 65– 81. DOI: 10.1144/SP336.5 0305-8719/10/$15.00 # The Geological Society of London 2010.
66
M. ROGERSON ET AL.
influences may be larger. In these locations, the petrographic character of seasonal laminae have been shown to be extremely variable (Pentecost 1978, 1987; Freytet & Verrecchia 1998; Janssen et al. 1999; Arp et al. 2001; Carthew et al. 2006), with no universal rule governing the character of summer or winter layers (Kano et al. 2003). Clearly, there is much complexity in these systems that remains to be unravelled. Physical modelling provides a means to assess the importance of individual processes without the complexity inherent to field sites. The literature concerning in vitro and/or ex situ work on tufa systems is small, but is growing rapidly (Merz 1992; Pedley 1994; Dittrich et al. 2003; Bissett et al. 2008b; Rogerson et al. 2008; Shiraishi et al. 2008a, b, c; Pedley et al. 2009) and provide a crucial testing grounds for hypotheses regarding the origin of physical and chemical characteristics of tufa carbonate. These studies confirm that photosynthesis drives a significant (10 –20%) increase in calcium flux to the water –calcite interface relative to noncolonized carbonate surfaces under artificial sunlight, and export of calcium from the surface in dark conditions (Shiraishi et al. 2008a). Perhaps surprisingly, however, they also indicate little change in calcium flux under macroenvironmental conditions between 4–17 8C or between pH 7.8– 8.4, indicating an unexpected ability of microbial mats to maintain constant internal (microenvironment) conditions (Bissett et al. 2008b). This study builds on these ground-breaking experiments by exploring diurnal pH and conductivity variability within a karst water system close to equilibrium with respect to calcite precipitation. The daily cycle of precipitation and dissolution is investigated, and the impact of varying day length and ambient temperature discussed. In the light of these experiments, we consider the implications for the origin of annual laminae in tufa stromatolites.
Methods The experiments described in this paper were performed within the mesocosm facility at the University of Hull described in previous publications (Rogerson et al. 2008; Pedley et al. 2009). The mesocosm is designed to re-circulate 40 L of water contained in a 60 L sump (clamp top barrel). The flume consisted of a 1 m length of 112 mm wide polycarbonate gutter, into which three low barriers were fixed to trap small water pools in an otherwise recycling, direct through-flow system. Water depth varies between 1 mm in riffles to a 5–6 cm in pools, allowing maximum interaction between the water and the biofilm and compares well with small tufa producing systems in the natural
environment. Water flow was driven by a ‘Hozelock Cascade Cyprio 1000’ variable rate submersible pump (Hozelock Cyprio Ltd., Aylesbury, UK). This system represents a recirculating flume, and therefore is not a precise analog for natural tufa systems which are essentially flow-through. Consequently, the magnitude of hydrochemical changes may be somewhat amplified in our data as a result of so-called ‘additive’ effects. However, relative changes within a system and between similar systems are unlikely to be altered by these passive influences and can be compared to natural systems with a high degree of confidence. Three identical Flumes (1, 2 and 3) were employed and a fourth system (Flume 4) which contained no transverse barriers, thereby reducing turbulence to a minimum. Each flume was sealed within a purpose-built transparent perspex tank with entry and exit connectors for the flexible pipe work and access for sampling. To ensure good air circulation, a ‘Hiblow HP40’ diaphragm-type air pump (Hiblow, Saline, MI, USA) was connected to the flume with a pre-flume filter and an open (slightly back pressured) exhaust port. The lighting for the experiments was provided by a timercontrolled Thorn-Lopak 250 W, HPS-T sodium lamp (Thorn Lighting, London, UK) for each mesocosm; this provided the fullest practicable light spectrum for photosynthesis. Experimental temperatures were maintained by means of in-line ‘Titan 150 minicooler’ chiller units (Aqua Medic, Bissendorf, Germany) in combination with refrigeration units placed around the sump. The ambient room temperature was buffered at 16 8C throughout the experiments by means of an ‘Airforce Climate Control’ air conditioning unit (10 000 BTU h21; cooling capacity 2.9 kW; Airconwarehouse, Stockport, UK). Flume 1 had about half the flow rate (0.04 L s21) of the other flumes (0.09, 0.08 and 0.07 L s21 respectively). Flumes 1 and 2 were colonized with biofilm recovered from the River Lathkill (Derbyshire). Initial communities were imported attached to plastic mesh pads that were placed on the river bed and attached to the bottom of the flume by wire springs. Subsequent to import of this community, the biofilm is allowed a colonisation period of 12 weeks to establish throughout the flume liner after which experiments were initiated. After this period, biofilm coverings of surfaces were between 1 mm and 5 cm. Throughout the experiments described in this manuscript, pH and conductivity were monitored by means of submerged probes (2–3 cm depth) linked to ‘Pinpoint’ digital metres (American Marine Inc., Ridgefield, CT, USA). All data were relayed (every 10 min on a 24/7 basis) via a webcam, to a PC for archiving. Certain experiments
MICROBIAL INFLUENCE ON TUFA STREAMS
were also systematically sampled for detailed water analysis by means of a Perkin-Elmer Optima 5300DV (Perkins-Elmer, Waltham, MA, USA) inductively coupled plasma optical emission spectrometer (ICP– OES) for dissolved metal ions and titration with 0.02 M hydrochloric acid using the ‘BDH 4.5’ indicator for bicarbonate. The water used for experiments was taken from a chalk spring in Welton (East Yorkshire, UK). This water was stored at 5 8C, sterilized by UV treatment and placed into the flume reservoirs within 1 week of collection. Water was not replenished during periods of equilibrium experimentation, and a small loss (,10%) volume was incurred via evaporation over this period. Detailed experiments concerning the impact of photoperiod and temperature on these systems revealed deviation of sterile Flume 4 from physicochemical expectation. It is therefore likely that some unknown chemical or biological agent infected this Flume during the experiment, meaning that the results of Flume 3 should be taken as a more reliable representation of sterile systems.
Experiment design Two experiments are reported in this manuscript; Experiment 1 was designed to investigate diurnal cyclicity in a system in approximate equilibrium, and Experiment 2 was designed to investigate the impact of changes in photoperiod and water temperature. In Experiment 1 all four flumes were run under ‘mid-summer conditions’, with water temperature set by the in-line chillers at 14 8C and the day length at 18 hours, and flume water was allowed to evolve freely over a period of three months. The initial evolution of water chemistry is dominated by carbonate precipitation, and this is described in detail elsewhere (Rogerson et al. 2008). Subsequent to attainment of approximate 2 equilibrium in terms of Ca2þ (aq) and HCO3(aq), the diurnal cycle of pH and conductivity was recorded for a further 2 months in order to establish whether diurnal changes were cyclic and what chemical and biological factors control this cyclicity. As this data represents near-equilibrium conditions, it reflects levels of calcite saturation slightly (SIcalcite ¼ lnf[Ca](aq)[CaCO22 3 ](aq)/kspg) below that usually associated with tufa precipitation (SI . 2.5; Pentecost 1992). Maintaining perfectly constant hydrochemistry under a condition of rapid calcite precipitation is extremely challenging in a recirculating flume, and inevitable small variability in alkalinity may have overprinted the photosynthetic variability this study is concerned with. Consequently, the flumes systems are relatively weakly buffered in comparison to many field sites,
67
Table 1. Temperature and photoperiod conditions used during treatments contributing to Experiment 2. Temperature was controlled via an in-line chiller unit, and lighting via automated over-head hydroponic lamps Experiment name ‘mid-summer’ ‘cold summer’ ‘warm winter’ ‘mid-winter’
Photoperiod length (hours)
Water temperature (8C)
18 18 6 6
14 8 14 8
thereby likely exaggerating oscillations driven by microbe metabolic processes. Nevertheless, the carbonate –biofilm–water chemical system these experiments represent is the same as at field sites, so this issue is likely only to affect the magnitudes of effects described. The mesocosm facilities at the University of Hull allow independent control on ambient temperature (via in-line chillers) and photoperiod length (via lamps) that cannot be achieved in field studies. Furthermore, comparison of the response of colonized Flumes (1 and 2) to sterilized Flumes (3 and 4) allows distinction of purely biological from physico-chemical effects. To exploit this, during Experiment 2 exposed the systems used in Experiment 1 to a series of ‘treatments’, the conditions for which are summarized in Table 1. These experiments were designed to reveal the independent and combined influences of temperature and photoperiod length on colonized and non-colonized systems. Each treatment ran for 3 weeks, with the first week of data disregarded for analysis. Treatments were separated by a 2-week period during which the flumes were maintained under ‘midsummer’ conditions.
Results Characterisation of the diurnal cycle in colonized and sterilized systems Figure 1a–d show the evolution of pH in Flumes 1 –4 respectively during Experiment 1. The impact of photoperiod cycling in Flumes 1 (Fig. 1a) and 2 (Fig. 1b) is clear, with the 6-hour period of dark consistently characterized by lower pH than the 18-hour period of light. The amplitude of day/night variations in both flumes increases with time, in close association with the decrease in dissolved calcium (Fig. 1e). Note that the decrease in dissolved calcium is primarily related to precipitation of calcite, and that evolution of bulk water chemistry over this experimental period is fully described elsewhere (Rogerson et al. 2008). Consequently the low
68
M. ROGERSON ET AL.
amplitude of diurnal changes in the earliest few days of the experiment reflect the high buffering capacity of the water, rather than low microbial activity. After c. 10 days, approximate equilibrium in terms of dissolved calcium concentration was reached and the amplitude of diurnal variability in pH in Flume 1 stabilized at approximately the same time. However, the amplitude of pH variability in Flume 2 continued to increase until 30 days, after which it
began to reduce. The sterilized Flumes 3 (Fig. 1c) and 4 (Fig. 1d) show an early period of raised pH followed by a smooth fall to a position of equilibrium. Some variability around this equilibrium can be found in both flumes, but this is essentially random noise. This experiment was fully replicated, with similar results. The pH data presented in Figure 1 and subsequent figures can be viewed as a proxy for the
Fig. 1. Data for Experiment 1. Macroenvironmental pH evolution of flume systems subsequent to addition of water from Welton Beck. The red vertical line indicates the start of the experiment. (a) Flume 1; (b) Flume 2; (c) Flume 3; 2þ (d) Flume 4; (e) [Ca2þ (aq)] in all four flumes. The decrease in [Ca(aq)] and corresponding increase in the amplitude of pH variability in Flumes 1 and 2 over the first 2–3 weeks reflect declining buffering capacity in the ambient water.
MICROBIAL INFLUENCE ON TUFA STREAMS
Fig. 2. Henderson–Hasselbach behaviour of carbonate species across pH range recorded in flume systems. 2 (pH210.32).CO þ H CO :HCO2 ¼ CO22 3 : HCO3 ¼ 10 2 2 3 3 1/(10(pH26.5)).
69
state of the carbonate system in solution. As temperatures and ionic strength vary little within and between experiments, single values for the first and second ionisation coefficient of carbonic acid can be used for all the data presented (Patterson et al. 1984). Using k1 and k2 values of 6.5 and 22 10.32, the ratio of CO2 þ H2CO3 :HCO2 3 :CO3 will vary across the total range of pHs recorded as shown in Figure 2. Figure 3a –d shows the diurnal cycles of all four flumes after approximate equilibration with respect to the precipitation of calcium carbonate. In each case the graph shows mean and 1s variability for data gathered over a period of 1 month for each 10 minute period in the 24 hour cycle (taken from data in Fig. 1). Figure 4a– d shows the diurnal cycle in water conductivity measured synchronously with the pH data shown in Figure 3. A slight variability (c. 5 mS) can be observed in Flumes 3 and 4 (sterilized), which probably relates to a slight variability in
Fig. 3. Data for Experiment 1. Mean and 1s data for stacked diurnal macroenvironmental pH variability over a 1 month period of the experiment shown in Figure 1. (a) Flume 1; (b) Flume 2; (c) Flume 3; (d) Flume 4.
70
M. ROGERSON ET AL.
Fig. 4. Data for Experiment 1. Mean and 1s data for stacked diurnal macroenvironmental conductivity over the 1 month period shown in Figure 1 (the same period as used for compilation of fig. 2). (a) Flume 1; (b) Flume 2; (c) Flume 3; (d) Flume 4.
macroenvironmental temperature. An approximately 5 mS variability in these systems indicates temperature changes of c. 0.8 8C, assuming that conductivity changes by c. 2% 8C21, as is generally the case for naturally occurring solutions (Sorensen et al. 1987). This variability will be induced by the thermal regime of the building in which the flume systems are housed, though it is heavily damped by our precautions to thermally buffer these systems. Flumes 1 and 2 (colonized), which are in the same room as Flumes 3 and 4 and therefore experience the same diurnal variability in air temperature, do not demonstrate the same thermallydriven cycling. Consequently, some additional process must be occurring within these systems that provides sufficient diurnal variability to overwhelm this physical control on conductivity. Minimum conductivity in both systems is achieved after the lights have switched on in the morning, and maximum conductivity in the middle part of the day.
Investigation of the independent and combined impacts of temperature and photoperiod length (Experiment 2) The impact of the experimental treatments on ambient water pH in sterilized systems are shown in Figures 6 and 7. The results for sterilized flumes (3 and 4) show lower pH under relatively
cold conditions (compare ‘mid-summer’ with ‘mid-winter’ and ‘cold summer’ in Figs 5 & 6). Relative changes in pH values are presented as offset values for treatment data from mid-summer data, so that negative offsets indicate lower pH in the treatment. This reflects the higher solubility of CO2(aq) at lower temperature, and consequently is a predictable component of the response of all of these systems (Weyl 1958). Less predictably, the pH changed in both flumes under reduced light conditions, with Flume 3 showing an offset of 0.1 (Fig. 5) and Flume 4 an offset of 20.2 (Fig. 6). Colonized flume (1 and 2) pH data remain dominated by photosynthetic cycles, regardless of the experimental conditions (Figs 7 & 8). However, changing both light and temperature conditions alter the state of this diurnal cycle profoundly. Under ‘Cold summer’ conditions, both flumes water experience a drop in night-time pH (Flume 1 0.1 and Flume 2 0.15) relative to ‘mid-summer’ (Figs 7 & 8), which are similar to the offsets found within sterilized systems (Figs 5 & 6). Daytime pH is increased by 0.1 in the data for Flume 1 (Fig. 7) and by 0.25 for Flume 2 (Fig. 8). ‘Warm winter’ conditions result a slight increase in night-time pH (,0.1) for Flume 1 (Fig. 7) and c. 0 change for Flume 2 (Fig. 8). However, in both flumes day-time pH is increased, by 0.15 for Flume 1 (Fig. 7) and 0.2 for Flume 2 (Fig. 8). Despite differences between the flumes in other experiments, both colonized systems show the
MICROBIAL INFLUENCE ON TUFA STREAMS
71
Fig. 5. Results for Flume 3 from Experiment 2. Conditions set for the different treatments are shown in Table 1. Dataset in grey represents ‘mid-summer’ data with ‘Experimental’ data overlaid in black for comparison. Experimental data is as follows: (a) ‘warm winter’; (b) ‘cold summer’; (c) ‘mid-winter’.
72
M. ROGERSON ET AL.
Fig. 6. Results for Flume 4 from Experiment 2. Conditions set for the different treatments are shown in Table 1. Dataset in grey represents ‘mid-summer’ data with ‘Experimental’ data overlaid in black for comparison. Experimental data is as follows: (a) ‘warm winter’; (b) ‘cold summer’; (c) ‘mid-winter’.
MICROBIAL INFLUENCE ON TUFA STREAMS
73
Fig. 7. Results for Flume 1 from Experiment 2. Conditions set for the different treatments are shown in Table 1. Dataset in grey represents ‘mid-summer’ data with ‘Experimental’ data overlaid in black for comparison. Experimental data is as follows: (a) ‘warm winter’; (b) ‘cold summer’; (c) ‘mid-winter’.
74
M. ROGERSON ET AL.
Fig. 8. Results for Flume 2 from Experiment 2. Conditions set for the different treatments are shown in Table 1. Dataset in grey represents ‘mid-summer’ data with ‘Experimental’ data overlaid in black for comparison. Experimental data is as follows: (a) ‘warm winter’; (b) ‘cold summer’; (c) ‘mid-winter’.
MICROBIAL INFLUENCE ON TUFA STREAMS
same change between mid-summer and mid-winter conditions with night-time pH increased by c. 0.1 and day-time pH increased, with maximum values increased by 0.2 and 0.1 for Flumes 1 (Fig. 7) and 2 (Fig. 8) respectively.
Discussion Photosynthetic influence on macroenvironment water chemistry The pH of alkaline streams is regulated by the dynamic equilibrium between the rates of photosynthesis and respiration and the degree of buffering provided by ambient alkalinity (Liu et al. 2008). Due to the pH range found for tufa streams, and in our flume systems, most of this alkalinity will be as bicarbonate (HCO2 3(aq)) ions. Consequently, phototrophic and heterotrophic activity within these systems are intimately linked with the carbonate system, forming a ‘bioreactor’ (Visscher & Stolz 2005) potentially driving either precipitation or dissolution. In our systems, which are dominated by cyanobacteria (Pedley et al. 2009), the dominant metabolic effect can be expressed following Visscher & Stolz (2005) by: Respiration 2þ ! CH2 O(cyanobacterial) þ O2(gas) þ CaCO3(solid) 2HCO 3(aq) þ Ca(aq) ! Photosynthesis (1)
It should be noted that free CO2(gas) and CO22 3(aq) do not appear in this equation. This is because the pH (generally c. 8) ensures that .90% of the dissolved inorganic carbon will be present as HCO2 3(aq), and consequently it is primarily this inventory from which carbon for both precipitation and microbial metabolic processes is taken (Visscher & Stolz 2005). Figure 1 shows an increase in day to night pH variability in the colonized flumes (1 and 2), over the first 10 days. This is driven by the precipitation 2 of calcite decreasing Ca2þ (aq) and HCO3(aq) concentrations and consequently reducing the buffering capacity of the system. In this early phase of the experiment, most of the metabolic activity of the biofilm is compensated by the buffering of the ambient water. Although it is counter-intuitive, it is therefore during this period that the highest diurnal changes in calcium flux via equation 1 will be occurring. In both data sets, the amplitude in Flume 2 is characteristically higher than in Flume 1. Given that Flume 1 contains substantially more bryophyte coverage and significantly greater biofilm development (established by visual
75
inspection), this result is distinctly counterintuitive. When it is further considered that the source water in the two systems is identical and that their chemical evolutions through the experiment are similar, with rapid precipitation of CaCO3(solid) during the first 1 –2 weeks followed by approximate equilibrium (Fig. 1e), it must be assumed that the buffering capacity of the two colonized systems should be approximately the same. Consequently, it must be assumed that although it may be smaller in terms of biomass, the Flume 2 biofilm is capable of achieving higher CO2 fluxes through photosynthesis and respiration. The nature of the linked microbialcarbonate system (equation 1) means that it is likely that this Flume also experiences enhanced diurnal flux of calcium between solid and aqueous phases (Bissett et al. 2008b; Liu et al. 2008; Shiraishi et al. 2008a). It is therefore significant to note that Flume 2 experienced enhanced loss of calcium from solution compared to Flume 1 over the course of this experiment (see Fig. 1e and Rogerson et al. 2008). Indeed, not only was the difference in Ca2þ (aq) loss between Flume 2 and the sterilized systems (Flumes 3 and 4) double that of Flume 1, but it continued to be lost throughout the experiment and never reached true equilibrium, suggesting that the diurnal calcium cycle was asymmetric in this system with a small mass of calcium added to the calcite inventory during the average cycle. It may be coincidence that this asymmetry occurs only in the system with the higher amplitude diurnal cycle, however the inference that photosynthetically driven precipitation is enhanced under fast flow conditions compares favourably with recent findings in German carbonate creeks (Bissett et al. 2008a). Increasingly, flow rate is becoming an important, perhaps the most important, parameter to consider with regard to precipitation kinetics (see Pedley et al. 2010 and Hammer et al. 2010). The co-occurrence of high amplitude diurnal pH variability and enhanced precipitation of calcite in Flume 3 provides indirect evidence that photosynthesis may be promoting precipitation in these systems. Figure 3, which shows mean pH in Flumes 1– 4 over 1 month under approximately equilibrium conditions, provide a means of assessing exchange of ions between solid and aqueous phases more directly. From this data compilation, it becomes clear that diurnal change in macroenvironmental pH occurs as two opposed asymptotes resembling titration curves. Indeed, these opposed curves almost certainly do reflect the response of the HCO2 3(aq) buffered system to a change in the relative rates of microbial carbon import and export, reflected in movement of the chemical system described in equation 1 to the right when photosynthesis becomes active at the start of the day and vice versa. The consequence of this observation is that
76
M. ROGERSON ET AL.
precipitation of calcite must be occurring during every early morning in close analogy to the suggestion of Liu et al. (2008).
which can, in turn, be simplified to:
Insights into precipitation from conductivity
where IPA(Ca,HCO3 ) is the ion activity product of 2 [Ca2þ (aq)] and [HCO3(aq)]. Usefully, in this form conductivity is linked to calcite saturation (SI ¼ ln(IAP(Ca,CO3 ) =ksp ), where ksp is the solubility product constant for calcite. We established the relationship between conductivity and IAP(Ca,HCO3 ) for Welton spring water (sterilized by the same procedure as used in all the described experiments) via sequential dilution at 25 8C (Fig. 9). The relationship between is nonlinear, approximating to Lc ¼ 5.38 104 IAP0.4094 over the IAP range 1027 to 1025. Changing pH (via 2 decreased [Hþ (aq)] and increased [OH(aq)] and 22 [CO3(aq)]), metal/Ca ratios and temperature will cause drift from this calibration, with pH being of particular importance due to the substantially higher molar conductivity of CO22 3(aq) compared to HCO2 3(aq). This is reflected in a limiting molar 24 conductivity (L0) for CO22 3(aq) of 69.5 10 2 21 24 2 21 m mol compared to 44.5 10 m mol for HCO2 3(aq) (Aguado et al. 2006). Sterilized flumes (3 and 4) both show fairly constant conductivity values, with a c. 2 mS cm21 diurnal variability with maximum values in the
Figure 4 shows electrical conductivity data (mS cm21) measured synchronously with the pH records in Figure 3. Conductivity of water is controlled by Kohlrausch’s Law, which states that conductivity is related to concentration via: Lc ¼
n X
v i Li
(2)
i¼0
where Lc is the conductivity of the solution, vi and Li are respectively the molar concentration and molar conductivity of ionic species i, irrespective of charge sign. In Welton spring water, which is taken directly from a chalk aquifer, HCO2 3(aq) represents the dominant anion and Ca2þ (aq) the dominant cation, and these are characteristically present in a 2:1 molar ratio (M. Rogerson, unpublished data). Consequently, for Welton spring water: 2þ € Lc [HCO 3(aq) ]LHCO3(aq) þ [Ca(aq) ]ECa2þ
(aq)
(3)
Lc IAP(Ca,HCO3 ) LCa(HCO3 )2
(4)
Fig. 9. Calibration of electrical conductivity against Ion Activity Product of calcium bicarbonate derived from serial dilution of Welton Beck water.
MICROBIAL INFLUENCE ON TUFA STREAMS
early evening (roughly 20.00) and minimum values in the early morning (04.00 to 06.00). The timing and amplitude of this variability is consistent with a small variability in temperature driven by the temperature of the room in which the facility is housed. Colonized flumes show larger amplitudes of diurnal variability (c. 5 mS cm21) and a different pattern, with maximum values in the afternoon (roughly 16.00) and minimum values in the midmorning (roughly 08.00). As it is different to the temperature cycle, this cyclicity has to be driven by repeatable chemical changes, and the peak in conductivity during the day, when pH is high, 22 suggest that conversion of HCO2 3(aq) to CO3(aq) is playing a dominant role. However, this would suggest that conductivity should rise synchronously with pH, and this is not the case. The conductivity minimum at the night–day transition must therefore reflect loss of ions from solution, most likely a reduction in IAP(Ca,HCO3 ) , and provides very strong confirmation for the photosynthesis-driven precipitation of calcite inferred from the pH measurements. As a change of 5 mS cm21 is the equivalent of 4.56 1024 in IAP(Ca,HCO3 ) this change is subtle, c. 1% of background values, but this reflects a relatively undersaturated solution. We anticipate that for a system maintained at higher IAP, the diurnal effect on conductivity may well be higher than in our Flumes, which is consistent with the data presented by Liu et al. (2008).
Light and temperature conditions and their effect on pH The sterilized Flumes (3 and 4) (Figs 5 & 6) show a decrease in pH from summer to winter conditions (0.15 for Flume 3 and 0.4 for Flume 4) and from ‘mid-summer’ to ‘cold summer’ conditions (0.15 for Flume 3 and 0.05 for Flume 4). The response of Flume 3 is consistent between these two experiments and also with expectations for the solubility of CO2, which will decrease by 2.54 1023 M 8C21 for simple aqueous solutions in this temperature range (Duan & Sun 2003). Flume 4, though displaying lower pH under lower temperature, does not show the same response between these experiments, and this is largely due to an unexpected response in this system to reduced light conditions. The ‘warm winter’ experiment produces minimal change in pH in Flume 3, but a significant change (20.15) in Flume 4. The cause of this may be either be evidence of compromised sterility in Flume 4, or alternatively be reflecting a change in an unknown photochemical reaction not present in Flume 3. In either case, data from Flume 4 must be interpreted with caution.
77
Reducing the length of the day has a significant effect on pH in the colonized systems (Figs 7 & 8), though the fundamental character of the diurnal pH curve (i.e. two opposed asymptotes) remains the same. In addition to the expected extension of the low pH conditions throughout the dark period, the daytime pH curve in Flume 1 changes its shape slightly. During the ‘mid-summer’ and ‘cold summer’ experiments, approximate equilibrium is achieved in ,5 hours in this system. However, under both ‘warm winter’ and ‘midwinter’ conditions a similar equilibrium is not achieved during the entire 6-hour daylight period, which lies in contrast to Flume 2, which achieves equilibrium in around 3 hours of ‘dawn’ regardless of experimental conditions. The failure of Flume 1 to achieve equilibrium may reflect an inability of photosynthesis to remove the respired CO2(aq) accumulated during the night. As with the day-night amplitude differences noted above, this apparently enhanced buffering of the Flume 1 chemical system reflects its longer overturning time (i.e. slower flow rate). In terms of comparison between pH values during the day and during the night, reducing temperature or photoperiod in isolation or simultaneously either has no effect on colonized Flumes (1 and 2) or tends to cause a small increase. Only during night-time in the ‘cold summer’ experiment in Flume 1 is pH marginally lower than during the ‘mid-summer’ run. Moreover, night-time pH during the ‘mid-winter’ experiment is roughly 0.05 higher than during the ‘warm winter’ experiment in both flumes. The tendency of macroenvironmental pH in colonized flumes to increase with cooling stands in contrast to the sterile systems, and indicates that temperature control on photosynthesis is smaller than that of respiration. A significant consequence of this is that where biofilm is present, changes in microbial metabolism dominate over the impact of CO2 solubility.
Geological and geochemical implications Microbial influence on precipitation Field-based research has given much emphasis on the role of bulk water saturation index in regulating the occurrence of precipitation (Chen et al. 2004), despite the apparent failure of this parameter to effectively predict the spatial occurrence of tufa deposits (Pentecost 1992). The discovery that precipitation of tufa-like precipitates within laboratory systems requires the presence of biofilm, that precipitates derive from biofilm interstitial water rather than directly from bulk ambient water (Pedley et al. 2009; Rogerson et al. 2008), and that biofilms organisms exert a dominant control on the chemistry of this fluid (Bissett et al.
78
M. ROGERSON ET AL.
Fig. 10. A conceptual four-phase model for dissolved ions within a tufa system, showing pathways of mass exchange.
2008a, b; Shiraishi et al. 2008a) provides an explanation for this, supporting a biomediated view for tufa precipitates. Consequently, rather than considering tufa systems a two-phase system (i.e. water–calcite) a process-oriented understanding of precipitation will require consideration of the four-phase system outlined in Figure 10. In this context, our conductivity data should be seen as indicating flux of ions into/out of the Ambient Water inventory and field data seen as point analyses of the state of this part of the tufa system. In a recent study in which the hydrochemistry of a pool within a Chinese tufa forming stream was monitored over a 48 hour period, Liu et al. (2008) show that pH varies by 0.8– 1 on a diurnal timescale, and that this change was dominantly driven by the
balance of photosynthesis and respiration. Moreover, via conductivity measurements they demonstrate that precipitation of calcite occurs during daytime pH maxima, and a degree of dissolution occurs during night time pH minima. Our results endorse this view, and comparison of the data provide significant new insights into the role of photosynthesis in promoting precipitation. Unlike the data presented here, the Chinese study show pH (conductivity) fairly smoothly increasing (decreasing) towards midday followed by a fairly smooth decrease (increase) back to night time values. This different character is a consequence of the different lighting regimes, as unlike natural sunlight our lamps are either on or off. Unfortunately, Liu et al. (2008) do not present light intensity
MICROBIAL INFLUENCE ON TUFA STREAMS
data, but we argue that the two different datasets can be unified by the concept that photosynthetically induced precipitation dominantly occurs in both settings during periods of rising pH independent of its absolute value. The period of maximum precipitation (i.e. ion flux out of the Ambient Water inventory, Fig. 10) reflected in a period of minimum conductivity, is succeeded by a period of return to background conditions, reflected by slowly rising conductivity in the Ambient Water inventory. The length of this recovery period will be dependent on the buffering capacity of the system (i.e. the capacity for microbiologically-induced fluxes to significantly alter the state of the Ambient Water inventory) and the variability of the light source (i.e. occurrence of period of increasing photosynthetic activity), both of which tend to lead to a longer recovery period for the Chinese system. A further consequence of precipitation under rising pH conditions is that the pH of water actively precipitating calcite will be biased towards values lower than at midday. This may be significant for studies of d18Ocalcite, as fractionation of oxygen iso22 topes between HCO2 3(aq) and CO3(aq) means that the equilibrium position for d18Ocalcite is expected to change approximately in proportion with the change in pH (Zeebe 1999). Consequently, incorporation of significant photosynthetically induced precipitate will tend to cause a reduction in tufa d18Ocalcite. The apparent reliability of the tufa d18Ocalcite archive (Andrews et al. 1997) may therefore require some reinvestigation using interstitial water from the biofilm rather than ambient water from the river. Impact on tufa d13Ccalcite would be expected to be even more significant, as much of the carbon incorporated is likely to be that liberated by respiration in the preceding night period. As with d18O, this will tend to cause photosynthetically induced precipitate to have relatively low d13C values.
Implications for seasonal lamination On initial inspection, the extended period of low pH in the colonized systems under short daylight periods is consistent with the expectation of impeded precipitation during the winter (Arp et al. 2001). However, this apparent agreement is undermined if precipitation occurs during the parabolic part of the pH curve and by the fact that the maximum pH reached in the diurnal cycle is actually higher under low temperature conditions. These results indicate that changing conditions within the microenvironment of the biofilm are capable of overcoming physicochemical changes in the macroenvironment of the entire flume system, and thus build on previous work indicating that the reverse does not happen (Bissett et al.
79
2008b). It seems that biofilms are more effective at modifying the streams they live within than the streams are at modifying them. In combination, these experiments therefore imply that externally forced changes in photosynthesis are not a likely explanation for reduced winter precipitation. For biological effects to be a significant factor in development of seasonal laminae, a change in the microbial assemblage may be necessary. It has previously been reported that the ecological composition of tufa biofilm is seasonally variable (Arp et al. 2001), and recent evidence for microbial partitioning of encrusting/non encrusting surfaces within a single river system indicates that this may prove to be a critical factor (Ledger et al. 2008) in determining the occurrence of laminae.
Conclusions Physical modelling of photosynthetic effects on tufa stream water reveals that precipitation of calcite occurs preferentially under conditions of rising pH, and consequently at the night–day transition. The amplitude of diurnal changes are regulated by the buffering capacity (i.e. alkalinity) of the ambient water and by the daytime balance of photosynthesis and respiration. Respiration is shown to be strongly affected by temperature, whereas photosynthesis is found to be limited by nutrient and/or DIC availability making temperature impacts minor. Consequently, macroenvironment pH during both day and night-time tend to be higher under lower temperatures, in contrast to expectation. These observations have potential implications for the isotopic geochemistry of tufa carbonate, promoting slightly lower d18O, due to the carbonate ion effect (Zeebe 1999), and more significantly negative d13C, due to incorporation of respired CO2 accumulated during the night. The observation that long periods of daylight are not necessarily needed for photosynthetically induced precipitation to occur confirm that seasonal lamination requires either strong variability in ambient physicochemical activity or an ecological change in the microbial assemblage, and cannot be ascribed to reduced temperature and light intensity (Bissett et al. 2008b).
References A GUADO , D., M ONTOYA , T., F ERRER , J. & S ECO , A. 2006. Relating ions concentration variations to conductivity variations in a sequencing batch reactor operated for enhanced biological phosphorus removal. Environmental Modelling & Software, 21(6), 845–851. A NDREWS , J. E. & B RASIER , A. T. 2005. Seasonal records of climatic change in annually laminated tufas: short review and future prospects. Journal of Quaternary Science, 20(5), 411 –421.
80
M. ROGERSON ET AL.
A NDREWS , J. E., R IDING , R. & D ENNIS , P. F. 1997. The stable isotope record of environmental and climatic signals in modern terrestrial microbial carbonates from Europe. Palaeogeography Palaeoclimatology Palaeoecology, 129(1–2), 171– 189. A RP , G., W EDEMEYER , N. & R EITNER , J. 2001. Fluvial tufa formation in a hard-water creek (Deinschwanger Bach, Franconian Alb, Germany). Facies, 44, 1 –22. B AYARI , C. S. & K URTTAS , T. 1995. Algae: an important agent in deposition of karstic travertines: observations on natural-bridge yerkopru travertines, Aladaglar, eastern Taurids, Turkey. In: G UNAY , G. J. A. I. (ed.) 5th International Symposium and Field Seminar on Karst Waters and Environmental Impacts. A. A. Balkema, Antalya, Turkey, 269–280. B ISSETT , A., D EBEER , D., S CHOON , R., S HIRAISHI , F., R EIMER , A. & A RP , G. 2008a. Microbial mediation of tufa formation in karst-water creeks. Limnology and Oceanography, 53, 1159–1168. B ISSETT , A., R EIMER , A., D E B EER , D., S HIRAISHI , F. & A RP , G. 2008b. Metabolic microenvironmental control by photosynthetic biofilms under changing macroenvironmental temperature and pH conditions. Applied Environmental Micropalaeontology, 74(20), 6306–6312. C ARTHEW , K. D., T AYLOR , M. P. & D RYSDALE , R. N. 2006. An environmental model of fluvial tufas in the monsoonal tropics, barkly karst, northern Australia. Geomorphology, 73, 78– 100. C HEN , J. A., Z HANG , D. D., W ANG , S. J., X IAO , T. F. & H UANG , R. G. 2004. Factors controlling tufa deposition in natural waters at waterfall sites. Sedimentary Geology, 166(3–4), 353– 366. D ITTRICH , M., M ULLER , B., M AVROCORDATOS , D. & W EHRLI , B. 2003. Induced calcite precipitation by cyanobacterium Synechococcus. Acta Hydrochimica et Hydrobiologica, 31(2), 162–169. D RYSDALE , R., L UCAS , S. & C ARTHEW , K. 2003. The influence of diurnal temperatures on the hydrochemistry of a tufa-depositing stream. Hydrological Processes, 17(17), 3421–3441. D RYSDALE , R. N. & G ALE , S. J. 1997. The Indarri Falls travertine dam, Lawn Hill Creek, northwest Queensland, Australia. Earth Surface Processes and Landforms, 22(4), 413–418. D UAN , Z. & S UN , R. 2003. An improved model calculating CO2 solubility in pure water and aqueous nacl solutions from 273 to 533 K and from 0 to 2000 bar. Chemical Geology, 193(3– 4), 257–271. F REYTET , P. & V ERRECCHIA , E. P. 1998. Freshwater organisms that build stromatolites: a synopsis of biocrystallization by prokaryotic and eukaryotic algae. Sedimentology, 45(3), 535– 563. H AMMER , Ø., D YSTHE , D. K. & J AMTVEIT , B. 2010. Travertine terracing: patterns and mechanisms. In: P EDLEY , H. M. & R OGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 345–355. H EVESI , A. 1970. Az algal ek mohak szerepe a bukki forrasmeszko kepzodeseben. Botanikai Ko¨zleme´nyek, 57, 233– 244. I RION , G. & M U¨ LLER , G. 1968. Mineralogy, petrology and chemical composition of some calcareous tufa
from swabische alb, germany. In: M U¨ LLER , G. & F REIDMAN , G. M. (eds) Recent developments in carbonate sedimentology in Central Europe. SpringerVerlag, New York, 156 –171. J ANSSEN , A., S WENNEN , R., P ODOOR , N. & K EPPENS , E. 1999. Biological and diagenetic influence in recent and fossil tufa deposits from Belgium. Sedimentary Geology, 126(1– 4), 75–95. K ANO , A., M ATSUOKA , J., K OJO , T. & F UJII , H. 2003. Origin of annual laminations in tufa deposits, southwest Japan. Palaeogeography Palaeoclimatology Palaeoecology, 191(2), 243–262. K AWAI , T., K ANO , A., M ATSUOKA , J. & I HARA , T. 2006. Seasonal variation in water chemistry and depositional processes in a tufa-bearing stream in SW-Japan, based on 5 years of monthly observations. Chemical Geology, 232(1– 2), 33–53. L EDGER , M. E., H ARRIS , R. M. L., A RMITAGE , P. D. & M ILNER , A. M. 2008. Disturbance frequency influences patch dynamics in stream benthic algal communities. Oecologia, 155(4), 809–819. L IU , Z. H., L IU , X. L. & L IAO , C. J. 2008. Daytime deposition and nighttime dissolution of calcium carbonate controlled by submerged plants in a karst spring-fed pool: Insights from high time-resolution monitoring of physico-chemistry of water. Environmental Geology, 55(6), 1159–1168. M ERZ , M. 1992. The biology of carbonate precipitation by cyanobacteria. Facies, 26(1), 81– 101. P ATTERSON , C. S., B USEY , R. H. & M ESMER , R. E. 1984. Second ionization of carbonic acid in NaCl media to 250 8C. Kaupia-Darmsta¨dter Beitra¨ge Naturgeschichte, 13(9), 647– 661. P EDLEY , H. M. 1994. Prokaryote microphyte biofilms: a sedimentological perspective. Kaupia-Darmsta¨dter Beitra¨ge Naturgeschichte, 4, 45–60. P EDLEY , H. M., R OGERSON , M. & M IDDLETON , R. 2009. The growth and morphology of freshwater calcite precipitates from in vitro mesocosm flume experiments. Sedimentology, 56, 511–527. P EDLEY , M. & R OGERSON , M. 2010. In vitro investigations of the impact of different temperature and flow velocity conditions on tufa microfabric. In: P EDLEY , H. M. & R OGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 145– 162. P ENTECOST , A. 1978. Blue– green-algae and freshwater carbonate deposits. Proceedings of the Royal Society of London Series B-Biological Sciences, 200(1138), 43–61. P ENTECOST , A. 1987. Growth and calcification of the fresh-water cyanobacterium Rivularia haematites. Proceedings of the Royal Society of London Series B-Biological Sciences, 232(1266), 125–136. P ENTECOST , A. 1992. Carbonate chemistry of surface waters in a temperate karst region - the southern Yorkshire dales, UK. Journal of Hydrology, 139(1–4), 211–232. R OGERSON , M., P EDLEY , H. M., W ADHAWAN , J. D. & M IDDLETON , R. 2008. New insights into biological influence on the geochemistry of freshwater carbonate deposits. Geochimica et Cosmochimica Acta, 72, 4976– 4987.
MICROBIAL INFLUENCE ON TUFA STREAMS S HIRAISHI , F., B ISSETT , A., D E B EER , D., R EIMER , A. & A RP , G. 2008a. Photosynthesis, respiration and exopolymer calcium-binding in biofilm calcification (Westerho¨fer and Deinschwanger Creek, Germany). Geomicrobiology Journal, 25(2), 83– 94. S HIRAISHI , F., R EIMER , A., B ISSETT , A., D E B EER , D. & A RP , G. 2008b. Microbial effects on biofilm calcification, ambient water chemistry and stable isotope records in a highly supersaturated setting (Westerho¨fer Bach, Germany). Palaeogeography Palaeoclimatology Palaeoecology, 262(1– 2), 91– 106. S HIRAISHI , F., Z IPPEL , B., N EU , T. R. & A RP , G. 2008c. In situ detection of bacteria in calcified biofilms using fish and card-fish. Journal of Microbiological Methods, 75(1), 103–108. S ORENSEN , J. A. & G LASS , G. E. 1987. Ion and temperature dependence of electrical conductance for natural waters. Analytical Chemistry, 59(13), 1594–1597.
81
S PIRO , B. & P ENTECOST , A. 1991. One day in the life of a stream - a diurnal inorganic carbon mass balance for a travertine-depositing stream (waterfall beck, Yorkshire). Geomicrobiology Journal, 9(1), 1– 11. T AKASHIMA , C. & K ANO , A. 2008. Microbial processes forming daily lamination in a stromatolitic travertine. Sedimentary Geology, 208(3–4), 114 –119. V ISSCHER , P. T. & S TOLZ , J. F. 2005. Microbial mats as bioreactors: Populations, processes, and products. Palaeogeography Palaeoclimatology Palaeoecology, 219(1– 2), 87–100. W EYL , P. K. 1958. The solution kinetics of calcite. The Journal of Geology, 66, 163–176. Z EEBE , R. E. 1999. An explanation of the effect of seawater carbonate concentration on foraminiferal oxygen isotopes. Geochimica et Cosmochimica Acta, 63(13– 14), 2001–2007.
Tufa-forming biofilms of German karstwater streams: microorganisms, exopolymers, hydrochemistry and calcification GERNOT ARP1*, ANDREW BISSETT2†, NICOLE BRINKMANN3, SYLVIE COUSIN4#, DIRK DE BEER2, THOMAS FRIEDL3, KATHRIN I. MOHR3‡, THOMAS R. NEU5, ANDREAS REIMER1, FUMITO SHIRAISHI1§, ERKO STACKEBRANDT4 & BARBARA ZIPPEL5 1
University of Go¨ttingen, Geoscience Centre, Goldschmidtstraße 3, D-37077 Go¨ttingen, Germany
2
Max Planck Institute for Marine Microbiology, Celsiusstraße 1, D-28359 Bremen, Germany 3
University of Go¨ttingen, Albrecht-von-Haller-Institute for Plant Sciences, Experimental Phycology and Culture Collection of Algae (SAG), Nikolausberger Weg 18, D-37073 Go¨ttingen 4
German Collection of Microorganisms and Cell Cultures DSMZ, Inhoffenstraße 7 B, D-38124 Braunschweig, Germany
5
†
Helmholtz Centre for Environmental Research UFZ, Department of River Ecology, Bru¨ckstraße 3a, D-39114 Magdeburg, Germany
Present address: CSIRO, Plant Industry, P. O. Box 1600, Canberra, ACT 2601, Australia ‡
Present address: Helmholtz Centre for Infection Research, Inhoffenstraße 7, D-38124 Braunschweig, Germany §
Present address: Division of Evolution of Earth Environment, Graduate School of Social and Cultural Studies, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan #
Institut Pasteur, Genopole de l’Ile de France, PF1, Paris, France *Corresponding author (e-mail:
[email protected])
Abstract: To understand mechanisms of tufa biofilm calcification, selected karstwater stream stromatolites in Germany have been investigated with regard to their hydrochemistry, biofilm community, exopolymers, physicochemical microgradients, calcification pattern and lamination. In stream waters, CO2 degassing drives the increase in calcite saturation to maximum values of approximately 10-fold, independent from the initial Ca2þ/alkalinity ratio. For the cyanobacteria of tufa biofilms, a culture-independent molecular approach showed that microscopy of resinembedded biofilm thin sections underestimated the actual diversity of cyanobacteria, i.e. the six cyanobacteria morphotypes were opposed to nine different lineages of the 16S rDNA phylogeny. The same morphotype may even represent two genetically distant cyanobacteria and the closest relatives of tufa biofilm cyanobacteria may be from quite different habitats. Diatom diversity was even higher in the biofilm at the studied exemplar site than that of the cyanobacteria, i.e. 13 diatom species opposed to 9 cyanobacterial lineages. The non-phototrophic prokaryotic biofilm community is clearly different from the soil-derived community of the stream waters, and largely composed of flavobacteria, firmicutes, proteobacteria and actinobacteria. The exopolymeric biofilm matrix can be divided into three structural domains by fluorescence lectin-binding analysis. Seasonal and spatial variability of these structural EPS domains is low in the investigated streams. As indicated by microsensor data, biofilm photosynthesis is the driving mechanism in tufa stromatolite formation. However, photosynthesis-induced biofilm calcification accounts for only 10–20% of the total Ca2þ loss in the streams, and occurs in parallel to inorganic precipitation driven by CO2-degassing within the water column and on biofilm-free surfaces. Annual stromatolite laminae reflect seasonal changes in temperature and light supply. The stable carbon isotope composition of the laminae is not affected by photosynthesis-induced microgradients, but
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 83– 118. DOI: 10.1144/SP336.6 0305-8719/10/$15.00 # The Geological Society of London 2010.
84
G. ARP ET AL. mirrors that of the bulk water body only reflecting climate fluctuations. Tufa stromatolites with their cyanobacterial–photosynthesis-related calcification fabrics form an analogue to porostromate cyanobacterial stromatolites in fossil settings high in CaCO3 mineral supersaturation but comparatively low in dissolved inorganic carbon. Here, the sum-effect of heterotrophic exopolymerdegradation and secondary Ca2þ-release rather decreases calcite saturation, contrary to settings high in dissolved inorganic carbon such as soda lakes.
Geobiological significance of tufa stromatolites Tufa-forming phototrophic biofilms are widespread in freshwater streams in European regions where Palaeozoic, Mesozoic and Cenozoic limestone formations are subject to present-day karstification (Ford & Pedley 1996). With respect to Germany, areas of outcropping Devonian reef carbonates (Herzynian Fold Belt with Rhenish Slate Mountains and Harz Mountains; e.g. Sommermeier 1913), Middle Triassic and Upper Jurassic epicontinental limestones (escarpments and plateaus of the
Muschelkalk Group: Usdowski et al. 1979; Salzwedel 1992; Malm Group in Weser and Leine Hills: Menzel 1909; Weißjura Group of the FranconianSwabian Alb: Burger 1911; Stirn 1964; Gru¨ninger 1965; Merz-Preiß & Riding 1999; Arp et al. 2001b), and Upper Cretaceous chalk deposits (e.g. Mu¨nsterland: Hartkopf-Fro¨der et al. 1989) are favourable for present-day fluvial tufa deposition (Fig. 1). In addition, a number of small streams and rivers of the Northern Alpine Molasse Basin show tufa stromatolites and oncoids due to Ca – HCO3-dominated groundwaters discharging from Miocene-Pliocene gravel and Quarternary till
Fig. 1. Location of the studied tufa-forming karstwater streams in Germany.
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
derived from the Northern Limestone Alps (e.g. Rott 1991; Ha¨gele et al. 2006; Arp 2008). Indeed, tufa-forming biofilms are one of the few natural biofilm systems in Europe where processes involved in biofilm calcification and present-day stromatolite formation can be studied. Also, tufa deposits form important climate archives because of high depositional rates and annual lamination (e.g. Andrews et al. 2000; Matsuoka et al. 2001; Garnett et al. 2004; Kano et al. 2004; Andrews & Brasier 2005; Pentecost 2005). Tufa deposits in Germany were first recognized by scientists in the 18th century (Baier 1708: ‘tophus’; Schu¨tte 1761: ‘Tuff’; Walch 1773: ‘Tophsteine’) and were described as porous friable lime deposits of freshwater streams encrusting moss, leaves and plant stems. Later, four different processes have been suggested as the primary cause of tufa (i.e. cool water travertine) formation: (1) Trapping of calcite particles, either derived from older limestone formations (e.g. von Buch 1809: 24), or precipitated within the water column (La Touche 1913: 326), (2) physicochemical CO2 degassing (e.g. Unger 1861: 509; Burger 1911; Schu¨rmann 1918), (3) photosynthetic CO2 assimilation by cyanobacteria, mosses and other aquatic plants (e.g. Cohn 1864: 592; Pokorny 1865: 35), and (4) adsoption [of Ca2þ] to plant surfaces (Kla¨hn 1923: 303). Subsequently, authors either emphasized the impact of photosynthesis on the carbonate equilibrium (e.g. Wallner 1934: 12), or that of inorganic CO2 degassing (e.g. Usdowski et al. 1979; Herman & Lorah 1987; see also von Pia 1933: 18). In fact, conclusions summarized by Gru¨ninger (1965) do not differ too much from the present-day view (e.g. Golubic´ 1973; Pentecost 1978, 1993, 2005; Ford 1989; Merz-Preiß & Riding 1999; Pedley 2000; Arp et al. 2001b; Zhang et al. 2001): tufa formation is driven by evasion of carbon dioxide, independent from day –night cycles of light and temperatures. Aquatic cryptogams (cyanobacteria) provide the site of calcite crystal growth within their mucus (Gru¨ninger 1965: 92). In addition, moss plants enhance calcite deposition by baffling of tiny calcite crystals (‘fyke nets’) from the water column, and by forming steps in the water course with increased turbulence and hence CO2 evasion (Gru¨ninger 1965: 93). More recent investigations focussed on the quantification of the relative impact of the different processes by stable isotope analysis (Spiro & Pentecost 1991), the potential effect of exopolymers on calcite nucleation (Borowitzka 1982; Pentecost 1985) and the possible impact of non-phototrophic prokaryotes on calcite precipitation (Caudwell 1987; Pentecost & Therry 1988; Szulc & Smyk 1994).
85
In parallel to the studies mentioned above, there was an increasing interest in the species composition of the algal community involved in tufa formation. However, while some descriptions mention only a limited number of ‘key species’ (e.g. Phormidium incrustatum (Naegeli) Gomont to form major parts of the biofilms (Fritsch 1949; Arp et al. 2001b), other authors have demonstrated a much larger biodiversity (Freytet & Verrecchia 1998; Rott 1994; Reichardt 1994; Pentecost & Whitton 2000). So far, non-phototrophic prokaryotes were investigated only rarely (Caudwell 1987; Pentecost & Therry 1988; Szulc & Smyk 1994; Ng et al. 2006). Biofilm composition may change with seasons, and corresponding tufa stromatolites (Riding 1991) display a clear lamination. However, the assignment of laminae to seasons, important for palaeoclimate reconstructions, differs between study sites (e.g. Stirn 1964: 13; Golubic´ 1969; Geurts 1976; Monty 1976; Freytet & Plet 1996; Janssen et al. 1999; Pentecost & Whitton 2000; Arp et al. 2001b; Kano et al. 2003), either because of differing ecological parameters, or different methods of investigation applied. Consequently, key questions in the study of biofilms and tufa formation presented here were as follows: (1) What is the identity and diversity of the most abundant phototrophic and non-phototrophic microorganisms within the tufa-forming biofilms? Which non-phototrophic prokaryotes are present and what kind of metabolic properties can be derived from phylogenetic assessment? (2) Is calcification linked to specific clusters of extracellular polymeric substances (EPS) within the biofilms? Are these EPS cell-surface associated or detached from cells and are they related to specific species? (3) Does tufa stromatolite formation result from passive calcite impregnation of biofilm due to high calcite supersaturation by physicochemical CO2 degassing, or does cyanobacterial photosynthesis drive or assist calcite precipitation within the biofilms? (4) Do non-phototrophic prokaryotes support or inhibit calcite precipitation within the biofilms by their physiological (heterotrophic) activity, e.g. exopolymer degradation. (5) Do geochemical signatures (especially d13C) directly reflect macroenvironmental changes only (climate), or are they modified by microbial activity within the biofilms? This paper summarizes published (Table 1) and unpublished results of tufa biofilm studies carried out by members of the DFG (German Research Foundation) research unit 571 ‘Geobiology of Organo- and Biofilms’.
86
G. ARP ET AL.
Table 1. Published investigations on tufa stromatolite biofilms of the Deinschwanger Bach (DB) and Westerho¨fer Bach (WB) Topic
Location
Hydrochemistry of stream waters (selected data) Diatom biodiversity (preliminary) Cyanobacterial morphotypes
DB, WB
Non-phototrophic prokaryote biodiversity of stream waters Non-phototrophic prokaryote biodiversity of tufa stromatolite (preliminary) Novel species of non-phototrophic prokaryotes
WB
Extracellular polymeric substances Physicochemical microgradients and mass balances Stable carbon and oxygen isotopes of tufa carbonate and stream waters Hydrochemical numerical simulations
DB, WB
Methods
References
electrodes, titration, ion chromatography, photometry 18S rDNA sequences from clones and cultures epifluorescence microscopy of resin-embedded biofilm sections 16S rDNA sequences from cultures
Arp et al. 2001b; Shiraishi et al. 2008a, b Brinkmann et al. 2007a, b
WB
16S rDNA sequences from clones and cultures
Cousin et al. 2007
WB
Cousin et al. 2007; Muurholm 16S rDNA sequences from et al. 2007; Stackebrandt cultures, GþC content of et al. 2007; Ali et al., DNA, fatty acid methyl ester in press analysis; cultural, biochemical and morphological analyses lectin-binding, LSM Zippel et al., submitted
DB, WB DB
Arp et al. 2001b Brambilla et al. 2007; Cousin et al. 2008
DB, WB
microsensors and ion flux calculations
Bissett et al. 2008a; Shiraishi et al. 2008a, b
DB, WB
mass spectrometry
Arp et al. 2001b; Shiraishi et al. 2008b
marine & non-marine
computer program PHREEQC (Parkhurst & Appelo 1999)
Arp et al. 2001a, 2003
Locations Two karstwater streams in Germany have been studied in detail (Fig. 1). Coordinates are given according to the Gauss–Kru¨ger system. The Westerho¨fer Bach (WB, Fig. 2a) is located to the west of the Harz Mountains, c. 27 km NNE of Go¨ttingen (Geological Map of Lower Saxony 1:25 000 Sheet 4226 Northeim– Ost, east 35 75 320, north 57 37 118 [main spring; 199 m above S.L.] to east 35 75 625, north 57 37 010 [lower stream section; 154 m above S.L.]). The stream is less than 2 m wide and receives its water from a spring area discharging from the Middle Triassic Muschelkalk-Group aquifer. The catchment area is approximately 3.5 km2 and covered exclusively by deciduous and coniferous forest. The investigated upstream section of the stream is approximately 340 m long with a difference of 40 m in altitude. This stream previously has been studied by Jacobson & Usdowski (1975), Usdowski et al. (1979) and Dreybrodt et al. (1992) with regard to hydrochemistry and isotope geochemistry.
The Deinschwanger Bach (DB, Fig. 2b) is located at the western margin of the Franconian Alb, approximately 30 km ESE of Nu¨rnberg (Geological Map of Bavaria 1:25 000 Sheet 6634 Altdorf bei Nu¨rnberg, east 44 63 700, north 54 72 450 [main spring; 519 m above S.L.] to east 44 61 500, north 54 72 250 [low-turbulent, well-illuminated lower stream section; 462 m above S.L.]). The main spring as well as most side springs discharge from the base of the Weißjura-Group aquifer, where thick Upper Jurassic limestones (here: ‘Weißjura a and b’, Dietfurt Formation) are underlain by a few metres of clays of the Middle Jurassic Ornatenton Formation (‘Braunjura 13 and z ’) (Schmidt-Kaler 1974). It is one of the most important spring horizons of the Franconian Alb. The stream has been investigated at a total length of 2800 m, and attains a maximum width of 2 m. The total catchment area of the investigated stream section including all side springs is approximately 8 km2. Valley sides are vegetated by mixed forests while the surrounding karst plateau is intensively used for agriculture.
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
87
Fig. 2. Field views of tufa-forming karstwater streams in Germany investigated in this study. (a) Westerho¨fer Bach, western Harz Foreland. (b) Deinschwanger Bach, Franconian Alb. (c) Reinsgraben at the Schillerwiese in Go¨ttingen. (d) ‘Steinerne Rinne’ at Erasbach, Franconian Alb. Image courtesy of Ch. Heim.
Additional hydrochemical analyses have been carried out at two further streams (Fig. 1): (1)
(2)
The Reinsgraben (RG, Fig. 2c) is situated at the eastern city margin of Go¨ttingen, Leine Hills (Geological Map of Lower Saxony 1:25 000 Sheet 4425 Go¨ttingen, east 35 66 625, north 57 11 675 [spring, 215 m above S.L.] to east 35 65 780, north 57 11 500 [lower stream section, 172 m above S.L.]). The spring is fed from the Middle Triassic Muschelkalk-Group aquifer (i.e. similar to Westerho¨fer Bach), at the eastern main fault of the Leinetalgraben, where impermeable Lower Jurassic claystones force the groundwater to discharge (von Koenen 1907; Stille 1932). The catchment area is approximately 1 km2 and covered by deciduous forest. The beck has been investigated at a length of 720 m, where it crosses the park area of the Schillerwiese. The Steinerne Rinne (SR, Fig. 2d) is situated 1.3 km south of the village Erasbach, southern Franconian Alb (Geological Map of Bavaria 1:25 000 Sheet 6834 Berching, east 44 57
750, north 54 44 250 [spring; 509 m above S.L.] to east 44 57 790, north 54 44 360 [lower end of tufa canal; 486 m above S.L.]). The spring of the Steinerne Rinne at Erasbach discharges from the base of the Weißjuraaquifer, similar to the Deinschwanger Bach (Meyer & Schmidt-Kaler 1983; Glassl & Schieber 1990; Baier 2002). The catchment area is less than 1 km2, covered exclusively by deciduous and coniferous forest. This ‘selfbuilt tufa canal’ of the Erasbach rivulet has a total length of c. 80 m and attains a maximum height of 77 cm (Baier 2002), and is now protected as a nature monument. There are no additional side springs or seepage sites contributing water to this tiny rivulet.
Hydrochemistry of stream waters General characteristics of spring and stream waters The investigated karstwater streams either belong to the calcium-bicarbonate (Deinschwanger Bach
88
G. ARP ET AL.
and Steinerne Rinne Erasbach) or the calciumbicarbonate-(sulphate) type (Westerho¨fer Bach and Reinsgraben) characterized by Ca2þ concentrations of 2– 5 mmol L21 and high total alkalinities ranging from 4–6 meq L21 (Table 2; Jacobsen & Usdowski 1975; Usdowski et al. 1979; Dreybrodt et al. 1992; Arp et al. 2001b; Baier 2002; Shiraishi et al. 2008a, b). Both streams originating in the evaporitecontaining Muschelkalk-Group aquifer (WB and RG) are comparatively enriched in Ca2þ (3.0– 5.2 mmol L21), Mg2þ (1.1–1.8 mmol L21), and (1.8–3.9 mmol L21). Also, these streams SO22 4 have high concentrations of strontium and barium. The Deinschwanger Bach and Erasbacher rivulet (SR) discharge from Mg-poor limestone aquifers of the Weißjura-Group resulting in lower Ca2þ concentrations (1.8–3.5 mmol L21) and Mg2þ concentrations (0.1–1.2 mmol L21), and an order concentrations. On the of magnitude lower SO22 4 other hand, enhanced nitrate, nitrite, and phosphate concentrations in case of the Deinschwanger Bach are due to agriculture in its catchment area (Table 2).
Hydrochemical trends during the course of the streams The common hydrochemical evolution, especially that of the carbonate system, along the flow path of the investigated streams is displayed in Figure 3. In general, the streams start at initially near-neutral pH (7.0–7.3), high partial CO2 pressure (PCO2 ¼ 10000–15000 matm), and consequently low saturation with respect to calcite, the solely precipitating carbonate mineral phase. Calcite saturation of the spring waters, defined as the saturation index SICalcite ¼ log VCalcite ¼ log(ion activity product fCa2þg fCO22 3 g/ amounts to solubility product Kcalcite), SICalcite ¼ 0.0– 0.25, indicating that spring waters are already saturated to slightly supersaturated. Because of CO2-degassing, the initially high PCO2 of the spring waters rapidly decreases to a level of about 800 –1100 matm along a flow path of some tenths to some hundreds of meters. This PCO2 decrease is coupled with an increase of pH from c. 7.3 at the spring sites to maximum values of 8.3–8.4. According to the rise in pH, calcite supersaturation increases along the flow path attaining maximum values of SICalcite ¼ 0.9–1.2 (i.e. 8- to 15-fold supersaturation VCalcite), and CaCO3 precipitation becomes evident by the remarkable loss in Ca2þ and total alkalinity. First occurrences of tufa carbonates coincide with the beginning of this decrease in calcium and alkalinity. Both, calcium and total alkalinity further decline downstream in fast flowing reaches or at major tufa cascades where PCO2 drop is enhanced. At downstream
sites where high calcium loss is achieved equilibration of the carbonate system in course of precipitation accounts for a minor decrease in saturation values and pH (Fig. 3). Similar to the calcium profile, strontium and barium concentrations show a downstream decrease suggesting that minor amounts are co-precipitated with the calcite. While the concentrations of ions such as Naþ, Kþ, Mg2þ and Cl2 keep very constant throughout the course of all streams, in case of the Westerho¨fer Bach, the detected decrease of SO22 4 concentrations also indicates incorporation of sulphate during carbonate precipitation. From elemental analysis of the tufa carbonate, sulphur incorporated in calcite here amounts to 0.2– 0.3 decreases percent by weight. Additionally, PO32 4 during the course of the stream (WB), which is either consumed by primary production or also bound to the precipitated carbonate phases. Where 2 analysed, NO2 3 and NO2 concentrations appear to be constant (WB, DB, RG), indicating that nitrogen is not limiting with regard to primary production.
Annual and diurnal cycles Waters emanating from the spring sites are rather constant in pH and temperature throughout the year. Depending on air temperature, stream water temperatures increase downstream in spring and summer time and decrease in winter time. Yet, main hydrochemical parameters generally show only minor seasonal variation (Jacobson & Usdowski 1975; Shiraishi et al. 2008b). Non-cyclic changes in water chemistry in case of the Westerho¨fer Bach reflect a dependence on rainfall and runoff intensity rather than seasonal fluctuations (Fig. 4). Here, intense rainfall in the catchment area leads to enhanced underground dissolution of evaporites within the Muschelkalk-strata resulting in increased sulphate and calcium concentrations which are, in turn, balanced by slightly lower total alkalinity (Fig. 4). Nevertheless, the different hydrochemical conditions hardly affect the commonly observable pattern of calcite supersaturation and calcium loss along the flow path (Jacobson & Usdowski 1975; Dreybrodt et al. 1992; Shiraishi et al. 2008b; Fig. 3). Also, the highly turbulent streams do not reveal a clear diurnal change in bulk water hydrochemistry with exception of a diurnal temperature cycle (Shiraishi et al. 2008b).
Conclusions (1)
Spring waters of all investigated streams have a much higher CO2 pressure than the atmosphere and rapidly start degassing of CO2 when leaving the spring site. Equilibration of the carbonate system results in
Table 2. Hydrochemical data of spring and stream waters of the investigated karstwater streams. Saturation index SICalcite ¼ log VCalcite ¼ log(ion activity product fCa 2þg fCO22 3 g/solubility product Kcalcite) Sr2þ
Ba2þ [mmol L21]
Cl2
0.344 0.338 0.341 0.335 0.335 0.336 0.336 0.336 0.335 0.333 0.338
0.052 0.055 0.053 0.054 0.053 0.053 0.053 0.054 0.054 0.052 0.053
0.0190 0.0190 0.0186 0.0188 0.0187 0.0182 0.0180 0.0180 0.0179 0.0177 0.0175
0.00024 0.00024 0.00025 0.00025 0.00025 0.00022 0.00023 0.00022 0.00023 0.00021 0.00022
0.298 0.299 0.297 0.293 0.296 0.293 0.291 0.291 0.294 0.292 0.291
1.27 1.09 1.03 1.03
0.314 0.335 0.320 0.319
0.0304 0.0312 0.0284 0.0289
0.00032 0.00038 0.00043 0.00043
0.00013 0.00013 0.00011 0.00011
2.28 2.27 2.23 2.17 2.11
0.91 0.91 0.86 0.86 0.86
0.348 0.349 0.331 0.334 0.334
0.0582 0.0578 0.0573 0.0568 0.0575
0.00048 0.00048 0.00055 0.00054 0.00054
2.09 2.07
1.00 1.03
0.322 0.316
EC 25 8C [mS cm21]
O2 [mmol L21]
p1
pH
8.9 9.0 9.1 9.1 9.3 9.4 9.7 10.0 10.3 9.9 10.1
1037 1016 1025 1011 1004 994 975 965 960 903 951
0.197 0.181 0.289 0.309 0.313 0.324 0.323 0.324 0.322 0.321 0.331
9.32 9.33 8.83 8.40 8.00 7.96 7.76 7.52 7.17 7.75 7.32
7.33 7.32 7.82 8.11 8.25 8.30 8.30 8.27 8.29 8.10 8.22
5.40 5.38 5.36 5.32 5.30 5.24 4.92 4.86 4.82 4.84 4.80
3.95 3.91 3.90 3.88 3.85 3.80 3.68 3.64 3.61 3.58 3.59
1.72 1.72 1.72 1.71 1.71 1.71 1.71 1.70 1.71 1.67 1.71
Deinschwanger Bach 9–11th October 2006 upper main stream: 03DB-14 0 03DB-12 1631 03DB-10 1916 03DB-05 2576
8.4 8.5 9.3 9.5
614 603 602 597
0.293 0.340 0.326 0.326
9.32 8.69 8.15 7.73
7.36 8.35 8.37 8.42
5.22 5.14 5.12 5.08
1.87 2.03 2.10 2.07
northern tributary: 03DB-07 0 03DB-08 53 03DB-01 93 03DB-02 132 03DB-03 173
9.0 9.2 9.7 9.9 11.5
637 628 605 594 584
0.321 0.325 0.327 0.317 0.302
7.69 7.40 7.24 7.26 7.87
7.99 8.46 8.50 8.48 8.44
5.14 5.08 4.92 4.80 4.68
lower main stream: 03DB-04 2599 03DB-06 2826
9.5 8.6
596 596
0.322 0.334
7.64 7.40
8.42 8.46
5.04 5.04
Distance from spring [m]
Si
NO2 3
2.94 2.93 2.95 2.93 2.92 2.91 2.90 2.89 2.89 2.81 2.87
0.159 0.159 0.158 0.159 0.159 0.158 0.158 0.159 0.159 0.157 0.158
0.093 0.090 0.092 0.092 0.092 0.091 0.090 0.087 0.088 0.083 0.083
0.00025 11482 0.00024 12023 0.00024 3631 0.00025 1862 0.00019 1318 0.00016 1148 0.00014 1096 0.00013 1175 0.00014 1096 0.00019 1738 0.00014 1318
0.21 0.20 0.69 0.96 1.10 1.14 1.10 1.07 1.09 0.90 1.01
0.550 0.562 0.558 0.556
0.19 0.18 0.18 0.18
0.102 0.101 0.101 0.101
0.504 0.485 0.483 0.482
0.00148 10715 0.00203 1047 0.00191 1000 0.00182 871
0.01 1.00 1.04 1.09
0.00012 0.00013 0.00012 0.00012 0.00012
0.650 0.650 0.617 0.615 0.614
0.17 0.17 0.17 0.17 0.17
0.097 0.097 0.098 0.098 0.099
0.617 0.615 0.598 0.595 0.592
0.00284 0.00295 0.00239 0.00231 0.00228
2455 794 692 708 794
0.72 1.15 1.19 1.15 1.11
0.0327 0.00044 0.00011 0.0303 0.00042 0.00011
0.559 0.557
0.18 0.18
0.101 0.101
0.489 0.00184 0.486 0.00185
871 776
1.09 1.11
SO22 4
PO32 4 !j
PCO2 SICalcite [matm] [log IAP/ KT]
Westerho¨fer Bach 23rd May 2006 03WB-W01 03WB-W02 03WB-W03 03WB-W04 03WB-W05 03WB-W06 03WB-W07 03WB-W08 03WB-W09 03WB-W10 03WB-W11
0.5 3 50 130 183 234 262 295 308 310 339
89
(Continued)
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
Kþ
Alkalinity Ca2þ Mg2þ Naþ [meq L21] j
T [8C]
Sample
90
Table 2. Continued Sample
Distance from spring [m]
T [8C]
EC 25 8C [mS cm21]
O2 [mmol L21]
11.4 11.3 11.4 10.8 10.9 10.8 10.8 11.8 12.8 12.8 13.2 13.4 14.3 14.1 14.5 14.7
1230 1226 1229 1222 1207 1218 1214 1202 1200 1193 1175 1175 1138 1147 1155 1144
0.200 0.236 0.250 0.280 0.272 0.281 0.313 0.337 0.342 0.373 0.381 0.378 0.371 0.344 0.360 0.344
8.2 8.6 8.7 8.9 9.2 10.0 10.2 10.9
642 645 635 627 610 578 570 524
0.308 0.357 0.370 0.371 0.369 0.369 0.360 0.350
p1
pH
Alkalinity Ca2þ Mg2þ Naþ [meq L21] j
Kþ
Sr2þ
Ba2þ [mmol L21]
Cl2
SO22 4
Si
NO2 3
0.150 0.154 0.153 0.153 0.153 0.153 0.155 0.154 0.153 0.153 0.151 0.152 0.154 0.148 0.148 0.148
0.058 0.059 0.057 0.060 0.059 0.060 0.058 0.060 0.059 0.060 0.059 0.061 0.062 0.062 0.061 0.063
PO32 4 !j
PCO2 SICalcite [matm] [log IAP/ KT]
Reinsgraben 6th August 2002 55 82 108 113 138 144 154 187 268 315 406 455 541 581 616 708
7.31 7.38 7.46 7.49 7.60 7.62 7.69 7.78 8.10 8.21 8.27 8.28 8.26 8.24 8.26 8.23
5.20 5.18 5.20 5.20 5.17 5.17 5.15 5.13 5.04 4.97 4.88 4.77 4.65 4.66 4.66 4.63
5.19 5.20 5.19 5.18 5.17 5.16 5.16 5.15 5.12 5.08 5.01 4.97 4.93 4.91 4.90 4.89
1.13 1.12 1.13 1.12 1.11 1.11 1.11 1.12 1.12 1.13 1.12 1.11 1.11 1.11 1.12 1.13
0.610 0.607 0.606 0.610 0.619 0.619 0.619 0.613 0.617 0.621 0.619 0.613 0.621 0.617 0.623 0.623
0.044 0.044 0.045 0.044 0.044 0.044 0.044 0.044 0.044 0.045 0.044 0.046 0.045 0.044 0.043 0.043
7.10 7.41 7.78 7.96 7.97 8.01 8.03 7.99
6.10 6.08 5.98 5.48 5.27 4.86 4.72 4.42
3.52 3.52 3.50 3.43 3.33 3.13 3.07 2.86
0.137 0.138 0.135 0.134 0.134 0.133 0.139 0.137
0.202 0.134 0.191 0.188 0.191 0.182 0.127 0.175
0.014 0.022 0.016 0.017 0.016 0.016 0.021 0.015
0.593 0.594 0.593 0.593 0.586 0.594 0.585 0.589 0.584 0.592 0.596 0.572 0.548 0.556 0.546 0.558
3.80 3.78 3.79 3.78 3.79 3.78 3.76 3.77 3.76 3.74 3.76 3.73 3.71 3.69 3.63 3.64
0.00126 11749 0.00119 10000 0.00171 8318 0.00163 7762 0.00139 6026 0.00155 5754 0.00159 4786 0.00155 3890 0.00137 1820 0.00124 1380 0.00063 1175 0.00048 1122 0.00064 1175 0.00039 1230 0.00041 1175 0.00045 1259
0.31 0.38 0.46 0.48 0.58 0.60 0.67 0.77 1.08 1.18 1.22 1.22 1.21 1.19 1.21 1.18
0.140* 0.27** 0.084** 0.081* 0.00263* 22909 10965 4571 2754 2630 2239 2089 2138
0.08 0.39 0.75 0.88 0.87 0.87 0.87 0.79
Steinerne Rinne Erasbach 13th October 2001 SRE 01/0 SRE 01/1 SRE 01/2 SRE 01/3 SRE 01/4 SRE 01/5 SRE 01/6 SRE 01/7
0 25 50 75 100 125 132 150
9.43 9.35 9.25 9.16 9.20 9.14 9.12 9.22
0.00156 0.00153 0.00153 0.00153 0.00154 0.00152 0.00151 0.00151
*From Baier (2002). **Average values of karstwaters springs from the Weißjura-aquifer of this area, map sheet 6834 Berching (Schmidt-Kaler 1981).
G. ARP ET AL.
P04 P06 P07 P08 P09 P10 P11 P12 P13 P14 P15 P16 P17 P18 P19 P20
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS 91
Fig. 3. Hydrochemical evolution of karstwaters from spring sites downstream, during successive CO2 degassing. (a) Westerho¨fer Bach, western Harz Foreland. (b) Deinschwanger Bach, Franconian Alb. (c) Reinsgraben at the Schillerwiese in Go¨ttingen. (d) ‘Steinerne Rinne’ of the Erasbach rivulet, Franconian Alb. Note that the hydrology of the Deinschwanger Bach is more complex than shown, with additional influx from further small tributaries and seepage sites from the Eisensandstein Formation.
92
G. ARP ET AL.
Fig. 4. Variations of Ca2þ concentrations and calcite supersaturation index (SICalcite) at the Westerho¨fer Bach dependent on run-off. Higher water run-off (August 2007: c. 6 –7 L s21) results in higher Ca2þ concentrations due to dissolution of evaporites in the catchment area, compared to times of lower run-off (May 2006: c. 3– 4 L s21; March 2007: c. 2 L s21). SICalcite reaches maximum values later during the course of the stream at times of higher run-off.
(2)
rising pH, and while HCO2 3 , which is several orders of magnitude higher in concentration, remains virtually constant, CO22 3 activity has to increase in the bulk water. At constant Ca2þ then rising CO22 3 steadily increases the ion activity product of calcium carbonate and supersaturation of the stream water with respect to calcite. There is no detectable impact of photosynthesis on bulk water hydrochemistry of the investigated karstwater streams. Due to a nucleation barrier (De Yoreo & Vekilov 2003), calcite precipitation does obviously not commence significantly until a supersaturation of about 10-fold (SICalcite of 0.9 to 1.1) is reached, similar to several other field investigations of karstwater streams (e.g. Dandurand et al. 1982; Suarez 1983; Herman & Lorah 1987, 1988; Liu et al. 1995; Merz-Preiß & Riding 1999; Kano et al. 2003; Kawai et al. 2006). In the downstream sections, calcite precipitation is evident from decreasing Ca2þ concentrations and total alkalinity. Precipitation produces an equivalent amount of CO2 that simultaneously escapes from the water. At comparable rates of precipitation and degassing then, pH, calcite saturation, and PCO2 keep nearly constant as observed in the lowermost investigated downstream reaches of the streams, where waters were yet not in equilibrium with atmospheric PCO2.
(3)
From comparison of the four investigated stream data, the maximum saturation level does not depend on the initial Ca2þ to alkalinity ratio. Depending on the charge balance (i.e. via ions other than Ca2þ, HCO2 3 and CO22 3 ) streams may start degassing at a different Ca2þ/alkalinity ratio, but will reach the same saturation state (around SICalcite ¼ 1.0) necessary to start significant precipitation. Thereby, those waters with initially lower Ca2þ concentration and higher alkalinity will be higher in pH and lower in PCO2. On the other hand, contribution of foreign ions with inhibiting effect on calcite precipitation, such (Reddy 1986; Liu et al. as Mg2þ and SO22 4 1995), are the likely cause of shifting the maximum saturation level above 10-fold. Such a foreign ion effect is suggested to explain the higher maximum saturation index of 1.15-1.20 in case of the Westerho¨fer Bach, Deinschwanger Bach, and Reinsgraben which are higher in Mg2þ and SO22 4 concentrations. Instead, in the Erasbach rivulet, where Mg2þ and SO22 4 concentrations are negligible, maximum saturation index is 0.9.
Microorganisms Cyanobacteria, diatoms and non-photrophic prokaryotes, all potentially involved in CaCO3 mineral nucleation via exopolymers and/or alterating
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
CaCO3 mineral saturation within microenvironments, have been investigated at several sites of the karstwater streams WB and DB. Most comprehensive studies have been carried out at the site WB5 (Fig. 3a), which shows typical tufa stromatolites growing in highly calcite supersaturated karstwaters, that is, conditions representative for the tufa stromatolites of all the investigated streams. Results shown here focus on biofilm samples and a drilling core of 5 cm length taken at this site WB5.
Cyanobacteria and diatoms Cyanobacteria, many of them calcifying, are the most abundant primary producers in tufa-forming biofilms where they outrange eukaryotic algae (e.g. von Pia 1933; Wallner 1934; Stirn 1964; Gru¨ninger 1965; Golubic´ 1973; Pentecost 1978, 1985). Among eukaryotic microalgae, diatoms are by far the most dominant eukaryotic microalgal group in tufa-forming biofilms (e.g. Gomes 1985; Golubic´ et al. 2008) where they outnumber other heterokont algae (Xanthophyceae, Eustigmatophyceae) and green algae in these habitats. In diatoms, calcite precipitation appears to be associated with exopolymer stalks rather than photosynthetic activity (Freytet & Verrecchia 1998; Golubic´ et al. 2008). Many diatoms, however, may rather inhibit calcification by their mucilage (see e.g. Arp et al. 2001b). Identification of diatoms in tufa biofilms is an important prerequisite to further study their role in calcification processes. Biofilms of the Westerho¨fer Bach and Deinschwanger Bach are being investigated in an ongoing study for their cyanobacterial and algal diversities using phylogenetic analyses of 16S and 18S
93
rDNA sequences to unambiguously distinguish the taxa. A culture-independent cloning/sequencing approach has been employed to detect also those species that are difficult or impossible to culture. Here we report about original results of molecular studies in assessing the diversities of cyanobacteria and diatoms obtained from the site WB5, in comparison with results obtained by morphological studies using resin-embedded hardpart sections of the biofilms. Cyanobacteria: a morphological approach. Cyanobacterial morphotypes observed in this study at the site WB5 (Fig. 3a) closely correspond to morphotypes already described from DB by Arp et al. (2001b). Six cyanobacterial morphotypes can be distinguished. Phormidium incrustatum (Na¨geli) Gomont ex Gomont was the predominant morphotype at the tufa surfaces. It formed densely arranged erect filaments with trichomes 3 –5 mm wide that were enclosed in a thin firm sheath (Fig. 5a). Cells were half long as wide to isometric and cell walls were laterally inserted; there were no constrictions at cell walls. As seen by light microscopy, the thylakoids were arranged peripherically. Morphotype Lyngbya sp. also exhibited erect filaments (Fig. 5a). The trichomes were enclosed in a thin firm sheath as well, but were wider (5–7 mm). The cells appeared more disc-shaped, a quarter to half long as wide. Cell walls were also laterally inserted and showed no constrictions at cell walls. Lyngbya sp. occurred scattered between numerous filaments of P. incrustatum in the upper biofilm parts, and sometimes even on top of the biofilms (Fig. 5a). A cultured strain (WBK15) exhibited features identical to those of the Lyngbya sp. morphotype (Fig. 6). The Leptolyngbya foveolarum
Fig. 5. Laser-scanning micrographs (excitation ex 543–633 nm, emission em 565–615, 639–704 nm) of cyanobacterial morphotypes from Westerho¨fer Bach biofilm sections. (a) Phormidium incrustatum (Pho), Leptolyngbya aff. foveolarum (Lep) and Lyngbya sp. (Lyn). Lower stream site WB 5 tufa stromatolite biofilm top. 23.05.2006. (b) Pseudoanabaena sp. (Pse), Aphanothece castagnei (Aph) and Phormidium incrustatum (Pho). Lower stream site WB 5 tufa stromatolite, c. 400 mm below biofilm top. 23.05.2006. (c) Hyella fontana at the top of a limestone cobble. Spring site WB 1. 09.06.2005.
94
G. ARP ET AL.
Fig. 6. Morphology of cultured representative of the Lyngbya sp. morphotype (strain WBK15, a member of the ‘Unidentified A’ lineage, see Fig. 7). Note trichomes enclosed within a firm sheath, cells with disc-shaped appearance. Cells walls laterally inserted and no constrictions at the cell boundaries. Scale bar is 25 mm.
(Gomont) Anagnostidis & Koma´rek morphotype (P. foveolarum Montagne ex Gomont in Arp et al. 2001b; Freytet & Verrecchia 1998; Janssen et al. 1999) comprised also densely arranged erect filaments with a thin, firm sheath. The trichomes were rather thin, that is, 1.2–2.0 mm wide (Fig. 5a). Their cells were half long as wide to isometric to slightly longer than wide and slightly constricted at cell walls. The apical cells were simple, rounded or slightly tapering. The L. foveolarum morphotype formed dense meadows of erect filaments; it was particularly abundant in autumn. Pseudanabaena sp. represented another filamentous morphotype. It had a chain-like appearance, ovoid to rounded barrel-shaped cells which were 2 mm long and 1 mm wide (Fig. 5b). The filaments were without sheath. The Pseudanabaena sp. morphotype was found in the deeper biofilm parts (Fig. 5b). The Hyella fontana Huber & Jadin morphotype formed pseudofilaments from multiple fissions that were bi- or triserial at the base, but uniserial at the distal tips (Fig. 5c). Cells were rounded-polygonal to elongated, 1.5–4.0 mm in size. H. fontana occurred in deeper biofilm parts and at growth discontinuities of the tufa stromatolites. Aphanothece castagnei (Bre´bisson) Rabenhorst, mentioned previously by Thunmark (1926) and Freytet & Plet (1996), was a unicellular morphotype at both karstwater streams. It exhibited coccoid, oval-elongated cells, 2.5– 3.5 mm wide
and 4–6 mm long (Fig. 5b). The cells showed abundant cell inclusions at their periphery, exhibited a simple binary fission and were arranged in loose colonies. The sheath was diffluent and hardly visible. A. castagnei occurred only in deeper biofilm parts of the tufa stromatolites (Fig. 5b). In biofilms other than those on the tufa stromatolite surfaces, additional morphotypes were found, e.g. Chamaesiphon at moss tufa with its spar crusts and various coccoid and endolithic filamentous cyanobacteria at the non-calcifying spring sites. Often the identification of cyanobacteria morphotypes in tufa is not unambiguous in the literature, i.e. the same morphotype may be assigned to different species by different authors. Lyngbya sp. when showing disc-shaped cells and a more distinct sheath may be assigned to Lyngbya martensiana (e.g. Stirn 1964). Also, the L. foveolarum morphotype may include other filamentous cyanobacteria exhibiting filaments enclosed in a common sheath in tufa biofilms, that is, Schizothrix pulvinata (Ku¨tzing) Gomont (in Freytet & Plet 1996), S. fasciculata (Na¨geli) Gomont (in Rott 1994) and S. calcicola (C. Agardh) Gomont in Szulc & Smyk (1994). Phormidium calcareum Ku¨tzing ex Starmach is distinguished from P. incrustatum by the absence of a calyptra (Kann 1973) and, therefore, both are treated as distinct taxa by Koma´rek & Anagnostidis (2005), whereas Pentecost (2003) considers both as synonymous. Cyanobacteria: a molecular approach. 16S rDNA cloning and sequencing of environmental DNA was performed at the site WB5 to investigate the diversity of cyanobacteria of the biofilm (BF) and the following annual stromatolite laminae couplets (Core layers ‘CL’) from a 5 cm deep drilling core (Fig. 10). A total of 711 cyanobacterial clones was obtained and phylogenetically analyzed. The clones were distributed on nine independent lineages and clades of the 16S rDNA phylogeny (Fig. 7). Each lineage has also been found at other sites of the two karstwater streams in our ongoing study. At other sampling sites only three to seven cyanobacterial lineages have been retrieved from the biofilm. Altogether, a total of 12 cyanobacteria lineages which occurred at two and more sites have been recovered. Identification of the recovered cyanobacteria, however, appeared difficult because for most lineages no closest neighbouring sequences from named reference strains have yet become available. Six cyanobacteria lineages (94% of all clones) were retrieved from the biofilm (BF) with frequencies higher than 1% of all clones (Fig. 7). The biofilm was dominated by the lineages ‘Unidentified B’ and Tychonema sp. (41% and 29% of all BF clones). The Chamaesiphon sp., ‘Unidentified D’ and ‘Unidentified E’ lineages were exclusively
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
found in the biofilm on the tufa surface. The other cyanobacteria lineages were also present in the core layers (CL) below the biofilm, but in rather low frequencies (5% or less), except for the lineages ‘Unidentified C’ (of which still 25% of the clones were found in CL 1–3 and 5) and Pseudanabaena sp. The latter even seemed to have a preference for layers below the biofilm, because 80% of the Pseudanabaena sp. clones were from CL 2 and 5; only a single clone was found in the biofilm. A Pseudanabaena morphotype has also been reported for the lower parts of the tufa biofilms at the DB (Arp et al. 2001b). Also ‘Unidentified F’ was retrieved only in a deeper layer, CL 5, but not found in the biofilm at the WB5 site. From ‘Unidentified A’, a single clone was even recovered from the lowest layer CL 6. Likely, these two unidentified cyanobacteria may exhibit an endo- or chasmolithic life style. Identifications of the environmental clones at the generic level could be inferred only for three cyanobacteria lineages, i.e. where close relationships with reference strains were resolved in the 16S rDNA phylogeny which corresponded to high phenetic sequence similarities. A large fraction (29%) of the cyanobacterial clones from the WB5 site had a very high phenetic sequence similarity of 99.9% with Tychonema bourrellyi Anagnostidis & Koma´rek strain CCAP 1459/11B which is the authentic strain of the type (Suda et al. 2002). The full sequence obtained for one of these clones was together with T. bourrellyi CCAP 1459/11B in a well supported clade (Fig. 7). Therefore, the clones were identified as members of Tychonema Anagnostidis & Koma´rek. Members of the ‘Unidentified A’ lineage, however, were more distantly related to strain T. bourrellyi CCAP 1459/11B in the phylogeny corresponding to the 97.9 –98.2% phenetic sequence similarity level and thus were not assigned to Tychonema. The majority of clones sampled from the WB5 site (41%) formed a lineage which could not be identified (‘Unidentified B’ in Fig. 7). The lineage had a sister-group relationship with the Tychonema/P. autumnale/Microcoleus clade in the 16S rDNA phylogeny (Fig. 7). Environmental clones identified as Chamaesiphon and Pseudanabaena formed well supported clades with corresponding reference sequences (Fig. 5). They had phenetic sequence similarities of at least 98.4% with C. subglobosus (Rostafinski) Lemmermann PCC 7430 and 96.8% with Pseudanabaena sp. PCC 6802, respectively. In contrast, for the four lineages, ‘Unidentified C’, ‘Unidentified D’, ‘Unidentified E’ and ‘Unidentified F’, no closest relatives were available and phenetic sequence similarities with the next named sequences were below 96.5%. However, an uncultured unnamed cyanobacterium (clone H-B02, sequence accession
95
number DQ181686; Taton et al. 2006) from Antarctic lakes was the next closest relative with lineage ‘unidentified C’ (98.4% phenetic sequence similarity) and an uncultured cyanobacterium (sequence accession number EF111085) from the Bogota River, Columbia, with lineage ‘unidentified F’ (98.1% similarity). Neither the morphology nor the taxonomic identity of these operational taxonomic units (OTUs) is known yet. The two lineages, ‘Unidentified D’ and ‘Unidentified E’, did not even group with any close relative in the phylogeny; they were different at the 90 –95% phenetic sequence similarity level to the next neighbouring available cyanobacteria sequences. Cyanobacterial rDNA lineages and corresponding morphotypes. Though the discussion of the observed morphotypes in the light of the molecular results is tempting, it remains still speculative at present. Sequence comparisons of the cultured biofilm cyanobacteria with their counterparts from the environmental clone libraries will be essential to link morphotypes and rDNA signatures. However, the same cyanobacterium may exhibit morphological features in the resin-embedded hardpart sections that are absent or different from those in culture. For ‘Unidentified A’, which was recovered from biofilms at almost all sites of both karstwater streams, a cultured strain (WBK15) exhibited features identical to those of the Lyngbya sp. morphotype (Fig. 6). Therefore, the Lyngbya sp. morphotype likely belongs to lineage ‘Unidentified A’ (Fig. 7). However, the same morphological features also match the description of Tychonema tenue (Skuja) Anagostidis & Koma´rek (formerly Oscillatoria bornetii f. tenuis Skuja) as given in Koma´rek & Anagnostidis (2005). A large fraction of the WB5 environmental clones were very close relatives with T. tenue strain SAG 4.82 and T. bourrellyi strain CCAP 1459/11B (Fig. 7). Interestingly, species of Tychonema are not yet known from epilithic habitats such as tufa surfaces. T. bourrellyi has been described from phytoplankton (Anagnostidis & Koma´rek 1988; Suda et al. 2002) and T. tenue is a major component of phytoplankton of the Norwegian fjord lake Mjøsa (World Lakes Database of the International Lake Environment Committee, www.ilec.or.jp/ database/database.html). From these findings it is concluded that the same morphotype may even represent genetically distant cyanobacteria and the closest relatives of tufa biofilm cyanobacteria may be from quite different habitats. Furthermore, a clear distinction of species and genera within the Tychonema/P. autumnale/Microcoleus clade as based on currently used morphological criteria appears impossible, for example, the distinction of Phormidium autumnale, T. bourrellyi and Microcoleus spp. as discussed in Palinska & Marquardt
96
Fig. 7.
G. ARP ET AL.
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
(2008). The genera Phormidium, Microcoleus as well as Leptolyngbya as presently circumscribed are polyphyletic in the 16S rDNA phylogenies (Fig. 7); they are in urgent need of taxonomic revision. Oscillatoriales (sensu Anagnostidis & Koma´rek 1988) that are associated with tufa formation in freshwaters are often described as Phormidium incrustatum, P. calcareum and P. foveolarum (e.g. Freytet & Verrecchia 1998; Arp et al. 2001b; Pentecost 2003). No cultures and no sequences are yet available as references for P. incrustatum and P. calcareum to clarify their distinction. However, the P. incrustatum/calcareum morphotype has been found very frequently in both karstwater streams by microscopy (e.g. Fig. 5). It is possibly represented by the ‘Unidentified B’ lineage in the environmental rDNA clone libraries (Fig. 7) because here the latter lineage was most frequently recovered. Leptolyngbya foveolarum, although frequently observed by microscopy in samples from both hard water streams, may not have been recovered in our molecular analyses yet. None of the analysed cyanobacterial clones appeared to be closely related to the only available reference strain of L. foveolarum (strain Koma´rek 1964/112; Fig. 7). Our molecular analyses also failed to recover cyanobacteria that represent the Aphanothece and Hyella morphotypes (Fig. 5), i.e. no sequences from members of Pleurocapsales/Chroococcales or Chroococcidiopsis were retrieved at the WB5 site. The Chamaesiphon morphotype, easy to detect by its characteristic polar growth with formation of exocytes, has been detected in resin-embedded biofilm thin sections from the Deinschwanger Bach and Erasbacher rivulet, where it formed a characteristic biofilm constituent on spar-cemented moss surfaces. The molecular analyses revealed Chamaesiphon in the biofilm at the WB5 site, and it was also observed in raw cultures developed from tufa biofilm material. However, this morphotype could not be detected in resin-embedded biofilm sections from WB5. This might indicate that Chamaesiphon may not always exhibit its typical growth form in the biofilms and further underlines the importance of a molecular investigation of cyanobacterial diversity of the tufa biofilms.
97
Diatoms. The high abundance and microscopic diversity of diatoms in biofilms associated with tufa formation has already been recognized more than hundred years ago by Reichelt (1899) and has been reported continuously since then (von Pia 1934; Gru¨ninger 1965; Rott 1994; Reichardt 1995; Freytet & Verrecchia 1998). A large variety of diatoms has been reported from European tufa biofilms, i.e. 29 genera with 169 species have been distinguished by microscopic features of their frustules (Rott 1994; Reichardt 1995; Freytet & Verrecchia 1998). Several diatom genera (e.g. Amphora Ehrenberg ex Ku¨tzing, Gomphonema Ehrenberg, Nitzschia Bory) are conspicuous in karstwater stream biofilms as they exhibit calcite precipitation around their stalks (e.g. Gomphonema olivaceum var. calcareum (Cleve) Cleve-Euler; Winsborough & Golubic´ 1987). In an ongoing study the diatom diversity has been assessed by a combination of both a culture and a culture-independent approach. At the site WB5 thirteen different lineages of raphid pennate diatoms were detected (Fig. 8); they may represent eight different genera and most likely 13 species (Table 3). Biofilms at only two other tufa sampling sites of both karstwater creeks have been found to harbour a larger diatom diversity of their biofilms, the range of diatom lineages per site has been from 1 to 16 lineages. Taken all sites of both karstwater streams together, a total of 17 different lineages which occurred at two and more sites have been recovered. They belong to a large monophyletic clade representing the raphid pennate diatoms in the 18S rDNA phylogenies, but also three species of araphid pennate diatoms are found (Brinkmann et al. 2007a, b). At phenetic sequence similarities of 99% and higher, diatoms were identified at the species level (Table 3). Each diatom lineage was represented one to five times in the 20 sequenced clones which indicated a low coverage, i.e. that the detected diversity was still below the saturation level. Four diatom species were also recovered by cultures (Table 3); examples are shown in Figure 9. Most of the WB5 diatoms were also recovered from other sites of both karstwater streams; only three species were found exclusively at the
Fig. 7. Maximum likelihood phylogenetic analysis of cyanobacteria from tufa biofilms and stromatolite laminae of the Westerho¨fer Bach site WB5 (shown in bold) using 78 complete 16S rDNA sequences (1401 bp long; 667 variable/ 466 informative sites). For cyanobacteria recovered from the tufa core, their location at a certain layer and the percentage of clones from the 711 analysed clones of the core sample (BF, biofilm, CL 1-6, core layers 1– 6; see Fig. 10) is given. The tentative allocation of four morphotypes to certain cyanobacterial lineages is indicated. The Tychonema/P. autumnale/Microcoleus clade has been drawn to a larger scale to show the genetic distances more clearly. A circle indicates the various independent lineages to which members of the genera Phormidium, Leptolyngbya and Microcoleus have been assigned. The sequences of Escherichia coli (AE000129), Bacillus subtilis (AJ276351) and Chlorobium tepidum (M58468) were used to root the phylogeny, but have been pruned away from the graphic. For details of cloning, sequencing and phylogenetic tree analysis see Appendix.
98
G. ARP ET AL.
Fig. 8. Maximum likelihood phylogenetic analysis of pennate raphid diatoms from tufa biofilms of the Westerho¨fer Bach site WB5 (shown in bold) using 49 complete 18S rDNA sequences (1700 bp long; 559 variable/351 informative sites). For details of cloning, sequencing and phylogenetic tree analysis see Appendix.
Table 3. Diatom species recovered from tufa stromatolite biofilms at the sampling site WB5, their next closest relatives from the NCBI GenBank sequence data base and their phenetic sequence similarities Diatom species Achnanthidium minutissimum c Achnanthidium sp. Amphora pediculus cf. Amphora* Gomphonema sp. A Gomphonema sp. B Navicula sp. A Navicula sp. B Navicula tripunctata Navicula veneta* Nitzschia palea c Pinnularia sp.* Surirella brebissonii c
Closest relative sequence
Phenetic sequence similarity to closest relative [%]
Achnanthidium minutissimum AM502032 Pauliella toeniata AY485528 Amphora pediculus AM501960 Phaeodactylum tricornutum EF140622 Gomphonema micropus AM501964 Gomphonema productum AM501993 Navicula cryptotenella AM502029 Navicula radiosa AM501972 Navicula tripunctata AM502028 Navicula veneta AM501975 Nitzschia palea AJ867003 Pinnularia rupestris AJ867027 Surirella brebissonii AJ867029
99.1 98.9 99.1 97.8 99.5 96.5 98.8 98.3 99.8 99.8 99.8 98.5 99.4
*species that were recovered only from the WB5 site. c species also studied by cultures.
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
99
Fig. 9. (a) Navicula cf. tripunctata in a squash preparation of the biofilm (scale 20 mm). (b) Scanning electron micrograph of a Nitzschia cf. palea frustule from a culture (scale 4 mm).
WB5 site. One of the WB5 species, Navicula cryptotenella Lange-Bertalot (corresponding to ‘Navicula sp. 5’ of Brinkmann et al. 2007b) was found at all tufa sampling sites of both karstwater streams. Conclusions and perspective. For the cyanobacteria, the morphological approach may underestimate the actual diversity of cyanobacteria associated with tufa. Six cyanobacteria morphotypes detected in biofilms of both karstwater streams together were opposed to nine different lineages of the 16S rDNA phylogeny which were retrieved just from a single representative site. Interestingly, the cyanobacteria exhibited a certain spatial distribution with several taxa likely to live endo- or chasmolithically in the tufa layers below the biofilm. The detected diatom diversity is considerable; it is even higher than that of the cyanobacteria. This may indicate that diatoms are involved in the calcification processes as well. Molecular signatures provided through environmental rDNA libraries allow to unambiguously distinguish and detect cyanobacteria and diatoms present in rock tufa biofilms. While 18S rDNA sequence comparisons allowed to identify most diatoms even at the the species level, the majority of cyanobacteria were left without genus and species names because the cyanobacterial diversity of tufa biofilms is still poorly represented by the currently available 16S rDNA sequences. Cultures of the tufa cyanobacteria will be required to allocate species names commonly used for the morphotypes in the literature to certain lineages in the 16S rDNA phylogenies.
Non-phototrophic prokaryotes It is well documented (Amann et al. 1995; Pace 1997) that neither cultured organisms nor the molecular assessment of diversity, either alone or in combination, is able to describe the full diversity of a natural habitat. While most studies today
concentrate exclusively on DNA-based assessments, we attempted to obtain a broader overview by targeting molecular diversity and the diversity of aerobic and heterotrophic bacteria along the rivulet in which the tufa is deposited. Only then it is possible to meaningfully explain the occurrence of the same and other types of bacteria in the biofilm and underlying annual stromatolite laminae couplets, here investigated using a tufa core sample (Fig. 10). The majority of the 960 rivulet organisms belong to the genus Flavobacterium (56%), 40% were affiliated to the Gammaproteobacteria (e.g. pseudomonads, aeromonads) and 4% to the Betaproteobacteria (Brambilla et al. 2007; Cousin 2009; Cousin et al. 2008). Several novel species and genera have been found to thrive in the river environment (Cousin et al. 2007; Muurholm et al. 2007; Stackebrandt et al. 2007; Ali et al. 2009). Cultivation approach. 133 aerobic and heterotrophic strains were isolated from the biofilm located on top of the tufa core sample (Fig. 10). The difference in composition between water and biofilm was significant. Of total bacteria isolated the percentage of flavobacteria in the biofilm decreased from 56% in the water to only 22%. Only in a few cases the flavobacteria sequences of the biofilm matched those of the rivulet waters. The majority of isolates (40%) belong to the phylum Firmicutes (mainly Bacillus and Paenibacillus), while the water contained only 1% of these organisms. The percentage of Proteobacteria present in the water and in the biofilm was 40% and 24%, respectively. Actinobacteria, encountered rarely in water samples, were present at 14%. Without going into details of the phylogenetic affiliation of isolates at the level of genera and species, the taxon distribution demonstrates qualitative and quantitative differences between water and biofilm samples, indicating that certain metabolic types were enriched in the biofilm.
100
G. ARP ET AL.
Fig. 10. Tentative affiliation of clone sequences to phyla from the tufa biofilm (BF) and in the 6 layers of the tufa core (CL1– CL6). The phylogenetic affiliation of isolates from the biofilm is included for comparison. Rare phyla: Gemmatimonadetes, Lentisphaerae, Dictyoglomi, Chrysiogenetes, Chlamydiae, Spirochaetes, Thermodesulfobacteria, Aquificae and Thermomicrobia and Candidate phyla OP10, OD1, WS3, SR1, BRC1, TM7.
1008 isolates were recovered from the tufa core layers beneath the biofilm and their phylogenetic affiliation determined (unpublished). Flavobacteria were present in CL 1 (15%) but rarely in the deeper core layers. A second most obvious difference is the presence of Actinobacteria (mainly Arthrobacter spp.) in all core samples. While members of Alphaproteobacteria and Betaproteobacteria are consistently found in small numbers throughout the core layers, members of Gammaproteobacteria are mainly present in the biofilm and the following 4 core layers; in the deepest 3 layers they constitute only a minor part of the population. Members of Firmicutes dominate all layers, especially the deepest 3 layers where they constitute more than 75% of the bacterial communities (Table 4). Molecular approach. In contrast to the rather narrow view of diversity obtained by isolating cultures on R2A medium the isolation of DNA
from the microbial community, subsequent amplification and cloning of the 16S rRNA genes provides a significantly broader assessment of phylotypes, as this approach will also include those organisms that failed to grow on artificial growth media provided. A preliminary phylogenetic analysis revealed membership of clone sequences to more than 25 phyla and candidate phyla, among which the most dominant (more than 100 clone sequences) were members of Cyanobacteria (mainly biofilm and core layer 1; see above), and Proteobacteria. Other phyla, like Acidobacteria, Actinobacteria, Nitrospira, Chloroflexi, and Firmicutes, encompassed between 20 and 100 clones each. Yet other phyla included less than 20 clones and were affiliated to candidate phyla, such as BRC1, TM7, SR1, WS3, OD1 and OP10. About 45 clones from layers 2 through 6 stood isolated in the tree and might represent putative novel candidate phyla.
Table 4. Examples of affiliation of isolates to non-phototrophic prokaryote phyla (in percentage) from rivulet water, biofilm and different tufa core layers from the Westerho¨fer Bach lower section (site WB5). One tufa layer represents one annual laminae couplet Sample Rivulet water Biofilm Tufa core layer 1 Tufa core layer 2 Tufa core layer 3 Tufa core layer 4 Tufa core layer 5 Tufa core layer 6/7
flavobacteria
alphaproteobacteria
betaproteobacteria
gammaproteobacteria
bacilli
actinobacteria
56 22 15 – 1 1 – 1
– 3 6 5 4 13 2 6
4 4 3 1 5 4 – 2
40 17 38 33 38 43 2 0
– 40 24 48 41 25 75 71
– 14 14 13 11 14 21 20
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
Figure 10 gives a tentative overview of the distribution of clone sequences among the biofilm and the core layers. It is obvious that the biofilm is composed mostly of cyanobacterial OTUs (operational taxonomic units), discriminating the presence of other taxa, possibly caused by PCR bias (von Wintzingerode et al. 1997) and by the insufficient number of clones analysed. Indeed, though several hundred clones were analysed from each layer, rarefaction curves (unpublished) indicated that none of the seven clone libraries were sufficiently sampled to reach diversity saturation. While cyanobacterial are still present in large numbers in layer 1, they are insignificant in deeper layers. Proteobacteria emerge as a major taxon in layer 1 and dominate the deeper layers. Other taxa appear in low (2– 12%) but rather constant percentage in layers 3 through 6, i.e. Bacteriodetes, Acidobacteria, Planctomycetes, and Actinobacteria. OTUs of different candidate phyla especially emerge through layers 1 to 6 but were not found in each layer. Molecular assessment and interpretation. It is not surprising that the distribution of sequences of clones and isolates differ significantly, inasmuch the isolation did not attempt to recover phototrophic (e.g. Chloroflexi), autotrophic, lithotrophic and anaerobic organisms. The swarming capacities of many fast growing flavobacteria and colonies with slimy appearance also hindered the growth of slowly growing heterotrophic cultures (e.g. Acidobacteria, Planctomycetes, Verrucomicrobia, Gemmatimonadetes) and the inability of members of candidate phyla to grow on any of the artificial growth media provided is well known. The fact that Firmicutes, e.g. Bacillus and Paenibacillus, are present among the isolates while almost absent in the clone libraries (Fig. 10) can be attributed to the spore stage in which these organisms may rest in the tufa core matrix. Spores are known to resist most of the less harsh DNA isolation techniques, applied to avoid shearing of DNA which would support the formation of chimeric PCR structures. Those sequences clustering with candidate phyla or representing novel candidate phyla will be fully sequenced in order to better analyse the presence of chimeric structure. Comparison with the results of Ng et al. (2006) on non-cyanobacterial taxa from three freshwater tufa environments in Taiwan is hardly possible. In contrast to this study, not individual layers but 3 homogenized core samples were pooled and analysed. For comparison we therefore present the Westerho¨fer rivulet (WB) data as if they were analysed from a pooled sample as well. Though the number of phyla listed by Ng et al. (2006) is much lower, comparison is possible with a few phyla
101
common to both studies. Acidobacteria are major contributors to diversity in all 4 tufa sites, ranging between 7.3 and 12.8% in the Taiwanese samples and 10.4% in the WB sample. The percentage values are also high for Alphaproteobacteria (6.0– 12.9% versus 19.5% for the WB tufa) and Betaproteobacteria (4.5–8.8% versus 9.5%). On the other hand, Gammaproteobacteria constitute a lower part of the WB community (5% versus 9.6– 15.9%), while the presence of Firmicutes of the WB tufa (6.6%) only matches that of the Eternal Spring Shrine tufa sample (6.4%). No Firmicutes were identified in the other two Taiwanese samples. Phylum affiliation of clones between the geographically different freshwater samples may be higher in case the high portion of the unassigned clone sequences (.30%) of the Taiwanese studies will be more thoroughly analysed.
Extracellular polymeric substances As in other mineralizing biofilm systems, such as marine and lacustrine cyanobacterial microbialites (e.g. Trichet & De´farge 1995; De´farge et al. 1996; Arp et al. 1998; Kawaguchi & Decho 2002; Gautret et al. 2004; Decho et al. 2005), extracellular polymeric substances (EPS) are considered to play a crucial role by providing mineral nucleation sites in tufa-forming biofilms of karstwater streams (Pentecost 1985). Microbiologically produced EPS are involved in calcium carbonate precipitation by providing diffusion-limited microenvironments that create alkalinity gradients in response to metabolic processes, and by attracting and binding of calcium ions to negatively charged sites. This would result in an inhibition of precipitation (Kawaguchi & Decho 2002). However, EPS is under constant modification through physico-chemical alteration (e.g. by UV radiation, pH, free radicals) and/or microbial degradation (e.g. hydrolysis, decarboxylation) (for review see Dupraz & Visscher 2005). In the present study, biofilm structure and spatial distribution of bacteria and phototrophic organisms as well as extracellular polymeric substances (EPS) within tufa-forming biofilms were investigated in the fully hydrated state using multi-channel confocal laser scanning microscopy (CLSM). Analysis of tufa-forming biofilms by CLSM revealed that extracellular polymeric substances can be divided into three major structural domains. Lectin-binding analysis firstly allowed the detection of EPS glycoconjugates (i.e. polysaccharides, including those ones covalently linked to proteins and/or lipids) which were clearly associated with phototrophic organisms (Fig. 11a, d, g). Secondly, network-like EPS glycoconjugates were detectable as extended sheet-like structures (Fig. 11b, e, h). Thirdly, glycoconjugates were found as more diffuse and
102
G. ARP ET AL.
Fig. 11. Different structural EPS domains in tufa-forming biofilms of both karstwater streams, Deinschwanger and Westerho¨fer Bach. a, d, g: Cell-associated EPS glycoconjugates, stained with N-acetylglucosamine-specific lectins, originated of filamentous cyanobacteria. b, e, h: Sheet-like EPS glycoconjugates stained with fucose-specific lectin form structures between cells of cyanobacteria and non-phototrophic prokaryotes. c, f, i: Diffuse and cloud-like EPS glycoconjugates stained with amino sugar- and fucose-specific lectins. a–c, f, h– i: Westerho¨fer Bach lower stream site WB3; d –e, g: Deinschwanger Bach northern side tributary site DB 2. Colour allocation: red, lectin-specific EPS glycoconjugates; blue, autofluorescence of chla; pink, autofluorescence of cyanobacteria for d –f; green, bacteria (SYTO9) for g –i; green, calcein-stained calcite. All images were shown as Maximum Intensity Projections of LSM data sets. Scale bar is 20 mm.
cloud-like signals (Fig. 11c, f, i) which covered more or less large areas of tufa-forming biofilms. Most of the lectins suitable for the detection of cyanobacteria-associated EPS glycoconjugates bind also to the cloud-like or sheet-like structures (Fig. 11b, d, g). One possible explanation is that these flat sheets have their origin at the end of filamentous cyanobacterial cells forming extended 3-dimensional structures. Alternatively, the formation of these sheet-like structures may be the result of sloughing and release of EPS produced
by cyanobacteria due to their mobility. Another possibility is that the activity of potential grazers like protozoa or nematodes could have an influence on the spatial arrangement of these extended sheetlike structures. Some lectins bound to cloud-like structures but also to diatom-associated EPS glycoconjugates. Diatoms formed during winter and early spring time a substantial part of the photosynthetic active community, but were quantitatively of lower importance during the rest of the seasonal succession. Stalk-like EPS structures similar to those
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
shown for various diatom cultures by Wustman et al. (1998), were not detected in the tufa-forming biofilms investigated. A combined staining approach of bacterial cells and EPS glycoconjugates revealed that especially the cloud-like EPS glycoconjugates were densely colonized by bacterial cells (Fig. 11f). If bacteria are located outside of such a cloud-like structure it seems likely that the EPS glycoconjugates are degraded by bacteria. Whereas if these EPS glycoconjugates were produced by bacteria, single cells or cell clusters should be located inside the EPS glycoconjugates. In the future the origin of these cloudlike EPS structures will be investigated in more detail in order to clarify their function in terms of mineral formation within tufa-forming biofilms. Fucose and amino sugars (N-acetyl-glucosamine and N-acetyl-galactosamine) are main constituents of the detected EPS glycoconjugates produced by cyanobacteria (¼ cell associated EPS) and sheetlike EPS glycoconjugates, according to Sharon & Lis (2003) and the information about lectin specificity provided by the supplier (Vector Inc., EY Laboratorities, Inc.). Diffuse lectin-specific EPS material not associated with cells suggested sialic acids, amino sugars and galactose as predominating compounds (Zippel et al., submitted). Detection of EPS in calcifying communities by using single or a few fluorochrome-labelled lectins in combination with CLSM has been shown by various authors. For instance, EPS glycoconjugates secreted by Schizothrix ep. were detected after ConA-FITC staining and were associated with carbonate sand grains (e.g. Kawaguchi & Decho 2002). In a study on the architecture and structure of stromatolites from a Mediterranean stream it was shown that an extensive glycoconjugate network (stained with succ-ConA-FITC) was present throughout the carbonated structures (Sabater 2000). Other exopolysaccharides produced by cyanobacteria isolated from Polynesian microbial mats (Kopara) were characterized by chemical analysis (Richert et al. 2005). It was found that both capsular and released EPS consisted of 7–10 different monosaccharides with neutral sugars dominating. For example, fucose, xylose and glucose were present in all EPS fractions. Glucosamine residues were found in the released EPS of strain Rhabdoderma cf. rubrum, and in both fractions of Chroococcus submarinus. In both streams studied, Westerho¨fer and Deinschwanger, the seasonal variability (April, July, and October 2007 and 2008) of EPS glycoconjugates within the tufa-forming biofilms was low, based on a panel of six selected lectins (AAL_Alexa488; HMA_Alexa488; PNA_FITC; LEA_FITC; LcH_FITC; and WGA_FITC). Surprisingly, despite the different hydrochemical
103
conditions in terms of relevant nutrients (P, N) in the stream water, the lectins investigated showed in most cases the same binding pattern in both tufaforming biofilms (Zippel et al., submitted). This implies that, independent of the trophic status of the habitats, the dominant EPS structures consisted more or less of the same glycoconjugates. Nevertheless, the overall occurrence of lectin-specific EPS glycoconjugates was up to 50% higher in the nutrient-poor stream (WB). A detailed analysis of structural EPS composition in combination with competive inhibition assays revealed that at least sialic acid-specific components of diffuse EPS as well as N-acetylglucosamine-specific compounds produced by cyanobacteria were up to 50% higher under nutrient limited condition in Westerho¨fer Bach (Zippel et al., submitted). On the one hand, this could be caused by higher cyanobacterial EPS production under nutrient-limited conditions. A stimulation of carbohydrate synthesis caused by N- or/and P-limitation was also found for example, in unicellular cyanobacteria isolated from hypersaline habitats (de Philippis et al. 1993), in benthic diatoms or intertidal mudflats (Stal 2003). It was highlighted that the increased exudation of EPS was caused by unbalanced growth and represented an overflow metabolism. On the other hand, it might be possible that higher metabolic activity of the heterotrophic community induces an increased degradation of the EPS compounds under nutrient-enriched conditions. A large variety of bacterial morphotypes was detected by laser scanning microscopy as shown in Figure 11d–f. Single cells of filament-, coccoidas well as rod-shaped bacteria colonized either the diffuse EPS matrix or discrete EPS structures like the empty sheaths of filamentous cyanobacteria. Furthermore, different bacterial morphotypes were detected within network-like EPS structures (Fig. 11e) which sometimes covered extended areas of the uppermost biofilm layer. In addition, dense bacterial colonies were observed either as clusters which completely covered the diffuse EPS structures (Fig. 11f) or in the form of connected bands on top of the diffuse EPS matrix (image not shown). It is possible that the diffuse EPS matrix may be produced to some extent by the colonizing bacteria. Nevertheless, bacteria which colonize the surface of diffuse EPS glycoconjugates are likely to be involved in their degradation. A combined application of the lectin-binding approach and calcium carbonate staining with calcein revealed that calcium carbonate crystals were detectable within all defined structural EPS domains (Fig. 11g–i). In some cases, crystals of different size were closely associated with empty sheaths of cyanobacteria (Fig. 11g). In addition, sheet-like and cloud-like EPS structures contained
104
G. ARP ET AL.
or were to some extent completely covered by calcium carbonate crystals (Fig. 11h, i). A clear association of calcium carbonate crystals with diatoms was not detectable during any of the sampling periods.
Conclusions In summary, the lectin-binding analysis in tufa biofilms suggests that: (1) Detected EPS glycoconjugates can be distinguished into three major structural domains. Based on their morphological appearance glycoconjugates closely associated with phototrophic and heterotrophic microorganisms were found. However, at the moment a clear separation of bacterial and diffuse EPS glycoconjugates is not possible via lectin-binding approach; (2) Calcium carbonate crystals were detected in all three structural EPS domains. Crystals nucleated as embedded structures in diffuse or sheet-like EPS glycoconjugates or were associated with EPS glycoconjugates of cyanobacterial origin. It is assumed that heterotrophic degradation of EPS below the phototrophic biofilm part (i.e. major zone of primary production) causes a break-down of the inhibitory effect of EPS, thereby permitting calcite overgrowth on existing calcite crystals (neomorphism) within calcite supersaturated pore waters. In order to proof this hypothesis, pure culture studies under controlled environmental conditions are necessary. In addition, further studies on in situ structure and function of diatom EPS are required to elucidate their assumed inhibitory effect on calcite nucleation in the karstwater streams.
Physicochemical microgradients and mass balance In order to evaluate the effects of biofilm activity on microenvironmental physicochemistry at tufa stromatolite surfaces, pH, O2, Ca2þ and CO22 3 microsensor profiles were measured (Bissett et al. 2008a; Shiraishi et al. 2008a, b). These measurements were conducted in situ at Deinschwanger Bach and Westerho¨fer Bach, and ex situ, in the laboratory, using tufa stromatolite pieces removed from their natural site (Fig. 12). In-situ measurements and laboratory experiments showed similar results.
Microgradients Under illumination, pH and O2 and CO22 con3 centrations increased from the water column to the biofilm surface, while Ca2þ concentration decreased. Consequently, calcite supersaturation microprofiles calculated from Ca2þ and CO22 3 exhibited a strong increase from 8-fold in the water column to 37-fold at the biofilm surface (Fig. 12). Under these conditions, CaCO3 is precipitated, as indicated by radioactive isotope (45Ca2þ) uptake studies (Bissett et al. 2008a). In turn, under darkness, pH, O2 and CO22 3 decreased toward the biofilm surface, while Ca2þ increased mariginally. As a result, calcite supersaturation decreased from 8-fold in the water column to 6-fold at the biofilm surface (Fig. 12). Incubation experiments using 45 Ca2þ indicate that no CaCO3 is precipitated under darkness (Bissett et al. 2008a). Indeed, at a biofilm-free limestone surface used as a control
Fig. 12. pH, O2, Ca2þ and CO22 3 microprofiles at tufa biofilms measured in situ (Deinschwanger Bach, left) and ex situ (Westerho¨fer Bach, middle). As a control, pH and Ca2þ microprofiles measured at a biofilm-free limestone substrate are shown on the right hand side. Open circles indicate light profiles, and closed circles indicate dark profiles. Saturation state (V calcite) calculated from Ca2þ and CO22 3 profiles is also shown. From Bissett et al. (2008a), with permission of the American Society of Limnology and Oceanography, and Shiraishi et al. (2008a), with permission of Taylor & Francis Ltd.
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
sample no pH or Ca2þ microgradients were detected (Fig. 12). Additional microsensor measurements at moss surfaces as well as at spring site endolithic biofilms revealed only microgradients too low to induce CaCO3 precipitation (Shiraishi et al. 2008b). Further experimental studies (Bissett et al. 2008b), manipulating pH and temperature, indicated that biofilms exhibited control over chemical conditions within the microenvironment of the tufa surface and that precipitation continued under photosynthesis over a wide temperature and pH range (4– 17 oC and 7.8–8.9). Further, this work demonstrated that pH was maintained between 7.8 (dark) and 9.5 (light) regardless of the pH of the overlying water. The interpretation of these results is that: (1) under illumination calcite precipitation is driven by the photosynthetic activity of the cyanobacterial biofilm; and (2) in darkness, biofilm respiration decreases the oversaturation so much that calcite precipitation stops. The minor Ca2þ release under dark conditions possibly resulted from exopolymer degradation releasing complexed Ca2þ, as calcite dissolution appears unlikely because the calcite saturation state calculated from the microsensor data remains positive at the biofilm surface even in darkness. In conclusion, tufa stromatolites are formed biogenically by photosynthesis-induced precipitation, not by externally forced permineralization due to physicochemical CO2 degassing responsible for the high ambient calcite supersaturation (Bissett et al. 2008a; Shiraishi et al. 2008a, b). At first glance, this conclusion appears to be in conflict with the results from macroenvironmental hydrochemical analysis, which suggest that physicochemical CO2 degassing is the major factor in driving calcite precipitation in the investigated streams (see previous section on the ‘Hydrochemistry of stream waters’). In order to solve this discrepancy, mass balance calculations were carried out for the Westerho¨fer Bach, based on: (1) annual average flux of photosynthesis-induced CaCO3 deposition (c. 2 1026 mol m22 s21) estimated from the mean annual depositional rate of tufa stromatolite (c. 3900 g m22 year21 obtained by weighing at downstream site WB5); (2) the biofilm surface area within stream (94 m2 recorded by mapping in field); (3) water flow rate at site WB5 (c. 2.0 L s21); and (4) the Ca2þ loss from bulk water during the course of the stream (Shiraishi et al. 2008a, b).
Mass balance The mass balance (see Appendix for calculation) showed that the total Ca2þ loss in the stream is 2.2 104 mol year21, but photosynthesisinduced precipitation via the tufa stromatolite’s biofilms accounts for a Ca2þ loss of only
105
2.7 –4.2 103 mol year21. This means that only about 10 –20% of Ca2þ lost from the bulk stream water is bound via biofilm photosynthesis to form tufa stromatolite calcium carbonate. The remaining 80 –90% Ca2þ lost from the bulk stream waters, in turn, can only be explained by physicochemical precipitation (spar cement) on branches, leaves, tufa debris and as fine-grained calcite precipitated directly in the water column (Shiraishi et al. 2008b). This interpretation may help to resolve the long standing controversy on biogenic v. inorganic origin of tufa stromatolites.
Biofilm structure, calcification pattern and annual lamination Biofilm structure and calcification pattern The structure of tufa-forming biofilms and the corresponding calcification pattern were investigated in detail at the Deinschwanger and Westerho¨fer Bach using formol-fixed, resin-embedded hardpart sections (for methods see Arp et al. 1999) and tape-stabilized cryosections (Shiraishi et al. 2008c). Biofilms of the: (1) non-calcifying spring sites; (2) moss plant surfaces; and (3) tufa stromatolites at downstream stream sites show characteristic compositions: (1) Biofilms on hard substrates at spring sites are characterized by endolithic cyanobacteria, locally overgrown by epilithic coccoid cyanobacteria (e.g. Pleurocapsa minor morphotypes) and filamentous green algae (e.g. Gongrosira sp.). Spring waters show only minor calcite supersaturation (SICc ¼ 0.0 to 0.1) and no calcification was observed in the corresponding biofilms. (2) Farther downstream, moss plants dominate at margins of the stream and at cascades. Once the stream waters have surpassed a calcite supersaturation of at least SICc ¼ 0.8, mosses show initial calcite spar cements on their leaf surfaces. These crystals are commonly idiomorphic rhombs or palisade-like. In depressions of the moss plant surfaces and between the crystals, scattered filamentous cyanobacteria and diatoms occur. Here, endolithic filamentous cyanobacteria (e.g. morphotype Schizothrix perforans) and epilithic cyanobacteria such as Chamaesiphon subglobosus morphotype locally flourish. (3) Central flow paths of the streams in middle and lower stream sections are covered by tufa stromatolite-forming biofilms, except for sections of loose tufa gravel. In tufa stromatolite-forming biofilms, primary producers are cyanobacteria, most of them filamentous, and pennate diatoms (see section Cyanobacteria and diatoms). Both groups of microorganisms concentrate at the top
106
G. ARP ET AL.
Fig. 13. Biofilm structure, calcification pattern and annual lamination. (a) Thin section of a tufa stromatolite showing the annual lamination represented by couplets of porous and dense laminae. Fractures at porous levels are artifacts from sample core drilling. Note phototrophic biofilm on top of the section. Left: Crossed Nicols. Right: Epifluorescence view (ex 450–490 nm, em .520 nm). Site WB5, 23.05.2006. (b) Phototrophic biofilm of the tufa stromatolite surface, with densely arranged filamentous cyanobacteria of the Phormidium foveolarum morphotype (forming a dense calcite lamina) overgrowing less dense calcified cyanobacteria of the Phormidium incrustatum morphotype. Cyanobacterial autofluorescence shown in red. Overlay of laserscanning micrograph (ex 633 nm; em 639–704) and crossed Nicols view. Site WB5, 09.06.2005. (c) Fluorescence in situ Hybridization at tape-stabilized cryosection of the biofilm top, using oligonucleotide probe EUB 338 to demonstrate the presence of Bacteria (shown in green). Cyanobacterial autofluorescence is shown in red, calcite in white. Overlay of laserscanning micrograph after linear unmixing (ex 543 nm; em 544 –704 spectral detection) and transmitted light view. Site WB5, 30.01.2007. (d) Porous layer of a tufa stromatolite, 1.5 mm below biofilm surface, with few remaining cyanobacterial filaments (Phormidium incrustatum morphotype; yellow) and largely empty calcite tubes (green) of former Phormidium filaments. Note abundant organic detritus (red–brown) characteristic for winter laminae of the tufa stromatolites. Epifluorescence micrograph (ex 450– 490 nm, em .520 nm). Site WB5, 12.06.2006.
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
500 mm of the tufa stromatolites, thereby forming a phototrophic biofilm (Fig. 13a, b). These phototrophic microorganisms are associated with numerous non-phototrophic bacteria (Fig. 13c), which occur throughout the biofilm as well as below into several mm of depth. Living cyanobacteria have been observed, scattered in carbonate tubes, in 1.8 mm depth below the biofilm surface (Fig. 13d). Composition and calcification pattern in the investigated tufa stromatolite-forming biofilm change seasonally: Winter–spring biofilms (approximately November –May) are brownish to pale green, with abundant diatoms upon a summer-autumn lamina with scattered cyanobacteria. The diatom colonies remain free of calcite, but commonly show attached detrital quartz and organic debris. During spring times, the diatoms are successively replaced by moderately dense arranged erect Phormidium incrustatum filaments, which locally form bush-like arrays. These filamentous cyanobacteria are enclosed within dominantly microsparitic tubes, with the diatoms remaining within pore spaces of porous microspar layers. Summer–autumn biofilms (approximately June –October) are finally intensively green, with erect P. incrustatum within microcrystalline tubes. These are locally overgrown by a dense microcrystalline layer of closely arranged erect Leptolyngbya aff. foveolarum filaments. Limestone cobbles at the basis of tufa stromatolites, as well as growth interruptions within the tufa stromatolites, commonly show endolithic cyanobacteria of the Hyella fontana morphotype, which also occurs scattered throughout the biofilm. Nonphototrophic prokaryotes occur thoughout biofilm as well as below, i.e. within subfossil tufa stromatolite layers. While a vertical distribution gradient of major groups is evident from mm-scale sampling of single lamina-pairs (see chapter Nonphototrophic Prokaryotes), detailed microscale distribution pattern of phylogenetic groups and single phylotypes within the phototrophic biofilm (top 500 mm) remains to be investigated. First investigations using DAPI staining and CARD-FISH using general probes and tape-stabilized cryosections (Shiraishi et al. 2008c) demonstrate the presence of coccoid, rod-shaped (some in chain-like arrangement) and filamentous Bacteria (Fig. 13c). Their function, e.g. in exopolymer degradation, however remains to be elucidated. Tufa stromatolite surfaces free of biofilms, e.g. due to freezing and chipping off of top laminae in winter, are characterized by further, inorganic crystal growth. This aggrading neomorphism at the absence of biofilms and (less intensive) within deeper, porous tufa crust parts is the likely cause of palisade crystal laminae in the investigated
107
study sites. Internally calcified empty filaments, which locally occur in porous (winter– spring) laminae below the living phototrophic biofilm (e.g. Arp et al. 2001b), may in turn reflect calcite precipitation in empty cells still surrounded by an inhibiting exopolymer matrix.
Annual lamination Seasonal changes in the biofilm community go along with the formation of annual lamination in the tufa stromatolites: A single year of deposition, represented by a porous and a dense lamina, varies between 1.6–5.4 mm thickness at the lower section of the Westerho¨fer Bach (site WB5) and 3.9 –7.6 mm thickness at the lower section of the Deinschwanger northern side stream (site DB2). In both streams, the porous winter-spring layers, which include organic and quartz detritus, are less thick (WB5: 1.0 –1.5 mm; DB2: 1.3–3.6 mm) than the dense summer–autumn layers (WB5: 0.5– 2.0 mm; DB2: 1.4– 4.0 mm). In fact, the transition from porous winter-spring to dense summer– autumn laminae is gradual, while the bases of the porous winter-spring laminae are mechanically instable, i.e. commonly break during sampling (Fig. 13a). In addition, a minute sub-layering within the annual cycles is evident. This seasonal pattern observed in Deinschwanger and Westerho¨fer Bach tufa stromatolites coincides with the stable isotope composition of the carbonate (Fig. 14): Porous winter-spring laminae show less negative d18O values than dense summer-autumn laminae. Based on that, calculated palaeo-water temperatures for the Westerho¨fer Bach are approximately 7 to 9 8C for porous, and 9 to 10.5 8C for dense laminae (Shiraishi et al. 2008b). Seasonal changes in d13C of tufa stromatolite calcite, however, reflect a number of different factors, with lower values in warmer seasons due to stronger supply of 12C from respiration in soils of the catchment area (see e.g. Hori et al. 2008).
Discussion: implications for the fossil record Seasonal lamination and Quaternary palaeoclimate signals Results of investigating calcifying biofilms and tufa stromatolites in karstwater streams are of interest for the interpretation of palaeoclimate signals in subfossil and Quaternary tufa deposits (e.g. Andrews & Brasier 2005). The interpretation of palaeoclimate signals recorded in tufa stromatolites indeed relies on the correct assignment of laminae to specific seasons. However, seasonal lamination pattern described in some tufa stromatolite studies (e.g.
108
G. ARP ET AL.
Fig. 14. Carbonate stable oxygen and carbon isotope depth profiles of the top 7 mm of a tufa stromatolite from the lower section of the Westerho¨fer Bach (site WB4, 19.10.2007), and calculated palaeo-water temperatures. For comparison, measured stream water temperatures of this site for May 2006, August 2006 and January 2007 are indicated as dashed lines. Left hand side shows a cross polar view of the thin section analysed, with dense layers appearing darker (D) and porous layer appearing lighter (P). From Shiraishi et al. (2008b), modified, with permission of Elsevier Ltd.
England, Belgium and France) appear to be ‘reversed’ to that of other locations and studies (e.g. Germany, Japan) (Kano et al. 2003: 259). This may partly reflect different methods applied (palynomorphs: Geurts (1976: 18); insect larval housings: Stirn (1964: 14); stable isotopes: Matsuoka et al. 2001; seasonal sampling: Kano et al. 2003). For example, the initial attempt to use of palynomorphs for this purpose (Geurts 1976) has the potential risk of a time shift between attachment of pollen to the biofilm surface (e.g. in autumn) and their incorporation into tufa laminae by calcification (e.g. in the following spring time). Nonetheless, porous-microsparitic and dense-microcrystalline laminae may form at different seasons in different places, possibly dependent on the kind of vegetation (grassland, coniferous or deciduous woodland) and corresponding shading at the sites in question. For the Westerho¨fer and Deinschwanger Bach, seasonal sampling demonstrated that porous–microsparitic form in winter–spring time and densemicrocrystalline laminae in summer–autumn time (Arp et al. 2001b), substantiated by the stable oxygen isotope record (Shiraishi et al. 2008b). For these both streams, spring discharge and chemical composition reflect (seasonal and sporadic) changes in rain fall in the catchment area. Strongest and most regular seasonal fluctuations, however, have been observed with respect to light irradiance and stream water temperature. Although spot measurements of irradiation show strong fluctuations for
single tufa-forming stream sites in the deciduous woodlands, with values in October (after leaf fall) commonly similar to values in June and July (Fig. 15), the sum of irradiated light as dependent on day length might be one steering variable in the formation of seasonal laminae couplets. However, temperature seems to be of greater significance. Not only does it have an effect on physiological activity of the microorganisms (e.g. the intensity of photosynthesis; Bissett et al. 2008b), but also on diffusion coefficients (specifically Ca2þ): higher temperatures permit higher fluxes of Ca2þ because of increased diffusion rate, consequently higher precipitation rate and formation of more dense, microcrystalline laminae (Shiraishi et al. 2008b). Consequently, lamination in WB and DB tufa stromatolites is considered to reflect seasonal changes in irradiance (photosynthesis, biofilm composition) and temperatures (diffusion coefficients). Similarly, Kano et al. (2003) interpret the formation of dense summer– autumn laminae alternating with porous winter–spring laminae in tufa stromatolites of Shirokawa, Japan, as mainly driven by differences in temperature. In any case, a correct assignment of laminae to specific seasons has to be achieved for each tufa –stromatoliteforming stream by seasonal sampling (e.g. Kano et al. 2003). A second prerequisite for the interpretation of palaeoclimate signals is that the proxies used, here stable oxygen and carbon isotope ratios, reflect the
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
109
Fig. 15. Mean irradiance near tufa biofilm surfaces in German karstwater streams at different seasons during 2006– 2008. Filled circles: Westerho¨fer Bach, open circles: Deinschwanger Bach.
macroenvironmental conditions at the stream sites, not microenvironmental conditions within biofilms partially decoupled from the bulk hydrochemical system. While the tufa stromatolite-forming biofilms do not change oxygen isotope ratios, the carbon isotopic composition of the calcite laminae is potentially strongly affected by photosynthesis, which even drives their initial formation. In addition, heterotrophic CO2 release, as detected during night conditions, potentially alters or blurs carbon isotope values of the tufa stromatolite laminae via solid state diffusion. The stable isotope analysis of stream waters and tufa carbonates at the Westerho¨fer Bach however shows that, despite photosynthesis-induced precipitation within the biofilms, the effect on d13C in tufa stromatolite carbonate is not detectable (Shiraishi et al. 2008b). This observation might be explained by spatially too small gradients (i.e. isotope exchange still fast enough at these small distances) and/or isotope exchange between 12Cdepleted daytime precipitates and 12C-enriched nighttime biofilm water phase. Earlier studies on the stable isotopic composition of tufa carbonate also detected only a minor 12C-depletion, but considered this observation as indicative of a largely inorganically forced precipitation (Spiro & Pentecost 1991). In any case, the results underline the potential of tufa stromatolites for palaeoclimate reconstructions, with d13C reflecting the bulk water carbon system (i.e. influx of 12C-enriched fluids from soil versus photosynthesis effect of the bulk stream system; Rayleigh fractionation). This also implies that 12 C-depletion (if present at all) in fossil stromatolite laminae is rather the result of a photosynthesiseffect on the whole water body, and not of the stromatolite-forming biofilm itself (e.g. in Norian lacustrine stromatolites; Arp et al. 2005).
Continuous and fast growth with low diagenetic overprint of tufa stromatolites is certainly a further prerequisite for accurate palaeoclimate reconstructions at annual scale. In this regard, Westerho¨fer and Deinschwanger Bach tufa stromatolites are disappointing study objects, because the small-scale flow paths commonly change. On the other hand, the petrographic observations at samples from these streams indicate that aggrading neomorphism appears to be more significant in slowly growing tufa stromatolites and at growth interruptions, than in the few continuously and comparatively fast growing tufa stromatolites. Such diagenetically alterated laminae consisting of palisade crystals, formed by crystal growth and recrystallisation in supersaturated water and lack of inhibiting exopolymers, therefore likely contain mixed isotope signals. Lateral groundwater flow along porous laminae may contribute to this effect as well. Fast and continuously growing tufa crust such as that of Japan (Kano et al. 2003, 2004) therefore appear to be particularly suitable for palaeoclimate reconstructions, while the palaeoclimate record of tufa stromatolites at the Westerho¨fer and Deinschwanger Bach rarely exceeds several years.
Microbialite fabrics and dissolved inorganic carbon concentrations in fresh- and seawater In addition to Quaternary palaeoclimate studies, tufa stromatolites and other non-marine microbialites provide insights for the interpretation of marine microbialite fabrics as related to seawater composition, which varied during the Phanerozoic and Precambrian (e.g. Arp et al. 2001a). Microbialite fabrics form dependent on the microbial community involved and the mechanisms of CaCO3 precipitation, both reflecting
110
G. ARP ET AL.
Fig. 16. Numerical simulations demonstrating the effect of photosynthetic CO2 assimilation (black solid line) and heterotrophic exopolymer degradation associated with secondary Ca2þ release (grey dotted line) on calcite saturation, as dependent on the bulk phase dissolve inorganic carbon concentration. For details of calculation and assumptions see Arp et al. (2001a, 2003).
environmental conditions (Riding 2000). Following investigations of biofilm calcification in presentday saline and freshwater settings and model calculations using a given CO2 removal and an initial 10-fold calcite supersaturation (equivalent to a 7-fold aragonite supersaturation), it has been demonstrated that cyanobacterial photosynthesis causes calcareous tubular microfossils to form only in Ca2þ-rich and comparatively poorly pHbuffered settings, (Fig. 16; Arp et al. 2001a). This relation reflects the solubility product and pH buffering by the dissolved inorganic carbon pool. Consequently, calcareous cyanobacterial microfossils can be used to trace secular changes in seawater Ca2þ throughout the Phanerozoic when taking into account palaeo-partial pressure curves for carbon dioxide (Arp et al. 2001a). In turn, the enigmatic lack of calcified cyanobacteria in stromatolitebearing Precambrian sequences can be explained as a result of high dissolved inorganic carbon concentrations (Arp et al. 2001a), alternatively to other hypothesis (e.g. Riding 1982, 2006; Knoll et al. 1993). The results from studying tufa-forming biofilms in karstwater streams support this view. Although it was initially assumed that cyanobacterial calcite tubes in tufa stromatolites reflect passive encrustation (inorganic mineralization) (e.g. Riding 1991: 32; Arp et al. 2001a: p. 1702), the present microsensor data indicate that these tubes are directly formed
as a result of cyanobacterial photosynthetic activity at comparatively low dissolved inorganic carbon concentrations and low pH buffering. On the other hand, the enormously diverse and abundant non-phototrophic microbial community apparently does not promote calcium carbonate precipitation in tufa-forming biofilms of karstwater streams. Although detailed pathways of exopolymer as well as low-molecular-weight substrate consumption remain to be investigated, heterotrophic activity of prokaryotes rather promotes the maintenance of porosity in tufa biofilms, and possibly break-down of inhibiting exopolymers at tufa stromatolites discontinuities. Here, overgrowth of existing calcite crystals (neomorphism) in observed. Strikingly, the non-phototrophic prokaryotes detected in the investigated tufa biofilms are almost exclusively aerobic. Indeed, model calculations suggest that the pH decreasing effect of CO2 generation by far surpasses the effect of ion activity product increase by secondary Ca2þ release from degraded exopolymers (Fig. 16). This is due to the relatively low pH buffering in tufa stream waters when compared to soda lake waters. In such lakes, exopolymer degradation and secondary Ca2þ release effectively causes CaCO3 precipitation because the simultaneously released CO2 is buffered by the high dissolved inorganic carbon concentrations. The corresponding microbialite fabrics, especially sickle-cell like shrinkage voids or net-like calcified EPS structures, therefore
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS
appear to be characteristic of settings high in dissolved inorganic carbon (Arp et al. 1998, 2003), and may be traceable in the fossil marine record as well (e.g. Archaean ‘cuspate microbialites’ with vertical supports and draping laminae; Sumner 1997). The transfer suggested above requires annotations because marine and non-marine (particularly freshwater) ecosystems are commonly regarded as almost unrelated: (1) Hydrochemical principles such as ion activity products defining mineral saturations and pH buffering by weak acids and their conjugate bases are valid in all aqueous settings. The crucial point for a transfer from non-marine to marine settings is that different concentrations of chemical species have to be included in model calculations by using computer programs accounting for ionic strength, ion pairing, and pH-dependent dissociations (e.g. PHREEQC; Parkhurst & Appelo 1999). (2) Seawater composition changed significantly during the Precambrian as well as during the Phanerozoic (Hardie 1996; Lowenstein et al. 2001; Habicht et al. 2002; Berner 2004) and present-day seawater calcite and aragonite supersaturation is lower than during most periods of the Phanerozoic (e.g. Riding & Liang 2005). Consequently, various non-marine calcifying biofilm systems may be more suitable to elucidate factors and mechanisms in the formation of fine-grained and skeletal fossil marine stromatolites than the coarse-agglutinated stromatolites of the present-day marine. (3) While the detected prokaryote species (cyanobacteria and – as far as cultivated – bacilli, arthrobacteria, pseudomonads, flavobacteria) in the karstwater streams find their closest relatives in other freshwater or soil habitats (e.g. Claus & Berkeley 1986; Jones & Keddie 1992; Taton et al. 2003; Bernardet & Nakagawa 2006), their function in biogeochemical cycles should nonetheless be the same in the marine and non-marine (e.g. oxygenic photosynthesis by cyanobacteria, degradation of high-molecular-weight organic compounds by flavobacteria). Major differences exist with respect the potential effect of bacterial sulphate-reduction between freshwater and marine stromatolite formation. Sulphate-reducing bacteria have to date not been detected in significant numbers in the investigated karstwater stream biofilms – which may either reflect methological failure or the combination of high O2 and low SO22 4 . However, the effect of sulphate-reduction on CaCO3 precipitation in marine settings is controversially discussed (Bosak & Newman 2003; Aloisi 2008) and may rather relate to removal of kinetic inhithan to increasing CO22 bitor SO22 4 3 . Indeed, future studies combining results from sulphaterich and sulphate-poor microbialite-forming settings will likely provide substancial insights in
111
formation mechanisms and palaeoenvironment of fossil microbialites, rather than focussing on single settings alone.
General conclusions (1) Independent from the initial Ca2þ/alkalinity ratio, loss of CO2 goes hand in hand with an increase in calcite supersaturation, approaching saturation index maximum values of c. 1.0, not surpassed due to concomitantly increasing calcite precipitation. Stream waters with higher concentrations of Mg2þ and SO22 4 reach higher calcite saturation values than stream waters with low Mg2þ and concentrations. Physicochemical precipiSO22 4 tation seems to be most effective in the Erasbach are rivulet, where inhibiting Mg2þ and SO22 4 lowest and a self-build tufa-canal largely composed of spar-cemented moss plants forms. Generally, there is no detectable impact of biofilm photosynthesis on bulk water chemical composition, neither seasonal nor diurnal. (2) For the cyanobacteria, the most dominant photoautotrophic organisms in the biofilms on tufa surfaces, a culture-independent molecular approach showed that microscopy of resin-embedded biofilm thin sections underestimated the actual diversity of cyanobacteria, i.e. the six cyanobacteria morphotypes were opposed to nine different lineages of the 16S rDNA phylogeny. The same morphotype may even represent two genetically distant cyanobacteria and the closest relatives of tufa biofilm cyanobacteria may be from quite different habitats. A drilling core 5 cm deep into the tufa rock revealed that the vast majority of cyanobacteria was in fact located within the biofilm on the tufa surface, but a few cyanobacteria (obviously with an endo- or chasmolithically life style) were recovered even from deeper layers below the biofilm. The diversity of biofilm diatoms was even higher than that of the cyanobacteria at the studied exemplar site, i.e. 13 diatom species opposed to 9 cyanobacterial lineages. (3) Based on cultivation, the non-phototrophic prokaryotic community of the stream waters, which is largely derived from adjacent soils, significantly differs from that of the tufa-forming biofilms. Based on clone numbers from tufa stromatolite laminae below the Cyanobacteria-dominated tufasurface biofilm, an enormous variety of taxa were identified which are likely to represent a highly diverse metabolic and morphological spectrum of bacteria. However, only a few strictly anaerobic prokaryotes were detected. (4) Lectin-binding analysis indicate the presence of at least three structural EPS glycoconjugate domains: cell (cyanobacteria/diatom/bacteria)associated, as well as sheet-like and cloud-like EPS domains. Seasonal and spatial variability of these
112
G. ARP ET AL.
structural EPS domains was low in the investigated streams, and calcium carbonate crystals were detected in all of them. However, a clear association of calcium carbonate crystals with diatoms or their stalks was not detectable during any of the sampling periods. (5) Microsensor measurements demonstrate that tufa biofilm calcification is, contrary to previous assumptions, controlled by the photosynthetic activity of the biofilms. Cyanobacterial tubes in karstwater streams therefore reflect photosynthesis-induced precipitation, not passive impregnation due to high external supersaturation as a result of physicochemical CO2 degassing. Mass balance calculations, however, suggest that biofilm photosynthesis is responsible for only 10 to 20% of Ca2þ loss in the stream, while remaining Ca2þ loss may derive from physicochemical precipitation on branches, leaves and fine-grained calcite particles. (6) Annual laminae couplets of tufa stromatolites in the investigated karstwater streams reflect seasonal changes, mainly driven by temperature and irradiation: Porous microspar layers formed by winter–spring biofilms with abundant diatoms and scattered cyanobacteria alternated with dense microcrystalline calcite layers formed by summer– autumn biofilms with erect cyanobacterial filaments. (7) Photosynthesis-induced microgradients in tufa-forming biofilms do not cause 12C-depletion in the precipitated carbonate. Consequently, stable carbon isotope values of tufa stromatolite carbonate reflect that of water column, underlining the potential of tufa stromatolites for palaeoclimate reconstructions. (8) Slowly or discontinuously growing tufa stromatolites are subject to early neomorphism, i.e. palisade crystal formation, within the calcite supersaturated stream and laterally moving pore waters, once biofilm exopolymeric matrix is removed. Therefore, only fast and continuously growing tufa crust appear to be suitable for palaeoclimate reconstructions. (9) While photosynthesis is the major mechanism in cyanobacterial calcification and tufa stromatolite formation within the karstwater streams, heterotrophic activity including exopolymer-degradation and secondary Ca2þ-release rather decreases calcite saturation, contrary to settings high in dissolved inorganic carbon such as soda lakes. Consequently, tufa stromatolites show cyanobacteria-related calcification fabrics, while soda lake microbialites show fabrics related to exopolymer degradation, possibly analogous to marine counterparts. The study is part of the research unit ‘Geobiology of Biofilms’ (DFG - FOR 571; publication #47), funded by the German Research Foundation, with the subprojects
‘Biodiversity and DNA-Taxonomy of Algae and Cyanobacteria in calcifying Biofilms’(Fr 905/13, Ar 335/6), ‘Control of Mineralization Processes by heterotrophic and autotrophic Prokaryotes in high-PCO2-Biofilms of tufa systems’(Ar 335/5, Sta 184/19), ‘In situ Structure and Function of Biofilm-Systems during Mineralization of Carbonates’ (Ne 904/2), and ‘Influence of Temperature, Oxygen Concentration and Concentration of organic Substrates on marine microbial Carbonate Dissolution’ (Be 2167/7). NB, KIM and TF thank Gabriele Schauermann and Ajoze Marrero-Calı´co for their skilful assistance in the sequencing work, Jessica Ramm for provision of yet unpublished results, as well as Regine Jahn, FU Berlin, for SEM support and training of NB in diatom identification. The work of TF was also supported by the German Federal Ministry of Education and Research, BMBF (AlgaTerra project, grant 01 LC 0026) within the BIOLOG program. SC and ES thank Evelyne Brambilla, DSMZ, for the skillful isolation of isolates from water and tufa. We thank Christine Heim for providing the field image of the Erasbach rivulet. Two anonymous reviewer provided detailed and helpful comments and suggestions.
Appendix: methods Hydrochemistry: Water samples for titration of total alkalinity were collected in Schott glass bottles, and for determination of main anions and cations (Ca2þ, Mg2þ, Naþ, Kþ) in pre-cleaned PE-bottles. Samples for cation analysis were filtered in the field through 0.8 mm membrane filters (Millipore) and fixed by acidification. Samples were stored cool and dark until laboratory measurements. Temperature, electrical conductivity, pH, and redox potential of water samples were recorded in-situ using a portable pH meter (WTW GmbH) equipped with a Schott pH-electrode calibrated against standard buffers (pH 7.010 and 10.010; HANNA instruments), and a portable conductivity meter (WTW GmbH). Dissolved oxygen was analysed titrimetrically following the Winkler method. Total alkalinity was determined by acid-base titration immediately after sampling using a hand-held titrator and 1.6 N H2SO4 cartridges as titrant (Hach Corporation). Main cations (Ca2þ, Mg2þ, Naþ and Kþ) were analysed either by flame AAS (atomic absorption spectroscopy, Philips-Unicam) in case of Reinsgraben and Erasbach, or by ion chromatography with suppressed conductivity detection (Dionex Corporation) in case of Westerho¨fer Bach and Deinschwanger Bach. ICP-OES (Perkin Elmer) was used to determine Sr2þ and Ba2þ. Anion concentrations (Cl2, SO22 and NO2 4 3 ) were measured either by ion chromatography with indirect photometric detection (Waters) in case of Reinsgraben and Erasbach, or by ion chromatography with suppressed conductivity detection (Dionex Corporation) in case of Westerho¨fer Bach and Deinschwanger Bach. Dissolved phosphate and dissolved silica concentrations were measured by spectrophotometric methods (Unicam). Measured values were processed with the computer program PHREEQC (Parkhurst & Appelo 1999) in order
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS to calculate ion activities and PCO2 of the water samples as well as saturation state with respect to calcite. Hydrochemical model calculations have been carried out using the same program. Contrary to an incorrect statement in Altermann et al. (2006: 156), the model calculations take into account the formation of ion pairs and the pH 22 control of HCO2 ratios. Total carbon, nitrogen 3 : CO3 and sulfur determination on tufa carbonate samples was conducted with a EuroEA CNS-analyser (Hekatech), and organic and inorganic carbon content was analysed with a Multiphase Carbon Determinator LECO RC– 412. SSU rDNA cloning and sequence analyses of diatoms and cyanobacteria: In brief, SSU rDNAs were amplified using PCR primers that preferentially bind to rDNAs of cyanobacteria (Wilmotte et al. 1993) or diatoms (N. Brinkmann, unpubl.) from DNA directly extracted of biofilm samples, cloned and sequenced. Phylogenetic trees were generated using all cyanobacterial 16S and diatom (raphid and araphid pennate) 18S rDNA sequences as currently (May 2009) available from public databases within the ARB database programme (Ludwig et al. 2004; www.arb-home.de). To this the newly determined sequences were added using the parsimony algorithm tool in ARB. An alignment of full length sequences extracted from ARB was then subjected to phylogenetic analyses using the Maximum Likelihood method. The phylogenies were obtained with the programme Treefinder (Jobb 2008). Confidence values for the obtained groups (edge support) were inferred from expected-likelihood weights (Strimmer & Rambaut 2002) applied to local rearrangements of tree topology as provided in Treefinder. Only values at or above 87% were recorded and support at internal branch lengths (edges) is indicated by a filled circle. Non-phototrophic prokaryotes: In order to determine whether the composition of a tufa biofilm differs from that of the rivulet water to which it is exposed, a core sample was drilled from tufa deposited at a downstream site (23.05.2006). The core, 4 cm Ø and 5 cm deep, was obtained using a modified Stihl motorsaw equiped with a coring device. The core was immediately rinsed with sterile 0.7% NaCl solution, transported to the laboratory on ice, and then frozen at 280 8C. The tufa core sample was separated into several layers (CL, core layers) of c. 4–5 mm thickness, each of them corresponding to an annual laminae couple of the tufa stromatolite; thus the core represents the CaCO3 deposition of about 10 years. The assessment of the taxonomic status of isolates from the biofilm and tufa layers followed isolation on R2A medium. Partial 16S rRNA gene sequence analysis (c. 500 bp) were analysed by BLAST (Altschul et al. 1997) which allows a rough affiliation of sequences to homologous sequences of described taxa which are deposited in public databases. For the identification and biodiversity assessment of uncultivated prokaryotes, DNA was isolated from the biofilm and from several 4– 5 mm layers of the tufa. Following sequence analysis and exclusion of chimeric sequences (RDP) 2.351 partial sequences were
113
analysed by the RDP Classifier program (http://rdp.cme. msu.edu/classifier/classifier.jsp). Laser-scanning microscopy and lectin-binding: Due to the occurrence of different pigments in cyanobacteria (chlorophyll a, phycocyanin and phycoerythrin) and algae (chlorophyll a), the specific autofluorescence of various phototrophic organisms was recorded in two different channels. Signals from phycocyanin and phycoerythrin were excited at 561 nm and detected in the range of 585 –625 nm. Autofluorescence of chlorophyll a (chla) was detected in a second channel (excitation 633 nm; emission 650-800 nm) (Neu et al. 2004). Bacterial cell distribution and biomass were determined after staining with nucleic acid-specific fluorochrome SYTO 9 (Molecular Probes, final concentration 1 mg ml21). Glycoconjugate distribution within biofilms was measured by using fluorescence lectin-binding analysis (FLBA) (Neu et al. 2001). A screening with 76 commercially available lectins revealed that approximately 30 lectins with different specificities were suitable for the detection of EPS within tufa-forming biofilms of both creeks investigated. Detailed information of methodological procedures and results are given in Zippel et al. (submitted). Calcein (Sigma) was used for the detection of calcium carbonate present in tufa-forming biofilms. Tufa pieces were stained with the calcein solution (10 mg l21) for 12 h at room temperature (Moran 2000). Methods for embedding and sectioning of tufa biofilm samples have been described in Arp et al. (2001b, 2003) and Shiraishi et al. (2008c). For microelectrode measurements and flux calculations see Bissett et al. (2008a, b) and Shiraishi et al. (2008a, b). Mass balance calculations of Ca2þ loss via photosynthetic biofilms versus Ca2þ loss from bulk water system have been calculated as follows: (1) The annual Ca2þ loss via biofilms [mol year21] is the annual average flux of PS-induced CaCO3 deposition (1.85–2.86 1026 mol m22 s21), multiplied by the biofilm surface area recorded by field mapping (94 m2) and time (60 60 12 365 seconds per year): 2.7– 4.2 103 mol year21; (2) The total annual Ca2þ loss from creek waters [mol year21] is the decrease in Ca2þ concentration during the course of the stream (0.35 mmol L21), multiplied by the water flow rate at site WB 5 (c. 2.0 L s21) and time (60 60 24 365 seconds per year): 2.2 104 mol year21. As a result, the annual Ca2þ loss via biofilms is only 12–19% of the total annual Ca2þ loss from stream waters.
References A LI , Z., C OUSIN , S., F RU¨ HLING , A., S CHUMANN , P., Y ANG , J. & S TACKEBRANDT , E. 2009. Description of Flavobacterium rivuli, sp. nov., Flavobacterium subsaxonicum sp. nov., Flavobacterium swingsii sp. nov., and Flavobacterium reichenbachii sp. nov., isolated from a hard water rivulet. International Journal of Systematic and Evolutionary Microbiology, 59, 2610– 2617.
114
G. ARP ET AL.
A LOISI , G. 2008. The calcium carbonate saturation state in cyanobacterial mats throughout Earth’s history. Geochimica et Cosmochimica Acta, 72, 6037–6060. A LTERMANN , W., K AZMIERCZAK , J., O REN , A. & W RIGHT , D. T. 2006. Cyanobacterial calcification and its rock-building potential during 3.5 billion years of Earth history. Geobiology, 4, 147– 166. A LTSCHUL , S. F., M ADDEN , T. L., S CHAFFER , A. A., Z HANG , J., Z HANG , Z., M ILLER , W. & L IPMAN , D. J. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research, 25, 3389– 3402. A MANN , R. I., L UDWIG , W. & S CHLEIFER , K. H. 1995. Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiological Reviews, 59, 143–169. A NAGNOSTIDIS , K. & K OMA´ REK , J. 1988. Modern approach to the classification system of cyanophytes. 3. Oscillatoriales. Archiv fu¨r Hydrobiologie Supplementband, 80 (Algological Studies, 50– 53), 327–472. A NDREWS , J. E. & B RASIER , A. T. 2005. Seasonal records of climatic change in annually laminated tufas: short review and future prospects. Journal of Quaternary Science, 20, 411–421. A NDREWS , J. E., P EDLEY , H. M. & D ENNIS , P. F. 2000. Palaeoenvironmental records in Holocene Spanish tufas: a stable isotope approach in search of reliable climatic archives. Sedimentology, 47, 961–978. A RP , G., 2008. Fossil and present-day stromatolites of southern Germany. In: R EITNER , J., Q UE´ RIC , N.-V. & R EICH , M. (eds) Geobiology of Stromatolites. International Kalkowsky-Symposium October 4– 11, 2008, Go¨ttingen, Excursion B: 168–202. A RP , G., H OFMANN , J. & R EITNER , J. 1998. Microbial Fabric Formation in Spring Mounds (‘Microbialites’) of Alkaline Salt Lakes in the Badain Jaran Sand Sea, PR China. Palaios, 13, 581–592. A RP , G., T HIEL , V., R EIMER , A., M ICHAELIS , W. & R EITNER , J. 1999. Biofilm exopolymers control microbialite formation at thermal springs discharging into the alkaline Pyramid Lake, Nevada, USA. Sedimentary Geology, 126, 159–176. A RP , G., R EIMER , A. & R EITNER , J. 2001a. Photosynthesis-induced biofilm calcification and calcium concentrations in Phanerozoic oceans. Science, 292, 1701– 1704. A RP , G., W EDEMEYER , N. & R EITNER , J. 2001b. Fluvial tufa formation in the hardwater creek Deinschwanger Bach (Franconian Alb, Germany). Facies, 44, 1 –22. A RP , G., R EIMER , A. & R EITNER , J. 2003. Microbialite formation in seawater of increased alkalinity, Satonda Crater Lake, Indonesia. Journal of Sedimentary Research, 73, 105 –127. A RP , G., B IELERT , F., H OFFMANN , V.-E. & L O¨ FFLER , T. 2005. Palaeoenvironmental significance of lacustrine stromatolites of the Arnstadt Formation (‘Steinmergelkeuper’, Upper Triassic, N-Germany). Facies, 51, 419– 441. B AIER , A. 2002. Die ‘Steinerne Rinne’ am Berg su¨dlich Erasbach/Opf. - eine Untersuchung zur Hydrologie und -chemie des Seichten Karstes. Geologische Bla¨tter fu¨r Nordost-Bayern, 52, 139– 194. B AIER , J. J. 1708. Oryktographia Norica sive rervm fossilivm et ad minerale regnvm pertinentivm. Impensis
Wolfgangi Michahellis, Norimbergæ. [VON F REYBERG , B. 1958. Johann Jacob Baiers ‘Oryktographia ¨ bersetung Norica’ nebst Supplementen. In der U aus dem Lateinischen von Herrmann Hornung und unter pala¨ontologischer Mitarbeit von Florian Heller. Erlanger geologische Abhandlungen, 29, 1– 133. Erlangen.] B ERNARDET , J. F. & N AKAGAWA , Y. 2006. An Introduction to the Family Flavobacteriaceae. In: F ALKOW , S., R OSENBERG , E., S CHLEIFER , K. H., S TACKEBRANDT , E. & D WORKIN , M. (eds) The Prokaryotes. Vol. 7, 3rd edn. Springer Verlag, New York, 455– 480. B ERNER , R. A. 2004. A model for calcium, magnesium and sulfate in seawater over Phanerozoic time. American Journal of Science, 304, 438–453. B ISSETT , A., DE B EER , D., S CHOON , R., S HIRAISHI , F., R EIMER , A. & A RP , G. 2008a. Microbial mediation of stromatolite formation in karst-water creeks. Limnology and Oceanography, 53, 1159– 1168. B ISSETT , A., R EIMER , A., DE B EER , D., S HIRAISHI , F. & A RP , G. 2008b. Metabolic microenvironmental control by photosynthetic biofilms under changing macroenvironmental temperature and pH conditions. Applied and Environmental Microbiology, 74, 6306– 6312. B OROWITZKA , M. A. 1982. Morphological and cytological aspects of algal calcification. International Review of Cytology, 74, 127 –162. B OSAK , T. & N EWMAN , D. K. 2003. Microbial nucleation of calcium carbonate in the Precambrian. Geology, 31, 577–580. B RAMBILLA , E., P A¨ UKER , O., C OUSIN , S., S TEINER , U., R EIMER , A. & S TACKEBRANDT , E. 2007. High phylogenetic diversity of Flavobacterium spp. isolated from a hardwater creek, Harz Mountains, Germany. Organisms, Diversity & Evolution, 7, 145– 154. B RINKMANN , N., B EHNKE , A., B RUNS , S., M OHR , K., J AHN , R. & F RIEDL , T. 2007a. Assessing the diversity of diatoms in calcifying biofilms of hard water creeks. In: K USBER , W. H. & J AHN , R. (eds) Proceedings of the 1st Central-European Diatom Meeting 2007. Botanic Garden and Botanical Museum Berlin– Dahlem Freie Universita¨t Berlin, Berlin, 11–14. B RINKMANN , N., B EHNKE , A., J AHN , R., A RP , G. & F RIEDL , T. 2007b. Diversity of diatoms dominating biofilms on tufa stromatolites of hard water creeks. In: S TACKEBRANDT , E., W OZNICZKA , M., W EIHS , V. & S IKORSKI , J. (eds) Proceedings of the 11th International Conference on Culture Collections 2007 (ICCC 11). DSMZ, Braunschweig, 41– 44. ¨ ber schwa¨bische Kalktuffe insbesonB URGER , O. 1911. U dere des Echaztales. Inaugural-Dissertation, Universita¨t Tu¨bingen. C AUDWELL , C. 1987. E´tude expe´rimentale de la formation de micrite et de sparite dans les stromatolites d’eau douce a` Rivularia. Bulletin de la Socie´te´ ge´ologique de France, 8, 299–309. C LAUS , D. & B ERKELEY , R. C. W. 1986. Genus Bacillus. In: S NEATH , P. H. A., M AIR , N . S. & H OLT , J. G. (eds) Bergey’s Manual of Systematic Bacteriology. Williams & Wilkins, Baltimore, 1105– 1139. ¨ ber die Entstehung des Travertin in den C OHN , F. 1864. U Wasserfa¨llen von Tivoli. Neues Jahrbuch fu¨r Mineralogie, Geologie und Palaeontologie, 1864, 580– 610.
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS C OUSIN , S. 2009. Flavobacterial community structure in a hardwater rivulet and adjacent forest soil, Harz Mountain, Germany. Current Microbiology, 85, 409–415. C OUSIN , S., P A¨ UKER , O. & S TACKEBRANDT , E. 2007. Flavobacterium aquidurense sp. nov. and Flavobacterium hercynium sp. nov., from a hard-water creek. International Journal of Systematic and Evolutionary Microbiology, 57, 243– 249. C OUSIN , S., B RAMBILLA , E., Y ANG , J. & S TACKEBRANDT , E. 2008. Culturable aerobic bacteria from the upstream region of a karst water rivulet. International Microbiology, 11, 91–100. D ANDURAND , J. L., G OUT , R., H OEFS , J., M ENSCHEL , G., S CHOTT , J. & U SDOWSKI , E. 1982. Kinetically controlled variation of major components and carbon and oxygen isotopes in a calcite-precipitating spring. Chemical Geology, 36, 299 –315. D E P HILIPPIS , R., M ARGHERI , M. C., P ELOSI , E. & V ENTURA , S. 1993. Exopolysaccharide production by a unicellular cyanobacterium isolated from a hypersaline habitat. Journal of Applied Phycology, 5, 387–394. D E Y OREO , J. J. & V EKILOV , P. G. 2003. Principles of crystal nucleation and growth. In: D OVE , P. M., D E Y OREO , J. J. & W EINER , S. (eds) Biomineralisation. Reviews in Mineralogy and Geochemistry, 54, 57–93. D ECHO , A. W., V ISSCHER , P. T. & R EID , R. P. 2005. Production and cycling of natural microbial exopolymers (EPS) within a marine stromatolite. Palaeogeography, Palaeoclimatology, Palaeoecology, 219, 71–86. D E´ FARGE , C., T RIBBLE , J., S ANSONE , F. J., T RICHET , J., J AUNET , A. M. & R OBERT , M. 1996. Texture of microbial sediments revealed by cryo-scanning electron microscopy. Journal of Sedimentary Research, 66, 935–947. D REYBRODT , W., B UHMANN , D., M ICHAELIS , J. & U SDOWSKI , E. 1992. Geochemically controlled calcite precipitation by CO2 outgassing: field measurements of precipitation rates in comparison to theoretical predictions. Chemical Geology, 97, 285– 294. D UPRAZ , C. & V ISSCHER , P. T. 2005. Microbial lithification in marine stromatolites and hypersaline mats. Trends in Microbiology, 13, 429– 438. F ORD , T. D. 1989. Tufa – the whole dam story. Cave Science, 16, 39–49. F ORD , T. D. & P EDLEY , H. M. 1996. A review of tufa and travertine deposits of the world. Earth-Science Reviews, 41, 117–175. F REYTET , P. & P LET , A. 1996. Modern freshwater microbial carbonates: the Phormidium stromatolites (tufa-travertine) of southeastern Burgundy (Paris basin, France). Facies, 34, 219–238. F REYTET , P. & V ERRECCHIA , E. P. 1998. Freshwater organisms that build stromatolites: a synopsis of biocrystallization by prokaryotic and eukaryotic algae. Sedimentology, 45, 535– 563. F RITSCH , F. E. 1949. The lime encrusted Phormidium community of British streams. Verhandlungen der Internationalen Vereinigung fu¨r Theoretische und Angewandte Limnologie, 10, 141– 144. G ARNETT , E. R., A NDREWS , J. E., P REECE , R. C. & D ENNIS , P. F. 2004. Climatic change recorded by
115
stable isotopes and trace elements in a British Holocene tufa. Journal of Quaternary Science, 19, 251 –262. G AUTRET , P., C AMOIN , G., G OLUBIC , S. & S PRACHTA , S. 2004. Biochemical Control of Calcium Carbonate Precipitation in Modern Lagoonal Microbialites, Tikehau Atoll, French Polynesia. Journal of Sedimentary Research, 74, 462– 478. G EURTS , M. A. 1976. Gene`se et stratigraphie des travertins de Fond de Vale´e en Belgique. Acta Geographica Lovaniensia, 16, 1– 66. G LASSL , R. & S CHIEBER , M. 1990. Die Tuffrinne von Erasbach. Archaeopteryx, 8, 127 –139. G OLUBIC´ , S. 1969. Cyclic and noncyclic mechanisms in the formation of travertine. Verhandlungen Internationale Vereinigung fu¨r Theoretische und Angewandte Limnologie, 17, 956–961. G OLUBIC´ , S. 1973. The relationship between blue-green algae and carbonate deposits. In: C ARR , N. E. & W HITTON , B. (eds) The Biology of Blue– Green Algae. Blackwell, Oxford, 434–472. G OLUBIC´ , S., V IOLANTE , C., P LENKOVIC´ -M ORAJ , A. & G RGASOVIC´ , T. 2008. Travertines and calcareous tufa deposits: an insight into diagenesis. Geologia Croatica, 61, 363–378. G OMES , N. A. DE N. C. 1985. Modern stromatolites in a karst structure from the Malmani Subgroup, Transvaal Sequence, South Africa. Transactions of the Geological Society of South Africa, 88, 1– 9. G RU¨ NINGER , W. 1965. Rezente Kalktuffbildung im Bereich der Uracher Wasserfa¨lle. Abhandlungen zur Karst- und Ho¨hlenkunde, Reihe E (Botanik), 2, 1 –113. H ABICHT , K. S., G ADE , M., T HAMDRUP , B., B ERG , P. & C ANFIELD , D. E. 2002. Calibration of sulfate levels in the Archean Ocean. Science, 298, 2372–2374. H A¨ GELE , D., K EINFELDER , R., G RAU , J., B URMEISTER , E. G. & S TRUCK , U. 2006. Oncoids from the river Alz (southern Germany): tiny ecosystems in a phosphorus-limited environment. Palaeogeography, Palaeoclimatology, Palaeoecology, 237, 378 –395. H ARDIE , L. A. 1996. Secular variation in seawater chemistry: An explanation for the coupled secular variation in the mineralogies of marine limestone and potash evaporites over the past 600 m.y. Geology, 24, 279– 283. H ARTKOPF -F RO¨ DER , C., H ISS , M. & L EINFELDER , R. R. 1989. Holoza¨ne Su¨ßwasserkalke im Alme und Aftetal su¨dlich von Bu¨ren (Kreis Paderborn, NordrheinWestfalen). Mu¨nstersche Forschungen zur Geologie und Pala¨ontologie, 69, 261–289. H ERMAN , J. S. & L ORAH , M. M. 1987. CO2 outgassing and calcite pricipitation in Falling Spring Creek, USA. Chemical Geology, 62, 251 –262. H ERMAN , J. S. & L ORAH , M. M. 1988. Calcite precipitation rates in the field: Measurement and prediction for a travertine-depositing stream. Geochimica et Cosmochimica Acta, 52, 2347–2355. H ORI , M., H OSHINO , K., O KUMURA , K. & K ANO , A. 2008. Seasonal patterns of carbon chemistry and isotopes in tufa depositing groundwaters of southwestern Japan. Geochimica et Cosmochimica Acta, 72, 480– 492. J ACOBSON , R. L. & U SDOWSKI , E. 1975. Geochemical controls on a calcite precipitating spring. Contributions to Mineralogy and Petrology, 51, 65–74.
116
G. ARP ET AL.
J ANSSEN , A., S WENNEN , R., P ODOOR , N. & K EPPENS , E. 1999. Biological and diagenetic influence in Recent and fossil tufa deposits from Belgium. Sedimentary Geology, 126, 75– 95. J OBB , G. 2008. TREEFINDER version of June 2008. Munich, Germany. Distributed by the author at www. treefinder.de. J ONES , D. & K EDDIE , R. M. 1992. The genus Arthrobacter. In: B ALOWS , A., T RU¨ PER , H. G., D WORKIN , M., H ARDER , W. & S CHLEIFER , K. H. (eds) The Prokaryotes. 2nd edn. Springer Verlag, New York, 1283–1299. ¨ koloK ANN , E. 1973. Bemerkungen zur Systematik und O gie einiger mit Kalk inkrustierter Phormidiumarten. Schweizerische Zeitschrift fu¨r Hydrologie, 35, 141– 151. K ANO , A., M ATSUOKA , J., K OJO , T. & F UJII , H. 2003. Origin of annual laminations in tufa deposits, southwest Japan. Palaeogeography, Palaeoclimatology, Palaeoecology, 191, 243– 262. K ANO , A., K AWAI , T., M ATSUOKA , J. & I HARA , T. 2004. High-resolution records of rainfall events from clay bands in tufa. Geology, 32, 793–796. K AWAGUCHI , T. & D ECHO , A. W. 2002. Isolation and biochemical characterization of extracellular polymeric secreations (EPS) from modern marine stromatolites and its inhibitory effect on CaCO3 precipitation. Preparative Biochemistry Biotechnology, 32, 51–63. K AWAI , T., K ANO , A., M ATSUOKA , J. & I HARA , T. 2006. Seasonal variation in water chemistry and depositional processes in a tufa-bearing stream in SW-Japan, based on 5 years of monthly observations. Chemical Geology, 232, 33– 53. K LA¨ HN , H. 1923. Die Petrogenese der Kalktuffe nebst einigen sich daraus ergebenden geologischen Problemen. Geologisches Archiv, 2, 298– 316. K NOLL , A. H., F AIRCHILD , I. J. & S WETT , K. 1993. Calcified microbes in Neoproterozoic carbonates: Implications for our understanding of the Proterozoic/ Cambrian Transition. Palaios, 8, 512– 525. K OMA´ REK , J. & A NAGNOSTIDIS , K. 2005. Cyanoprokaryota, part 2. Oscillatoriales. In: B U¨ DEL , B., G A¨ RTNER , G., K RIENITZ , L. & S CHAGERL , M. (eds) Su¨ßwasserflora von Mitteleuropa Band 19/2. Gustav Fischer, Jena, 1– 759. L A T OUCHE , T. H. D. 1913. Geology of the Northern Shan States. Memoirs of the Geological Survey of India, 39(2), 1– 397. L IU , Z., S VENSSON , U., D REYBRODT , W., Y UAN , D. & B UHMANN , D. 1995. Hydrodynamic control of inorganic calcite precipitation in Huanglong Ravine, China: Field measurements and theoretical prediction of deposition rates. Geochimica et Cosmochimica Acta, 59, 3087– 3097. L OWENSTEIN , T. K., T IMOFEEFF , M. N., B RENNAN , S. T., H ARDIE , L. A. & D EMICCO , R. V. 2001. Oscillations in Phanerozoic seawater chemistry: evidence from fluid inclusions. Science, 294, 1086– 1088. L UDWIG , W., S TRUNK , O. ET AL . 2004. ARB: a software environment for sequence data. Nucleic Acids Research, 32, 1363– 1371. M ATSUOKA , J., K ANO , A., O BA , T., W ATANABE , T., S AKAI , S. & S ETO , K. 2001. Seasonal variation of
stable isotopic compositions recorded in a laminated tufa, SW Japan. Earth and Planetary Science Letters, 192, 31– 44. M ENZEL , H. 1909. Beitra¨ge zur Kenntnis der Quarta¨rbildungen im su¨dlichen Hannover. 4. Das Kalktufflager von Lauenstein. Jahrbuch der preußischen geologischen Landesanstalt 29(I), 604– 609. M ERZ -P REI ß, M. U. E. & R IDING , R. 1999. Cyanobacterial tufa calcification in two freshwater streams: ambient environment, chemical thresholds and biological processes. Sedimentary Geology, 126, 103– 124. M EYER , R. & S CHMIDT -K ALER , H. 1983. Erdgeschichte sichtbar gemacht. Ein geologischer Fu¨hrer durch die Altmu¨hlalb. Bayerisches Geologisches Landesamt, Mu¨nchen. M ONTY , C. L. V. 1976. The origin and development of crytalgal fabrics. Developments in Sedimentology, 20, 193–249. M ORAN , A. L. 2000. Calcein as a marker in experimental studies newly-hatched gastropods. Marine Biology, 137, 893 –898. M UURHOLM , S., C OUSIN , S., P A¨ UKER , O., B RAMBILLA , E. & S TACKEBRANDT , E. 2007. Pedobacter duraquae sp. nov., Pedobacter westerhofensis sp. nov., Pedobacter metabolipauper sp. nov., Pedobacter hartonius sp. nov. Pedobacter steynii sp. nov., isolated from a hardwater rivulet. International Journal of Systematic and Evolutionary Microbiology, 57, 2221–2227. N EU , T. R., S WERHONE , G. D. W. & L AWRENCE , J. R. 2001. Assessment of lectin-binding analysis for in situ detection of glycoconjugates in biofilm systems. Microbiology, 147, 299–313. N EU , T. R., W OELFL , S. & L AWRENCE , J. R. 2004. Threedimensional differentiotation of photo-autotrophic biofilm constituents by multi-channel laser scanning microscopy (single-photon and two-photon excitation). Journal of Microbiological Methods, 56, 161–172. N G , C. C., H UANG , W. C., C HANG , C. C., T ZENG , W. S., C HEN , T. W., L IU , Y. S. & S HYU , Y. T. 2006. Tufa microbial diversity revealed by 16S rRNA cloning in Taroko National Park, Taiwan. Soil Biology & Biochemistry, 38, 342–348. P ACE , N. R. 1997. A molecular view of microbial diversity and the biosphere. Science, 276, 734–740. P ALINSKA , K. A. & M ARQUARDT , J. 2008. Genotypic and phenotypic analysis of strains assigned to the widespread cyanobacterial morphospecies Phormidium autumnale (Oscillatoriales). Archives of Microbiology, 189, 325 –335. P ARKHURST , D. L. & A PPELO , C. A. J. 1999. User’s guide to PHREEQC (version 2) – a computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations. WaterResources Investigations Report, 99– 4259, 1– 312. U.S. Geological Survey, Denver. P EDLEY , H. M. 2000. Ambient temperature freshwater microbial tufas. In: R IDING , R. E. & A WRAMIK , S. M. (eds) Microbial Sediments. Springer, Berlin, 179–186. P ENTECOST , A. 1978. Blue– green algae and freshwater carbonate deposits. Proceedings of the Royal Society of London, Series B (Biological Sciences), 200, 43– 61.
TUFA-FORMING BIOFILMS IN KARSTWATER STREAMS P ENTECOST , A. 1985. Association of cyanobacteria with tufa deposits: identity, enumeration, and nature of the sheath material revealed by histochemistry. Geomicrobiology Journal, 4, 285– 298. P ENTECOST , A. 1993. British travertines: a review. Proceedings of the Geologists’ Association, 104, 23– 39. P ENTECOST , A. 2003. Taxonomic identity, ecology and distribution of the calcite-depositing cyanobacterium Phormidium incrustatum (Oscillatoriaceae). Cryptogamie/Algologie, 24, 307–321. P ENTECOST , A. 2005. Travertine. Springer, Berlin. P ENTECOST , A. & T HERRY , C. 1988. Inability to demonstrate calcite precipitation by bacterial isolates from travertine. Geomicrobiology Journal, 6, 185–194. P ENTECOST , A. & W HITTON , B. A. 2000. Limestones. In: W HITTON , B. A. & P OTTS , M. (eds) The Ecology of Cyanobacteria. Kluver, Dordrecht, 257 –279. P OKORNY , A. 1865. Die geologische Bedeutung der Laubmoose. Schriften des Vereins zur Verbreitung naturwissenschaftlicher Kenntnisse in Wien 4 (1863– 1884), 25. R EDDY , M. M. 1986. Effect of magnesium ions on calcium carbonate nucleation and crystal growth in dilute aqueous solutions at 25 8C. In: M UMPTON , F. A. (ed.) Studies in Diagenesis. U.S. Geological Survey Bulletin, 1578, 169–182. R EICHARDT , E. 1994. Zur Diatomeenflora (Bacillariophyceae) tuffabscheidender Quellen und Ba¨che im Su¨dlichen Frankenjura. Berichte der Bayerischen Botanischen Gesellschaft, 64, 119– 133. R EICHARDT , E. 1995. Die Kieselalgenflora (Bacillariophyceae) des Wachsenden Steins von Usterling. Berichte der Bayerischen Botanischen Gesellschaft, 65, 87– 92. ¨ ber fossile Diatomeen aus NordR EICHELT , H. 1899. U bo¨hmen. Berichte der Naturforschenden Gesellschaft zu Leipzig, 1899– 1900, 27–35. R ICHERT , L., G OLUBIC´ , S., L E G UE´ DE` S , R., R ATISKOL , J., P AYRI , C. & G UEZENNEC , J. 2005. Characterization of exopolysaccharides produced by cyanobacteria isolated from Polynesian microbial mats. Current Microbiology, 51, 379– 384. R IDING , R. 1982. Cyanophyte calcification and changes in ocean chemistry. Nature, 299, 814–815. R IDING , R. 1991. Classification of microbial carbonates. In: R IDING , R. (ed.) Calcareous algae and stromatolites. Springer, Berlin, 21– 51. R IDING , R. 2000. Microbial carbonates: the geological record of calcified bacterial-algal mats and biofilms. Sedimentology, 47(Supplement 1), 179– 214. R IDING , R. 2006. Cyanobacterial calcification, carbon dioxide concentrating mechanisms, and Proterozoic– Cambrian changes in atmospheric composition. Geobiology, 4, 299–316. R IDING , R. & L IANG , L. 2005. Geobiology of microbial carbonates: metazoan and seawater saturation state influences on secular trends during the Phanerozoic. Palaeogeography, Palaeoclimatology, Palaeoecology, 219, 101– 115. R OTT , E. 1991. Oncoids from the summer-warm River Alz (Bavaria) – morphology and dominant cyanophytes. Archiv fu¨r Hydrobiologie Supplementband, 92 (Algological Studies 64), 469–482.
117
R OTT , E. 1994. Der Algenaufwuchs in der Oberen Alz (Oberbayern). Berichte des Naturwissenschaftlich-Medizinischen Vereins in Innsbruck, 81, 229– 253. S ABATER , S. 2000. Structure and architecture of a stromatolite from a Mediterranean stream. Aquatic Microbial Ecology, 21, 161– 168. S ALZWEDEL , U. 1992. Kalktuffvorkommen im Eschweger Raum. Berliner Geographische Abhandlungen Beihefte, 1, 1 –85. S CHMIDT -K ALER , H. 1974. Geologische Karte von Bayern 1:25 000 Erla¨uterungen zum Blatt Nr. 6634 Altdorf. Bayerisches Geologisches Landesamt, Mu¨nchen. S CHU¨ RMANN , E. 1918. Die chemisch-geologischen Vorga¨nge bei der Bildung des Uracher Wasserfalls. Jahreshefte des Vereins fu¨r vaterla¨ndische Naturkunde in Wu¨rttemberg, 74, 58–68. S CHU¨ TTE , J. 1761. Oryctographia Ienensis sive fossilium et mineralium in agro ienensi brevissima descriptio. Gu¨the, Fickelscherr, Ienae. S HARON , N. & L IS , H. 2003. Lectins. Kluwer Academic Publishers, Dordrecht, Boston, London. S HIRAISHI , F., B ISSETT , A., DE B EER , D., R EIMER , A. & A RP , G. 2008a. Photosynthesis, respiration and exopolymer calcium-binding in biofilm calcification (Westerho¨fer and Deinschwanger Creek, Germany). Geomicrobiology Journal, 25, 83–94. S HIRAISHI , F., R EIMER , A., B ISSETT , A., DE B EER , D. & A RP , G. 2008b. Microbial effects on biofilm calcification, ambient water chemistry and stable isotope records (Westerho¨fer Bach, Germany). Palaeogeography, Palaeoclimatology, Palaeoecology, 262, 91–106. S HIRAISHI , F., Z IPPEL , B., N EU , T. R. & A RP , G. 2008c. In situ detection of bacteria in calcified biofilms using FISH and CARD-FISH. Journal of Microbiological Methods, 75, 103– 108. S OMMERMEIER , L. 1913. Der Kartstein und der Kalktuff von Dreimu¨hlen bei Eiserfey in der Eifel. Verhandlungen des Naturhistorischen Vereins der Preussischen Rheinlande und Westfalens, 70, 303– 333. S PIRO , B. & P ENTECOST , A. 1991. One day in the life of a stream – a diurnal inorganic carbon mass balance for a travertine-depositing stream (Waterfall Beck, Yorkshire). Geomicrobiology Journal, 9, 1 –11. S TACKEBRANDT , E., L ANG , E., C OUSIN , S., P A¨ UKER , O., B RAMBILLA , E., K ROPPENSTEDT , R. & L U¨ NSDORF , H. 2007. Deefgea rivuli gen. nov., sp. nov., a member of the class Betaproteobacteria. International Journal of Systematic and Evolutionary Microbiology, 57, 639 –645. S TAL , L. 2003. Microphytobenthos, their extracellular polymeric substances, and the morphogenesis of intertidal sediments. Geomicrobiology Journal, 20, 463– 478. S TILLE , H. 1932. Geologische Karte von Preußen und benachbarten deutschen La¨ndern. Erla¨uterungen zu Blatt Go¨ttingen Nr. 2520. 3rd ed., Preußische Geologische Landesanstalt, Berlin. S TIRN , A. 1964. Kalktuffvorkommen und Kalktufftypen der Schwa¨bischen Alb. Abhandlungen zur Karst- und Ho¨hlenkunde, Reihe E (Botanik), 1, 1 –92.
118
G. ARP ET AL.
S TRIMMER , K. & R AMBAUT , A. 2002. Inferring confidence sets of possibly misspecified gene trees. Proceedings of the Royal Society of London, Series B, 269, 137– 142. S UAREZ , D. L. 1983. Calcite supersaturation and precipitation kinetics in the lower Colorado River, All-American Canal and East Highland Canal. Water Resources Research, 19, 653– 661. S UDA , S., W ATANABE , M. M., O TSUKA , S., M AHAKAHANT , A., Y ONGMANITCHAI , W., N OPARTNARAPORN , N., L IU , Y. & D AY , J. G. 2002. Taxonomic revision of water bloom-forming species of oscillatorioid cyanobacteria. International Journal of Systematic and Evolutionary Microbiology, 52, 1577–1595. S UMNER , D. Y. 1997. Late Archean calcite-microbe interactions: two morphologically distinct microbial communities that affected calcite nucleation differently. Palaios, 12, 302– 318. S ZULC , J. & S MYK , B. 1994. Bacterially controlled calcification of freshwater Schizothrix-stromatolites: an example from the Pieniny Mts., Southern Poland. In: B ERTRAND -S ARFATI , J. & M ONTY , C. (eds) Phanerozoic Stromatolites II. Kluwer, Dordrecht, 31–51. T ATON , A., G RUBISIC , S. ET AL . 2006. Polyphasic study of Antarctic cyanobacterial strains. Journal of Phycology, 42, 1257– 1270. T HUNMARK , S. 1926. Bidrag till Kannendomen om recenta Kalktuffer. Geologiska Foreningens i Stockholm Forhandlingar, 48, 541–583. T RICHET , J. & D E´ FARGE , C. 1995. Non-biologically supported organomineralization. Bulletin de l’Institut Oce´anographique Monaco, no. spe´c., 14, 203– 236. U NGER , F. 1861. Beitra¨ge zur (Anatomie und) Physiologie ¨ ber den anatomischen Bau des der Pflanzen. VII. U ¨ ber die kalkausscheidenden Moosstammes. VIII. U Organe der Saxifraga enistala. Sitzungsberichte der Akademie der Wissenschaften Wien, mathematischnaturwissenschaftliche Klasse, 43 II (1. Ha¨lfte), 497– 524. U SDOWSKI , E., H OEFS , J. & M ENSCHEL , G. 1979. Relationship between 13C and 18O fractionation and changes in major element composition in a Recent calcite-depositing spring – A model of chemical variations with inorganic CaCO3 precipitation. Earth and Planetary Science Letters, 42, 267–276.
B UCH , L. 1809. Geognostische Beobachtungen auf Reisen durch Deutschland und Italien. Band II. Haude & Spener, Berlin. ¨ uterungen zur geologischen VON K OENEN , A. 1907. Erla Specialkarte von Preussen und den Thu¨ringischen Staaten. Blatt No. 28 Go¨ttingen. 2nd edn. Schropp, Berlin. VON P IA , J. 1933. Die rezenten Kalksteine. Zeitschrift fu¨r Kristallographie, Mineralogie und Petrographie, Abteilung B, Erga¨nzungsband, 1, 1– 420. VON P IA , J. 1934. Die Kalkbildung durch Pflanzen. Beihefte zum botanischen Zentralblatt Kassel A, 52, 1 –72. ¨ BEL , U. & S TACKEVON W INTZINGERODE , F., G O BRANDT , E. 1997. Determination of microbial diversity in enviromental samples: pitfalls of PCR-based rRNA analysis. FEMS Microbiology Reviews, 21, 213–229. W ALCH , J. E. I. 1773. Von den incrustierten Ko¨rpern. Die Naturgeschichte der Versteinerungen Band 1, Capitel 6, Nu¨rnberg (Felßecker). ¨ ber die Beteiligung kalkablagernder W ALLNER , J. 1934. U Pflanzen bei der Bildung su¨dbayrischer Tuffe. Bibliotheca Botanica, 110, 1 –30. W ILMOTTE , A., V AN DER A UWERA , G. & D E W ACHTER , R. 1993. Structure of the 16S ribosomal RNA of the thermophilic cyanobacterium Chlorogloeopsis HTF (‘Mastigocladus laminosus HTF’) strain PCC7518, and phylogenetic analysis. FEBS Letters, 317, 96– 100. W INSBOROUGH , B. M. & G OLUBIC´ , S. 1987. The role of diatoms in stromatolite growth: two examples from Modern freshwater settings. Journal of Phycology, 43, 195– 201. W USTMAN , B. A., L IND , J., W ETHERBEE , R. & G RETZ , M. R. 1998. Extracellular Matrix Assembly in Diatoms (Bacillariophyceae). III. Organization of Fucoglucuronogalactans within the Adhesive Stalks of Achnanthes longipes. Plant Physiology, 116, 1431–1441. Z HANG , D. D., Z HANG , Y., Z HU , A. & C HENG , X. 2001. Physical mechanisms of river waterfall tufa (travertine) formation. Journal of Sedimentary Research, 71, 205– 216. Z IPPEL , B. & N EU , T. R. submitted. EPS glycoconjugates in tufa forming biofilms characterized by fluorescence techniques and laser scanning microscopy. Applied and Environmental Microbiology. VON
Seasonal record from recent fluvial tufa deposits (Monasterio de Piedra, NE Spain): sedimentological and stable isotope data ´ CAR, C. SANCHO, M. VA ´ ZQUEZ-URBEZ, C. ARENAS*, C. OSA ´ L. AUQUE & G. PARDO Department of Earth Science, University of Zaragoza, Calle Pedro Cerbuna 12, E-50009 Zaragoza, Spain *Corresponding author (e-mail:
[email protected]) Abstract: Physical and hydrochemical parameters and sedimentation rates were monitored twice a year from August 1999 to March 2003 at the Monasterio de Piedra area (NE Spain). Different tufa facies related to distinct fluvial subenvironments were characterized and the isotopic composition of water was analysed seasonally. Sedimentary features (thickness, texture and structure) and stable isotope composition of the seasonal record on tablets were analysed. The seasonal intervals were identified from six-monthly thickness measurements on tablets. Sedimentation rates had a strong seasonal pattern with higher values in warm periods than in cool ones, although erosive events and sporadic, warmer-than-normal climate conditions altered it. Three main types of fluvial facies were studied in detail: dense, stromatolitic tufa; dense to porous, massive tufa; and spongy, moss- and alga-bearing, crudely laminated tufa. Textural features of deposits from warm and cool periods had a variable pattern. The sediment d18O composition showed a rhythmic variation, with higher values in cool periods and lower in warm ones, caused by the fractionation due to seasonal temperature variations. The calculated temperatures for a theoretical equilibrium precipitation accord with the actual measured temperatures. The sediment d13C composition had an irregular pattern, indicating that other parameters than temperature intervened in the fluvial system.
The study of the short-term (seasonal) behaviour pattern of active tufa system allows: (1) to advance in the knowledge of the mechanical degassing versus photosynthesis role in tufa deposition processes; (2) to analyse the problems arisen from correlating stable isotope composition of water and derived tufaceous sediment; (3) to asses the use of Quaternary tufa deposits for high-resolution paleoclimate studies; and (4) to increase the limited knowledge of present-day tufa activity under different environmental conditions. Understanding present-day dynamics of fluvial tufa systems requires field monitoring of tufa sediment formation. Measurement of tufa deposition rates and analysis of sediment, including both textural features and stable isotope composition, are some of the most important parameters in order to establish the characteristics of the tufa formation process (Chafetz et al. 1991; Drysdale & Gilleson 1997; Lu et al. 2000; Matsuoka et al. 2001; Ihlenfeld et al. 2003; Kano et al. 2004, 2007; Lojen et al. 2004; O’Brien et al. 2006; Shiraishi et al. 2008). Due to the intense tufaceous activity observed, the fluvial system within the Monasterio de Piedra Natural Park is a highly suitable scenary to measure sedimentation rates and to analyse the deposits
(sedimentological and geochemical studies) along with their environmental conditions. In addition, both in the Park and in the surrounding areas, the presence of many fossil tufa buildings proves the high tufa sedimentation during some Quaternary intervals until the present. The aim of this contribution is to analyse seasonal variations in sedimentological features recorded on artificial substrates installed in different fluvial subenvironments, and the significance of the correlative carbon and oxygen stable isotope data of sediment and water. The study is based on: (1) monitoring of physical and hydrochemical parameters, and characterization of the type of substrates; (2) sedimentation rates measured seasonally from tablets installed from August 1999 to March 2003; (3) correlation of these thickness values with the sedimentary record on crosssections of tablets once these were removed from the field; and (4) textural and structural characteristics and stable isotope composition of the several seasonal intervals identified in the tablet records. The study is focused on the analysis of the factors that control present-day tufa sedimentation in the Monasterio de Piedra Natural Park and aims to obtain environmental and climatic information that
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 119–142. DOI: 10.1144/SP336.7 0305-8719/10/$15.00 # The Geological Society of London 2010.
120
C. ARENAS ET AL.
Fig. 1. Location of the study area. (a) Geographical location of the River Piedra and the Monasterio de Piedra Natural Park. (b) Geological map of the area, showing the course of the River Piedra and its entry into the Tranquera Reservoir.
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN
121
Table 1. Values of water and air temperature, rainfall and water discharge. Water temperature is the mean of 5 sampling sites at the end of December and June. Air temperature, rainfall and water discharge are expressed as the mean of the values corresponding to the same months that define the cool and warm periods of the sedimentation rate measurements. Data provided by the Agencia Estatal de Meteorologı´a and Confederacio´n Hidrogra´fica del Ebro (gauging point of Nue´valos) Periods
Air T (8C)
Water T (8C)
Precipitation (mm)
Water discharge (m3/s)
Oct. 02–Mar. 03 Apr. 02–Sept. 02 Oct. 01–Mar. 02 Apr. 01–Sept. 01 Nov. 00–Mar. 01 Apr. 00–Oct. 00 Nov. 99–Mar. 00
8.70 18.63 6.72 18.99 8.54 18.37 6.70
6.9 18.9 10.1 16.5 10.1 14.5 11.1
326.5 442.3 119.0 189.0 249.8 416.4 178.6
1.14 0.62 0.69 0.75 1.03 0.71 1.12
help understand the sedimentary processes that govern present and ancient tufa formation in Mediterranean regions.
Geological context, climate and hydrology The Monasterio de Piedra Natural Park is an exceptional place near the town of Calatayud, between Madrid and Zaragoza (NE Spain) (Fig. 1a). The Cistercian monastery (built in 1195AD) is at an altitude of 786 m, although the surrounding uplands range between 850 and 1000 m. From a geological point of view the study area is within the Iberian Range (Fig. 1a), a NW– SE trending alpine intraplate fold belt. Near the Park, the following sequence is found (Fig. 1b): Palaeozoic shales and quartzites, Triassic sandstones, dolostones and gypsum-rich marls, Lower Jurassic dolostones, Lower Cretaceous sands and a thick Upper Cretaceous succession of limestones and dolostones. Miocene conglomerates, sandstones and mudstones overlay this sequence. All these units are slightly deformed by NW–SE trending folds and faults. During the Quaternary the River Piedra formed a number of gorges and small canyons in which thick Upper Pleistocene and Holocene tufa outcrops are common (Arenas et al. 2004). The River Piedra flows from south to north through the Iberian Range (Fig. 1b). Its longitudinal profile shows several important knickpoints with large waterfalls, two of which (c. 15 and 35 m high) lie within the Park (Fig. 1c). At the lower part of the Park there is a shallow lake – Lake Espejo – that reaches a depth of some 2 m at its
centre and occupies an area of around 2500 m2. The lake empties into the River Piedra via an artificial channel downstream of the Park (Fig. 1c). The climate of the area is of the continental Mediterranean type with strong seasonal contrasts. For the studied period, the mean annual air temperature was around 12 8C, with mean montly values of 5 8C in January and 23 8C in July–August. Water temperature ranged from 6.9–11.1 8C in autumn/ winter and from 14.5 –18.9 8C in spring/summer. The mean annual rainfall was around 400 mm, irregularly distributed, with maxima in spring and autumn (Table 1). The drainage area of the River Piedra is about 1545 km2. From 1999–2003 the mean water discharge ranged between 0.62–1.14, with a mean of 0.87 m3/s (Table 1). The river receives important water inputs from a Mesozoic carbonate aquifer (Lower Jurassic and Upper Cretaceous limestones and dolostones around 500 m thick). The strong influence of underground water makes water discharge show only slight monthly variations. These small monthly variations have a cyclical pattern with maximum values in winter and minimum values in summer. Some exceptional floods caused by heavy rain in spring and summer had a strong influence on the tufa sedimentary system by triggering strong erosion processes that affected the sedimentary record. At present the river presents highly active tufa formation, particularly near the Monasterio de Piedra Natural Park, as shown by present-day monitoring (Arenas et al. 2004; Va´zquez-Urbez et al. 2005, Va´zquez-Urbez et al. in press).
Fig. 1. (Continued) (c) Course of the River Piedra, with location of the vertical waterfalls and Lake Espejo within the Park and location of tablets. Tablets 1 to 6 were set on the river bed, tablets 10 and 11 in stepped waterfalls on the river banks, tablets 7, 8, 9, 12 and 13 in stepped waterfalls outside the main stream area, and tablet 14 in the Iris Cave behind the Cola de Caballo waterfall.
122
C. ARENAS ET AL.
Methods Several sites that represent different fluvial subenvironments (mostly defined by the physical flow characteristics and biofacies) within the Monasterio de Piedra Natural Park were choosen for periodic monitoring from August 1999 onwards. This contribution deals with data obtained between August 1999–March 2003. The study comprised sixmonthly monitoring of properties and parameters of water and sediment. In this study, spring/ summer was the warm period and autumn/winter the cool period. These periods correspond to the astronomical seasons. During the first year of monitoring, the warm period included October. Some physical and chemical parameters of the water (e.g. velocity, depth, temperature, conductivity and pH) were measured in situ in five selected sampling sites along the river at the end of June and December from 1999 to March 2003. At the same time, water samples were taken at those sites for chemical (alkalinity, Cl, SO4, Ca, Mg, Na and K) determinations and d18O analyses. Water analyses were carried out at the Petrology and Geochemistry Laboratory of the University of Zaragoza. Speciation–solubility calculations (e.g. to calculate pCO2, Total Disolved Inorganic Carbon, and calcite saturation index) were performed with the code PHREEQC (Parkhurst & Appelo 1999) and the WATEQ4F thermodynamic database (Ball & Nordstrom 2001) (see details of procedures in Va´zquez-Urbez et al. in press). Determinations of d18O in water were based on the CO2 –H2O equilibrium according to Epstein & Mayeda (1953) method, adding pure CO2 and keeping at 25 8C during 12 hours. Water stable isotope determinations were made at the Stable Isotope Analysis Laboratory of the University of Salamanca (Spain). The results are expressed in d ‰ units and reported against V-SMOW. In addition, tufa sediment was sampled every six months (end of June and end of December) in areas adjacent to the tablets that had identical physical flow and sediment characteristics to those of the tablets. The most recent parts were selected, then ground and sieved (53 mm) for mineralogical analyses and determination of Mg content (x-ray diffraction with a Phillips PW 1729 diffractometer) at the Mineralogy and Cristallography Laboratory of the University of Zaragoza. Sediment characteristics (texture, flora, mineralogy, d13C and d18O) and sedimentation rates were obtained from sediment deposited on limestone tablets (25 15 2 cm) installed in different subenvironments along the River Piedra, laying parallel to the floor. The tablets and mesurement device were based on those designed by Drysdale & Gillieson (1997). The measurement device
(Micro-Erosion Meter, MEM) consisted of a 5 10 point coordinate on a rectangular framework on which a digital micrometer was installed, allowing sediment height to be measured at 50 coordinates (Fig. 2). The tablets were removed from the river at the end of the warm (end of March) and cool (end of September) periods. In the laboratory, the tablets were first air-dried for 5 days, and then measurements of sediment accumulation by the aforementioned device were made. About a week after their removal, the tablets were put back in their original position. The differences in sediment height (mean of the 50 points per tablet) between consecutive measuring times represented the sixmonthly sedimentation rates at each site. Measurements are expressed as millimetres per six-monthly period. Up to 14 tablets were monitored at 13 sites from August 1999–March 2003. By March 2003 all tablets were definitively removed. In the laboratory, the tablets were cut perpendicularly to the accumulation surfaces by means of a power saw. Crosssection cuttings were made coinciding with one of the ten five-point rows of measurement. In each case, one of the two parts was used to obtain thin sections and samples for scanning electron microscopy and stable isotope analysis. Prior to sediment sampling, each tablet section was analysed to identify the several seasonal sedimentation intervals. Each tablet section was scanned and then the six-monthly measured accumulation data of the five-point row that coincided with the section were plotted on the corresponding scanned tablet cross-section. Once the different six-monthly accumulation periods were identified, textural analysis and sampling of sediment for stable isotope analyses and Scanning Electron Microscopy observations were made. Textural observations of the sediment were made using a petrographic microscope (employing thin sections of the deposits on the tablets) and by scanning electron microscopy performed at the Electron Microscope Service of the University of Zaragoza (using a JEOL JSM 6400). Sediment samples for stable isotope analysis (d18O and d13C) were selected from each sixmonthly interval with a punch, then ground and sieved (32 mm). The samples were reacted with 103% phosphoric acid at 25 8C, following standard methods (McCrea 1950). Sediment stable isotope determinations were made by the Stable Isotope Analysis Service of the University of Salamanca (Spain). The analytical precision was better than 0.1‰ for d18O and d13C. The results are expressed in d ‰ units and compared against VPDB. Additionally, water discharge of the Piedra River, rainfall and air temperature data of the studied area were obtained from the Confederacio´n
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN
123
Fig. 2. Measurement device (Micro-Erosion Meter, MEM) used to measure sediment thickness on the tablets. Tablet is 25 15 2 cm.
124
C. ARENAS ET AL.
hidrogra´fica del Ebro (Zaragoza) and the Agencia Estatal de Meteorologı´a (Table 1).
Hydrochemistry The River Piedra water is of calcium-bicarbonate type, with pH values always between 8–8.5 (except for two values c. 7.4 in September 2000), conductivity between 630 –670 mS/cm, alkalinity between 250 –310 mg/L and Ca concentration between 85 –102 mg/L (for detailed hydrochemical data, see Va´zquez-Urbez et al. in press). Alkalinity and calcium concentrations showed a more or less gradual decrease from the upstream site to the downstream site, indicating the effective precipitation of calcium carbonate. The water is permanently saturated with respect to calcite, with saturation index [log IAP/K(T)] values close to þ1 (without clear downstream variations) and, therefore, at levels sufficient to overcome the carbonate kinetic precipitation barrier (Jacobson & Usdowski 1975; Dandurand et al. 1982; Drysdale et al. 2002). All these trends are similar to those observed in other tufa-depositing rivers and streams (Herman & Lorah 1987; Lu et al. 2000; Drysdale et al. 2002; Malusa et al. 2003). Differences in the hydrochemical data between warm and cool periods only were clearly seen for calcium and alkalinity (Osa´car et al. 2003; Va´zquez-Urbez et al. in press), with a steeper decreasing pattern along the river in the warm periods. These differences indicate a more intense calcite precipitation process in warm periods. This seasonal pattern related to temperature changes has also been observed in similar fluvial systems (e.g. Kano et al. 2003; Kawai et al. 2006).
Sedimentary subenvironments of the fluvial system Topographic features, flow characteristics and biofacies of the River Piedra in the Monasterio de Piedra Natural Park allowed Va´zquez-Urbez et al. (in press) to distinguish several tufa sedimentation subenvironments, some of which have been described from other fluvial systems (e.g. Pedley 1990, 2009). The main characteristics of the several subenvironments in the Monasterio de Piedra are summarized as follows (Fig. 3): (1) Areas of fast flowing water, including steeper stretches of the river bed (e.g. small jumps and stepped falls up to 2 m high) (Fig. 3a, b, c). Mean water velocity ranged between 109.33 and 215.67 cm/s, and depth between a 3 and 10 cm. The sediment consisted of dense laminated tufa mostly formed of microbial tube-like bodies made of calcite crystals.
(2) Areas of slow flowing to standing water: these appeared upstream and downstream of waterfalls and small barrages along the river (Fig. 3d). Water velocity ranged between almost 0 (although at the sites monitored the minimum was 32 cm/s) to 55.33 cm/s, and water depth from 10 to 30 cm. Lime mud, diverse carbonate grains and in situ coated plants in palustrine conditions were present. Microbial filaments and tube-like bodies appear in areas of faster flow. (3) Stepped waterfalls and barrages with bryophytes: these include small waterfalls and barrages (up to 3 m high) on the river bed, and stepped waterfalls (up to about 8 m high) (Fig. 3e, f). All of them had a thin water lamine with turbulent flow. They are mainly formed of mosses and filamentous green algae, with minor cyanobacteria and herbaceous plants, all of which were coated by calcite producing porous deposits. These included trapped coated phytoclasts and hanging coated plants. (4) Spray and splash areas beside waterfalls, with mats of mosses, cyanobacteria and filamentous algae coated and impregnated by calcite. These constituted thin, generally non-laminated deposits. (5) Waterfalls with vertical drops (e.g. waterfalls Caprichosa and Cola de Caballo, 10 and 35 m high) (Fig. 3c). Herbaceous hanging plants and moss mats that grew in parts of the river section with low water flow were coated by calcite constituting banded deposits with steep inclination. (6) Caves behind waterfalls (e.g. Cave Iris). These were shady areas with great water seepage, in which bryophyte and microbial mats coated diverse cave surfaces. Stalactites developed there, sometimes from hanging plants coated by calcite. In addition, in most fluvial deposits diatoms, aquatic insects, worms and crustaceans were present.
Analysis of sedimentation patterns To carry out the analysis of the seasonal variations of sedimentological (thickness, structure, texture) and geochemical (stable isotopes) features, data of sedimentation rates obtained from six-monthly thickness measurements with the MEM on tablets were used to identify seasonal intervals of the tufa deposited on the tablets installed from August 1999–March 2003. Data on sedimentation rates for a longer period and their relationships to environmental factors, such as hydrology, hydrochemistry, climate and biological activity, are dealt with in detail in the paper by Va´zquez-Urbez et al. (in press). The following section is based on the information of that paper that is essential for comprehension of the present research.
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN
125
Fig. 3. Field views of several fluvial subenvironments. (a) and (b) Fast flowing water areas. Note in (b) the location of tablet 3 in fast flow conditions on the river bed. (c) Small fall and Caprichosa waterfall. Note tablet 2 in a fall (about 2 m high) along the river bed. In (c) the Caprichosa vertical waterfall (15 m high) is seen in the background. (d) Slow flowing water areas upstream of small falls. (e) and (f) Stepped waterfall with bryophytes and algae. In (e), tablets installed in February 2003 to substitute the ones installed in 1999 (tablets 7 and 8), which were removed at the end of March 2003 (these cannot be distinguished in the photograph). Note in (f) tablet 8 with mosses and algae.
126
C. ARENAS ET AL.
Sedimentation rates from six-monthly measurements on tablets Mean sedimentation rates recorded by the 14 tablets installed are expressed in Table 2. Measurements with the MEM indicated that the greater rates were those of tablets in areas of fast flowing water on the river bed and in shallow and turbulent water areas such as stepped waterfalls. Tablets in standing to slow flowing water and spray areas had lower rates. The minimun corresponded to spray areas beside waterfalls and in the cave Iris. Detail sedimentation rates from August 1999– March 2003 of seven selected tablets are shown in Table 3. These tablets are representative of the two subenvironments that recorded greater sedimentation, the deposits of which are suitable for textural and stable isotope analysis. Va´zquez-Urbez et al. (in press) suggested that flow velocity and turbulence exerted an important control on calcite precipitation through mechanical CO2 loss on the river settings. Similar results were obtained in several present-day tufa-depositing streams by Lorah & Herman (1988), Liu et al. (1995), Drysdale & Gillieson (1997), Merz-Preiß & Riding (1999), Lu et al. (2000) and Chen et al. (2004). On the other hand, the amount of water and its dynamics (e.g. spray versus running water) controlled calcite precipitation in spray and drip conditions beside waterfalls. Sedimentation rates presented a clear periodic pattern that was observed in all subenvironments (Tables 2 and 3): higher values in the warm periods and lower values in the cool ones. Exceptions to this pattern were seen in the two last periods of monitoring (Table 3): tablets 1, 3, 5 and 6 recorded lower rates than expected in the warm period (April 2002–September 2002). In contrast, in the following cool period (October 2002– March 2003) some of these (tablets 1, 3 and 5) had values higher than expected, close to those of
the warm periods. On the other hand, during the same warm period, tablets 2 and 7 presented sedimentation rates much higher than those of previous warm periods and in the following cool period the values were much higher than those of previous cool periods. Tablet 8 recorded a strong erosion (221.28 mm). Some exceptions to the periodic pattern and the negative values recorded during the cool periods were associated with erosion due to strong storms and/or floodings (e.g. in May and July 2002), whose effects could be recognized in the field. Except for these occasional high discharge events, the flow characteristics of the river (e.g. water velocity) and hydrochemical parameters did not experience great modification from warm to cool periods. Temperature was the only variable that showed a clear seasonal variation. Va´zquez-Urbez et al. (in press) attributed the periodic time variation in sedimentation rates to changes in temperature between the warm and cool periods. These temperature changes controlled both the inorganic calcite precipitation rate and the development of flora; the latter through variations in biomass and the correlative photosynthetic activity. A biological participation, even in mechanical CO2-degassing dominated systems, has been demonstrated in recent experimental studies (Rogerson et al. 2008; Shiraishi et al. 2008; Pedley et al. 2009). Therefore, despite the physiological activity of aquatic flora is very difficult to evaluate (Merz-Preiß & Riding 1999; Emeis et al. 1987; Shiraishi et al. 2008), the high sedimentation rate recorded during the last cool period is thought to be linked at least partially to such activity. That is it, the mild temperatures recorded in autumn of 2002 (mean temperature was 3.2 8C higher in November and 5.2 8C higher in December with respect to the same months in 2001) would have favoured the biological activity, producing greater flora mass and associated calcite precipitation in this period.
Table 2. Mean sediment thickness recorded by 14 tablets in the different depositional subenvironments Environments
Fast flowing water areas Slow and standing flowing water areas Stepped cascades with continuous jet Spray areas in cascades Spray and dripping in a shaddy cave
Tablet number
Mean of warm periods (mm)
Mean of cool periods (mm)
Mean yearly accumulation (mm)
2, 3, 4, 5 1, 6
8.89 1.58
4.21 1.00
13.10 2.58
7, 8, 9, 10, 13
7.85
1.88
9.73
11, 12 14
2.32 0.02
20.73 0.02
1.59 0.04
Table 3. Sedimentation rates measured with the MEM from seven selected tablets, their physical environmental characteristics and bio- and lithofacies. Estimates of the mean yearly accumulation were calculated from the sum of the means of the cool and warm periods. The total accumulation includes the first colonization values Tablet num.
2
3
5
1
6
Location and characteristics
Downstream of Caprichosa vertical waterfall. Fast flowing water on a fall along the river bed. Inclination: 578 – .328. Ban˜o de Diana. Fast flowing water on the river bed. Subhorizontal: 108. Upstream of Cola de Caballo vertical waterfall. Fast flowing water on a small jump on the river bed. Inclination: 268 Upstream of Caprichosa vertical waterfall. Slow to standing water on left palustrine riverbank. Inclination: 08. Downstream of Cola de Caballo vertical waterfall. Slow flowing water on the river bed. Subhorizontal: 48.
Depth (cm)
Flow vel. (cm/s)
Warm Cool
Warm Cool
6.30 6.50
Bio and lithofacies on tablets
6.00 5.00
215.67 Mostly dense, 195.00 laminated micrite and spar deposits, mainly made of tube-like microbial calcite bodies. Diatoms. 142.67 109.33
5.33 5.00
185.67 183.33
Coloniz. (mm)
Accumulation (mm)
Aug. 99 – Nov. 99 – Apr. 00 – Nov. 00 – Apr. 01 – Oct. 99 Mar. 00 Oct. 00 Mar. 01 Sept. 01
Oct. 01 – Mar. 02
Apr. 02 – Sept. 02
Mean Mean Mean yearly Total of of cool accumulation accumulation warm periods (mm) (mm) Oct. 02 – periods (mm) Mar. 03 (mm)
1.64
8.08
3.71
14.13
12.04
11.11
5.80
16.90
39.60
2.89
1.54
10.23
1.73
9.76
1.55
1.93
7.45
7.31
3.07
10.37
37.08
2.26
1.93
7.37
2.17
7.17
3.37
3.61
6.23
6.05
3.43
9.48
34.11
17.00 13.5
32.00 Massive lime mud, 55.33 with coated phytoclasts and gastropods.
0.72
0.98
1.38
21.57
1.85
1.23
0.08
4.06
1.10
1.18
2.28
8.73
15.00 20.00
58.00 Massive to poorly 49.00 laminated calcite with microbial films. Diatoms.
2.28
0.76
2.50
0.64
3.63
0.98
0.02
0.91
2.05
0.82
2.87
11.72
(Continued)
Table 3. Continued Tablet num.
7
8
Location and characteristics
Stepped waterfall on riverbank. Discontinous flow + splashing at the base of the waterfall. Inclination: 238 – . 428. Stepped waterfall on riverbank. Stepped continuous flow. Inclination: 608 – . 488.
Depth (cm)
Flow vel. (cm/s)
Warm Cool
Warm Cool
0 to 1–2
1 to 2
Bio and lithofacies on tablets
Coloniz. (mm)
Accumulation (mm)
Aug. 99 – Nov. 99 – Apr. 00 – Nov. 00 – Apr. 01 – Oct. 99 Mar. 00 Oct. 00 Mar. 01 Sept. 01 Mostly porous, massive to crudely laminated micrite and spar deposits. Calcified mosses, cyanobacteria and algae.
Mean of each period Monthly mean ¼ mean of each period/no. of months Six-monthly mean calculated as monthly mean 6
Oct. 01 – Mar. 02
Apr. 02 – Sept. 02
Mean Mean Mean yearly Total of of cool accumulation accumulation warm periods (mm) (mm) Oct. 02 – periods (mm) Mar. 03 (mm)
1.09
1.52
5.71
21.5
5.35
3.69
10.46
6.67
7.17
2.60
9.77
32.99
0.36
1.12
11.81
21.87
16.72
9.45
13.95
221.28
14.16
23.15
11.02
30.26
1.6
1.31 0.26 1.92
6.50 1.08 6.48
0.18 0.03 0.20
7.51 1.25 7.50
3.43 0.57 3.42
6.31 1.05 6.30
2.30 0.38 2.28
6.99
1.96
8.96
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN
129
Cross-sections of tufa deposited on tablets: identification of six-monthly intervals
events; and (2) presence or absence of lamination and/or textural variations.
The identification of the different sedimentation intervals (i.e. the periodic thickness variations obtained with the MEM) was made by plotting on each tablet section the successive thickness data measured every six months with the MEM that corresponded to the five-point row that coincided with the cross-section cutting (Fig. 4). The recognition of the different sedimentation intervals presented a number of problems depending on: (1) thickness of the preserved deposits, in some cases closely related to the intensity and/or frequency of erosion
Sedimentary record from tablets in fast flowing water areas In general, the laminated record of tablets in fast flowing conditions had a rather good correlation with the several thickness data obtained seasonally with the MEM (Figs 4a & 5b, d, f). The smoothly domed morphology typical of this laminated facies caused some lack of coincidence between the five-point lines obtained with the MEM and the intervals recognized on the tablet sections.
Fig. 4. Cross-sections of (a) tablets 3 (fast flow conditions on the river bed) and (b) 7 (stepped waterfall) and thickness of each cool and warm period. For each case, white lines joint the thickness data measured seasonally with the MEM at the five points of the row closest to the tablet cross-section cutting. In (b), numbers help see the succesive seasonal measures; dashed lines represent erosion, deduced as negative values of thickness (e.g. lines 4 and 8) with respect to the previous measures (e.g. lines 3 and 7).
130
C. ARENAS ET AL.
Fig. 5. Views of tufa deposited on tablets 2, 3 and 5 placed in fast flowing water areas from August 1999– March 2003. Tablet 2 was installed at the end of October 2000. (a), (c) and (e): overhead views with indication of the cross-section cuttings shown in (b), (d) and (f). In (d) and (f) the colonization phase at the base corresponds to August– November 1999. The sediment thickness of the cool and warm periods shown in (b), (d) and (f) is that obtained with the MEM at the corresponding lines of measurement. Notice the lamination in the three of them.
There are two thick light intervals that correspond to the warm periods of 2000 and 2001 (Fig. 5d, f) and up to three thin dark intervals formed in cool periods. In general, the alternating light and dark intervals made identification of six-monthly intervals easier. However, the identification of other intervals on some sections was not possible without the thickness data obtained with the MEM, as those did not follow the thick light and thin dark pattern. This is the case of the record of April–September 2002 in tablet 3 (Fig. 5d), which corresponded to a thin slightly dark interval. Another case is the last six-month cool period
(October 2002–March 2003) that recorded higher-than-normal thickness in almost all studied tablets. This record appears as a thick light deposit that parallels the previous light warm period deposit without any variation in colour or other morphological features (e.g. in tablets 2 and 5, Fig. 5b, f).
Sedimentary record from tablets in standing to slow flowing water areas No distinct sedimentation intervals could be distinguised from the tablet cross-sections alone
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN
(Fig. 6b, d). The deposits preserved on the two tablets were thin and had a rather massive or very poor laminated character. The profiles of thickness data obtained seasonally with the MEM did not match any particular surface of the cross-sections. Indeed, some of these lines crossed among them indicating frequent erosion processes, although only in one case (cool period) negative values of accumulation were obtained with the MEM (Table 3). As a result, these deposits were not considered suitable for seasonal isotopic studies.
Sedimentary record from tablets in stepped waterfalls The identification of the cool and warm intervals was difficult in the cross-sections from tablets in stepped waterfalls (Figs 4b & 7b, d), in part because lamination was absent or very poor. The generally irregular, mostly domed bed morphology of this facies (Fig. 7a, c) favoured the profiles of measurement lines and the intervals seen in crosssections not to be fully coincident. The identification of the several periods was only possible thanks to the thickness data taken seasonally with the MEM, that revealed the existence of several erosion phases (Table 3, Figs 4b & 7b, d). In tablet 7, these erosion processes occurred almost through all cool and warm periods, although some record was preserved from each period. In contrast, in tablet 8, erosion during the last cool period caused the sediment record from October 2001–March 2003 to be totally eliminated (Table 3, Fig. 7d).
131
Sedimentary record from tablets in spray areas and caves The very thin thickness of sediment made impossible recognize distinct sedimentation intervals on cross-sections. These tablets were not used for seasonal isotopic studies.
Conclusion on sedimentary records In conclusion, sedimentation rates could be identified on cross-sections of tablets thanks to data obtained from seasonal measurements with the MEM, although with variable degree of accuracy mostly depending on subenvironments and erosive processes. In some cases, the lack of accurate or fine coincidence between thickness values from the MEM and thickness of distinct intervals in cross-sections was attributed to the fact that the profiles of five-point rows (four straight lines) did not reflect the real, complete morphology of each interval, and, hence, to the lateral variations in thickness caused by the geometrical features of each facies (e.g. smooth to conspicuous domes; e.g. see Fig. 5b). In other cases, however, the lack of coincidence (cf. Figs 4b & 7b) was greater and a clear indicator of erosion in some intervals. The MEM measurements revealed that the pattern of alternating thin dark and thick light intervals was not always present. The reversals in thickness data can be associated either with erosive processes or particular climate conditions. In the case of the last cool period
Fig. 6. Views of tufa deposited on tablets 1 and 6 placed in slow flowing water areas from August 1999–March 2003. (a) and (c): overhead views with indication of the cross-section cuttings shown in (b) and (d). The colonization phase at the base corresponds to August–November 1999. The sediment thickness of the cool and warm periods shown in (b) and (d) is that obtained with the MEM at the corresponding rows of measurement. Notice the lack of lamination and the loose appearance of the deposit.
132
C. ARENAS ET AL.
Fig. 7. Views of tufa deposited on tablets 7 and 8 placed in stepped waterfalls from August 1999 to March 2003. (a) and (c): overhead views with indication of the cross-section cuttings shown in (b) and (d). The colonization phase at the base corresponds to August–November 1999. The sediment thickness of the cool and warm periods shown in (b) and (d) is that obtained with the MEM at the corresponding lines of measurement. Several erosion phases can be seen (e.g. the cool period of 2000–2001).
(2002–2003) with thicker-than-normal sedimentation, the slightly higher mean temperature during the first part of the cool period favoured the high sedimentation rates recorded in all cases devoid of erosion. In addition, this explains the lack of a sharp textural change with respect to the preceeding warm period sediment. The most suitable sedimentary record from tablets for the seasonal study of textural features and stable isotope composition was that of dense, laminated deposits of fast flowing water areas (facies 1), and in some cases, of spongy, crudely laminated, mainly moss and alga deposits of stepped waterfalls (facies 3). On the other hand, it is important to note that despite the duration of the warm and cool periods chosen for monitoring may not exactly coincide with the duration of the natural processes/cycles that caused periodic variations in sedimentation rates, in most cases the thickness values of sixmonthly measurements with the MEM and of the intervals recognized on tablet sections are very close. So are measurements of warm intervals of seven months and of cool intervals of five months, as monitored during the first year of this study. It must then be concluded that the cycles of processes
that caused periodic changes in sedimentation characteristics (i.e. sedimentation rates) are close to six-month periods. Variations can be expected according to climate and hydrological conditions of some months at the beginning or end of the sixmonthly periods. For instance, when warmer-thannormal conditions extend through most of autumn.
Tufa facies The structural and textural characteristics of the sediments deposited in the several subenvironments allowed five main types of fluvial facies to be distinguished, based on field observations and the sedimentary record on tablets. The difficulty to identify seasonal (close to six-month) intervals in tablet cross-sections from slow flowing, spray and cave areas led the seasonal study of texture to be mostly focused on the facies formed in two distinct subenvironments: fast flowing water areas on the river bed; and stepped waterfalls (see section of Sedimentary subenvironments of the fluvial system). (1) Dense laminated tufa (Figs 5 & 8), that is, stromatolitic tufa, formed in fast flowing water areas (subenvironment a) without growth of
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN
133
Fig. 8. Photomicrographs of sediment of tablets in fast flowing water areas. (a), (b) and (c): from thin sections. The thickness of the cool and warm period sediment obtained with the MEM is indicated. (a) and (b) from tablet 3. (c) from tablet 5. In (c) arrows point to the dense cyanobacterial fabric of the upper half of the warm period of 2001. (d), (e), (f), (g) and (h): images from SEM. (d): Subperpendicular calcite microbial tubes arranged as laminae. The fine tubes at the base correspond to the warm period of 2002 and the overlying thick ones (arranged as bush-like groups) to the cool period of 2002–2003. An erosion phase separates them (arrowed surface). (e) and (f): Thick calcite microbial tubes arranged in subvertical, coalescent bush-like groups (e) and in subhorizontal groups as a massive network (f). Calcified filaments (fi) indicated by arrows. Both correspond to sediment of cool periods. Notice in (f) the overgrowth of the tubes. (g) and (h): Thin calcite microbial tubes arranged in subvertical, coalescent groups (g) and in subhorizontal groups in a massive network (h). Both correspond to sediment of warm periods.
mosses or other macrophytes. This facies is made of millimetre to 1–2 centimetre-thick laminae and layers, commonly marked by differences in colour and/or thickness, and ocassionally by variations in porosity, crystal size and biological components. Laminae range from roughly tabular to smoothly domed (Fig. 5b, d, f). It is formed of micrite and
spar calcite tube-like bodies, in which the empty center correspond to decayed microbes, presumably cyanobacteria, and less common of calcified microbial filaments and pennated diatoms; spar and less common micrite calcite is present among these components (Fig. 8a, b, c). The cyanobacterial tube-like bodies are either perpendicular to
134
C. ARENAS ET AL.
subperpendicular to the accumulation surface forming bush-like groups that may coalesce laterally (Fig. 8d, e, g), or non oriented constituting a massive network (Fig. 8f, h). Tube-like bodies are 15–60 mm in diameter (inner diameter: 4 –20 mm; coating thickness: 3– 20 mm) and 200 –250 mm long, rarely up to 500 mm. Microbial filaments of different sizes are among the tubes, but are not abundant. Intervals from the warm periods, as compared to those of cool ones, are commonly thicker, and the sediment is usually lighter in colour, more porous and has more abundant micrite calcite, although these features may change from site to site and depending on the intervals (Fig. 8a, b, c). Porosity is variable depending upon the type, relative proportion and arrangement of the biological components that are present. Empty cavities produced from decayed fauna (e.g. crustacean and aquatic insects) are more abundant in warm perids (e.g. Fig. 8b, c). The cyanobacterial tube-like bodies constitute either massive, open networks, or bush-like groups that in places form laminae. The latter may be more abundant and/or thicker in the warm than in the cool periods (Fig. 8c, upper half of warm period 2001) and, in some cases, present more dense fabrics in the warm than in the cool period cyanobacterial laminae (Fig. 8e, g). The thickness of the microbial tubes commonly is thinner in the warm (3– 5, rarely up to 7 mm; Fig. 8g, h) than in the cool intervals (3 to 20 mm) (Fig. 8e, f). Intervals from the cool periods locally show more dense fabrics and present thicker coatings of the tubes (Fig. 8a, b, f); however, in particular when these tubes form bush-like groups arranged in laminae, the fabric can be less dense than that of the warm period laminae (Fig. 8e). Ocassionally, palisades of large blade calcite crystals associated to
thick microbial tubes appear at the base of some cool intervals (Fig. 10a, cool period 1999–2000). (2) Dense to porous, commonly loose, massive tufa (Figs 6 & 9), formed in slow flowing, dammed and/or palustrine conditions (subenvironment b). The sediment consists of lime mud, coated phytoclasts and rare gastropods, unidentifiable tufa fragments, non-coated plant-remains and rare, up to 10 cm long oncolites. Lime mud and boundstones of plants are common in palustrine conditions. Calcified microbial filaments of several sizes, calcite microbial tube-like bodies and diatoms appear in areas of faster flow. Cyanobacterial filaments and diatoms are sometimes abundant and commonly have no calcite coatings. Micrite and spar calcite is present both coating the components and among them (Fig. 9b, c). The sediment of these tablets did not allow to differenciate textural characteristics between warm and cool periods, given the difficulty to identify the corresponding intervals, due to poor preservation and almost nule contrast between them. (3) Porous, spongy tufa, crudely laminated or without lamination (Figs 7 & 10), deposited in stepped waterfalls, small falls and barrages, and in spray and splash areas (subenvironments c and d). This facies consists of mosses, filamentous green algae, cianobacterial mats, diatoms and herbaceous plants, which are impregnated and/or coated by micrite and spar calcite, producing porous deposits mostly made of moss and filamentous algae boundstones, and associated cyanobacterial mats. These include trapped coated phytoclasts and other intraclasts (Fig. 10a) and hanging coated plants. In spray areas, the thin deposits formed there are mostly made of microbial mats. Moss and alga intervals produce soft, generally domed deposits, in which the plants are coated
Fig. 9. Photomicrographs of sediment of tablets in slow flowing water areas. (a): from a thin section (tablet 6). The identification of the cool and warm period sediment is only clear at the top. The rest of the intervals are indicated as approximate and cannot be dated. (b) and (c): images from SEM. Notice the massive appearance of calcite tubes (tu), diatoms (di) and calcite crystals. Diatoms at the base of (a) are oriented with their long axes vertical.
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN
135
Fig. 10. Photomicrographs of sediment of tablets in stepped waterfalls. (a) and (b): from thin sections of tablet 7. The thickness of the cool and warm period sediment obtained with the MEM is indicated. Notice in (a) and (b) the cyanobacterial-rich laminae (cy) in some warm perids and large crystals at the base of the cool period 1999–2000. Notice in (b) filamentous algae (al). (c), (d), (e), (f) and (g): images from SEM. (c) and (d): Sediment from the cool period 2002– 2003. Notice in (c) calcite crystals on a moss leaf and clumps of filamentous microbes and diatoms. (d): detail of one of the clumps. (e) and (f): Massive network of calcite microbial tubes, filaments and films from the warm period of 2000. (g): Coalescent bush-like bodies of thin calcite microbial tubes from the warm period of 2001.
as they grow from subperpendicular to subparallel to the accumulation surface depending on water flow intensity. Associated filamentous algae (macroscopic) are always parallel to the flow direction. Cyanobacterial mats consist of tube-like bodies, 20– 80 mm in diameter (inner diameter: 6–20 mm; coating thickness: 3 –20 mm, rarely 60 mm) and 200 –350 mm long (Fig. 10e, f). In some cases the tubes are arranged as hemi-domic to bush-like forms that may coalesce to form laminae (Fig. 10a, b, g). Calcified microbial filaments are of several sizes (Fig. 10d, e, f). Spar and less common micrite calcite is present among these components. In general, moss- and filamentous alga-bearing layers are thicker in the sediment from warm periods, although erosion can partially or
completely eliminate them. The microbial bush-like body-bearing intervals are usually preserved at the base of the warm intervals and are thicker than those of the cool periods (Fig. 10a). However, no clear textural differences between warm and cool periods have been found. Neither the shape nor size of calcite crystals present a clear pattern of seasonal variation. Coatings around in situ flora appear thicker in some cool intervals, but there are many exceptions. (4) Steep banded deposits developed in vertical waterfalls (subenvironment e), made of inclined moss-rich layers, with similar characteristics to facies 3, that include or alternate with curtains of hanging herbaceous plants, which become rapidly coated by calcite and constitute deposits with steep accumulation surfaces. The plant coatings
27.45 27.04 27.50 27.07 27.36 27.29 27.39
28.04 27.92 27.46 28.59 27.21 28.54 27.75
d C (‰ PDB) d18O (‰ PDB)
28.50 28.08 28.94 28.57 27.90 27.21 28.55 27.33 28.75 27.45 28.71 27.74 27.28 28.07 26.88 27.45 27.09 27.44 27.84
d C (‰ PDB) d O (‰ PDB) d C (‰ PDB)
(‰ PDB) (‰ SMOW)
d O
13 18
P-3
28.44 28.02 27.22 28.81 27.92
13
28.30 27.24 27.03 28.01 28.16
18
P-2
27.75 27.97 27.12 27.52
28.70 28.09 28.83 28.67 28.60 27.49 28.27 28.32 28.68 27.35 27.78 27.71 27.49
(‰ PDB) (‰ SMOW)
d O
d C (‰ PDB)
Oct. 02 –Mar. 03 Apr. 02 – Sept. 02 Oct. 01 –Mar. 02 Apr. 01 – Sept. 01 Nov. 00 – Mar. 01 Apr. 00 – Oct. 00 Nov. 99 – Mar. 00
The water d18O values showed little variation either among subenvironments or downstream (Table 4, Figs 11 & 12). Values ranged from – 9.08 to –7.90‰ SMOW in the cool periods and from –8.78 to –8.08‰ SMOW in the warm periods. Oscillation for each site was also small: standard deviations were lower than 5% of the mean value. Values of warm periods seem to be slightly heavier than those of cool ones: mean values were –8.65‰ for cool periods and –8.43‰ for warm periods. This trend is opposite to the one displayed by d18O composition of the sediment from tablets (Figs 11 & 12). These compositional changes can be related to the common seasonal variation of the isotopic composition of rainfall in continental sites, with lower values in cool periods than in warm ones (Leng & Marshall 2004). Higher temperatures can also have an influence through the loss of 16O due to an increase of evaporation. Water input in the River Piedra comes mainly from groundwater, as shown by the contrast between the marked seasonal character of rainfall and the almost constant water discharge. Thus, the water isotopic composition should only change due to the limited variability of the aquifer isotopic signature and to the small changes produced by the runoff supply, whose relevance is small with respect to the groundwater supply.
13
Oxygen isotopes of water
18
Mineralogical analyses were performed on the sediment sampled at the same time than water in areas adjacent to the tablets. X-ray diffraction analysis showed that the sediment invariably consisted of low-magnesium calcite (average Mg content ¼ 1.99% M), with quartz and ocasionally phyllosilicates and dolomite as trace components. No significant variations in mineralogy were observed through space or time.
P-7
Mineralogy
13
consist of massive or laminated micrite and spar calcite with calcite filamentous microbial remains. No tablets were installed in this subenvironment. (5) Dense, hard, laminated spar and micrite calcite deposits formed in caves (subenvironment f). Commonly the laminae are micrometre to millimetre-thick and include calcite microbial remains. The present surfaces of the caves include mats of mosses, liverworts, bacteria and cyanobacteria that are slowly coated by calcite. The tablet showed a very thin, discontinuous, greenish and pinkish laminae of microbial origin. Some bryophyte boundstones may form at the outer part of the caves associated with facies 4.
P-5
C. ARENAS ET AL.
Table 4. Stable isotope composition (d13C and d18O, ‰ PDB) of tufa sediment taken from tablets 2, 3, 5 and 7 from November 2000 – March 2003, and d18O (‰ SMOW) of water sampled at 5 sites in June and December from December 2000 – December 2002
136
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN
137
Fig. 11. Stable isotope composition (d13C and d18O, ‰ PDB) of cool and warm period sediment taken from tablets 2, 3, 5 and 7. Mean d18O (‰ SMOW) of water sampled in June and December from December 2000– December 2002 is shown.
Fig. 12. Stable isotope composition (d13C and d18O, ‰ PDB) of cool and warm period sediment taken from tablets in a stepped waterfall (a) and in fast flowing water areas (b), (c) and (d). d18O (‰ SMOW) values of water of the sites of tablets 7 and 3 are plotted in (a) and (c). (e) Mean d18O (‰ SMOW) of water sampling sites. (f) Water temperature expressed as the mean of the sites of the tablets for each period. Mean air temperature for the same periods. Mean of the calculated temperatures (for theoretical precipitation equilibrium) from sediment d18O of tablets 7, 2, 3 and 5. (g) The calculated temperature for the measured sediment d18O of tablet 3 is plotted along with the measured water temperature for this site.
138
C. ARENAS ET AL.
Carbon and oxygen stable isotopes of the tufa record The sedimentary record on tablets that allowed identification of warm and cool depositional intervals was sampled for analysis of stable isotope composition of sediment of every warm and cool period. This record corresponded to laminated deposits formed in fast flowing water areas and to mossand alga-rich, crudely laminated deposits formed in stepped waterfalls. Four tablets were selected for this purpose: three with laminated deposits and one with moss and alga deposits. The total number of intervals anlysed was 25: 14 ascribed to cool periods and 11 to warm periods, which correspond to those indicated by six-monthly thickness measurements with the MEM. The isotopic composition of the sediment analysed in the four tablets ranged, for the warm period samples from 28.07 to 27.07‰ PDB for d13C and from 28.81 to 27.92‰ PDB for d18O, and for the cool period samples from 28.30 to 26.88‰ PDB for d13C and from 28.32 to 27.21‰ PDB for d18O. Mean isotopic composition was d13C ¼ 27.45 ‰ and d18O ¼ 28.42‰ PDB for the warm periods and d13C ¼ 27.49 and d18O ¼ 27.67‰ PDB for the cool periods (Table 4; Fig. 11). Although the fields of warm and cool period samples overlap each other, the samples are partially set apart into two fields by d18O values. The difference in mean d18O values between warm and cool periods was 0.75‰. In contrast, d13C values do not show separate isotopic fields of the two periods. The difference in mean d13C values between warm and cool periods was 0.04‰. These d18O and d13C values are within the range of other recent tufas (Andrews et al. 1997; Ihlenfeld et al. 2003; Andrews & Brasier 2005). As in many other fluvial carbonate systems, in the study case correlation between d18O and d13C values was very poor, which accords with fractionation processes driven by different environmental parameters.
Oxygen isotopes The d18O composition of the sediment of the four tablets showed a rhythmic variation, with higher values in cool periods and lower values in warm periods (Fig. 12). Some exceptions to this pattern occurred in some intervals of tablet 7, and in tablets 2 and 5 the sediment of the warm period of 2002 had no isotopic variation or slightly higher values than in the following cool period. As for d18O variations among the three tablets corresponding to fast flowing conditions (tablets 2, 3 and 5; Fig. 12), these were small and did not show any spatial pattern for any of the warm and cool periods. In contrast, samples from tablet 7,
which corresponds to a stepped waterfall with moss development, displayed a more irregular seasonal d18O pattern and the isotopic differences between warm and cool periods were smaller than in the samples from fast flowing conditions (Fig. 12). Waterfalls can produce a strong degassing effect than can influence tufa d18O formed in waterfalls and in areas downstream of these (Andrews et al. 1997). In the present study, this process does not seem to have had a significant effect in the tufa produced downstream of the waterfalls. The periodic variations in d18O composition are thought to be mainly caused by the fractionation effects due to changes in temperature between the warm and cool periods, as interpreted in other fluvial systems that present little seasonal variation in d18O composition of water. Similar conclusions have been reported for other modern and subrecent seasonal laminated tufas in semiarid conditions (Chafetz et al. 1991; Matsuoka et al. 2001; O’Brien et al. 2006; Kano et al. 2007), with differences between sediment from warm and cool periods of about 1–1.5‰ PDB. These values are close to those obtained in this study (mean d18O difference is 0.75‰). The small d18O variability and lack of a spatial pattern among sediment from the three tablets in fast flow conditions, along with the little d18O variability of water, leads to infer that seasonal changes in d18O primarily reflect the water temperature variations. Exceptions seen in tablet 7 (warm period of 2000 and cool period of 2001– 2002) might be related to sampling problems, due to the common erosive processes that make difficult the identification of the sixmonthly record in the stepped waterfalls (Figs 4b and 7). These problems might also account for the less contrasted seasonal d18O differences in this subenvironment. On the other hand, CO2 degassing in waterfalls can locally alter the sediment d18O, both by varying the water d18O and the fractionation process. In contrast, exceptions in tablets 2 and 5, mainly the transition between the last two periods, where the seasonal trend was not so evident, cannot correspond to sampling errors and the explanation of the anomaly is not clear.
Carbon isotopes The d13C composition of the sediment of the four tablets did not show a persistent rhythmic pattern as that of d18O (Fig. 12). Only in tablet 3 lower values of warm periods alternated with higher values of cool periods. In the rest, the six-monthly d13C variation was irregular, so that both high and low values were recorded during cool and warm periods. In addition, no significant variations in d13C were seen between the two subenvironments
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN
considered. Neither were there regular variations through space among the samples of the three tablets in fast flow conditions. Although the rhythmic pattern is not the rule for d13C, some recent tufa systems have registered it. In the River Piedra, a few intervals recorded it (e.g. tablet 3). In the case studied by Chafetz et al. (1991) the d13C seasonal pattern presented higher values in the summer sediment than in winter sediment, which was attributed to the increase in 12C loss by degassing in summer. In contrast, the seasonal pattern of the example studied by Matsuoka et al. (2001) had low summer values, which were explained by local seasonal phenomena that were able to compensate the degassing effect. In general, the d13C behaviour is more complex because it can suffer the influence of several factors, both biotic and abiotic. The rather irregular variation of d13C composition through time and space for most of the studied intervals in such a short stretch of the River Piedra can be attributed to changing influence of the several variables affecting tufa formation on a local scale. Changes in d13C of the stream water that affect the entire fluvial system do not explain such irregular variation and must be disregarded. This is the case of seasonal changes in temperature, and of variations in water d13C caused by input of soilderived CO2, introduction of atmospheric CO2 into soil in arid conditions and variations in aquifer d13C composition or in aquifer contribution to the stream. Only processes that operate on a local scale can be invoked: for example, 13C enrichment caused by a local increase of mechanical CO2 degassing or local changes of biofacies (type of flora and its physiological activity). However, the expected slight increase in d13C composition due to photosynthesis is not always traceable (Shiraisi et al. 2008). In particular, in fast flow environments, the effect of this process commonly becomes masked by the strong influence of the mechanical degassing, which increases the 13C content in water. To sum up, the generally irregular d13C pattern through space and time might indicate that the several variables that affect the d13C composition, such as the fractionation processes mentioned above, had changing contributions through the studied warm and cool periods. For instance, it cannot be excluded that the sediment sampled at each interval in a particular subenvironment could present some slight differences in terms of biofacies.
Temperature calculations The theoretical water temperature at which the sediment of warm and cool periods was formed, in a theoretical equilibrium precipitation, was
139
calculated using the average measured water d18O values for each period, by means of the O’Brien et al. (2006) formula: T (8C) ¼ 15:310 4:478(d18 OcalcitePDB d18 OwaterSMOW ) þ 0:14[0:227 1:0412(d18 OcalcitePDB d18 OwaterSMOW )] In the case studied by these authors, the temperature range of tufa formation was similar to that of the Parque del Monasterio de Piedra. In the River Piedra, the calculated temperatures, which correspond to a theoretical equilibrium precipitation, show a high degree of agreement with the measured temperatures, displaying the same seasonal trend, although the calculated temperatures present a narrower range than the measured ones (Fig. 12). Similar differences have been reported by Lojen et al. (2004), although a larger disagreement between calculated and measured temperatures was found; this was linked to warmer temperatures with wider ranges. In the River Piedra, the measured water temperature range was 12.0 8C in the studied period. For the measured water d18O composition, which remained almost constant through the studied period, this temperature range corresponds to a variation in d18O of the carbonate sediment of 2.9‰ PDB, assuming an approximate change of 0.24‰ per 8C for the equilibrium precipitation (Andrews 2006). The average measured d18O range in the sediment from tablets is 1.6‰, which roughly corresponds to a temperature range of 6.7 8C. These data are closer to the range of calculated water temperatures (5.4 8C) than to the measured ones (12 8C). To some extent, this difference can be explained by the fact that the analysed sediment corresponded to a period of several months, while the measured water temperatures represented punctual conditions within the periods. In spite of that difference, the calculated water temperatures are quite similar to the actual ones. The cool period of 2002–2003 and the warm period of 2002 yield calculated water temperatures too high and too low, respectively, for the kind of periods; this is related to the above mentioned anomaly in d18O detected in tablets 2 and 5. As discussed above, high sedimentation rates in this cool period are linked to the warmer-than-normal mean temperature of November and December of 2002 (about 4 8C higher compared to the same months in 2001), so that most sediment of this cool interval probably was deposited under warmer conditions than the average of the period. However, the warmer temperature
140
C. ARENAS ET AL.
effects should be recorded by the d18O of all sediment samples of this period. Thus, other type of factors must have accounted for the anomaly of sediment d18O in the last year.
Conclusions The 3.5-year monitoring period of the River Piedra within the Monasterio de Piedra Natural Park showed an intense tufaceous sedimentation in this area of Mediterranean environmental conditions. The Park environment enables seasonal control of physical, hydrochemical and biological parameters, characterization of different subenvironments and facies, measurement of sedimentation rates and analysis of stable isotope composition of water and of the associated deposits. The different sedimentation intervals were identified in the tufa cross-sections on tablets by means of the thickness measurements taken every six months with the MEM. The following conclusions can be drawn from the six-monthly sedimentological and geochemical study of the sedimentary record on tablets and the analysis of the associated environmental parameters monitored in this natural laboratory from 1999–2003: (1) Five main types of fluvial facies were distinguished: (1) Dense stromatolitic tufa, formed in fast flowing water areas along the river bed; (2) Dense to porous, commonly loose, massive tufa, deposited in slow flowing water areas of the river; (3) Porous, spongy, coarsely laminated tufa, formed in stepped waterfalls with bryphytes, algae and cyanobacteria; (4) Steep banded deposits developed in vertical waterfalls; and (5) Dense, hard laminated deposits formed in caves. The seasonal study from tablet records was made in the first three facies. (2) The most suitable tufa record for seasonal studies was that of laminated facies formed in fast flowing water areas because this facies had the most continuous sedimentation record with the highest rates, and the sedimentological and geochemical features were seasonally more marked. Although stepped waterfalls also record high sedimentation rates, erosion events prevented a continuous sedimentary record. Facies from slow water flowing areas presented low sedimentation rates and frequent erosion events. (3) Erosion can locally alter the seasonal pattern of tufa sedimentation rates. So can slight variations of climatic conditions (e.g. of temperature) with respect to normal conditions during some seasons. (4) Sedimentological analysis revealed some differences between sediments from warm and cool periods, although no regular patterns exist, apart from general thickness variations. Neither
the shape or size of calcite crystals nor the density of the deposits presented definite seasonal variations. In some stromatolitic deposits, the sediment from warm periods, compared to that from cool periods, was thicker, lighter in colour, less dense, and the thickness of the tube-like microbial bodies was thinner. In spongy, moss- and alga-rich tufa, seasonal differences of these features were less clear. (5) The d18O composition of the sediment showed a rhythmic variation, with higher values in cool periods and lower values in warm periods. Mean d18O isotopic composition was 28.42‰ PDB for the warm periods and 27.67‰ PDB for the cool periods. Given the small variability of water d18O, these periodic variations in d18O composition were mainly caused by the fractionation effects linked to stream temperature changes between seasonal periods. Mean d13C isotopic composition was 27.45 PDB for the warm periods and 27.49 PDB for the cool periods. The d13C composition of the sediment did not show a clear rhythmic pattern. This is attributed to changing contributions of the fractionation processes on a local scale during the studied periods. (6) The calculated temperatures, for a theoretical equilibrium precipitation, accord with the measured temperatures and display the same seasonal trend. As a consequence: (a) tufa d18O values can be used to estimate relative paleotemperature variations in fluvial carbonate deposits; and (b) laminated tufa sediments can be useful as high-resolution records of temperature change. This work was funded by project REN3575/CLI of the Spanish Government and European Regional Development Fund, and forms part of the activities of the Continental Sedimentary Basin Analysis Group (Arago´n Government). We thank the Confederacio´n Hidrogra´fica del Ebro for providing data on water discharge, and the Stable Isotope Analysis Laboratory of the University of Salamanca and the department of SEM of the University of Zaragoza for scientific and technical assistance. We are grateful to the Monasterio de Piedra Park management and staff who allowed and facilitated the fieldwork. The editors of the volume and two reviewers are thanked for their suggestions and meticulous revision.
References A NDREWS , J. E. 2006. Paleoclimatic records from stable isotopes in riverine tufas: Synthesis and review. Earth Science Review, 75, 85–104. A NDREWS , J. E. & B RASIER , A. T. 2005. Seasonal records of climatic change in annually laminated tufas: short review and future prospects. Journal of Quaternary Science, 20, 411– 421. A NDREWS , J. E., R IDING , R. & D ENNIS , P. F. 1997. The stable isotope record of environmental and climatic signals in modern terrestrial microbial carbonates
SEASONAL RECORD IN RECENT FLUVIAL TUFAS, SPAIN from Europe. Palaeogeography, Palaeoclimatology, Palaeoecology, 129, 171 –189. A RENAS , C., S ANCHO , C., O SA´ CAR , M. C., V A´ ZQUEZ , M. & A UQUE´ , L. F. 2004. La sedimentacio´n toba´cea actual en el Parque del Monasterio de Piedra (Zaragoza). VI Congreso Geolo´gico de Espan˜a; tomo 2. Geotemas, 6, 27–30. B ALL , J. W. & N ORDSTROM , D. K. 2001. User’s manual for WATEQ4F, with revised thermodynamic data base and test cases for calculating speciation of major, trace, and redox elements in natural waters. U.S. Geological Survey, Open File Report 91– 183. C HAFETZ , H. S., U TECH , N. M. & F ITZMAURICE , S. P. 1991. Differences in the d18O and d13C signatures of seasonal laminae comprising travertine stromatolites. Journal of Sedimentary Petrology, 61, 1015– 1028. C HEN , J., Z HANG , D. D., S HIJIE , W., T ANGFU , X. & R ONGGUI , H. 2004. Factors controlling tufa deposition in natural waters at waterfall sites. Sedimentary Geology, 166, 353–366. D ANDURAND , J.-L., G OUT , R., H OEFS , J., M ENSCHEL , G., S CHOTT , J. & U SDOWSKI , E. 1982. Kinetically controlled variations of major components and carbon and oxygen isotopes in a calcite-precipitating spring. Chemical Geology, 36, 299–315. D RYSDALE , R. N., T AYLOR , M. P. & I HLENFELD , C. 2002. Factors controlling the chemical evolution of travertine-depositing rivers of the Barkly karst, Northern Australia. Hydrological Processes, 16, 2941–2962. D RYSDALE , R. & G ILLIESON , D. 1997. Micro-erosion meter measurements of travertine deposition rates; a case study from Louie Creek, Northwest Queensland, Australia. Earth Surface Processes and Landforms, 22, 1037– 1051. E MEIS , K. C., R ICHNOW , H. H. & K EMPE , S. 1987. Travertine formation in Plitvice National Park, Yugoslavia: chemical versus biological control. Journal of Sedimentary Petrology, 34, 595–609. E PSTEIN , S. & M AYEDA , T. 1953. Variation of O18 content of waters from natural sources. Geochimica et Cosmochimica Acta, 4, 213–224 H ERMAN , J. S. & L ORAH , M. M. 1987. CO2 outgassing and calcite precipitation in falling spring creek, virginia, U.S.A. Chemical Geology, 62, 251–262. I HLENFELD , C., N ORMAN , M. D., G AGAN , M. K., D RYSDALE , R. N., M AAS , R. & W EBB , J. 2003. Climatic significance of seasonal trace element and stable isotope variations in a modern freshwater tufa. Geochimica et Cosmochimica Acta, 67, 2341– 2357. J ACOBSON , R. L. & U SDOWSKI , E. 1975. Geochemical controls on a calcite precipitating spring. Contributions to Mineralogy and Petrology, 51, 65– 74. K ANO , A., H AGIWARA , R., K AWAI , T., H ORI , M. & M ATSUOKA , J. 2007. Climatic conditions and hydrological change recorded in a high-resolution stableisotope profile of a Recent laminated tufa on a subtropical island, southern Japan. Journal of Sedimentary Research, 77, 59–67. K ANO , A., M ATSUOKA , J., K OJO , T. & F UJII , H. 2003. Origin of annual laminations in tufa deposits, southwest Japan. Palaeogeography, Palaeoclimatology, Palaeoecology, 191, 243 –262.
141
K ANO , A., K AWAI , T., M ATSUOKA , J. & I HARA , T. 2004. High-resolution records of rainfall events from clay bands in tufa. Geology, 32, 793– 796. K AWAI , T., K ANO , A., M ATSUOKA , J. & I HARA , T. 2006. Seasonal variation in water chemistry and depositional processes in a tufa-bearing stream in SW Japan, based on 5 years of monthly observations. Chemical Geology, 232, 33–53. L ENG , M. J. & M ARSHALL , J. D. 2004. Palaeoclimate interpretation of stable isotope data from lake sediment archives. Quaternary Science Reviews, 23, 811–831. L IU , Z., S VENSSON , U., D REYBRODT , W., D AOXIAN , Y. & B UHMANN , D. 1995. Hydrodynamic control of inorganic calcite precipitation in Huanglong Ravine, China; field measurements and theoretical prediction of deposition rates. Geochimica et Cosmochimica Acta, 59, 3087–3097. L OJEN , S., D OLENEC , T., V OKAL , B., C UKROV , N., M IHELCIC , G. & P APESCH , W. 2004. C and O stable isotope variability in recent freshwater carbonates (River Krka, Croatia). Sedimentology, 51, 361–375. L ORAH , M. M. & H ERMAN , J. S. 1988. The chemical evolution of a travertine-depositing stream; geochemical processes and mass transfer reactions. Water Resources Research, 24, 1541– 1552. L U , G., Z HENG , C., D ONAHOE , R. J. & L YONS , W. B. 2000. Controlling processes in a CaCO3 precipitating stream in Huanglong natural scenic district, Sichuan, China. Journal of Hydrology, 230, 34– 54. M ALUSA , J., O VERBY , S. T. & P ARNELL , R. A. 2003. Potential for travertine formation: Fossil Creek, Arizona. Applied Geochemistry, 18, 1081–1093. M ATSUOKA , J., K ANO , A., O BA , T., W ATANABE , T., S AKAI , S. & S ETO , K. 2001. Seasonal variation of stable isotopic compositions recorded in a laminated tufa, SW Japan. Earth Planetary Science Letters, 192, 31–44. M C C REA , J. M. 1950. On the isotope chemistry of carbonates and a paleotemperature scale. Journal of Chemical Physics, 18, 849–587. M ERZ -P REI b, M. & R IDING , R. 1999. Cyanobacterial tufa calcification in two freshwater streams: ambient environment, chemical thresholds and biological processes. Sedimentary Geology, 126, 103–124. O’B RIEN , G. R., K AUFMAN , D., S HARP , W. D., A TUDOREI , V., P ARNELL , R. A. & C ROSSEY , L. J. 2006. Oxygen isotope composition of annually banded modern and mid-Holocene travertine and evidence of paleomonsoon, Grand Canyon, Arizona, USA. Quaternary Research, 65, 366– 379. O SA´ CAR , M. C., A UQUE´ , L., S ANCHO , C., A RENAS , C. & V A´ ZQUEZ , M. 2003. Dina´mica de la precipitacio´n de calcita en las formaciones toba´ceas del Monasterio de Piedra (Zaragoza). Boletı´n de la Sociedad Espan˜ola de Mineralogı´a, 26A, 81–82. P ARKHURST , D. L. & A PPELO , C. A. J. 1999. User’s Guide to PHREEQC (Version 2), a computer program for speciation, batch-reaction, onedimensional transport, and inverse geochemical calculations. Water Resources Research Investigations, Report 99– 4259. P EDLEY , H. M. 1990. Classification and environmental models of cool freshwater tufas. Sedimentary Geology, 68, 143–154.
142
C. ARENAS ET AL.
P EDLEY , H. M. 2009. Tufas and travertines of the Mediterranean region: a testing ground for freshwater carbonate concepts and developments. Sedimentology, 56, 221– 246. P EDLEY , H. M., R OGERSON , M. & M IDDLETON , R. 2009. Freshwater calcite precipitates from in vitro mesocosm flume experiments: a case for biomediation of tufas. Sedimentology, 56, 511– 527. R OGERSON , M., P EDLEY , H. M., W ADHAWAN , J. D. & M IDDLETON , R. 2008. New insights into biological influence on the geochemistry of freshwater carbonate deposits. Geochimica et Cosmochimica Acta, 72, 4976–4987. S HIRAISHI , F., R EIMER , A., B ISSET , A., DE B EER , D. & A RP , G. 2008. Microbial effects on biofilm calcification, ambient water chemistry and stable isotope
records in a highly supersaturated setting (Westerho¨fer Bach, Germany). Palaeogeography, Palaeoclimatology, Palaeoecology, 262, 91–106. V A´ ZQUEZ -U RBEZ , M., A RENAS , C., S ANCHO , C., O SA´ CAR , C., A UQUE´ , L. & P ARDO , G. in press. Factors controlling present-day tufa dynamics in the Monasterio de Piedra Natural Park (Iberian Range, Spain): depositional environmental settings, sedimentation rates and hydrochemistry. International Journal of Earth Sciences, doi: 10.1007/s00531-009-0444-2. V A´ ZQUEZ -U RBEZ , M., O SA´ CAR , C., A RENAS , C., S ANCHO , C., A UQUE´ , L. & P ARDO , G. 2005. Variabilidad de la sen˜al isoto´pica (d13C y d18O) del sistema toba´ceo actual del Parque del Monasterio de Piedra (provincia de Zaragoza). Geo-Temas, 8, 119–123.
Factors controlling growth of modern tufa: results of a field experiment ´ SKI MICHAŁ GRADZIN Institute of Geological Sciences, Jagiellonian University, Oleandry 2a, 30-063 Krako´w, Poland (e-mail:
[email protected]) Abstract: A field experiment was performed at four sites in Slovakia and Poland in order to identify factors that influence the growth rate and textures of modern tufas. Two pairs of tablets were placed at each study point, every pair consisting of a limestone tablet and a copper tablet. One pair at each site was changed every three or four months, while the second pair was left for approximately 14 months. Each tablet was weighed before placing and after removal to find the amount of the tufa growth. Tufa growth rate was found to depend on SIcalc. of parent water, though deposition of tufa on limestone tablets was substantially faster than on copper tablets. This result indicates that micro-organisms are essential for more efficient growth of tufa. Tufa growth rate was higher in fast-flowing water than in nearby sluggish flow settings. Fast-growing tufa has crystalline texture or consists of highly encrusted algal filaments. The latter texture is due to faster growth of micro-organisms, forced by rapid crystallization of calcite on their cells. The slow-growing tufa exhibits mainly micritic textures. Clotted micrite with numerous diatoms forms mostly in winter, while encrusted algal filaments are typical for spring and summer growth.
Freshwater carbonate deposits form in headwater streams worldwide (Ford & Pedley 1996). They originate under various conditions, from different types of waters and therefore may form bodies of various thicknesses, spreading over the distances from a few metres to several kilometres. During their growth these deposits build evolved threedimensional structures such as dams, cascades, and they display various internal structures. Two terms are used for such deposits: calcareous tufa and travertine (Ford & Pedley 1996; Pedley 2009). The former term refers to deposits containing plant and algal moulds or imprints. Conversely, travertine hardly ever exhibits plant imprints and seldom displays recognizable biological texture. Both terms have strong genetic implications. Tufas commonly form in karst areas and are fed with water saturated with soil CO2, while travertines originate from deep circulation water, usually of elevated temperature and highly charged with CO2 of deep origin (Chafetz & Folk 1984). Other terms used for tufas and travertines are meteogene and thermogene travertines, respectively (Pentecost 1995, 2005). Many authors, including American and Spanish ones, call all spring-related carbonates ‘travertines’ regardless of their characteristics and origin. Here, the terms tufa and travertine will be used to denote lithology of deposits, regardless of their origin and temperature of parent fluids. The growth of tufa and travertine is one of a few geological phenomena which can easily be noted
and observed by an ordinary man. Therefore, it is no wonder that it has attracted men’s interest for ages, as is recorded in European fine arts and literature as early as in the Middle Ages. For example, Dante in Divine Comedy at least twice mentioned a travertine deposited in streams. Self-building ridges were illustrated in a 15th century altar scene of Christ’s baptism in the St. Jacob’s church in Usterling, Bavaria. The actively growing ridge may still be seen nearby. Many researchers tried to assess the rate of tufa growth. Average growth rates of Quaternary (mostly Holocene) tufa were estimated using isotope ratios, mainly radiocarbon. The rate of 0.49 mm a21 resulted from dating of Flandrian tufa from Holywell Coombe (Kent) made by Kerney et al. (1980). Pazdur et al. (1988) found the growth rate of Holocene tufa in the Krako´w Upland to depend strongly on facies. In stromatolitic facies it reached 10 mm a21 while in marly facies only 1 –2.5 mm a21. The study by Heimann & Sass (1989) from north Israel yielded value of 0.32 mm a21. Preece & Day (1994) obtained the average rate of growth of the Holocene tufa in Oxfordshire between 0.13 and 0.54 mm a21, while Meyrick & Preece (2001) calculated a rate ranging from 0.27 to 0.7 mm a21 for tufas in English Midlands. A higher rate, varying between 1.5 and 2.5 mm a21, was recorded by Pedley et al. (1996) from lake tufa in southern Spain. Limondin-Lozouet & Preece (2004) recorded the growth of tufa in
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 143–191. DOI: 10.1144/SP336.8 0305-8719/10/$15.00 # The Geological Society of London 2010.
144
´ SKI M. GRADZIN
Normandy at a rate between 1.44 and 1.65 mm a21. All these data are mean values. The studied profiles may include surfaces of omission or erosion, so the maximum growth rate may have been substantially higher. Moreover, these calculations carry the error of dating methods. Therefore, these data are difficult to be directly compared with the growth rate of modern tufa. Many subfossil tufa show regular internal lamination. An assumption that it is annual in origin along with direct measuring of lamina thickness provide tools for assessing the growth rate. Irion & Mu¨ller (1968) claimed that cyanobacterial tufa grows with the rate between 0.6 and 7 mm a21. The mean values obtained by Schnitzer (1974) and Statham (1977) fall within the lower limit of the above range. Ohle (1937), using an intricate calculation, achieved similar values for tufa in north Germany. A precise analysis of annually laminated tufa led to the conclusion that the mean rate of tufa growth in Japan mainland is 3.8 mm, while in subtropical Miyako Island (the Ryukyu Islands) it varies between 4.3 and 9.8 mm. However, the latter value is influenced by human-induced changes of the hydrologic regime (Kano et al. 2007). Tubes of Chironomid larvae form every spring in tufa sections. They allowed Wallner (1934a) to estimate the annual growth of tufa composed of calcified Vaucheria at 7–14 mm. Ma¨gdefrau (1956) mentioned that the recent moss tufa near Aringsdorf grew with the rate of 1–2 cm per year. The extremely high growth rate reaching up to 140 mm a21 was reported from layered moss tufa in Spain by Weijermars et al. (1986). Direct measuring of thickness of growing tufa seems to be a more reliable method. Drysdale & Gillieson (1997) applied micro-erosion metre to determine tufa growth rate in a tropical karst area of north Australia. They obtained the maximum value of 36.22 mm a21 and the mean value between 0.19 and 4.38 depending on environmental conditions. Merz-Preiß & Riding (1999) determined maximum annual rate of tufa growth at 2.21 mm in the Fleisbrunenbach in the Schwa¨bische Alb. The neighbouring area of the Frankische Alb was studied by Arp et al. (2001), who stated that various substrates immersed in a hard-water creek were covered with a 1.5 mm tufa layer in a 10-month period, which after recalculation gives 1.8 mm layer per year. Kano et al. (2003) found a growth rate between 1.5– 4.0 mm a21 in a Japanese locality studied throughout 3 years. The rate of tufa growth expressed as increase in thickness in time cannot be directly recalculated to mass growth in time because of extreme variations in porosity of tufas, from below 10% up to more than 70% (Pentecost 2005, 31). Thus, mass increment in time on known surface area has been
measured by some authors. Bono et al. (2001) came out with a tufa (travertine in their terminology) deposition rate up to 1.1 1027 mmol/cm2/s in a stream fed by a karst spring in central Italy. Zhang et al. (2001) in an artificial pool constantly flushed with stream water obtained maximum daily tufa growth rate as high as 0.0728 mg/cm2. However, this method has been applied mostly to travertines. Liu et al. (1995) and Lu et al. (2000) studied depositional rate of travertine in Hunaglong ravine and they obtained values between less than 8.647 1024 and 0.8647 mg/cm2/day, and between 0.15 and 5.4 mg/cm2/day, respectively. Recently, Pentecost & Coletta (2007) have applied a similar method to La Zittelle spring near Viterbo in Italy. Their results show travertine precipitation rate between 13.4 and 30.9 mg/cm2/day. Micro-organisms – various bacteria and algae – are abundant in environments of tufa growth (e.g. Pia 1934; Symoens 1957; Golubic´ 1967; Freytet & Verrecchia 1998; Pentecost 2005). Further on in this paper the term ‘algae’ is used collectively for cyanobacteria and algae unless it concerns a particular taxon or a structure connected with a particular group of organisms. The term ‘micro-organisms’ is used for bacteria and algae, including the socalled macro-algae, as for example Vaucheria. Micro-organisms are commonly encrusted with calcium carbonate; this has triggered a long-lasting discussion on their role in the origin of tufa. Some authors do not regard micro-organisms as an active agent in carbonate deposition and point out that degassing of CO2, leading to supersaturation of parent water, exerts a fundamental control on it (Jacobson & Usdowski 1975; Dandurand et al. 1982; Michaelis et al. 1985; Herman & Lorah 1987; Dreybrodt et al. 1992; Liu et al. 1995). Others imply that micro-organisms colonizing hard-water channels provide a suitable substrate for nucleation of carbonate minerals and hence passively stimulate the growth of tufa and also influence its texture (Merz-Preiß & Riding 1999; Janssen et al. 1999; Pentecost & Whitton 2000; Pentecost 2005). Some authors consider micro-organisms as important agents which can actively stimulate the tufa growth, at least in some chemical and hydrodynamic conditions (Pedley 1992, 1994, 2000). This view has been recently confirmed by the microelectrode measurements and mass balance calculations by Shiraishi et al. (2008) and by laboratory experiments (Pedley et al. 2009). These results clearly show that micro-organisms and microbial biofilms actively contribute to the tufa growth. Tufa is important as a palaeoenvironmental archive that records hydrological and climatic changes (Andrews & Brasier 2005; Andrews 2006). A huge body of literature concerning the seasonality of tufa growth has been published in the last
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
twenty years. Differences in textures deposited during cold and warm seasons were recognized (e.g. Szulc & Smyk 1994; Arp et al. 2001; Kano et al. 2003). However, there was no attempt to elucidate the difference of tufa growth rate or tufa mass increment along with changes of its texture. Since tufas are fed by water saturated with CO2 of soil origin, their growth depends strongly on the development of a vegetation cover, which in turn is related to climatic condition. Therefore, Hennig et al. (1983) concluded that abundance of tufa in the geological column is an indicator of periods of relatively warm and humid climate. Pentecost (1995) claimed that the majority of tufas in Europe occur in the areas characterized by mean annual temperatures between 5–15 8C and mean annual rainfall exceeding 500 mm. No comparison of the tufa growth rates in the same observational period in places located in various climatic condition has been undertaken. The purpose of this study was to measure the growth rate of modern tufa in some selected sites in Central Europe. The sites are spread across the Central Carpathians and one site lies in the northern foreland of the Carpathians. They differ in climatic conditions, type and origin of feeding water, modern hydrology as well as in type of vegetation around them. Main chemical parameters of parent water
Fig. 1. Location of the studied sites.
145
were studied to examine the conditions of tufa growth. Petrography of the actively growing tufa was also examined. The data were used to assess seasonality of tufa growth and to compare growth rates, textures, and environmental factors governing their formation.
Study sites Four sites of active tufa deposition have been chosen for this study. They are scattered across the Carpathian range, which is the main topographic barrier in Central Europe (Fig. 1), acting as the continental divide between the Baltic and Mediterranean seas and as an important climatic barrier. The northernmost site – Karwo´w – is located in the Middle Polish Uplands, approximately 250 km north of the main ridge of the Central Carpathians. Two sites are situated in the main part of the Central Carpathians: Za´zrı´va´ in a narrow mountain valley and Lu´cˇky at the border of a wide intramontane basin. The fourth site – Ha´j – lies on the southfacing slopes of the Central Carpathian chain.
Karwo´w The Karwo´w site is located on the Sandomierz Upland c. 25 km NWW of the city of Sandomierz
fluvial tufa (waterfalls and dam)
*cascade point; †upper waterfall point; ‡lower waterfall point; §dam point.
8.6 8C
19824 11.7
00
208500 53.200 208500 51.400 208500 51.600 488380 29.800 488380 11.800 488380 09.600 Ha´j
375 m† 325 m‡ 315 m§
630 mm
mixed forest, beech dominated deciduous forest, oak and beech dominated, xerothermic plants 800 mm 5.8 8C 49807 47.16
Lu´cˇky
600 m
mixed forest 1000 mm 6.3 8C
00 0
00
198090 14.2200 498140 33.90 Za´zriva´
560 m
mixed infiltration and deep circulation infiltration, karst circulation
fluvial tufa (dams and cascade) perched tufa (tufa fan) fluvial tufa (waterfall) infiltration, high residence time infiltration agricultural area 550 – 600 mm 7.0 8C 218280 26.000 508460 05.500 Karwo´w
240 m*
Type of feeding water Type of vegetation Mean annual rainfall Mean annual temperature Altitude Longitude
The Za´zriva´ site is situated in northern Slovakia, c. 12 km NWW of the city of Dolny´ Kubı´n in the Mala´ Fatra Mountains, an alpine mountain chain (Fig. 1, Table 1). The Mala´ Fatra is built of crystalline core, its autochthonous cover and a sequence of nappes (Hasˇko & Pola´k 1978). The autochthonous cover and the nappes are built of Mesozoic rocks which include many carbonate units. The neighbouring peaks reach altitudes between 930 – 1600 m. Mountain meadows occupy the upper parts of their slopes, while the lower parts are forested with mixed coniferous and deciduous trees. The object of the present study was a perched tufa fan attached to the western slope of deeply incised Za´zrivka stream valley (Figs 2b & 3c).
Latitude
Za´zriva´
Study site
and 1 km south of the small village of Karwo´w in the Opato´wka river valley (Fig. 1; Table 1). A small stream emerges there in a spring, called by local people the Blessed Kadłubek spring. The area around is agriculturally used, with small groves in close surrounding. The area is covered by loess, up to a few metres thick. Older tufa, probably of Holocene age, is locally present as debris spread over the fields. The spring issues at the contact between folded Cambrian siliciclastics and horizontally lying Miocene detrital limestone which builds a local hill (Dowgiałło 1974). The origin of tufa in this area was ascribed to shallow water circulation and dissolution of Miocene limestones (Samsonowicz 1929) or Middle Devonian carbonates that occur in close vicinity (Dowgiałło 1974). Szulc (1984) in his unpublished paper suggests that the water may also ascend from greater depth along the S´wie˛tokrzyski Fault that separates Devonian carbonates from Cambrian siliciclastics. The opinion on deep circulation origin of the water is confirmed by low tritium concentration, which suggests great residence time of the discussed water (M. Dulin´ski, pers. comm., 2003). The spring expels between 0.8 and 1.0 l/s (Ga˛gol & Urban 2000; M. Gradzin´ski, unpublished data, 2003). Tufa deposits were studied along the uppermost 120 m section of the stream which first meanders over a gentle slope, then abruptly turns to the right and flows swiftly down toward a small gorge (Fig. 2a). Here the first growing tufa appears in form of small dams across the stream (Fig. 3a). The stream then flows down to the junction with a side ravine cut in loess, dry throughout almost all the year, except thaw periods and heavy rainfalls. At the junction of the stream and the ravine a small fan-shaped tufa cascade forms, c. 2 m high (Fig. 3b). The cascade and the dams are deeply shaded in summer by a leaf canopy.
Type of tufa
´ SKI M. GRADZIN
Table 1. Location and main characteristics of sites studied
146
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
147
Fig. 2. General topography of the sites studied, the points where tablets were installed and station of water sampling are indicated (a) Karwo´w; (b) Za´zriva´; (c) Lu´cˇky; (d) Ha´j.
The tufa fan is 12 m high and its lateral extent at the base exceeds 30 m; the average dip of the fan is 358. The fan is built of growing tufa that partly cements older slope scree. The fan is fed by numerous small springs and seeps from which water flows down the fan in small rivulets. The water flows also as a thin film through the moss cushions that cover the fan. The tufa fan is densely overgown by mosses and algae. The algal flora is dominated by desmid Oocardium stratum. It also comprises numerous
diatoms, cyanobacteria (for instance Rivularia haematites), and green algae, including Cladophorophyceae (T. Mrozin´ska, pers. comm., 2003). Locally patches of bushes and small willows occur on the tufa fan. The fan is shaded by neighbouring coniferous trees. The tufa fan in Za´zriva´ was not included in the lists of Slovak tufas and travertines (Kovanda 1971). It was recognized during the geological fieldwork by the geologists from Department of Geology and Palaeontology, Comenius University in Bratislava.
148
´ SKI M. GRADZIN
Fig. 3. General view of the sites studied; the tablet positions are arrowed (a) Karwo´w dam, in the dam crest the tablets are visible, photo taken in August 2002 just after tablet installation; (b) Karwo´w cascade; (c) Za´zriva´; (d) Lu´cˇky; (e) Ha´j upper waterfall in March 2003; (f) Ha´j dam in June 2003, (g) dam in October 2003 during low-flow conditions.
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
Lu´cˇky The Lu´cˇky site is located in the village of Lu´cˇky in Slovakia. The village lies at the outlet of a valley incised into the Chocˇ Mountains to the Liptov Basin, 14 km west of Liptovsky´ Mikula´sˇ (Fig. 1, Table 1). The mountain creek flows along the valley and drains an area with maximum elevation of 1611 m. In the centre of the village the stream drops in a picturesque 15 m high waterfall, where calcium carbonate is being vigorously precipitated (Nemejc 1928; Kovanda 1971). The studied site is at the waterfall (Figs 2c & 3d). The Chocˇ Mountains are built mainly of folded and thrusted Triassic carbonates of the Chocˇ Nappe. They adjoin the neighbouring Liptov Basin along a master fault zone which has approximately longitudinal orientation (Gross 1980). The basin is developed in less resistant clastis, mainly of flysch type, called the Central Carpathian Palaeogene that occupy a tectonic depression. Their thickness reaches 1500 m. The Central Carpathian Palaeogene is underlain by various Mesozoic deposits, mainly carbonates, of the Chocˇ and Krı´zˇna nappes. They crop out to the south of the basin and build the extensive Low Tatra Mountains range there. The Central Carpathian Palaeogene has very low permeability, and hence it acts as a confining bed. The artesian water migrates up along the faults that cut the flysch rocks, mainly along the master fault zone which separates the Liptov Basin from the Chocˇ Mountains (Franko & Hanzel 1980). It has been suggest that the recharge zone of the artesian basin is located in the Low Tatra Mountains (Hynie 1963). However, the basin may be also fed by water from the Chocˇ Mountains, which is suggested by the open hydrological balance of this carbonate massif (see Franko & Hanzel 1980). The water in the stream in the studied waterfall is a mixture of two components. One represents mountain stream water draining the Chocˇ Mountains. The stream is additionally fed by deeply-circulating water ascending along the faults bordering the Liptov Basin, which constitutes the second component. Both components mix near a spa, approximately 300 –500 m upstream from the waterfall. Nowadays only a part of deep-circulating water reaches a stream, because it is used in a spa which collects water from natural springs and shallow boreholes. The reconnaissance measurements done in October 2008 show that the discharge of the stream feeding the waterfall is around 60 l/s. The waterfall occurs at the upstream (northern) end of a southward widening ravine sharply incised into older tufa. Neˇmejc (1928) attributed Holocene, mainly Atlantic age to the tufa, basing on floral remains. This has been confirmed by exploratory 14C dating in the course of the present
149
study (M. Gradzin´ski, unpublished data, 2003). The waterfall has a complex morphology. Its main wall is 12 m high with a central projecting pillar bordered by two depressions where water cascades down. The wall is covered with cyanobacteria and algae, moss curtains and some liverworts and grasses overhanging from the waterfall head. Two small primary caves are present at the base of the pillar. They formed by progradation of the waterfall wall. The occurrence of presently prograding tufa waterfall over a cliff built of eroded Holocene tufas proves that tufa growth was interrupted during the Holocene, and the stream eroded the ravine then.
Ha´j The Ha´j site lies in southern Slovakia in the eastern part of the Slovak Karst area, about 30 km west of Kosˇice (Fig. 1, Table 1). The upper section of a fast-flowing stream was examined (Fig. 2d). The section is 900 m long with vertical drop of c. 80 m. The surrounding plateaus rise to altitudes 650–800 m and the narrow valley is incised c. 150–200 m deep. The plateaus are built of Triassic limestone of Wetterstein type, belonging to the Silica Nappe (Mello 1996). The limestone crops out in numerous cliffs and crags on the valley slopes. The valley is partly filled with inactive tufa over 25 m thick. Lozˇek (1957) determined its age as Holocene on the base of malacofauna. An exploratory 14C dating done in the course of the present study has confirmed the above view (M. Gradzin´ski, unpublished data, 2003). The stream flowing along the valley is incised into older tufa and forms a narrow gorge up to 10 m deep, which is shaded owing to its topography and a dense canopy of tree and bush leaves during the vegetation period. The stream is densely colonized by cyanobcteria and algae and in some places also by liverworts and mosses. The stream forms four waterfalls whose distribution clearly mirrors the location of older Holocene tufa dams. This is why Lozˇek (1957) wrote that the stream now only erodes Holocene tufa. However, a reconnaissance study clearly showed that calcium carbonate actively precipitates along the stream course. The stream is fed by springs draining a karst aquifer. The reconnaissance measurements done in October 2008 show that the stream carries between 6 and 15 litres of water per second, depending on the studied place. The springs are active throughout almost all the year, however their discharges vary in time. Two springs are situated above the studied section while the third one within it (Fig. 2d). Three points were selected for the study (Figs 2d & 3e– g).
150
´ SKI M. GRADZIN
Methods Measurements of water chemistry In the course of the experiment, water samples were collected between April 2002 and December 2003 during the 5–8 campaigns, depending on the site. A detailed hydrochemical study is beyond the scope of this paper, hence only selected data are presented. They concern springs that feed the tufa precipitating streams and the stations located directly above the place where tablets were exposed. At the Lu´cˇky site only chemical composition of water directly feeding the waterfall was taken into consideration. The specific electrical conductance (SEC) and pH were measured in the field using calibrated equipment: a WTW product. Total alkalinity (as bicarbonate HCO3) was determined using 0.05 molar HCl acid by Gran titration also in the field immediately after sampling of water. Temperature was read with a mercury thermometer. The resolution was 0.02%, 0.01 and 0.1 8C for SEC, pH and temperature, respectively. Chloride (Cl) contents were determined by the method of Mohr, using 0.01 molar AgNO3 within two days after sampling. Concentration of other components – Ca, Mg, Na, K, Li, SO4, Sr, Ba – was determined by inductively coupled plasma-atomic emission spectroscopy (ICP-AES) using a Perkin-Elmer product. Total sulphur was calculated as sulphates. All laboratory analyses were done at the AGH University of Science and Technology in Krako´w. The total dissolved solids (TDS) were calculated by summing concentrations of ions. The PHREEQC Interactive program was used to calculate the equilibria of water samples (Parkhurst & Appelo 1999; Charlton & Parkhurst 2002). Calcite saturation index (SIcalc.) was applied as a measure of equilibrium, according to the formula: SI ¼ log(IAP/KT), where: IAP – Ion Activity Product for ions forming minerals soluble in the given water solution, calculated according to the law of mass action for the given reaction; KT – equilibrium constant for a given reaction.
Measurement of tufa growth rate Tablets made of limestone and copper were used in the experiment. The limestone ones were made of Givetian dark limestone from the vicinities of Krako´w, South Poland, known as the De˛bnik Limestone. It has low porosity and deep-grey to black colour which allows recognizing easily in the field if the tablet is covered with a freshly deposited tufa. The tablets were approximately 80 40 8 mm in size. The faces of the tablets were smoothed. A circular hole 5 mm across was bored through each tablet to enable their fastening
in streams. The copper tablets were c. 75 40 0.7 or 1 mm, and had two holes, each 5 mm across. The dimensions of each tablet were measured with the precision of 0.1 mm and their overall surface area was calculated. Before placing in streams the tablets were dried up at 50 8C for 48 h and weighed with precision of 0.01 g. The tablets were installed using steel nails, in pairs containing copper and limestone ones. At least two pairs were placed at each study site in August 2002. One pair at each site remained there until the end of the experiment (that is until October 2003). The second pair was changed four times a year, every three or four months on average. At the Karwo´w site two pairs of tablets were installed in the crest of the small tufa dam (Karwo´w dam point; Fig. 3a). Other two pairs were placed on the vertical wall of a small cascade a few metres downstream from the dam (Karwo´w cascade point; Fig. 3b). At the Za´zriva´ site only one point was studied (Figs 2b & 3c). Two pairs of tablets were installed on the horizontal part of the tufa fan. Conversely, three points were selected for tablet installation at both the Lu´cˇky and Ha´j sites. In Lu´cˇky all the points were located on the waterfall (Figs 2c & 3d). The first of them (Lu´cˇky top point) was located in horizontal riverbed over the waterfall about 2 m from the waterfall crest. The subsequent points (Lu´cˇky E and W) were on the vertical walls of the waterfall. The water depth at the Lu´cˇky top point was between 6–10 cm. The second point (Lu´cˇky E) was flushed by water intensely falling down, while the third one (Lu´cˇky W) was supplied by a thin film of constantly flowing water. The Lu´cˇky E point was shaded by a vertical cliff of older tufa and by trees, the Lu´cˇky top only by trees while Lu´cˇky W facing to the south was almost not shaded, even during the development of canopy in late spring and summer. At the Ha´j site the tablets were installed along the course of a stream over a distance of more than 900 m (Fig. 2d). The uppermost point was at the Ha´j upper waterfall. Two pairs of tablets were installed on the waterfall face, on a moss curtain that dips at an angle of about 608 (Fig. 3e). The second point was at the lower waterfall. Similarly to the upper one, tablets were installed on its face dipping at an angle of 808. The lowermost point – Ha´j dam point – is located about 50 m downstream from the lower waterfall. The two pairs of tablets and one extra limestone tablet were placed there in the inclined downstream-facing part of the dam that is 0.7 m high (Fig. 3f, g). Collected tablets were transported to the laboratory carefully packed to prevent any loss of tufa, dried at 50 8C for 48 h and weighted again. A test showed that subsequent drying in temperature of 85 8C throughout next 12 h did not cause any
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
change in sample weight. The difference between the tablet weight after and before the experiment was accepted as the mass increment of tufa. The rate of tufa growth was calculated by dividing the mass increment by the total surface area of the tablet and the exposition time.
Texture The textures of the studied tufa were examined. All the limestone tablets after weighting were cut parallel to their shorter edge and the standard thin sections were made. The additional sections were made from the tufa grown on some copper tablets. In the latter case the deposits were carefully scaled off and sunken in boxes filled with ARALDITE. After the resin solidified, thin sections were made. Small portions of tufa were sampled from the tablets to be studied under scanning electron microscope (SEM) JEOL 5410, coupled with a microprobe (EDS) Voyager 3100 (Noran product). As these samples were previously heated in 50 8C their internal structure could be destroyed. To study the original structure of fresh tufa additional samples were collected in the close vicinity of the tablets. To prevent collapse of the organic structure some samples were lyophilized prior to examination. The samples were mounted on SEM holders with silver glue and coated with C or Au.
Water chemistry Chemistry of water that fed growing tufas differed markedly between the sites. The Lu´cˇky site (L3 station) had the highest values of TDS, and Ca content: 968.45 mg/l and 145.9 mg/l, respectively (Table 2). The water at this site represents the Ca –Mg– SO4 –HCO3 type. The water at the Karwo´w site belongs to Ca –Mg –HCO3 type while in other sites to the Ca –HCO3 type. Studied waters were dominated by Ca ion, the mean annual concentration of which ranges from 84.4 at Karwo´w (K1 station) to 67.4 in Za´zriva´ (Z1 station). However, this content is nearly twice lower than that at the Lu´cˇky site. At the sites with lower Ca content concentration of other ions is also lower, which is reflected in TDS values ranging between 522.2 (Karwo´w site) and 305.9 (Za´zriva´ site). Mean SIcalc. values were calculated for the water collected in the nearest stations upstream of the points where the tablets were exposed. Mean value of this parameter is highest at the Lu´cˇky site, equalling 0.94 (L3 station). In other places SIcalc. varies from 0.87 (H6 station) to 0.48 (Z2 station). Some parameters change seasonally (Table 2). The general characteristic is lowering of the water temperature between autumn and spring, which obviously mirrors the lower air temperature in that
151
time. The Ca concentration and TDS are covariant in all sites; however, there is no evident correlation between these two parameters and seasonality. At the Ha´j site they reach their maximum values in November 2002 and March 2002/2003. Conversely, in Karwo´w the highest concentration of Ca was noted in October 2003, while at the Za´zriva´ site in November 2002. The Za´zriva´ site is characterized by the smaller variation of Ca concentration in the course of a year. The Lu´cˇky site experienced the most dramatic changes in Ca concentration (Fig. 4). These reflect fluctuation of mixing proportion between water derived from shallow and deep circulation. For example, the Ca concentration decreased from 129.9 to 97.2 mg/l within a two-week period in March 2003. It resulted from significant influx of surface runoff due to snow melting in the mountains. The Ca concentration was highly elevated in June –December 2003 reaching up to 201.1 mg/l. This, in turn, was the effect of limited surface runoff, which resulted in greater proportion of the deep-circulation water substantially elevating the TDS of the surface stream (Table 2). The SIcalc. values also fluctuated in particular stations over the period of the experiment (Fig. 4). All but one calculated values of this parameter were positive, clearly indicating the overall tendency of calcite to precipitate. The lowest value, 20.03, was calculated for water from the Za´zriva´ site (Z2 station) collected in December 2003. The outstanding positive values, elevated up to 1.22– 1.27, characterize the water from the Lu´cˇky site (L3 station) in early March and June 2003 respectively. It clearly reflects the domination of deep circulating component in the water supplying the studied site.
Tufa deposited on tablets and its rate of growth In the course of the experiment almost all tablets were covered with tufa. Substantial differences could be noticed even in the field between the amounts of tufa growing on different tablets, between individual points and between individual sites studied (Figs 5 & 6). The water level in the studied sites also changed. In the autumn 2003 it drastically decreased in response to prolonged drought. The tablets at the Ha´j upper waterfall and dam points, and at the Za´zriva´ site became exposed and tufa growth on them stopped (Figs 3g & 6b). The rate of tufa growth is calculated from the tufa mass increments (Figs 7 & 8). The results are of semiquantitative character only due to the following reasons: (i) asymmetrical growth of tufa on the tablets; (ii) unknown real area of nucleation
´ SKI M. GRADZIN
152
Table 2. Major parameters of studied water Sample* K 1.1 K 1.2 K 2.1 K 2.2 K 3.1 K 3.2 K 4.1 K 4.2 K 5.1 K 5.2 K 6.1 K 6.2 K 7.1 K 7.2 L 1.3 L 2.3 L 3.3 L 4.3 L 5.3 L 6.3 L 8.3 Z 1.1 Z 1.2 Z 2.1 Z 2.2 Z 3.1 Z 3.2 Z 4.1 Z 4.2 Z 5.1 Z 5.2 Z 6.1 Z 6.2 H 1.1 H 1.2 H 1.6 H 2.1 H 2.2 H 2.6 H 2.10 H 3.1 H 3.2 H 3.6 H 3.10 H 4.1 H 4.2 H 4.6 H 4.10 H 5.1 H 5.2 H 5.6 H 5.10 H 6.6 H 6.10
Sampling date
pH
t [8C]
Ca [mg/l]
Mg [mg/l]
HCO3 [mg/l]
TDS [mg/l]
SIcalc.
24.04.2002 24.04.2002 21.06.2002 21.06.2002 30.08.2002 30.08.2002 03.12.2002 03.12.2002 03.03.2003 03.03.2003 12.06.2003 12.06.2003 03.10.2003 03.10.2003 30.04.2002 22.08.2002 13.11.2002 14.03.2003 31.03.2003 22.06.2003 03.12.2003 18.04.2002 18.04.2002 21.08.2002 21.08.2002 13.11.2002 13.11.2002 14.03.2003 14.03.2003 22.06.2003 22.06.2003 03.12.2003 03.12.2003 03.05.2002 03.05.2002 03.05.2002 23.08.2002 23.08.2002 23.08.2002 23.08.2002 10.11.2002 10.11.2002 10.11.2002 10.11.2002 06.03.2003 06.03.2003 06.03.2003 06.03.2003 29.06.2003 29.06.2003 29.06.2003 29.06.2003 16.10.2003 16.10.2003
7.22 7.93 7.26 7.76 7.44 7.74 7.03 7.66 7.36 8.02 7.19 8.07 7.10 7.74 7.96 8.10 8.01 8.50 8.22 8.04 7.52 7.99 8.10 7.82 8.16 7.32 8.09 7.68 8.04 7.87 8.36 7.65 7.67 7.64 7.95 8.24 8.18 8.12 8.25 8.41 7.81 8.27 8.36 8.26 8.18 8.25 8.25 8.37 7.76 8.20 8.43 8.44 8.36 8.47
8.9 9.4 9.5 10.5 9.5 10.2 8.9 8.6 8.7 7.8 9.1 10.4 9.4 9.7 15.8 14.5 9.3 5.8 4.7 19.8 14 7.5 9.0 7.6 9.4 7.3 7.0 7.2 6.8 7.9 9.5 7.3 5.6 7.9 9.0 11.7 10.1 10.4 11.7 11.5 8.2 2.0 6.9 7.1 6.3 6.2 5.3 6.3 10.6 10.8 12.5 12.9 7.6 7.3
79.3 81.0 78.1 75.6 87.7 74.3 74.3 77.6 90.7 84.6 86.2 90.5 94.6 94.3 137.0 125.0 135.0 129.0 97.2 201.1 196.7 69.6 60.9 62.3 56.8 71.3 70.5 70.1 65.5 70.5 62.0 60.6 55.0 68.3 66.9 57.4 80.5 74.0 65.2 64.2 85.3 88.6 80.3 78.4 89.1 85.9 79.7 77.1 76.9 77.3 76.8 69.8 73.9 73.0
27.1 27.7 22.3 21.4 25.7 25.1 29.2 29.5 25.5 25.6 29.3 31.6 31.6 30.8 43.5 27.8 31.6 26.5 17.2 51.0 45.7 7.1 6.9 5.6 4.9 5.1 5.6 4.7 5.4 7.1 6.8 5.7 3.7 9.0 8.8 8.6 6.3 6.1 6.1 6.0 7.1 8.2 8.1 8.2 7.7 7.5 7.3 7.3 6.9 7.3 7.5 7.4 7.4 7.4
345.7 347.5 362.2 343.6 364.2 349.9 345.3 341.3 401.5 390.5 364.9 379.6 390.1 401.5 289.7 301.7 306.9 315.5 286.8 450.3 430.2 209.8 189.3 206.8 190.4 206.5 218 205.7 212.2 229.4 206.2 218.3 189.3 252.4 249.5 213.7 273.0 238.6 238.6 235.1 258.2 259.6 258.1 238.6 289.7 286.8 263.9 258.1 283.4 285.1 269.6 252.4 260.2 258.1
489.5 491.6 505.0 482.9 522.1 494.6 493.4 493.0 558.2 540.7 526.6 547.1 560.8 569.9 830.2 683.2 672.2 654.5 507.1 1023.6 973.9 309.9 278.4 291.2 269.2 303.0 317.1 298.0 301.3 326.9 294.6 306.5 270.2 365.2 360.7 314.7 383.0 338.2 328.4 323.4 379.7 386.4 375.5 354.8 421.1 413.9 383.8 376.3 389.6 393.4 374.8 350.1 371.9 369.3
20.06 0.66 0.00 0.48 0.23 0.45 20.28 0.35 0.19 0.78 20.04 0.88 20.06 0.59 0.81 0.94 0.82 1.22 0.84 1.27 0.67 0.45 0.49 0.24 0.53 20.22 0.55 0.13 0.46 0.38 0.78 0.07 20.03 0.17 0.48 0.68 0.83 0.7 0.79 0.93 0.44 0.9 0.93 0.8 0.83 0.88 0.8 0.91 0.43 0.86 1.08 1.03 0.91 1.0
*first digit signifies number of sampling campaigns, second one sampling station; K, Karwo´w; Z, Za´zriva´; L, Lu´cˇky; H, Ha´j.
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
Fig. 4. Seasonal temperature, HCO3, Ca and SIcalc. data for Za´zriva´ and Lu´cˇky sites.
153
continually the increase of the surface available for the growth of subsequent crystals (cf. Liu et al. 1995). The tablets were colonized by algae, mosses and housing producing larvae, all of which enlarge the effective nucleation surface. Algae may form freely hanging three-dimensional garlands. They are subsequently incorporated into tufa which overgrows a tablet (Fig. 9b, c). Moreover, many different foreign objects are incorporated into growing tufa. They include detrital mineral grains and plant debris, especially leaves. A substantial part of their surface also acts as a substrate for nucleation. Taking into account that all the above processes are ubiquitous in natural tufa depositing system the initial surface of the tablet may be employed while calculating the tufa rate of growth, though the obtained value must be regarded as semi-quantitative only. The foreign objects mentioned above are included into the tufa increasing its mass, hence the mass of tufa employed into the calculation of the growth rate is somewhat overestimated. Table 3 lists the mass increments and the calculated rates of tufa growth on all tablets. The obtained data set displays some regularity. The highest growth rate was at the Lu´cˇky site and the smallest at the Za´zriva´ site and at Karwo´w dam point. Tufa growth rate was almost always quicker on the limestone tablets than on respective copper ones. Tufa growth was also more rapid at settings with rapid flow, as on waterfalls. A comparison between the tufa growth in mass and in thickness clearly shows that there is no simple relation. A good example is the comparison of tufa growing throughout the whole experiment at the Karwo´w cascade point and at the Lu´cˇky E point. Although at Karwo´w tufa grew to the maximum thickness of 20.04 mm, its mass is only 771.8 mg/cm2, the tufa at Lu´cˇky attained a similar thickness of 24 mm, but a mass of 2 739.5 mg/ cm2 (Table 3). This discrepancy is due to the textural difference between them, reflected in their different porosity (cf. Fig. 9a, c).
Tufa textures surface; and (iii) incorporation of foreign allochthonous components. Tufa asymmetrically overgrew the tablets (Fig. 9). Much more tufa was deposited on the airfacing surface. Since the total area of the tablets was applied to the calculations, the obtained rate of growth is underestimated. The total area of the tablet was assumed as the surface available for nucleation and deposition. In fact, the surface that served as substrate for nucleation is much larger and the increase is hard to be estimated. Even the constantly growing crystals cause
The studied tufa deposited on the tablets during the experiment is built of various components and exhibits a great spectrum of textures (Table 3). X-ray diffraction analyses confirmed that the only autochthonous carbonate mineral phase in all the studied tufa samples is calcite.
Biogenic components and their affinity Filaments. Filaments are the most common biogenic components. On the surface of growing tufa and in its shallow subsurface they are commonly built of
154
´ SKI M. GRADZIN
Fig. 5. Proceeding growth of tufa on pair of tablets exposed at the Karwo´w cascade point throughout the whole experiment, in each photo the limestone tablet (upper one) is much encrusted than copper tablet, (a) in March 2003; (b) in June 2003; (c) in October 2003 after 204, 286 and 399 days of exposition respectively, upper arrow – limestone tablet lower arrow copper tablet, in c pencil tip shows a lower left corner of copper tablet.
organic matter. Deeper on, they are preserved as moulds encrusted with calcite. According to filament size and shape, three morphotypes can be distinguished. The first morphotype comprises non-branching filaments with diameters ranging from 20 to 30 mm and the lengths exceeding even a few centimetres (Fig. 10a, b). They are preserved as smoothwalled moulds without traces of segmentation. The moulds are encrusted predominantly with sparry calcite crystals which commonly constitute the continuous cover on the filament. The smallest crystals, growing directly on the filaments are 1–5 mm in size. They exhibit platy or isometric habit (Fig. 10c, d). The moulds are not filled with calcite. The filaments exhibit different orientation from random to parallel to each other (Fig. 10e). They are parallel or subparallel to the basement, hardly ever erected and never oriented at an angle of 908 to the basement. They were formed mostly in the area of vigorous flow, though are less common on waterfall or cascade faces, where water freely falls down. The tufa from the Lu´cˇky and Karwo´w sites abounds in such filaments which also occur, but subordinately, at the Ha´j site. The size and shape of these filaments suggest their affinity to Vaucheria, an alga belonging to Xanthophyceae, which commonly lives in modern tufa-depositing environments. It was reported, for instance, from Germany (Wallner 1934a, b; Irion & Mu¨ller 1968; Flajs 1977), Belgium (Symoens 1957; Janssen et al. 1999), Poland (Szulc 1983,
1984), France (Freytet & Plet 1991), Italy (Anzalone et al. 2007) and Croatia (Golubic´ et al. 2008). Vaucheria colonizes the fast flowing section of the stream where its filaments are arranged paralel to water flow (Wallner 1934b; Freytet & Verrecchia 1998). Living Vaucheria is abundant at the Karwo´w and Lu´cˇky sites, and has been also found at the Ha´j site (T. Mrozin´ska, pers. comm., 2003). Vaucheria at Karwo´w was also noted by Szulc (1984). Freytet & Verecchia (1998) noted that Vaucheria are encrusted exclusively with platy crystals. The Vaucheria filaments from the Lu´cˇky site are initially covered with small isometric or platy crystals. These crystals develop into rhombohedra oriented perpendicular to the filament which acts as the base for nucleation (Fig. 10a, f; cf. Golubic´ et al. 2008). The finding of Vaucheria filaments, completely overgrown with calcite crystals even in the youngest part of tufa samples deposited within a three-month period, proves that overgrowing is a very quick process. At the Lu´cˇky site it probably lasts only a few days. Some filaments seem to be collapsed or broken, and in thin sections they display semicircular cross-sections. The second type of filaments are of smaller diameter – 5–8 mm – while their lengths attain several dozen micrometres. They are preserved as moulds encrusted with crystals of micritic or sparitic size (Fig. 11a –c). The organic filament serves as nucleation sites for the crystals (Fig. 11d–f ). They can also be entombed in sparry calcite crystals;
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
Fig. 6. (a) Limestone and copper tablets (small arrows) at the Lu´cˇky top point completely covered with tufa after 120 days of exposition, March 2003; (b) tablets at the Za´zriva´ site, left hand pair after 122 days of exposition, right-hand pair after 429 days of exposition, the much more encrustation of the limestone tablet than paired copper one is visible; the tablets are dry owing to low-flow condition, October 2003; big arrows indicate direction of water flow.
however, they do not seem to serve as nucleation surfaces (Fig. 12a). The filament moulds form parallel or upward radiating elongated inclusions, which in some cases are arranged in horizons within sparry crystals (Fig. 12b, c). Analogous preservation of filamentous algae in modern tufa was reported from Oklahoma by Love & Chafetz (1988) and from the British Islands by Pedley (1994, fig. 20). Filaments belonging to this type are variously oriented. They may form fans composed of several upward radiating filaments (Fig. 12d –f ). The filaments do not seem to exhibit branching, however the number of filaments in the fan top appears to be larger than at the base. Biological affinity of this type of filaments is not clear. Their size implies that they belong to cyanobacteria. Upward radiating orientation of filaments shares an affinity with some species of Phormidium. Freytet & Verrecchia (1998) depicted and illustrated Phormidium incrustatum and P. faveolarum which are characterized by alike radial arrangement of filaments. Both cyanobacteria are noted from many
155
Fig. 7. Growth rate of tufa on tablets exposed at Karwo´w and Za´zriva´ sites.
tufa-depositing streams in Europe (e.g. Golubic´ 1967; Penetcost & Spiro 1990; Merz-Preiß & Riding 1999; Arp et al. 2001; Pentecost 2003, 2005). Several species of Phormidium inhabit the studied sites (T. Mrozin´ska, pers. comm., 2003). Thus, the filaments forming radiate fans probably represent this genus. However, other filaments of similar size may be related to any filamentous cyanobacteria. The third type of filaments comprises short forms up to 40 mm in length. Diameter of individual filament varies and maximally reaches 7 mm (Figs 12c & 13a –b). Filaments of this type are preserved as irregularly curved moulds. Their dimensions and co-occurrence with diatom frustules suggest that they represent diatom stalks (cf. Winsborough & Golubic´ 1987; Freytet & Verrecchia 1998). Under SEM they are visible as pits in faces of sparrry crystals (Fig. 13c–d; cf. Pedley 2000). Desmid Oocardium stratum. Sparitic tubes building complicated, upward branching bushes are
156
´ SKI M. GRADZIN
Fig. 8. Growth rate of tufa on tablets exposed at Lu´cˇky and Ha´j sites.
found at the Za´zriva´ site (Fig. 14a, b). The most mature form of the bushes reaches the height of 0.8 mm and width of more than 1 mm. It comprises several branching tubes, thus the whole bush systematically widens upward. Its peripheral zones overhang and the whole structure obtains an umbrella shape. Such bushes overgrew the tablets exposed between August 2002 and October 2003. Their initial forms appear on tablets exposed seasonally. They have a form of a single tube, only 30 mm in height, oriented perpendicular to the substrate tablet, or small bushes with a limited number of ramification (Fig. 14c). The tubes forming the bush occur separate or coalesce with the neighbouring ones. The form of twin-tube was particularly common in the spring of 2003 (Fig. 14d).
The rounded, ovoid cells of algae located at the tube tips were observed under SEM in lyophilized samples (Fig. 14b). Organic mucus was detected between the cells and the calcitic tube. The internal diameter of tubes is constant and varies between 17 and 20 mm. Each tube is built of a single sparry calcite crystal composed of small paralel subcrystals. The spatial arrangement and diameter of tubes, as well as the cell shape prove that the structures are composed of unicellular desmid alga Oocardium stratum. This alga is one of a few freshwater species intimately connected with calcite precipitation (Pentecost 2005, 164–165). It was characterized by Wallner (1935a, b) from Germany, who described the ecology of the species and the
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
157
of Oocardium stratum cells (Fig. 14d; Golubic et al. 1993). Diatoms. The diatom flora is abundant in the studied sites (T. Mrozin´ska, pers. comm., 2003). It is documented by SEM observations of growing tufa surface (Fig. 15a, b). The diatoms are associated there with copious amounts of EPS (extracellular polymeric secretions), some filamentous cyanobacteria and other algae. The diatoms belong to different morphological groups. Particularly common are pennate diatoms, some of which produce slimy stalks. The diatom frustules are discernible also within the deposited tufa. They are particularly common in micrite tufa, but have been also found in sparry crystals (Fig. 15c). They are engulfed by calcite but seldom act as a substrate for nucleation of calcite crystals (Fig. 15d). Diatom frustules without calcite nucleated on them was reported from many modern tufa sites, and it seems to be a rule (e.g. Merz-Preiss & Riding 1999; Lu et al. 2000). In the course of the present study, nucleation of crystals on diatom frustules was observed exclusively in some samples from the Lu´cˇky site. In the deeper zones of the studied tufa samples, frustules become less abundant, though their moulds are found under the SEM (Fig. 15e; cf. Pedley 1994; Szulc & Smyk 1994). It implies that silica frustules can be dissolved within three months or less.
Fig. 9. Scanned thin sections of tufa formed on tablets, (a) Karwo´w cascade point, tablet exposed between August 2002 and October 2003, tufa displays high porosity; (b) Lu´cˇky E point, tablet exposed between June 2003 and October 2003; (c) Lu´cˇky E point, tablet exposed between August 2002 and October 2003; arrows in b and c indicate leaves incorporated into tufa and acting as additional nucleation surface.
growth mode of its calcified tubes. After World War II many other recent tufa sites abounding in Oocardium stratum were recognized in Croatia, France, Switzerland, Belgium, Poland, Great Britain, USA and China (Golubic´ & Marcˇenko 1958; Pentecost 1990, 1991; Freytet & Plet 1991; Mrozin´ska 1992; Golubic et al. 1993; Pentecost & Zhang 2000; Golubic´ et al. 2008). The observed twin-tubes are a record of divisions
Other microalgae. In the tufa sample from the Lu´cˇky E point, single calcified ovoid bodies 20– 25 mm across were found. Inside, each of them contained several small globules, 1–4 mm in diameter (Fig. 15f). The bodies strongly resemble colonies of unicellular cyanobacteria that are common in tufa-forming environment. At Lu´cˇky they were found in a high-energy setting in the waterfall wall, in a shelter cavity formed by a leaf projecting from the tablet and cemented to it. It seems probable that they were earlier transported by the stream. Extracellular polymeric secretions. The extracellular polymeric secretions (EPS) cover the surface of all samples. Under SEM they are visible as irregularly twisted filaments or a three-dimensional reticulate structure (Fig. 16a, b; cf. De´farge et al. 1996). The filaments are between ,1 mm and c. 15 mm across. In some cases EPS construct a dense layer which completely obliterates their substrate. The EPS are spatially related to diatoms and filamentous cyanobacteria. It also suggests a genetic relationship, because some diatoms and cyanobacteria are known to be efficient EPS producers (Riding 2000). Moss stems. Moss stems were occasionally found within the tufa deposited on the tablets. The moss
158
Table 3. Mass increment, growth rate, thickness and texture of tufa deposited on tablets, asterisk denotes that thickness were measured with slide gauge, thickness of other samples was measured under petrographical microscope Exposition period
Exposition time [days]
Mass increment [g]
Mass increment [mg/cm2]
Growth rate [mg/cm2/day]
Karwo´w, dam, limestone Karwo´w, dam, limestone Karwo´w, dam, limestone Karwo´w, dam, limestone Karwo´w, dam, copper Karwo´w, dam, copper Karwo´w, dam, copper Karwo´w, dam, copper Karwo´w, cascade, limestone Karwo´w, cascade, limestone Karwo´w, cascade, limestone Karwo´w, cascade, limestone Karwo´w, cascade, limestone Karwo´w, cascade, copper Karwo´w, cascade, copper Karwo´w, cascade, copper Karwo´w, cascade, copper Karwo´w, cascade, copper Za´zriva´, limestone Za´zriva´, limestone Za´zriva´, limestone Za´zriva´, limestone Za´zriva´, limestone Za´zriva´, copper Za´zriva´, copper Za´zriva´, copper Za´zriva´, copper Za´zriva´, copper Lu´cˇky top, limestone Lu´cˇky top, limestone
08–12.2002 12.2002–06.2003 06–10.2003 08.2002–10.2003 08–12.2002 12.2002–06.2003 06–10.2003 08.2002–10.2003 08–12.2002 12.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08–12.2002 12.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003
95 191 113 303 95 191 113 303 95 109 82 113 399 95 109 82 113 399 84 119 102 122 427 84 119 102 122 427 84 119
0.08 0.77 0.36 2.22 0.05 2 0.94 0.14 21.67 4.35 1.54 4.91 13.49 62.67 0.43 0.08 0.1 4.22 5.06 0.35 0.31 21.05 0.15 5.43 0.24 20.02 0.09 20.08 3.17 7.82 8.61
0.9858 9.5926 4.5249 27.4107 0.8050 – 2.2624 – 52.7785 19.2284 61.1533 174.7636 771.7980 7.1227 1.3249 1.6316 69.637 83.8165 4.3088 3.979 214.1720 1.9118 68.3105 3.8785 – 1.4329 – 51.2282 112.3079 104.9360
0.0104 1.0392 1024 0.0502 5.0160 1024 0.0400 4.0000 1024 0.0905 9.0428 1024 8.4739 1023 8.4678 1024 – – 0.0200 2.0001 1024 – – 0.5556 5.5516 1023 0.1763 1.761 1023 0.7458 7.4520 1023 1.5468 0.0155 1.9343 0.0193 0.0750 7.4940 1024 0.0122 1.2190 1024 0.0199 1.9884 1024 0.6163 6.1580 1023 0.2101 2.0993 1023 0.0513 5.1259 1024 0.0334 3.3410 1024 20.1389 21.3879 1023 0.0157 1.5687 1024 0.1599 1.5984 1023 0.0462 4.6163 1024 – – 0.0140 1.3989 1024 – – 0.1199 1.1988 1023 1.337 0.0134 0.8818 8.811 1023
Growth rate [mmol/cm2/day]
Thickness [mm] 0 – 0.02 0.1– 1.1 0 – 0.38 0 – 077 na – na – 0 – 0.46 0 – 0.57 0 – 0.95 0.57– 1.33 1.14– 20.04 0 – 0.6* 0 – 0.6* 0 – 0.2* 0.38– 1.4* 0 – 2.28 0.02– 0.04 0.02– 0.2 0 – 0.2 0 – 0.1 0.1– 1.9 0 – 0.1* – 0 – 0.1* – 0.06– 1.1 1.25– 2.09 1.52
Main texture
M S M, HLM HML, S na – na – HLM M HLM S HML, F, S na na na S S S, B S, B S, B S, B S, B na na na na S, B S, M M, S, F
´ SKI M. GRADZIN
Tablet location, type
03–06.2003 06–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 11.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 11.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003
102 121 426 84 119 102 121 426 84 119 102 121 426 84 119 102 121 426 84 119 102 121 342 84 119 102 121 342 80 115 116 108 419 80 115 116
tablet lost 63.45 127.2 2.82 2.12 2.36 40.34 tablet lost 46.15 25.25 11.39 98.78 223.82 19.73 12.02 5.6 tablet lost 119.21 tablet lost tablet lost 1.41 tablet lost 124.69 tablet lost 0.54 0.44 68.81 93.89 1.79 0.41 0.52 2.65 8.12 tablet lost 0.17 0.06
– 845.5490 1643.83 46.5303 34.2599 38.9439 651.9069 – 568.3498 313.6256 145.0955 1296.2866 2739.5349 317.6622 190.4309 91.3689 – 2007.9164 – – 18.0978 – 1604.7619 – 8.9109 7.0053 1095.5262 1549.3399 22.0362 5.0774 6.8647 34.2466 101.4493 – 2.7473 0.9553
– 6.988 3.8588 0.5539 0.2879 0.3818 5.6199 – 6.7661 2.6356 1.4225 10.7131 6.4308 3.7817 1.6033 0.8958 – 4.7135 – – 0.1774 – 4.6923 – 0.0749 0.0687 9.0539 4.5302 0.2755 0.0442 0.059 0.3171 0.2421 – 0.0239 8.82353 10 – 3
– 0.0698 0.0386 5.5346 1023 2.8767 1023 3.815 1023 0.0562 – 0.0676 0.0263 0.0142 0.107 0.0643 0.0378 0.016 8.9506 1023 – 0.0471 – – 1.7729 1023 – 0.0469 – 7.4822 1024 6.8625 1024 0.0905 0.0453 2.7528 1023 0.4165 1024 5.8953 1024 3.1685 1023 2.4193 1023 – 2.3881 1024 8.2288 1025
– .14.2* .18.4* 1.44– 1.86 0.96– 1.71 0.3* 4.94– 6.27 – 0.76– 0.09 0.76– 1.36 0.57– 0.86 2.28– 17.1 2.28– 24.0 1.79 0.53– 1.33 0.57– 0.82 – 0.5– 2.7 – – 0 – 0.15* – 3.0 – 9.9 – na na 2.3– 7.4* 3.7– 17.6* 0.1– 2.9 0 – 0.5 0 – 0.1 0.1– 0.48 0.48– 1.78 – 0 – 0.2* 0 – 0.2*
– S, F S, F M, S M, S na S, F – S S S S S, M, F S S S – S, F – – na – S, F S, M M, S, F S, F S, F S S HLM S S, M
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
Lu´cˇky top, limestone Lu´cˇky top, limestone Lu´cˇky top, limestone Lu´cˇky top, copper Lu´cˇky top, copper Lu´cˇky top, copper Lu´cˇky top, copper Lu´cˇky top, copper Lu´cˇky E, limestone Lu´cˇky E, limestone Lu´cˇky E, limestone Lu´cˇky E, limestone Lu´cˇky E, limestone Lu´cˇky E, copper Lu´cˇky E, copper Lu´cˇky E, copper Lu´cˇky E, copper Lu´cˇky E, copper Lu´cˇky W, limestone Lu´cˇky W, limestone Lu´cˇky W, limestone Lu´cˇky W, limestone Lu´cˇky W, limestone Lu´cˇky W, copper Lu´cˇky W, copper Lu´cˇky W, copper Lu´cˇky W, copper Lu´cˇky W, copper Ha´j, upper waterfall, limestone Ha´j, upper waterfall, limestone Ha´j, upper waterfall, limestone Ha´j, upper waterfall, limestone Ha´j, upper waterfall, limestone Ha´j, upper waterfall, copper Ha´j, upper waterfall, copper Ha´j, upper waterfall, copper
– na na (Continued)
159
160
Table 3. Continued Exposition period
Exposition time [days]
Mass increment [g]
Mass increment [mg/cm2]
Ha´j, upper waterfall, copper Ha´j, upper waterfall, copper Ha´j, lower waterfall, limestones Ha´j, lower waterfall, limestone Ha´j, lower waterfall, limestone Ha´j, lower waterfall, limestone Ha´j, lower waterfall, limestone Ha´j, lower waterfall, limestone Ha´j, lower waterfall, copper Ha´j, lower waterfall, copper Ha´j, lower waterfall, copper Ha´j, lower waterfall, copper Ha´j, lower waterfall, copper Ha´j, dam, limestone Ha´j, dam, limestone Ha´j, dam, limestone Ha´j, dam, limestone Ha´j, dam, limestone Ha´j, dam, limestone Ha´j, dam, copper Ha´j, dam, copper Ha´j, dam, copper Ha´j, dam, copper Ha´j, dam, copper
06–10.2003 08.2002–102003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08.2002–06.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003 08.2002–10.2003 08–11.2002 11.2002–03.2003 03–06.2003 06–10.2003 08.2002–10.2003
108 419 80 115 116 108 419 311 80 115 116 108 419 80 115 116 108 419 419 80 115 116 108 419
0.38 2.36 tablet lost 11.84 6.33 9.66 tablet lost 16.94 5.89 tablet lost 0.55 0.32 tablet lost 3.43 5.06 2.67 1.78 11.47 11.70 2.32 2.53 0.24 0.06 8.9
6.05 37.9171 – 153.5070 78.3125 124.8385 – 213.5653 97.5650 – 8.7566 5.0947 – 40.3867 61.9263 33.5596 22.2388 143.3929 142.3531 37.3531 40.8856 3.8210 0.9553 141.0013
Growth rate [mg/cm2/day] 0.0560 0.0905 – 1.3348 0.6751 1.1559 – 0.6867 1.2196 – 0.0755 0.0499 – 0.5048 0.5385 0.2893 0.2059 0.3422 0.3397 0.4669 0.3555 0.0329 8.8453 1023 0.3365
na, not analyzed; B, sparite bushes; F, fibrous; HLM, hemiphaerically laminated micrite; M, clotted micrite; S, sparry crystals.
Growth rate [mmol/cm2/day] 5.5955 1024 9.0422 1024 – 0.0133 6.7457 1023 0.0115 – 6.8616 1023 0.0122 – 7.5428 1024 4.9860 1024 – 5.0440 1023 5.3807 1023 2.8908 1023 2.0574 1023 3.4195 1023 3.3947 1023 4.6653 1023 3.5522 1023 3.2913 1024 8.8383 1025 3.3625 1023
Thickness [mm] 0 – 2.3* 0.4– 0.5* – 0.39– 0.91 0.07– 0.77 0.57– 1.1 – 0.64– 3.04 0 – 1.1* – 0 – 0.7* 0 – 1.4* – 0.1– 0.76 0.2– 2.66 0.1– 0.48 0.1– 0.34 0.2– 2.28 0.1– 0.76 0.2– 0.7* 0 – 0.2* 0 – 0.2* 0 – 0.6* 0.1– 1.9
Main texture
na na – S S S, M – S, M na – na na – S, M S, M S, M S, M S, F, M S, M na na na na na
´ SKI M. GRADZIN
Tablet location, type
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
161
Fig. 10. Calcified Vaucheria filaments: (a) cross section of a filament, December 2003; (b) cross-section of filaments, June 2003; (c) living Vaucheria filament initially covered with calcite crystals, October 2003; (d) initial crystals forming a filament, June 2003; (e) parallel orientation of filaments, June 2003; (f) cross section through more heavily encrusted filaments, Lu´cˇky W point, tufa formed on Cu tablet exposed between August 2002 and October 2003; all but (f) Lu´cˇky top point, a, d – SEM images, lyophilized samples, b, c, e – thin sections, c – algological sample under microscope.
stems were cemented to the air-facing surface of the limestone tablet exposed throughout the whole time of the experiment at the Ha´j upper waterfall point (Fig. 16c). The stem represents Brachythecium rutabulum (R. Ochyra, pers. comm., 2003). This species of moss grows in the water running down the
waterfall and forms there, together with other species, a 2 m long moss curtain. The stem was covered by epiphytic diatoms and cyanobacteria and was encrusted with calcium carbonate (Fig. 16d). Cross sections of diatom stems were also detected in the tufa growing on limestone
162
´ SKI M. GRADZIN
Fig. 11. Calcified cyanobacterial filaments: (a) micrite encrusted filament, Karwo´w cascade point, December 2002; (b) sparite encrusted filament, Lu´cˇky top point, tufa formed on limestone tablet exposed between August 2002 and October 2003; (c) longitudinal section of filaments, Lu´cˇky E point, tufa formed on Cu tablet exposed between August 2002 and October 2003; (d, e) filaments serving as nucleation surfaces for calcite crystals, d – Lu´cˇky E point, tufa formed on limestone tablet exposed between August 2002 and October 2003, e – Lu´cˇky E point, tufa formed on Cu tablet exposed between August 2002 and October 2003; (f) parallely arranged living filamentous cyanobacteria partly covered with calcite crystals, Lu´cˇky E point, March 2003; all but d SEM images, d – thin section, f – lyophilized sample.
tablets at the Ha´j lower waterfall point and at the Karwo´w cascade point. Larval housings. Vaulted voids up to 0.5 mm wide and 0.35 mm high occur within the tufa deposited on some tablets (Fig. 17). The voids have circular or semicircular cross-section with convex-up ceilings. They are covered with a thin lamina of micrite, which in turn is overlain by sparry calcite crystals displaying competitive growth pattern
(Fig. 17a–c). The voids were detected at the Ha´j lower waterfall, Ha´j dam points and at the Lu´cˇky site. They may be arranged in horizons within the tufa samples as was the case at the Ha´j site or distributed randomly within the tufa section (Fig. 17b, c). Some voids are less regular; they were probably corroded (see Golubic´ 1969). The voids are comparable to the larval housings described from many tufa sites in Europe, for instance in Germany (Wallner 1934a; Irion &
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
163
Fig. 12. Calcified cyanobacterial filaments: (a) sub-vertically arranged filaments entombed in growing sparry crystals comprising several needle-shaped subcrystals, openings of some filament moulds are arrowed, tufa formed at Lu´cˇky E point on Cu tablet between August and November 2002, (b) sparry calcite crystals with entombed filaments, Karwo´w cascade point, tufa formed on Cu tablet exposed between August 2002 and October 2003; (c) cross-section of filament openings differing in size, small openings (arrows) probably represent diatom stalks, crystal broken surface, the same sample as in b; (d) fan of upward radiating filaments, tufa formed at Karwo´w cascade point between March and June 2003; (e) detail from d; (f) top view of the fan, scattered moulds of cyanobacteria and detritic quartz grains (arrows) incorporated into fan are visible, tufa formed at Karwo´w dam point on limestone tablet exposed between August 2002 and October 2003; a, c, f – SEM images, b, d, e – thin sections.
Mu¨ller 1968; Du¨rrenfeldt 1978), Belgium (Symoens 1957; Janssen et al. 1999), France (Freytet & Plet 1996), Italy (Golubic et al. 1993) and Greece (Andrews & Brasier 2005). They are particularly abundant at the Ha´j study site, where they dominated tufa micromorphology in June 2003
(Fig. 17d). Such housings may be produced by Trichoptera and Chironomidae larvae (see Wallner 1934a; Du¨rrenfeldt 1978; Drysdale 1999). However, it is impossible to precisely ascertain their producers in the studied case as a reconnaissance examination carried out at the Ha´j site proved the
164
´ SKI M. GRADZIN
Fig. 13. Small filaments – probably encrusted diatom stalks or fibres of EPS: (a) small curved moulds within sparry crystals, the moulds are more than 20 mm above the basement, tufa formed on Cu tablet, Za´zriva´ site, between August 2002 and October 2003, arrow indicates position of tablet before its detachment; (b) moulds in sparry crystals lying directly on limestone tablet, arrow indicates its upper boundary, Lu´cˇky top point, tufa on limestone tablet exposed between August and November 2002; (c) transversal and oblique cross-sections of filament moulds; (d) pit in a sparry crystal having been an attachment point of diatom stalk, Lu´cˇky top point, December 2003; a, b – thin sections, c, d – SEM images.
occurrence of both groups, Trichoptera and Chironomidae there (A. Kownacki, pers. comm., 2003).
Carbonate precipitates and their textures Clotted micrite. Micrite displaying clotted texture is common in the studied tufa samples (Fig. 18). It consists of anhedral crystals approximately a few micrometres across on average. The crystals agglomerate and form rugged clots several dozen micrometres across. Micrite builds laminae up to 0.5 mm thick, which commonly fill irregularities in its basement, such as intercrystalline porosity of underlying sparite laminae (Fig. 18d). It also engulfs sparite crystals (Fig. 18e). In some samples the substrate of micrite layer is corroded (Fig. 19a), which is better visible when the substratum is composed of sparry crystals. Diatom frustules or their moulds and cyanobacterial filaments abundantly co-occur with micrite (Figs 15b & 18f ). On the surface of growing tufa the minute carbonate crystals building micrite are
accompanied by diatoms, cyanobacteria and their EPS. Clotted micrite commonly hosts larval housings. Clotted micrite is found in nearly all studied samples. It predominates in samples growing in sluggish flow conditions and is rare in waterfall settings (cf. Pedley 1992). If it occurs in the latter settings, it forms thin laminae alternating with sparitic ones or it fills shelter cavities, for instance between sparite crystals. Clotted micrite never forms the lowermost layer of tufa directly on Cu tablets. Although the lowermost layer may be built of minute crystals, it does not display clotted texture or the presence of algae (Fig. 18c). Such crystals quickly augment upward, growing up according to the rule of competitive crystal growth (cf. Gonza´lez et al. 1992). The above observations seem to confirm the view by Pedley (1992, 1994, 2000) that micrite in tufa is genetically connected with micro-organisms. Microbial origin of micrite was, for example, postulated also by Jones & Kahle (1995) who studied
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
165
Fig. 14. Sparitic tubes of Oocardium stratum, all samples from the Za´zriva´ site (a) bushy form comprising several branching tubes that grew on Cu tablet exposed between August 2002 and October 2003, the tubes are divided from Cu substratum by a sparry layer; (b) surface of Oocardium stratum colony seen from above, the cells are visible at the tube tips, uncalcified diatom frustules occur on the sample surface, October 2003; (c) initial tube formed on limestone tablet between August and October 2002; (d) twin tubes marking the cell division moment, surface of tufa, sample collected in March 2003; a, c – thin sections, b, d – SEM image, lyophilized samples.
the Cayman Islands. However, they put forward also other possible origins of micrite in a karst environment. Hemispherically layered micrite. Micrite also forms hemispheres up to 1 mm high and a few millimetres wide (Fig. 19b, c). The hemispheres exhibit subtle convex-up internal lamination visible most probably due to changes in microporosity. They comprise radially oriented cyanobacterial filaments (Fig. 12d –f). Minute calcite crystals are present between the filaments. Detrital grains, including quartz, are trapped there and between the hemispheres (Figs 12f & 19d). The hemispheres occur mainly at both the studied points of the Karwo´w site. They occur in patches, but tend to be concentrated along the tablet edges (Fig. 19c, e, f ). They predominate on the air-facing side of tablets but some are also located on the substrate-facing one. On the tablet exposed throughout the whole period of the experiment the neighbouring hemispheres coalesce into a continuous layer (Fig. 19e, f). The micrite hemispheres are akin to tuft-shaped colonies of Phormidium incrustatum. What they have in common is the size, internal lamination and the presence of radially arranged cyanobacterial
filaments inside (Freytet & Verrecchia 1998; Pentecost & Whitton 2000). Also the general outline of calcitized Phormidium colonies resembles the described hemispheres (Pentecost 2003; Golubic´ et al. 2008). The coalescence and overlapping of hemispheres result from horizontal invasion of cyanobacteria (Pitois et al. 2001). Fibrous micrite/sparite. Crystals from a few micrometres to .100 mm in size, encrusting algal filaments, build porous tufa of fibrous texture (Fig. 20a, b). The texture may be visible with the naked eye (Fig. 20c). The filaments acting as a substrate for nucleation belong to Vaucheria and some cyanobacteria. Diatom frustules occurring within this type of tufa become less numerous downward the samples, probably because of dissolution. The filaments are preserved as empty tubes. Crystal size in filaments tends to increase outwards (Fig. 20d). Though intercrystalline porosity is an important characteristic of this texture, sizes of pores are variable. They depend on the degree of cementation and on the primary arrangement of the filaments (Golubic´ et al. 2008). If the filaments are tightly packed, a small amount of sparry calcite can effectively cement the whole
166
´ SKI M. GRADZIN
Fig. 15. (a) Diatoms on tufa surface, Karwo´w cascade point March 2003; (b) diatoms and EPS on tufa surface, note small clots of anhedral calcite crystals, Ha´j dam point, May 2002; (c) diatom rich layer in the topmost part of tufa deposited between November 2002 and March 2003 in Ha´j dam point on limestone tablet; (d) diatoms on tufa surface, Lu´cˇky E site, note the nucleation of small calcite crystals directly on the frustules (arrows), March 2003; (e) diatom frustule mould (arrow), Lu´cˇky top point, tufa formed on limestone tablet between August 2002 and October 2003; (f) calcified colony of coccoid cyanobacteria, Lu´cˇky E site, tufa on limestone tablet exposed between August 2002 and October 2003; a, b, d – lyophilized samples, SEM images, c, f – thin sections.
structure. Conversely, loosely arranged filaments need considerably higher amount of calcite to be cemented completely into a compact sparry layer (Figs 10e & 20e). Some sparry crystals do not exhibit preferred orientation (Fig. 20a, b, f ). They are between 5– 15 mm across. They lack enfacial junctions and compromise boundaries as well. They are loosely cemented and look like detached from their substrate. The detachment is easiest after the
decomposition of the algal tissue acting as a nucleation surface for the crystals. The whole process probably took place below a living algal or cyanobacterial mat, which prevented removal of the loose crystals. They were later cemented and become a component of the fibrous texture. Sparite bushes. The sparite bushes are intimately connected with Oocardium stratum. They were present only at the Za´zriva´ site. Loose sparite
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
167
Fig. 16. (a) Elongated cyanobacterial filaments covered with EPS, note small clots of anhedral calcite crystals, Lu´cˇky E point, March 2003; (b) three dimension network of EPS, Ha´j upper waterfall point, June 2003; (c) moss stem on the tablet exposed between August 2002 and November 2003 at the Ha´j upper waterfall point, arrow indicates direction of water flow; (d) cross section of moss stems, the same sample as in c; a, b – lyophilized sample, SEM images; d – thin section.
tubes of Oocardium stratum are overgrown syntaxially and cemented together into a more compact structure (Figs 14a, b & 21a). However, the bushy shape is still retained. The Oocardium stratum tubes inherit their crystal properties from their parent tubes (Golubic et al. 1993). Hence, many neighbouring tubes display uniform optical orientation and they, while being overgrown, form bigger sparite crystals (Fig. 21b, c). However, some defects must occur, probably during the division of the cells, since the bush that is a progeny of one tube, is composed not of one but a few sparite crystals slightly differing in optical orientation (Fig. 21d). Cementation of neighbouring tubes is a rapid process. On a tablet exposed in the course of 14 months, more than half of the tubes were cemented together; however, some primary inter-tube voids within bushes still remain empty. The tube interiors survived empty as well. Sparry calcite. Sparitic crystals are common components of the tufa formed during the experiment (Fig. 22). The crystals are elongated, they range from several micrometres to a few millimetres in length. Their terminations are flat, convex-up or, rarely, steeply sided (Figs 18e & 22a, b, f ). Most of the crystals grew normal to the substrate;
however, their orientation becomes variable just over the substrate because of competitive growth fabrics (Figs 18c, 22, 23a, b, d; cf. Gonza´lez et al. 1992). The sparry crystals occur in laminae or lenses, the latter commonly surrounded by micrite. The thicknesses of individual laminae range from a few micrometres to more than one millimetre. Some laminae are uniform in thickness, while others vary in thickness as a result of their uneven tops (Fig. 18d). The uneven tops are characteristic mostly of thicker laminae, composed of fan-shaped crystals. Such crystals display undulose extinction and comprise several smaller radially arranged subcrystals. The fan-shaped crystals may coalesce laterally or are separated from each other. Some sparry laminae exhibit high intercrystalline porosity (Fig. 23a, b). The porosity is higher in sparry lamina from the Karwo´w and Ha´j sites than in those formed at the Lu´cˇky site. Sparry crystals abound in moulds of filamentous cyanobacteria, diatom stalks and sporadic diatom frustules (Fig. 12a, b). The filaments are mostly normal to the crystal substrate or form upward radiating fans. Laminae built of sparry calcite are common components of tufa. Pedley (1992, 2000) noted that they form mainly in high-energy settings.
168
´ SKI M. GRADZIN
Fig. 17. Larval housings: (a) larval housing within tufa composed of calcified cyanobacterial filaments, note sparry crystals displaying competitive growth pattern which cover the housing, tufa formed on limestone tablet at Lu´cˇky E site between August 2002 and October 2003; (b) housings arranged one over another, tufa formed on limestone tablet at Lu´cˇky E point between August 2002 and October 2003; (c) horizontally arranged housings, tufa formed at Ha´j lower waterfall point between August 2002 and June 2003; (d) field photo of tufa at Ha´j site covered by larval housings with meandering pattern, June 2003, coin is 22 mm across; a – SEM image; b, c – thin sections.
Moreover, they are regarded as a typical ‘winter deposits’ (e.g. Pedley 1992, 1994). During the present experiment such laminae grew at all the studied sites. They always build the lowermost layer deposited on the copper tablets, where they lack biogenic components (Figs 13a, 14a & 23c). They were formed also during the winter season, for example at the Ha´j and Karwo´w sites. Sparry laminae abound in samples from waterfalls, as at Lu´cˇky, Ha´j and Karwo´w (Figs 16d, 22 & 23a). Sparry crystals commonly develop on plant debris, especially leaves, and larval housings, and on any other objects introduced into the tufa-depositing environment (Fig. 23d). Detritic components. Although all the studied tufas are autochthonous, they include some admixture of detritic components. Fragments of redeposited tufa are probably the most common, but they are hard to be distinguished. Fine sand- and silt-sized quartz grains are found at all the studied sites, except Za´zriva´. They are particularly common at the Karwo´w site, especially in the tufa deposited
during autumn and early spring (Figs 12f & 19d). The quartz grains at that site derive from a loess cover which is easily eroded. As carbonate is a ubiquitous component of the nearby loess, it most probably enters the tufa-depositing stream and is introduced into growing deposits as well. Other allogenic components: clay minerals and iron oxides, were detected by EDS analyses. Plant debris, especially leaves and small twigs are also commonly incorporated into the growing tufa (Figs 9b, c & 23d). Lamination. Three types of lamination were recorded within the tufa that grew on the tablets. The most common type of lamination is one with alternating laminae of sparry crystals and micrite (Fig. 18a–d). Boundary between the laminae may be accentuated by corrosion. Corrosion is best visible on the upper surfaces of crystalline laminae underlying the micritic ones. Subtle changes in crystal size, probably in conjunction with changes in porosity are responsible for the lamination visible within hemisphaerically layered micrite (Fig. 19b, c).
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
169
Fig. 18. (a) Micrite lamina overlying sparry crystals, Lu´cˇky top point, tufa formed on limestone tablet exposed between August and November 2002; (b) clotted micrite, Lu´cˇky top point, upper part of the sample presented in a; (c) sparry crystals displaying competitive growth rule covered with micrite, the smooth surface (arrow) is the basement of sample visible due to detaching of the tablet, Lu´cˇky top point, tufa formed on copper tablet exposed between August and November 2002; (d) micrite surrounds fan of sparry crystals, Lu´cˇky top point tufa formed on limestone tablet exposed between August and November 2002; (e) sparry crystal engulfed by micrite, arrow indicates EPS, tufa surface, Ha´j dam site, October 2003; (f) clotted micrite rich in diatoms, the same sample as in c; a, b thin sections, c–f – SEM images, e – lyophilized sample.
170
´ SKI M. GRADZIN
Fig. 19. (a) Corroded sparry crystals below micrite (arrows), tufa formed on limestone tablet at Karwo´w cascade point between June and October 2003; (b, c) laminated hemisphere, tufa formed on limestone tablets at Karwo´w dam point between August 2002 and October 2003 (b) and Karwo´w cascade point between March and June 2003 (c); (d) laminated hemisphere surrounded by detritic quartz grains, Karwo´w dam point, tufa on limestone tablet exposed between August 2002 and November 2003; (e) linked hemispheres constituting continuous tufa cover on limestone tablet exposed at Karwo´w dam point between August 2002 and October 2003; (f) isolated hemispheres, limestone tablet exposed at Karwo´w cascade point between August and December 2002; a –d thin sections, arrows in e and f indicate direction of water flow.
Another type of lamination is visible because of changes in porosity within tufa of fibrous texture (Fig. 24a– c). It is akin to lamination in algal mats. This type of lamination is especially well developed at the Lu´cˇky top and E points, in the tufa with fibrous texture. More compact laminae alternate with those being more porous. The dimensions of crystals in compact laminae gradually grow
upwards (Fig. 24b). Both types of laminae are built of moulds of filamentous algae, highly encrusted with calcite. In the compact laminae, spacing between particular algal filaments is small, so that cement largely fills the gaps between the filaments (Fig 24d). In the porous laminae the number of algae is considerably smaller and the spaces between them remain open. Thus, the lamination
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
171
Fig. 20. Fibrous texture: (a) encrusted Vaucheria filaments, Lu´cˇky top point, December 2003; (b) encrusted cyanobacterial filaments, Lu´cˇky E point, tufa formed on limestone tablet between August 2002 and October 2003; (c) sample built of Vaucheria filaments, macroscopic view, Lu´cˇky top point, June 2003; (d) crystal size increases outwards the Vaucheria filament, Lu´cˇky top point, October 2003; (e) loosely cemented cyanobacterial filaments showing parallel orientation Lu´cˇky E point, tufa formed on limestone tablet between August 2002 and October 2003; (f ) loose crystals (arrows) trapped and cemented between filaments, the same sample as in e; a, b, d, f – SEM images, e – thin section.
reflects variations in primary density of algae inhabiting the surface of the growing tufa. More than 60 couplets of compact and dense laminae are present in tufa that grew on a limestone tablet exposed between August 2002 and October 2003. The laminated tufa is underlain by tufa with crystalline texture, so its deposition took place only for a part of the time of the tablet exposition. The third type of lamination is manifest by concentration of detritic non-carbonate components or
organic matter. When detritic components are deposited within porous tufa, they are dispersed. The tufa that grew at the Karwo´w cascade throughout the whole experiment serves as an example. Conversely, compact, crystalline texture of tufa favours the origin of clearly visible laminae, built of fine-grained detritic components (Fig. 22a– e). The laminae drape underlying crystals and accentuate the shape of crystal terminations. The laminae range in thickness from less than 0.5 to 5.0 mm
172
´ SKI M. GRADZIN
Fig. 21. (a) Neighbouring tubes of Oocardium stratum cemented together forming one sparry crystal accompanied by uncalcified diatom frustules; (b) upper surface of sparite bush, all tubes form one monocrystal, the cells are still situated at tube tips; (c) almost uniform extinction of sparite bushes built of Oocardium stratum tubes; (d) different optic orientation of crystals within one Oocardium stratum bush, the same sample as in c; all samples from Za´zriva´ site, a, b – October 2003, SEM images, lyophilized samples, c, d – tufa formed on copper tablet between August 2002 and October 2003, thin sections, crossed nicols.
and the thicker ones display granular fabrics. Calcite crystals do not change their optic orientation crosscutting many of the thin laminae, probably because of the low content of detritic grains within the laminae. Thicker laminae hindered calcite crystal growth, which is reflected in competitive growth fabrics of sparry crystals constituting the younger sparry lamina. Laminae of this type were formed especially at the Lu´cˇky site during autumn and spring seasons. Less distinctive and thin laminae of this type were found at the Karwo´w and Ha´j sites (Fig. 23a). Such laminae were previously known from crystalline tufa and are related to the concentration of organic matter (Irion & Mu¨ller 1968; Pedley 1992). The enrichment in fine-grained clastics was thought to cause similar lamination in
speleothems (Gradzin´ski et al. 1997; see also Baker et al. 2008 and references herein).
Discussion Influence of micro-organisms on the rate of tufa growth The field experiments conducted in the course of this study allow the semi-quantitative estimation of the difference between the rate of microbially and inorganically driven tufa growth in various natural settings. The interpretation is based on differences between the rates of tufa growth on two different substrates (limestone and copper), in various settings and during various seasons, and
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
173
Fig. 22. (a, b) Highly elongated crystals building fans formed on limestone (a) and copper (b) tablets between November 2002 and March 2003; (c) fans showing undulose extinction, the same sample as in a, crossed nicols; (d, e) highly elongated crystals building fans, d – parallel nicols, e – crossed nicols, tufa formed between June and October 2003, (f) highly elongated crystals, tufa formed on Cu tablet between August 2002 and October 2003, dark laminae visible in a–c originated due to increased surface run-off during autumn rains; all samples from Lu´cˇky E point, a–e – thin sections, f – SEM image.
on the textures of tufa deposited in these variable conditions. The phenomenon that copper prevents algal colonization by being toxic to algae (e.g. Anagdostidis & Roussomoustakai 1988; Gorbushina & Krumbain 2000, fig. 2) was previously employed in several experiments with growth of tufa and
travertine. Chafetz et al. (1991a) regarded calcite deposited directly on copper as purely inorganic in origin. Emeis et al. (1987) did not notice calcium carbonate deposition on copper sponge immersed in tufa-depositing water in the Plitvice National Park, Croatia. Similarly, no tufa was deposited on copper plates installed in some hard-water
174
´ SKI M. GRADZIN
Fig. 23. (a) Intercrystalline porosity between sparry crystals, thin laminae are visible within the crystals (arrows), tufa formed on limestone tablet at Karwo´w cascade point between June and October 2003; (b) sparry crystals developed directly on copper tablet exposed between August 2002 and October 2003, note the high intercrystalline porosity and small pits in the crystals being traces of cyanobacteria or diatom stalks, sparite is capped by micrite, Ha´j dam point (c) compact sparry lamina formed directly on copper tablet, the smooth surface is the basement of sample visible due to detaching of the tablet, tufa formed at Karwo´w cascade point between June and October 2003; (d) sparry crystals developed on both sides of a leaf incorporated into tufa formed on limestone tablet at Lu´cˇky E point between June and October 2003, compare Figure 9b; a, d – thin sections, b, c – SEM images.
streams in Belgium by Janssen et al. (1999) and in Germany by Merz-Preiß & Riding (1999). Folk (1994) found travertine in a Tuscany spring growing considerably slower on copper than on other substrates. All these facts were ascribed to the toxicity of copper towards algae, and this view is accepted herein. The fact that the weight of tufa deposited on limestone tablets was substantially and persistently higher than on copper tablets exposed in the same conditions for the same time supports the idea that copper inhibits tufa growth (Figs 7 & 8; Table 3). The difference in growth rates was always significant when tufa deposited on a limestone tablet contained biogenic components. In extreme case copper tablets were not covered by tufa at all. This suggests that micro-organisms inhabiting the limestone tablets promote faster and more efficient tufa growth.
Other causes that can affect the difference between the rate of tufa growth on copper and limestone tablets are rate of nucleation and toxicity of copper ions in solution. Copper basement can be quickly overgrown by calcite, which has been proved in a series of experiments carried out by Chibowski et al. (2003). Nonetheless, in the same chemical conditions, calcite is prone to nucleate quicker on limestone than on copper because of structural similarity between the nucleating phase and its substratum (see Bathurst 1976, 437). The difference in the weight of tufa formed on paired limestone and copper tablets may be used for estimation of this process only when tufa covering both tablets displays crystalline fabrics without substantial traces of algae, that is when any significant influence of algae on tufa growth can be ruled out. This is the case at the Lu´cˇky E site. Tufa growing there on both tablets constituting a pair between
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
175
Fig. 24. (a) Compact and porous laminae formed by highly encrusted cyanobacterial filaments, tufa formed on limestone tablet exposed between August 2002 and October 2003; (b) crystal sizes asymmetry in compact laminae, tufa formed on Cu tablet exposed between August 2002 and October 2003; (c) Alternating compact and porous laminae, tufa formed on limestone tablet between August 2002 and October 2003; (d) compact laminae comprising cemented cyanobacterial filaments, the same sample as in c; all samples from Lu´cˇky E point, a, b – thin sections; c, d – SEM images.
November 2002 and March 2003 exhibits exactly the same texture, even in terms of sequence of laminae (Fig. 22a, b). In this case the difference in growth rates is relatively small. Pentecost (2005, 198), commenting on the earlier experiments on tufa and travertine growth, argued that in some conditions copper ions can inhibit inorganic calcite precipitation. Indeed, such a phenomenon was experimentally confirmed (Parsiegla & Katz 1999, 2000). However, one can rule out the importance of Cu ions in the conducted experiments because if any Cu ions were liberated during a possible dissolution of copper tablets, they could be instantly removed by flowing water. The higher concentration is possible only within a
diffusion boundary layer adjacent to the copper surface. In the stream conditions, the thickness of such a layer may be estimated at 0.01 cm (Liu et al. 1995). The thickness of this layer decreases or the layer even disappears over elevations of the substrate. Thus, any freshly precipitated carbonate patches protruding from the copper surface reduce the thickness of the diffusion boundary layer, and become a favourable place for algae to settle. As a consequence, algae are able to colonize small patches of calcite previously crystallized on a copper tablet, even if calcite does not form a continuous layer insulating entirely the copper substratum. Hence, algae can grow in close vicinity to bare copper surface. It suggests that, in the studied
176
´ SKI M. GRADZIN
case, a key inhibitor of the algal growth is the direct contact with copper substrate rather than the amount of toxic copper ions in bulk water. The longer the exposition time of a pair of tablets, the smaller is the difference in weight of growing tufa between the copper and limestone tablet. This implies that processes immediately following immersion of the tablets are of crucial importance. Microscopic observations shed some light on those processes. Tufa deposited directly on copper substrata is always built of calcite crystals without any traces of algae, such as moulds of filamentous algae or diatom frustules, which is a result of the copper toxicity to algae discussed above (Figs 13a, 14a & 23c). Tufa with biofabrics if appears, it does it only 10 –20 mm or more over the copper surface. Conversely, tufa with biogenic components can directly overlie the surface of a limestone tablet. Thus, just after immersion of a pair of tablets different processes take place on the limestone and copper ones. Algae can start to settle the limestone tablet, which promotes rapid tufa growth, whereas algal colonization and growth on a copper tablet is always delayed until the protective calcite layer or patches crystallize inorganically. The amount of tufa which grew on limestone tablets exposed for more than one year almost always exceeds the total amount of tufa originating on tablets seasonally exposed in the same place (Fig. 25). It points out that after placing, every tablet experienced some delay in deposition. Every tablet is subjected to the delay only once, however the effects of particular delays sum up when calculating the total amount of tufa on four seasonally exposed tablets. A similar phenomenon was noted by Pentecost & Coletta (2007) at La Zitelle spring in Italy. They called it ‘lag period’ and explained
Fig. 25. Comparison between tufa increment on limestone tablets exposed during the whole experiment period and total increment on tablets exposed seasonally.
it by a slower initial rate of growth due to difference in crystal sizes between marble substrate and overgrowing travertine, which affected nucleation process. Nucleation of calcite crystals can also play a role in the observed delay, but its effect probably overlaps with the effect of the rate of colonization by algae which actively or passively intensify tufa growth. Any bare surface immersed in stream water experiences algal colonization, due to immigration from surrounding algal colonies, mainly those located upstream. It was studied especially with regard to recolonization after severe floods that scoured older algal community (e.g. Fisher et al. 1982). The rate of succession depends on several factors, such as temperature, light intensity, roughness of surface and adaptation of particular species. It takes a few weeks before a bare surface is covered by a stable epiphytic community (Korte & Blinn 1983; Peterson & Hoagland 1990; Kralj et al. 2006). As a consequence, only inorganic precipitation can operate between the immersion of a tablet and the moment, when a tablet is colonized by at least a pioneer community. Bacteria can quickly settle on bare substrata (Goulder 1988). As for algae, diatoms are the most frequent pioneer colonizers; they are followed by cyanobacteria and other algae (Fisher et al. 1982; Rushford et al. 1986; Winsborough et al. 1994). Colonization by Vaucheria lasts a four-week period (Wilde 1982). A similar succession has been recorded in the material obtained during the present experiment. In some of the tablets the first layer is composed of sparry calcite crystals with only scarce traces of algae with the diatoms or traces of their stalks being more common in this layer (Fig. 13a, b). Further up from the substrate, encrusted moulds of filamentous algae abound. The prolonged succession of algae, along with rapid inorganically driven crystal growth, probably accounts for inorganic texture obtained during some earlier experiments. For example Liu et al. (1995, fig. 5) observed that a tablet was covered with calcite crystals without any traces of algae. However, the immersion period was only 5 days, which, in the light of the present discussion, was insufficient for algae to colonize the substrate. During the present experiment, most of the copper tablets were covered with calcium carbonate though four tablets remained uncovered. Similarly, in earlier experiments reported in literature some copper surfaces were covered with calcium carbonate and other were not. Thus, conditions favouring the overgrowth of copper tablets by tufa need to be discussed. It is significant that during the experiment only the tablets at the Za´zriva´ site exposed during the winter season and at the Karwo´w dam point exposed during the winter–spring season were not covered with calcite at all. Earlier
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
experiments conducted in thermal water gave other results; rapid formation of carbonate minerals on copper substratum was noted. Chafetz et al. (1991a) and Chafetz & Lawrence (1994) observed calcium carbonate precipitation on copper substratum in hot-springs in USA and Italy. Similarly, Folk (1994) in La Zitelle spring found that carbonate crust 0.2–0.3 mm thick grew on a copper tablet within 22 hours. SIcalc. is the main parameter differing the previously studied sites from those where the encrustation of copper was not discerned throughout the recent experiment. Water at the studied sites has low SIcalc. values: 0.35 at Karwo´w and 0.55 at Za´zriva´, in the late autumn of 2002 (Table 2). This parameter is highly elevated in the case of previously studied thermal travertine sites, where it reaches 1.8 (Chafetz et al. 1991a). The results of experiments carried out in ambient temperature water are ambiguous. Emeis et al. (1987) and Malusa et al. (2003) did not observe precipitation on copper sponges in the Plitvice lake system and Fossil Creek in Arizona, respectively. Conversely, Lojen et al. (2004) detected tufa growth on copper substratum during an experiment carried out in the Krka river geographically close to Plitvice. The SIcalc. values of water were similar in all the sites. Thus, the different times of exposition could be responsible for differences. Exposure lasted a few weeks in Plitvice and around three months in Krka. Similarly, a copper substrate was covered with travertine (tufa in terminology adopted here) in the course of three-month experiments carried out by Chafetz et al. (1991b) in the Honey and Falls creeks in Oklahoma. During an experiment conducted by Janssen et al. (1999) in Belgium no carbonate precipitates were detected on copper plates placed in tufa-depositing streams. It might have been caused by short exposition time and/or by low SIcalc. values which at some locations were as low as 0.44. Thus, two factors appear to be relevant for inorganic precipitation of carbonate minerals on copper substratum – time of exposition and SIcalc. value. Both of them stimulate precipitation of an insulating layer or at least of small patches of carbonate directly on copper substratum, essential for further algal colonization. Unless the insulating calcite is present, algal colonization is not possible.
Differences between tufa growing at various settings The location of various study points along the course of the same stream allowed to observe differences in the tufa growth and texture related to microenvironmental conditions. All the study sites, except for Za´zriva´, provide a good opportunity to discuss this difference.
177
Growth rates of tufa deposited during the same time in various points differed substantially. A strong tendency is present to faster deposition of tufa in high-energy settings. An impressive example is provided by the Karwo´w site. The annual rate of tufa growth at the dam point is 0.09 mg/cm2/day while at the cascade, only a few metres downstream, the rate is 1.93 mg/cm2/day (Table 3). Tufa at both points includes biogenic components. At the dam it has hemispherically layered texture while in the cascade the texture is mainly hemispherically layered and fibrous. Similar difference, but not so profound, was observed at the Ha´j site, between the lower waterfall and dam points. The net rates of growth equalled 0.69 mg/cm2/day and 0.34 mg/cm2/day, respectively. The rate at the upper waterfall point is slightly lower than that at the dam, but the waterfall is located in the upper reaches of the stream, close to the spring, hence the rate of tufa growth may be affected be the length of flow too short for water to degas (cf. Michaelis et al. 1985). Moreover, the stream between the upper and lower waterfalls is additionally fed by other springs, which may change the water chemistry and make a meaningful comparison impossible. At the Lu´cˇky site, higher annual growth of 6.43 mg/cm2/day was found at the Lu´cˇky E point, that is on the waterfall face, where the tablets were constantly flushed by water falling down. In the same time period, the tufa growth rate at the Lu´cˇky top point, located in the streambed just over the waterfall head, was substantially slower and reached 3.86 mg/cm2/day. The examples mentioned above depict semiquantitatively the tendency to a faster growth of calcite in settings with fast flow than in calm ones. A similar tendency was earlier detected by Liu et al. (1995) and Lu et al. (2000) in the famous Huanglong ravine in China, by Drysdale & Gillieson (1997) in Louie Creek in northern Australia as well as by Bono et al. (2001) in Tartare karst spring in Italy. Liu et al. (1995) explained this phenomenon by the smaller thickness of a diffusionboundary layer in fast-flowing water, which facilitates ion migration toward the growing crystals. This phenomenon occurs particularly in waterfall settings. Zhang et al. (2001) coined the term ‘waterfall effect’ to describe it, while Chen et al. (2004) put forward aeration, low pressure and jet-flow effects as stimulators of calcite precipitation. Recently Hammer et al. (2008) have shown that the most decisive mechanism governing the calcite growth in flowing water is advection of ions towards the crystal surface. All these mechanisms are based on inorganically driven processes. However, the tufa textures imply that biogenic activity may also influence tufa growth in moderate and
178
´ SKI M. GRADZIN
even high-energy settings. The essence of this biologically driven mechanism is explained in the following sections. Apart from energy, the mode of water supply seems to be relevant to the rate of tufa growth. This is evident at the Lu´cˇky waterfall. The tablets placed in the south-facing wall of the waterfall (Lu´cˇky W point) obtained slowly seeping water and periodically intense spray. The tufa growth rate on these tablets was slower by a factor of 2 than on the tablets exposed on the vertical face in the eastern part of a waterfall (Lu´cˇky E point), constantly flushed by water falling freely down. Drysdale & Gillieson (1997), while studying Loui Creek tufa in Australia, observed a similar tendency to the strong dependence of the tufa growth rate on water-flow regime. Differences in textures were also observed between tufas deposited in close vicinity, between the tufa deposited on tablets exposed throughout the whole period of experiment and on tablets exposed seasonally. The Lu´cˇky E point provides an example. The younger part of the tufa deposited on the limestone tablet exposed there between August 2002 and October 2003 is built of laminated tufa of fibrous texture with highly encrusted cyanobacterial filaments (Fig. 9c). In contrast, the limestone tablet exposed at the same point from June 2003 to October 2003 is covered exclusively with sparry calcite (Figs 9b & 22d, e). Both tablets were exposed close to each other in almost the same conditions during more than three months. Despite the fact that the distance between their locations was only 35 cm, the tufas differ in texture (compare Fig. 9b, c). A similar difference was noted at the Karwo´w cascade site. Tufa with fibrous texture was deposited in spring and probably in the beginning of summer of 2003 on the tablet exposed there throughout the whole experiment. During the same time, tufa with hemispherically layered micrite and with sparry fringe cement with entombed algal filaments was formed on the seasonally exposed tablets located in close vicinity. Hence, the difference in texture may be explained by physical conditions differing over a very small distance or by reduced capacity of algae to colonize a fresh tablet because of the relatively fast growth of crystals. The latter possibility is considered to be more probable.
Water chemistry, tufa growth rate and algal calcification There is a general tendency that the higher SIcalc. value is, the faster tufa grows. However, correlation between these two factors is weak (Fig. 26) for two reasons. Firstly, SIcalc. values were calculated for stations located at some distance upstream of the
Fig. 26. Relationship between tufa growth rate and SIcalc. value of feeding water, used SI calc. value is calculated as a mean values of the tablet exposure time.
tablet location. Stream environment is very unstable from the chemical point of view, hence the measured chemical parameters may be slightly different from those at the point where tufa samples grew. This reason seems to be essential, though one can also take into consideration that chemical analyses of water were carried out only four or five times a year. Thus, all the above parameters are discrete values, extrapolated to the whole period of experiment, while the real parameters are unknown. Nonetheless, it seems that any departures from real values are not significant. Where traces of filamentous algae are visible in any of the described textures, always perfectly preserved moulds of filaments are present. This implies that calcification proceeded outside the cyanobacterial sheaths. It was thus, external surficial precipitation sensu Riding (1991), typical of freshwater environment with SIcalc. higher than 0.8 (Merz-Preiß & Riding 1999). A similar value is accepted as the threshold for effective precipitation of calcium carbonate in freshwater stream environments (Michaelis et al. 1985; Herman & Lorah 1987) and in soda lakes (Kempe & Kaz´mierczak 1990, 1993). In the studied sites SIcalc. values around 0.8 and higher favoured the origin of sparry crystals or fibrous texture. However, the experiment shows that the values lower than 0.8 also promote external surficial precipitation of calcite around cyanobacterial filaments. This was the case at the Karwo´w site in the summer and autumn of 2002, when SIcalc. systematically decreased from 0.45 to 0.35 (Table 2). In this case, crystals precipitated around the
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
filaments are micritic rather than sparitic in size. Nonetheless, the filament moulds have smooth walls, regular cross section and constant diameters. This strongly suggests that the moulds in question are a product of calcification around and not within cyanobacterial sheaths, that is of encrustation, not impregnation type (cf. Merz 1992). Merz-Preiß & Riding (1999) stated, based on the study of recent tufa in Germany and on previous observation carried out by Merz (1992) in the Everglades, that encrustation takes place under SIcalc. .0.8 while impregnation proceeds under SIcalc. ,0.3. Therefore, the present data imply that the former process may take place in SIcalc. range slightly lower than previously postulated. The sparitic crystals in the discussed tufa are interpreted as precipitated from solution. They do not seem to be formed by agradational neomporhism as described by Love & Chafetz (1988). This interpretation is strongly supported by the competitive crystal growth fabrics displayed by sparitic crystals in the studied samples (Fig. 22a –e). The crystal fans in some samples are laterally separated by micrite with algal remnants, which suggests that both types of tufa grew simultaneously (Fig. 18d, e; cf. Jones & Renaut 1994). SIcalc. values exceeding 0.8 make the growth of sparitic crystals plausible, as it is uncommon under lower SIcalc.. The elongation of sparry crystals changes along with the changes in SI values. Higher SI values favour the origin of highly elongated crystals. The tufa sample that grew on a limestone tablet at the Lu´cˇky E point between June and October 2003 provides a good example (Figs 9b & 22d, e). It is built of highly elongated crystals arranged in closely coalescent fans. This tablet was installed when SIcalc. reached 1.27, that is the highest value recorded during the experiment (Table 2). Other samples grown at the same point under lower SIcalc. between 0.82 and 0.94 display crystals of shorter elongation also grouped in tightly packed fans (Fig. 22a– c). This is in line with the opinion of Given & Wilkinson (1985) that growth rate of crystals governs the crystal nature. The molar Mg/Ca ratio appears to be unimportant because this parameter is constantly low. It varied during the exposition of the tablets within a small range: between 0.29 and 0.42 at station L3. Gonza´lez et al. (1992) claimed that changing flow velocity determined speleothem textures and that fast flow conditions caused formation of fibrous fabrics of highly elongated crystals. However, this rule does not explain the textural differences between particular samples from the Lu´cˇky E point, because all the tablets in this point were in comparable conditions of constant and intense water flushing. The elongated sparry crystals grew at the Karwo´w and Ha´j sites under SIcalc. values of
179
0.35 –0.88 and 0.7 –1.08, respectively (Fig. 23a, b). Tufa composed of these crystals has high intercrystalline porosity. Thus, the tendency that higher SIcalc. values support grow of compact tufa with lower intercystalline porosity (Kano et al. 2003) seems to be confirmed. The extremely high SIcalc., probably around 1 or more, enables nucleation of calcite crystals directly on diatom frustules, which is a rare phenomenon (Fig. 15d; cf. Merz-Preiß & Riding 1999; Lu et al. 2000). One can suppose that such a high supersaturation allowed to break the barrier resulting from structural differences between the silica nucleation surface and the carbonate precipitate. The growth of calcite tubes of Oocardium stratum and formation of sparitic bush texture took place under relatively low SIcalc. values, which only once – in June 2003 – reached 0.78, but otherwise did not exceed 0.55 (Table 2). In spite of rather widespread occurrence of this alga in European hard-water streams, the data on chemistry of its growth environment are scarce. Pentecost (1991) quoted the values of SIcalc. between 0.28 –0.81, which correspond well to the present data.
Environmental factors controlling tufa textures The origin of tufa textures is controlled by a number of factors interacting in a complex way. The principal factors and their interactions are characterized below. Water energy. The distribution of tufa facies points to a clear relationship between facies and local environmental conditions. The most important of them is energy of flow, which not only governs the rate of tufa growth, as it is discussed above, but also exerts a fundamental control on the tufa texture. Sparry crystals are the most common texture on the tablets exposed in extreme energy on the vertical face of the Lu´cˇky waterfall, regardless of the type of tablet (Figs 9b, c, 11d, 12a, 22 & 23d). By contrast, some samples of tufa deposited at the same time on horizontal streambed, close to the waterfall head, display miciritic texture with abundant algal remnants (Fig. 18a, b). Similarly, the tufa growing at the waterfalls in Ha´j displays abundant sparry textures. In a high-energy setting the growth of microorganisms inhabiting the streams is impeded, which enables formation of sparry calcite. This relationship was earlier pointed out by Pedley (2000). Highvelocity flow causes shear stress at the bottom which prevents colonization by micro-organisms. The experimental work by Horner et al. (1990) implied that at flow velocity over 80 cm/s algal periphyton
180
´ SKI M. GRADZIN
is efficiently removed. Pedley (1992), following the experimental work by Lorch & Ottow (1985), stated that even bacterial biofilm can be damaged and scrubbed off when water flow is too fast. Mineral grains transported in traction substantially enhance this process. Tufa deposited on the tablets exposed in waterfalls and in fast-flowing sections of the streams, includes also commonly porous tufa composed of highly encrusted algal filaments. Orientation of the filaments is governed by energy of the stream. In higher energy conditions, the filaments are aligned with the flow, parallel to each other and more or less parallel to their substrate. Both Vaucheria and cyanobacteria filaments display such orientation (Figs 10e & 20a, e). The dependence of Vaucheria filament orientation on the current direction was noted earlier (Wallner 1934b; Golubic´ et al. 2008). It seems to be a rule applying to all filamentous algae. Whitton (1975), for instance, described Cladophora, an alga commonly found in tufadepositing milieu, forming elongated structures, exhibiting parallel orientation to the flow, in conditions of higher current velocity. Some filaments in the studied samples form laminae loosely cemented with the basement, which indicates that cementation proceeded within a cyanobacterial or algal mat partly floating in the current. This was probably forced by shear stress of flowing water and vulnerability of the upper part of the algal mat to dislodgement because the lower part of the mat looses its ability to support the overlying part, as a result of senescence, shading or metabolic processes (Allan 1995, 96). The tufa which grew on the tablets at the Lu´cˇky E point exhibits such a structure (Fig. 24). Conversely, in a lower-energy setting, such as the Lu´cˇky top and W points or in places where water is supplied only as small rivulets, as at the Karwo´w cascade, the filaments are not stuck by water, which results in their random orientation. Tufa rich in micrite forms in low energy settings. It displays clotted or hemispherically layered texture. In the former case the micritic crystals grow within or on the EPS. Hence, its origin probably reflects conditions suitable for production of the copious amount of EPS (Pedley 1992, 2000). The diatoms which accompanied the EPS are probably their main producer (Winsborough 2000). However, diatoms were detected also in other samples. They accompany the filaments that build the porous tufa growing in fast-flowing water. Diatoms are able to withstand high-velocity currents; hence they also inhabit high-energy milieus (Allan 1995, 95). Therefore, the presence of diatoms is not limited to the conditions in which micrite tufa forms (see Winsborough & Golubic´ 1987; Arp et al. 2001).
The extreme situation favouring the origin of clotted micrite rich in diatoms and their EPS is slowing down of water flow, coupled with systematic drying up of the stream. Such a process took place in the summer of 2003, when the upper segment of the stream in Ha´j dried up completely and the water level in its lower segment lowered substantially, resulting in subaerial exposition of tablets at the Ha´j upper waterfall and dam points. The occurrence of layers abounding in diatoms formed during the drought was noted from Cuatro Cie´negas area in Mexico by Winsborough & Golubic´ (1987). The growth form of upward-radiating filaments of cyanobacteria, most probably Phormidium sp., controls the hemispherically layered texture (Figs 11a, 12d–f, 19). It dominated at the Karwo´w dam and Karwo´w cascade points where minute water rivulets seeped or trickled down. Thus, this textural type seems to be connected with different energy conditions but in a relatively small stream. However, it occurs also commonly in bigger streams (Golubic et al. 1993; Freytet & Plet 1996; Freytet & Verrecchia 1998 and references herein). Interestingly, the sparite bushes connected with the calcification of Oocardium stratum originate in sluggish flow condition. At the earlier described localities, this desmid alga inhabits fast-flowing water. Wallner (1935b) mentioned, based on his experiments, that in the sluggish flow colonization of Oocardium stratum was not observed. Conversely, he found vigorous growth of Oocardium tufa in relatively fast-flowing water. Similarly, Pentecost (1991), describing two sites with Oocardium stratum from Great Britain and one from Belgium, pointed out that each of them experienced moderately fast flow. Temperature. The temperature of water influences principal physicochemical conditions governing SIcalc. values, which in turn rules the tufa growth rate (Pedley et al. 1996; Pentecost 2005). However, the temperature at all the studied sites changes within narrow range and thus any effect of temperature may be overridden by the influence of other factors (Merz-Preiß & Riding 1999). Nevertheless, temperature increase probably contributes to SIcalc. elevation in summer seasons, for instance at the Za´zriva´ site (Fig. 4). Conversely, the substantial decrease of SIcalc. at this site in cold seasons resulted probably from temperature decline. This was possible because of the relatively small amount of water feeding the discussed site. In some cases, higher temperatures between spring and autumn may favour development of tufa with abundant moulds of cyanobacterial or other organism (except diatoms), because these algae seem to prefer slightly higher temperature
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
(Whitton 1975). The Ha´j dam site may serve as an example. Light intensity. The light intensity is decisive for biological activity in streams. Particular taxonomic groups demand different light intensity, which is an important factor controlling the distribution of algae in streams (Canfield & Hoyer 1988; Allan 1995, 88–89). Therefore, the amount of available light exerts a fundamental control on the tufa texture. Bearing in mind annual changes of day-time length at the European latitudes, the shortest period of insolation is in winter. However, all the studied sites are shaded by the vegetation which consists mainly of deciduous species. This effect decreases solar irradiance in summer. Thus, the overlapping of these two effects causes that the best light condition is in spring just before leaves fully develop (McIntire 1968; Pentecost 2003). Vaucheria is especially vulnerable to low intensity of light (Ensminger et al. 2005). This clearly explains the superabundance of tufa displaying fibrous texture and built of this alga in spring at the Karwo´w and Lu´cˇky sites. At Lu´cˇky this type of tufa developed also in summer and autumn, but only on the tablets installed at the south-facing wall of the Lu´cˇky waterfall (Lu´cˇky W point), that is in the well irradiated location. At other points and at Karwo´w it ceased to grow. Green algae and cyanobacteria are more tolerant to shading (Merz 1992; Ensminger et al. 2005). This results in the development of tufa with cyanobacterial moulds at sites exposed to twilight, located in narrow and relatively deep gorges, as for example at the Ha´j site and to predominance of this type of tufa during summer season in streams shaded by leaf canopies, as at the Karwo´w site and at the Lu´cˇky top and E points. Formation of tufa with diatom moulds seems to be independent from light conditions as a result of the tolerance of diatoms to low light intensity (McIntire 1968). Moreover, diatoms probably can cope well with changing irradiation (Allan 1995, 68). Therefore, their distribution is independent of light conditions and they dominate in tufa deposited in low-light conditions, inhospitable to other algae. Rate of calcite growth. The rate of calcite growth obviously depends on the chemistry of parent solution, as discussed above. However, the results of the experiment described here prove that calcite growth rate, in turn, exerts a significant influence on the texture of deposits by controlling diversity and density of algal communities. A similar idea was previously put forward by Pedley (1992, 2000). The high rate of tufa growth favours the crystalline texture. The Lu´cˇky site provides a good example. The extremely high rate of tufa growth
181
ever detected during the described experiments was at the Lu´cˇky site in front of the waterfall (Lu´cˇky E point) between July–October 2003. The layer of compact tufa up to 17.1 mm thick, displaying crystalline texture, was deposited there (Fig. 9b). The mean rate of tufa growth equals there 10.7 mg/ cm2/day or 0.142 mm/day. The real rate just after placing the tablet was probably lower, bearing in mind the ‘lag period’ resulting from nucleation process (Pentecost & Coletta 2007). Nonetheless, such a high rate of inorganically driven crystal growth significantly impedes growth of algal flora (Golubic´ 1969; Kano & Fuji 2000). Any alga which appears on the crystal surface is instantly covered with calcite, either being trapped by expanded older crystals or providing an ideal place for nucleation of the new ones (Figs 11b–f & 12a, b). Hence, attempts of colonization are abortive; their traces are visible as moulds of filamentous algae, completely entombed within sparry calcite crystals. Impossibility of colonization due to high rate of carbonate mineral growth results in a scarcity of algal structures in travertines. There it works in combination with other factors as for instance temperature and toxic elements dissolved in water. Conversely, under the condition of low crystal growth rate, tufa built of clotted micrite or hemisperically layered micrite textures was deposited. For example, at the Karwo´w cascade between December 2002– March 2003, when the growth rate equalled 0.18 mg/cm2/day, tufa built of clotted micrite was formed. Similar tufa originated at the Ha´j dam point during spring and summer 2003. The occurrence of tufa with fibrous texture is intimately connected with high rate of tufa growth. The samples from the Karwo´w cascade point and those from the Lu´cˇky top and E points exhibit such a texture. The tufa growth rate of these samples was 1.93, 3.87 and 6.43 mg/cm2/day respectively. The samples with fibrous texture formed in settings with relatively high-energy, but not under extreme energy conditions. At the Lu´cˇky top point they were formed in a fast-flowing stream while at Karwo´w on the face of a small cascade. Formation of this texture at the Lu´cˇky E point, where water falls freely along the vertical wall, needs further explanation. Energy there seems to be unfavourable for algae to colonize. However, the separation of flow at the head of the waterfall caused a shadow zone between the wall and water in the place where the tablets were installed. Although water constantly flushed the tablets, conditions were not so severe to completely preclude algal colonization. Tufa with fibrous texture originates due to external calcification on filamentous algae. This process isolates the organisms from their environment, and hence impedes the access of nutrient and light and,
182
´ SKI M. GRADZIN
eventually, makes the basic life processes impossible (Golubic´ 1969; Kano & Fuji 2000). However, the filamentous organisms due to their apical growth may effectively emerge from a calcified zone. This process is of crucial importance for the origin of fibrous texture. Vaucheria appears to be especially prone to constructing such a texture. This alga can colonize fast-flowing water where calcification proceeds at a high rate. Moreover, it is particularly prone to being forced to faster growth and ‘escaping’ because of its susceptibility to decreased irradiance (see Ensminger et al. 2005). Such a texture built of filamentous cyanobacteria was detected in samples characterized by a high rate of tufa growth from the Lu´cˇky site. The texture may be attributed to higher photsynthetic plasticity of cyanobacteria (Pentecost 1978; Merz 1992) which require a relatively thick calcite cover to be forced to grow. Tufa with fibrous texture includes only scarce diatom frustules, while coccoid cyanobacteria are absent. For these algae, entombing with calcite may be critical as they do not show the ability to ‘escape’ when being covered with calcite crystals. The traces of diatoms are probably only the effect of abortive colonization of growing crystals by these algae as the first colonizers (Fisher et al. 1982) that did not contribute to the formation of the texture itself. It is noteworthy that algal growth is particularly vigorous in such settings where a reduction of diffusion boundary layer facilitates access to dissolved nutrients for algae (Horner et al. 1990; Raven 1992). At the same places inorganic precipitation effectively proceeds (Liu et al. 1995; Zhang et al. 2001). This implies that a kind of specific positive feedback exists between the growth rate of algae and the rate of calcification. The proceeding calcification encrusts the algae and forces them to ‘escape’ by faster growth (Kano et al. 2003). In turn, the faster growth of algae promotes the faster growth of tufa, which not only influences its texture and thickness, but also affects its mass increase by several coexisting mechanisms mentioned below. The appearance of new parts of algal filaments solely augments the surface of the new substrate for calcite nucleation. A pierced algal construction constantly flushed by water disturbs the water flow, accelerates outgassing of CO2 and in some cases also leads to evaporation and to reduced diffusion boundary layer. Each of these processes stimulates calcite precipitation. Moreover, many authors claim that algae and their EPS may in several ways passively stimulate calcite growth (Emeis et al. 1987; Riding 2000; Turner & Jones 2005). Finally, recent experimental studies by Shiraishi et al. (2008) and Pedley et al. (2009) demonstrate that algae actively influence calcite growth. The influence of the above feedback
mechanism on the growth rate of tufa is indirectly confirmed by the unexpected growth rate – 1.93 mg/cm2/day – of the tufa sample with fibrous texture that formed at the Karwo´w cascade site during the whole period of the experiment. Similarly, the samples from the Lu´cˇky site, which show fibrous texture, grew at the rate between 0.87 –6.43 mg/cm2/day. Hence, the experiment invites an interpretation that the tufa exhibiting fibrous texture, which is a product of the discussed feedback mechanism, was formed mainly under high SIcalc. values. Such a tufa did not originate either in Za´zriva´ or in Ha´j sites, where SIcalc. values exceeding 0.78 and 1.08 were never detected, but it grew at the Karwo´w and Lu´cˇky sites under SIcalc. values of 0.88 and .1.2. Hence, conditions favourable for algae to grow are also needed, that is light intensity and temperature. In other cases, under the same SIcalc. values tufa showing other texture grows. Tufa with fibrous texture was formed predominantly on the tablets exposed all year around, for instance at the Karwo´w and Lu´cˇky sites. At the same sites the tufa formed at the same points on seasonally exposed tablets displays crystal texture containing only some entombed algal filaments. This suggests that the high SIcalc. values, causing high crystal growth rate, which favour formation of fibrous texture, preclude algal colonization on a newly installed tablet. Hence, the feedback mechanism presumably demands earlier colonization of a given area by algae when conditions are not so severe and crystal growth is a little slower. The feedback mechanism works particularly efficiently as long as the rate of calcite nucleation and crystallization is equal or slightly lower than the rate of algal growth. When calcite growth accelerates, algal filaments become entombed completely, and sparry calcite without biogenic texture grows on them (see Pentecost 1998). Deposition may switch to pure sparite growth by increase in values of SIcalc. caused by some independent factors, such as higher temperature and higher flow rate, which stimulate degassing, or by various factors impeding algal growth. Moss tufa grows in an analogous manner, as postulated by Pentecost (1998). The feedback mechanism probably resulted in exceptionally fast growth rate of moss tufa of 140 mm a21 noted by Weijermars et al. (1986) from Spain. The above described feedback mechanism results in the formation of fibrous tufa facies, with encrusted calcitized algal filaments, known from several modern and fossil localities (Fig. 20c; e.g. Irion & Mu¨ller 1968; Szulc 1983). This facies, composed of long and friable fibres, which in some cases are not completely cemented together, evinces the tendency to be easily eroded. Thus, it may provide
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
small sand- or silt-sized intraclasts, which after redeposition may constitute important component of other tufa facies or build independently intraclast tufa deposits (sensu Pedley 1990).
Seasonality of tufa growth The seasonality of tufa growth can manifest itself in two ways. The first is seasonally varying rate of growth recorded as varying mass of tufa deposited in different seasons, the second is seasonal alternation of tufa textures. The experiments described here shed some light on both manifestations of seasonality. Seasonal mass increment. The comparison of the tufa amount which grew during particular seasons in particular study sites lead to the conclusion that there is no one common trend in tufa growth rate (Table 3, Figs 7, 8). At the Za´zriva´ and Karwo´w study sites the greatest mass increment was recorded in the summer of 2003. At two points of the Ha´j site more tufa grew during the winter of 2002 –2003, while in the third point tufa grown in summer 2003 slightly prevails. At all points studied at the Lu´cˇky site, overwhelmingly dominates the tufa increment from the summer 2003. This points to the importance of local factors, specific for particular sites, which effectively govern the tufa growth rate. At the Lu´cˇky E point tufa grew at a relatively low rate – only 6.77 mg/cm2/day – between August– November 2002, while between June– October 2003 the growth rate was as high as 10.71 mg/ cm2/day. At other points at Lu´cˇky waterfall the trend is similar regardless the tablet type (Table 3). Changing water chemistry can fully account for the above difference. The SIcalc. value for the water feeding the studied points in summer 2003 reached 1.27, the highest value ever recorded for this point. The elevated SIcalc. is associated with elevated TDS value. It is adequately explained by the changing ratio of deep and shallow contribution to stream water that feeds the Lu´cˇky waterfall. The change was probably due to a weather anomaly; the summer 2003 was exceptionally dry and hot in central Europe, with the amount of rainfall lower than average (Scha¨r & Jendritzky 2004). Hence, the proportion of deep water increased, altering the chemistry of feeding water, and stimulating vigorous growth of tufa. At the Ha´j site, the rate of tufa growth at both studied points in the lower segment of the stream was highest in the winter season 2002–2003. This was independent of SIcalc. values calculated for November 2002 and March 2003, which are slightly lower than those for the summer season. The vigorous tufa growth in the winter season coincides with a
183
relatively high SIcalc. value. However, the SIcalc. reached higher values in summer 2003, when the rate of growth was slightly slower. Hence, the discussed phenomenon is hard to envisage. One possible explanation is the specific winter hydrological condition in the Slovak Karst area. Although the Ha´j site is characterized by continental climate, with maximum concentration of rains in spring and summer, evapotranspiration causes that the water flow in stream may be higher in cold season (Cˇerma´k 1994). The lower growth rate of tufa at the Ha´j dam site in summer 2003 was caused by periodical emersion of the tablet during reduced flow conditions in summer 2003. Water flew only through the spillway in the tufa dam, leaving the rest of the crest, where the tablets were placed, emerged (Fig. 3g). The seasonal tufa growth rate and SIcalc. changes are generally covariant at the Karwo´w site. The SIcalc. calculated for the station just upstream of the studied points reached maximal value of 0.94 in June 2003, which is consistent with maximum growth rate of tufa in summer 2003. The SIcalc. trend is generally similar at the Za´zriva´ site with the maximum value in late June. It is seemingly contradictory to the tufa growth rate, which implies the maximal growth in autumn 2002. However, in the spring of 2003 the seasonally exposed tablet experienced significant period of corrosion followed by precipitation, which is documented by newly formed tufa deposited on corroded surface of the limestone tablet. Moreover, the water level in the rivulet feeding the tablet between June and October 2003 fall drastically, which resulted in drying up of the limestone tablet (Fig. 6b). Thus, the growth rate obtained for the period June – October 2003 is significantly underestimated. To sum up, the only one general rule detected during the experiment was the response of tufa depositing streams to exceptional weather conditions in summer 2003. Central Europe experienced prolonged drought and hot weather (Scha¨r & Jendritzky 2004). This resulted in the elevated SIcalc. values in all studied streams, regardless of the type of water supplying the tufa and in the episode of more efficient tufa growth. This is not valid for the points where tablets were emerged after the fall of water level, that is for the Za´zriva´ site and the Ha´j upper waterfall and dam points. Seasonal textural sequence. The textures of growing tufa, as was shown above are controlled by closely interrelated factors like temperature, irradiance, biological activity and hydrology: especially intensity of a water flow. All these factors change in the course of a year. One could thus expect a similar sequence of tufa textures in a one-year increment in all the studied sites, reflecting the periodical
184
´ SKI M. GRADZIN
changes. However, there is no one common sequence of textures in the studied samples but instead, the sequences of tufa textures deposited on the tablets exposed throughout the whole experimental period in each of the sites differ and are characteristic for each site. This proves that deposition of tufa at each site is governed by a different local factor. Nonetheless, some regularity is present in the seasonal sequence of textures. Laminae built of sparry calcite crystals were commonly formed in cold season. The Ha´j and Lu´cˇky sites provide good examples. Similar regularity was noted in tufas in Great Britain (Pedley 1992). Vigorous water flow in winter, along with algal vegetation impeded by low irradiance and low temperature, may lead to formation of tufa composed of sparry calcite crystals. However, the formation of sparry calcite crystals is strongly controlled by local environment as they form in fast-flowing settings almost irrespectively of season, especially if biological activity is impeded. Conversely, moulds of filamentous algae mark periods when algae thrived. Thus, tufa abounding in such components originates mainly in spring and summer. Despite of this, the overshadowing of the streams in summer may lead to deposition of a tufa built of sparry calicte or to a change from deposition of Vaucheria-dominated to cyanobacteriadominated tufa, as recorded at the Karwo´w cascade point and at the Lu´cˇky top and E points. Tufa with micritic texture rich in diatoms is commonly deposited in cold seasons as at the Ha´j dam point at the turn of winter and spring of 2003. Deposition of such tufa is favoured by the predominance of diatoms. A similar phenomenon was reported earlier from different regions of Europe. For instance, Szulc & Smyk (1994) noted predominance of bacteria and diatoms in winter laminae of tufa stromatolites in southern Poland. Analogously, Arp et al. (2001) found diatom-rich biofilm in winter in the tufa-depositing Deinschwanger Bach in the Franconian Alb. A similar conclusion was reached by Plenkovic´-Moraj et al. (2002) who studied the Plitvice lake system. Sabater et al. (2000), investigating La Solana stream in Spain, detected the predominance of diatom community in a spring season. The most characteristic structures originating during late spring– summer are larval housings. Irion & Mu¨ller (1968) and Janssen et al. (1999) pointed at seasonal significance of larval housings in tufas in Germany and Belgium, respectively, and regarded them as a marker of a spring season. However, voids of similar dimensions, sharing most probably the same origin, are detected in the tufa sample that grew between late June –October 2003 at the Lu´cˇky top site. They are distributed in the whole, up to 14 mm thick, section of this
sample, which implies that the housings formed throughout the entire period of sample growth. Alike voids, similar in size but less regular in shape, were formed within the tufa deposited at Ha´j between November 2002–March 2003. Their irregular shapes may have been caused by postdepositional modification (see Golubic´ 1969). Therefore, though the housings are most common at the turn of spring and summer, they cannot be considered an unequivocal index of this season (see also Du¨rrenfeldt 1978). Laminae built of detritic non-carbonate components, visible within sparry crystals, are an accurate index of hydrological conditions (cf. Kano et al. 2004). They record increased surface runoff to tufa-depositing streams, which caused pronounced supply of fine-grained clastics and organic matter from soils. The concentration of such laminae in the tufa formed in Lu´cˇky clearly marks late autumn rainy season and spring thaw periods (Fig. 22a, b).
Differences between rate of tufa growth in the studied sites – influence of climatic conditions and water origin The annual tufa growth rate for each studied site was assessed by calculating the mean growth rate for the sites where observation was conducted in more than one point (Karwo´w, Lu´cˇky and Ha´j). This minimalized local effects caused by microenvironmental conditions, such as the ‘waterfall effect’. Figure 27 shows the relation between the mean annual growth rates, chemical parameters of feeding water and mean annual temperatures for all sites. The mean growth rates differ markedly between the sites. Exceptional is the mean growth rate at Lu´cˇky: 4.994 mg/cm2/day. It is not only higher than values for the other sites studied herein, but also higher than any values for tufa growth, known in literature. For example, the highest value for tufa obtained by Bono et al. (2001) was 1.8635 mg/cm2/day, substantially lower than the values from the Lu´cˇky site. The rate of tufa growth obtained by Zhang et al. (2001, their table 3) in their experiment with the Ya He river water is also lower. On the contrary, the growth rate in Lu´cˇky is lower than those measured for the majority of travertines. Pentecost & Colletta (2007) gave a value ranging from 13.4– 30.9 mg/cm2/day for the travertine growing in La Zitelle spring in Tuscany. Other data from travertine sites are provided in centimetres per year. For instance, Kitano (1963) estimated the growth rate of Futamata travertine in Japan between 200–1000 mm. Values given by
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT
185
Folk (1994) from La Zitelle spring fall approximately within the same range. Slightly lower values, varying from 2.8–56.5 cm a21, were reported by Bargar (1978, 13) from Mammoth Hot Springs. Guo & Riding (1992) measured thickness of laminae in Terma San Giovanni and accepted daily character of lamination. Since a lamina is 0.5 mm thick one can calculate the thickness of annual increment to c. 18 cm. The measurements carried out during the present experiment are based predominantly on mass increment, which impedes direct comparisons. However, the maximum thickness of tufa formed within 14 months at the Lu´cˇky site is 24 mm. It is lower than the above quoted values for travertines. The tufa growth rate at the Lu´cˇky site is therefore intermediate between the values typical of tufas and travertines. This may be explained by the origin of water feeding the studied tufa at Lu´cˇky. The stream water is a mixture of shallow and deep circulating components. Consequently, the water feeding the tufa abounds in TDS, including bicarbonate ions and dissolved CO2, which results in relatively high SIcalc. values, reaching 1.27 (Table 3). Chemical composition of water is thus responsible for the unexpectedly high rate of tufa growth at Lu´cˇky. Slightly lower is the rate of growth of the Hunaglong travertine complex in China, which also is fed by a mixture of waters from deep and shallow circulation. The mean growth rate of travertine in a fastflow setting equals there 0.9252 mg/cm2/day according to recalculated data by Liu et al. (1995). Lu et al. (2000) gave slightly higher values for travertine growth rate in a fast-flow setting in Hunaglong, which ranged from 0.75 to 5.4 mg/cm2/day. Another travertine depositing system fed by a mixture of different waters has been described by Liu et al. (2006) from Yunnan province in China. The travertine growth varies there between 6.5 and 10.2 mm a21. Regardless of the origin of water, the tufa recently precipitated at the Lu´cˇky site is similar in texture to tufas fed exclusively by water from shallow karst aquifers. The same is true for the Pleistocene tufa at Lu´cˇky (Gradzin´ski et al. 2008).
Fig. 27. Relationship between mean annual growth rate of tufa and mean annual air temperature as well as selected physicochemical parameters of feeding water; mean annual growth rate for Karwo´w, Lu´cˇky and Ha´j sites is calculated as mean increment on yearly or longer than seasonally exposed limestone tablets (see Table 3); data on Huanglong and Tartare sites, from Liu et al. (1995) and Bono et al. (2001) respectively, are multiply by 4 to unify the methods of calculation with the present data
186
´ SKI M. GRADZIN
The mean tufa growth rate at the Karwo´w site is 1.0124 mg/cm2/day, less than at Lu´cˇky but more than in the two other sites. Mineralization of water (mean TDS of 522.2 mg/l) seems the most likely cause of the relatively high tufa growth rate at the Karwo´w site. Thus, the case of Karwo´w is to some extent similar to that of Lu´cˇky. The Karwo´w site is fed by water ascending along a fault; however the water temperature, TDS and probably residence time are all significantly lower than those of the deep-circulating component of water that feeds the Lu´cˇky site. It is worth mentioning that other tufa-depositing streams are present in the vicinity. Moreover, Gruszczyn´ski et al. (2004) described an inactive travertine dome 7 m high and 60 m across, which started to grow in early Holocene and lies 30 km west of Karwo´w also on a fault line. Mean annual rates of tufa growth differ radically also between the Za´zriva´ and Ha´j sites: 0.1592 mg/ cm2/day and 0.4027 mg/cm2/day, respectively. Although the data concern only two sites, some tentative explanations can be offered. The mean annual temperatures and the SIcalc. values of parent fluids clearly influenced tufa growth rate at the studied sites (Fig. 27). Such influence was suggested in literature, as based on the interrelation between temperature, amount of soil CO2, the rate of carbonate bedrock dissolution and SIcalc. of parent water (e.g. Hennig et al. 1983; Pedley et al. 1996; Pentecost 1995, 2005, p. 281). Direct comparison of the data presented here with those of Bono et al. (2001), who studied Tartare karst spring in Italy, leads to a similar conclusion. In the Tartare spring, the mean growth rate measured by Bono et al. (2001) for dam settings, so similar to those presently studied, was 1.86 mg/cm2/day. The mean annual temperature in the vicinity of Tartare karst spring is over 12 8C. Hence, both the tufa growth and temperature are higher there than the studied herein. A mean rate of tufa growth in a tropical karst region in northern Australia reaches 4.15 mm a21 (Drysdale & Gillieson 1997). In that region the mean annual temperature is around 25 8C, which fits well to the above tendency. Similar relationship was noted in Japan (Kano et al. 2007). However, a further study is needed in this field, because no other tufa sites were studied systematically in this respect that could be used for comparison.
Conclusions The field experiment carried out in Slovakia and Poland in 2002 and 2003 have led to the following conclusions: 1. Micro-organisms and high values of SIcalc. of parent water favour growth of tufa.
2. 3.
4.
5.
6.
Deposition of tufa is more efficient in setting with fast flow of water in comparison with neighbouring setting with sluggish flow. Growing tufa contains many different biogenic components and displays a wide spectrum of textures. Fast-growing tufa exhibits crystalline texture or is composed of highly encrusted algae building fibrous texture. The crystalline texture results from inhibition of microbial colonization by quickly growing crystals, while the fibrous texture is the result of accelerated growth of micro-organisms forced by quick crystallization of calcite on their cells. Both textures developed under SIcalc. . 0.8 and both are characteristic of fast-flow setting. Slowly growing tufa shows mainly micritic texture. Algae were calcified on the surface of their cells, while cyanobacteria on their sheaths. Thus, the process is one of external surficial precipitation. It took place under the condition of SIcalc. . 0.45. Unicellular desmid Oocardium stratum calcified under SIcalc. , 0.78. Diatom frustules were commonly overgrown by calcite but they acted as nucleation surface only when SIcalc. value was higher than 1. No universal common seasonal trend was detected in tufa growth rate. This demonstrates the importance of local factors, specific for particular sites. A clear seasonal sequence of textures was not detected, either, though some textures, such as clotted micrite with numerous diatoms, are common in winter, and encrusted algal filaments are abundant in spring and summer. The high growth rate of tufa in the summer of 2003 was the result of exceptionally warm and dry conditions prevailing at that time. Chemistry of parent water is the main agent governing the tufa growth. This is confirmed by the higher net growth rate at the Lu´cˇky and Karwo´w sites (4.994 mg/cm2/day and 1.0124 mg/cm2/day respectively), where water had the highest content of TDS, which reflects the deep-circulation origin of this water. The net tufa growth rates at the mountain Za´zriva´ site and at the upland Ha´j site are 0.1592 mg/cm2/day and 0.4027 mg/cm2/ day, respectively, which suggests that this process may be influenced by local climatic conditions.
The study is financed by the State Committee of Scientific Research grant 6PO4D 001 21. While conducting the experiment the author was supported by the Foundation for Polish Science in the frame of Grant for Researchers to Professor Jo´zef Kaz´mierczak. Professor Wolfgang Dreybrodt let the author know how he calculated the tufa growth rate in Hunaglong and Tartare sites. Roman
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT Aubrecht informed about new locality of tufa in Za´zriva´, Mariusz Czop and Jacek Motyka conducted water analyses while Anna Lewandowska ran XRD analyses. Professor Teresa Mrozin´ska determined algae, Professor Ryszard Ochyra bryophytes, and Dr Andrzej Kownacki invertebrates. Annette Go¨tz (Darmstadt) and Jolanta Bednarczyk (Dallas) provided the author with literature that is inaccessible in Poland. Michał Banas´, Ryszard Gradzin´ski, Renata Jach, Paweł Ramatowski, Maciej Tomaszek, and Tomasz Tomaszek enthusiastically helped in the fieldwork. Stanisław Olbrych, Andrzej Szumny and Jarosław Szumny skillfully prepared thin sections. Zuza Banach associated with SEM work. Renata Jach prepared the figures. Robert Szczepanek helped in calculation of the stream discharge. Administration of Slovak Caves, Slovak Museum of Environment Protection and Speleology, speleological club Cassovia and Miroslav Terray provided logistic support during work in Slovakia. The author’s participation in the Hull tufa meeting was supported by the Ferens Fund. The author is deeply indebted to all of the above mentioned individuals and institutions as well as to Marek Dulin´ski for stimulating ¨ zkul and an discussion. The author also thanks Mehmet O anonymous reviewer for critical suggestion on the earlier version of the manuscript, as well as Grzegorz Haczewski for assistance with the English.
References A LLAN , J. D. 1995. Stream Ecology: Structure and Function of Running Waters. Chapman & Hall, London. A NAGDOSTIDIS , K. & R OUSSOMOUSTAKAI , M. 1988. Cyanophytes from metal burdened substrate. Archiv fu¨r Hydrobiologie, Supplement Band, 80, Algological Studies, 50– 53, 561– 654. A NDREWS , J. E. 2006. Palaeoclimatic record from stable isotopes in riverine tufas: synthesis and review. EarthScience Reviews, 75, 85– 104. A NDREWS , J. E. & B RASIER , A. T. 2005. Seasonal records of climatic change in annually laminated tufas: short review and future prospects. Journal of Quaternary Science, 20, 411–421. A NZALONE , E., F ERRERI , V., S PROVIERI , M. & D’A RGENIO , B. 2007. Travertines as hydrologic archives: the case of the Pontecagnano deposits (southern Italy). Advances in Water Resources, 30, 2159–2175. A RP , G., W EDEMEYER , N. & R EITNER , J. 2001. Fluvial tufa formation in a hard-water creek (Deinschwanger Bach, Franconian Alb, Germany). Facies, 44, 1– 22. B AKER , A., S MITH , C. L., J EX , C., F AIRCHILD , I. J., G ENTY , D. & F ULLER , L. 2008. Annually laminated speleothems: a review. International Journal of Speleology, 37, 193– 206. B ARGAR , K. E. 1978. Geology and thermal history of Mammoth Hot Springs, Yellowstone National Park, Wyoming. Geological Survey Bulletin, 1444, 1 –55. B ATHURST , R. G. C. 1976. Carbonate Sediments and their Diagenesis. Developments in Sedimentology, 14. Elsevier, Amsterdam. B ONO , P., D REYBRODT , W., E RCOLE , S., P ERCAPO , C. & V OSBECK , K. 2001. Inorganic calcite precipitation in
187
Tartare karstic spring (Lazio, central Italy), field measurements and theoretical prediction on depositional rates. Environmental Geology, 41, 305–313. C ANFIELD , D. E., JR & H OYER , M. V. 1988. Influence of nutrient enrichment and light availability on the abundance of aquatic macrophytes in Florida streams. Canadian Journal of Fisheries and Aquatic Sciences, 45, 1467–1472. Cˇ ERMA´ K , Cˇ. 1994. Klimaticke´ pomery. In: R OZLOZˇ NIK , M. & K ARASOVA´ , A. (eds) Slovensky´ Kras. Osveta, Martin, 36– 47. (In Slovak.) C HAFETZ , H. S. & F OLK , R. L. 1984. Travertines: depositional morphology and the bacterially constructed constituents. Journal of Sedimentary Petrology, 54, 289– 316. C HAFETZ , H. S., R USH , P. F. & U TECH , N. M. 1991a. Microenvironmental controls on mineralogy and habit of CaCO3 precipitates: an example from an active travertine system. Sedimentology, 38, 107–126. C HAFETZ , H. S., U TECH , N. M. & F ITZMAURICE , S. P. 1991b. Differences in the d13C signatures of seasonal laminae comprising travertine stromatolites. Journal of Sedimentary Petrology, 61, 1015–1028. C HAFETZ , H. S. & L AWRENCE , J. R. 1994. Stable isotopic varaiability within modern travertines. Ge´ographie Physique et Quaternaire, 48, 257–273. C HARLTON , S. R. & P ARKHURST , D. L. 2002. PhreeqcI – A graphical user interface to the geochemical model PHREEQC. U.S. Geological Survey Fact Sheet, FS – 031– 02, 1– 2. C HEN , J., Z HANG , D. D., W ANG , S., X IAO , T. & H UANG , R. 2004. Factors controlling tufa deposition in natural waters at waterfall sites. Sedimentary Geology, 166, 353– 366. C HIBOWSKI , E., H OŁYSZ , L. & S ZCZES´ , A. 2003. Adhesion of in situ precipitated calcium carbonate in the presence and absence of magnetic field in quiescent conditions on different solid surfaces. Water Research, 37, 4685–4692. D ANDURAND , J. L., G OUT , R., H OEFS , J., M ENSCHEL , G., S CHOTT , J. & U SDOWSKI , E. 1982. Kinetically controlled variations of major components and carbon and oxygen isotopes in a calcite-precipitating spring. Chemical Geology, 36, 299– 315. D E´ FARGE , C., T RICHET , J., J AUNET , A.-M., R OBERT , M., T RIBBLE , J. & S ANSONE , F. 1996. Texture of microbial sediments revealed by cryo-scanning electron microscopy. Journal of Sedimentary Research, 66, 935 –947. D OWGIAŁŁO , W. D. 1974. Objas´nienia do szczego´łowej mapy geologicznej Polski, arkusz Opato´w (854) 1:50 000. Wydawnictwa Geologiczne, Warszawa. (In Polish.) D REYBRODT , W., B UHMANN , D., M ICHAELIS , J. & U SDOWSKI , E. 1992. Geochemically controlled calcite precipitation by CO2 outgassing: field measurements of precipitation rates in comparison to theoretical predictions. Chemical Geology, 97, 285– 294. D RYSDALE , R. N. 1999. The sedimentological significance of hydropsychid caddis-fly larvae (order: Trichoptera) in travertine-depositing stream: Louie Creek, northwest Queensland, Australia. Journal of Sedimentary Research, 69, 145–150.
188
´ SKI M. GRADZIN
D RYSDALE , R. & G ILLIESON , D. 1997. Micro-erosion meter measurements of travertine deposition rates: a case study from Louie Creek, Northwest Queensland, Australia. Earth Surface Processes and Landforms, 22, 1037– 1051 D U¨ RRENFELDT , A. 1978. Untersuchungen zur Besiedlungsbiologie von Kalktuff – faunistische, o¨kologische und elektronenmikroskopische Befunde. Archiv fu¨r Hydrobiologie, Supplementband, 54, 1– 79. E MEIS , K.-C., R ICHNOW , H.-H. & K EMPE , S. 1987. Travertine formation in Plitvice National Park, Yugoslavia: chemical versus biological control. Sedimentology, 34, 595–609. E NSMINGER , I., F OESTER , J., H AGEN , C. & B RAUNE , W. 2005. Plasticity and acclimation to light reflected in temporal and spatial changes of small-scale macroalgal distribution in a stream. Journal of Experimental Botany, 56, 2047–2058. F ISHER , S. G., G RAY , L. J., G RIMM , N. B. & B USH , D. E. 1982. Temporal succession in a desert stream ecosystem following flash flooding. Ecological Monographs, 52, 91– 110. F LAJS , G. 1977. Die Ultrastrukturen des Kalkalgenskeletts. Palaeontographica, 160B, 69– 128. F OLK , R. L. 1994. Interaction between bacteria, nannobacteria, and mineral precipitation in hot springs of central Italy. Ge´ographie Physique et Quaternaire, 48, 233– 246. F ORD , T. D. & P EDLEY , H. M. 1996. A review of tufa and travertine deposits of the world. Earth-Science Reviews, 41, 117– 175. F RANKO , O. & H ANZEL , V. 1980. Hydrogeologicke´ pomery Liptovskej kotliny. In: G ROSS , P. & K O¨ HLER , E. (eds) Geolo´gia Liptovskej Kotliny. Geolo´ stav Diony´za Sˇtu´ra, Bratislava, 126– 137. (In gicky´ U Slovak.) F REYTET , P. & P LET , A. 1991. Les formations stromatolitiques (tufs calcaires) re´centes de la region de Tournus (Saoˆne et Loire). Geobios, 24, 123– 139. F REYTET , P. & P LET , A. 1996. Modern freshwater microbial carbonates: the Phormidium stromatolites (tufa-travertine) of southwestern Burgundy (Paris Basin, France). Facies, 34, 219–238. F REYTET , P. & V ERRECCHIA , E. P. 1998. Freshwater organisms that build stromatolites: a synopsis of biocrystallization by prokaryotic and eukaryotic algae. Sedimentology, 45, 535– 563. G A˛GOL , J. & U RBAN , J. 2000. Jerzy Bogumił Pusch prekursorem ochrony geologicznych stanowisk dokumentacyjnych. Przegla˛d Geologiczny, 48, 55– 56. (In Polish.) G IVEN , R. K. & W ILKINSON , B. H. 1985. Kinetic control of morphology, composition and mineralogy of abiotic sedimentary carbonates. Journal of Sedimentary Petrology, 55, 109– 119. G OLUBIC´ , S. 1967. Algenvegetation der Felsen. Die Binnengewa¨sser, 23, 1 –183. G OLUBIC´ , S. 1969. Cyclic and noncyclic mechanisms in the formation of travertine. Internationale Vereinigung fu¨r Theoretische und Angeweine Limnologie, 17, 956– 961. G OLUBIC´ , S. & M ARCˇ ENKO , E. 1958. Zur Morphologie und Taxonomie der Desmiceengattung Oocardium.
Schweizerische Zeitschrift fu¨r Hydrologie, 20, 177–185. G OLUBIC , S., V IOLANTE , C., F ERRERI , V. & D’A RGENIO , B. 1993. Algal control and early diagenesis in Quaternary travertine formation (Rocchetta a Volturno, Central Apennines). In: B ARATTOLO , F. (ed.) Studies on Fossil Benthic Algae. Bolletino della Societa Paleontologica Italina, Spec. Vol. 1, 231– 247. G OLUBIC´ , S., V IOLANTE , C., P LENKOVIC´ -M ORAJ , A. & G RGASOVIC´ , T. 2008. Travertines and calcareous tufa deposits: an insight into diagenesis. Geologia Croatica, 61, 363– 378. G ONZA´ LEZ , L., C ARPENTER , S. J. & L OHMANN , K. C. 1992. Inorganic calcite morphology: roles of fluid chemistry and fluid flow. Journal of Sedimentary Petrology, 62, 382–399. G ORBUSHINA , A. A. & K RUMBAIN , W. E. 2000. Subaerial microbial mats and their effects on soil and rocks. In: R IDING , R. E. & A WRAMIK , S. M. (eds) Microbial Sediments. Springer, Berlin, 161 –170. G OULDER , R. 1988. Eplithic bacteria in an acid and calcareous headstream. Freshwater Biology, 19, 405–416. G RADZIN´ SKI , M., R OSPONDEK , M. & S ZULC , J. 1997. Paleoenvironmental controls and microfacies variability of the flowstone cover from Zvoniva´ Cave in the Slovakian Karst. Slovak Geological Magazine, 3, 299–313. G RADZIN´ SKI , M., D ULIN´ SKI , M. ET AL . 2008. Facies and age of travertines from Spisˇ and Liptiov regions (Slovakia) – preliminary results. Slovensky´ Kras, 46, 31–40. G ROSS , P. 1980. Tektonika. In: G ROSS , P. & K O¨ HLER , E. ´ stav (eds) Geolo´gia Liptovskej Kotliny. Geologicky´ U Diony´za Sˇtu´ra, Bratislava, 116–121. (In Slovak.) G RUSZCZYN´ SKI , M., K OWALSKI , B. J., S OŁTYSIK , R. & H ERCMAN , H. 2004. Tectonic origin of the unique Holocene travertine from the Holy Cross Mts.: microbially and abiologically mediated calcium carbonate, and manganese oxide precipitation. Acta Geologica Polonica, 54, 61– 76. G UO , L. & R IDING , R. 1992. Aragonite laminae in hot water travertine, Rapolano Terme, Italy. Sedimentology, 39, 1067–1079. H AMMER , Ø., D YSTHE , D. K., L ELU , B., L UND , H., M EAKIN , P. & J AMTVEIT , B. 2008. Calcite precipitation instability under laminar, open channel flow. Geochimica et Casmochimica Acta, 72, 5009– 5021. H ASˇ KO , J. & P OLA´ K , M. 1978. Geological map of the Kysucke´ Vrchy and Krı´vanska Mala´ Fatra Mts. Geolo´ stav Diony´za Sˇtu´ra, Bratislava. gicky´ U H EIMANN , A. & S ASS , E 1989. Travertine in the northern Hulla Valley, Israel. Sedimentology, 36, 95–108. H ENNIG , G. J., G RU¨ N , R. & B RUNNACKER , K. 1983. Speleothems, travertines and paleoclimates. Quaternary Research, 20, 1 –29. H ERMAN , J. & L ORAH , M. M. 1987. CO2 outgassing and calcite precipitation in Falling Spring Creek, Virginia, U.S.A. Chemical Geology, 62, 251–262. H ORNER , R. R., W ELCH , E. B., S EELEY , M. R. & J ACOBY , J. M. 1990. Responses of periphyton to changes in current velocity, suspended sediment and
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT phosphorus concencentration. Freshwater Biology, 24, 215–232. ˇ SSR, II, Minera´lni vody. H YNIE , O. 1963. Hydrogeologie C ˇ eskoslovenske´ Akademie Ve´d, Praha. Nakladatelstvı´ C (In Czech.) I RION , G. & M U¨ LLER , G. 1968. Mineralogy, petrology and chemical composition of some calcareous tufa from the Schwa¨bische Alb, Germany. In: M U¨ LLER , G. & F RIEDMAN , G. M. (eds) Carbonate Sedimentology in Central Europe. Springer, New York, 157– 171. J ACOBSON , R. L. & U SDOWSKI , E. 1975. Geochemical controls on a calcite precipitating spring. Contribution to Mineralogy and Petrology, 51, 65– 74. J ANSSEN , A., S WENNEN , R., P ODOOR , N. & K EPPEN , E. 1999. Biological and diagenetic influence in Recent and fossil tufa deposits from Belgium. Sedimentary Geology, 126, 75– 95. J ONES , B. & K AHLE , C. F. 1995. Origin of endogenetic micrite in karst terrains: a case study from the Cayman Islands. Journal of Sedimentary Research, A65, 281– 293. J ONES , B. & R ENAUT , R. W. 1994. Crystal fabrics and microbiota in large pisoliths from Laguna Pastos Grandes, Bolivia. Sedimentology, 41, 1171– 1202. K ANO , A. & F UJII , H. 2000. Origin of the gross morphology and internal texture of tufas of Shirokawa Town, Ehime Prefecture, southwest Japan. Journal of Geological Society of Japan, 106, 397–441. K ANO , A., H AGIWARA , R., K AWAI , T., H ORI , M. & M ATSUOKA , J. 2007. Climatic conditions and hydrological change recorded in a high-resolution stable-isotope profile of a Recent laminated tufa on a subtropical island, southwest Japan. Journal of Sedimentary Research, 77, 59–67. K ANO , A., K AWAI , T., M ATSUOKA , J. & I HARA , T. 2004. High-resolution records of rainfall events from clay bands in tufa. Geology, 32, 793–796. K ANO , A., M ATSUOKA , J., K OJO , T. & F UJII , H. 2003. Origin of annual lamination in tufa deposits, southwest Japan. Palaeogeography, Palaeoclimatology, Palaeoecology, 191, 243–262. K EMPE , S. & K AZ´ MIERCZAK , J. 1990. Chemistry and stromatolites of the sea-linked Satonda Crater Lake, Indonesia: A recent model for the Precambrian sea? Chemical Geology, 81, 299 –310. K EMPE , S. & K AZ´ MIERCZAK , J. 1993. Satonda Crater Lake, Indonesia: hydrogeochemistry and bicarbonates. Facies, 28, 1 –32. K ERNEY , M. P., P REECE , R. C. & T URNER , C. 1980. Molluscan and plant biostratigraphy of some Late Devensian and Flandrian deposits in Kent. Philosophical Transactions of the Royal Society of London, 291B, 1– 43. K ITANO , Y. 1963. Geochemistry of calcareous deposits found in hot springs. Journal of Earth Sciences Nagoya University, 11, 68–100. K ORTE , V. L. & B LINN , B. W. 1983. Diatom colonization on artificial substrata in pool and riffle zones studied by light and scanning microscopy. Journal of Phycology, 19, 332–341. K OVANDA , J. 1971. Quaternary limestones of Czechoslovakia. Sbornik Geologicky´ch Ve´d, Antropozoikum, 7A, 7–256. (In Czech, English summary.)
189
K RALJ , K., P LENKOVIC´ -M ORAJ , A., G LIGORA , M., P RIMC -H ABDIJA , B. & Sˇ IPOSˇ , L. 2006. Structure of periphytic community on artificial substrata: influence of depth, slide orientation and colonisation time in karstic Lake Visovacˇko, Croatia. Hydrobiologia, 560, 249– 258. L IMONDIN -L OZOUET , N. & P REECE , R. C. 2004. Molluscan successions from the Holocene tufa of St Germain-le-Vasson, Normandy (France) and their biogeographical significance. Journal of Quaternary Science, 19, 55– 71. L IU , Z., L I , H., Y OU , C., W AN , N. & S UN , H. 2006. Thickness and stable isotopic characteristics of modern seasonal climate-controlled sub-annual travertine laminas in a travertine-depositing stream at Baishuitai, SW China: implications for paleoclimate reconstruction. Environmental Geology, 51, 257– 265. L IU , Z., S VENSSON , U., D REYBRODT , W., D AOXIAN , Y. & B UHMANN , D. 1995. Hydrodynamic control of inorganic calcite precipitation in Huanglong Ravine, China: Field measurements and theoretical prediction of deposition rates. Geochimica et Cosmochimica Acta, 59, 3087–3097. L OJEN , S., D OLENEC , T., V OKAL , B., C UKROVS , N., M IHELCˇ IC´ , C. & P APESCH , W. 2004. C and O stable isotope variability in recent freshwater carbonates (River Krka, Croatia). Sedimentology, 51, 361– 375. L ORCH , H.-J. & O TTOW , C. G. 1985. Use of bacteria attached to submerged macrophytes and glass slides as indicators of an increasing water pollution. Internationale Vereinigung fu¨r Theoretische und Angeweine Limnologie, 22, 2297– 2302. L OVE , K. M. & C HAFETZ , H. S. 1988. Diagenesis of laminated travertine crusts, Arbuckle Mountains, Oklahoma. Journal of Sedimentary Petrology, 58, 441– 445. L OZˇ EK , V. 1957. Stratigraphie und Weichtiere der holoza¨nen Travertine in Ha´j bei Turnˇa. Anthropozoikum, 7, 27–36. (In Czech, German summary). L U , G., Z HENG , C., D ONAHOE , R. J. & L YONS , W. B. 2000. Controlling processes in a CaCO3 precipitating stream in Huanlong Natural Scenic District, Sichuan, China. Journal of Hydrology, 230, 34– 54 M A¨ GDEFRAU , K. 1956. Pala¨obiologie der Pflanzen. Gustav Fischer Verlag, Jena. M ALUSA , J., O VERBY , S. T. & P ARNELL , R. A. 2003. Potential for travertine formation; Fossil creek, Arizona. Applied Geochemistry, 18, 1081–1093. M C I NTIRE , C. D. 1968. Structural characteristic of benthic algal communities in laboratory streams. Ecology, 49, 520– 537. M ELLO , J. (ed.) 1996. Geological map of the Slovensky´ Kras Mts. Ministerstvo Zˇivotne´ho Prostredia Slovenskej Republiky. Geologicka´ Sluzˇba Slovenskej Republiky, Bratislava. M ERZ , M. U. E. 1992. The biology of carbonate precipitation by cyanobacteria. Facies, 26, 81–102. M ERZ -P REI ß, M. & R IDING , R. 1999. Cyanobacterial tufa calcification in two freshwater streams: ambient environment, chemical thresholds and biological processes. Sedimentary Geology, 126, 103–124.
190
´ SKI M. GRADZIN
M EYRICK , R. A. & P REECE , R. C. 2001. Molluscan successions from two Holocene tufas near Northampton, English Midlands. Journal of Biogeography, 28, 77–93. M ICHAELIS , J., U SDOWSKI , E. & M ENSCHEL , G. 1985. Partitioning of 13C and 12C on the degassing of CO2 and the precipitation of calcite: Rayleigh-type fractionation and a kinetic model. American Journal of Science, 285, 318– 327. M ROZIN´ SKA , T. 1992. Algae of the Pieniny National Park (S. Poland). Vero¨ffentlichungen des Geobotanischen Institut der ETH, 107, 218 –237. N Eˇ MEJC , F. 1928. Palaeobotanical investigation in the travertine-complex around the village of Lu˚cˇky near Ru˚zˇomberok in Slovakia. Bulletin International de l’Academie des Sciences de Boheˆme, 37, 35, 1–19. O HLE , W. 1937. Kalksystematik unserer Binnengewa¨sser und der Kalkgehalt Ru¨gener Ba¨che. Geologie der Meere und Binnengewa¨sser, 1, 291–316. P ARKHURST , D. L. & A PPELO , C. A. J. 1999. User’s guide to PHREEQC (Version 2) – a computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations. U.S. Geological Survey Water-Resources Investigations Report, 99– 4259, 1 –312. P ARSIEGLA , K. I. & K ATZ , J. L. 1999. Calcite growth inhibition by copper(II). I. Effect of supersaturation. Journal of Crystal Growth, 200, 213– 226. P ARSIEGLA , K. I. & K ATZ , J. L. 2000. Calcite growth inhibition by copper(II). II. Effect of solution composition. Journal of Crystal Growth, 213, 368 –380. P AZDUR , A., P AZDUR , F. & S ZULC , J. 1988. Radiocarbon dating of Holocene calcareous tufa in southern Poland. Radiocarbon, 2, 133–152. P EDLEY , H. M. 1990. Classification and environmental models of cool freshwater tufas. Sedimentary Geology, 68, 143– 154. P EDLEY , H. M. 1994. Procaryote-microphyte biofilms and tufas: a sedimentogical perspective. Kaupia. Darmsta¨dter Beitrage zur Naturgeschichte, 4, 45– 60. P EDLEY , M. 1992. Freshwater (phytoherm) reefs: the role of biofilms and their bearing on marine reef cementation. Sedimentary Geology, 79, 255–274. P EDLEY , M. 2000. Ambient temperature freshwater microbial tufas. In: R IDING , R. E. & A WRAMIK , S. M. (eds) Microbial Sediments. Springer, Berlin, 179–186. P EDLEY , M. 2009. Tufas and travertines of the Mediterranean region: a testing ground for freshwater carbonate concepts and developments. Sedimentology, 56, 221– 246. P EDLEY , M., A NDREWS , J., O RDONEZ , S., G ARCIA DEL C URA , M. A., G ONZALEZ -M ARTIN , J. A. & T AYLOR , D. 1996. Does climate control the morphological fabric of freshwater carbonates? A comparative study of Holocene barrage tufas from Spain and Britain. Palaeogeography, Palaeoclimatology, Palaeoecology, 121, 239– 257. P EDLEY , M., R OGERSON , M. & M IDDLETON , R. 2009. Freshwater calcite precipitates from in vitro mesocosm flume experiments: a case for biomediation of tufas. Sedimentology, 56, 511– 527. P ENTECOST , A. 1978. Blue-green algae and freshwater carbonate deposits. Proceedings of the Royal Society of London, B, 200, 43– 61.
P ENTECOST , A. 1990. The algal flora of travertine: an overview. In: H ERMAN , J. S. & H UBBARD , D. A., JR (eds) Travertine-marl: stream deposits in Virginia. Department of Mines, Minerals, and Energy, Division of Mineral Resources, Charlottesville, Virginia, 117–127. P ENTECOST , A. 1991. A new and interesting site for the calcite-encrusted desmid Oocardium stratum Naeg in the British Isles. British Phycological Journal, 26, 297–301. P ENTECOST , A. 1995. The Quaternary travertine deposits of Europe and Asia Minor. Quaternary Science Review, 14, 1005–1028. P ENTECOST , A. 1998. The significance of calcite (travertine) formation by algae in a moss-dominated travertine from Matlock Bath, England. Archiv fu¨r Hydrobiologie, 143, 487– 509. P ENTECOST , A. 2003. Taxonomic identity, ecology and distribution of the calcite-depositing cyanobacterium Phormidium incrustatum (Oscillatoriaceae). Cryptogamie Algologie, 24, 307– 321. P ENTECOST , A. 2005. Travertine. Springer, Berlin. P ENTECOST , A. & C OLETTA , P. 2007. The role of photosynthesis and CO2 evasion in travertine formation: a quantitative investigation at an important travertine-depositing hot spring, Le Zitelle, Lazio, Italy. Journal of the Geological Society, London, 164, 843 –853. P ENTECOST , A. & S PIRO , B. 1990. Stable carbon and oxygen isotope composition of calcites associated with modern freshwater cyanobacteria and algae. Geomicrobiology Journal, 8, 17–26. P ENTECOST , A. & W HITTON , B. A. 2000. Limestones. In: W HITTON , B. A. & P OTTS , M. (eds) The Ecology of Cyanobacteria. Kluwer, Amsterdam, 257–279. P ENTECOST , A. & Z HANG , Z. 2000. The travertine flora of Juizhaigou and Munigou, China, and its relationship with calcium deposition. Cave and Karst Science, 27, 71–78. P ETERSON , C. G. & H OAGLAND , K. D. 1990. Effects of wind-induced turbulence and algal mat development on epilithic diatom succesion in a large reservoir. Archiv fu¨r Hydrobiologie, 118, 47–68. P IA , J. 1934. Die Kalkbildung durch Pflanzen. Beihefte zum Botanischen Zentralblatt, 52A, 1 –72. P ITOIS , F., J IGOREL , A. & B ERTRU , G. 2001. Colonization dynamics of an encrusting cyanobacterial mat in a hardwater river (Eaulne, France). Geomicrobiology Journal, 18, 139–155. P LENKOVIC´ -M ORAJ , A., HORVANTINCˇ IC´ , N. & P RIMC H ABDIJA , B. 2002. Periphyton and its role in tufa deposition in karstic waters (Plitvice Lakes, Croatia). Biologia, 57, 423– 431. P REECE , R. C. & D AY , S. P. 1994. Comparison of Postglacial molluscan and vegetational successions from a radiocarbon-dated tufa sequence in Oxfordshire. Journal of Biogeography, 21, 463 –478. R AVEN , J. A. 1992. How benthic macroalgae cope with flowing freshwater: resource acquisition and retention. Journal of Phycology, 28, 133– 146. R IDING , R. 1991. Classification of microbial carbonates. In: R IDING , R. (ed.) Calcareous Algae and Stromatolites. Springer, Berlin, 21–51.
GROWTH OF TUFA: RESULTS OF A FIELD EXPERIMENT R IDING , R. 2000. Microbial carbonates: the geological record of calcified bacterial–algal mats and biofilms. Sedimentology, 47, Suppl. 1, 179– 214. R USHFORD , S. R., S QUIRES , L. E. & C USHING , C. E. 1986. Algal communities of springs and streams in the Mt. St. Helens region, Washington, U.S.A. following the May 1980 eruption. Journal of Phycology, 22, 129–137. S ABATER , S., G UASCH , H., R OMANI , A. & M UN˜ OZ , I. 2000. Stromatolitic communities in Mediterranean streams: adaptation to a changing environment. Biodiversity and Conservation, 9, 379–392. S AMSONOWICZ , J. 1929. Sprawozdanie z badan´ geologicznych, wykonanych w roku 1928-ym w okolicach Klimontowa na arkuszu Sandomierz. Posiedzenia Naukowe Pan´stwowego Instytutu Geologicznego, 24, 13– 14. (In Polish.) S CHA¨ R , C. & J ENDRITZKY , G. 2004. Hot news from summer 2003. Nature, 432, 559–560. S CHNITZER , W. A. 1974. Kalkinkrustation und Kalksinterknollen in Lias-Quellwa¨ssern bei Elsenberg (Bl. Erlangen-Nord). Geologische Blatter fu¨r Nordost Bayern und Angrenzende Gebiete, 24, 188–191. S HIRAISHI , F., R EIMER , A., B ISSETT , A., DE B EER , D. & A RP , G. 2008. Microbial effects on biofilm calcification, ambient water chemistry and stable isotope records in a highly supersaturated setting (Westerho¨fer Bach, Germany). Palaeogeography, Palaeoclimatology, Palaeoecology, 262, 91–106. S TATHAM , I. 1977. A note on tufa depositing springs in Glenasmole, Co. Dublin. Irish Geographer, 10, 14– 18. S YMOENS , J.-J. 1957. Les eaux douces de l’Ardenne et re´gions Voisines: les milieux et leur ve´ge´tation algae. Bulletin de la Socie´te´ Royale de Botanique de Belgique, 89, 111– 314. S ZULC , J. 1983. Genesis and classification of calcareous sinter deposits. Przegla˛d Geologiczny, 31, 231–237. (In Polish, English summary.) S ZULC , J. 1984. Sedymentacja czwartorze˛dowych martwic wapiennych Polski południowej. PhD thesis. Polska Akademia Nauk, Instytut Nauk Geologicznych (In Polish). S ZULC , J. & S MYK , B. 1994. Bacterially controlled calcification of freshwater Schizotrix-stromatolites: an example from the Pieniny Mts., Southern Poland. In: B ERTRAND -S ARFATI , J. & M ONTY , C. (eds)
191
Phanerozoic Stromatolites II. Kluwer Academic Publishers, Dordrecht, 31–51. T URNER , E. C. & J ONES , B. 2005. Microscopis calcite dendrites in cold-water tufa: implications for nucleation of micrite and cement. Sedimentology, 52, 1043– 1066. ¨ ber die Bedeutung der W ALLNER , J. 1934a. U sog. Chironomidentuffe fu¨r die Messung der ja¨hrlichen Kalkproduktion durch Algen. Hedwigia, 74, 176 –181. W ALLNER , J. 1934b. Beitrag zur Kenntnis der Vaucheria-Tuffe. Zentralblatt fu¨r Bakteriologie, Parasitenkunde und Infektionskrankheiten, Abt. II, 90, 150 –154. W ALLNER , J. 1935a. Eine gesteinebildende Su¨sswasserAlge Deutchlands. Natur und Volk, 66, 85– 91. ¨ ber die Verbreitungso¨kologie der W ALLNER , J. 1935b. U Desmidiacee Oocardium. Planta, 23, 249–263. W EIJERMARS , R., M ULDER -B LANKEN , C. W. & W IEGERS , J. 1986. Growth rate observation from the moss-built Checa travertine terrace, central Spain. Geological Magazine, 123, 279– 286. W HITTON , B. A. 1975. Algae. In: W HITTON , B. A. (ed.) River Ecology. University of California Press, Berkeley, 81– 105. W ILDE , E. W. 1982. Responses of attached algal communites to termination of thermal pollution. Hydrobiologia, 94, 135– 138. W INSBOROUGH , B. M. 2000. Diatoms and benthic microbial carbonates. In: R IDING , R. E. & A WRAMIK , S. M. (eds) Microbial Sediments. Springer, Berlin, 76–83. W INSBOROUGH , B. M. & G OLUBIC´ , S. 1987. The role of diatoms in stromatolite growth: two examples from modern freshwater settings. Journal of Phycology, 23, 195 –201. W INSBOROUGH , B. M., S EELER , J.-S., G OLUBIC´ , S., F OLK , R. L. & M AGUIRE , B., JR . 1994. Recent freshwater lacustrine stromatolites, stromtolitic mats and oncoids from northeastern Mexico. In: B ERTRAND S ARFATI , J. & M ONTY , C. (eds) Phanerozoic Stromatolites II. Kluwer Academic Publishers, Dordrecht, 71–100. Z HANG , D. D., Z HANG , Y., Z HU , A. & C HENG , X. 2001. Physical mechanisms of river waterfall tufa (travertine) formation. Journal of Sedimentary Research, 71, 205 –216.
In vitro investigations of the impact of different temperature and flow velocity conditions on tufa microfabric H. MARTYN PEDLEY* & MIKE ROGERSON Department of Geography, University of Hull, Cottingham Road, Hull, HU6 7RX, UK *Corresponding author (e-mail:
[email protected]) Abstract: A series of experiments on freshwater carbonates (tufas) involving biofilm colonization in both fast-flow and slow-flow mesocosms was carried out in order to assess the changing nature of biofilm and associated precipitates under contrasting conditions. A thin biofilm developed over 14 weeks during the ‘summer’ experimental run contained a basal calcite layer overlain by small calcite crystals suspended within the Extracellular Polymeric Substances (EPS). The ‘autumn’ biofilm, however, showed the development of multi-laminated calcite precipitates within the EPS despite constant environmental conditions throughout the run. The experiments also showed that the largest volume of calcite precipitate developed in the fast-flow flumes regardless of temperature control. Development of an extensive calcite layer at the base of EPS in conditions of complete darkness within the sump was also observed. This study provides increased weight for the concepts: (1) that fresh- and saltwater stromatolites appear to be highly comparable multi-laminated systems with precipitation strongly influenced by both phototrophic and heterotrophic microbes; and (2) that microbial precipitation may be more common within aphotic (including cave, lake bottom and soil) systems than has previously been considered.
Field investigations show that freshwater carbonate (tufa) deposition is affected by climate. Pedley et al. (1996) showed that barrage tufas in the UK differ from those developed under semi-arid conditions in Spain in the relative rates of aggradation and progradation in their barrages. Cool-humid barrages well seen in Derbyshire (UK), but typical of present day north western Europe, show strong progradation but only weak aggradation, resulting in the development of shallow pools upstream of the barrages which may be maintained for millennia in some cases (e.g. Taylor et al. 1994). Ultimately, however, these 1–3 m deep pools became infilled with lime muds and marginal paludal creep-out facies. In contrast, the barrage tufa developments at Ruidera Natural Park (Spain), have developed under a semi-arid climate (Pedley et al. 1996). This appears to have encouraged almost vertical barrage aggradation but little or no progradation (Fig. 1). This mode of barrage accumulation involved progressive deepening of the upstream pools which today can be over 20 m deep and over one kilometer in length (e.g. Pedley et al. 1996). An extreme example of this mode of arid climate vertical barrage wall build-up is seen in the Band-e-Emir tufas (Afghanistan) (Brett 1970; Bourrouilh-Le et al. 2007). Similar climatic influences can be seen when comparing paludal tufa systems at Ddol in North Wales (UK) with those from semi-arid areas of Central Spain (Pedley et al. 2003). The climatic effect on accumulation rates provides a potential means of extracting
qualitative palaeoclimate information from tufa systems on the basis of their architecture alone. However, the precise climatic factors forcing the cool-humid and semi-arid models are currently too poorly constrained to unlock this important source of information about the past. The variables potentially affecting precipitation are many and include physico-chemical (including degassing rate of carbon dioxide and flow rate) and biochemical (temperature, nutrient availability, water flow rate, daylight hours) probably in complex, variable combination. Progressively, the approaches to solving these problems have shifted from field based outcrop modelling (Pedley 1990), via sampling for the geochemical (Francˇisˇkovic´-Bilinski et al. 2004; Andrews & Brasier 2005) and hydrochemical characterization of tufas (Zaihua et al. 1995; Lu et al. 2000; Kano et al. 2003), to an increasing awareness that the living biofilm plays a major role in freshwater carbonate precipitation (Pedley 1994; Freytet & Verrecchia 1998; Caudwell et al. 2001; Bisset et al. 2008; Shiraishi et al. 2008; Pedley et al. 2009). Many recent studies have involved placing artificial substrates in static water (Pedley 1994; Benzerara et al. 2006) or collecting living river bed material (Kano et al. 2003) in order to sample the biofilm products. Although most in situ biofilm studies have been carried out on marine stromatolites (Reid et al. 2000; Decho et al. 2005), efforts on freshwater systems are currently advancing at considerable pace. Shiraishi
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 193–210. DOI: 10.1144/SP336.9 0305-8719/10/$15.00 # The Geological Society of London 2010.
194
M. PEDLEY & M. ROGERSON
Fig. 1. Profiles of barrage tufa constructions from two dissimilar climatic regimes. Upper sketch represents a tufa barrage at Caerwys, North Wales, UK and associated pools. This is typical of deposits in cool-humid climates. Note that barrage progradation and aggradation rates are similar leading to relatively shallow upstream pool development prone to lateral infill by paludal facies. Stipple symbol indicates well-cemented areas at the barrage crest; black areas are sapropel deposits no symbol is lime mud. Lower sketch represents a tufa barrage in Ruidera Natural Park, Spain. This is typical of deposits in warm, semi-arid climates. Note the development of the barrage mainly by aggradation of stromatolites (laminated symbol). In consequence, a deep lake is developed immediately upstream of the barrage. Accumulation style of barrage development is shown in the top left inset of both sketches.
et al. (2008) have successfully studied the effects of photosynthesis on in-situ fluvial biofilms and Bissett et al. (2008) have studied precipitation products in ex-situ fluvial biofilms. Collectively, these works indicate tufa precipitation is not the straightforward process of physico-chemical CO2 degassing at waterfalls and riffles but is much complicated by the precipitating activities of microbial biofilms in both freshwater and marine static and flowing water situations. In order to circumvent many of the problems inherent to field work, a controllable system capable of supporting fluvial biofilms and tufa precipitation has been constructed in the laboratory during the present study. This laboratory-based experimental approach has the benefit over field site studies of sampling accessibility and independent control. Single environmental variables can be changed progressively through a series of experiments run under the same underlying conditions. Previously reported flowing mesocosm experiments (Rogerson et al. 2008; Pedley et al. 2009) have demonstrated the feasibility of such work and the close comparison between laboratory and natural calcite precipitates (cf. Turner & Jones 2005). The present study extends this experimental work into investigating the effects of variable day length, ambient temperature and flow rate on biofilm development and calcium carbonate precipitation. It also
considers the partitioning of carbonate products within the communal EPS and presents a new model for the freshwater tufa biofilm.
Mesocosm and experimental apparatus The mesocosm apparatus used in this set of experiments is closely comparable to that used in Pedley et al. (2009). However, in the current experiments two, 1 m long flumes were run simultaneously from a continuous flow of recycled water held in a closed sump. The system was driven by a 4000 l/h Blagdon Hydratech Torrent HTT 4000 submersible pump which lifted the water 0.5 m from the 40 litre sump to the flumes. The flow rate to the first flume (slow-flow) was set at c. 100 litres/hour while the second flume (fastflow) received c. 1000 litres/hour. Illumination was provided by a 250 watt Thorn–Lopack sodium lamp sited 1.5 m above the flumes. Temperature was controlled by an in-line Aquamedic ‘minicooler’ chiller unit directly linked to the pump outflow pipe. Water was diverted to each of the flumes immediately down-flow of the chiller unit. Natural spring water for the system obtained from a Cretaceous Chalk spring at Welton, East Yorkshire. In order to adequately compensate for a progressive fall in calcium ion supply during the
MESOCOSM MICROFABRICS
experimental runs the sump was connected to a pair of ‘Korallin C501’ calcium reactors. These supplied calcium ions derived by dissolution of aragonitic coral debris within the reactor towers. The supply rate of calcium ions was adjusted by controlling the flow of carbon dioxide gas delivered to each of the reactor towers. The flume waters were maintained at a conductivity level of 300 –400 microsiemans (ms), this conductivity principally being generated from calcium and bicarbonate ions in solution, and a pH was typically c. 8.4. Calibration of conductivity to calcium hardness (established via titration with EDTA) was undertaken during a previous experiment (Pedley et al. 2009) and is shown as Figure 2a. Consequently, it can be assumed the system ranged within calcium hardnesses of 250– 350 mg L21. The Corallin system typically yields magnesium hardnesses one order of magnitude lower than calcium hardness. Rapid biofilm colonization of the mesocosm was achieved by seeding the fast flow flume with part of a living Crataneuton moss colony and associated biofilms freshly taken from a tufa producing stream (River Lathkill, N. Derbyshire). The biofilms contained an abundant diverse range of Oscillitoriacea and other cyanobacteria, diatoms and filamentous green algae and heterotrophic bacteria (see accounts in Pedley 1994). The reader is referred to Figure 1 in Pedley et al. (2009) for other details of the mesocosm construction and operational procedures. Building on experimental work reported in Pedley et al. (2009); Rogerson et al. (2008) a sequence of medium term experiments were set up to better assess the role of the biofilm in calcium carbonate precipitation under controlled conditions. Each experiment commenced after 3–4 weeks, once extensive biofilm colonization of the flume surface had been established. During the experimental period the biofilm community was also encouraged to colonize 1 cm2 glass sampling slides which were inserted behind small transverse barriers within the flumes (see Pedley et al. 2009 for details). The glass slides (Fig. 2b) provided a convenient means of obtaining representative surface colonized samples of known age which could be air dried and examined by means of scanning electron microscopy (SEM) using a Zeiss SMG EVO 60 SEM with an Oxford Instruments INCA energy 350 energy dispersive x-ray (EDX) spectrometer probe. Additional crystalline calcite analyses were carried out on a Seimens D5000 X-ray Diffractometer (XRD) using CuK alpha radiation. Most specimens were either gold or carbon coated prior to study, or remained uncoated for backscatter examination. In addition, living biofilm was examined by stereoscopic optical microscopy. This enabled many of the 3D interrelationships
195
within the biofilm to be resolved. Acradine Orange which stains all DNA luminous green but selectively also stains the DNA of heterotrophs orange. However, focus at high magnifications made recording details difficult in comparison to SEM photomicrography, consequently, the latter is used for all figures in this account. Some biofilm was also examined using a wet stage attachment to the SEM which helped in interpreting EPS distributions and morphologies prior to desiccation shrinkage. Water in the system was changed at the end of each experiment and replaced with fresh spring water in order to avoid build up of magnesium and trace elements derived from the calcium reactors and any toxic metabolic byproducts. The established biofilms, however, were kept alive and used for re-seeding the flumes. Nutrients were supplied at 3 week intervals by the addition of 200 ml of dilute slurry extracted from submerged, decayed leaves taken from the Welton catchment.
Experimental procedures The first experiment simulated conditions comparable to ‘summer’ in central England (18 hours of ‘daylight’ and a water temperature of 18 degrees maximum during the daylight period). The parallel flume arrangement permitted high-flow and lowflow conditions to be run simultaneously. Once set up the experiment continued for 14 weeks. In the second experiment flow conditions were maintained identically to those established for the summer experiment, however, daylight was reduced to 6 hours and temperatures were reduced to 10 degrees. This equates to ‘autumn’ conditions in Central England. Once set up the experiment was left to run for 17 weeks.
Results Biofilm structure Once established, the summer biofilm was pale green in colour tending towards yellow –green. The colour generally darkened slightly after nutrients were added. Biofilm EPS coated all available surfaces and was dominated by diatoms and filamentous algae, though cyanobacteria and heterotrophic bacteria were also ubiquitous (Fig. 2b). The filamentous algae showed a tendency to develop into trailing filament blanket weed blooms beneath which an additional diatom, cyanobacterial and heterotrophic community developed. Similar developments occurred under both fast and slowflow conditions. At the start of the autumn experiment the sump water was refilled with fresh Welton spring water
196
M. PEDLEY & M. ROGERSON
Fig. 2. (a) Calibration of conductivity and calcium hardness in Corallin Calcium Reactor generated flumes water. (b) Typical biofilm and associated carbonate precipitates developed in the flume experiments after 14 weeks under summer fast-flow conditions. Note the moss framework aiding the development of a micro-barrage and the trailing
MESOCOSM MICROFABRICS
and re-seeded with biofilm coated moss. Once established the resulting autumn biofilm was much darker green in colour and contained few filamentous green algae. A significantly thicker amount of viscous EPS in the biofilm (up to 6 mm thick, Fig. 2c, d) developed in association with filamentous cyanobacteria colonization. Irregular, centimetrescale hemispherical undulations developed on the exposed EPS surface on both fast-flow and slowflow biofilms (Fig. 2c). The detailed interrelationships between the microbes and the EPS are difficult to demonstrate in the thin summer biofilm and this was exacerbated by shrinkage during dehydration, in preparation for SEM work. More detail was seen when wet samples were viewed using stereoscopic optical microscopy. In the summer biofilm the EPS consisted of a ,100 mm thick sheet attached to the substrate from which filamentous algae extended. In contrast, the autumn biofilm was up to 6 mm thick. Algae were relatively infrequent in the autumn biofilm whereas cyanobacteria were dominant, diatoms were common and coccoid and short-rod varieties of heterotrophic bacteria were widespread especially within and in close proximity to the associated calcite precipitates. This microbial association conforms most closely to the ‘Group 1 aquatic community of (Freytet & Verrecchia 1998). Importantly, the autumn EPS (especially when living samples were viewed normal to substrate under light microscopy) had an internal structure of closely aligned cyanobacterial filaments (mainly Phormidium but also with abundant Nostoc) all arranged normal to substrate which were anchored on to the flume base. Many filaments extended beyond the outer EPS surface into the overlying water. In addition, the outermost few hundred mm of the EPS also contained a surficial zone of coiled cyanobacterial filaments giving a ‘woolly’ appearance. When the autumn biofilm was viewed parallel to substrate the microbial community was also seen to be partitioned within the EPS into clumps of oscillitoriacean filaments associated with relatively few coccoid heterotrophs. These clumps were separated from each other by narrow, vertical zones arranged into polygons which were frequently dominated by larger filamentous cyanobacteria cf. Nostoc (best seen under light microscopy but well shown in Fig. 5b). Small (200 –500 nm) coccoid and short rod bacteria (see Fig. 5b) were also
197
abundant in the polygonal zones and were identified, using Acridine Orange stain, predominantly as heterotrophs. Significantly, during both summer and autumn experiments a sparse, aphotic biofilm community also developed on the submerged walls of the unlit water sump and on the pump. This community developed in total darkness and did not contain phototrophs. Rather, it appeared as a sparsely distributed heterotroph community associated with very little visible EPS but abundant calcite precipitates.
Calcite precipitates within the photic biofilm Considerable quantities of calcite were precipitated within the biofilm during both summer and autumn experiments. These precipitates, exclusively developed within the EPS, add further support to the conclusions of Ercole et al. (2007); Pedley et al. (2009); Shiraishi et al. (2008) that EPS plays a significant role calcite precipitation in association with microbial metabolic activities. Five XRD analyses showed that the experimental precipitates were all composed of pure calcite. There was no evidence either of aragonite or vaterite being present even when analyses were carried out within 5 hours of sampling the living biofilm. Unlike the poikilotopic lamellar calcite crystals associated with obligate calcifying Rivulariacea these precipitates were composed of 200–500 nm diameter, apparently solid, calcite nanospherulites that are morphologically similar to vaterite spherules in experiments by Braissant et al. (2003) and Nehrke & Van Cappellen (2005). These were arranged into packed groupings to form micropeloids and small imperfectly-faced (anhedral) microspar crystals (Fig. 3a). Typically, all precipitates grew with extensive perforations (Fig. 3b–d) which were occupied during their development by EPS strands and frequently (Fig. 3d) by small coccoid (possibly heterotrophic) bacteria as previously illustrated and reported in Pedley et al. (2008, see fig. 10b), cf. figure 2b of Freytet & Verrecchia (1998). Significantly, these micropeloid precursors and subsequent microspar were not randomly distributed within the biofilm EPS. Initially, they developed within the basal part of the biofilm forming a basal calcite layer (Fig. 4a) which was associated with heterotrophic bacteria, cyanobacteria and algae.
Fig. 2. (Continued) filamentous algae downstream of the barrage. Pale coloured areas are calcite. (c) Typical EPS-dominated biofilm developed under autumn slow-flow conditions after 17 weeks. These biofilms show a tendency to develop a mammilated upper surface as a result of partial detachment of the EPS from the substrate. (d) EPSdominated biofilm developed under autumn fast-flow flume conditions after 17 weeks. The biofilm is partly detached from the substrate and has become torn under the fast-flow conditions revealing the basal calcite layer (1). The paler biofilm colour compared to (b) is caused by the greater quantities of calcite precipitates within the fast-flow EPS. N.B. both flumes were drained before photographing.
198
M. PEDLEY & M. ROGERSON
Fig. 3. (a) Initial development of crystal faces on a nanospherulite aggregate, all intimately associated with EPS (smooth areas especially on the left). Note the apparently random stacking of nanospherulites. Nevertheless, incipient rhombic face development can also be seen. Autumn, fast-flow experiment, SEM view. (b) Initiation of calcite precipitation within the biofilm (glass slide substrate). Micropeloids, composed of nanospherulites and well formed calcite rhombs are visible, all linked by dehydrated strands of EPS. Summer, slow-flow experiment, air dried SEM sample after 28 days. (c) Details of the early calcite precipitates showing the ubiquitous development of irregular to elliptical cavities within the embryonic calcite crystals. During continued crystal growth these EPS areas are progressively occluded by further calcite precipitation, both internally and also on to the outer surface of the crystals. SEM view. (d) Cavities within a microspar crystal. The central surface depression was originally occupied by EPS (now dehydrated) and coccoid heterotrophic bacteria. The deeper cavity to the left was occupied by a strand of EPS. Autumn, slow-flow experiment. Air dried SEM sample.
Typically, this layer developed into an interlocking crystalline calcite fabric which frequently was seen to be physically attached to the substrate (Fig. 4b). Nevertheless, abundant EPS strands and living microbial attachments continued to extend through the crystals and around them imparting a microcavernous fabric to the basal calcite layer (Fig. 4). Other calcite precipitates also developed, apparently randomly, within the thin EPS sheet and above the basal calcite layer in the summer biofilm. In the thicker autumn biofilm, however, these precipitates were focused within the EPS into narrow polygonal zones (Fig. 5a) associated with high concentrations of heterotrophic bacteria (Fig. 5b). Individual micropeloids and anhedral microspar within these narrow polygonal zones were largest and most densely packed close to the basal calcite layer (see Fig. 5c) where the precipitates had intergrown to form vertical septae. At their base, these septae also appeared to intergrow
with other microspar crystals precipitated within the base of the EPS layer directly above the basal calcite layer (Figs 4a & 5a).
Calcite precipitates in the aphotic (sump) EPS A well formed rhombic calcite mosaic also developed ubiquitously in all submerged sump areas from the air water interface down to the sump bottom and also on the surface of the pump (Fig. 6a). This distribution of precipitates argues against precipitation solely via degassing of CO2, which tends to occur at the air–water interface (Rogerson et al. 2008). The tightly interlocking calcite crystals were pale brown in colour (cf. pale grey precipitates in the photic zone) and were commonly buried in EPS (Fig. 6b). Typically, crystals were pockmarked with cavities and 1–5 mm holes in the same characteristic manner as seen in biomediated precipitates developed within photic
MESOCOSM MICROFABRICS
199
Fig. 4. (a) Well developed, flat basal calcite layer (attached to the flume surface when sampled). Note the relatively precipitate-free EPS layer (now dehydrated into a cavity) above the basal calcite layer. Filamentous phototrophs and associated EPS extend from the base of the calcite layer up into the former EPS area [details in (b)]. Autumn, fast-flow experiment. Air dried SEM sample. (b) View of the underside of the basal calcite layer after detatchment from the flume surface. Note abundant interlinked sinuous EPS strands (e.g. top centre and right middle) and filamentous phototrophs (medium grey threads with ill-defined margins which lie parallel to the attachment surface but also extend within the crystal precipitates and away from the precipitate base). This EPS network was anchored to the flume surface throughout crystal growth. The dark grey EPS layer (into which the EPS strands attach, and seen at lower left hand side of image) directly overlies the basal calcite layer. Autumn, fast-flow experiment. SEM wet stage view, backscatter image.
200
M. PEDLEY & M. ROGERSON
Fig. 5. (a) Details of the polygonal structures (septae are here composed of dehydrated EPS) which develop within the EPS (the basal calcite layer uniformly covers the glass sample slide). These septae are the sites of further calcite precipitation and the development of polygonal walls of crystals which attach to the previously established basal calcite layer. Autumn, fast-flow experiment. Air dried SEM sample. (b) Details of the area at the base of a microspar septum wall. The filamentous bacteria are Oscillitoriacea. Note the abundant 200–500 nm diameter coccoid and short-rod heterotrophic bacteria (pale raised patches now apparently lying on the glass slide in the lower part of the view but originally suspended within the EPS and intimately associated with the precipitates). Autumn, slow-flow experiment. Air dried SEM sample. (c) Details of calcite microspar suspended within EPS in a septum. Note how the anhedral microspar becomes larger towards the base of the septal zone (bottom of the view). Also note the subparallel orientation of the crystal long-axes in the centre and top (these are parallel to the cyanobacterial filaments which are in approximate life position). Summer, slow-flow experiment. Air dried SEM sample.
EPS. The precipitates continued to grow without morphological modification under both summer and autumn conditions.
Flow Rates and precipitates There was no observable difference in morphology between biofilm microstructure developed in
flumes 1 (slow-flow) and 2 (fast-flow) during the summer experiments. The higher flow rates in flume 2 however, created trailing filamentous cyanobacterial EPS with unusual precipitate morphologies developed on to downstream-pointing biofilm ‘fingers’. SEM studies revealed the calcite crystals to have accelerated epitaxial growth in areas where water flow was fastest (Fig. 7a, b).
MESOCOSM MICROFABRICS
201
Fig. 6. (a) Microspar crystals developed within thick biofilm which coated the submerged wall of the water sump. This sealed sump received no light therefore the biofilm must have been generated exclusively by heterotrophic bacteria (note the total absence of filamentous cyanobacteria and algae). Summer, slow-flow experiment. SEM view. (b) Detail of a similar sump-grown calcite rhombs without EPS. Note the abundant cavities within these crystals which are occupied by EPS in (a). Summer, slow-flow experiment. Air dried SEM sample.
However, crystal growth along the c-axis (normal to the water surface) was minimal (Fig. 7b) but fast lateral growth resulted in a dense interlocking mosaic. Away from the fastest flow the precipitates were unremarkable in their morphologies. However, there was significantly more observable calcite precipitate associated with the flume 2 (fast-flow) biofilm compared to flume 1 (cf. Fig. 8a, b). During the autumn experiments similar biofilm volumes developed in flumes 1 and 2. In both flumes the biofilm initially established attachment to the flume substrate. However, the hemispherical biofilm topography best developed where the biofilm became locally detached from the substrate. In flume 2 (fast-flow) this partial detachment encouraged areas to tear and become carried away in the flow. Again, at the end of the experiments
much more calcite precipitation was observed associated with the flume 2 (fast-flow) biofilm.
Discussion The experimental runs show that temperature rather than flow rate controlled the composition of the microbial community. Nevertheless, biofilm colonization rates were about the same in both summer and autumn biofilm experiments. It may be that biofilm adhesion and consequent colonization rates were also influenced by other factors such as calcium ion availability (Fletcher & Floodgate 1976) and nutrient availability but these are yet to be investigated. The experiments also supported the findings of Bissett et al. (2008) that although increased biological activity, ion diffusion rates and daylight
Fig. 7. (a) Undulose-topped calcite crystals coating trailing ‘fingers’ of biofilm at the flume outflow. The topography of these crystalline surfaces has been limited by the thin film of fast flowing water which covered them during growth. The resulting crystals show rapid lateral growth but almost no growth parallel to c-axis. Summer, fast-flow experiment. Air dried SEM sample. (b) Details of the broad lateral growth increments which appear to have grown above the EPS sheet. Summer, fast-flow experiment. Air dried SEM sample.
202
M. PEDLEY & M. ROGERSON
Fig. 8. (a) Typical diatom dominated summer slow-flow biofilm (glass slide substrate) associated with micropeloids and well formed rhombic calcite. SEM view. (b) Summer fast-flow biofilm and precipitates (glass slide substrate) submerged in the same water and under the same sunlight conditions as (a). Note the relative abundance of calcite precipitate and the larger crystal size produced under fast-flow conditions [cf. (a)]. Air dried SEM sample.
calcification, might be expected at higher temperatures, these are balanced by the buffering capacity of the biofilm itself, causing precipitation rates to remain fairly constant within a wide temperature range. In contrast, however, flow rate proved to be consistently the dominant influence in determining the calcite precipitate yield. This confirms results from field sites studies by Zaihua et al. (1995) and Lu et al. (2000), also accommodating the data from Kano et al. (2003), which recorded fastest growth in the summer– autumn period when river flow rates were highest. These observations are also endorsed by recent theoretical insights (Hammer 2008; Hammer et al. 2008; Veysey & Goldenfeld 2008). Clearly, if ion flux, moderated by flow rate, is the overriding control of precipitation rate, then a number of long-held assumptions about the relative roles played by turbulence and degassing (Zhang et al. 2001; Chen et al. 2004) and microbial biomediation (Bissett et al. 2008; Shiraishi et al. 2008) need to be reassessed. However, our experiments indicate there may be a threshold in this flow rate regulation system, as in the fast-flow flume experiment biofilm tended to partly slough off and erode; we anticipate that had the flume velocities been much higher, biofilm colonization would have been severely restricted. True tufa precipitation (in terms of microfabrics) appears only to occur in association with biofilm EPS (Pedley et al. 2009) therefore, at the limiting shear stress for biofilm development a significant change in depositional fabric should occur with microbialites giving way to dense, massive developments of calcite. This was seen in Figure 7a, b where large spar crystals rapidly developed, apparently external to the EPS, under laminar flow conditions. This also
appears to explain a common characteristic of fossil tufa barrages, in that the spillover point of the barrage is generally characterized by horizontally laminated, massive calcite with less dominance by microbial fabrics (also seen on the stoss side of erect semi-aquatic vegetation, fig. 21 in Pedley 1994). This observation raises the critical question of whether aggradation by the microbial colony (largely by calcite biomediation) is more important than physico-chemical precipitation (largely controlled by ion flux) in tufa growth at spillover points. Answering this question is likely the key step in understanding why cool-humid and semiarid barrages have different architectures.
Biofilm structure The freshwater biofilm structure generated during experiments was initiated as a single EPS sheet which rapidly developed a basal calcite layer. However, during the autumn fast-flow experiments the biofilm also developed a multilayered structure involving alternations of crystalline calcite layers and cellular (polygonal) EPS zones. Figure 9 shows the basal calcite layer of inter-grown crystals and the overlying polygonal zone EPS (see enlargement of this part in Fig. 4a). In addition, a second layer of more loosely associated calcite crystals occurs above the EPS zone and is succeeded by a second EPS zone also with a cellular structure. This is succeeded by a further calcite layer followed by a less well defined layer of cellular EPS. This multi-layered structure appears to have much in common with marine stromatolite biofilms described by Reid et al. (2000); Ku¨hl et al. (2003); and Decho et al. (2005). In addition, the polygonal structures within the freshwater EPS appear to fulfil a similar role to the micro-polygonal fabrics
MESOCOSM MICROFABRICS
203
Fig. 9. Cross section of a laminated tufa biofilm grown under the uniform conditions throughout its 14 weeks development (6 hours daylight and 10 8C temperature). Note the alternating layering between calcite dominated (crystalline aggregates) and EPS dominated (dehydrated and cellular) areas. The biofilm shows a composite three layer structure in which progressively younger basal calcite layers overbuilt the older cellular EPS fabrics to produce a multilayered freshwater stromatolite. Autumn, fast-flow experiment. Air dried SEM sample.
described by Dupraz et al. (2004) in marine stromatolites. Although the images presented by Dupraz et al. (2004) showed these to be about 2 mm wide (cf. 500 mm in the freshwater mesocosm experiments) they share a similar style of calcite nanospherulite development embedded within polygonal EPS structures to that found in our experiments. These polygonal structures appear too similar to be coincidental and the difference in scaling could be an important point to unravel, particularly if it is dependent on ionic diffusion gradients within the EPS which appears to be key to the regulation of the water– biofilm–carbonate system (Bissett et al. 2008). Light microscopy of the living freshwater biofilm revealed that the development of additional layers (apparently achieved by calcite precipitation within the outer surface of the biofilm) was intimately related to the location of the tangle of dead and coiled filamentous microbes. Precipitation in the surficial part of the EPS has previously been reported by Reid et al. (2000); Dupaz et al. (2004), though from saline and hypersaline sites. The driver for precipitation within the mesocosm biofilms is unclear though it may also be triggered
by the EPS templating and dead cell degradation by heterotrophic bacteria. Importantly, the multilayered mesocosm biofilm (Fig. 9) was seeded and developed during a period of continuing fast-flow, (autumn conditions) without any breaks to calcium ion supply or in flow rate. This suggests that the mesocosm freshwater stromatolite lamination was controlled by organic factors internal to the biofilm rather than by external seasonal variables (see review in Andrews & Brasier 2005). Shiraishi et al. (2008) has suggested that Ca2þ flux within the EPS, caused by photosynthesis-induced calcite precipitation coupled with microbial metabolic functions, as the drivers for developing laminations. Further experimentation is required before drawing firm conclusions.
Growth of crystals within the mesocosm EPS The microspar was predominantly developed from nanospherulite precursors precipitated within the EPS (cf. Fig. 3a). In few cases were nanospherulites seen to be directly in contact with bacterial or cyanobacterial sheath material nor was any replacement of bacteria or cyanobacterial sheath observed.
204
M. PEDLEY & M. ROGERSON
However, 200 –300 mm coccoid and short-rod shaped bacteria frequently occurred in close proximity to the micropeloids and larger micropeloids typically contain EPS-filled cavities many of which also host the same coccoid bacteria. If these bacteria were chemo-organic (heterotrophs) feeding on the EPS (Arp et al. 2001; Turner & Jones 2005) their metabolic activity should inhibit precipitation (Visscher & Stolz 2005) via CO2 production. If they were responsible for the precipitation, as seems likely, this would suggest they were either anaerobes involved in reducing sulphate or iron (Visscher & Stolz 2005), or EPS digestion was releasing sufficient calcium to supersaturate the system despite the presence of abundant respired CO2, or that organisms are manipulating the microenvironment of biofilm interstitial fluid in a manner that promotes precipitation for example via production of extracellular carbonic anhydrases (Li et al. 2005; Kupriyanova et al. 2007). The basal calcite layer also developed from discrete nanospherulite nucleation points within the
EPS. Initially, unrestricted but close stacking of EPS-supported nanospherulites occurred. Each developed progressively into anhedral crystals which eventually became attached to the flume substrate (see Fig. 4b). SEM studies show that crystal growth involved the progressive occlusion of EPS from between and within the growing nanospherulite aggregates and anhedral crystals, though the process rarely reached completion. Consequently, EPS frequently was still present as strands between and within the microspar throughout the biofilm (see Fig. 3a) and must have played a fundamental role in the epitaxial growth of the crystals. The associated crystal perforations (best seen after removal of the EPS) are so common that Pedley et al. (2009) considered this feature was diagnostic of biomediated calcite wherever it occurred naturally. Significantly, the development from nanospherulite to euhedral calcite was not by random chance but rather by a systematic addition of nanospherulites onto a well ordered calcite lattice. Ultimately,
Fig. 10. (a) Nanospherulite crystal building blocks within the EPS. Although apparently randomly arranged these 200–500 nm diameter calcite nanospherulites are pre-arranged into sheets associated with a well ordered calcite atomic lattice. This is illustrated at the right side of the view by the presence of a well developed tabular crystal face. Summer, fast-flow experiment. Air dried SEM sample. (b) Enlarged view of the crystal face showing the nanospherulites arranged into sheets parallel to the rhombic face. Summer, fast-flow experiment. Air dried SEM sample. (c) Basal calcite layer (base is at top) showing the tightly interlocking nature of calcite crystals and the striated compromise boundaries on the central crystal. Autumn, fast-flow experiment. Air dried SEM sample. (d) Close-up of the striated compromise boundary in Figure 10c (same orientation). This reveals that the striations coincide with rows of nanospherulites. The larger spheres scattered on the striated surface are coccoid heterotrophic bacteria.
MESOCOSM MICROFABRICS
by the time that these aggregates reached 2 –10 mm in size many had developed well-formed rhombic crystal faces by tabular stacking of the nanospherulites (see Fig. 10a, b; Pedley et al. 2009). Continued crystal growth, and further reduction of accommodation space within the EPS, ultimately led to the development of a competitive growth fabric in most basal calcite layers. Interestingly, the striated compromise boundary surfaces between adjacent crystals also show that tabular nanospherulite stacking lies parallel to these striations (Fig. 10c, d). Ultimately, many crystals bonded with the flume substrate beneath the EPS. However, some biofilm areas of the basal carbonate layer failed to attach and the floating basal calcite layer developed rhombic crystal terminations at the base of the crystals (cf. Fig. 11a with Fig. 4b). Similarly, the floating calcite microspar crystals within the EPS septal zones also developed free crystal faces (Fig. 11b) although a competitive growth fabric did develop towards the base of the septae and eventually buried the upper surface of the basal layer (see Fig. 4a). Despite the floating nature of crystal growth it is clear that the majority of crystals grew with sub-parallel c-axes which were all orientated normal to the flume surface even though many never became physically attached to it. This process raises a number of fundamental issues related to the development of crystals within freshwater EPS. From the previous discussion it is clear that initially, the cyanobacterial and algal filaments were the only elements within the biofilm to be anchored to and orientated normal to substrate. Subsequently, the polygonal septal
205
zones, most easily defined by nanospherulite micropeloid development within the EPS, also developed normal to biofilm base. Finally, the microspar crystals which developed from these micropeloids also developed long-axis alignment normal to the base of the biofilm. Clearly, the underlying solid substrate exerted no direct control on the primary orientation of the nanospherulite stacking patterns. Consequently, the development of the subsequent calcite crystal orientations cannot be entirely accounted for by the ‘heterogeneous nucleation’ mechanism outlined in Turner & Jones (2005). It is also unlikely that cyanobacterial filaments exert any direct orientation control (cf. template hypothesis of Monty & Mas 1979) because of the general lack of contact between crystal and microbial sheath material, a feature also observed in marine stromatolite mats by Dupraz et al. (2004). However, coccoid (heterotrophic) bacteria are frequently present in surface depressions on and within the precipitates (e.g. Fig. 3d). In addition, there appeared to be dominance of them in the vicinity of the polygonal septae within the EPS. Equally significant was the ramification of EPS throughout these embryonic crystals. Precipitation mechanisms driving crystal growth are becoming better understood (see Rogerson et al. 2010). Briefly, studies of the biological influence on diurnal variations in pH (Rogerson et al. 2008; Liu et al. 2008; Shiraishi et al. 2008; Bissett et al. 2008) indicate that an excess of CO2(aq) is generated throughout the night as a metabolic byproduct of respiration, and consequently promoted a low
Fig. 11. (a) Lower surface of a basal calcite layer which never attached to the flume substrate during development. Consequently, individual crystals within it have developed free rhombic faces by continued growth along their c-axes. The localized absence of basal calcite layer attachment encouraged the development of an undulose upper surface to the biofilm. Note that calcite rhombs at the base of the basal calcite layer have never been in direct contact with the underlying flume surface. Autumn, fast-flow experiment. Air dried SEM sample. (b) An area of basal calcite layer (attached surface to the left) examined immediately after being detached from the flume substrate. Note how freedom of crystal growth on the upper surface of the basal calcite layer is now compromised by downward growth of larger spar crystals within the overlying EPS cover (also with their c-axes orientated normal to substrate). When all EPS between these two crystal sets has become occluded the crystal compromise boundaries will be indistinguishable from a physico-chemically precipitated two generation cement mosaic grown directly on to a solid substrate.
206
M. PEDLEY & M. ROGERSON
nocturnal pH within the EPS. However, at the start of each day the biofilm switches to a dominantly phototrophic metabolism and much of this CO2(aq) is consumed, raising pH and driving a simultaneous 22 conversion of HCO2 3(aq) to CO3(aq). The result of this is an adjustment of the chemical system towards precipitation, effectively ‘mopping up’ additional heterotroph-derived carbon and also the Ca2þ ions, liberated from chelation within the EPS by heterotroph digestion (see Chekroun et al. 2004; Dupraz & Visscher 2005 for further consideration of precipitation processes). The mesocosm experiments show that freshwater biofilms have many similarities with the marine stromatolite model of Reid et al. (2000) and Decho et al. (2005). In both scenarios, in addition to defending the community from physical attack, EPS provides a medium within which ion transportation takes place allowing movement of excess dissolved solids away from individuals to stabilized ‘waste’ sites. In the marine systems, the ‘waste disposal’ occurs at the base of the phototrophic layer, and is enhanced by the presence of anaerobes. Ions are therefore mainly transported vertically, and the biofilm system as a whole shows significant vertical and little horizontal variability. However, in freshwater communities the vertically arranged septae caused calcite to form a mechanically rigid scaffold (Fig. 12) with ‘walls’ enclosing polygonal spaces above a carbonate ‘floor’ developed in close analogy to the marine system. To take the analogy further, the occurrence of multiple layers of EPS implies that some ‘walls’ have ‘roofs’ which are also the ‘floor’ for the next layer. Though the spatial components of elements of the microbial consortium are not fully resolved, it is likely that ‘rooms’ are dominantly occupied by the phototrophs which generate much of the EPS, with boundaries between ‘rooms’ being occupied by aerobic and anaerobic heterotroph communities. Consequently, disposal of EPS waste occurs throughout the biofilm system in freshwater settings providing both vertical and horizontal transportation of ions and, therefore, both vertical and horizontal variability in structure.
Significance of sump grown calcite The original heterotroph colonizers within the sump undoubtedly were derived by biofilm erosion (Rittmann 1989) and re-adhesion (Stolzenbach 1989; Turner & Jones 2005), especially in flume 2 where the high flow velocity created fluid shear at the biofilm-water interface (Characklis 1990). Precipitation in the absence of light has consequences for both tufa and speleothem systems, in addition to precipitation in lakes and soil. Consequently, this precipitate deserves some additional
Fig. 12. Block diagram of the freshwater (tufa) biofilm constructed from optical binocular microscope images. Note the basal calcite layer (black layer at the figure base) initially grows within the EPS and does not always attach to the substrate. Extending vertically away from the basal calcite layer are narrow zones (septae) containing suspended calcite microspar crystals which decrease in size away from the basal calcite layer. At their base these septae extend laterally into a less densely packed crystal layer lying close above the basal calcite layer (unconnected black shapes above the basal calcite layer). When viewed from above the septal zones appear as polygonal zones within the transparent EPS. Inset lower right side: Calcite crystals growing anywhere within the biofilm EPS all grow perfect rhombic terminations and crystal c-axes appear dominantly orientated parallel to the septae (i.e. normal to underlying substrate) despite their lack of direct attachment with a solid substrate.
consideration. Classic considerations of physical degassing and photosynthesis driven enhancement of pH as a precursor to precipitation are probably not sufficient to explain precipitation, as the sump did not have the same potential to degass as the flume does and was unlit. Even the slight diurnal changes imposed on the entire system by photosynthesis external to the sump (Rogerson et al. 2010) is likely to be exceeded by respiration in the sump water. Carbonate buffering of certain anaerobic metabolisms (Visscher & Stolz 2005) may provide a partial explanation for the sump precipitates, with the EPS providing the required nucleation sites. However, it should be noted that nutrition for sump organisms can only be provided by the constant supply of degraded EPS in suspension derived from the photic parts of the flumes. This is an important characteristic to note as digestion of
MESOCOSM MICROFABRICS
this EPS will liberate chelated ions resulting in high [Ca2þ (aq)], and therefore enhanced supersaturation, within sump biofilm interstitial water. The result is that the degraded EPS molecules act as colloids that filter metal ions from the water within the flume and transport these ions down into the sump, where they are released back into solution. If sufficient ions are transported, some of them will eventually be incorporated into precipitates. This colloid filtration mechanism can be expected to be active within lake systems (transporting ions from surface water to the lake bottom) and in caves (transporting ions from the vadose zone into drips and groundwater) and would be a fruitful topic for further investigation.
Precipitate products The absence of vaterite in the mesocosm is perplexing as it is a common microbial associated precipitate in other experiments which have yielded nanospherulites (e.g. Braissant et al. 2003; Nehrke & Van Cappellen 2006). However, in nature though vaterite can occur in micro-oncoids (‘biscuits’ of Giralt et al. 2001) it is more typically recorded within the ‘opalescent’ milky precipitates (whitings) developed within the water column of some tufa pools (e.g. Rolands & Webster 1971; Lu et al. 2000). It is important to note that these data are all effectively from still waters. Consequently, the actively flowing regime, and specifically the high calcium ion supply rate within the mesocosm experiments, may be critical factors controlling mineral species growth. Significantly, vaterite nanospherulites were also absent from fast-flowing natural site samples at Big Hill Springs Provincial Park, Canada (Turner & Jones 2005). Consequently, it is possible that the ion supply rate is a limiting factor in determining the presence or absence of vaterite.
Conclusions The laboratory mesocosm provides easier access to living biofilms and their precipitation products than do natural field sites and allows for the manipulation of environment parameters independently and in incremental stages. The mesocosm experiments described herein successfully simulated natural tufa precipitating conditions and enabled a close study of biofilm development and the initial stages of freshwater stromatolite development. The flowing water experiments reported in this paper provide the following conclusions. The Mesocosm freshwater biofilm community closely parallels the natural community with phototrophs including filamentous green algae and filamentous cyanobacteria
207
(all orientated with filaments normal to substrate); heterotrophic bacteria (coccoids and short rods) and diatoms. Although temperature strongly influences the composition of the biofilm community it is flow rate that influences calcite precipitation rate. Overall, however, there is a fine balance between maximized precipitation and biofilm erosion. Under very high flow (and ion supply) regimes unusual and apparently physico-chemical precipitate morphologies may develop beyond the surface of the biofilm. Consequently, not only does flow rate-regulated ion supply account for the development of barrages, it can also provide a cause for the occurrence of additional, laminated, and coarsely crystalline carbonates at barrage spillover points. Biomediated calcite precipitates commence as micro-peloidal aggregates of nanospherulites suspended within the EPS. These grow rapidly into calcite crystals, by a process involving occlusion of EPS and the addition of further sheets of nanospherulites, to a well defined calcite lattice. The earliest EPS precipitated fabrics form a basal calcite layer within the EPS which becomes attached to the flume substrate. Subsequently, calcite precipitation within the EPS generally is towards the base of the EPS sheet. However, some is localised along linear, vertical polygonal zones. Here, biomediated precipitates (microspar) grow closely associated with bacteria. Typically, the microspar crystals show c-axis orientations normal to the biofilm base. Internally, in the upper part of longer established EPS sheets a new zone of calcite crystals can form and provide a vaulted ‘roof’ to cap the polygon walls. The trigger for the initiation of a new basal layer is not clear though over an 8-month period subsequent to the initial experiments 5 such repeated basal calcite layers were developed, each separated by varying thicknesses of green EPS containing ill-defined polygonal fabrics (Fig. 13a, b). As these basal calcite layers became buried by successively younger biofilm layers they coarsened in grain-size and became coherently cemented into thrombolitic tufa (Fig. 13c). In this way, the development of a multilayered fabric composed of alternations of calcite laminae separated by EPS leads directly to the development of a well laminated freshwater stromatolite. Progressively, as the earlier laminae became more deeply buried the green colour (phototrophs) disappeared and the EPS became darker. These darker EPS zones are often the sites of further calcite precipitates which progressively occlude the EPS. Complete occlusion appears rarely to occur and the older EPS disappears from the fabric leaving open cavities which contribute to the characteristic cellular microfabric seen in ancient
208
M. PEDLEY & M. ROGERSON
Fig. 13. (a) Profile of the flume barrage (hachures) with biofilm growth (stipple) and cemented tufa (black). This sketch was made after 8 months of continuous development under the same temperature, calcium ion supply and flow regime (autumn experiment, fast-flow). The profile of the tufa illustrates the tendency of the biofilm to prograde down-flow at the barrage top. Note that all sketches in Figure 13 are drawn from hydrated, living material taken directly from the flume experiments. (b) Profile of the biofilm immediately down-flow of the barrage (stippling represents the calcite crystal precipitates within the EPS. Note the large alveolus developed on the right side which is lined by EPS but filled with flume water. (c) Profile at the spill-over point of the barrage where the deposit is thickest. The upper 4 mm consists of living green coloured EPS containing several zones of fine calcite crystals dispersed within the EPS. The living biofilm transitionally overlies 5 mm of well cemented crystalline tufa with a thrombolitic texture. Note that the tufa layer is coarsely granular and totally without biofilm. The interconnected alveoli are filled with flume water. (d) Profile from the fast-flow flume down-flow of the barrage. Note the extensive development of water filled alveoli within the biofilm. This gives a flexible sheet which accommodated flow variability with time. It develops by the addition of extra laminae which prograde down-flow (one is shown on the right side of the sketch). The stippled areas represent calcite crystal precipitates within the EPS which is well developed in the older layers but generally is absent from the surface biofilm layer. The alveoli are elongated parallel to flow and are filled with flume water.
tufas. Alternatively, the cavities within the thrombolytic tufa fabric may develop from water filled alveoli within successive biofilm layers. These are best seen in the cushion-like biofilms developed under fast flow regimes (see Fig. 13d) where they irregularly separate successive biofilm layers. Calcite crystals develop both within the fenestral biofilm layers (coarse stipple in Fig. 13d) and can grow on the surface of the alveoli in deeper buried sites. Significantly, these multilayered stromatolite fabrics developed under constant conditions of temperature, illumination conductivity and flow regime during the experimental runs. Clearly,
diurnal or ‘seasonal’ changes are not necessary for their development. The only introduced variable was the addition of nutrients on five occasions. At the present time it would be speculative to correlate these with the 5 or more basal calcite layers present (see Fig. 13b) though it is quite likely that nutrient addition may change the overall microbial dominance within the biofilm. However, the only evidence at present is an apparent ‘greening’ of the biofilm over a few days subsequent to nutrients being added. Considerable precipitation is also recorded within the unlit areas (sump) in association with heterotroph dominated EPS. Such aphotic communities
MESOCOSM MICROFABRICS
are known to exist in natural subterranean situations (see Blyth & Frisia 2008). Consequently, lessons learnt from the study of microbially mediated precipitation in tufa systems may well be relevant to understanding the geochemical records of speleothems. Of particular importance may be the mechanism of ‘colloid filtration’, whereby the EPS molecules collect ions in one location and then concentrate and release them back into solution at a second location, potentially a significant distance away. This mechanism may also be assumed to play some role in promoting in-sediment precipitation within lakes. The authors are pleased to thank Dr John Adams for assistance with the staining and identification of the bacteria and phototrophic cyanobacteria. Thanks also to Mr Tony Sinclair for assistance with the scanning electron microscopy.
References A NDREWS , J. E. & B RASIER , A. T. 2005. Seasonal records of climatic change in annually laminated tufas: short review and future prospects. Journal of Quaternary Science, 20, 411–421. A RP , G., W EDEMEYER , N. & R EITMER , J. 2001. Fluvial tufa formationin a hard water creek (Deinschwanger Bach, Franconian Alb, Germany). Facies, 44, 1– 22. B ENZERARA , K., M ENGUY , N. ET AL . 2006. Nanoscale detection of organic signatures in carbonate microbialites. Proceedings National Acadamy of Science, 103, 9440–9445. B ISSETT , A., R EIMER , A., D E B EER , D., S HIRAISHI , F. & A RP , G. 2008. Metabolic microenvironmental control by photosynthetic biofilms under changing macroenvironmental temperature and pH conditions. Applied and Environmental Microbiology, 74, 6306–6312. B LYTH , A. J. & F RISIA , S. 2008. Molecular evidence for bacterial mediation of calcite formation in cold highaltitude caves. Geomicrobiology Journal, 25, 101– 111. B OURROUILH -L E , F. G., A KRAM , J. B. & S CHVOERER , M. 2007. Band-e-Amir Lakes and Dragon Valley (Bamiyan): myths and seismicity in Afghanistan. In: P ICCARDI , L. & M ASSE , W. B. (eds) Myth and Geology. Geological Society, London, Special Publications, 273, 121–132. B RAISSANT , O., C AILLEAU , G., D UPREZ , C. & V ERRECCHIA , E. P. 2003. Bacterially induced mineralization of calcium carbonate in terrestrial environments: the role of exopolysaccharides and amino acids. Journal of Sedimentary Research, 73, 485–490. B RETT , T. E. W. 1970. A Study of Tufa Deposition in the Band-e-Amir Lakes of Central Afghanistan. In: Afghanistan Expedition Report, University of Bristol, 14. C AUDWELL , C., L ANG , J. & P ASCAL , A. 2001. Lamination of swampy-rivulets Rivularia haematites stromatolites in a temperate climate. Sedimentary Geology, 143, 125–147. C HARACKLIS , W. G. 1990. Biofilm processes. In: C HARACKLIS , W. G. & M ARSHALL , K. C. (eds) Biofilms. John Wiley and Sons, New York, 195– 231.
209
C HEKROUN , K. B., R ODRIGUEZ -N AVARO , C., G ONZALEZ -M UNOZ , M. T., A RIAS , J. M., C ULTRONE , J. & R ODRIGUEZ -G ALLEGO , M. 2004. Precipitation and growth of calcium carbonate induced by Myxococcus xanthus: implications for recognition of bacterial carbonates. Journal of Sedimentary Research, 74, 868– 876. C HEN , J., Z HANG , D. D., W ANG , T., X IAO , T. & H UANG , R. 2004. Factors controlling tufa deposition in natural waters at waterfall sites. Sedimentary Geology, 166, 353– 366. D ECHO , A. W., V ISSER , P. T. & R EID , P. 2005. Productuion and cycling of natural microbial exopolymers (EPS) within a marine stromatolite. Palaeogeography, Palaeoclimatology, Palaeoecology, 219, 71–86. D UPRAZ , C. & V ISSCHER , P. T. 2005. Microbial lithification in marine stromatolites and hypersaline mats. Trends in Microbiology, 13, 429–438. D UPRAZ , C., V ISSCHER , P. T., B AUMGARTNER , L. K. & R EID , R. P. 2004. Microbe–mineral interactions: early precipitation in a hypersaline lake (Eluthera Island, Bahamas). Sedimentology, 51, 745–765. E RCOLE , C., C ACCHIO , P., B OTTA , A. L., C ENTI , V. & L EPIDI , A. 2007. Bacterially induced mineralization of calcium carbonate: the role of exopolysaccharides and capsular polysaccharides. Microscopy and Microanalysis, 13, 42– 50. F LETCHER , M. & F LOODGATE , G. D. 1976. The adhesion of bacteria to solid surfaces. In: F ULLER , R. & L OVELOCK , D. W. (eds) Microbial Ultrastructure. Academic Press, London, 101–107. F RANCˇ ISˇ KOVIC´ -B ELINSKI , S., B ARISˇ IC´ , D., V ERTACˇ NIK , A., B ILINSKI , H. & P ROHIC´ , E. 2004. Characterisaation of tufa from the Dinaric Karst of Croatia: mineralogy, geochemistry and discussion of climate conditions. Facies, 50, 183–193. F REYTET , P. & V ERRECCHIA , E. P. 1998. Freshwater organisms that build stromatolites: a synopsis of biocrystallization by prokaryotic and eukaryotic algae. Sedimentology, 45, 535–563. G IRALT , S., J ULIA` , R. & K LERKX , J. 2001. Microbial Biscuits of Vaterite in Lake Issyk-Kul (Republic of Kyrgyzstan). Journal of Sedimentary Research, 71, 430– 435. H AMMER , O. 2008. Watch your step. Nature Physics, 4, 265– 266. H AMMER , O., D YSTHE , D. K., L ELU , B., L UND , H., M EAKIN , P. & J AMTVEIT , B. 2008. Calcite precipitation instability under laminar, open-channel flow. Geochimica et Cosmochimica Acta, 72, 5009– 5021. K ANO , A., M ATSUOKA , J., K OJO , T. & F UJII , H. 2003. Origin of annual laminations in tufa deposits, southwest Japan. Palaeogeography, Palaeoclimatology, Palaeoecology, 191, 243– 262. K U¨ HL , M., F ENCHEL , T. & K AZMIERCZAK , J. 2003. Growth, structure and calcification potential of an artificial cyanobacterial mat. In: K RUMBEIN , W. E., P ATERSON , D. & Z AVARIN , G. (eds) Fossil and Recent Biofilms, a Natural History of Life on the Planet. Kluwer, Dordrecht, 77–87. K UPRIYANOVA , E., V ILLAREJO , A. ET AL . 2007. Extracellular carbonic anhydrases of the stromatolite-forming
210
M. PEDLEY & M. ROGERSON
cyanobacterium Microcoleus chthonoplastes. Microbiology, 153(4), 1149–1156. L I , W., Y U , L. J., H E , Q. F., W U , Y., Y UAN , D. X. & C AO , J. H. 2005. Effects of microbes and their carbonic anhydrase on Ca2þ and Mg2þ migration in column-built leached soil-limestone karst systems. Applied Soil Ecology, 29(3), 274–281. L IU , Z. H., L IU , X. L. & L IAO , C. J. 2008. Daytime deposition and nighttime dissolution of calcium carbonate controlled by submerged plants in a karst spring-fed pool: insights from high time-resolution monitoring of physico-chemistry of water. Environmental Geology, 55(6), 1159–1168. L U , G., Z HENG , C., D ONAHOE , R. J. & B ERRY L YONS , W. 2000. Controlling processes in a CaCO3 precipitating stream in Huanglong Natural Scenic District, Sichuan, China. Journal of Hydrology, 230, 34– 54. M ONTY , C. L. V. & M AS , J. R. 1979. Lower Cretaceous (Wealden) blue-green algal deposits of the province of Valencia, eastern Spain. In: M ONTY , C. V. L. (ed.) Phanerozoic Stromatolites. Springer Verlag, Berlin, 85–120. N EHRKE , G. & V AN C APPELLEN , P. 2006. Framboidal vaterite aggregates and their transformation into calcite: a morphological study. Journal of Crystal Growth, 287, 528– 530. P EDLEY , H. M. 1990. Classification and environment models of cool freshwater tufas. Sedimentary Geology, 68, 143– 154. P EDLEY , H. M. 1992. Freshwater Phytoherm reefs: the role of biofilms and their bearing on marine cementation. In: P EDLEY , H. M., H ARWOOD , G. & S ELLWOOD , B. (eds) Ramps and Reefs Sedimentary Geology, Special Volume, 79, 255– 274. P EDLEY , H. M. 1994. Prokaryote-microphyte biofilms: a sedimentological perspective. Kaupia, Darmsta¨dter Beitra¨ge zur Naturgeschichte, 4, 45–60. P EDLEY , H. M., A NDREWS , J., O RDONEZ , S., G ONZALES -M ARTIN , J. A., G ARCIA D EL C URA , M. A. & T AYLOR , D. 1996. Does climate control the morphological fabric of freshwater carbonates?: a comparative study of Holocene barrage tufas from Spain and Britain. Palaeogeography, Palaeoclimatology, Palaeoecology, 121, 239–257. P EDLEY , H. M., O RDONEZ , S., G ONZALES -M ARTIN , J. A. & G ARCIA D EL C URA , M. A. 2003. The sedimentology of perched springline and paludal tufas: criteria for recognition, with examples from Guadalajara Province, Spain. Sedimentology, 50, 23–44. P EDLEY , M., R OGERSON , M. & M IDDLETON , R. 2009. Freshwater calcite precipitates from in vitro mesocosm flume experiments: a case for biomediation of tufas. Sedimentology, 56, 511– 527. R EID , R. P., V ISSCHER , P. T. ET AL . 2000. The role of microbes in accretion, lamination and early
lithification of modern marine stromatolites. Nature, 406, 989 –992. R ITTMANN , B. E. 1989. Detachment from biofilms. In: C HARACKLIS , W. G. & W ILDER , P. A. (eds) Structure and Function of Biofilms. Wiley, Chichester, 49–58. R OGERSON , M., P EDLEY , H. M., W ADHAWAN , J. D. & M IDDLETON , R. 2008. New insights into biological influence on the geochemistry of freshwater carbonate deposits. Geochimica et Cosmochimica Acta, 72, 4976– 4987. R OGERSON , M., P EDLEY , H. M. & M IDDLETON , R. 2010. Microbial influence on macroenvironment chemical conditions in alkaline (tufa) streams: perspectives from in vitro experiments. In: P EDLEY , H. M. & R OGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 65–82. R OLANDS , D. L. G. & W EBSTER , R. K. 1971. Precipitation of vaterite in lake water. Nature (Physical Sciences), 299, 158. S HIRAISHI , F., R EIMER , A., B ISSETT , A., D E B EER , D. & A RP , G. 2008. Microbial effects on biofilm calcification, ambient water chemistry and stable isotope records in a highly supersaturated setting (Westerhofer Bach, Germany). Palaeogeography, Palaeoclimatology, Palaeoecology, 262, 91–106. S TOLZENBACH , K. D. 1989. Particle transport and attachment. In: C HARACKLIS , W. G. & W ILDER , P. A. (eds) Structure and Function of Biofilms. Wiley, Chichester, 33–48. T AYLOR , D. M., G RIFFITHS , H. I., P EDLEY , H. M. & P RINCE , I. 1994. Radiocarbon dated Holocene pollen and ostracod sequences from barrage tufa-dammed fluvial systems in the White Peak, Derbyshire. The Holocene, 4, 356– 364. T URNER , E. C. & J ONES , B. 2005. Microscopic calcite dendrites in cold-water tufa: implications for nucleation of micrite and cement. Sedimentology, 52, 1043– 1066. V EYSEY , J. & G OLDENFELD , N. 2008. Watching rocks grow. Nature (Physical Sciences), 4, 310–313. V ISSCHER , P. T. & S TOLZ , J. F. 2005. Microbial mats as bioreactors: populations, processes, and products. Palaeogeography, Palaeoclimatology, Palaeoecology, 219, 87– 100. Z AIHUA , L., S VENSSON , U., D REYBRODT , W., D AOXIAN , Y. & B UHMANN , D. 1995. Hydrodynamic control of inorganic calcite precipitation in Huanglong Ravine, China: field measurements and theoretical prediction of depositional rates. Geochimica et Cosmochimica Acta, 59, 3087– 3097. Z HANG , D. D., Z HANG , Y., Z HU , A. & C HENG , C. 2001. Physical Mechanisms of River Waterfall Tufa (Travertine) Formation. Journal of Sedimentary Research, 71, 205–216.
Nanotextures of aragonite in stromatolites from the quasi-marine Satonda crater lake, Indonesia KARIM BENZERARA1*, ANDERS MEIBOM2, QUENTIN GAUTIER1, ´ JOZEF KAZ´MIERCZAK3, JAROSŁAW STOLARSKI3, NICOLAS MENGUY1 & GORDON E. BROWN, JR4,5 1
Equipe Ge´obiosphe`re Actuelle et Primitive, IMPMC & IPGP, UMR 7590, CNRS, Universite´s Paris 6 et IPGP. 140, rue de Lourmel, 75015 Paris, France
2
Muse´um National d’Histoire Naturelle, Laboratoire de Mine´ralogie et Cosmochimie du Muse´um (LMCM), UMR 7202, Paris, France
3
Institute of Paleobiology, Polish Academy of Sciences, Twarda 51/55, 00818 Warsaw, Poland 4
Surface & Aqueous Geochemistry Group, Department of Geological & Environmental Sciences, Stanford University, Stanford, CA 94305-2115, USA
5
Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA *Corresponding author (e-mail:
[email protected]) Abstract: Stromatolites have been extensively used as indicators of ancient life on Earth. Although much work has been done on modern stromatolites, the extent to which biological processes control their structure, and the respective contributions of biological and abiotic processes in their formation are, however, still poorly constrained. A better description of the mineralogical textures of these formations at the submicrometre scale may help improve our understanding of how carbonates nucleate and grow in stromatolites. Here, we used a combination of microscopy and microspectroscopy techniques to study the chemical composition and the texture of aragonite in lacustrine stromatolites from the alkaline crator lake in Satonda, Indonesia. Several textural features are described, including morphological variations of aragonite from nanosized grains to micrometre-sized fibres, the presence of striations in the aragonite laminae showing a striking similarity with growth bands in corals, and clusters of small aragonite crystals sharing a common crystallographic orientation. These nanotextural features are compared with those observed in scleractinian corals, and possible processes involved in their formation are discussed.
Stromatolites are laminated sedimentary growth structures, usually composed of calcium carbonates, with growth initiated from a point or a limited surface (Semikhatov et al. 1979). They have been found throughout the geological record as far back as 3.5 Ga ago (e.g. Hofmann 2000). As modern stromatolites are systematically and intimately associated with microbial communities, ancient stromatolites have often been considered to be one of the oldest traces of life on Earth (Hofmann 2000; Altermann et al. 2006; Schopf et al. 2007). The biogenicity of many ancient stromatolites has, however, been questioned in particular because very few microfossils have been found in these samples (e.g. Lowe 1994; Grotzinger & Rothman 1996; Lepot et al. 2008) and similar morphological patterns can be produced abiotically (Grotzinger & Rothman 1996; McLoughlin et al. 2008). The
macroscopic morphology (e.g. Allwood et al. 2006) as well as the mesostructure (Shapiro 2000) of stromatolites are important tools presently used to infer the biogenicity of ancient stromatolites. New numerical models allowing exploration of the various possible morphologies of stromatolites and assessment of the respective roles of environmental and biological processes in their formation have been proposed by Dupraz et al. (2006). These authors note that a detailed mechanistic description of how stromatolites are formed is still warranted. Such a description may benefit from a more detailed characterization of the mineral and organic building units of stromatolites and their arrangement at the submicrometre-scale, that is, the scale at which nucleation and mineral growth take place. However, one difficulty is that this approach requires the use of specialized analytical tools, as the building units of
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 211–224. DOI: 10.1144/SP336.10 0305-8719/10/$15.00 # The Geological Society of London 2010.
212
K. BENZERARA ET AL.
carbonates can be very small (in the few tens of nanometre size range). In the present study, we focus on the morphology and arrangement of carbonate crystals in recent subfossil (possibly several hundreds of years old) lacustrine stromatolites. The combination of microscopy and microspectroscopy tools such as scanning and transmission electron microscopy (SEM and TEM), NanoSIMS and synchrotron-based scanning transmission X-ray microscopy (STXM) provides unique information on the compositional variations and mineral textures or microfabrics of stromatolites at the submicrometre scale. Several textural features show a remarkable similarity to those observed in modern corals. These similarities as well as differences are discussed in light of the mineralization processes by which modern stromatolites and corals form. Although these results are of interest to the scientific community working on microbialites, the approach and the basic mineral building units that are described should be of general interest to communities working on carbonate precipitates such as tufas and speleothems.
Material and methods Stromatolite samples Modern subfossil stromatolite samples were collected from the crater lake of Satonda, a small volcanic island in Indonesia previously described in detail by Kempe & Kaz´mierczak (1993, 2007), Arp et al. (2003) and Kazmierczak & Kempe (2004). We studied sample S-47, collected with from the stromatolitic reefs at station 10 above the water table (see Kempe & Kazmierczak, 1993). These stromatolites are actually subfossil and, although we do not know precisely their age, may be as old as several hundred years. Some features may have been modified by aging (e.g. some organic functions), but the transformation of mineral textures seems unlikely considering that: (1) the samples have remained at the surface; (2) they are still composed of aragonite and not calcite as would be expected from prolonged interaction with meteoric water; and (3) the same mineral textures are observed in other genuinely modern microbialites from alkaline lakes (e.g. from Lake Van in Turkey). As a comparison, the earliest steps of diagenesis of aragonite structures, which can be observed after a few years, have been studied for example by Perrin & Smith (2007). Satonda Lake is a slightly alkaline (pH 8.55) quasi-seawater system that harbors calcareous stromatolites along the shore. The laminations of Satonda stromatolites consist of alternating 50–500 mm thick aragonite laminae and 1 to a few tens of micrometres thin Mg –Si-containing
layers. Some fragments of the Satonda stromatolites were powdered for XRD and TEM analyses to characterize the nature and morphology of the crystals comprising them. In order to spatially resolve textural and chemical information within the structure of Satonda stromatolites, petrographic thin sections were prepared and gold-coated for SEM and NanoSIMS analyses. Some spots were selected for preparation of ultrathin electrontransparent foils (less than 100 nm in thickness) by Focused Ion Beam milling. These foils were further analysed at the few nanometer spatial scale by scanning transmission X-ray microscopy (STXM) and TEM.
Scanning electron microscopy SEM analyses were performed on the Zeiss Supra 55 SEM microscope in the Laboratory Magie at University Pierre et Marie Curie (Paris, France). The microscope was operated at 10 kV with a working distance of 3 mm. Two different detectors were used: an in-Lens detector for secondary electron imaging (nano-topography of the sample) and an Angle selective Backscattered (AsB) detector for low-angle backscattered electrons which provide a contrast more sensitive to crystal orientation. Following NanoSIMS analyses, the gold coating was removed from the petrographic sections and the samples were etched using a slightly acidic solution consisting of formic acid diluted to 0.1% in de-ionized milliQ water with 2% glutaraldehyde. The samples were re-coated with gold-palladium and observed by SEM.
Focused ion beam milling Focused ion beam (FIB) milling was performed with a FEI Model 200 TEM FIB system at the University Aix-Marseille III. The FIB lift-out method was used to prepare a cross-section across one Mg –Si-rich area surrounded by two aragonite laminae. This method is described in detail in Heaney et al. (2001) and Benzerara et al. (2005). A thin strip of platinum was deposited on the area of interest in order to protect it during the milling process. A 30 kV Gaþ beam operating at c. 20 nA excavated the sample from both sides of the Pt strip to a depth of 5 mm. Before removal of the thin foil, it was further thinned to c. 100 nm with a glancing angle beam at much lower beam currents of c. 100 pA. Finally, a line pattern was drawn with the ion beam along the side and bottom edges of the thin foil allowing its removal from the sample. The slide was transferred at room pressure with a micromanipulator onto the membrane of a formvar-coated 200 mesh copper grid for TEM and STXM analyses.
NANOTEXTURES OF ARAGONITE IN STROMATOLITES
Transmission electron microscopy Two types of samples were used for TEM work: (1) powders that were suspended for a few seconds in distilled water then deposited on the membrane of a lacey carbon-coated 200 mesh copper grid and air dried; and (2) the FIB ultrathin foil. TEM observations were carried out at IMPMC on a JEOL 2100F microscope operating at 200 kV, equipped with a field emission gun and a high resolution UHR pole piece. A double-tilt sample holder was used to orient samples during the collection of TEM images and electron diffraction patterns from the samples.
Scanning transmission X-ray microscopy (STXM) STXM is a type of transmission microscopy that uses a monochromated X-ray beam produced by synchrotron radiation. The rationale for STXM data acquisition and analysis and examples of applications can be found in Hitchcock (2001) and Bluhm et al. (2006). For STXM imaging, the X-ray beam is focused on an X-ray transparent sample using a zone plate, and a 2-D image is collected by scanning the sample at a fixed photon energy. The achieved spatial resolution is dependent on the zone plate and was c. 25 nm in the present study. The image contrast results from differential absorption of X-rays, which partly depends on the chemical composition of the sample. In addition to imaging, it is possible to perform at the same spatial resolution, near edge X-ray absorption fine structure (NEXAFS) spectroscopy at the carbon K-edge (or other absorption edges in the 80 –2000 eV energy range) which gives information on the speciation (i.e. type of functional group and bonding) of carbon (or other elements). In this present case, it was thus possible to detect an absorption feature at 290.3 eV corresponding to 1s ! p * electronic transitions in the carbonate group. It has been shown previously that the absorption intensity at 290.3 eV depends on the orientation of the aragonite c-axis (i.e. the direction of the carbonate p * orbitals) relative to the polarization of the synchrotron X-ray beam (e.g. Metzler et al. 2008). This phenomenon is known as linear dichroism. The absorption at 290.3 eV is maximum when the c-axis of aragonite parallels the polarization vector of the beam and minimum when they are perpendicular (see Fig. 6). At 290.3 eV, we acquired ten different images of the FIB foil, each at a different polarization of the synchrotron X-ray beam ranging from horizontal (noted as 08) to vertical (noted as 908). These images were converted to a linear absorbance (optical density, OD) scale OD ¼ ln(I=I0 )
(1)
213
where I is the transmitted intensity and I0 is the incident beam measured in a region adjacent to the foil that contains no sample. The resulting images were aligned carefully. The carbonate OD of each individual pixel was then fit using a routine written with the statistical software R as a function of the polarization angle using the following equation: OD(a) ¼ C þ A cos2 (a f)
(2)
where a is the polarization of the incident X-ray beam (from 0 to 908), f is the angular orientation of the c-axis of aragonite (from 0 to 1808), C is the in-plane polarization-independent fraction of the absorption, and A is the amplitude of the polarization-dependent absorption. From these measurements the crystallographic orientation of the c-axis (f) of aragonite was plotted for each individual pixel (here, at the 50-nm scale) on the whole FIB foil providing the map reported in Figure 6. STXM observations were performed at the Advanced Light Source (ALS) (Lawrence Berkeley National Laboratory) on MES branch line 11.0.2.2. During our measurements, the ALS storage ring operated at 1.9 GeV and 200–400 mA stored current. A 1200 l/mm grating and 40 mm exit slits in both dispersive and non-dispersive direction were used for carbon imaging and spectroscopy, providing a theoretical energy resolution of 72 meV. Energy calibration was accomplished using the well-resolved 3p Rydberg peak at 294.96 eV of gaseous CO2 for the C K-edge.
NanoSIMS Spatially resolved determination of Mg/Ca and Sr/Ca ratios along profiles within the carbonate laminae of the Satonda stromatolites was carried out using the Cameca NanoSIMS N50 at the Muse´um National d’Histoire Naturelle in Paris with the same settings as in Meibom et al. (2007, 2008). Using a primary beam of O2, secondary ions of 24Mgþ, 40Caþ, and 88Srþ were sputtered from the sample surface and detected simultaneously in electron multipliers at a mass resolving power of c. 4000. At this mass-resolution all potentially problematic interferences are resolved. The data were obtained from a pre-sputtered surface in a series of line-scans with the primary ions focused to a spot-size of c. 150 nm. Magnesium and Sr concentrations were calibrated against carbonate standards of known composition.
Results The modern subfossil stromatolites from Satonda that have been analyzed in the present study consist of aggregated mm-sized laminated circular
214
K. BENZERARA ET AL.
overgrowths on filamentous green algae thalli (Fig. 1). These samples show a facies that was named ‘stromatolitic-siphonocladalean’ and ‘peloidal zone’ by Kempe & Kaz´mierczak (1993) and
‘Green-Algal Microbialite’ by Arp et al. (2003). This facies comprises a significant part of the Satonda reef framework the bulk of which is composed of marine stenohaline calcareous red algae
Fig. 1. Observation and compositional analysis of carbonate laminae in Satonda stromatolites. (a) Optical microscopy image of the laminated circular overgrowths on filamentous green algae forming the framework of Satonda stromatolites. Concentric dark (Mg– Si rich) and bright (aragonite) laminae can be observed. (b) Optical microscopy image of one aragonite lamina at higher magnification showing striations parallel to the general direction of the lamina accretion. The average distance between striations is around 1 mm. (c) Secondary electron SEM image of the laminae which have been slightly etched. The fibrous texture of the aragonite laminae as well as numerous striations are visible. (d) NanoSIMS measurements of the Sr/Ca and Mg/Ca ratios along a 50-mm segment profile.
NANOTEXTURES OF ARAGONITE IN STROMATOLITES
(Peyssonnelia and Lithoporella) and nubecullinid foraminifers. The laminated overgrowths show alternation of a poorly crystallized Mg– Si-rich mineral phase (thin and dark in transmitted light optical microscopy) described by Arp et al. (2003) and aragonite, forming thick and bright laminae in transmitted light optical microscopy. We focused on the aragonite laminae for the textural analyses. A detailed description of the texture of aragonite laminae might help to better understand the potential role of microorganisms in their formation. Aragonite laminae can be as thick as a millimetre and most of them show fine striations in transmitted light microscopy parallel to the accretion planes of the laminae (Fig. 1). This striated texture could also be observed by SEM on samples etched gently with formic acid. These striations were also described by Kempe & Kaz´mierczak (1993), but their origin remains mysterious. Interestingly, these striations show a striking similarity with the growth bands observed in the fibrous aragonite skeleton of scleractinian corals (Fig. 2, e.g. Meibom et al. 2004). As these bands correlate with significant Mg and Sr variations in corals, we measured the same elemental profiles within the aragonite laminations by NanoSIMS (Fig. 1). In contrast to what has been observed for corals (e.g. Meibom et al. 2007, 2008), no significant variations of the Sr/Ca ratio were observed in the aragonite laminae. The slight increase of Mg/Ca at 36 microns was very small compared to what is usually observed in corals and moreover, this ratio was constant for the rest of the profile and did not show systematic variations within the striations observed by SEM. TEM observations on powdered samples of the crystal units forming the Satonda stromatolites showed only a poorly crystallized Mg –Si-rich phase and aragonite, consistent with bulk XRD
215
measurements. Although powdering destroys the textural relationship between crystals, these observations were nonetheless useful because they showed that the aragonite can have two different morphologies (Fig. 3): (1) some aragonite appears as single-crystal fibres that are a few micrometres in length and a few hundred nanometres in width; and (2) the remainder of the aragonite appears as small globules a few tens of nanometres in size and forms polycrystalline clusters. This latter texture is usually called micritic and has been observed by many authors in the past (e.g. Riding 2000; Dupraz et al. 2004). To obtain more information on the spatial distribution of these types of morphologies in the aragonite laminae of Satonda stromatolites, non-etched petrographic thin sections were examined by SEM. As previously reported by Kempe & Kaz´mierczak (1993) and Arp et al. (2003), the aragonite laminations appear at first sight to be mostly fibrous, with aragonite fibres perpendicular to the laminae surface (Fig. 4). However, at higher magnification, these fibres seem to consist of clusters of nanocrystals that are similar in size to the nanocrystals found in the powdered samples. SEM observations of such tiny domains may however be misleading as they probe the very near surface of the sample which might be very sensitive to the sample preparation protocol, to slight etching during the polishing stage and/or to artifactual nanograins that may have formed during the gold-palladium coating procedure (Gibbs & Powell 1996; Steele et al. 1998). The use of other techniques that require minimal sample preparation such as cryo-SEM and/or environmental SEM would be a useful approach to better address this issue (e.g. Dupraz et al. 2004). Another approach consists of preparing the samples for techniques that are not surface-sensitive such as TEM or STXM.
Fig. 2. Comparative SEM images of Porites sp. skeleton (a) and Satonda stromatolites (b). Both samples were gently etched with formic acid (1%). (a) Centres of rapid accretion can be observed as pits in the centre of the image. Concentric growth bands as well as aragonite fibres are visible on this etched sample. (b) Two pits can be observed at the center of the structure. They have classically been interpreted as remnants of green algae around which the carbonates precipitate. Concentric growth bands can be observed. Aragonite fibres can be observed at higher magnification (see Fig. 4).
216
K. BENZERARA ET AL.
Fig. 3. TEM observation of powders of the Satonda stromatolites. (a) Bright field image showing aragonite fibres as well as clusters of very small aragonite grains. The circles indicate the locations where electron diffraction patters were measured (upper one on the cluster, lower one on a fibre). (b) The powder pattern shows spacings at 4.2, 3.4, 2.9, ˚ , which can be indexed as due to diffraction from the (110), (111), (002), (121) and (200) planes of 2.7 and 2.5 A aragonite. (c) The single crystal diffraction pattern is consistent with aragonite and was measured along the [1– 10] zone axis.
We investigated the crystallographic orientation of the aragonite in more detail by TEM and STXM using a FIB-milled ultrathin foil, which was cut across a Mg–Si-rich layer (Fig. 5). On both sides of this layer, we observed the top of the underlying aragonite laminae (left in Fig. 5) and the bottom of the overlying aragonite laminae (right in Fig. 5). The two distinct morphologies of aragonite observed in the powdered samples were also observed in this foil. In the top part of the underlying aragonite laminae, aragonite appears as bundles of single-crystal fibres, with their growth axes parallel to each other and perpendicular to the laminae. In contrast, aragonite in the overlying aragonite laminae appears as a massive cluster of tiny aragonite crystals (c. 50 nm in size). The orientation of the aragonite crystals in the FIB foil can be assessed by electron diffraction. Single-crystal electron diffraction patterns were obtained using a 100-nm large aperture on the fibre area. These patterns show unambiguously that aragonite fibres share a common crystallographic orientation (Fig. 5). Powder electron diffraction patterns were obtained on the clusters of tiny aragonite crystals using a 1-mm large aperture. The clustering of some spots on some rings suggests some orientation of the
nanocrystals. It is difficult, however, to get a more comprehensive view of the crystallographic orientation of aragonite in the foil. To obtain a more complete view of the crystallographic orientation of these very small aragonite domains, we thus carried out STXM imaging and polarizationdependent imaging contrast on the entire FIB-milled foil. STXM has classically been used in the Earth Sciences to characterize the speciation of diverse elements such as C, N, O and heavy metals and metalloids such as at the nanoscale (e.g. Haberstroh et al. 2006; Bernard et al. 2007; Benzerara et al. 2008; Lepot et al. 2008). Some organic carbon could be detected by STXM in the FIB foil (data not shown); however, it was not possible to decipher whether this organic carbon was indigenous or was an artifact resulting from impregnation of the sample with epoxy. Here, we used the sensitivity of X-ray absorption spectroscopy to the crystallographic orientation of aragonite, an effect commonly referred to as X-ray linear dichroism (e.g. Metzler et al. 2008; Zhou et al. 2008). By varying the direction of the polarization vector of the X-ray beam from 0 to 908 in 108 steps and measuring the absorption of X-rays at 290.3 eV for each 50-nm pixel of the FIB-milled foil, it was possible
NANOTEXTURES OF ARAGONITE IN STROMATOLITES
217
Fig. 4. SEM observation of polished sections of the Satonda stromatolites. (a) Image of an aragonite lamina collected in the low angle backscattered electron mode and revealing crystal directions. The fibrous texture is visible and the aragonite crystals are mostly oriented NNE–SSW. (b) Image of the same area collected in the secondary electron mode and showing the aragonite fibres. (c) At higher magnification, these fibres show a rough surface and seem to consist of an aggregation of tiny crystals.
to map the orientation of the in-plane projection of the c-axis of aragonite (Fig. 6). Interestingly, although TEM provides a better spatial resolution for imaging than STXM, it is limited for electron diffraction by the size of the smallest aperture, that is, in this case, 100 nm. This is larger than the spatial resolution achieved by STXM for the mapping of the orientation of aragonite c-axis as shown hereafter. Finally, STXM polarization-dependent imaging contrast is much less sensitive to variations of the orientation of aragonite than electron diffraction, in particular out-of-plane variations, and can thus be more efficient in providing an average view of the crystallographic texture of aragonite in the whole FIB foil. From this dataset, several observations can be made. First, the aragonite laminae are composed of micrometre-sized domains that respond homogeneously to the polarization changes (see areas 1, 2, 3 and 4 in Fig. 6). This response means that the in-plane projections of the c-axes
of the grains present within these areas share a common orientation. Area 4 corresponds to the bundle of aragonite fibres observed by TEM in the aragonite laminae underlying the Mg –Si-rich laminae. The whole area is more absorbing at 08 than at 908. The fit of the variation of absorption in A cos2(u 2 f) gives a single value for the orientation of the c-axis of aragonite (f) which is around 1558 (equivalent to 2258). The STXM data thus confirm that the c-axes of these crystals (which are also the growth axes of aragonite) are roughly parallel to each other and parallel to the growth axis of the fibres as observed by TEM. Area 3 corresponds to the massive cluster of nanocrystals observed in Figure 5 at the bottom of the underlying aragonite layer. Interestingly, this area also shows a homogeneous response to the variations of polarization direction, with a f value of 1078. This observation suggests that the c-axes of these nanocrystals also show a preferential orientation. It is
218
K. BENZERARA ET AL.
Fig. 5. TEM imaging of the FIB foil. (a) Negative image (bright field mode) of the whole FIB foil. The Si– Mg-rich lamina appears as a smooth dark grey area on the left hand side of the foil (area delimited by the dashed curves). Three electron diffraction patterns were obtained from different areas in the fibre area with a 100-nm wide aperture showing a common crystallographic orientation. One electron diffraction pattern was measured on the right hand side of the Mg–Si-rich lamina using a 1-mm wide aperture. On this powder pattern, the clustering of diffraction spots shows a common crystallographic orientation of different grains within the probed area. On the left hand side of the Mg–Si-rich lamina aragonite appears as fibres (see magnified image). (b) Arrows indicate the general direction of the fibres. On the right hand side, aragonite appears as clustered small domains [some grains are indicated by arrows in the magnified image (c)].
possible that these nanocrystals form larger-scale fibres by clustering, as supported by SEM images, but this cannot be ascertained based on TEM and the STXM images. STXM offers, however, a view differing from that of SEM. While SEM images show a general fibrous texture within the carbonate laminae with fibres perpendicular on average to the laminae surfaces, STXM suggests that the orientation of the in-plane projection of the c-axis of aragonite varies significantly within a lamina from the bottom to the top.
Discussion The present study provides a basic description at the submicrometre-scale of the mineral texture of aragonite in Satonda stromatolites. Some textural
features have been previously observed using more classical methods and include the heterogeneous microfabrics with either micritic (finegrained) or sparitic (fibres) textures (e.g. Riding 2000) that have been related to various biotic or abiotic mineralization processes. Yet, some of these features are interesting in the light of their formation mechanism. Stromatolites have often been considered as biologically induced formations, that is, resulting from chemical variations (e.g. pH, alkalinity) indirectly induced by the metabolic activity of diverse microorganisms such as cyanobacteria and/or sulphate-reducing bacteria (e.g. Dupraz & Visscher 2005; Altermann et al. 2006). In addition to these metabolic processes, it has been shown that extracellular polymeric substances (EPS) may play a complex but significant role in the
NANOTEXTURES OF ARAGONITE IN STROMATOLITES
precipitation of calcium carbonates (e.g. Braissant et al. 2007). In contrast, coral skeletons are usually considered as biologically controlled deposits, with specific genes controlling the achievement of the coral architecture, although a significant phenotypic plasticity due to environmental conditions is also observed (Todd 2008). The main conclusions of this general view are not contested here. There are obvious structural differences between corals and stromatolites such as the absence of centers of rapid accretion (i.e. centres of calcification) in stromatolites, variations in the orientation of the c-axes of aragonite within a single stromatolite lamina that are not visible in corals, or the absence of Mg– Si-rich laminae in corals. However, in the present study we show that some textural features, which have been attributed specifically to corals in the past and interpreted as the result of a biologically controlled process (e.g. clusters of nanocrystals within the centres of rapid accretion, crystallographic alignment of aragonite crystals, striations interpreted as growth bands), can also be found in stromatolites. Two alternative conclusions can be drawn from this comparison: (1) biologically controlled mineralization processes may also play a role in the formation of Satonda stromatolites, at least to a certain extent; or (2) these textural features for which the precise mode of formation is unknown cannot be used as signatures of bio-controlled processes in corals. Using a reductionist but possibly heuristic approach, we discuss below some of the similarities and differences in the mineralogical textures of corals v. stromatolites at the submicrometre-scale. From a mineralogical point of view, it is interesting to note that corals are assemblages of basic units consisting of aragonite fibres, aragonite nanoglobules (forming centres of calcification in scleractinian corals), and organic polymers (associated with the fibres as well as centers of rapid accretion, e.g. Stolarski 2003; Cuif & Dauphin 2005). No centre of rapid accretion (i.e. micrometre-sized clusters of aragonite nanocrystals with a very precise crystallographic alignment), was observed in Satonda stromatolites. However, Satonda stromatolites show similar fibrous, single-crystal aragonite as well as aragonite nanocrystals. Moreover, several studies have previously shown that aragonite or calcite nanocrystals are often associated with organic polymers in stromatolites (e.g. Kawaguchi & Decho 2002; Ku¨hl et al. 2003; Kazmierczak et al. 2004; Dupraz & Visscher 2005; Benzerara et al. 2006; Kremer et al. 2008). The reason why two types of aragonite morphologies (fibrous v. micritic) are observed in these objects is not clear. Several authors have proposed for Satonda stromatolites as well as for corals that the micritic texture in both coral and stromatolites may be the
219
result of precipitation in organic-enriched microenvironments (centers of rapid accretion in corals, or in living or partially degraded biofilms in stromatolites), while fibres would form in an organic-poor environment (e.g. Kempe & Kaz´mierczak 1993; Arp et al. 2003; Stolarski 2003; Kempe & Kaz´mierczak 2007). This proposal is supported by observations that aragonite nanocrystals within microbialites are surrounded by organic polymers (e.g. Benzerara et al. 2006) and that the experimental precipitation of carbonates in an organic matrix results in the formation of numerous and very small grains (Sethmann et al. 2005; Aloisi et al. 2006). In stromatolites from the Bahamas, three different types of microbial communities exhibiting different physical structures were identified. Micritic laminae composed of small aragonite fibres and not nanospheroids were shown to be specific to one of these communities (Reid et al. 2000; Visscher et al. 2000; Petrisor et al. 2004). It should be noted that in these open marine stromatolites, aragonite fibres can nucleate within EPS-rich biofilms, balancing the idea that fibres would always be associated with organic-deprived environments (Reid et al. 2000; Visscher et al. 2000). Finally, it has been shown that different EPS have various affinities for cations and may produce various polymorphs of calcium carbonate (e.g. Braissant et al. 2003), suggesting an additional source of variety in the texture of calcium carbonates. There are genuine single-crystal aragonite fibres within stromatolites as observed in the present study (e.g. Fig. 3). They share a preferential crystallographic orientation along the growth axis (c-axis), perpendicular to the laminae, suggesting an abiotic growth process that initiates from an underlying surface. In contrast, some fibres may have formed by the clustering of aragonite nanocrystals resulting in what have been called mesocrystals, when nanodomains share a common crystallographic orientation. The existence of such mesocrystals in stromatolites is suggested by the STXM polarizationdependent images in the present study. Their presence in microbialites may have been overlooked in the past and will have to be assessed more thoroughly in the future. Interestingly, the aggregation of oriented vaterite nanocrystals with c-axis normal to the bacterial cell wall has been observed in cultures of Myxococcus xanthus during bacterially induced calcification (Rodriguez-Navarro et al. 2007). This is one of the very few studies showing such a pattern involving prokaryote cells. In contrast, abundant mesocrystals have been observed in several eukaryote biomineralizing systems such as corals (Cuif & Dauphin 2005; Przeniosło et al. 2008; Vielzeuf et al. 2008), including oriented aragonite nanocrystals within centers of
220
K. BENZERARA ET AL.
Fig. 6. STXM mapping of the orientation of the c-axis of aragonite in the FIB foil. (a) XANES spectra at the C K-edge of a single aragonite crystal measured at two perpendicular orientations of the beam polarization, either parallel or perpendicular to the c-axis of aragonite. The dashed line indicates the absorption peak at 290.3 eV attributed to 1s ! p * electronic transitions of carbonate groups. Absorption at this energy is higher when the polarization vector of the beam is parallel to the c-axis of aragonite, and lower when perpendicular. (b– g) STXM images taken at 290.3 eV of the FIB foil with varying directions of the polarization of the incident X-ray beam (0, 20, 40, 60, 70 and 908),
NANOTEXTURES OF ARAGONITE IN STROMATOLITES
rapid accretion of corals (Benzerara et al. in rev.), sponge spicules (Sethmann et al. 2006), and mollusc nacre (Rousseau et al. 2005) as well as in abiotic systems deprived of organics (e.g. Niederberger & Colfen 2006). Although it has been proposed that the formation of such mesocrystals may involve specific and controlled biological processes, the stromatolite and the abiotic occurrences of aragonite mesocrystals underline the fact that as long as the molecular processes leading to crystallographic alignment of these nanodomains are ignored, little can be concluded about the biospecificity of such objects. The evolution of the crystallographic orientation of the crystals within aragonite laminae is an additional issue to discuss. Observations of the Satonda stromatolites by transmitted light microscopy suggest that aragonite fibres are globally oriented perpendicular to the laminae. Interestingly, such a structure with well-oriented aragonite crystals is also observed in nacre for example, in which aragonite tablets grow in alternation with organic matrix sheets. Recently, Coppersmith et al. (2009) proposed a model to explain how a greater orientational order develops over a distance of several aragonite tablets. The model is based on two assumptions: (1) well-oriented tablets grow faster than misoriented ones; and (2) the crystal orientation of a tablet from a given layer is highly probable to be the same as that of the tablet directly below its nucleation site. This latter assumption reflects the presence of pores in the organic matrix that allow some kind of ‘communication’ between the successive aragonite tablet layers. In Satonda stromatolites, the Mg –Si-rich layers separate the aragonite layers. However, the map of the orientation of aragonite c-axis (Fig. 6) and TEM images do not show any ‘transmission’ of the crystal orientation through the Mg –Si-rich layer. On the FIB foil observed in the present study, only a few crystals were found to have their c-axes oriented perpendicular to the laminae, except aragonite fibres at the top of the underlying laminae. Two issues remain unresolved: (1) how aragonite crystals become eventually oriented within an aragonite laminae? And (2) why aragonite crystals in a given cluster do not influence the orientation of
221
aragonite crystals forming on top? More observations of whole aragonite laminae, also involving additional techniques that provide a larger scale view of the crystallographic orientation of aragonite, will be needed in order to better understand how the general orientation of aragonite crystals eventually evolved towards a preferred crystallographic orientation of the fibres within a single lamina. Finally, the striations observed within the aragonite laminae and parallel to them remain unexplained. Texturally, they look similar to the growth bands observed in corals (Fig. 2) and hence there could be a general mechanism responsible for their formation in stromatolites and corals. This mechanism would be important to decipher in order to understand whether it is related to variations in the biological activity and/or whether it is associated with some temporal periodicity of the environment. However, we could not detect in the Satonda stromatolites the significant Mg and Sr variations observed in corals. The meaning of such compositional variations is, however, still discussed in the coral literature (e.g. Meibom et al. 2004), and further studies detailing the ultrastructural variations responsible for their formation are required to better assess the similarities and differences of these features in corals and stromatolites. One possibility that will require a more systematic investigation is that these rings correspond to the 1 –2 mm thick areas evidenced by STXM in which the orientation of the aragonite crystals is roughly the same (Fig. 6). They would correspond to successive episodes of precipitation with variations in the growth direction of aragonite crystals. In conclusion, the present study provides (1) a new description of Satonda stromatolites at the nm-scale, which is, in our opinion, an important basis for future comparisons; and (2) a new methodological framework offering an unprecedented description down to the nm-scale of the mineralogy and texture of carbonate minerals and that might be of interest to the scientific communities studying stromatolites, speleothems, travertines and tufas. Finally, the present study proposes speculative yet heuristic working hypotheses based on a tentative
Fig. 6. (Continued) indicated by arrows. Brighter areas absorb more, darker areas absorb less. For a single pixel, the variation of intensity between the different images is due to linear dichroism. Pixels showing similar variations have a common orientation of the in-plane projection of the c-axis of aragonite. Dashed curves in (b) delimit areas in which grains roughly show a consistent crystallographic orientation of the aragonite c-axis with significant deviations toward the upper right corner. (h) Plot of the variation of absorption with the direction of the polarization vector for different areas chosen in the FIB foil. The curves can be fit as A cos2(u 2 f) þ B. The maximum of absorption gives the angular direction of the c-axis of aragonite (f). Maximum absorption for areas 1, 2, 3, 4 are at 188, 548, 1078 and 1558, respectively. (i) Map of the orientation of the in-plane projection of the c-axis of aragonite in the FIB foil. Areas in black do not contain aragonite and were not fitted. The spatial resolution (i.e. the size of the pixels that were fitted) of this map is 50 nm.
222
K. BENZERARA ET AL.
comparison between stromatolites and corals that question the mechanisms of nucleation and growth of carbonates in these structures and the way these structures are or are not controlled at the mesoand macroscale. It is obvious that our observations and the resulting understanding of the texture of carbonates in stromatolites at the submicrometre-scale are still very limited. Additional studies of this type on other samples should offer new insights about the biological and abiotic processes involved in the formation of such structures with significant implications regarding the search for traces of ancient life on Earth. This work was partly funded by the ANR ‘Jeunes Chercheurs’ grant NanoGeobio #06-JCJC-0158-02 (Anders Meibom and Karim Benzerara). This work was also supported in part by NSF Grant CHE-0431425 (Stanford Environmental Molecular Science Institute).The microbialites were sampled during German-Indonesian Expedition to Satonda Lake 1993 (led by Stephan Kempe, Darmstadt) supported by the Deutsche Forschungsgemeinschaft. JS was funded by the Polish Ministry of Science and Higher Education, project N307-015733. The National NanoSIMS facility at the Muse´um National d’Histoire Naturelle was established by funds from the CNRS, Re´gion ˆIle de France, Ministe`re de´le´gue´ a` l’Enseignement supe´rieur et a` la Recherche, and the Muse´um itself. We thank Omar Boudouma from the Laboratory Magie at University Pierre et Marie Curie (Paris, France) who performed the SEM analyses on the Zeiss Supra 55 SEM microscope. Finally, we thank Emmanuelle Porcher (UMR 7254, MNHN Paris, France) for programming the R routine fitting the STXM polarization-dependent data.
References A LLWOOD , A. C., W ALTER , M. R., K AMBER , B. S., M ARSHALL , C. P. & B URCH , I. W. 2006. Stromatolite reef from the Early Archaean era of Australia. Nature, 441, 714– 718. A LOISI , G., G LOTER , A., K ROGER , M., W ALLMANN , K., G UYOT , F. & Z UDDAS , P. 2006. Nucleation of calcium carbonate on bacterial nanoglobules. Geology, 34, 1017–1020. A LTERMANN , W., K AZ´ MIERCZAK , J., O REN , A. & W RIGHT , D. T. 2006. Cyanobacterial calcification and its rock-building potential during 3.5 billion years of Earth history. Geobiology, 4, 147– 166. A RP , G., R EIMER , A. & R EITNER , J. 2003. Microbialite formation in seawater of increased alkalinity, Satonda crater lake, Indonesia. Journal of Sedimentary Research, 73, 105–127. B ENZERARA , K., M ENGUY , N., G UYOT , F., V ANNI , C. & G ILLET , P. 2005. TEM study of a silicate-carbonatemicrobe interface prepared by focused ion beam milling. Geochimica et Cosmochimica Acta, 69, 1413–1422. B ENZERARA , K., M ENGUY , N., L OPEZ -G ARCIA , P., Y OON , T. H., K AZ´ MIERCZAK , J., T YLISZCZAK , T., G UYOT , F. & B ROWN , G. E. J R . 2006. Nanoscale
detection of organic signatures in carbonate microbialites. Proceedings of the National Academy of Sciences of the United States of America, 103, 9440– 9445. B ENZERARA , K., M ORIN , G., Y OON , T. H., M IOT , J., T YLISZCZAK , T., C ASIOT , C., B RUNEEL , O., F ARGES , F. & B ROWN , G. E. J R . 2008. Nanoscale study of As biomineralization in an acid mine drainage system. Geochimica et Cosmochimica Acta, 72, 3949– 3963. B ENZERARA , K., M ENGUY , N., O BST , M., S TOLARSKI , J., M AZUR , M., T YLISCZAK , T., B ROWN , G. E. J R . & M EIBOM , A. Nanoscale skeletal architecture of a scleractinian coral. Submitted to Proceedings of the Royal Society B. B ERNARD , S., B ENZERARA , K., B EYSSAC , O., M ENGUY , N., G UYOT , F., B ROWN , G. E. J R . & G OFFE , B. 2007. Exceptional preservation of fossil plant spores in highpressure metamorphic rocks. Earth and Planetary Science Letters, 262, 257– 272. B LUHM , H. K., A NDERSSON , T. ET AL . 2006. Soft X-ray microscopy and spectroscopy at the molecular environmental science beamline at the Advanced Light Source. Journal of Electron Spectroscopy and Related Phenomena, 150, 86–104. B RAISSANT , O., C AILLEAU , G., D UPRAZ , C. & V ERRECCHIA , E. P. 2003. Bacterially induced mineralization of calcium carbonate in terrestrial environments: the role of exopolysaccharides and amino acids. Journal of Sedimentary Research, 73, 485– 490. B RAISSANT , O., D ECHO , A. W., D UPRAZ , C., G LUNK , C., P RZEKOP , K. M. & V ISSCHER , P. T. 2007. Exopolymeric substances of sulfate-reducing bacteria: Interactions with calcium at alkaline pH and implication for formation of carbonate minerals. Geobiology, 5, 401–411. C OPPERSMITH , S. N., G ILBERT , P. U. P. A. & M ETZLER , R. A. 2009. Theoretical characterization of a model of aragonite crystal orientation in red abalone nacre. Journal of Physics A – Mathematical and Theoretical, 42, Article Number: 125101. C UIF , J. P. & D AUPHIN , Y. 2005. The two-step mode of growth in the Scleractinian coral skeletons from the micrometre to the overall scale. Journal of Structural Biology, 150, 319–331. D UPRAZ , C. & V ISSCHER , P. T. 2005. Microbial lithification in marine stromatolites and hypersaline mats. Trends in Microbiology, 13, 429– 438. D UPRAZ , C., V ISSCHER , P. T., B AUMGARTNER , L. K. & R EID , R. P. 2004. Microbe– mineral interactions: early carbonate precipitation in a hypersaline lake (Eleuthera Island, Bahamas). Sedimentology, 51, 745 –765. D UPRAZ , C., P ATTISINA , R. & V ERRECCHIA , E. P. 2006. Translation of energy into morphology: simulation of stromatolite morphospace using a stochastic model. Sedimentary Geology, 185, 185– 203. G IBBS , W. W. & P OWELL , C. S. 1996. Bugs in the data? The controversy over Martian life is just beginning. Scientific American, 275, 20. G ROTZINGER , J. P. & R OTHMAN , D. H. 1996. An abiotic model for stromatolite morphogenesis. Nature, 383, 423–425. H ABERSTROH , P. R., B RANDES , J. A., G ELINAS , Y., D ICKENS , A. F., W IRICK , S. & C ODY , G. 2006.
NANOTEXTURES OF ARAGONITE IN STROMATOLITES Chemical composition of the graphitic black carbon fraction in riverine and marine sediments at submicron scales using carbon X-ray spectromicroscopy. Geochimica et Cosmochimica Acta, 70, 1483– 1494. H EANEY , P. J., V ICENZI , E. P., G IANNUZZI , L. A. & L IVI , K. L. T. 2001. Focused ion beam milling: a method of site-specific sample extraction for microanalysis of Earth and planetary materials. American Mineralogist, 86, 1094– 1099. H ITCHCOCK , A. P. 2001. Soft X-ray spectromicroscopy of polymers and biopolymer interfaces. Journal of Synchrotron Radiation, 8, 66– 71. H OFMANN , H. J. 2000. Archean stromatolites as microbial archives. In: R IDING , R. & A WRAMIK , S. M. (eds) Microbial Sediments. Springer-Verlag, Berlin, Heidelberg, 315– 327. K AWAGUCHI , T. & D ECHO , A. W. 2002. In situ microspatial imaging using two-photon and confocal laser scanning microscopy of bacteria and extracellular polymeric secretions (EPS) within marine stromatolites. Marine Biotechnology, 4, 127– 131. K AZ´ MIERCZAK , J. & K EMPE , S. 2004. Microbialite formation in seawater of increased alkalinity, Satonda Crater Lake, Indonesia – Discussion. Journal of Sedimentary Research, 74, 314 –317. K AZ´ MIERCZAK , J., K EMPE , S. & A LTERMANN , W. 2004. Microbial origin of Precambrian carbonates: lessons from modern analogues. In: E RIKSSON , P. G., A LTERMANN , W., N ELSON , D. R., M UELLER , W. U. & C ATUNEANU , O. (eds) The Precambrian Earth: Tempos and Events. Elsevier, Amsterdam, 545–564. K EMPE , S. & K AZ´ MIERCZAK , J. 1993. Satonda Crater Lake, Indonesia: Hydrogeochemistry and Biocarbonates. Facies, 28, 1– 31. K EMPE , S. & K AZ´ MIERCZAK , J. 2007. Hydrochemical key to the genesis of calcareous non-laminated and laminated cyanobacterial microbialites. In: S ECKBACH , J. (ed.) Algae and Cyanobacteria in Extreme Environments. Springer, Dordrecht, 239– 264. K REMER , B., K AZ´ MIERCZAK , J. & S TAL , L. J. 2008. Calcium carbonate precipitation in cyanobacterial mats from sandy tidal flats of the North Sea. Geobiology, 6, 46–56. K U¨ HL , M., F ENCHEL , T. & K AZMIERCZAK , J. 2003. Structure, function, and calcification potential of an artificial cyanobacterial mat. In: K RUMBEIN , W. E., P ATERSON , D. & Z AVARZIN , G. (eds) Fossil and Recent Biofilms – A Natural History of Life on Earth. Kluver Academic Publishers, Dordrecht, 77– 102. L EPOT , K., B ENZERARA , K., B ROWN , G. E. J R . & P HILIPPOT , P. 2008. Microbially influenced formation of 2,724-million-year-old stromatolites. Nature Geoscience, 1, 118–121. L OWE , D. R. 1994. Abiological origin of described stromatolites older than 3.2 Ga. Geology, 22, 387–390. M C L OUGHLIN , N., W ILSON , L. A. & B RASIER , M. D. 2008. Growth of synthetic stromatolites and wrinkle structures in the absence of microbes – implications for the early fossil record. Geobiology, 6, 95–105. M EIBOM , A., C UIF , J. P., H ILLION , F. O., C ONSTANTZ , B. R., J UILLET -L ECLERC , A., D AUPHIN , Y., W ATANABE , T. & D UNBAR , R. B. 2004. Distribution
223
of magnesium in coral skeleton. Geophysical Research Letters, 31. M EIBOM , A., M OSTEFAOUI , S., C UIF , J. P., D AUPHIN , Y., H OULBREQUE , F., D UNBAR , R. B. & C ONSTANTZ , B. R. 2007. Biological forcing controls the chemistry of reef-building coral skeleton. Geophysical Research Letters, 34, #L02601. M EIBOM , A., C UIF , J. P., H OULBREQUE , F., O STEFAOUI , S., D AUPHIN , Y., M EIBOM , K. L. & D UNBAR , R. B. 2008. Compositional variations at ultra-structure length scales in coral skeleton. Geochimica et Cosmochimica Acta, 72, 1555– 1569. M ETZLER , R. A. D., Z HOU , M. ET AL . 2008. Polarizationdependent imaging contrast in abalone shells. Physical Review B, 77, #064110. N IEDERBERGER , M. & C OLFEN , H. 2006. Oriented attachment and mesocrystals: Non-classical crystallization mechanisms based on nanoparticle assembly. Physical Chemistry Chemical Physics, 8, 3271– 3287. P ERRIN , C. & S MITH , D. C. 2007. Earliest steps of diagenesis in living scleractinian corals: Evidence from ultrastructural pattern and Raman spectroscopy. Journal of Sedimentary Research, 77, 495–507. P ETRISOR , A. I., K AWAGUCHI , T. & D ECHO , A. W. 2004. Quantifying CaCO3 microprecipitates within developing surface mats of marine stromatolites using GIS and digital image analysis. Geomicrobiology Journal, 21, 491– 496. P RZENIOSŁO , R., S TOLARSKI , J., M AZUR , M. & B RUNELLI , M. 2008. Hierarchically structured scleractinian coral biocrystals. Journal of Structural Biology, 161, 74–82. R EID , R. P. P. T., V ISSCHER , A. W. ET AL . 2000. The role of microbes in accretion, lamination and early lithification of modern marine stromatolites. Nature, 406, 989– 992. R IDING , R. 2000. Microbial carbonates: the geological record of calcified bacterial– algal mats and biofilms. Sedimentology, 47, 179–214. R ODRIGUEZ -N AVARRO , C., J IMENEZ -L OPEZ , C., R ODRIGUEZ -N AVARRO , A., G ONZALEZ -M UNOZ , M. T. & R ODRIGUEZ -G ALLEGO , M. 2007. Bacterially mediated mineralization of vaterite. Geochimica et Cosmochimica Acta, 71, 1197–1213. R OUSSEAU , M., L OPEZ , E., S TEMPFLE , P., B RENDLE , M., F RANKE , L., G UETTE , A., N ASLAIN , R. & B OURRAT , X. 2005. Multiscale structure of sheet nacre. Biomaterials, 26, 6254–6262. S CHOPF , J. W., K UDRYAVTSEV , A. B., C ZAJA , A. D. & T RIPATHI , A. B. 2007. Evidence of archean life: stromatolites and microfossils. Precambrian Research, 158, 141–155. S EMIKHATOV , M. A., G EBELEIN , C. D., C LOUD , P., A WRAMIK , S. M. & B ENMORE , W. C. 1979. Stromatolite morphogenesis – progress and problems. Canadian Journal of Earth Sciences, 16, 992–1015. S ETHMANN , I., P UTNIS , A., G RASSMANN , O. & L OBMANN , P. 2005. Observation of nano-clustered calcite growth via a transient phase mediated by organic polyanions: a close match for biomineralization. American Mineralogist, 90, 1213–1217. S ETHMANN , I., H INRICHS , R., W ORHEIDE , G. & P UTNIS , A. 2006. Nano-cluster composite structure of calcitic sponge spicules – a case study of basic characteristics
224
K. BENZERARA ET AL.
of biominerals. Journal of Inorganic Biochemistry, 100, 88–96. S HAPIRO , R. S. 2000. A comment on the systematic confusion of thrombolites. Palaios, 15, 166–169. S TEELE , A., G ODDARD , D., B EECH , I. B., T APPER , R. C., S TAPLETON , D. & S MITH , J. R. 1998. Atomic force microscopy imaging of fragments from the Martian meteorite ALH84001. Journal of Microscopy, Oxford, 189, 2 –7. S TOLARSKI , J. 2003. Three-dimensional micro- and nanostructural characteristics of the scleractinian coral skeleton: A biocalcification proxy. Acta Palaeontologica Polonica, 48, 497–530. T ODD , P. A. 2008. Morphological plasticity in scleractinian corals. Biological Reviews, 83, 315– 337.
V IELZEUF , D., G ARRABOU , J., B ARONNET , A., G RAUBY , O. & M ARSCHAL , C. 2008. Nano to macroscale biomineral architecture of red coral (Corallium rubrum). American Mineralogist, 93, 1799–1815. V ISSCHER , P. T., R EID , R. P. & B EBOUT , B. M. 2000. Microscale observations of sulfate reduction: Correlation of microbial activity with lithified micritic laminae in modern marine stromatolites. Geology, 28, 919–922. Z HOU , D., M ETZLER , R. A., T YLISZCZAK , T., G UO , J. H., A BRECHT , M., C OPPERSMITH , S. N. & G ILBERT , P. 2008. Assignment of Polarization-Dependent Peaks in Carbon K-Edge Spectra from Biogenic and Geologic Aragonite. Journal of Physical Chemistry B, 112, 13 128 –13 135.
Calcitic nanofibres in soils and caves: a putative fungal contribution to carbonatogenesis SASKIA BINDSCHEDLER1, L. MILLIE`RE1, G. CAILLEAU1, D. JOB2 & E. P. VERRECCHIA1* 1
Institut de Ge´ologie et de Pale´ontologie, Universite´ de Lausanne, Anthropole, 1015 Lausanne, Switzerland 2
Institut de Biologie, Universite´ de Neuchaˆtel, Rue Emile Argand 11, 2007 Neuchaˆtel, Switzerland *Corresponding author (e-mail:
[email protected])
Abstract: The origin of soil mineralized nanofibres remains controversial. It is attributed to either biogenic factors or physicochemical processes. Scanning electron microscope and transmission electron microscope observations show that nanofibres could originate from the breakdown of fungal hyphae, especially its cell wall. It is hypothesized that during the decay of organic matter, cell wall microfibrils are released in the soil where they are exposed to mineralizing pore fluids, leading to their calcitic pseudomorphosis and/or are used as a template for calcite precipitation. When associated with needle fibre calcite bundles, nanofibres could indicate the relict of an organic sheath in which calcite has precipitated. This paper emphasizes the important roles of both organic matter and fungi in carbonatogenesis, and consequently in the soil carbon cycle.
Natural nanofibres have been observed in various environments such as subtropical and temperate soils (Verrecchia & Verrecchia 1994; Cailleau et al. 2005) and cave deposits such as moonmilk (Borsato et al. 2000; Can˜averas et al. 2006). They are often associated with Needle Fibre Calcite (NFC; Verrecchia & Verrecchia 1994; Borsato et al. 2000). The aim of this study is to provide new insight into the processes at the origin of nanofibres. In order to differentiate between organic and mineral nanofibres, the term ‘organic nanofibre’ will be used for nanofibres whose organic nature has been determined by analytical methods. The term ‘mineral nanofibre’ will be used for: (i) for nanofibres observed in scanning electron microscopy (SEM), in absence of specific labelling (see ‘Materials and methods’ section); and (ii) for nanofibres diffracting the electron beam under transmission electron microscopy (TEM). The term nanofibre alone refers only to a shape or an object and therefore used for morphological descriptions.
Previous work on mineral nanofibres Since 1980, many authors have reported mineral nanofibres from various environments (Table 1). The four following authors have specifically observed organized structures related to mineral nanofibres: (i) filamentous, ramified, microscopic
structure composed by a dense nanofibre scaffolding (Borsato et al. 2000; Richter et al. 2008); (ii) a straight macro-structural alignment (3– 5 mm wide and .70 mm long) of unordered nanofibres observed close to an organic filament (possibly actinomycetes, cyanobacteria, or fungi; Benzerara et al. 2003); and (iii) 3 mm wide and .50 mm long filaments interpreted as ‘calcified filaments with needles in grain-coating needle mat’ have also been observed by Jones & Ng (1988).
Filamentous organisms and structures in soils and caves Filamentous organisms living within the soil or in caves must be heterotrophic organisms. Algae and cyanobacteria are photosynthetic organisms and thus are present only at the soil surface or in rock fractures near a light source. Indeed, in mineral substrates that are far away from any light, these organisms are absent due to the lack of their energy source. Accordingly, filamentous fabrics present in these environments could be fungi, filamentous bacteria (in soils and caves mostly actinomycetes), and roots (Paul & Clark 1996; Gobat et al. 2003). Taking into account their sizes and morphologies summarized in Table 2, fungi are the most suitable organisms associated with nanofibres and NFC.
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 225–238. DOI: 10.1144/SP336.11 0305-8719/10/$15.00 # The Geological Society of London 2010.
226
S. BINDSCHEDLER ET AL.
Table 1. Review of nanofibres in the literature Authors Pouget & Rambaud (1980) Verge`s et al. (1982) Ducloux et al. (1984) Phillips & Self (1987) Phillips et al. (1987) Jones & Ng (1988)
Nomenclature Calcite ‘en baˆtonnets’ Small needle-shaped crystals Calcite ‘en baˆtonnets’ Micro-rods Submicron size rods Needles
Geological setting Soil with calcareous crust Calcareous soils Developed on screeslope Pedogenic calcrete Pedogenic calcrete Rhizolith from the Pleistocene Ironshore Formation
Verrecchia & Verrecchia (1994) Loisy et al. (1999)
Micro-rods
Quaternary calcretes, Israel
Micro-rods
Borsato et al. (2000)
Nanofibres
Carbonate paleosol in scree deposits Moonmilk (cave deposits)
Benzerara et al. (2003)
Nanobacteria-like rods
At the surface of the Tataouine meteorite
Cailleau et al. (2005)
Micro-rods
Orthox soils
Jeong & Chun (2006)
Nanofibre calcite
Richter et al. (2008)
Nanofibres
Aerosols coming from loess plateau and desert Moonmilk (cave deposits)
Interpretation/context Mesh of monocrystaline calcite crystals Tangled crystals Covering larger needle-fibre calcite crystals Interpreted as calcified rod-shaped bacteria Interpreted as calcified rod-shaped bacteria Calcified filaments coated withneedles (i.e. nanofibres) Disordered mesh Mineralized threadlike and bacilliform bacteria Probably abiogenic precipitation Straight micro-alignment of nanofibres; possible organic origin Observed on burnt oxalate crystals embedded in tree tissues – –
Note: Review of nanofibres present in soils and caves: nomenclature, occurrence and interpretation in the literature.
Fungal presence and activity in soil and caves Fungi are present in large amounts in soils. As an example, one metre square of fertile soil can contain a 10 000-km long fungal network (Gobat et al. 2003). About 80% of land plant species are colonized by arbuscular mycorrhizal fungi (endomycorrhiza), and around 3% of phanerogam species are colonized by ectomycorrhizal fungi (EcM), especially plants with a large distribution at a global scale (Pinaceae, Fagaceae). In soils, a vertical distribution can be distinguished regarding fungal type in terms of their ecology. Organic layers are mostly colonized by saprophytic fungi, whereas mineral layers are colonized by EcM fungi (van Scho¨ll et al. 2008). The latter has been demonstrated as being a significant agent of mineral weathering of ecosystem-wide importance (van Scho¨ll et al. 2008). The mycelial network is able to efficiently translocate nutrients in solution from one place to another (Gobat et al. 2003). Basidiomycetes, and among them EcM fungi, are able to build structures
named fungal strands that can extend meters away from the roots (Finlay & So¨derstro¨m 1992; van Scho¨ll et al. 2008). Thus, the presence of mycorrhized roots, fungal hyphae and strands in deep mineral layers or in caves is not surprising. Canadell et al. (1996) showed an average rooting depth of 4.6 m þ/20.5 m, with a maximum depth of 7.0 m þ/21.2 m for trees. In their review, only the root itself is taken into account. Considering the mycorrhiza, it can considerably extend the root network (Timonen & Marschner 2006). Observations of roots at depths below 2–3 m in caves have also been observed (Canadell et al. 1996). Jasinska et al. (1996) demonstrated that root mats could be the sole source of food for faunal communities in an Australian cave.
Cave geomicrobiology Caves are nutrient-limited environments due to the absence of light that prevents primary production through photosynthesis, contrary to other common environments on Earth. Thus, in terms of presence of life, this kind of environment can be considered
FUNGAL ORIGIN FOR CALCITIC NANOFIBRES
227
Table 2. Characteristics of filamentous organisms in soils and caves Diameters Fungi
3–6 mm on average 1 mm (min)– 30 mm (max) Strands: 0.02 2. 1 mm
Actinomycetes 0.5 –1 mm in average, max 2 mm in some genus
Fine roots
Morphologies/ structures
Cell wall morphology
Cell wall biochemistry
Hyphae with or without septum, more or less ramified. Bundles of differentiated hyphae forming linear organs, fungal strands
Thick-walled (up to 1 mm) and thin-walled (100 – 200 nm) hyphae
Two layers: a fibrillar component with chitin and b-glucans and an amorphous component with glycoproteins specific to taxonomic groups
Ramified mycelium, sometimes fragmented
Wall 20 – 80 nm thick
Gram positives bacteria with one homogenous layer of peptidoglycan (murein). Four types of peptidoglycans depending on genus
Primary wall: network Highly variable in Ramified structure ,2–0.2 mm of fibrous cellulose thickness, from Single conducting composed, at a and hemicellulose 0.1 mm in young vessel 10–30 mm micromorphological embedded in a matrix cells to 100 mm in level, of complex of pectin. mature cells arrangements of Secondary wall: only in linear vessels mature cells, can contain lignin
Note: Review of the different filamentous organisms in soils synthesising characteristics such as filament diameter and morphology and their cell wall morphology and biochemistry. References from Jones 1970; Dix & Webster 1995; Paul & Clark 1996; Bouma et al. 2001; Carlile et al. 2001; Pregitzer et al. 2002; Prescott et al. 2003; Coleman et al. 2004; Pessoni et al. 2005; Hishi 2007.
as ‘extreme’. On the other hand, physicochemical parameters tend to be buffered and constant throughout the year (e.g. mild temperature normally equals MAST (Mean Annual Surface Temperature) and fairly high humidity). These extreme but constant conditions allow the presence of underground ecosystems, which may or may not be connected to aboveground energy-sources (Jasinska et al. 1996; Sarbu et al. 1996). Prokaryotes and fungi are the most common organisms that can be encountered in caves, and their link in speleothem formation is often proposed and debated. Moonmilk is a common speleothem mineral, and its biological or physico-chemical origin has long been discussed. Today most of the theories involve microbial mediation in its formation, but the exact role that microorganisms play, whether it is bacterial or fungal, is still being discussed (Gradzinski et al. 1997; Northup & Lavoie 2001; Can˜averas et al. 2006; Barton & Northup 2007). In caves, moonmilk is more likely to be present in the vicinity of soils (Gradzinski, pers. comm.). This increases the probability of cave access for roots and their fungal associates, and consequently their involvement in the genesis of moonmilk.
Fungal strands and hyphae ultrastructure Fungal hyphae size Single fungal hyphae diameter is highly variable, depending on the taxonomic position, environmental conditions, age, and function of hyphae. Nevertheless, for functional hyphae, an average diameter of 3– 6 mm is found in the literature (Dix & Webster 1995; Carlile et al. 2001). Nonfunctional hyphae, such as those from the cortex of fungal strands, can have a diameter of 1 mm, with an inner diameter of ,0.5 mm.
Fungal strands Basically, a fungal strand is a bundle of juxtaposed linear hyphae. They are organs produced by fungi to explore their environment and to translocate nutrients from one place to another. They have the ability to extend over long distances, that is up to 30 m. In addition, macromorphologically, they exhibit a variable diameter, depending mostly on their age and remoteness from nutrient sources. This diameter ranges from a few mm up to 4 mm
228
S. BINDSCHEDLER ET AL.
in some wood decaying species (Thompson 1984). The structure of the fungal strand is composed of: (i) an outer layer (the cortex) composed of a thick layer of narrow thick-walled multiseptate dead hyphae (average diameter of 1 mm); and (ii) an inner layer (the medulla) composed of a few linear wide thin-walled sparsely septate living hyphae (average diameter of 6–10 mm). The latter seems to be less resistant to hydrostatic pressure (Watkinson 1979) and thus is more rapidly exposed to decay processes than hyphae with a thick melanized wall in the outer layer. The inner part often collapses, leaving fungal strands composed of a thick layer of narrow hyphae with empty wide channels in the middle (Watkinson 1984). The cortex of the fungal strands makes it an impermeable organ where fluids can be bidirectionally translocated (Watkinson 1979; Dix & Webster 1995).
Composition and structure of the fungal cell wall The thickness of the cell wall also shows great variability depending on physiological processes and the function of hyphae. Single hyphae have an average cell wall thickness of 150 nm (Jones 1970; Beckett
et al. 1974; Farkas 1979; Ruiz-Herrera et al. 1996; Pessoni et al. 2005). On the other hand, hyphae from the cortex of a fungal strand can exhibit very thick cell walls, up to 500 nm. The wall thickening can even occlude the hypha lumen (Watkinson 1984). Moreover, the walls of these hyphae often present a high degree of melanization, which is also a factor in impermeabilization (Paul & Clark 1996). The fungal cell wall can be described by two main types of materials, an outer layer composed of amorphous material (mainly mannoproteins), and an inner layer composed of fibrous material, chitin and b-glucans (Burnett 1979; Ruiz-Herrera 1992; Bowman & Free 2006). Chitin is a polymer of a polysaccharide, N-acetyl-glucosamine. It is present in the form of long microfibrils, sometimes over 1 mm, with a diameter of 10– 25 nm. It is located in the innermost part of the wall, arranged as an intertwined mesh embedded in an amorphous matrix (Aronson & Preston 1960; Carlile et al. 2001). b-glucans are homopolysaccharides of glucose. In the fungal wall, it is present either as b (1-3) glucan or in a lesser amount as b (1-6) glucan. They are found in greater amounts than chitin (Carlile et al. 2001; Farkas 2003). Figure 1 shows a sketch of the fungal cell wall.
Fig. 1. Sketch of the fungal cell wall (modified from Latge´ 2003). Note the fibrous layer composed by chitin and b-glucan fibres and the amorphous layer. In order to give an orientation to this sketch, the plasma membrane of the fungal cell has been represented by the phospholipids bilayer.
FUNGAL ORIGIN FOR CALCITIC NANOFIBRES
Cell wall material can be a significant part of the resistant organic matter in soils. Depending on physico-chemical parameters, enzymatic degradation of polymerous substances from the cell wall may or may not be possible (Paul & Clark 1996; Coleman et al. 2004).
Materials and methods Secondary carbonate accumulations have been sampled at four sites: (i) in the mineral layers (calcic horizon of a calcisol; IUSS Working Group WRB 2006) at a quarry near Villiers (Swiss Jura Mountains, 478040 N, 68590 E), as well as (ii) at an outcrop near Aı¨nsa (Spain, 428210 N; 08040 E); (iii) in the narrow entrance of a cave near les Cornettes de Bises (Swiss Alps, 468190 N, 168480 E); and (iv) in the vers chez le Brandt cave, inside a wide chamber at 100 metres from the entrance (Swiss Jura Mountains, 468560 N, 68280 E). At the first two sites (soils developed on scree slopes), samples exhibit two different morphologies directly visible in the outcroup: (i) cotton-ball-like NFC that accumulates in the soil pores; and (ii) coatings on grains and centimetric to decimetric cryoclasts. When wet, these coatings constitute a plastic paste, which becomes pulverulent when dry. At the third and fourth sites (caves), only moonmilk deposits were sampled in the form of a wet plastic wall coating, up to 30 cm thick. In addition to crystals, fungal strands associated with different macroscopic morphologies of NFC have been sampled for electron microscope observations. Strands have been taken from cotton-ball-like NFC, associated with rock fragment coatings, or free in the soil pores. All samples were collected using polypropylene tweezers and stored in sterilized 50 mL tubes at low temperature. Bulk samples were analysed by X-ray diffraction (XRD) using a Scintag diffractometer in order to determine their mineralogical nature. In each sample, quartz powder was added in order to normalize and compare samples with each other. In order to detect organic from mineral material, prior to observations, each macromorphological sample was stained using a 4% osmium aqueous solution from a modified Pearson et al. (2004) protocol. However, using this labelling method, the presence of the organic matter can be determined but not its nature. Samples were gold-coated (10 nm) and observed using a Phillips ESEM-FEG XL30 Field Emission Gun Scanning Electron Microscope (FEG-SEM). Osmium staining was detected with an EDAX Energy Dispersive Spectrometer (EDS) coupled to the electron microscope. With natural non-flat samples and absence of a standard, EDS spectra only give qualitative information.
229
In order to check possible artifacts due to highvacuum, some representative samples were observed using XL30 SEM in its LTSEM cryo mode (Low-temperature SEM). Some fungal strands and coatings were embedded in an epoxy resin, and ultrathin sections were performed using a Reichert Ultracut S (Leica) microtome with a diamond knife. Ultrathin sections (200 –220 nm thick) were observed using a Phillips CM-200 Transmission Electron Microscope (TEM) with a voltage of 200 kV. Crystal properties were determined using microdiffraction.
Results XRD analysis of the three types of samples (cotton-ball-like structures, coatings on grains, and moonmilk) shows that their mineralogy is calcitic in nature (Fig. 2a–c). Moreover, the absence of shift (expected in presence of Mg in the crystal lattice) after normalization with quartz powder characterizes the presence of low magnesium calcite (LMC) at these sites. SEM observations of soil samples show recursive associations between NFC, unidentified nanofibres, and fungal strands. NFC is characterized by a width between 0.5–2.0 mm and a length ,100 mm. Some epitactic growths are present but no important development is observed (no big euhedral crystals due to epitactic cementation). The NFC is present either as random meshes, or as bundles, 3–30 mm in diameter (average 10 mm; Fig. 3a). This microscopic feature constitutes the macroscopic cotton-ball-like structures. Bundles contain some nanofibres, occasionally associated with amorphous matter assumed to be an organic veil (Fig. 3b). These nanofibres are rarely observed on NFC when the latter are randomly oriented and/or strongly modified by epitactic growth. At a microscopic scale, coatings from soil grains exhibit various amounts of NFC and nanofibres, in which needles are less represented than in the cotton-ball-like morphology. NFC often shows random orientations, whereas nanofibres are packed in clusters. This morphology shows great similarity with microstructures of moonmilk samples, in which NFC is even less represented. SEM measurements show that nanofibres are up to 6 mm long (the shortest is 0.2 mm long) with an average width of 78.6 nm (standard deviation of 22.5 nm, based on 106 measurements; Fig. 3c). They are characterized by a high flexibility (as mentioned by Borsato et al. 2000), leading to spectacular curvature (Fig. 3c). They appear smooth under TEM. Two kinds of structures are observed: (i) a randomly-oriented framework of nanofibres, in which widespread putative organic veils and calcitic micro-aggregates are present; these meshes are
230
S. BINDSCHEDLER ET AL.
Fig. 2. X-ray diffractogramme showing the low-magnesium calcite nature of (a) cotton-ball like NFC from the Swiss Jura Mountains, (b) coatings on blocks from Swiss Jura Mountains and (c) moonmilk from the Vers chez le Brandt cave in the Swiss Jura Mountains samples. Dotted lines correspond to the main peaks of low-magnesium calcite (CaCO3), the other peaks mainly correspond to quartz (quartz added prior to analysis to allow peak correction).
FUNGAL ORIGIN FOR CALCITIC NANOFIBRES
231
Fig. 3. (a –d) SEM photomicrographs. (a) Bundle of NFC present in a sample from the Spanish site of cotton-ball-like NFC associated with a fungal strand (observed macroscopically). (b) Bundle of NFC covered by putative organic veils. Some organic nanofibres are also present. Sample from a grain coating associated with a fungal strand and some cotton-ball like NFC. (c) Close-up of an organized mesh composed by nanofibres (moonmilk, Swiss Alps). Preferential orientations of nanofibres are shown by the crossed double-headed arrows. Some nanofibres are curved (arrows). This characteristic indicates a contact-deformation. (d) Mesh composed by randomly-oriented nanofibres associated with NFC (white arrows) and putative organic veils (black arrows). Swiss Jura Mountains, coatings on block.
often associated with other components (i.e. NFC, fungal strands, hyphae, etc.) as observed in moonmilk deposits (Fig. 3d); and (ii) organized structures of nanofibres (Fig. 4), either as small pieces that have apparently undergone a breakdown, or as a tubular/circular microscopic network (Fig. 4a, c). The term organized refers to a non-random distribution of nanofibres, whatever their nature. These networks are composed by intertwined nanofibres oriented in two main directions (Fig. 4b, d). Another main component is frequently observed associated with soil samples: macroscopic, brown organic filaments, identified as fungal strands. Their average diameter can reach 100 mm and they are composed of the two typical mycelial strand structures, an external part made of several narrow fungal hyphae with a thick cell wall and an inner part characterized by a few wide thinwalled hyphae, which often lack in our observations due to their ability to be rapidly decayed
(Fig. 5a, b). Nanofibres are abundant all along the macroscopic filament where fungal strands seem to break down. A cross-section of a fungal hypha shows that the fungal wall is composed of two layers, an inner part composed of fibrous material and an external part composed of an amorphous material (Fig. 5c, d). From these observations, it is obvious that there is an intimate relationship between the hyphae and the nanofibres (Fig. 5c, d). Optical observations, hydrochloric acid tests on moonmilk, as well as TEM microdiffractions (Borsato et al. 2000) indicate that the nanofibres are mineral in nature. In order to test this hypothesis, in-situ analyses were performed to distinguish organic from mineral matter using osmium labelling with EDS control on samples. The osmium stains only organic matter and not mineral material (Pearson et al. 2004). Osmium peaks indicate that non-organized frameworks composed of only nanofibres do not
232
S. BINDSCHEDLER ET AL.
Fig. 4. Samples from the Swiss Jura Mountains site. (a) SEM photograph showing a tubular-like structure resembling an organic filament (e.g. a fungal hypha) and composed by nanofibres (cotton-ball-like NFC associated with decaying organic matter). (b–d) Sample from coatings on a cryoclast associated with decaying fungal strands. (b) SEM photograph showing a network of intertwined nanofibres oriented in two main directions (white rectangle). White star indicates a NFC crystal. (c) TEM photomicrograph showing a section of a circular structure interpreted as a possible organic filament (e.g. a fungal hypha) and composed by nanofibres. (d) TEM photomicrograph showing a network of intertwined nanofibres oriented in two main directions. Also note the presence of a circular section similar to the one shown in Figure 2c (stars).
contain organic material (Fig. 6a, b) and organized meshes do contain organic matter (Fig. 6c, d).
Discussion The presence of nanofibres in various vadose environments has been widely observed. Their origin is either attributed to a biogenic factor (‘probably rod-shaped calcified bacteria’ Phillips & Self 1987; Ould Mohamed & Bruand 1994; ‘microrod attributed to bacteria or nuclei in gel’ Verrecchia & Verrecchia 1994; Loisy et al. 1999) or physicochemical processes [‘precipitation from pore filling fluids’, sometimes assoicated with
organic filaments, that is, nanofibres ‘cannot be fully attributed to direct organic activity’ (Jones & Ng 1988); ‘nanofibres show microstructures that are typical of inorganic, crystalline material’ (Borsato et al. 2000)]. Therefore, the origin of terrestrial accumulations of nanofibres remains controversial. The nanofibres discussed in this paper are similar to those described in Phillips & Self (1987), based on their size and flexibility. It is important to note that they identified the organic filaments associated with nanofibres as fungi, and the same conclusion is drawn in this study, based on their size and morphology, as well as presence of macrostructures
FUNGAL ORIGIN FOR CALCITIC NANOFIBRES
233
Fig. 5. SEM photomicrographs. (a, b) Samples from Aı¨nsa (Spain). (a) Fungal strand associated with cotton-ball-like NFC (star). Note the absence of the inner hyphae (arrow), which seem to have already undergone breakdown. (b) Close-up of another fungal strand showing the narrow and thick-walled hyphae from the outer part (black arrow). Note that the internal wide thin-walled hyphae are absent (white star). (c, d) Swiss Jura Mountains samples. (c) Close-up of a decaying hypha from a grain coating sample showing the release of cell wall fibrous material (e.g. chitin, b-glucan) from its cell wall (arrow). (d) Remains of a putative fungal hyphae on needle fibre calcite. Two textures can be distinguished, an outer one that appears to be smooth (black arrow) and an inner one that appears to be composed of nanofibres (white arrow). Note also the presence of a mesh of nanofibres along the other calcite needles in this picture (star). Sample from cotton-ball like NFC associated with decaying organic matter.
such as mycelial strands (Fig. 5a, b). Two layers are visible during the breakdown of the hyphal cell wall (Fig. 5d). It is known that the inner wall layer is composed of a hard microfibril framework (theoretically 10–25 nm in diameter, Carlile et al. 2001) made of chitin and b (1-3), b (1-6) glucan. Based on this fact and the recognition of the organic nature of some nanofibres (Fig. 6c, d), the organized meshes of nanofibres are considered as the result of a slightly destructive decay of the fungal fibrous cell wall material. At this stage, the organized meshes are interpreted as the first step in the breakdown of the fungal hyphae cell wall, whereas the non-organized mats represent the ultimate state of decay and reworking of this organic matter. In other words, the nanofibres could be interpreted as organic in origin and being the result of an incomplete decaying of fungal matter. During early
diagenesis, their calcitic pseudomorphosis (Cailleau 2005; Cailleau et al. 2005) and/or their role as a template for calcitic precipitation results from the release of the nanofibres in the soil environment, followed by their exposure to mineralizing pore filling fluids. This could explain why non-organized meshes (interpreted as the oldest decay product) are often composed of nanofibres of calcite, due to the longer exposure time to soil fluids. To conclude, the release of nanofibres may represent a partly destructive decay of the fibrous cell wall material. Moreover, this interpretation has an important implication for NFC origin. The observation of NFC inside organic sleeves and the presence of small mats of nanofibres on bundles of NFC (Fig. 5d (star); Cailleau et al. 2009) suggest a large contribution of organic matter for their genesis.
234 S. BINDSCHEDLER ET AL. Fig. 6. (a) and (c) SEM view of samples. The black window shows the area analysed. (b) and (d): EDS spectra, dotted lines correspond to Au peaks (samples coated with gold for SEM observations) and Os is the osmium labelling. (a, b) Analysis performed on a dense unorganized mesh of nanofibres (sample from coatings on a block). Note the absence of an osmium peak on the spectrum in the grey box due to the purely mineral nature of nanofibres. (c, d) Analysis performed on a dense non-random mesh of nanofibres (sample from a coating on a block associated with a decaying fungal strand). Note the presence of osmium peaks on the spectrum due to the organic nature of nanofibres.
FUNGAL ORIGIN FOR CALCITIC NANOFIBRES
235
Fig. 7. Hypothetical sketch recapitulating the potential processes of fungal organic matter decay and mineralization of cell wall fibrous material. This model is based on observations, analysis, and interpretations given in this paper. The first stage starts with a fungal hyphae or a group of hyphae forming a fungal strand. The cell wall is constituted by two main layers, an amorphous one (on the top) and a fibrous one (at the bottom; for more details see Fig. 1). It is assumed that the cell wall fibrous material is not decayed at the same rate than the amorphous material. When released in the soil microenvironment, these nanofibres could act as template for calcite nucleation, eventually leading to calcitic pseudomorphosis. The ultimate step is represented by possible reworking due to various processes (bioturbation, water movements, cryoturbation, etc.). Nanofibres are often associated with other calcitic features, such as NFC, leading to the complex microfabric observed in soil and cave deposits. O.M., organic matter.
236
S. BINDSCHEDLER ET AL.
As first noted by Phillips & Self (1987), NFC bundles could be the first step in the distribution of NFC in soil pores. Indeed, with time, the collapse of bundles due to various processes, such as weathering and bioturbation, would lead to a random distribution of NFC (mesh). Nanofibres on bundles would then be a relict of the organic sheath (assumed to be fungal in origin). The presence of mycelial strands is critical to understand the origin of the bundles. Strands and bundles are both organized as a tubular structure (Figs 3a & 5b) composed of sub-parallel components. They have similar diameters: 2–30 mm on average for the bundle and 8– 80 mm for the whole fungal strand. But the outer layer of the mycelial strand is usually wide and often represents between a third to a half of the strand section. Thus, only the inner diameter should be considered in this case. The outer layer is composed of hyphae with thick cell walls. Their central hole is probably too small to contain any NFC, whereas the inner part of the strand contains wider hyphae that would have enough room to allow the formation of a crystal such as a needle. One of the most important elements for fungal growth is calcium. Indeed, it is implicated in the apical growth control. Nevertheless, Ca2þ is considered as toxic when present in high concentrations (Gadd 1993). Consequently, its concentration within the fungal cell, and especially in the apex, must be under strict control of the organism in order to allow proper growth (Jackson & Heath 1993). Under hydrous stress conditions, the concentration of calcium could reach a high level, close to saturation. As it has been suggested for metal-oxalate (Whitney 1989; Gadd 1999), fungi could induce the precipitation of carbonate, possibly leading to a decrease of their internal calcium content (Gadd 2007). This process is documented for bacteria (Simkiss 1986; Schultze-Lam et al. 1996; Barton & Northup 2007). The inner layer of the fungal cell wall is composed by a large amount of chitin known to be a good template for calcite precipitation (Manoli et al. 1997). Consequently, nucleation of calcite crystals inside the inner functional hyphae from mycelial strands constitutes a serious hypothesis. The role of fungal hyphae as a crystal nucleation enhancer has already been suggested in the past (Went 1969; Northup & Lavoie 2001; Gadd 2007). Any other cell wall fibrous material or polymeric substance, for example b (1-3) glucan or a glycoprotein, may have the same effect (Burazerovic et al. 2007; Shen et al. 2007). To conclude, all our observations are recapitulated in a step-by-step hypothetical model (Fig. 7), showing the potential relationships between fungal organic matter and calcium carbonate precipitation.
Conclusion Considering previous hypotheses on the origin of nanofibres (i.e. biogenic or purely physicochemical), the results presented here indicate that nanofibres could also originate from the breakdown of fungal hyphae, especially their cell walls. During the decay of organic matter, microfibrils such as chitin or b (1-3) glucan, are released from the inner layer of the fungal cell wall. When these organic nanofibres are exposed to mineralizing pore fluids, they could undergo calcitic pseudomorphosis and/or be used as templates for calcitic precipitation. In the case of NFC bundles, which have an intimate relationship with nanofibres, these nanofeatures could indicate the relict of an organic sheath. As interpreted by Phillips & Self (1987), the implication of fungal strands in the genesis of NFC is now better supported. In other words: bundles could be the ultimate remains of the presence of a fungal strand. This hypothesis emphasizes the important role of organic matter in carbonatogenesis as well as the fundamental role of fungi in the terrestrial carbon cycle. The authors would like to thank Andre´ Villard and Miche`le Vlimant for their technical assistance for sample preparation, especially for TEM purposes; Dr Massoud Dadras, Dr Vladislav Spassov, and Mireille Leboeuf from CSEM for their assistance in using electron microscopy, and Dr. Thierry Adatte from UNIL for X-ray diffraction analysis. This work is supported by the Swiss National Science Foundation, Grant No. FN 205320-109497/1 and FN 205320-122171.
References A RONSON , J. M. & P RESTON , R. D. 1960. An electron microscopic and X-ray analysis of the walls of selected lower phycomycetes. Proceedings of the Royal Society of London. Series B, Biological Sciences, 152(948), 346–352. B ARTON , H. A. & N ORTHUP , D. E. 2007. Geomicrobiology in cave environments: past, current and future perspectives. Journal of Cave and Karst Studies, 69(1), 163– 178. B ECKETT , A., H EATH , I. B. & M C L AUGHLIN , D. J. 1974. An atlas of fungal ultrastructure. Longman Group Limited, London. B ENZERARA , K., M ENGUY , N., G UYOT , F., D OMINICI , D. & G ILLET , P. 2003. Nannobacteria-like calcites at surface of the Tatahouine meteorite. Proceedings of the National Academy of Sciences of the United States of America, 100, 7438– 7442. B ORSATO , A., F RISIA , S., J ONES , B. & V AN DER B ORG , K. 2000. Calcite moonmilk: crystal morphology and environment of formation in caves in the Italian Alps. Journal of Sedimentary Research, 70, 1179– 1190. B OUMA , T. J., N IELSEN , K. L., V AN H AL , J. & K OUTSTAAL , B. 2001. Root system topology and diameter
FUNGAL ORIGIN FOR CALCITIC NANOFIBRES distribution of species from habitats differing in inundation frequency. Functional Ecology, 15, 360– 369. B OWMAN , S. M. & F REE , S. J. 2006. The structure and synthesis of the fungal cell wall. BioEssays, 28, 799–808. B URAZEROVIC , S., G RADINARU , J., P IERRON , J. & W ARD , R. 2007. Hierarchical self-assembly of onedimensional streptavidin bundles as a collagen mimetic for the biomineralization of calcite. Angewandte Chemie, 46, 1 –5. B URNETT , J. H. 1979. Aspects of the structure and growth of hyphal walls. In: B URNETT , J. H. & T RINCI , A. P. J. (eds) Fungal Walls and Hyphal Growth. British Mycological Society Symposium. Cambridge University Press, Cambridge, 1– 25. C AILLEAU , G. 2005. Carbon cycle and carbonate biomineralization in terrestrial environments: diagenesis of oxalate– carbonate phases. PhD thesis, Neuchaˆtel University, Switzerland. C AILLEAU , G., B RAISSANT , O., D UPRAZ , C., A RAGNO , M. & V ERRECCHIA , E. P. 2005. Biologically induced accumulations of CaCO3 in orthox soils of Biga, Ivory Coast. Catena, 59, 1 –17. C AILLEAU , G., V ERRECCHIA , E. P., B RAISSANT , O. & E MMANUEL , L. 2009. The biogenic origin of needle fibre calcite. Sedimentology, 56, 1858–1875. C ANADELL , J., J ACKSON , R. B., E HLERINGER , J. R., M OONEY , H. A., S ALA , O. E. & S CHULZE , E. D. 1996. Maximum rooting depth of vegetation types at the global scale. Oecologia, 108(4), 583– 595. C AN˜ AVERAS , J. C., C UEVZA , S., S ANCHEZ -M ORAL , S., L ARIO , J., L AIZ , L., G ONZALEZ , J. M. & S AIZ J IMENEZ , C. 2006. On the origin of fibre calcite crystals in moonmilk deposits. Naturwissenschaften, 93, 27– 32. C ARLILE , M. J., W ATKINSON , S. C. & G OODAY , G. W. 2001. The Fungi, 2nd edn. Elsevier Academic Press, Amsterdam. C OLEMAN , D. C., C ROSSLEY , D. A., J R . & H ENDRIX , P. F. 2004. Fundamentals of Soil Ecology. Elsevier Academic Press, Amsterdam. D IX , N. J. & W EBSTER , J. 1995. Fungal Ecology. Chapman & Hall, London. D UCLOUX , J., B UTEL , P. & D UPUIS , T. 1984. Microse´quence mine´ralogique des carbonates de calcium dans une accumulation carbonate´e sous galets calcaires, dans l’ouest de la France. Pedologie, 34, 161 –177. F ARKAS , V. 1979. Biosynthesis of cell-walls of fungi. Microbiological Reviews, 43(2), 117 –144. F ARKAS , V. 2003. Structure and biosynthesis of fungal cell walls: methodological approaches. Folia Microbiologica, 48(4), 469– 478. F INLAY , R. D. & S O¨ DERSTRO¨ M , B. 1992. Mycorrhiza and carbon flow to the soil. In: A LLEN , M. (ed.) Mycorrhizal Functioning: an Integrative Plant–Fungal Process. Chapman and Hall, London, 134–160. G ADD , G. M. 1993. Interactions of fungi with toxic metals. New Phytologist, 124, 25–60. G ADD , G. M. 1999. Fungal production of citric and oxalic acid: importance in metal speciation, physiology and biogeochemical processes. Advances in Microbial Physiology, 41, 47– 92. G ADD , G. M. 2007. Geomycology: biogeochemical transformations of rocks, minerals, metals and
237
radionuclides by fungi, bioweathering and bioremediation. Mycological Research, 111, 3 –49. G OBAT , J.-M., A RAGNO , M. & M ATTHEY , W. 2003. Le Sol Vivant. 2nd edn. Presses polytechniques et universitaires romandes, Lausanne. G RADZINSKI , M., S ZULC , J. & S MYK , B. 1997. Microbial agents of moonmilk calcification. Proceedings of the 12th International Congress of Speleology, 1, 275– 278. H ISHI , T. 2007. Heterogeneity of individual roots within the fine root architecture: causal links between physiological and ecosystem functions. Journal of Forest Research, 12, 126–133. I USS W ORKING G ROUP W ORLD R EFERENCE B ASE . 2006. World Reference Base for Resources 2006. 2nd edn. World Soil Resources Reports No. 103. FAO, Rome. J ACKSON , S. L. & H EATH , I. B. 1993. Roles of calcium ions in hyphal tip growth. Microbiological Reviews, 57(2), 367–382. J ASINSKA , E. J., K NOTT , B. & M C C OMB , A. J. 1996. Root mats in ground water: a fauna-rich cave habitat. Journal of the North American Benthological Society, 15(4), 508– 519. J EONG , G. Y. & C HUN , Y. 2006. Nanofibre calcite in Asian dust and its atmospheric roles. Geophysical Research Letters, 33, L24802, doi: 10.1029/ 2006GL028280. J ONES , B. & N G , K.-C. 1988. The structure and diagenesis of rhizoliths from Cayman Brac, British West Indies. Journal of Sedimentary Petrology, 58, 457– 467. J ONES , D. 1970. Ultrastructure and composition of cell walls of Sclerotinia–Sclerotiorum. Transactions of the British Mycological Society, 54, 351 –360. L ATGE´ , J.-P. 2003. Activity Reports, Institute Pasteur, Paris. World Wide Web Address: http://www. pasteur.fr/recherche/RAR/RAR2003/print/Aspergilen.html. L OISY , C., V ERRECCHIA , E. P. & D UFOUR , P. 1999. Microbial origin for pedogenic micrite associated with a carbonate paleosol (Champagne, France). Sedimentary Geology, 126, 193– 204. M ANOLI , F., K OUTSOPOULOS , S. & D ALAS , E. 1997. Crystallization of calcite on chitin. Journal of Crystal Growth, 182, 116–124. N ORTHUP , D. E. & L AVOIE , K. H. 2001. Geomicrobiology of caves: a review. Geomicrobiology Journal, 18, 199 –222. O ULD M OHAMED , S. & B RUAND , A. 1994. Morphology and origin of secondary calcite in soils from Beauce, France. In: R INGROSE -V OASE , A. J. & H UMPHREYS , G. S. (eds) Soil Micromorphology: Studies in Management and Genesis. Proceedings of the IX International working Meeting on Soil Micromorphology, Townsville, Australia, July 1992. Developments in Soil Science, Elsevier, Amsterdam, 22, 27–36. P AUL , E. A. & C LARK , F. E. 1996. Soil Microbiology and Biochemistry. 2nd edn. Academic Press, San Diego. P EARSON , V. K., K EARSLEY , A. T. & S EPHTON , M. A. 2004. The in-situ detection of organic material in extraterrestrial samples. Microscopy and Analysis, 18, 5 –8. P ESSONI , R. A. B., F RESHOUR , G., F IGUEIREDO R IBEIRO , R. L., H AHN , M. G. & B RAGE , M. R.
238
S. BINDSCHEDLER ET AL.
2005. Cell-wall structure and composition of Penicillium janczewskii as affected by inulin. Mycologia, 97(2), 304–311. P HILLIPS , S. E. & S ELF , P. G. 1987. Morphology, crystallography and origin of needle-fibre calcite in Quaternary pedogenic calcretes of South Australia. Australian Journal of Soil Research, 25, 429– 444. P HILLIPS , S. E., M ILNES , A. R. & F OSTER , R. C. 1987. Calcified filaments: an example of biological influences in the formation of calcrete in South Australia. Australian Journal of Soil Research, 25, 405–428. P OUGET , M. & R AMBAUD , D. 1980. Quelques types de cristallisation de calcite dans les sols a` crouˆte calcaire (steppes alge´riennes). Apport de la microscopie e´lectronique. In: H UMBERT , L. (ed.) Cristallisation– De´formation–Dissolution des Carbonates. Universite´ de Bordeaux III, Institut de Ge´odynamique, 371–380. P REGITZER , K. S., D E F OREST , J. L., B URTON , A. J., A LLEN , M. F., R UESS , R. W. & H ENDRICK , R. L. 2002. Fine root architecture of nine North American trees. Ecological Monographs, 72(2), 293–309. P RESCOTT , L. M., H ARLEY , J. P. & K LEIN , D. A. 2003. Microbiologie. 2nd edn. De Boeck Universite´, Bruxelles. R ICHTER , D. K., I MMENHAUSER , A. & N EUSER , R. D. 2008. Electron backscatter diffraction documents randomly orientated c-axes in moonmilk calcite fibres: evidence for biologically induced precipitation. Sedimentology, 55, 487– 497. R UIZ -H ERRERA , J. 1992. Fungal Cell Wall: Structure, Synthesis and Assembly. CRC Press, Boca Raton, Florida. R UIZ -H ERRERA , J., L EON , C. G., C ARABEZ -T REJO , A. & R EYES -S ALINAS , E. 1996. Structure and chemical composition of the cell walls from the haploid yeast and mycelial forms of Ustilago maydis. Fungal Genetics and Biology, 20(2), 133– 142. S ARBU , S. M., K ANE , T. C. & K INKLE , B. K. 1996. A chemoautotrophically based cave ecosystem. Science, 272, 1953– 1955. S CHULTZE -L AM , S., F ORTIN , D., D AVIS , B. S. & B EVERIDGE , T. J. 1996. Mineralization of bacterial surfaces. Chemical Geology, 132, 171–181. S HEN , Y. H., X IE , A. J., Y ANG , Y. M., H UANG , F. Z., C HEN , L. & H UANG , J. T. 2007. Growth of calcium bilirubinate crystal controlled by functional polymer. Reactive and Functional Polymers, 67(3), 241–246.
S IMKISS , K. 1986. The process of biomineralization in lower plants and animals - an overview. In: L ANDBEATER , B. S. C. & R IDING , R. (eds) Biomineralization in Lower Plants and Animals. Systematics Association Special Volume, Oxford, Clarendon, 30, 19– 37. T HOMPSON , W. 1984. Distribution, development and fuctionning of mycelial cord systems of decomposer basidiomycetes of the deciduous woodland floor. In: J ENNINGS , D. H. & R AYNER , A. D. M. (eds) The Ecology and Physiology of the Fungal Mycelium. British Mycological Society Symposium, Cambridge University Press, Cambridge, 165–184. T IMONEN , S. & M ARSCHNER , P. 2006. Mycorrhizosphere Concept. In: M UKERJI , K. G., M ANOHARACHARY , C. & S INGH , J. (eds) Microbial Activity in the Rhizosphere. Soil Biology, Volume 7. Springer-Verlag, Berlin, Heidelberg, 155– 172. ¨ LL , L., K UYPER , T. W., S MITS , M. M., L ANDVAN S CHO EWEERT , R., H OFFLAND , E. & VAN B REEMEN , N. 2008. Rock-eating mycorrhizas: their role in plant nutrition and biogeochemical cycles. Plant and Soil, 303(1– 2), 35– 47. V ERGE` S , V., M ADON , M., B RUAND , A. & B OQUIER , G. 1982. Morphologie et crystallogene`se de monocristaux superge`nes de calcite en aiguilles. Bulletin of Mineralogy, 105, 351–356. V ERRECCHIA , E. P. & V ERRECCHIA , K. E. 1994. Needle fibre calcite: a critical review and a proposed classification. Journal of Sedimentary Research, 64, 650–664. W ATKINSON , S. C. 1979. Growth of rizhomorphs, mycelial strands, coremia and sclerotia. In: B URNETT , J. H. & T RINCI , A. P. J. (eds) Fungal Walls and Hyphal Growth. British Mycological Society Symposium, Cambridge University Press, Cambridge, 93– 113. W ATKINSON , S. C. 1984. Morphogenesis of the Serpula lacrimans colony in relation to its function in nature. In: J ENNINGS , D. H. & R AYNER , A. D. M. (eds) The Ecology and Physiology of the Fungal Mycelium. British Mycological Society Symposium, Cambridge University Press, Cambridge, 165–184. W ENT , F. W. 1969. Fungi associated with stalactite growth. Science, 166, 385–386. W HITNEY , K. D. 1989. Systems of biomineralization in the Fungi. In: C RICK , R. E. (ed.) Origin, Evolution and Modern Aspects of Biomineralization in Plants and Animals. Plenum Press, New York, 433–441.
The fractionation of phosphorus in some modern and late-Holocene calcareous tufas in North Yorkshire, UK ALLAN PENTECOST1,2 1
The Freshwater Biological Association, Ferry Landing, Far Sawrey, Ambleside, Cumbria, UK LA22 0LP (e-mail:
[email protected]) 2
King’s College, Strand, London UK WC2R 2LS
Abstract: Phosphorus fractionation studies were undertaken on seven UK tufas: three modern, and four old. The phosphorus in the carbonate fraction of the tufa averaged 19% of the total phosphorus in the modern material, and from 48–64% in the 4000 year-old deposits and the increase attributed to the mineralization of the contained organic matter. Two further phosphorus fractions contained significant amounts of P. The dithionite fraction ranged from 10– 27% of the total, and most of this fraction was probably associated with detrital iron minerals. The alkali-soluble fraction which removed most of the organically-bound P was highest in the modern bryophyte tufas. Total P levels in the tufas ranged from 33– 119 ppm. These are low values for stream sediments. Phosphorus uptake rates into tufa were estimated for these sites. They were less than 5% of the incoming P and deemed to have a negligible effect on the aquatic biota.
Actively-forming deposits of calcareous tufa are often festooned with aquatic plants such as algae and bryophytes. While our knowledge of the flora, and to some extent, the fauna of these deposits in Europe is good (Durrenfeldt 1978), the nutritional ecology of the streams in which the deposits and the associated biota occur has received little attention. This is in contrast with lake and stream systems in general where a large amount of information exists on the relationships between nutrient levels and biota (e.g. Wetzel 2001; Prior & Johnes 2002). Phosphorus is one of the most important micronutrients for aquatic organisms, as it is often in such short supply that it limits the growth rates of communities. When present in excess through human activity it can lead to dramatic changes in the biomass and composition of aquatic ecosystems (Dodds 2002). The low solubility of most inorganic phosphorus compounds contrasts strongly with another important nutrient, nitrogen, and results in the deposition of phosphorus in freshwater sediments. These sediments, under certain conditions, can release phosphorus back into water where it becomes available to plants, and freshwater sediments can provide a major reservoir for this element. Tufa, as a freshwater sediment, may also behave in this manner. Phosphorus fractionation studies, using a range of extractants, permit some understanding of the association of this element with the different mineralogical and organic components of sediments. From this it is often possible to predict the mobility of this element within the sediment and also understand the phosphorus pathway from water to sediment. In order to provide some preliminary data that will help answer these questions, a small number
of samples from modern and late Holocene tufas have been investigated, to determine their total phosphorus content, and the fractionation of the element within them.
Materials and methods Seven samples of tufa were collected from deposits in the Yorkshire Dales, UK. They consisted of two modern bryophyte-dominated samples, two modern alga-dominated and three inactive deposits dating from 0– c. 4000 years in age (Table 1). They were chosen to provide a representative sample of the dominant plant associations, modern and fossil within this area. Five separate subsamples from each site were pooled to provide a single sample for analysis. All samples were retained moist for the fractionation studies, since anomalous results have been reported from samples that have been dried (Golterman 1996). To determine gross composition, sub-samples were first dried at 105 8C for 2h then lightly ground and weighed, and the calcium carbonate removed by dissolution in M HCl. The residues were washed in distilled water, ashed at 550 8C for 2h and the organic and acid-insoluble inorganic residue determined by difference. For the phosphorus fractionation about 260 mg of material was used and the P extracted according to the method of Golterman (1996) for lake sediments. This method is based upon a series of chemical extractions to remove the phosphorus associated with the main mineral and organic components. An initial cold water extraction of the material (25 ml)
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 239–244. DOI: 10.1144/SP336.12 0305-8719/10/$15.00 # The Geological Society of London 2010.
240
A. PENTECOST
Table 1. Site descriptions, tufa composition and total P content Number
1 2 3 4 5 6 7
Description
Nat Grid ref
Age yr
CaCO3% dry wt.
Organic matter %
Acid-insol. inorganic residue %
Total P ppm
Schizothrix crust, Gordale Rivularia crust, Gordale Fossil Rivularia, Gordale Eucladium crust, Gordale Fossil Eucladium, Gordale Palustriella mat, Waterfall Beck Palustriella tufa, Waterfall Back 24 yr old
34/915643 34/914643 34/915641 34/915643 34/915641 34/908698
0 0 4000 – 4500 0 4000 –4500 0
98.0 94.5 98.4 91.1 98.9 95.4
2.0 5.1 0.5 6.4 0.3 2.7
0.1 0.4 1.1 2.4 0.8 1.8
33 64 104 119 67 84
34/908698
26
94.6
3.0
2.3
88
for 24 h was followed by centrifugation, to determine loosely-bound dissolved orthophosphate using the standard molybdenum blue method. This was followed by an extraction with 0.1 M Na2EDTA (disodium ethylene diamine tetraacetic acid) at pH 4.5 for 24 h followed by centrifugation. The extraction was repeated three times to remove all of the carbonate from the sample. The combined extract was then analysed by the modified molybdenum blue method as described by Golterman (1996). The residue was then extracted with sodium dithionite/EDTA solution for a further 24 h and analysed as above to remove phosphorus (P) associated with iron minerals. There followed extractions using 0.5 M H2SO4 and 2 M NaOH to remove P associated with the organic fraction. Some modifications to the method had to be made owing to the high calcium carbonate content. These consisted of: (a) interposing an extraction using 5 M HCl between the 0.5 M H2SO4 stage and the 2 M NaOH stage; and (b) a final fusion of the residue with Na2CO3 on platinum foil to determine the recalcitrant mineral fraction and total phosphorus content. Total P was determined by adding together all of the above fractions and also by determining the total P upon ignition and digestion of a separate tufa sample. The two estimates did not differ significantly.
Results There are no obvious trends in total P between the samples (Table 1). Values ranged from 33 ppm (modern Schizothrix) to 119 ppm (modern Eucladium). The results of the fractionation study are presented in Figure 1 and Tables 1 and 2. Seven fractions were obtained and these will be discussed in turn.
Water-soluble P This was a small fraction with the highest values found in the modern cyanobacterium deposits. Although in lake sediments it represents the dissolved interstitial P and that weakly sorbed onto surfaces, in this study it probably includes some P resulting from cell damage in the modern deposits. This was due to the grinding process necessary to ensure that the calcite was efficiently dissolved by EDTA fractionation, below.
EDTA fraction Extraction with EDTA at pH 4.5 is considered the most effective method for measuring P associated with carbonate minerals (Golterman 1996). However, it is a difficult procedure since precautions are needed to remove the interference of EDTA in the analysis of orthophosphate. This fraction includes P that is sorbed or precipitated onto calcite. Although calcium carbonate is the dominant component of these deposits, (.90% dry wt), the P associated with it is proportionately less than that associated with the acid-insoluble mineral and organic matter. The proportion of total P in the modern tufas averaged 19% but the P associated with carbonate was higher in the fossil tufas, and this was particularly noticeable in the case of the Rivularia and Eucladium facies where it was 64% and 48% of the total respectively (Table 2).
EDTA-dithionite fraction This fraction includes P that is released when samples are extracted under the reducing conditions imposed by sodium dithionite. The amounts extracted were significant, ranging from 10 –27% of the total. There were no obvious trends between samples.
THE FRACTIONATION OF PHOSPHORUS IN SOME CALCAREOUS TUFAS IN UK
241
Fig. 1. Fractionation of phosphorus in the seven tufa samples (as ppm P dry wt.). Euclad., Eucladium verticillatum; Palust., Palustriella commutata; Riv., Rivularia haematites; Schiz., Schizothrix species.
Acid fractions These were all small and ranged between 1.5–15% of the total, being lowest in the fossil material. The fraction was subdivided into a strong (5 N HCl) and weak (0.5 N HCl) acid extraction and most of the P was removed by the strong acid. These extractions will remove some P associated with detrital minerals and some acid – hydrolysable organic P from the organic matter. Since the highest fractions appear in the modern material it
is assumed that most of this P was associated with organic matter.
Alkali fraction This was a large fraction, particularly in the modern bryophyte tufas. The procedure removed most of the remaining organically-bound P, and it is clear that this is a significant fraction even in the fossil material.
Table 2. Percentage fractionation of phosphorus in the tufas Fraction Water-soluble EDTA EDTA-dithionite H2SO4 0.5 N HCl 5 N Alkali Na2CO3
1 Riv. modern
2 Riv. fossil
3 Schiz. modern
4 Palust. modern
5 Palust. 26 yr
6 Euclad. modern
7 Euclad. fossil
7.9 17.3 26.8 0.8 14.2 31.5 1.6
1.4 64.4 22.1 0.5 1.0 9.6 1.0
7.6 12.1 21.2 1.5 12.1 39.4 6.1
1.8 39.3 21.4 0.6 9.5 23.8 3.6
2.3 32.9 13.6 0.1 2.3 47.7 1.1
1.7 8.4 10.0 0.3 2.5 71.2 5.9
3.0 48.0 16.5 0.1 3.0 28.5 0.8
242
A. PENTECOST
Sodium carbonate fraction This component is not normally determined in P fractionation studies but was included here to remove all the remaining P from the sample. It can be regarded as ‘recalcitrant P’ and will include occluded or strongly-bound mineral P plus any organic P not extracted by other methods. It is generally regarded as ‘biologically unavailable’ and it seen to be an insignificant fraction in all of the samples, ranging from 1–6% of the total.
Discussion There have been few reports of P analyses from tufas and travertines. Demovic et al. (1972) reported some values from Central Europe and in a previous study of UK tufas, Pentecost (1993) found that the UK mean P content was 210 ppm with no significant difference between the levels in modern and fossil material. These data were summarised in Pentecost (2005) where an absolute range of 8–950 ppm was noted. Similar amounts of P have been reported from speleothem by Huang et al. (2001) and Borsato et al. (2007). In an Italian speleothem displaying annual laminations, it was found that phosphorus levels varied seasonally and were associated with an organic-rich layer thought to represent an autumnal infiltration phase, but there appear to be no P fractionation studies of speleothems. In tufas, organic matter content may also vary seasonally owing to variations in plant productivity, but high resolution studies of P levels have not yet been made. House et al. (1986) found Ca:P ratios of 200– 2500 in some calcareous stream deposits depending on the total P in solution in one experimental channel in the southern UK. These ratios are far lower than those found here, which ranged from 2400–8900. River sediments generally contain much higher levels of total P than tufas. A global survey by Froelich (1988) found most to be in the range 1150–1250 ppm. The low values found in the tufa probably reflect the small organic matter and clay content, and the oligotrophic nature of the associated waters. Fractionation studies of phosphorus were first applied to soils and lake sediments to gain an understanding of the form of the element in these materials and their availability to plants for growth. The water-soluble fraction is considered part of the ‘biologically available sediment P’. It is in fact the most available fraction to plants. In the tufa, a small but significant proportion appears to be present that could be available to aquatic plants, particularly the bryophytes since they develop rhizoids that can penetrate the surface layers of the deposit.
The carbonate fraction might be expected to contain most of the phosphorus since it is by far the greatest component of tufa, and phosphorus co-precipitation readily occurs. However the results demonstrated that in only one sample was this the case. The older tufas contained a greater proportion of P in the carbonate fraction and this suggests that some additional P is sorbed or co-precipitated after deposition. The process appears to take place over a long time, because in the case of the Palustriella facies, the carbonate and other fractions remain similar over a period of 26 years. The 26-year old tufa sample was collected from a deposit that began to form in 1983 and it was evident that while there were no obvious intact bryophyte remains within it, the gross composition of the material had otherwise changed little. In Malham Tarn, a marl lake less than 5 km from these sites with waters derived from the same limestone formation, calcium carbonate, the major component of the sediment, contained only 14% of the total phosphorus (Pentecost 1998). Here, the total P content of the sediment was an order of magnitude higher than that of the tufas. In the Tarn sediments however, the proportion of organic matter was much higher owing to the retention of lake seston. Little is known of the form of the phosphorus in these carbonates although it appears likely that most exist as a calcium phosphate co-precipitate (House 1990). Studies of speleothem using NMR spectroscopy indicate the existence of monetite (CaHPO4) and incorporated PO4 ions in the calcite lattice (Mason et al. 2007). There has been much interest in the association of phosphorus with calcium carbonate in freshwaters (e.g. Koschel 1983; Istva´novics et al. (1989); Talling & Parker 2002). Pettersson & Bostro¨m (1986) in a detailed study of phosphorus in the sediments of calcareous Lake Balaton, Hungary, estimated a phosphorus co-precipitation rate of 0.6–2.2 mgP/m2/day onto calcium carbonate. Estimates may also be made for the tufa sites. Deposition rates of tufa in the Yorkshire streams are known to be about 3 mm/year. Given the total P values in Table 1, total P removal rates to tufa would amount to about 1 mgP/m2/day, although only about 20% of this would be associated with the calcite. Gordale Beck, one of the sampling sites, has a 2 km length where tufa is deposited extensively. With a width of about 3 m, a total P content of c. 10 mg/l and a discharge of about 200 l/sec, the daily P removal rate to tufa would be about 3 g. This is only about 4% of the total amount of P entering with the stream, suggesting that P uptake by tufa is unlikely to have a significant effect on the stream biota. Similar values were obtained for the other sampled stream, Waterfall Beck.
THE FRACTIONATION OF PHOSPHORUS IN SOME CALCAREOUS TUFAS IN UK
In lake and river sediments, the dithionite fraction is associated with ferric oxy- hydroxides, to which P is strongly sorbed under conditions of circumneutral pH and the presence of oxygen. It is an important fraction in lakes, since ferric iron becomes reduced to ferrous iron in sediments devoid of free oxygen resulting in the release of bioavailable phosphorus. This phosphorus may lead to rapid plant growth and nuisance algal blooms. Tufas however, are much lower in their iron content when compared with most lake sediments and in the fast-flowing rivers where tufas occur, such enrichment is highly unlikely. Pentecost & Zhang (2008) found a mean Fe content of 280 ppm in the fossil Eucladium and 200 ppm in the fossil Rivularia sampled in this study. Slightly higher levels were found in the modern deposits but the overall atomic Fe:P ratios are around 2 for these deposits. Total Fe:Total P ratios are known for several lake sediments (e.g. Baccini et al. 1985; Jensen et al. 1992) and are mostly in the range 5–25 so the tufas have a low ratio in comparison. Of more interest is the total Fe:P ratio of the dithionite fraction. For the tufas this is about 10, a value which according to Roden & Edmonds (1997) is sufficient to saturate Fe(III) oxides with sorbed P. Based upon their acid and alkali fractions, it is evident that although the organic matter content of tufas declines with age, they still retain significant amounts of organically-bound P. This fraction is sometimes considered as part of the ‘biologically available P’ as much of the organically-bound P of plants can be released enzymatically during decay, although substantial amounts are clearly retained in tufa deposits. Pentecost (2005) estimated that the complete phosphate mineralization of a tufa containing 4% organic matter could raise the mineral P content by about 40 ppm. The modern samples analysed here contained from 2–6.4% organic matter, and although not all was mineralised after burial, the difference in the P content between the more labile fractions in the young and old material is certainly of the same order, suggesting that some mineralization has occurred.
Conclusions This brief investigation has demonstrated that tufas provide interesting contrasts in their phosphorus contents and fractionation when compared with other freshwater sediments. The main findings are: 1.
2.
The tufas possessed low concentrations of phosphorus when compared with other types of sediment although they were similar to those reported from speleothems. The carbonate fraction was not normally the most significant in terms of P content, but
3.
4.
243
older deposits appeared to contain more which is presumed to result from some organic matter mineralization. The dithionite fraction was similar to the carbonate fraction despite the low levels of iron, suggesting that the contained iron oxides contained a large proportion of sorbed P. Viewed as an ecosystem, it appears that tufa deposition in streams is unlikely to influence the dynamics of phosphorus significantly unless deposition is rapid and extensive.
References B ACCINI , P. 1985. Phosphate interactions at the sedimentwater interface. In: S TUMM , W. (ed.) Chemical Processes in Lakes. John Wiley & Sons, New York, 189– 205. B ORSATO , A., F RISIA , S., F AIRCHILD , I. J., S OMOGYI , A. & S USINI , J. 2007. Trace element distributions in annual stalagmite laminae mapped by micrometerresolution X-ray fluorescence: Implications for incorporation of environmentally significant species. Geochimica et Cosmochimica Acta, 71, 1499–1512. D EMOVIC , R., H OEFS , J. & W EDEPOHL , K. H. 1972. Geochemische untersuchungen an travertineen der Slowakei. Contributions to Mineralogy and Petrology, 37, 15–28. D ODDS , W. K. 2002. Freshwater Ecology: Concepts and Environmental Applications. Academic Press, San Diego. D URRENFELDT , A. 1978. Untersuchungen zur Besiedlungsbiologie von kalktuff faunistosche, o¨kologische und elekronenmikroscopische Befunde. Archiv fu¨r Hydrobiologie (Supplement Bande), 54, 1 –79. F ROELICH , P. N. 1988. Kinetic control of dissolved phosphate in natural rivers and estuaries: a primer on the phosphate buffer mechanism. Limnology and Oceanography, 33, 649– 668. G OLTERMAN , H. 1996. Fractionation of sediment phosphate with chelating compounds: a simplification, and comparison with other methods. Hydrobiologia, 335, 87–95. H OUSE , W. A. 1990. The prediction of phosphate coprecipitation with calcite in freshwater. Water Research, 24, 1017–1023. H OUSE , W. A., C ASEY , H. & S MITH , S. 1986. Factors affecting the coprecipitation of inorganic phosphate with calcite in hardwaters – 1. Water Research, 20, 917– 922. H UANG , Y. M., F AIRCHILD , I. J., B ORSATO , A., F RISIA , S., C ASSIDY , N. J. & M C D ERMOTT , F. 2001. Seasonal variation in Sr, Mg and P in modern speleothems (Grotta di Ernesto, Italy). Chemical Geology, 175, 429– 448. I STVA´ NOVICS , V., H ERODEK , S. & S ZILAGYI , F. 1989. Phosphate adsorption by different sediment fractions in Lake Balaton and its protecting reservoirs. Water Research, 23, 1357– 1366. J ENSEN , H. S., K RISTIANSEN , P., J EPPESEN , E. & S KYTTHE , A. 1992. Iron:phosphorus ratio in surface sediment as an indicator of phosphate release from
244
A. PENTECOST
aerobic sediments in shallow lakes. Hydrobiologia, 235/236, 731– 743. K OSCHEL , R., B ENNDORF , J., P ROFT , G. & R ECKNAGEL , F. 1983. Calcite precipitation as a natural control mechanism of eutrophication. Archiv fu¨r Hydrobiologie (Supplement Bande), 98, 380– 408. M ASON , H. E., F RISIA , S., T ANG , Y., R EEDER , R. J. & P HILLIPS , B. L. 2007. Phosphorus speciation in calcite speleothems determined from solid-state NMR spectroscopy. Earth and Planetary Science Letters, 254, 313–322. P ENTECOST , A. 1993. British travertines: a review. Proceedings of the Geologists’ Association, 104, 23– 39. P ENTECOST , A. 1998. Phosphorus fractionation in the sediments of Malham Tarn, North Yorkshire. Field Studies, 9, 337– 342. P ENTECOST , A. 2005. Travertine. Springer-Verlag, Berlin, Heidelberg. P ENTECOST , A. & Z HANG , Z. 2008. Microfossils and geochemistry of some modern, Holocene and Pleistocene travertines from North Yorkshire and Derbyshire.
Proceedings of the Yorkshire Geological Society, 57, 79–94. P ETTERSSON , K. & B OSTRO¨ M , B. 1986. Phosphorus exchange between sediment and water in L. Balaton. In: S LY , P. G. (ed.) Sediments and Water Interactions. Springer, New York, 427–435. P RIOR , H. & J OHNES , P. J. 2002. Regulation of surface water quality in a Cretaceous Chalk catchment, UK: an assessment of the relative importance of instream and wetland processes. Science of the Total Environment, 282–283, 159–174. R ODEN , E. E. & E DMONDS , J. W. 1997. Microbial Fe(III) oxide reduction versus iron sulphide formation. Archiv fu¨r Hydrobiologie (Supplement Bande), 139, 347– 378. T ALLING , J. F. & P ARKER , J. E. 2002. Seasonal dynamics and phytoplankton and phytobenthos, and associated chemical interactions, in a shallow upland lake (Malham Tarn, northern England). Hydrobiologia, 487, 167 –181. W ETZEL , R. G. 2001. Limnology. 3rd edn. Academic Press, San Diego.
Depositional properties and geochemistry of Holocene perched springline tufa deposits and associated spring waters: a case study from the Denizli Province, Western Turkey ´2 ¨ ZKUL1*, ALI˙ GO ¨ KGO ¨ Z1 & NADA HORVATINCˇIC MEHMET O 1
Pamukkale University Engineering Faculty Department of Geological Engineering, 20070 Kınıklı campus, Denizli, Turkey 2
Ruder Bosˇkovic´ Institute, Zagreb, Croatia
*Corresponding author (e-mail:
[email protected]) Abstract: The Gu¨ney waterfall area is a perched springline tufa site developed on the southeast slope of the Bu¨yu¨k Menderes River near Gu¨ney town, in the Denizli province, Western Turkey. The site is 12 km away from Gu¨ney and 72 km from the city centre of Denizli. The spring waters emerge from the boundary between Palaeozoic marble and micaschist and precipitated tufa deposits downslope at the altitudes ranging from 220 to 400 metres. The tufa deposits cover an area of about 20 hectares. Flat upper surfaces of the deposits are indicative of mature stage. The waters are of the Ca–HCO3 type and supersaturated with respect to CaCO3. The stable isotope values of the spring waters are 249.94 for d2H and 27.15 for d18O. The d13C and d18O values of active and passive tufa samples are in the range from 29.13 to 26.0‰, and from 28.44 to 27.40‰, respectively. These isotopic values are typical for fresh water tufa. The passive tufas give the 14C age in the range from 2000 to 5800 yr BP. According to the 14C age data, passive tufas are not older than Holocene. The stable isotope composition is similar south European examples.
Tufa and travertine are common carbonate deposits in Quaternary and present-day depositional systems (Chafetz & Folk 1984; Ford & Pedley 1996; Guo & Riding 1998; Arenas et al. 2000; Horvatincˇic´ et al. 2000; Glover & Robertson 2003; Bonny & Jones 2003) and are known from Neogene and older successions (Arp 1995; Evans 1999; Cole et al. 2004). In this paper, ‘tufa’ was considered as subaerial deposits produced from ambient temperature waters and contains typically the remains of micro- and macrophytes, invertebrates and bacteria (Ford & Pedley 1996; Arenas et al. 2000; Pedley et al. 2003). On the other hand, travertine is hydrothermal in origin (Chafetz & Folk 1984; Guo & Riding 1998; Minissale et al. 2002). Ford & Pedley (1996) divided tufa deposits into: (1) perched springline, (2) fluvial; (3) lacustrine; and (4) paludal depositional systems. Subsequently, Pedley (2009) added the cascade model. Perched springline tufa deposits, which occur on valley slopes in mountainous or hilly countries, are part of the tufa depositional continuum (Ford & Pedley 1996; Pedley 2009; Pedley et al. 2003). These deposits have been used as indicators of palaeohydrogeological evolution in karstic massifs and used to evaluate climatic change in the Mediterranean areas (Martı´n-Algarra et al. 2003). Although outlines of these deposits have been presented previously in Pedley (1990), Ford & Pedley (1996) and Pedley et al. (2003), there are few
well-documented sites (Andreo et al. 1999; Martı´n-Algarra et al. 2003; Anzalone et al. 2007). Although, the Denizli province in the western Turkey is well-known for its travertine deposits and associated geothermal fields (Altunel & Hancock ¨ zkul 1993; Go¨kgo¨z 1998; S¸ims¸ek et al. 2000; O et al. 2002; S¸ims¸ek 2003), there are also a few tufa sites (Ceylan 2000; Horvatincˇic´ et al. 2005). In this study, the depositional features and geochemistry of a unique example of perched springline tufa deposits and accompanying spring waters near the Gu¨ney town, Denizli, SW Turkey are examined.
Location and geological setting The site is located on the SE slope of the Bu¨yu¨k Menderes river valley near the town of Gu¨ney in the NW part of the Denizli basin, western Turkey (Figs 1 & 2). The site is 12 km from Gu¨ney and 72 km north of the city of Denizli. Based on long-term data, the mean annual temperature and the rainfall are about 13.5 8C and 548.8 mm, respectively. The tufa deposits rest on marble and schist of the Palaeozoic metamorphic bedrock that belong to the Menderes masif (Bozkurt & Oberha¨nsli 2001; Erdog˘an & Gu¨ngo¨r 2004). The bedrock is unconformably overlain by a late Miocene lacustrine succession (Koc¸yig˘it 2005; Kaymakc¸ı 2006; Alc¸ic¸ek et al. 2007) at higher altitude of 900 m where the Cindere
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 245–262. DOI: 10.1144/SP336.13 0305-8719/10/$15.00 # The Geological Society of London 2010.
246
¨ ZKUL ET AL. M. O
Fig. 1. (a) Geographical setting of the Denizli province in Turkey; (b) Location of the Gu¨ney springline tufa site (a) and other two tufa sites (b: Sakızcılar and c: Honaz) in the province and (c) Local geological map of the Gu¨ney springline tufa site. Sp-1, 2, 3 and 4 are spring resurgences.
village located (Fig. 2). The tufa deposits cover an area of about 20 hectares. A road goes up through the tufa deposits from the river bed towards Cindere (Fig. 1c). The Gu¨ney waterfall site is an area of unique geological/natural heritage that is visited by many people, especially
during spring and summer seasons. There are mainly four springs and numerous seeps that emerge from the metamorphic bedrock that has been fractured intensely. The spring waters have formed coalescent tufa bodies that occur at elevations between 220–400 m above the Bu¨yu¨k Menderes river bed (Fig. 1c).
PERCHED SPRINGLINE TUFAS, SW TURKEY
247
Fig. 2. Panaromic view (a) and drawing interpretation (b) of the Gu¨ney perched springline tufa site. The late Miocene lacustrine sediments rest unconformably on the bedrock composed of schist and marble belong to Menderes massif of Palaeozoic age around Cindere village.
Methods Fieldwork The Gu¨ney tufa site was visited seasonally from July 2003 to December 2005. During the early stages of the study, geological and depositional features of the tufa site were investigated and a local geological
map was prepared (Fig. 1c). Some physicochemical analyses and measurements of the waters were carried out at several points during midday along the flow paths in order to monitor the seasonal variations at physical and chemical parameters. Temperature and pH values were measured using a HACH Sension 2 pH meter at the time of sampling. Electrical conductivity was measured
248
¨ ZKUL ET AL. M. O
Fig. 3. Field view of the active waterfall and lobe-top terrace areas developed in front of Sp-4. The flat lobe-top terrace is presently used as an agricultural area by local people. The water of Sp-4 flowing through an elevated channel (dotted sinuous line) on the terrace area reaches to the spill-over point of the active waterfall-top. A primary cave (P.C., arrowed) has developed within the active waterfall tufa body.
with a HACH Sension 5 conductivity meter. CO2 and HCO3 were determined volumetrically by titration using NaOH and H2SO4, respectively. Water samples were collected for chemical analysis and each water sample consists of two 500 ml polyethylene bottles. One of the bottles was acidified with ultra pure HNO3 (pH , 2) after filtering with 0.45 mm cellulose membrane filters for the determination of cations and the other was unacidified for anion analyses which is kept at a temperature equal to lower than þ4 8C. Both active and passive tufa samples were collected for mineralogical, petrographical and geochemical analysis (including major and trace elements, SEM investigations and isotope composition).
Laboratory work Major constituents of water were determined at the geochemistry laboratory of the Geological Engineering Department at Pamukkale University. Ca, Mg, Na and K concentrations of the water
samples were determined using a UNICAM atomic absorption spectrometer. Cl and SO4 were determined by a HACH DR/4000 visible spectrophotometer. The stable isotope (d2H, d18O and d13C in DIC of water) and d3H analyses of some waters were performed at the laboratory of the Faculty of Earth and Life Science in Vrije University (Amsterdam) and at the Geological Engineering Department in Hacettepe University (Ankara), respectively. Mineral saturation indices of hydrothermal minerals were calculated using the PHREEQC computer code (Parkhurst & Appelo 1999). The tufa samples were examined by optical microscope in thin section and some selected samples were studied on a scanning electron microscope (SEM; Jeol JSM 6490 LV). SEM investigations were performed at Turkish Petroleum Corporation (TPAO) in Ankara. The mineralogy of the tufa samples was determined by X-ray diffraction (XRD) using a Phillips PW 1729 diffractometer at the Geological Engineering Department, Hacettepe University. Major and trace elements
PERCHED SPRINGLINE TUFAS, SW TURKEY
249
Fig. 4. Field photographs of the Gu¨ney perched springline tufa site. (a) The active waterfall area; (b) Frontal part of the extinct waterfall area adjacent to active part; (c, d) Close views of the active waterfall front accompanied by overhanged bryophyte (moss) curtains and pouring water column; (e) Lower slope tufa deposits exposed at the extinct waterfall foot along the roadside section, note that the steep inactive waterfall tufa facing downslope; (f ) An inside view of the primary cave located behind the tufa body of the active waterfall. A shallow pool took place at the bottom and the cave walls were coated by speleothems, note the trace of highstand water level above present water surface.
250
¨ ZKUL ET AL. M. O
Fig. 5. Field views of different tufa components at the Gu¨ney perched springline tufa site. (a) Encrusted algal mass associated with macrophyte pieces was enclosed by fine grained detrital tufa layer (upper right corner); (b) The light-coloured and laminated tufa facies, lower slope, roadside section; (c) Calcified, fragile bryophyte cluster in the
PERCHED SPRINGLINE TUFAS, SW TURKEY
251
Fig. 6. A topographic cross section from the Sp-4 to the Bu¨yu¨k Menderes river valley in north– south direction. Details of the waterfall area between GW-1 and GW-2 were shown at the upper right. Sp-1 and Sp-4 are spring numbers, GW-1, GW-2 and GW-3 are in-situ measurement points at waterfall head (¼ splash zone), waterfall bottom and the lowermost slope, respectively.
were carried out on 11 recent and old tufa samples from different positions. Analyses were performed at the XRAL Laboratories in Ontario, Canada, by ICP-AAS. The isotope analyses of 14C activity were made at Rudjer Boskovic Institute, in Radiocarbon and Tritium Laboratory, Zagreb, Croatia. The 14C measurement was performed by liquid scintillation counter (LSC), Quantulus 1220 using two methods of sample preparation, a benzene synthesis method (LSC-B) and a CO2 absorption method (LSC-D) (Horvatincˇic´ et al. 2004). Mass spectrometry measurements of stable isotopes content, d13C and d18O were performed using a Finnigan Delta XP mass spectrometer at Joanneum Research Institute of Water Resources Management Hydrogeology and Geophysics Stable Isotope Laboratory, Graz, Austria. d13C and d18O in carbonates is expressed in ‰ deviations from the international standard PDB.
Depositional features of the perched springline tufa deposits Perched springline tufas show distinct depositional morphologies in hilly country. The deposits are fanshaped in plan and wedge-like in profile (Chafetz & Folk 1984; Pedley 1990; Ford & Pedley 1996; Carthew et al. 2003; Pedley et al. 2003). In the study area, the perched springline tufa site has been mainly divided into lobe-top terrace, waterfall (cascade) and lower slope zones.
Lobe-top terrace area At the study site, there are two distinct lobe-top terraces developed in different elevations. One formed in front of Sp-2 whereas the other one formed in front of Sp-4 (Fig. 1c). The horizontal terraces are relatively narrow, long and restricted in areal extent. An elevated channel was developed on the lobe-top terrace (Fig. 3). This kind of channel previously was named as self-built channel (Bean ¨ zkul et al. 1971; Altunel & Hancock 1993; O 2002), catwalk/suspended channel (Violante et al. 1994; Pedley et al. 2003; Pedley 2009) at travertine and tufa sites. The water from Sp-4 flows along this gently sinuous channel for about 140 m across the subhorizontal terrace area before reaching the active waterfall head area or spill over point (GW-1 in Fig. 6). The channel was raised up from a few decimetres to metre scale above the pedestal of the lobe-top area. Lower half of the channel length is more elevated. Local people use both terraces for agricultural purposes.
Waterfall zone In the study site, the waterfall area is located in the upper part of the system. The active cascade area is fed by the waters of Sp-4 and immediately SW of the active side is extinct presently (Fig. 4b). Another extinct cascade was developed in front of Sp-2 to the NE of the tufa site (Fig. 1c). The steep face of the waterfall is c. 35 m high and covered
Fig. 5. (Continued) base of waterfall front; (d) Cylindrical macrophyte pieces in vertical position embedded in tufa body, lower slope, roadside section; (e) Fine- grained detrital tufas in the middle and lower right, distal slope on the roadside section; (f) Calcified bryophytes and coated stems; (g) Phytoherm framestone facies, immediately above the roadside section, lower slope; (h) Coated stems and micritic tufa crust underlain by detrital tufa. Scales: Camera cover is 13 cm, pen is 14 cm and pocket-knife is 5.5 cm.
252
¨ ZKUL ET AL. M. O
mostly by the bryophytes (Fig. 4a, c, d). The bryophytes and other organic entities, which are in growth position, are main components of the phytoherm framestone facies (Fig. 4c, d). The exposed surface is green whereas the inner parts of the tufa are yellow because of the calcite cement that coats the surfaces. There is an overhang on the uppermost part of the waterfall face (Fig. 4a, c) from which blocks of variable sizes commonly detach and fall to the foot of the waterfall. Blocks fell and a fracture was created in the overhang when the Buldan earthquakes of 5.2–5.6 magnitude occured close the Buldan town between 23 and 26 July in 2005 (Kumsar et al. 2008, p. 95 –98). The springline tufa body contains numerous primary cavities up to cave size in some cases. There are two main caves in the tufa site (Fig. 1c). One of them is under the splash zone within the active waterfall area (Fig. 3). A pool at the cave bottom is about 10 m wide and filled by percolating and drip waters (Fig. 4f). The second cave, to the NE, is located in the inactive tufa body that is close to Sp-2 (Fig. 1c). The pool in the bottom of this cave has been partly filled by wet lime mud and tufa fragments. Both caves have been partly or completely coated by various speleothems (Fig. 4f).
evidence of earlier deposits, although some sites as in the Tajuna valley (central Spain) demonstrate considerable preservation potential. As the river incises into its valley, there is often a corresponding reduction in the elevation of springlines along the valley size. There is, therefore, the possibility that deposits high on the upper valley sides are older than those lower down the slopes. In the case of multi-spring resurgences, age relationships between individual tufa mounds are more complex (Pedley et al. 2003). At the Gu¨ney perched springline tufa site, tufa lobes at different levels have coalesced each others. There are inactive lobes at various elevations above and below of the active waterfall area. Therefore, a regular downslope decrease in ages of the tufa lobes was not expected. The preliminary 14C dating of the some extinct tufa samples taken randomly gave 14C age in the range from 2000 yr BP to 5800 yr BP (Table 2). The 14C results are not corrected for the initial 14C activity (Ao) of the carbonate (the ‘reservoir’ effect). According to the 14C age data, extinct tufas are not older than Holocene (Horvatincˇic´ et al. 2005). Further age dating is required before the precise ages of deposition can be determined.
Spring water physiochemistry Lower slope area The lower slope (or distal slope) lies below the foot of the waterfall where the slope gradient changes abruptly. In plan view, the lower slope tufas are fan-like and cover a broad area in comparison with the waterfall. The most of these deposits are composed of intraclast tufa. The lower slope deposits were exposed clearly along a road cuts approaching from the Cindere dam to the south –southwest and river bad (Fig. 4e). At the road cut surfaces, the lower slope deposits are consist of dominantly detrital tufas (Fig. 5a, e, h) and associated laminated tufa in some places (Fig. 5b) and macrophyte fragments eroded from the cascade area (Fig. 5d, e, h). The detrital tufa deposits have changed from silt to block in size. The light-coloured laminated tufa facies is composed of bryophyte layers and they should have been precipitated along flow paths on the slope (Fig. 5b). The macrophytes (¼ phytoherm framestone facies) in some cases are in-situ life position on the lower slope (Fig. 5g).
Age relationships of the coalescent tufa lobes Perched springline tufas typically develop from single or multi-spring resurgences emerging on hill slopes. Subsequent valley incision may remove
There are four main springs (Fig. 1c) and many smaller seeps in the Gu¨ney tufa site. The total flow rate of the springs is about 80 L/s. The discharge elevations are more or less the same for two springs (Sp-2 at 415 m and Sp-3 at 417 m a.s.l.; see Fig. 1c for spring location), whereas the other two discharge at lower elevations (Sp-1 at 320 m and Sp-4 at 355 m a.s.l.). The waters of Sp-2, Sp-3, and Sp-4 emerge from the boundary between the tufa deposit and schist – marble alternation forming aquifer rocks. The spring water of Sp-4 reaches the active waterfall area via a 140 m long channel. In-situ measurements show that the water physiochemistry of springs Sp-2, Sp-3 and Sp-4 are similar and that there are not distinct seasonal variations. Hydrochemical features of the tufa precipitating waters are summarized in Table 1. The main spring waters discharge at 18.7 –18.8 8C (Table 1) throughout the year. The water temperature of Sp-1 is lower (15.9 8C in winter to 16.7 8C in summer) than the others. The water temperature in the splash zone of the waterfall lobe-top area shows a regular variation that is linked to seasonal air temperatures. The seasonal fluctuations in the water temperature are about 0.5 8C in January and November while 0.2–0.5 8C in April and July. The biggest variation in water temperature occurs in the waterfall area due to water falling. The water temperature at the cascade bottom was decreased 0.5 8C in summer, 2.1 8C in
Table 1. Chemistry of spring water at the Gu¨ney perched springline tufa site T 8C
EC
pH
CO2
Ca
Mg
Na
K
Cl
SO4
HCO3
DIC
SIC
log pCO2
Mg/Ca
October 2004
Sp-1 Sp-2 Sp-3 Sp-4 GW-1 GW-2
16.7 18.7 18.7 18.8 18.8 17.2
551 479 482 477 474 391
7.2 7.3 7.4 7.3 7.9 8.2
71 48 44 43 32 24
81.0 64.2 65.2 64.0 61.0 47.0
12.7 11.0 10.4 10.8 10.6 7.8
4.9 1.9 1.9 1.8 1.8 1.9
1.7 0.8 0.8 0.8 0.8 0.8
4.8 3.9 3.9 3.9 4.2 4.3
5.8 1.6 2.0 0.6 1.2 0.6
291 236 232 236 225 178
66.8 52.2 50.6 51.8 45.9 35.83
0.00 20.02 0.04 0.01 0.51 0.59
21.74 21.94 22.01 21.97 22.54 22.95
0.26 0.29 0.27 0.28 0.29 0.28
January 2005
Sp-4 GW-1 GW-2 GW-3
18.7 18.3 13.5 13.2
482 475 392 376
7.2 7.7 8.1 8.3
65 56 38 6
64.8 62.4 45.2 35.6
9.8 9.7 9.3 9.2
1.8 1.8 1.8 1.8
0.8 0.8 0.8 0.8
4.0 4.1 4.1 4.2
2.0 2.0 1.6 2.6
243 222 182 152
54.7 45.9 36.7 30.6
20.07 0.35 0.47 0.44
21.86 22.38 22.91 23.13
0.25 0.26 0.34 0.43
April 2005
Sp-4 GW-1 GW-2 GW-3
18.7 18.9 16.8 16.9
481 472 400 381
7.3 7.7 8.1 8.3
53 38 26 5
63.7 58.9 51.1 35.2
10.1 9.8 9.2 9.1
1.7 1.8 1.8 1.9
0.8 0.8 0.8 0.8
4.3 4.2 4.2 4.5
1.8 1.8 1.9 2.2
240 216 188 149
53.0 44.6 38.0 29.9
20.01 0.32 0.52 0.51
21.94 22.39 22.81 23.15
0.26 0.28 0.30 0.43
July 2005
Sp-4 GW-1 GW-2 GW-3
18.8 19.3 18.8 21
482 470 408 356
7.4 8.0 8.2 8.3
40 31 23 4
63.8 60.7 48.1 46.3
10.7 10.4 8.2 8.1
1.8 1.8 1.8 1.8
0.8 0.8 0.8 0.8
3.9 4.0 3.9 4.1
2.1 1.9 2.0 2.2
234 222 182 171
51.2 45.1 36.7 34.4
0.02 0.58 0.60 0.71
21.99 22.62 22.91 23.05
0.28 0.29 0.28 0.29
November 2005
Sp-4 GW-1 GW-2 GW-3
18.7 18.2 13.6 14.1
486 478 421 385
7.4 8.0 8.3 8.5
31 25 19 3
62.6 58.2 45.1 32.8
10.5 10.3 8.9 8.2
1.8 1.8 1.7 1.8
0.8 0.8 0.8 0.8
4.0 4.1 4.1 4.0
1.1 1.2 1.1 1.4
229 221 178 137
49.5 44.7 35.8 27.5
0.06 0.62 0.61 0.58
22.06 22.70 23.07 23.39
0.28 0.30 0.33 0.42
PERCHED SPRINGLINE TUFAS, SW TURKEY
No
Date
SIC, calcite saturation index; EC (electrical conductivity), mS/cm; pH, standard unit; DIC (dissolved inorganic carbon), ppmC; log pCO2, atm.; Mg/Ca, molar ratio; other constituents are in ppm.
253
254
Table 2. Major, trace element and isotope compositions of the Gu¨ney perched springline tufa deposits Sample no*
Si
Al
Fe
Mg
Ca
Ba
Sr
2150
580
980
1860
383500
97
236
0.80
2
9380
4230
2800
3120
341400
108
222
1.50
3 4
140 2000
50 890
700 840
3180 3240
379200 368000
152 125
249 230
1.38 1.45
5 6
930 1910
420 370
490 1050
2220 5940
380900 377300
143 136
225 237
0.96 2.56
7 8 9
1440 8770 190
580 2060 160
770 1680 770
3300 7440 1380
378300 352200 388900
33 183 192
262 285 322
1.43 3.40 0.59
10
6910
1790
1960
2220
375800
150
222
0.97
980
320
490
1680
381700
138
253
0.73
3164
1041
1139
3235
373382
133
249
1.43
d13C
C measurement
Method
pMC
yr BP
B49 D172 B56 D193 B51 B52 D182 B68 B57 D190 D228 D225 B59 D200 B61 D202 D248 B55
63.8 + 0.5 60.5 + 2.0 64.1 + 0.6 68.5 + 2.0 65.7 + 0.5 65.5 + 0.5 65.4 + 1.9 76.0 + 0.4 73.6 + 0.6 77.9 + 2.1 68.0 + 2.1 62.7 + 1.9 48.6 + 0.5 50.5 + 1.4 65.0 + 0.6 67.0 + 1.6 67.9 + 1.8 70.6 + 0.6
3610 + 60 4030 + 270 3565 + 70 3040 + 230 3375 + 60 3395 + 60 3415 + 230 2200 + 45 2463 + 70 2010 + 220 3096 + 245 3748 + 249 5790 + 80 5490 + 220 3460 + 70 3220 + 190 3108 + 215 2800 + 70
d18O
(‰ PDB) 29.13
28.44
27.71
28.16
27.69 27.85
27.90 28.10
26.25 26.00
28.14 27.40
27.99 28.36 27.69
27.59 28.11 28.12
28.19
28.03
26.53 26.93 27.52
27.65 27.88 27.96
*The samples 1, 3, 4 and 12 are recent tufas. Elements are in ppm, MgCO3 values (in mol%) were calculated from Mg and Ca data assuming these are the only cations in the carbonate fraction. 14C activity was determined with LSC method with 2 sample preparation procedures (B- benzene synthesis and D- CO2 absorption method). 14C results were expressed as 14C activity in percent of modern carbon (pMC) and as 14 C age in yr BP (not corrected for the initial 14C activity Ao).
¨ ZKUL ET AL. M. O
1
11 12 Average
14
MgCO3
PERCHED SPRINGLINE TUFAS, SW TURKEY
spring, 1.6–4.6 8C in autumn and 4.8 8C in winter with respect to the splash zone. These variations in water temperature are dependent on seasonal air temperature. In the city centre of Denizli according to long-term meteorological data, average air temperatures are 5.9 8C in January, 14.5 8C in April, 27.5 8C in July and 11 8C in November. The pH value increases in downflow direction, from 7.2–7.4 at the spring orifice (measurements at Sp-4), to 7.7 at waterfall head (point GW-1 in Fig. 6), and to 8.0 at the waterfall bottom (point GW-2) as a result of CO2-degassing. The biggest pH variation has occured during the flow from top to bottom (7.7–8.1) of the waterfall in winter season (Table 1). Similarly, the HCO3 value also decreases rapidly downflow, from Sp-4 to GW-3 (see Fig. 6). The CO2 value in the water of Sp-4 is higher (53–65 ppm) in winter and spring seasons. It is lower (6–15 ppm) in summer and autumn seasons. Maximum CO2 – degassing from the water occured in January. The calcium content of the Sp-4 was almost constant with 64 ppm in five sampling period. The Ca values decrease sharply in the splash zone. The Mg ratio ranged from 7.8– 12.7 (Table 1). Mg/Ca molar ratios of the Sp-4 are lower in wet season compared with dry period because of dilution. According to hydrochemical analysis, the dominant ions in the spring waters are calcium and bicarbonate (Table 1) and consequently, all spring waters in the tufa site are type of Ca–HCO3. The calcite saturation indices (SIC) in the spring orifice are within borderline scale potential (20.02– 0.06); however, they increase rapidly along the flow path because of CO2-degassing and the waters are supersaturated with respect to calcite at splash zone of the waterfall (Table 1). These values reach up to 0.6 at the waterfall bottom. The SIC values are highest in the summer season and lowest during the winter. Similarly, the pCO2 and DIC values of the spring water change naturally like the SIC values. The CO2 partial pressure of the water in GW-3 location is close to the atmospheric pressure. The tritium value of the spring Sp-1 and Sp-4 is 5.9 TU. These isotopic data indicate that the spring waters are of meteoric origin and ascend to the surface following shallow circulation. The d13C in dissolved inorganic carbon (DIC) of the Sp-4 is 23.71‰, indicating that the inorganic carbon was derived mainly by dissolution of marine carbonate.
Mineralogical, chemical and biological composition of the tufa deposits Mineralogical composition Tufa samples are mainly composed of calcite (Fig. 7) with minor amound of quartz and feldspar.
255
The calcite is mostly micrite and microspar (5– 20 m) in size. The microspar calcite occurred as cement readily distinguishable from the micritic groundmass, and commonly fills pores (Fig. 7e–g). Micropores on some of the crystal surfaces are possibly from dissolution of inclusions (perhaps loss of organic fragments) (Fig. 7h). The quartz grains, which are scattered throughout the micritic ground and commonly single quartz grains in silt size, probably derived from the schists around the tufa site.
Biological composition Apart from hydrophytic macrophytes, the tufa deposits also include bryophytes, cyanobacterial filaments, and diatoms. Bryophytes dominate in the waterfall area. The diatom frustules were determined as species Cymbella sp. and Synedra sp. (Figs 7b & 8e, f –h). In some cases, pulmonate gastropods were observed occasionally within the fine grained detrital tufa of the lower slope deposits.
Chemical composition The eleven recent and old tufa samples were taken from different parts of the site. The results of major and trace element analyses of the samples were given in Table 2. The Mg content ranged from 1380 –7440 with an average of 3235 ppm. The MgCO3 of the tufa deposits ranges from 0.59 –3.40 mol% with an average of 1.43 mol%. According to these values, the calcite in the tufa is clearly low magnesian calcite. The Mg contents of the tufa samples are more or less similar with the Urrea de Jalo˜n tufa deposit, Ebro basin, NE Spain (Arenas et al. 2000), whereas the higher values were reported from some travertine occurances in hiydrothermal origin (Minissale et al. 2002). Ba and Sr compositions of the tufa samples are 33 – 192 ppm and 222–322 ppm, respectively. The Ba and Sr contents are similar or slightly higher than those in fluvio-lacustrine tufas in the central Ebro Depression, NE Spain (Arenas et al. 2000), while Ba content in Quaternary tufa stromatolite from central Greece is also in similar value (Andrews & Brasier 2005).
Radiocarbon and stable-isotope composition Preliminary results of 14C and stable-isotope (d13C and d18O) composition of these tufa deposits were reported in Horvatincˇic´ et al. (2005). The 14C activity of recent tufa samples in the Gu¨ney springline tufa site are 60 –70% of modern carbon (pMC) (Table 2, sample 1, 3, 4 and 12). This Ao value is
256
¨ ZKUL ET AL. M. O
Fig. 7. SEM images of mineral precipitates from the Gu¨ney perched springline tufa site. (a) Surface of rough and porous tufa deposits; (b) Tufa deposits in (a) associated with rod like microbial components (arrowed); (c) Bundles of calcite crystals with blade and spearhead form; (d) Euhedral microcrystalline calcite; (e) Calcite rhombs; (f) Secondary calcite crystals in blade shape developed in a pore; (g) Close view of calcite rhombs in (e); (h) Calcite crystals with smooth surfaces and micropores.
PERCHED SPRINGLINE TUFAS, SW TURKEY
257
Fig. 8. SEM images of microbial components from the tufa deposits. (a) Longitudinal, fibrous microbial component and adhered calcite crystals on its surface; (b) A network of microbial filaments and calcified extracellular polymeric substances (EPS); (c) Curved leaf (between arrows) and diatoms in the middle; (d) Longitudinal cellular bryophyte tubes; (e) Lenticular diatom frustules (Cymbella sp.), some diatoms deformed and broken; (f) Rod-like diatom, Synedra sp.; (g) Synedra sp. attached on calcite rhombs; (h) A close view of the diatom Synedra sp. in (g).
258
¨ ZKUL ET AL. M. O
typical for carbonates precipitated from the fresh water system where 14C activity of dissolved inorganic carbon (DIC) is result of mixture of biogenic/atmospheric CO2 and dissolved mineral carbonates, for example, limestone. The d13C and d18O values of active and passive tufa samples are in the range from 29.13 to 26.0‰, and from 28.44 to 27.40‰, respectively (Table 2). More negative d13C values of the Gu¨ney site indicate that most of the 14C is of atmospheric CO2 and/or biogenic origin.
Discussion The prominent features of the studied site were compared with some well-known tufa sites throughout Europe in Table 3. The Gu¨ney waterfall site is composed of coalescent tufa bodies and includes an active waterfall area. The two distinct lobe-top terrace areas took place above the active waterfall. Horizontal– subhorizontal lobe-top area is an indication of mature stage of perched springline tufa deposits (Pedley 1990; Ford & Pedley 1996; Pedley et al. 2003). Consequently, some of the tufa lobes have reached a mature stage. In the study area, the flat terraces possibly developed in consequence of prograding lobe fronts and issue of antecedent topography. A self-built channel has been developed at one of the lobe-top terraces. This kind of channels was observed at thermal areas in Pamukkale and surroundings in the Denizli basin, western Turkey ¨ zkul et al. (Bean 1971; Altunel & Hancock 1993; O 2002). Similar channels have been mentioned from another perched springline tufa site (i.e. Rochetta a Volturno, Italy; Violante et al. 1994), but not seen in the Spanish examples (Pedley et al. 2003). The inactive Holocene tufa samples are from 2000–5800 yr BP old, based on the 14C dating method (Horvatincˇic´ et al. 2005). Age relationships between coalescent tufa bodies in perched springline sites are quite complex (Martı´n-Algarra et al. 2003; Pedley et al. 2003). In the study area, the present multi-spring resurgences and individual tufa lobes coalesced occur at various elevations. Thus, a regular downslope trend in ages is not expected. One of the reasons of this situation may be heterogenity within the aquifer rocks. Earthquake activity in the region (Kumsar et al. 2008, p. 95 –98) may have been changed the settings of the spring resurgences through time. In some Spanish sites, regular trends have been recorded in downslope directions. Gonza´les Martı´n et al. (1989), for example, indicated that the higest deposits at Tajun˜a sites, NE of Madrid, are no longer active, whereas those in the valley bottom are still developing. This suggests a decrease in age of individual
mounds in a downslope direction. This is not always the case, hovever, as new tufa lobes can develop above older sites of deposition (Pedley et al. 2003). Furthermore, Martı´n-Algarra et al. (2003) pointed out that well-defined tufa steps in Granada basin (south Spain) were deposited by springs that migrated downslope in time. Tufa-depositing spring waters have distinct chemical compositions. Most waters emerging from karstic massifs commonly are of Ca–HCO3 and Ca –HCO3 –SO4 types and supersaturated with respect to calcite (Andreo et al. 1999; Horvatincˇic´ et al. 2005). In some areas, however, the chemical compositions of karstic waters may be slightly different as in the Honaz site, Denizli, Turkey (Horvatincˇic´ et al. 2005). In addition, Pentecost (2000) assumed that the Matlock Bath deposit and associated spring waters, Derbyshire, UK were ‘thermogene travertine’ and ‘thermal water’, respectively, depending on chemical compositions. The Matlock Bath waters are really deep-cycled meteoric waters, which are only a few degrees above ambient temperature (pers. comm., Pedley 2009). The Matlock Bath site is a good example of a perched springline tufa model in aspect of its formation and morphology and fits his classification (Pedley 1990) very well. At the Gu¨ney site, the hydrochemical parameters of the tufa-precipitating waters were monitored seasonally for more than one year (Table 1). Downflow, the pH, SIC, Mg/Ca values increased whereas the CO2, EC, Ca, Mg, HCO3, and DIC decreased. The Na, K, Cl and SO4 values did not change during the study period. The DIC concentrations were higher in January (winter season) and lower in November such as in the recent freshwater carbonates of River Krka, Croatia (Lojen et al. 2004). Apart from the Gu¨ney site, there are two other tufa sites in the Denizli basin, one at Sakızcılar and the other at Honaz. The tufa deposits and spring waters of Gu¨ney and Sakızcılar sites show similar characteristics in water chemistry. In contrast, the water and tufa chemistry at the Honaz site, which is located along a normal fault zone (Bozkus¸ et al. 2000), is different. The water temperature, electrical conductivity, dissolved anion and cation values of the Honaz site are higher than the other two tufa sites because of longer residence time. Waterfall tufa area is one of the most important components of perched springline and fluvial models as investigated in some European sites previously. Also this area takes place in front of tufa barrage downstream in a fluvial model and passes into lake deposits upstream. The most rapid tufa precipitation occurs in waterfall front of a perched springline tufa site (Pedley et al. 1996; Pentecost 2000; Pedley et al. 2003; Carthew et al. 2003; Andrews 2006; Pedley 2009).
Table 3. Comparative features of some European perched springline tufa sites well-documented Violanta et al. 1994, Rochetta a Volturno, central Italy
Pedley et al. 2003, Rio Tajun˜a valley, central Spain
Martin-Algarra et al. 2003, Granada basin, central Spain
This study, Gu¨ney S¸elalesi, Denizli, SW Turkey
Temperate Rainfall 1000 mm pa* 100– 130 m Multi-spring resurgences, water temp. close to 20 8C Ca-Mg-HCO3 type waters Western lower slope of the River Derwent with a few cascades
Semi-arid
Semi-arid Rainfall ,490 mm pa* 750– 800 m Multi-spring resurgences
Semi-arid Rainfall ,490 mm pa* 950– 1120 m Multi-spring resurgences, migrating springs downslope in time
Semi-arid Rainfall 548 mm pa, ann mean air temp. 13.5 8C 200– 500 m Multi-spring resurgences, spring water temp. ¼ 18.7 8C, Ca– Mg–HCO3 type water
Four well-defined steps succeeding each other in downslope direction
Lobate or fan-shaped mounds in plan, lenticular, wedge-shaped in profile, flat top (mature stage)
Lob top surface
Suspended channels
Flat upper surface, Abundant oncoid up to gravel size
Areal extent
6 hectars
Terraced, flat upper surface associated with suspended channel (catwalk) and very shallow lakes 1500 hectars
150 hectars (1.5 km2)
Flat upper surface (maturity stage) associated with suspended channel, stand of trees, used as agricultural area 20 hectars
Primary cavities
Present
Mature deposits are characterized by lenticular to wedge-shaped, downslope-facing lobate profiles Gently convex or flat upper surface in mature stage. Organic-rich and oncoidal deposits From a few square metres up to hundreds of hectars Abundant, speleothem-lined
Decimeter to meter-size voids partially speleothem-lined
Abundant cavities up to cave size with speleothem-lined
Flat-topped stromotolite heads below bryophyte curtain Moss (mainly bryophyte), some grasses and trees
Thin, planar to undulose stromotolite crust Decimeter-sized moss mound, patches of canes, leaves of Salix sp., Quercus sp., and herbaceous veget. Gastropod
Thin, undulose stromotolite crust Some grasses and trees, moss (mainly bryophyte), diatom Gastropod
Common as calcified plant detritus on distal slope
Common on distal slopes and interlobe areas
Tufa steps get younger downslope with time
No regular age trend downslope
Climate Altitude Spring
Morphology
Stromotolite Flora
Bryophyte (moss), diatom, mollusca, ostracoda, pollen
450 –550 m Ambient temperature or low thermal waters
Lobe-shaped configirations and minor terraces are visible
Large cavities and several caves, speleothem-lined cave walls Stromotolites encrusring travertine intraclasts Moss (bryophyte), cyanobacteria
Fauna Tufa breccia, palaeosol Age relationship
No regular age trend downslope
259
*Rainfall data taken from Viles et al. 2007.
Filled up erosional channels with intraclasts on upper slope associated with palaeosols
Pulmonate gastropods usually dominant in distal slope facies by Detrital tufas associated with palaeosols
PERCHED SPRINGLINE TUFAS, SW TURKEY
Pentecost 2000; Pedley et al. 2003 Matlock Bath, Derbyshire, UK
260
¨ ZKUL ET AL. M. O
Fig. 9. Regional setting of the Gu¨ney springline tufa deposits base on the d18O and d13C stable isotope data plotted on figure 3 of Andrews (2006). The tufa deposits in this study correspond to ‘wooded mountain tufa’. See for details Andrews (2006).
Stable isotope signatures in tufas have useful environmental and climatic information (Andrews et al. 1997). Based on stable isotope distribution (Fig. 9), the Gu¨ney springline tufa deposits, which
are located below 500 m in altitude, correspond to the ‘wooded mountain tufa’ of Andrews (2006). In the same time, the distribution of stable isotope data of the tufa deposits show similarity with Dinaric
Fig. 10. Continental setting of stable isotope data of the Gu¨ney perched springline tufa deposits plotted on Figure 7 of Andrews (2006). The stable isotopes (d18O and d13C) plotting of the tufas show similarity with Dinaric karst, Czech Republic and Spanish precipitates in European scale.
PERCHED SPRINGLINE TUFAS, SW TURKEY
karst (Horvatincˇic´ et al. 2003), Czech Republic and Spanish precipitates in European scale (Fig. 10).
Conclusions The Gu¨ney tufa site is a typical representative of perched springline tufa model in eastern Mediterranean. The site with lobe-top terraces partly reached in mature stage. Numerous cavities have developed behind of the tufa lobes; some of them reached more than 10 m in size. In the Gu¨ney site, tufa deposition began at least 5800 yr BP. The tufa-precipitating waters supersaturated in calcite outflow from multi-spring points and are of Ca –HCO3 in origin. According to hydrochemical and isotopic composition, the site waters are clearly meteoric and shallow circulated. Radiocarbon and stable-isotope composition of the tufa deposits support meteoric origin. Additionally, the stableisotope values show similar distribution with south European countries. This work was supported by the Scientific Research Grants of Pamukkale University, Denizli, Turkey (Project grant: 2003MHF012) and the Ministry of Science and Technology of the Republic of Croatia (Project grant: 05MP014). The authors thank Arzu Gu¨l, Melek Tin and Salih Kıyak for their assistance. We are grateful to Brian Jones for his review of the SEM images, manuscript and linguistics. We thank also Bernie Owens for identifcation of diatoms. The governorship of the city of Denizli provided logistic support for monitoring.
References ¨ ZKUL , M. 2007. SedimenA LC¸ I˙ C¸ EK , H., V AROL , B. & O tary facies, depositional environments and palaeogeographic evolution of the Neogene Denizli Basin of SW Anatolia, Turkey. Sedimentary Geology, 202, 596–637. A LTUNEL , E. & H ANCOCK , P. L. 1993. Morphology and structural setting of Quaternary Travertines at Pamukkale, Turkey. Geological Journal, 28, 335–346. A NDREO , B., M ARTI´ N -M ARTI´ N , M. & M ARTI´ N A LGARRA , A. 1999. Hydrochemistry of spring water associated with travertines. Examples of the Sierra de la Alfaguara (Granada, southern Spain). Comptes Rendus de l’Academie des Sciences Paris, Sciences de la Terre et des Plane`tes, 328, 745– 750. A NDREWS , J. E. 2006. Paleoclimatic record from stable isotopes in riverine tufas: synthesis and review. EarthScience Review, 75, 85–104. A NDREWS , J. E. & B RASIER , A. T. 2005. Seasonal records of climatic change in annually laminated tufas: short review and future prospects. Journal of Quaternary Science, 20, 411–421. A NDREWS , J. E., R IDING , R. & D ENNIS , P. F. 1997. The stable isotope record of environmental and climatic signals in modern terrestrial microbial carbonates from Europe. Palaeogeography, Palaeoclimatology, Palaeoecology, 129, 171 –189.
261
A NZALONE , E., F ERRERI , V., S PROVIERI , M. & D’A RGENIO , B. 2007. Travertines as hydrologic archives: The case of the Pontecagnano deposits (southern Italy). Advances in Water Resources, 30, 2159– 2175. A RENAS , C., G UTIERREZ , F., O SA´ CAR , C. & S ANCHO , C. 2000. Sedimentology and geochemistry of fluviolacustrine tufa deposits controlled by evaporate solution subsidence in the central Ebro Depression, NE Spain. Sedimentology, 47, 883 –909. A RP , G. 1995. Lacustrine bioherms, spring mounds, marginal carbonates of the Ries-impact-crater (Miocene, southern Germany). Facies, 33, 35– 90. B EAN , G. E. 1971. Turkey beyond the Meander. Ernest Benn, London (Revised and edited by J. Bean [1980] and published by John Murray, London [1989]). B ONNY , S. & J ONES , B. 2003. Relict tufa at Miette hot springs, Jasper National Park, Alberta, Canada. Canadian Journal of Earth Science, 40, 1459– 1481. B OZKURT , E. & O BERHA¨ NSLI , R. 2001. Menderes masif (Western Turkey): structural, metamorphic and magmatic evolution – a synthesis. International Earth Sciences, 89, 679–708. ¨ ZKUL , M. & H ANC¸ ER , M. B OZKUS¸ , C., K UMSAR , H., O 2000. Seismicity of Active Honaz fault under an exten¨ ZGENC¸ , ˙I. & sional tectonic regime. In: D ORA , O., O S O¨ ZBILIR , H. (eds) International Earth Science Colloquium on the Aegean Region. Proceedings of Dokuz Eylul University, Deptartment of Geology, I˙zmirTurkey, 7 –16. C ARTHEW , K. D., T AYLOR , M. P. & D RYSDALE , R. N. 2003. Are current models of tufa sedimentary environments applicable to tropical systems? A case study from the Gregory River. Sedimentary Geology, 162, 199– 218. C EYLAN , M. A. 2000. The recreational importance of the Gu¨ney waterfall. Dog˘u Cog˘rafya Dergisi, 3, 61– 76. C HAFETZ , H. S. & F OLK , R. L. 1984. Travertines: depositional morphology and the bacterially constructed constituents. Journal of Sedimentary Petrology, 54, 289– 316. C OLE , J. M., R ASBURY , E., T ROY , M. I. P., P EDONE , V. A., L ANZIROTTI A NTONIO , H. & G ILBERT , N. 2004. Petrographic and trace element analysis of uraniumrich tufa calcite, middle Miocene Barstow Formation, California, USA. Sedimentology, 51, 433– 453. E RDOG˘ AN , B. & G U¨ NGO¨ R , T. 2004. The problem of the core-cover boundary of the Menderes massif and an emplacement mechanism for regionally extensive gneissic granites, western Anatolia (Turkey). Turkish Journal of Earth Sciences, 13, 15–36. E VANS , J. E. 1999. Recognation and implications of Eocene tufas and travertines in the Chadron Formation, White River Group, Badlands of South Dakota. Sedimentology, 46, 771–789. F ORD , T. D. & P EDLEY , H. M. 1996. A review of tufa and travertine deposits of the world. Earth Science Reviews, 41, 117– 175. G LOVER , C. P. & R OBERTSON , A. H. F. 2003. Origin of tufa (cool-water carbonate) and related terraces in the Antalya area, SW Turkey. Geological Journal, 38, 329– 358. G ONZA´ LEZ M ARTI´ N , J. A., G ARCI´ A DEL C URA , M. A. & O RDO´ N˜ EZ , S. 1989. Formaciones tobaceas en los
262
¨ ZKUL ET AL. M. O
valles Tajun˜a y Tajo, Excursion C-4. Reunion del Cuaternario Iberico, Madrid 25– 29 September 1989. Association of Espan˜ola Estudio del Cuaternario (AEQUA) – Grupo Trabalho Portugies Estudodo Quaternario (GTPEQ). G O¨ KGO¨ Z , A. 1998. Geochemistry of the KızıldereTekkehamam-Buldan-Pamukkale Geothermal Fields. United Nations University, Geothermal Training Programme, Reports, 115–156. Reykjavik, Iceland. G UO , L. & R IDING , R. 1998. Hot-spring travertine facies and sequences, Late Pleistocene Rapolano Terme, Italy. Sedimentology, 45, 163– 180. H ORVATINCˇ IC´ , N., Cˇ ALIC´ , R. & G EYH , M. 2000. Interglacial growth of tufa in Croatia. Quaternary Research, 53, 185–195. H ORVATINCˇ IC´ , N., K RAJCAR , B RONIC´ , I. & O BELIC´ , B. 2003. Differences in the 14C age, 13C and 18O of Holocene tufa and speleothem in the Dinaric Karst. Palaeogeography, Palaeoclimatology, Palaeoecology, 193, 139– 157. H ORVATINCˇ IC´ , N., B ARESˇ IC´ , J., K RAJCAR B RONIC´ , I. & O BELIC´ , B. 2004. Measurement of low 14C activities in liquid scintillation counter in the Zagreb Radiocarbon Laboratory. Radiocarbon, 46, 105–116. ¨ ZKUL , M., G O¨ KGO¨ Z , A. & H ORVATINCˇ IC´ , N., O B ARESˇ IC´ , J. 2005. Isotopic and geochemical investigation of tufa in Denizli province, Turkey. In: ¨ ZKUL , M., Y AG˘ IZ , S. & J ONES , B. (eds) Proceedings O of International Symposium on Travertine, Sedimentology Session. 21– 25 September 2005, Denizli, Tu¨rkiye, 162– 170. K AYMAKC¸ I , N. 2006. Kinematic development and paleostress analysis of the Denizli Basin (Western Turkey): implications of spatial varition of relative paleostress magnitudes and orientations. Journal of Asian Earth Sciences, 27, 207–222. K OC¸ YIG˘ IT , A. 2005. The Denizli graben-horst system and the eastern limit of western Anatolian continental extension: basin-fill, structure, deformational mode, throw amount and episodic evolutionary history, SW Turkey. Geodinamica Acta, 18, 167– 208. ¨ ., T ANO , H. & C ¸ ELIK , S. B. 2008. K UMSAR , H., A YDAN , O An investigation of Buldan (Denizli) earthquakes of July 23– 26, 2003. Ekin Yayınevi, Bursa. [In Turkish with English abstract.] L OJEN , S., D OLENEC , T., V OKAL , B., C UKROV , N., M IHELCˇ IC´ , G. & P APESCH , W. 2004. C and O stable isotope variability in recent freshwater carbonates (River Krka, Croatia). Sedimentology, 51, 361–375. M ARTI´ N -A LGARRA , A., M ARTI´ N -M ARTI´ N , M., A NDREO , B., J ULIA , R. & G ONZALEZ -G OMEZ , C. 2003. Sedimentary patterns in perched spring travertines near Granada (Spain) as indicators of the
paleohydrological and palaeoclimatical evolution of a karst massif. Sedimentary Geology, 161, 217– 228. M INISSALE , A., K ERRICK , D. M. ET AL . 2002. Geochemistry of Quaternary travertines in the region north of Rome (Italy): structural, hydrologic and paleoclimatologic implications. Earth and Planetary Science Letters, 203, 709– 728. ¨ ZKUL , M., V AROL , B. & A LC¸ I˙ C¸ EK , M. C. 2002. DeposiO tional environments and petrography of the Denizli travertines. Bulletin of the Mineral Research and Exploration, 125, 13–29. P ARKHURST , D. L. & A PPELO , C. A. J. 1999. User’s guide to PHREEQC (version 2) – A computer program for speciation, batchreaction, one-dimensional transport, and inverse geochemical calculations. US Geological Survey Water-Resources Investigations, Report 99– 4259. P EDLEY , H. M. 1990. Classification and environmental models of cool freshwater tufas. Sedimentary Geology, 68, 143 –154. P EDLEY , H. M. 2009. Tufas and travertines of the Mediterranean region: a testing ground for freshwater carbonate concepts and developments. Sedimentology, 56, 221–246. P EDLEY , H. M., O RDO´ N˜ EZ , S., G ONZA´ LEZ M ARTI´ N , J. A. & G ARCI´ A DEL C URA , M. A. 1996. Does climate control the morphological fabric of freshwater carbonates? A comparative study of Holocene barrage tufas from Spain and Britain. Palaeogeography, Palaeoclimatology, Palaeoecology, 121, 239– 257. P EDLEY , H. M., O RDONEZ , S., G ONZALES -M ARTIN , J. A. & G ARCIA DEL C URA , M. A. 2003. Sedimentology of Quaternary perched springline and paludal tufas: criteria for recognition, with examples from Guadalajara Province, Spain. Sedimentology, 50, 23–44. P ENTECOST , A. 2000. The origin and development of the travertines and associated thermal waters at Matlock Bath, Derbyshire. Proceedings of the Geologists’ Association, 110, 217– 232. S¸ ˙I MS¸ EK , S¸. 2003. Hydrogeological and isotopic survey of geothermal fields in the Bu¨yu¨k Menderes graben, Turkey. Geothermics, 32, 669– 678. S¸ I˙ MS¸ EK , S¸., G U¨ NAY , G., E LHATIP , H. & E KMEKCI , M. 2000. Environmental protection of geothermal waters and travertines at Pamukkale, Turkey. Geothermics, 29, 557– 572. V IOLANTE , C., F ERRERI , V., D’A RGENIO , B. & G OLUBIC´ , S. 1994. Quaternary travertines at Rochetta a Volturno (Isernia, Central Italy). Facies analysis and sedimentary model of an organogenic carbonate system. PreMeeting Fieldtrip Guidebook, A1. International Association of Sedimentologist, Ischia ’94, 15th regional meeting, Italy, 3– 23.
Calcareous tufa as indicators of climatic variability: a case study from southern Tuscany (Italy) ENRICO CAPEZZUOLI*, ANNA GANDIN & FABIO SANDRELLI Dipartimento di Scienze della Terra, University of Siena, Via Laterina 8, 53100 Siena, Italy *Corresponding author (e-mail:
[email protected]) Abstract: A carbonate terraced succession mainly consisting of fluvial/palustrine calcareous tufa and of lacustrine limestone was deposited during recent Quaternary in a limited segment of the Valdelsa fluvial pattern (southern Tuscany, Italy). The radiometric data obtained from three carbonate terraces indicate that the depositional/ erosional history of the Valdelsa succession during Late Pleistocene– Holocene, has been constrained by the same cyclic events observed in coeval detrital lacustrine successions of Central Italy. At least three of the Valdelsa carbonate Synthems and the interposed erosional phases can be correlated with the major climatic changes recognized in the European– Mediterranean area, from the Last Glacial Interstadial through the Younger Dryas to the Atlantic ‘Optimum Climatic’, the Sub-Boreal and finally at 2.5 ka the last Sub-Atlantic oscillation. This climatic correlation and the radiometric data imply that the deposition of calcareous tufa in Valdelsa was mainly dependent on rainfall availability and, consequently, was active during the milder oscillations within the cold periods.
Interest in the climatic oscillations on the Earth, and the extent to which humankind is changing the biosphere, has never been as great as at present since climate controls all the surface processes on our planet including the formation of sediments/rocks (Pentecost 2005). So far, acquisition of data on the climatic conditions of recent terrestrial deposits has been generally performed in mainly terrigenous, lacustrine sediments (Digerfeldt et al. 1997; Magri 1999; Giraudi 2000, 2001; Magny et al. 2002, 2006; Sadori et al. 2004; Wulf et al. 2004; DrescherSchneider et al. 2007). Actually the study of past lake-level changes provides an excellent proxy for palaeoclimatic reconstructions because of: (i) a potentially continuous record covering a time span from the present day back through several glacial/ interglacial cycles; (ii) a high temporal resolution due to high sedimentation rates and the preservation of annual laminations; and (iii) considerable potential for tephrochronological investigation, enabled by contemporary volcanic activity (Wulf et al. 2004). Fluvial deposits have a lower potential for preservation owing to the high-erosive potential of the flowing waters and high erodibility of the associated detrital deposits. However, terraced, fluviatile carbonate deposits appears to be particularly promising for the understanding of the environmental and palaeoclimatic evolution of the area, since the morphology of carbonate terraces and their depositional features are generally rather well-preserved due to early lithification of the carbonate sediments
(Ordo`n˜ez et al. 2005; Capezzuoli & Sandrelli 2006; Ortiz et al. 2009). Terrestrial carbonates comprise a number of deposits formed on emerged land in a variety of depositional settings where old carbonate rocks are exposed. Their perceived importance has greatly increased since it has been revealed that they are particularly sensitive indicators of past environmental conditions. Recent studies, in fact, have demonstrated that extended and detailed palaeoclimatic records can be obtained from the sedimentological/stratigraphic and geochemical analysis of speleothems, calcrete and lacustrine limestone (Henning et al. 1983; Livnat & Kronfeld 1985; Bar-Matthews et al. 1997; Von Grafenstein et al. 2000; Leng & Marshall 2004; McDermott 2004; Fleitmann et al. 2004; Drysdale et al. 2004; Dworkin et al. 2005). Comparable information is now becoming available from the study of travertines and tufas precipitated from calciumbicarbonate-rich stream waters flowing under subaerial conditions, from thermal or karstic springs (Andrews 2006; Andrews et al. 1993, 1994, 1997, 2000; Frank et al. 2000; Makhnach et al. 2000, 2004; Matsuoka et al. 2001; Soligo et al. 2002; Ihlenfeld et al. 2003; Garnett et al. 2004; Smith et al. 2004; Andrews & Brasier 2005). The geomorphological evolution of a river valley can also reflect the main palaeoclimatic variations. In particular the down cutting of a series of terraces is generally related to major climate phases in the region. It has been noted that many
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 263–281. DOI: 10.1144/SP336.14 0305-8719/10/$15.00 # The Geological Society of London 2010.
264
E. CAPEZZUOLI ET AL.
rivers did form new terraces during warm periods or cold-to-warm transitions. In contrast, they seemed to cut down and deepen their valleys following interglacial periods, which reflect responses driven by climate change, mainly at orbital (Milankovitch) frequencies (Bridgland et al. 2004a, b; Bridgland & Westaway 2008; Ortiz et al. 2009). Consequently, terraced carbonate deposits hold the potential to provide significant palaeoclimatic data in karstic regions and, when preserved they retain more accessible genetic information compared with the poorly lithified detrital terraced deposits that, being easily eroded and reworked by weathering, pedogenic processes and landslides, do not retain unequivocal evidence of their geomorphologic evolution. An interesting case study on this subject is offered by the results of the geomorphological, sedimentological investigation and dating of calcareous tufa terraced deposits exposed in Elsa Valley (Southern Tuscany Italy). Aim of this study was to demonstrate that among subaerial continental deposits, not only detrital lacustrine deposits, but also terraced fluvial carbonates/calcareous tufa can be considered reliable registers of climate variations, since their development seems to be mainly controlled by water availability more than ambient temperature. In this paper the four depositional and downcutting events recognized and dated in the calcareous bodies of Valdelsa terraced succession are reported and correlated with the major climate changes occurred during the Late Quaternary glacial period in the European/Mediterranean area and in particular in Central Italy (Giraudi 2000, 2001, 2004; Narcisi 2001; Sadori et al. 2004; Magny et al. 2006; Drescher-Schneider et al. 2007).
Geological setting The southern Elsa River Valley, a NW–SE tectonic depression known as Valdelsa Basin (Bossio et al. 1995b), filled up by late Miocene and Pliocene continental and marine sediments (up to 2000 m; Ghelardoni et al. 1968) is bounded by the MiddleTuscan Ridge to the West and South and by the Chianti Ridge to the East (Bossio et al. 2000) (Fig. 1). While the Chianti Ridge is mainly composed of detrital –carbonate flysch successions pertaining to the Late Cretaceous –Eocene Ligurian tectonic units, the Middle-Tuscan Ridge is composed of formations of the Tuscan Nappe resting on the metamorphic Middle Triassic Verrucano red beds through the ‘Calcare cavernoso’, a karstified tectonic breccia (Rauhwacke) derived from Late Triassic evaporites (Gandin et al. 2000). It constitutes the carbonate rich, catchments basin of the main aquifer that supplies the base flow of the
Elsa River (Canuti & Tacconi 1975; Barazzuoli et al. 2002) which is the main tributary of the Arno River. The southern portion of the Valdelsa Basin is characterized by a complex Quaternary succession of subaerial deposits (Capezzuoli & Sandrelli 2004; Capezzuoli et al. 2007, 2008) made of mainly carbonate, unconformity-bounded, units that can be referred to Synthems that is, the basic unconformity-bounded unit of ‘Unconformable Boundary Stratigraphic Units’ [(UBSU); ISSC 1987; Salvador 1994]. The oldest units, the Early–Middle Pleistocene Strove (STR) and Campiglia dei Foci (CDF) Synthems (Capezzuoli et al. 2007), forming subhorizontal plateaux that characterize the local landscape (Fig. 2a), represent two consecutive episodes of a wide lacustrine–paludal environment. The general uplifting of the region marked the end of the Early–Middle Pleistocene lacustrine sedimentation that was followed by the progressive incision and selective erosion of the carbonate plateaux down to the underlying Pliocene sands, and resulted in an inverted relief in this area (Bartolini & Peccerillo 2002) and a new hydrographical pattern. Along this renewed fluvial network, four subsequent fluvial–palustrine episodes gave rise to the sedimentation in the river valleys, of four unconformity-bounded units composed of detrital, alluvial materials (sands, silts and pebbles) locally replaced by fresh water carbonates. The four synthems (Fig. 2b): Abbadia (ABB), Calcinaia (CAL), Torrente Foci (FOC) and Bellavista (BEL), are represented by terraces located on the slopes of the main valleys of the area (Capezzuoli & Sandrelli 2004), that can be easily identified in the landscape on the basis of their elevation above the present thalweg position (þ20–30 m/ ABB, þ12 –25 m/CAL, þ8–15 m/FOC, þ2– 8 m/BEL). The younger alluvial deposits locally well-developed in the valley floors of several rivers, form the present Poggibonsi Synthem (POG). In discrete/specific segments of Elsa River and its tributary valleys, the detrital alluvial deposits of the four terraced systems are replaced by carbonates in an area of over 50 Km2 (Fig. 1). Tufa bodies have been deposited in tracts of Elsa River (between Gracciano and Poggibonsi) and Staggia Creek (between Castellina Scalo and Poggibonsi) about 8–9 km long, while their extension along Foci Creek (between San Gimignano and Campiglia dei Foci) and Imbotroni Creek is shorter, no more than 4 km long. Reduced calcareous crusts are currently forming locally, around few active small springs of low thermal water (Capezzuoli et al. 2007). The four calcareous bodies, previously described as ‘recent
CALCAREOUS TUFA AS INDICATORS OF CLIMATIC VARIABILITY 265
Fig. 1. (a) Location of the Valdelsa area and of the Late Quaternary–Holocene lacustrine successions of Central Apennine cited in the text. (b) Geological scheme of the southern Valdelsa Basin.
266
E. CAPEZZUOLI ET AL.
Fig. 2. (a) Schematic reconstruction of the Quaternary stratigraphic units in southern Valdelsa (not in scale), after Capezzuoli et al. 2007, slightly modified. (b) View of the Torrente Foci valley, east of Campiglia dei Foci, showing some of the older surfaces of the Quaternary synthems.
travertines’ of thermal origin and referred to the Holocene (Merla & Bortolotti 1967), were later related to calcareous tufa and assigned to the Late Pleistocene –Holocene by Capezzuoli & Sandrelli (2004).
Materials and methods The geological/geomorphological survey on the field, supported by aerial photographs resulted in the reconstruction of the stratigraphic succession of the fluvial terraces as well as the distribution and geometry of the carbonate bodies. The study of the internal structures and facies associations has been carried out on outcrops and road cuts and refined with the help of the sedimentological and petrographic analysis. Epigean ‘flowstones’, described in many works under the general term ‘travertine’ (Buccino et al. 1978; Julia` 1983; D’Argenio et al. 1983; Chafetz & Folk 1984; Ferreri 1985; Brancaccio et al. 1986; D’Argenio & Ferreri 1988; Lang et al. 1992; Golubic et al. 1993; Demicco & Hardie 1994; Violante et al. 1994, 1996; Pentecost & Whitton 2000; Martı`nAlgarra et al. 2003) or ‘thermogene/meteogene travertine’ (Pentecost & Viles 1994; Pentecost 1995, 2005), today are commonly separated into travertine and calcareous tufa on the basis of their constructional, petrological, geochemical and isotopic characteristics; all of which reflect the contrasting conditions of deposition related to the physico-chemical proprieties of the parent waters respectively derived by an hydrothermal system or from a fluvial/paludal system fed by cool karstic springs (Pedley 1990, 2009; Riding 1991; Ford & Pedley 1996; Guo & Riding 1998; Pedley et al. 2003; Glover & Robertson 2003; Capezzuoli & Gandin 2004). According to the results of recent research on active epigean systems (Pedley 1990, 2009; Ford & Pedley 1996; Gandin & Capezzuoli 2008), the term travertine applies to a compact, well-bedded and laminated limestone, forming
wedge-shaped bodies in the surroundings of hydrothermal spring systems. It is characterized by dominantly crystalline fabrics and minor biotic contribution which reflects the lethal effect of temperature and sulphides on the life of macrophytes (grasses and trees) and most of the microphytes (algae and bryophytes). Conversely, calcareous tufa corresponds to a highly porous, locally chalky, poorly bedded limestone forming irregularly staked lenticular bodies of massive phytohermal or stromatolitic buildups deposited by ambient temperature waters in fluvial/palustrine systems (Pedley 1990, 2009 with references therein; Riding 1991; Ford & Pedley 1996; Pedley et al. 2003; Glover & Robertson 2003; Capezzuoli & Gandin 2004). The luxuriant vegetation of emergent as well as aquatic plants promoted by the abundant water availability and release of CO2, mostly act in a passive role of support of the microcrystalline precipitate and rapidly decays upon death, leaving characteristically porous phytohermal or phytoclastic structures. Cool water tufa and thermal travertine are not always mutually exclusive in occurrence, since in the distal end of the thermal-spring systems cooled waters merge with surface rainwater or normal-water streams, forming pools where transitional paludal-like tufas are formed (Ford & Pedley 1996; Capezzuoli & Gandin 2005; Capezzuoli et al. 2008; Pedley 2009). Terminology adopted to distinguish the tufa facies follows the more recent genetic classification which is based on the textural criteria first proposed by Buccino et al. (1978) (D’Argenio et al. 1981; Ordo`n˜ez & Garcia del Cura 1983; Ferreri 1985; Viles & Goudie 1990; Pentecost & Viles 1994; Violante et al. 1994, 1996; Pentecost 1995), and developed in terms of depositional conditions by Pedley (1990) and later reviewed by Ford & Pedley (1996) and Pedley (2009). Absolute dating has been employed in order to better constrain the age of the four carbonate bodies. Dating methods most commonly used to
CALCAREOUS TUFA AS INDICATORS OF CLIMATIC VARIABILITY
determine terrestrial carbonate ages are 14C and U/ Th (Henning et al. 1983; Livnat & Kronfeld 1985; Ordo`n˜ez et al. 1990; Arenas et al. 2000; Horvatincˇic et al. 2000; Garnett et al. 2004). However, the range of the radiocarbon method (c. 30– 40 ka) is a serious limitation and U/Th dating presents constraints linked to U –geochemistry (input/output) as well as to detrital thorium (232Th) presence and the method range. In an attempt to cross-check the resulting data, two dating methods, the U/Th (Quinif method at CERAK Laboratory, Belgium) and the 14C (AMS method at Beta Laboratory, USA) have been applied on two sets of samples collected in each carbonate unit.
The Valdelsa terraced carbonate deposits The four carbonate terraced successions, resting unconformably above a stepped substrate mainly made of Neogene silicoclastic deposits, are exposed in different proportions along intermediate tracts of the valleys progressively replacing upstream and downstream the equivalent but exclusively detrital/lithoclastic deposits (Fig. 3). The calcareous bodies, ranging in thickness from 15– 20 m, are poorly exposed so that observations on the internal organization and lateral distribution of the depositional facies is often approximate whereas identification and facies description derive from field observations implemented by petrologic analysis carried out in a number of small sites cropping out from a commonly dense Mediterranean maquis. Most of the outcrops have reduced lateral and vertical extent (2– 3 m thick), while only in few places a complete natural section of a synthem can be observed. The lateral transitions to the detrital/lithoclastic alluvial deposits appear to be rather abrupt upstream where phytohermal pure limestone facies suddenly occur, while downstream chalky facies progressively mix with detrital material and then disappear. The basal transition directly on the substrate or above irregular erosion surfaces on underlying fining-upward,
267
alluvial deposits, is often marked by relatively thin, commonly lenticular layers of mixed carbonate –lithoclastic alluvial deposits containing wood fragments and/or peat lenses. Locally channels cut in the unconsolidated substrate are filled up of coarse materials comprising calcareous clasts/boulders manifestly deriving from the erosion of continental carbonates originated from a previous carbonated system eroded at the time of deposition of the synthem (Capezzuoli et al. 2008). The carbonate body consists of calcareous tufa forming irregularly superposed sequences, sometimes including lenticular layers of dark clays with preserved organic matter (peat), separated by erosional surfaces and ending with a tabular top commonly covered by brown soils (Capezzuoli & Sandrelli 2004). Their internal structure corresponds to a facies mosaic varying in composition according to their position in the valley profile. The three dominant facies associations bear the distinctive structures of calcareous tufa (facies associations A –C; Fig. 3). A fourth one (facies association D) showing peculiar depositional features that are not typical of calcareous tufa is locally found. Facies association A – massive lenticular bodies composed of rather well-cemented phytohermal boundstone and framestone (Fig. 4a) with vertical development of stromatolite heads and bryophyte hummocks (Fig. 4b –d). This phytohermal facies irregularly interfingers or vertically alternates with lenticular layers of phytoclastic and chalky sediments and locally of lithoclasic sands and gravel. This facies association represents the first carbonate sediments that replace upstream the terraced alluvial materials. It laterally interfingers with facies association B. It can be related to fast flowing water in a rather low gradient setting, corresponding to a fluvial barrage system (Pedley 1990, 2009) with relatively small dams and backward ponds. Facies association B – crudely stratified, irregularly alternated lenticular layers of granular and chalky facies consisting of calcareous sands and micritic/peloidal lime mud (Fig. 5a) with frequent
Fig. 3. Lateral relationships of the detrital and carbonate facies in the Late Pleistocene –Holocene Valdelsa Synthems. See text for description of the carbonate facies associations. Like in Bagnoli section, Facies association A is locally and/or partially replaced by Facies association D.
268
E. CAPEZZUOLI ET AL.
Fig. 4. Calcareous tufa facies in the Valdelsa synthems, Association A: (a) Phytohermal framestone with encrusted leaves (ABB Synthem, Colle Val d’Elsa). (b) Fossil abandoned knee-like rim-margin made of superposed phytohermal carbonate crusts (Diborrato Natural Park, south of Colle Val d’Elsa). (c) Superposed cushion-like phytohermal bodies represent a fossil cascade at the Diborrato Natural Park. (d) Laminated bryophyte boundstone in the BEL Synthem (Staggia Senese).
remains of ostracods and gastropods. Lenticular accumulations of oncoids or phytoclasts and thin lenticular layers bearing in place small phytohermal clusters (Fig. 5b) irregularly associated with calcite-encrusted cyanobacterial tufts or bryophyte mounds are locally interbedded with lenses of lithoclastic sands and gravel, and of sapropelitic clays as well as remnants of palaeosols. This facies association prevails in the intermediate tract of the terraced calcareous body and locally interfingers with both facies association A and C. It can be related to poorly drained conditions corresponding to a fluvial/paludal system (Pedley 1990, 2009). Facies association C – dominantly lithoclastic deposits (lithoclast sands and conglomerates) with irregularly distributed calcareous intercalations (Fig. 5c) made of lenticular accumulations of oncoid/cyanoliths and phytoclast/intraclasts (Fig. 5d), and layers of micritic/peloidal lime mud. Asymmetrical stromatolitic microhermes and more or less developed microbial coating of pebble substrates can be locally found. This facies
association, commonly transitional to facies association B, represents the end of the carbonate sedimentation in the valley where the last carbonate deposits were laid down by nearly depleted carbonate waters in a dominantly detrital alluvial setting. The depositional conditions of the carbonates, can be related to active fluvial flow in a braidplain system (Pedley 1990, 2009). In the upstream occurrence of the FOC Synthem calcareous body exposed in the Imbotroni Valley at Bagnoli (SE of San Gimignano, Fig. 1), the facies association A is replaced by a different facies association (facies association D); laterally and downstream this facies disappears and the calcareous tufa facies crops out (Capezzuoli et al. 2008). Facies association D, about 15 m thick, is characterized by the alternation of poorly bedded chalky tufa lithofacies (as in facies association B) and of laminated limestone made of compact laminated centimetric layers (Fig. 6a) forming sharply defined, well bedded, uniformly laterally extended planar packages. The laminae consisting of arborescent
CALCAREOUS TUFA AS INDICATORS OF CLIMATIC VARIABILITY
269
Fig. 5. Calcareous tufa facies in the Valdelsa synthems. Association B: (a) Alternated lime-mudstone and calcareous sands in a palustrine tufa succession (FOC Synthem, Gracciano Val d’Elsa). (b) Small phytoherms and clusters of phytoclastic material included in the sandy facies of the calcareous tufa (FOC Synthem, Bagnoli). Association C: (c) lithoclastic accumulation (lithoclast sands and conglomerates) with small irregularly distributed phytoherms (CAL Synthem, Colle Val d’Elsa). (d) Section of the CAL Synthem near Poggibonsi; the contact with the Pliocene substrate is marked by a broken line.
crystalline and shrublike vertical structures (Fig. 6b) alternated with micritic mudstone (Fig. 6c) locally enclosing lenses of accumulated fragments of calcite rafts (Fig. 6d). The chalky, porous facies made of peloidal micrite lacks any trace of macrophytes or faunal elements (ostracods and gastropods). Some of these features are identical to those illustrated for deposits forming in hydrothermal systems: the crystalline and shrublike vertical structures (Fig. 7), described by Chafetz & Guidry (1999) and the calcite rafts illustrated by Folk et al. (1985) and described as ‘paper thin rafts’ by Guo & Riding (1998). These lithologic features, unusual for the calcareous tufa characterized by roughly bedded, unlithified chalky deposits rich in micro- and macrophytes remains, could be interpreted as originated on carbonate-rich waters flowing from hydrothermal springs (Capezzuoli et al. 2008). Sedimentological and petrographic evidence suggest that the carbonate-rich waters flowing at Bagnoli in the Late Pleistocene Imbotroni Valley
were mixed waters in part of meteoric/karstic origin in part cooled waters derived from a thermal source located upstream and now covered by younger deposits. We can speculate that a thermal spring was building up a local small travertine system, as suggested by the occurrence of well-bedded shrub laminites and paper-thin rafts. Laterally this water should have been mixed with ambient temperature, low carbonate water of the Imbotroni Creek that upstream was unable to precipitate carbonate deposits (Fig. 1b). This new mixed, carbonate enriched water, could precipitate tufa deposits in ponds and marshes of the richly vegetated fluvial valley system. Ponds filled with carbonate mud inhabited by ostracods and gastropods were bounded by shallow barrages colonized by bryophytes and small cyanobacterial tufts able to entrap encrusted fragments of the surrounding vegetation. Thus, we can speculate that the laminated carbonates were precipitated from hydrothermal waters perhaps comparable to the low-temperature
270
E. CAPEZZUOLI ET AL.
Fig. 6. Facies association D recognized in the Bagnoli section: (a) well-bedded, laminated facies. (b) Shrub structures. (c) Alternated micritic and shrub laminae; (d) accumulation of calcite rafts.
resurgences at present active in the area. One of them can be observed at Bagnoli, near the limestone outcrop. The spring is presently characterized by a low water discharge (1 l/s) and temperature of 23 8C. Its encrusting capacity is now very reduced, limited to a thin carbonate coating forming only at
Fig. 7. Microscopic view of the dendritic fern-like crystalline/bacterial shrub structure illustrated in Figure 6b.
the spring orifice (Casagli et al. 1990). However, discharge and temperature could have varied during time. Other known active springs are the Vene di Onci, consisting of several emergences with a mean discharge of 800 l/s and a temperature of 21 8C and the Caldane di San Marziale, near Gracciano d’Elsa, which drain into the Elsa River with a total low-flow of abut 80 l/s and a temperature of 22 8C (Casagli et al. 1990; Minissale 2004) (Fig. 1). These springs that fall in the class of Hypothermal springs (208 , T , 358; Casagli et al. 1990; Renaut & Jones 2000; Minissale 2004) are probably involved with the active deposition of tufa that is presently taking place downstream at the Diborrato Falls (Capezzuoli & Sandrelli 2004, 2006; Capezzuoli et al. 2007). In all these springs water upwelling seems to be connected to fractures related to the local NE –SW Apennine trending fault system, which connects the deeper circulating network with the upper aquifer (Casagli et al. 1990; Barazzuoli et al. 2002; Capezzuoli & Sandrelli 2006). The hypothesis that the origin of the continental carbonates of southern Valdelsa Basin can be
–
1.082
T4
CERAK FOC 0.524 + 0.008 1.060 + 0.012 0.657 + 0.016 4.2 + 0.1 6433 CERAK ABB 0.283 + 0.003 1.169 + 0.014 1.313 + 0.023 4.13 + 0.09 6434 T3
1.32 + 0.016 2.213 + 0.085 CERAK BEL 0.136 + 0.001 6432 T2
6.4 + 0.48
–
–
–
Weak 230Th/232Th Isotopic ratio, 230 Th/234U isotopic ratio higher than 1. Impossible to determine an age Weak 230Th/232Th Isotopic ratio, 230 Th/234U isotopic ratio higher than 1. Impossible to determine an age Good analisys, but the 230Th/232Th isotopic ratio attest for a not truthful age Weak 230Th/232Th Isotopic ratio 230 Th/234U isotopic ratio higher than 1. Impossible to determine an age 0.23 + 0.003 1.152 + 0.013 1.438 + 0.05
4.5 + 0.2 CERAK CAL 6429 T1
–
Age (kyr) 230
Th/234U 230
˙ U/238U 234
[U]ppm Unit Lab. code Sample
Table 1. U/Th data from the Valdelsa Synthems
The existing chronology of the Valdelsa carbonate succession was estimated to span from Late Pleistocene to present on stratigraphic evidence (Capezzuoli & Sandrelli 2004). During this period, significant changes of the landscape occurred. Several stages of erosion and fluvial-palustrine re-establishment have been recognized. With the aim to date the different depositional events, four U/Th dates (performed at CERAK Laboratory, Belgium) were obtained from compact, micritic layers collected in the different terraces, while four 14C dates (performed with AMS methodology at Beta Laboratory, USA) were obtained from wood and organic-rich layers of different terraces (Table 1). Unfortunately the U/Th dating failed because three of the samples (T1, T2, T4) were affected by contamination from detrital or colloidal elements, as evidenced by the 230Th/232Th isotopic ratio and by the 230Th/234U isotopic ratio higher than 1 (Table 1). Detrital contamination is one of the main constraints of the U/Th dating and that often occur (Garnett et al. 2004). The fourth sample (T3), not affected by contamination, did provide however a low 230Th/232Th isotopic ratio attesting for an unreliable age (Table 1). On the contrary the results of the four AMS 14C dates (Table 2) provide evidences to constrain the time interval of development of the synthems, and record their cyclic evolution and high incision rates of the valley. These four 14C datings, performed on wood and organic matter found in the rare organic-rich layers, commonly located at the base of the carbonate bodies, are consistent with their relative stratigraphic positions and suggest a regular evolution, from the older to the younger terraces, of the Valdelsa sequence (Table 2). The reliability of absolute dating moreover appears to be confirmed by the range of variability of d13C values measured in the four samples (Table 2) reflecting the extent of isotopic fractionation in the fossil material after deposition. According to Walker (2005, 17– 55) the extent of isotopic fractionation ranges close to the mean isotopic
Th/232Th [234U/238U]t ¼ 0
Chronological dating
Comments of the laboratory
related to the activity over time of these thermal springs appears to be supported by their areal distribution. In fact, as the small Bagnoli spring, the Caldane di San Marziale and Vene di Onci springs are also located upstream of the segment of the Elsa River Valley where carbonates were deposited. It is possible to assume that the Pleistocene calcareous deposits occurring in southern Valdelsa Basin were connected with thermal springs that in part are now dried up or buried by the overlying calcareous tufa.
271
114.3[þ5.2/ 2 4.8] –
CALCAREOUS TUFA AS INDICATORS OF CLIMATIC VARIABILITY
Beta-170679
Beta-180158
Wood sample cored by a drilling in a well 20 centimetres thick organic-rich layer sampled near the base of the synthem 17 centimetres thick organic-rich layer sampled in the middle of the synthem 25 centimetres thick organic-rich layer sampled near the base of the synthem Beta-180156
Beta-180159
Description of the sample
E. CAPEZZUOLI ET AL.
Lab. code
272
composition of wood and peat (226‰). With regard to the dating of the sample of wood (Co2), it is possible to assume that the resulting measurement represents a maximum age, therefore including the possible residence time in soil and/or reworking of the wood, before C14 clock started. The 14C chronological data combined with the results of sedimentological analysis support a new model of the geomorphologic evolution of the valley network and the cyclic deposition of the tufa systems. The depositional events recognized in the Valdelsa terraced carbonates have been correlated with the main Late Quaternary climatic changes occurred in the European area and in particular in Central Italy.
2 28.4‰
2 25.2‰
2 25.7‰
2 25.2‰
8360 – 8160
16 950 – 16 240
Not available
Not available 25 690 + 180 Poggibonsi C1
CAL
21 080 + 100 Colle Val d’Elsa Co5
CAL
13 830 + 50 North of Campiglia dei Foci Ai3
FOC
7430 + 50 Colle Val d’Elsa
BEL
The study of the terraced carbonate succession in the southern Valdelsa Basin documents significant changes of the landscape during Late Quaternary. The four episodes of tufa development alternated with stages of erosion that were followed by the re-establishment of fluvial–palustrine conditions. For a better understanding of the focal mechanism triggering the erosional/depositional events in this sector of Tuscany, we try to outline the likely palaeoclimatic and/or tectonic regimes that controlled the origin of the Valdelsa terraced succession and related carbonate deposits.
Climatic conditions of deposition
Co2
Unit Locality Sample
Table 2. AMS 14C data from the Valdelsa Synthems
14 C age (yrs B.P.)
Calibrated age 2s (cal yrs B.P.)
13
C/12C ratio
Discussion
In order to document the palaeoclimatic context in which the calcareous tufa of the late Quaternary Valdelsa carbonate synthems were developed, an attempt has been made to correlate the depositional/erosional phases represented by the terrace succession, with the coeval climatic changes known to have occurred in Central Italy. Palaeoclimatic data from Late Pleistocene – Holocene lacustrine deposits of Central Apennine (Fig. 1a) are available in the literature (Narcisi & Anselmi 1998; Magri 1999; Magri & Sadori 1999; Giraudi 2000, 2001; Narcisi 2001; Sadori et al. 2004; Magny et al. 2006; Drescher-Schneider et al. 2007). Giraudi (2004) summing up the data concerning the volcanic lakes of Mezzano and Vico (northern Latium) and the drained tectonic Fucino Lake (Abruzzo) (Fig. 1a), evidence an unequivocal correspondence between the high and low water levels of these lakes. According to Giraudi (2004), the distribution of these Central Apennine lakes implies that the key factor controlling the cyclical alternation of the lake levels was conditioned by the hydrological balance (difference between precipitation and evaporation) and, consequently, by climatic oscillations. Comparing
CALCAREOUS TUFA AS INDICATORS OF CLIMATIC VARIABILITY
Giraudi’s reconstruction with the available data of the aggradational/erosional phases of the Valdelsa terraced sequence it is possible to appreciate the correspondence of events (Fig. 8). In particular, radiometric data obtained from the CAL Synthem show that this depositional phase can be readily related to the high lake-level stage connected with the 1st terrace of Mezzano Lake and the oldest terrace recognized in the Vico Lake. The radiometric data from the FOC Synthem correlates with the 2nd terrace of the Mezzano Lake and the youngest terrace of the Vico Lake. The same high lake-level event has been recognized in the Accesa Lake (southern Tuscany) (Fig. 1a) where a coeval rapid rise of the water table is attested by a remarkable peak of Chara oogonia production (Magny et al. 2006). The age obtained for the BEL Synthem coincides with the high lake-level stage recognized in the Fucino Lake and assumed in the Mezzano Lake. Even though the ABB, the POG Synthems, and all the interposed erosive stages lack actual dating, the correspondence between the succession of events reconstructed in Central Italy by Giraudi (2004) and those recognized in Valdelsa Basin enables the correlation of the two sequences. Consequently the erosional phase between CAL and FOC Synthems can be correlated with the low lake-level at about 17– 20 ka BP of the Mezzano, Vico and Fucino lakes (Giraudi 1989). The FOC-BEL erosional phase appears to be coeval with the fall of the lake-level recorded in the three Central Apennine lakes about 10 –12 ka BP. The BEL-POG transition represents the last major erosional phase recognized in the Valdelsa Basin. The correlation between this erosive stage and the low lake-level recognized in the Mezzano and Fucino lakes at c. 4 ka BP is likely. A coeval depositional phase at about 4,000 years BP (uncal. 14 C age) has been also recognized in the low-stand limestones of Accesa Lake by Magny et al. (2006). The last POG Synthem comprises the most recent alluvial deposits of the area. Unfortunately, no data are available for the oldest Valdelsa Synthem (ABB) and the associated erosional phase (ABB-CAL).
Tectonic involvement The presence of travertine-like facies in the basal part of the calcareous sequences exposed in the proximal part of the Valdelsa terraces suggests an alternative interpretation to climate being the dominant control on the genesis of continental carbonates in this area. In fact, according to some authors (Altunel & Hancock 1993a, b; Hancock et al. 1999; Atabey 2002; Brogi et al. 2005a; Brogi & Capezzuoli 2008), travertine bodies deposited
273
around high-temperature springs are considered good indicators of extensional tectonic activity. The strict relationship between travertine deposition and faulting, recently recognized and termed ‘Travitonics’ (Hancock et al. 1999), suggests that the age of travertine can be taken as a proxy for the age of faulting (Altunel 2005; Hancock et al. 1999). A similar tectonic control has been recognized in the Hungarian Danube area where several orders of terraces made of travertine have been interpreted as successive river incisions resulting from the combined uplift of the Hungarian Mountain Range and periodic climate changes (Ruszkiczay-Ru¨diger et al. 2005). The assumed tectonic activity in the Valdelsa Valley could be related to the extensional tectonic movements affecting Southern Tuscany since the Early –Middle Miocene (Bossio et al. 1995a; Carmignani et al. 1994). The tectonic evolution of the Pliocene –Quaternary basins in Southern Tuscany and in the Valdelsa Basin itself, appears to be mostly controlled by a rapid regional uplift (Dallmeyer & Liotta 1998) triggered by a widespread, mainly intrusive Middle Pliocene magmatism (Brogi et al. 2005b). The laminated carbonates, even if finally would be demonstrated to be caused by thermal waters, seem to be very limited and not sufficient to explain a significant tectonic control on the local fluvial deposition. In fact, the physical and geochemical features of the present day active low temperature springs suggest that the superficial aquifer is fed by waters coming from the ‘Calcare cavernoso’ reservoir (Casagli et al. 1990; Barazzuoli et al. 2002). The slight thermalism of the springs can be easily related to the local geothermal gradient, which is about 57 8C/km (Del Chicca et al. 1988). Moreover, during the Late Pleistocene– Holocene there are only minor evidences of tectonic activity in this area (Capezzuoli & Sandrelli 2006) and in other areas of Tuscany (Boccaletti et al. 1999). Certainly the Late Pleistocene–Holocene tectonic activity seems to be less significant than the Middle –Late Pleistocene movements of uplift that resulted in the rearrangement of the Arno Basin drainage, recognized by Bartolini & Pranzini (1981). This latter uplift phase has been recognized as a major event that changed an area of rather low relief into the Northern Apennine mountain chain (Bartolini 2003).
Depositional/erosional phases and late Quaternary climatic events: a tentative correlation Assuming that the Valdelsa carbonate deposition was mainly controlled by climate, a tentative correlation between the depositional/erosional Valdelsa
274 E. CAPEZZUOLI ET AL. Fig. 8. Dated Valdelsa terracing events correlated with the fluctuations of lake levels in the Central Apennine (data about Mezzano, Fucino and Vico Lake derived from Giraudi 2004, slightly modified).
CALCAREOUS TUFA AS INDICATORS OF CLIMATIC VARIABILITY
phases and the main climatic variations recognized in the European area during the Last Termination (Bjo¨rck et al. 1998) and Holocene can be attempted. The FOC Synthem appears to correspond to the Late Glacial Interstadial warming event (Bølling-Allerød interval of Mangerud et al. 1974 and GI-1 episode of Bjo¨rck et al. 1998) dated to about 13 –14 ka BP in the North Atlantic region (Lowe et al. 1994; Bjo¨rck et al. 1998) and to 15 ka BP in southern Europe (De Beaulieu et al. 1994). The successive FOC-BEL erosional phase could be related to the cold and dry Younger Dryas event (GS-1: Bjo¨rck et al. 1998) dated to c. 11 ka BP by several authors (Broecker 1994; Lowe et al. 1994; Bond 1995; Bjo¨rck et al. 1998). The same Late Glacial erosional phase documented in the Valdelsa terraced sequence has been also recognized in some Tuscan rivers (Arno River – Aguzzi et al. 2007; Ombrone River – Biserni et al. 2005). The BEL Synthem age corresponds to the ‘Optimum Climatic’ phase generally considered as a wet and warm period. It seems to have been an active period of carbonate deposition in Central Italy since coeval calcareous tufa (c. 7–8 ka BP) have been reported from the Adriatic side of Northern Apennine (Cilla et al. 1994; Calderoni et al. 1996). The BEL-POG transition can be easily correlated with the dry Sub-Boreal period. Evidence of this period is largely reported from Central Italy. At c. 4000 years BP (uncal. 14C age) deposition of limestone at low-stand lake-level has been found at Accesa Lake (Magny et al. 2006). A sudden drop in the amount of pollens from arboreal vegetation, related to a climatic deterioration, has been recognized in the same period at Lagaccione and Mezzano lakes (Magri 1999; Ramrath et al. 1999). The latest POG Synthem is characterized by the near disappearance of carbonate deposition. Actually thin carbonate crusts are found only in the immediate surroundings of only one of the lowthermal springs active in Valdelsa (Bagnoli spring). This drastic reduction of carbonate deposition may be related to the same event described in all the European area as ‘tufa decline’ (Goudie et al. 1993). Causes of this decline dated to the Sub-Atlantic Period (last 2.5 ka; Goudie et al. 1993) are not yet clear (for a full discussion see Goudie et al. 1993; Dramis et al. 1999) but their effects are evident also in the Valdelsa carbonate area. The oldest Valdelsa Synthem (ABB) has not yet been directly dated.
Main climatic regime According to Giraudi’s (2004) interpretation of the climatic control in the development of the
275
deposition/erosion cycles in Central Apennine lakes during the Late Pleistocene, it seems likely that a similar regime, primarily determined by the amount of rainfall and consequent variations of the hydrological balance, could have also occurred during the last 30 ka in the relatively near Valdelsa area. Under such a climatic regime, increase of water discharge should enhance karstic dissolution and consequently amplify tufa deposition because of the increased availability of calcium carbonate in stream waters. Conversely, reduction of rainfall would result in reduced limestone dissolution, less calcium carbonate being dissolved in stream waters, and consequently lower rates of tufa deposition (Goudie et al. 1993; Pentecost 2005). In the latter situation, calcareous tufa aggradations and vegetation are reduced, while downcutting is intensified. These conditions are reminiscent of the Erhart’s (1956) ‘biostasie’ and ‘rhexistasie’ scenario where periods of biologically-mediated landscape stability alternate with period of downcutting and erosion. Similar deductions derive from Vandour (1986) and Goudie et al. (1993), who infer a stable and extensive vegetation cover associated with stable soils as the key factor controlling tufa deposition. The close relation between climate and tufa deposition has been stressed by many studies (Srdocˇ et al. 1983; Henning et al. 1983; Carrara 1991; Griffiths & Pedley 1995; Bessanc¸on et al. 1997; Martin-Algarra et al. 2003) showing that tufa deposition reached a maximum during warm and wet interglacial periods. Pedley (1990) observes that Holocene tufas are better developed in humid, temperate climates, whereas their growth is slowed by cold conditions, and that semi-arid regimes can rarely maintain the high water tables necessary for sustained tufa deposition. Dissolution rate in the recharge area, which ultimately controls downstream tufa precipitation rates, can be affected by reduced rainfall, increased evaporation, and falling water tables (Ford & Pedley 1996). As evidenced by Evans (1999), the virtual absence of tufas in the Oligocene Brule Formation (South Dakota, USA) appears to result from palaeoclimatic conditions, since an increase of aridity across the Eocene– Oligocene boundary is well-documented in that region. The alternation of depositional and erosional phases in the Valdelsa terraced succession appears to have been strictly linked to the availability of water that is, to a rainy regime, since radiometric dating indicates that at least the older sequences (CAL, FOC and probably ABB Synthems) were deposited during a glacial period (O.I.S. 2). As a consequence, air temperature would not have significantly influenced the encrusting capacity of these waters.
276
E. CAPEZZUOLI ET AL.
This view is not shared by some authors who postulate that tufa growth in the Mediterranean area was limited to global warm climate and that no carbonate deposition could occur during glacial, stadial and interstadial periods (Vaudour 1988, 1994; Pen˜a et al. 2000; Horvatincˇic´ et al. 2000; Soligo et al. 2002; Pedley et al. 2003). Possible explanations of this contradiction are: (i) the glacial period in the southern segment of Valdelsa valley was not sufficiently cold to prevent the deposition of carbonate; (ii) the rainfall is arguably as a climatic factor affecting tufa deposition more important than supposed before; and (iii) the speculated low-thermal characteristics of the issuing waters had a positive influence on the production of carbonate deposits even during cool/cold periods. The latter interpretation, even if the climate –travertine relationships are not yet adequately substantiated, is in agreement with several authors which concur in the opinion that endogenic groundwaters may be less dependent on climate, since they have little opportunity to interact with the atmosphere before resurgence (Sturchio et al. 1994; Pentecost 1995; Capezzuoli & Gandin 2004; Faccenna et al. 2008). More isotopic and geochemical data are necessary for a comprehensive vision of this problem, but at the moment they are not available.
controlled by the climatic variability of the younger Late Pleistocene–Holocene. At least three of the Valdelsa carbonate Synthems and the interposed erosional phases can be correlated with the major climatic changes recognized in the European –Mediterranean area. The FOC Synthem apparently coincides with the Last Glacial Interstadial; the FOC/BEL erosional phase with the Younger Dryas; the BEL Synthem with the ‘Optimum Climatic’ (Atlantic); the BEL/POG erosional phase with the Sub-Boreal, and finally the POG Synthem seems to summarize the deposition of the last 2.5 ka (Sub-Atlantic). This climatic correlation and the radiometric data imply that the deposition of calcareous tufa in Valdelsa was mainly dependent on the rainfall availability and, consequently, active even during cold periods. In this reconstruction, evidence of Late Pleistocene–Holocene climatic variations in the Central Apennine have been inferred for the first time from terraced fluvial carbonates (calcareous tufa) rather than from detrital lacustrine deposits. This work has been financially supported by the University of Siena PAR Grant (F. Sandrelli). The authors would like to thank Susan Pedley for language ameliorations of the text. We are grateful also to Brian Jones and to the Volume Editor for critical reading of the manuscript and constructive comments suggesting major improvements.
Conclusions The southern segment of the Valdelsa Basin (Tuscany) is characterized by the presence of a carbonate terraced succession developed in a limited segment of the fluvial pattern. Sedimentological and petrographic analysis shows that these carbonates are mainly composed of calcareous tufa deposited in a fluviatile–paludal environment. Radiometric dates performed on organic-rich material indicate that at least the older terraced sequences were deposited during a cold climatic interval. A scenario that cannot be discarded to explain the occurrence of calcite precipitation even during cold periods is that the carbonate system could have been fed by slightly warm waters, even during cold periods. This is speculated on the presence of microbial laminated carbonate facies locally exposed at the base of the apical portion of these carbonates bodies and possibly related to thermal springs. Geomorphological and sedimentological analysis supported by the results the radiometric data show that the depositional/erosional evolution of the Valdelsa succession is constrained by major climate phases corresponding to those recognized in the coeval detrital lacustrine successions of Central Italy. Consequently, the evolution of the terraced succession appears to have been directly
References A GUZZI , M., A MOROSI , A., C OLALONGO , M. L., R ICCI L UCCHI , M., R OSSI , V., S ARTI , G. & V AIANI , C. 2007. Late Quaternary climatic evolution of the Arno coastal plain (Western Tuscany, Italy) from subsurface data. Sedimentary Geology, 202, 211–240. A LTUNEL , E. 2005. Travertines: neotectonic indicators. In: O ZKUL , M., Y AGIZ , S. & J ONES , B. (eds) Proceedings of 1st International Symposium on Travertine, September 21–25, 2005, Denizli-Turkey. Kozan Ofset, Ankara, 120–127. A LTUNEL , E. & H ANCOCK , L. 1993a. Morphology and structural setting of Quaternary travertines at Pamukkale, Turkey. Geological Journal, 28, 335– 346. A LTUNEL , E. & H ANCOCK , P. L. 1993b. Active fissuring and faulting in Quaternary travertines at Pamukkale, western turkey. Zeitschrift fuer Geomorphologie N.F., 94, 285– 302. A NDREWS , J. E. 2006. Palaeoclimatic records from stable isotopes in riverine tufas: synthesis and review. Earth Science Reviews, 75, 85– 104. A NDREWS , J. E. & B RASIER , A. T. 2005. Seasonal records of climate change in annually laminated tufas: short review and future prospects. Journal of Quaternary Science, 20, 411– 421. A NDREWS , J. E., R IDING , R. & D ENNIS , P. F. 1993. Stable isotopic compositions of recent cyanobacterial carbonates from the British Isles: local and regional environmental controls. Sedimentology, 40, 303– 314.
CALCAREOUS TUFA AS INDICATORS OF CLIMATIC VARIABILITY A NDREWS , J. E., P EDLEY , H. M. & D ENNIS , P. F. 1994. Stable isotope record of palaeoclimatic change in a British Holocene tufa. Holocene, 4, 349– 355. A NDREWS , J. E., R IDING , R. & D ENNIS , P. F. 1997. The stable isotope record of environmental and climatic signals in modern terrestrial carbonates from Europe. Palaeogeography, Palaeoclimatology, Palaeoecology, 129, 171– 189. A NDREWS , J. E., P EDLEY , M. & D ENNIS , P. F. 2000. Palaeoenvironmental records in Holocene Spanish tufas: a stable isotope approach in search of reliable climatic archives. Sedimentology, 47, 961– 978. A TABEY , E. 2002. The formation of fissure ridge type laminated travertine-tufa deposits microscopical characteristics and diagenesis, Kirs¸ehir, central Anatolia. Bulletin of the Mineral Research and Exploration, 123–124, 59–65. B ARAZZUOLI , P., M OCENNI , B., R IGATI , R. & S ALLEOLINI , M. 2002. Studio idrogeologico dei rapporti tra acque superficiali e profonde nella zona di Gracciano dell’Elsa (Prov. di Siena). Quaderni di Geologia Applicata, 9(2), 79–95. B AR -M ATTHEWS , M., A YALON , A. & K AUFMAN , A. 1997. Late Quaternary Palaeoclimate in the Eastern Mediterranean Region from Stable Isotope Analysis of Speleothems at Soreq Cave, Israel. Quaternary Research, 47(2), 155– 168. B ARTOLINI , C. 2003. When did the Northern Apennine become a mountain chain? Quaternary International, 101–102, 75–80. B ARTOLINI , C. & P ECCERILLO , A. 2002. I fattori geologici delle forme del rilievo. Lezioni di Geomorfologia Strutturale. Pitagora Editrice, Bologna. B ARTOLINI , C. & P RANZINI , G. 1981. Plio-Quaternary evolution of the Arno basin drainage. Zeitschrift fuer Geomorphologie N.F., 40, 77– 91. B ESSANC¸ ON , J., D ELGIOVINE , A., F ONTUGNE , M., L ALOU , C., S ANLAVILLE , P. & V AUDOUR , J. 1997. Mise en e´vidence et duration de phases humides du Ple´istoce`ne supe´rieur dans la re´gion de Palmyre (Syrie). Pale´orient, 23, 5– 23. B ISERNI , G., B ERENDSEN , H. J. A. & S ANDRELLI , F. 2005. Holocene evolution of the Ombrone alluvial plain (Tuscany, Central Italy). Bollettino della Societa` Geologica Italiana, 124, 465–474. B JO¨ RCK , S., W ALKER , M. J. C., C WYNAR , L. C., J OHNSEN , S., K NUDSEN , S., L OWE , J. J., W OHLFARTH , B. & INTIMATE MEMBERS . 1998. An event stratigraphy for the last Termination in the North Atlantic region based on the Greenland ice-core record: a proposal by the INTIMATE group. Journal of Quaternary Science, 13, 283–292. B OCCALETTI , M., B ONINI , M., M ORATTI , G. & S ANI , F. 1999. Compressive neogene-quaternary tectonics in the hinterland area of the northern Apennines. Journal of Petroleum Geology, 22(4), 37– 60. B OND , C. 1995. Climate and the conveyor. Nature, 377, 383–384. B OSSIO , A., C OSTANTINI , A. ET AL . 1995a. Studi preliminari sul sollevamento della Toscana meridionale dopo il Pliocene medio. Studi Geologici Camerti, Volume Speciale, 1995/1, 87– 91. B OSSIO , A., F ORESI , L. M., M AZZEI , R., S ALVATORINI , G. & S ANDRELLI , F. 1995b. Evoluzione
277
tettonico-sedimentaria neogenica lungo una trasversale ai Bacini di Volterra e della Val d’Elsa. Studi Geologici Camerti, Volume Speciale, 1995/1, 93–104. B OSSIO , A., M AZZEI , R., S ALVATORINI , G. & S ANDRELLI , F. 2000. Geologia dell’area compresa tra Siena, Poggibonsi e Castellina in Chianti (Prov. di Siena). Atti Societa` Toscana di Scienze Naturali, Memorie Serie A, 107, 69– 85. B RANCACCIO , L., D’A RGENIO , B., F ERRERI , V., P REITE M ARTINEZ , M., S TANZIONE , D. & T URI , B. 1986. Caratteri tessiturali e geochimica dei travertini di Rocchetta a Volturno (Molise). Bollettino della Societa` Geologica Italiana, 105, 265– 277. B RIDGLAND , D. R. & W ESTAWAY , R. 2008. Climatically controlled river terrace staircases: a worldwide Quaternary phenomenon. Geomorphology, 98(3– 4), 285– 315. B RIDGLAND , D. R., S CHREVE , D. C., K EEN , D. H., M EYRICK , R. & W ESTAWAY , R. 2004a. Biostratigraphical correlation between the late Quaternary sequence of the Thames and key fluvial localities in central Germany. Proceedings of the Geologists’ Association, 115, 125– 140. B RIDGLAND , D., M ADDY , D. & B ATES , M. 2004b. River terrace sequences: templates for Quaternary geochronology and marine–terrestrial correlation. Journal of Quaternary Science, 19, 203– 218. B ROECKER , W. 1994. Massive iceberg discharge as triggers for global climate change. Nature, 372, 421–424. B ROGI , A. & C APEZZUOLI , E. 2008. Travertine deposition and faulting: the fault-related travertine fissure-ridge at Terme S. Giovanni, Rapolano Terme (Italy). International Journal of Earth Science (Geol. Rundsch.), doi: 10.1007/s00531-007-0290-z. B ROGI , A., C APEZZUOLI , E., C OSTANTINI , A., G ANDIN , A. & L AZZAROTTO , A. 2005a. Tectonics and travertine relationships in the Rapolano Terme area (Northern Apennines, Italy). In: O ZKUL , M., Y AGIZ , S. & J ONES , B. (eds) Proceedings of 1st International Symposium on Travertine, September 21– 25, 2005, Denizli Turkey. Kozan Offset, Ankara, 142–148. B ROGI , A., L AZZAROTTO , A., L IOTTA , D., R ANALLI , G. & CROP 18 WORKING G ROUP . 2005b. Crustal structures in the geothermal areas of southern Tuscany (Italy): insights from the CROP 18 deep seismic reflection lines. Journal of Volcanology and Geothermal Research, 148, 60–80. B UCCINO , G., D’A RGENIO , B., F ERRERI , V., B RANCACCIO , L., F ERRERI , M., P ANICHI , C. & S TANZIONE , D. 1978. I travertini della bassa Val Tanagro (Campania) studio geomorfologico, sedimentologico e geochimico. Bollettino della Societa` Geologica Italiana, 97, 617 –646. C ALDERONI , G., C ILLA , G., D RAMIS , F., E SU , D., M AGNATI , M. & M ATERAZZI , M. 1996. La deposizione di travertino nelle aree prossimali dei Fiumi Esino, Potenza e Chienti durante l’Olocene antico (Appennino cantrale marchigiano). Italian Journal of Quaternary Sciences, 9(2), 481– 492. C ANUTI , P. & T ACCONI , P. 1975. Idrogeologia e risorse idriche del bacino del Fiume Arno: sottobacini dei fiumi Greve, Pesa, Elsa. Conference. ‘Risorse idriche ed assetto del territorio’ Provincia di Firenze, Studi di Geologia Applicata e Geologia dell’Ambiente.
278
E. CAPEZZUOLI ET AL.
C APEZZUOLI , E. & G ANDIN , A. 2004. I ‘travertini’ in Italia: proposta di una nuova nomenclatura basata sui caratteri genetici. Italian Journal of Quaternary Science, 17(2/1), 273– 284. C APEZZUOLI , E., C OSTANTINI , E. A. C., P RIORI , S. & S ANDRELLI , F. 2007. Stratigraphic aspects from the Middle Pleistocene continental deposits of the southern Valdelsa Basin. Rendiconti della Societa` Geologica Italiana, Nuova Serie, 5, 95–99. C APEZZUOLI , E. & S ANDRELLI , F. 2004. I sedimenti quaternari del settore meridionale della Valdelsa (Provincia di Siena). Italian Journal of Quaternary Science, 17(1), 33– 40. C APEZZUOLI , E. & G ANDIN , A. 2005. Facies distribution and microfacies of thermal-spring travertine from Tuscany. In: O ZKUL , M., Y AGIZ , S. & J ONES , B. (eds) Proceedings of 1st International Symposium on Travertine, September 21–25, 2005, Denizli Turkey. Kozan Offset, Ankara, 43–50. C APEZZUOLI , E., G ANDIN , A. & S ANDRELLI , F. 2008. Evidence of associated deposition of travertine and calcareous tufa in the Quaternary carbonates of Valdelsa Basin (Tuscany). Italian Journal of Quaternary Sciences, 21(B), 113– 124. C APEZZUOLI , E. & S ANDRELLI , F. 2006. Neotectonic evidence in the Quaternary continental carbonates from Southern Valdelsa Basin (Tuscany). Italian Journal of Quaternary Science, 19(1), 155–166. C ARMIGNANI , L., D ECANDIA , F. A., F ANTOZZI , P. L., L AZZAROTTO , A., L OTTA , D. & M ECCHERI , M. 1994. Tertiary extensional tectonics in Tuscany (Northern Apennines, Italy). Tectonophysics, 238, 295– 315. C ARRARA , C. 1991. Travertine deposits of the middle Liri valley (central Italy): geomorphological, sedimentological and palaeoclimatic implications. Italian Journal of Quaternary Science, 4, 55–84. C ASAGLI , N., D UCHI , V. & P RANZINI , G. 1990. Low temperature thermal springs of Colle Val d’Elsa (Tuscany, Central Italy). Memoires of the 22nd Congress of IAH, XXII, 635–644. C HAFETZ , H. S. & F OLK , R. L. 1984. Travertines: depositional morphology and the bacterially constructed costituents. Journal of Sedimentary Petrology, 54, 289– 316. C HAFETZ , H. S. & G UIDRY , S. A. 1999. Bacterial shrubs, crystal shrubs, and ray-crystal shrubs: bacterial vs. abiotic precipitation. Sedimentary Geology, 126, 57–74. C ILLA , G., C OLTORTI , M. & D RAMIS , F. 1994. Holocene fluvial dynamics in mountain areas: the case of the river Esino (Appennino Umbro-marchigiano). Geografia Fisica e Dinamica Quaternaria, 17, 163–174. D ALLMEYER , R. D. & L IOTTA , D. 1998. Extension, uplift of rocks and cooling ages in thinned crustal provinces: the Larderello geothermal area (inner Northern Apennines, Italy). Geological Magazine, 135, 193– 202. D’A RGENIO , B. & F ERRERI , V. 1988. Ambienti deposizionali e litofacies dei travertini quaternari dell’Italia centro-meridionale. Memorie della Societa` Geologica Italiana, 41, 861– 868. D’A RGENIO , B., F ERRERI , M., F ERRERI , V. & S TANZIONE , D. 1981. Travertins of Southern Italy.
Texture, geochemistry and sedimentary model. Proceedings 2nd Meeting of International Association of Sedimentologists, Bologna, 43–46. D’A RGENIO , B., F ERRERI , V., S TANZIONE , D., B RANCACCIO , L. & F ERRERI , M. 1983. I travertini di Pontecagnano (Campania): geomorfologia, sedimentologia, geochimica. Bollettino della Societa` Geologica Italiana, 102, 123–136. D E B EAULIEU , J.-L., A NDRIEU , V., L OWE , J. J., P ONEL , P. & R EILLE , M. 1994. The Weichselian Late-glacial in southwestern Europe (Iberian Penisula, Pyrenees, Massif Central, Northern Apennines). Journal of Quaternary Sciences, 9, 101–108. D EL C HICCA , R., G IORNI , R., S QUARCI , P. & T AFFI , L. 1988. Sulle risorse geotermiche a medio-bassa temperatura nella parte meridionale del Bacino neoautoctono della Val d’Elsa. Studi preliminari: Misure di Gradiente. 28 Seminario Informativo P.F.E. – S.E.G., SI-5, 563–571, C.N.R. D EMICCO , R. V. & H ARDIE , L. A. 1994. Sedimentary structures and early diagenetic features of shallow marine carbonate deposits. SEPM Atlas Series, 1, 142–158. D IGERFELDT , G., DE B EAULIEU , J.-L., G UIOT , J. & M OUTHON , J. 1997. Reconstruction and palaeoclimatic interpretation of Holocene lake-level changes in Lac de Saint-Leger, Haute Provence, southeast France. Palaeogeography, Palaeoclimatology, Palaeoecology, 136, 231–258. D RAMIS , F., M ATERAZZI , M. & C ILLA , G. 1999. Influence of climatic changes on freshwater travertine deposition: a new hypothesis. Physics and Chemistry of the Earth (A), 24(10), 893–897. D RESCHER -S CHNEIDER , R., DE B EAULIEU , J. L. ET AL . 2007. Vegetation history, climate and human impact over the last 15,000 years at Lago dell’Accesa (Tuscany, Central Italy). Vegetation History and Archaeobotany, 16, 279– 299. D RYSDALE , R. N., Z ANCHETTA , G. ET AL . 2004. Palaeoclimatic implications of the growth history and stable isotope (d18O and d13C) geochemistry of a Middle to Late Pleistocene stalagmite from centralwestern Italy. Earth and Planetary Science Letters, 227, 215 –229. D WORKIN , S. I., N ORDT , L. & A TCHLEY , S. 2005. Determining terrestrial palaeotemperature using the oxygen isptopic composition of pedogenetic carbonate. Earth and Planetary Science Letters, 237, 56–68. E RHART , H. 1956. La Genese des sols en tantque phomene geologique. Masson, Paris. E VANS , J. E. 1999. Recognition and implications of Eocene tufas and travertines in the Chadron Formation, White River Group, Badlands of South Dakota. Sedimentology, 46, 771– 789. F ACCENNA , C., S OLIGO , M., B ILLI , A., D E F ILIPPIS , L., F UNICIELLO , R., R OSSETTI , C. & T UCCIMEI , P. 2008. Late Pleistocene depositional cycles of the Lapis Tiburtinus travertine (Tivoli, Central Italy): Possible influence of climate and fault activity. Global and Planetary Change, doi: 10.1016/j.gloplacha.2008.06.006. F ERRERI , V. 1985. Criteri di analisi di facies e classificazione dei travertini pleistocenici dell’Italia Meridionale. Rendiconti Accademia Scienze Fisiche e Matematiche, 52, Napoli.
CALCAREOUS TUFA AS INDICATORS OF CLIMATIC VARIABILITY F LEITMANN , D., B URNS , S. J., N EFF , U., M UDELSEE , M., M ANGINI , A. & M ATTER , A. 2004. Palaeoclimatic interpretation of high-resolution oxygen isotope profiles derived from annually laminated speleothems from Southern Oman. Quaternary Science Reviews, 23, 935–945. F OLK , R. L., C HAFETZ , H. S. & T IEZZI , P. A. 1985. Bizarre forms of depositional and diagenetic calcite in hot-spring travertines, central Italy. In: S CHEIDEMANN , N. & H ARRIS , P. (eds) Carbonate Cements. SEPM, 36, 349–369. F ORD , T. D. & P EDLEY , H. M. 1996. A review of tufa and travertine deposits of the world. Earth Science Review, 41, 117– 175. F RANK , N., B RAUM , M., H AMBACH , U., M ANGINI , A. & W AGNER , G. 2000. Warm Period Growth of Travertine during the Last Interglaciation in Southern Germany. Quaternary Research, 54(1), 38– 48. G ANDIN , A. & C APEZZUOLI , E. 2008. Travertine versus Calcareous tufa: distinctive petrologic features and related stable isotopes segnature. Italian Journal of Quaternary Sciences, 21(B), 125–136. G ANDIN , A., G IAMELLO , M., G UASPARRI , G., M UGNAINI , S. & S ABATINI , G. 2000. The Calcare Cavernoso of the Montagnola Senese (Siena, Italy): mineralogical–petrographic and petrogenetic features. Mineralogica et Petrogrophica Acta, 43, 271–289. G ARNETT , E. R., A NDREWS , J. E., P REECE , R. C. & D ENNIS , P. F. 2004. Climatic change recorded by stable isotopes and trace elements in a British Holocene tufa. Journal of Quaternary Science, 19, 251–262. G HELARDONI , R., G IANNINI , E. & N ARDI , R. 1968. Ricostruzione palaeogeografica dei bacini neogenici e quaternari nella bassa valle dell’Arno sulla base dei sondaggi e dei rilievi sismici. Memorie della Societa` Geologica Italiana, 7, 91–106. G IRAUDI , C. 1989. Lake levels and climate for the last 30,000 years in the Fucino area (Abruzzo-Central Italy) – a review. Palaeogeography, Palaeoclimatology, Palaeoecology, 70, 249–260. G IRAUDI , C. 2000. Le oscillazioni di livello del Lago di Vico (Lazio) nell’ambito delle variazioni climaticoambientali dell’Italia centrale. Italian Journal of Quaternary Sciences, 13(1/2), 21–30. G IRAUDI , C. 2001. The Late Pleistocene and Holocene temporary lakes in the Abruzzo Parks and the Central Apennine. In: V ISCONTI , G., B ENISTON , M., I ANNORELLI , E. & B ARBA , D. (eds) Global Change and Protected Areas. Advances in Global Changes Research, 9, 445 –458. G IRAUDI , C. 2004. Le oscillazioni di livello del Lago di Mezzano (Valentano – VT): variazioni climatiche ed influenza antropica. Italian Journal of Quaternary Sciences, 17(2/1), 221–230. G LOVER , C. & R OBERTSON , A. H. F. 2003. Origin of tufa (cool-water carbonate) and related terraces in the Antalya area, SW Turkey. Geological Journal, 38, 329–358. G OLUBIC , S., V IOLANTE , C., F ERRERI , V. & D’A RGENIO , B. 1993. Algal control and early diagenesis in Quaternari formation (Rocchetta a Volturno, Central Apennines). In: B ARATTOLO , F. ET AL . (eds) Studies on Fossil Benthic Algae. Bollettino della
279
Societa` Palaeontologica Italiana, Special Volume 1, 231– 247. G OUDIE , A. S., V ILES , H. A. & P ENTECOST , A. 1993. The late-Holocene tufa decline in Europe. The Holocene, 3(2), 181– 186. G RIFFITHS , H. I. & P EDLEY , H. M. 1995. Did changes in the late last glacial and early Holocene atmosphere CO2 concentrations control the rates of tufa precipitation? Holocene, 5, 238. G UO , L. & R IDING , R. 1998. Hot-spring travertine facies and sequences, Late Pleistocene, Rapolano Terme, Italy. Sedimentology, 45, 163–180. H ANCOCK , P. L., C HALMERS , R. M. L., A LTUNEL , E. & C ¸ AKIR , Z. 1999. Travitonics: using travertines in active fault studies. Journal of Structural Geology, 21, 903 –916. H EMMING , S. R. 2004. Heinrich events: massive late Pleistocene detritus layers of the North Atlantic and their global climate imprint. Reviews of Geophysics, 42(1), 1–43. H ENNING , G. J., G RU¨ N , R. & B RUNNACKER , K. 1983. Speleothems, Travertines and Palaeoclimates. Quaternary Research, 20, 1– 29. H ORVATINCˇ IC , N., Cˇ ALIC , R. & G EYH , M. A. 2000. Interglacial growth of Tufa in Croatia. Quaternary Research, 53, 185–195. I HLENFELD , C., N ORMAN , M. D., G AGAN , M. K., D RYSDALE , R. N., M AAS , R. & W EBB , J. 2003. Climatic significance of seasonal trace element and stable isotope variations in a modern freshwater tufa. Geochimica et Cosmochimica Acta, 67(13), 2341–2357. ISSC. 1987. Unconformity-bounded stratigraphic units. Geological Society American Bulletin, 98(2), 232– 237. J ULIA` , R. 1983. Travertines. In: S CHOLLE , P. A., B EBOUT , D. G. & M OORE , C. H. (eds) Carbonate Depositional Environments. AAPG Memoir, 33, 62–72. L ANG , J., P ASCAL , A. & S ALOMON , J. 1992. Caracte´risation pe´trographique de divers carbonates continentaux holoce´nes du Jura franc¸ais (Arbois, Chalain, Dortan). Implications pale´oge´ographiques. Zeitschrift fuer Geomorphologie N.F., 36, 273– 291. L ENG , M. J. & M ARSHALL , J. D. 2004. Palaeoclimate interpretation of stable isotope data from lake sediment archives. Quaternary Science Reviews, 23, 811–831. L IVNAT , A. & K RONFELD , J. 1985. Palaeoclimatic implications of U-series dates for lake sediments and travertines in the Arava Rift Valley, Israel. Quaternary Research, 24, 164–172. L OWE , J. J., A MMAN , B. ET AL . 1994. Climatic changes in areas adjacent to the North Atlantic during the last glacial-interglacial transition (14–9 ka BP): a contribution to IGCP-253. Journal of Quaternary Science, 9, 185– 198. M AGNY , M., DE B EAULIEU , J. L. ET AL . 2006. Holocene climate changes in the central Mediterranean as recorded by lake-level fluctuations al Lake Accesa (Tuscany, Italy). Quaternary Science Reviews, 26, 1736– 1758. M AGNY , M., M IRAMONT , C. & S IVAN , O. 2002. Assessment of the impact of climate and anthropogenic factors on Holocene Mediterraneean vegetation in Europe on the basis of palaeohydrological records.
280
E. CAPEZZUOLI ET AL.
Palaeogeography, Palaeoclimatology, Palaeoecology, 186, 47–59. M AGRI , D. 1999. Late-Quaternary vegetation history at Lagacione near Lago di Bolsena (central Italy). Review of Palaeobotany and Palynology, 106, 171– 208. M AGRI , D. & S ADORI , L. 1999. Late Pleistocene and Holocene pollen stratigraphy at Lago di Vico (central Italy). Vegetation History and Archeobotany, 8, 247– 260. M AKHNACH , N., Z ERNITSKAYA , V., K OLOSOV , I., D EMENEVA , O. & S IMAKOVA , G. 2000. d18O and d13C in calcite of freshwater carbonate deposits as indicators of climatic and hydrological changes in the Late-Glacial and Holocene in Belarus. Journal of Geochemical Exploration, 69–70, 435 –439. M AKHNACH , N., Z ERNITSKAJA , V., K OLOSOV , I. & S IMAKOVA , G. 2004. Stable oxygen and carbon isotopes in Late Glacial– Holocene freshwater carbonates from Belarus and their palaeoclimatic implications. Palaeogeography, Palaeoclimatology, Palaeoecology, 209, 73–101. M ANGERUD , J., A NDERSEN , S. T., B ERGLUND , B. E. & D ONNER , J. J. 1974. Quaternary stratigraphy of Norden, a proposal for terminology and classification. Boreas, 4, 109–128. M ARTI` N -A LGARRA , A., M ARTI` N -M ARTI` N , M., A NDREO , B., J ULIA` , R. & G ONZALEZ -G OMEZ , C. 2003. Sedimentary patterns in perched springline travertines near Granada (Spain) as indicators of the palaeohydrological and palaeoclimatical evolution of a karst massif. Sedimentary Geology, 161, 217– 228. M ATSUOKA , J., K ANO , A., O BA , T., W ATANABE , T., S AKAI , S. & S ETO , K. 2001. Seasonal variation of stable isotopic compositions recorded in a laminated tufa, SW Japan. Earth and Planetary Science Letters, 192, 31–44. M C D ERMOTT , F. 2004. Palaeo-climate reconstruction from stable isotope variations in speleothems: a review. Quaternary Science Reviews, 23, 901–918. M ERLA , G. & B ORTOLOTTI , V. 1967. Note illustrative della Carta Geologica d’Italia alla scala 1:100.000. Foglio 113 – Castelfiorentino. Nuova Tecnica Grafica, Roma. M INISSALE , A. 2004. Origin, transport and discharge of CO2 in central Italy: Earth-Science Reviews, 66, 89–141. N ARCISI , B. 2001. Palaeoenvironmental and palaoeclimatic implications from the Late-Quaternary sediment record of Vico volcanic lake (Central Italy). Journal of Quaternary Science, 16(3), 245–255. N ARCISI , B. & A NSELMI , B. 1998. Sedimentological investigation on a Late Quaternary lacustrine core from the Lagaccione crater (Central Italy): palaeoclimatic and palaeoenvironmental inferences. Quaternary International, 47/48, 21–28. O RDO` N˜ EZ , S. & G ARCIA D EL C URA , M. A. 1983. Recent and tertiary fluvial carbonates in Central Spain. Special Publications of International Association of Sedimentologists, 6, 485–497. O RDO` N˜ EZ , S., G ONZALEZ M ARTI` N , J. A., G ARCI` A D EL C URA , M. A. & P EDLEY , H. M. 2005. Temperate and semi-arid tufas in the Pleistocene to Recent fluvial barrage system in the Mediterranean area: the
Ruidera Lakes Natural Park (Central Park). Geomorphology, 69, 332– 350. O RTIZ , J. E., T ORRES , T., D ELGADO , A., R EYES , E. & D`I AZ -B AUTISTA , A. 2009. A review of the Tagus river tufa deposits (central Spain): age and palaeoenvironmental record. Quaternary Science Reviews, doi: 10.1016/j.quascirev.2008.12.007. P EDLEY , H. M. 1990. Classification and environmental models of cool freshwater tufas. Sedimentary Geology, 68, 143 –154. P EDLEY , M. 2009. Tufas and travertines of the Mediterranean region: a testing ground for freshwater carbonate concepts and developments. Sedimentology, 56, 221–246. P EDLEY , M., G ONZALE` Z M ARTI` N , J. A., O RDON˜ E` Z D ELGADO , S. & G ARCIA D EL C URA , M. A. 2003. Sedimentology of Quaternary perched springline and paludal tufas: criteria for recognition, with examples from Guadalajara Province, Spain. Sedimentology, 50, 23–44. P EN˜ A , J. L., S ANCHO , C. & L OZANO , M. V. 2000. Climatic and tectonic significance of Late Pleistocene and Holocene Tufa deposits in the Mijares River canyon, Eastern Iberian Range, Northeast Spain. Earth Surface Processes and Landforms, 25, 1403– 1417. P ENTECOST , A. 1995. The quaternary travertine deposits of Europe and Asia Minor. Quaternary Science Reviews, 14, 1005– 1028. P ENTECOST , A. 2005. Travertine. Springer-Verlag, Berlin, Heidelberg. P ENTECOST , A. & V ILES , H. A. 1994. A review and reassessment of travertine classification. Geographie physique et Quaternaire, 48, 305–314. P ENTECOST , A. & W HITTON , B. A. 2000. Limestone. In: W HITTON , B. A. & P OTTS , M. (eds) The Ecology of Cynobacteria. Kluwer Academic Publisher, Netherland, 257–279. R AMRATH , A., Z OLITSCHKA , B., W ULF , S. & N EGENDANK , J. F. W. 1999. Late Pleistocene climatic variations as recorded in two Italian lakes (Lago di Mezzano, Lago Grande di Monticchio). Quaternary Science Reviews, 18, 7, 977–992. R ENAUT , R. W. & J ONES , B. 2000. Microbial precipitates around continental hot springs and geysers. In: R IDING , R. E. & A WRAMIK , S. M. (eds) Microbial Sediments. Springer, Berlin, 187 –195. R IDING , R. 1991. Classification of microbial carbonates. In: R IDING , R. (ed.) Calcareous Algae and Stromatolites. Springer-Verlag, Berlin, Heidelberg, 21–51. R USZKICZAY -R U¨ DIGER , ZS ., F ODOR , L., B ADA , G., ¨ SSY , SZ ., H ORVA` TH , E. & D UNAI , T. J. L EE` L -O 2005. Quantification of Quaternary vertical movements in the central Pannonian Basin: A review of chronologic data along the Danube River, Hungary. Tectonophysics, 410, 157– 172. S ADORI , L., G IRAUDI , C., P ETITTI , P. & R AMRATH , A. 2004. Human impact at Lago di Mezzano (central Italy) during the Bronze Age: a multidisciplinary approach. Quaternary International, 113/1, 5– 17. S ALVADOR , A. 1994. International Stratigraphic Guide. A Guide to Stratigraphic Classification, Terminology and Procedure. ISSC 2nd edn., The International
CALCAREOUS TUFA AS INDICATORS OF CLIMATIC VARIABILITY Union Geological Sciences and The Geological Society of America, Inc., Tulsa. S MITH , J. R., G IEGENGACK , R. & S CHWARCZ , H. P. 2004. Constraints on Pleistocene pluvial climates through stable-isotope analysis of fossil-spring tufas and associated gastropods, Kharga Oasis, Egypt. Palaeogeography, Palaeoclimatology, Palaeoecology, 206, 157– 175. S OLIGO , M., T UCCIMEI , P., B ARBERI , R., D ELITALA , M. C., M ICCADEI , E. & T ADDEUCCI , A. 2002. U/Th dating of freshwater travertine from Middle Velino Valley (Central Italy): palaeoclimatic and geological implications. Palaeogeography, Palaeoclimatology, Palaeoecology, 184, 147– 161. S RDOCˇ , D., O BELIC´ , B., H ORVATINCˇ IC´ , N. & S LIEPCˇ EVIC´ , A. 1983. Radiocarbon dating of tufa in palaeoclimatic studies. Radiocarbon, 25, 421–427. S TURCHIO , N. P., P IERCE , K. L., M URREL , M. T. & S OREY , M. L. 1994. Uranium-series ages of travertines and timing of the Last Glaciation in the Northern Yellowstone Area, Wyoming Montana. Quaternary Research, 41, 265– 277. V AUDOUR , J. 1986. Travertins holocenes et pression anthropique. Mediterranee, 10, 168– 173. V AUDOUR , J. 1994. Evolution Holocene des travertins de valle`es dans le Midi Mediterraneen franc¸ais. Geographie Physique et Quaternaire, 48, 315– 326.
281
V ILES , H. A. & G OUDIE , A. S. 1990. Tufas, travertines and allied carbonate deposits. Progress in Physical Geography, 14, 19–41. V IOLANTE , C., F ERRERI , V. & D’A RGENIO , B. 1996. Modificazioni geomorforfiche controllate dalla deposizione di travertino. Il Quaternario, 9(1), 213– 216. V IOLANTE , C., F ERRERI , V., D’A RGENIO , B. & G OLUBIC , S. 1994. Quaternary travertines at Rocchetta a Volturno (Isernia, Central Italy). Facies analysis and sedimentary model of an organogenic carbonate system. In: C ARANNANTE , G. & T ONIELLI , R. (eds) I.A.S. 15th Reg. Meet., April, 1994, Ischia, Guide book to the field trip, 3 –23. V ON G RAFENSTEIN , U., E ICHER , U., E RLENKEUSER , H., R UCH , P., S CHWANDER , J. & A MMANN , B. 2000. Isotope signature of the Younger Dryas and two minor oscillations at Gerzensee (Switzerland): palaeoclimatic and palaeolimnologic interpretation based on bulk and biogenic carbonates. Palaeogeography, Palaeoclimatology, Palaeoecology, 159, 215 –229. W ALKER , M. J. C. 2005. Quaternary Dating Method. Wiley, UK, 286. W ULF , S., K RAML , M., B RAUER , A., K ELLER , J. & N EGENDANK , J. F. W. 2004. Tephrochronology of the 100 ka lacustrine sediment record of Lago Grande di Monticchio (southern Italy). Quaternary International, 122, 7– 30.
Cave atmosphere controls on stalagmite growth rate and palaeoclimate records JAMES U. L. BALDINI Department of Earth Sciences, University of Durham, Durham DH1 3LE, UK (e-mail:
[email protected]) Abstract: Cave atmosphere PCO2 partially controls calcite deposition on stalagmites by changing the thermodynamic drive of drip water to deposit calcite. The dissolved carbon dioxide contained in karstic percolation water is generally controlled by the soil PCO2, and this CO2 will degas in any void spaces with a lower PCO2, including caves. If void space PCO2 is higher than the PCO2 of the water, dissolution may occur. Measured cave air PCO2 ranges of several caves in different climate regimes suggest that soil temperature is a major control on cave air PCO2, but that the observed trend deviates from the modelled trend when soil carbon dioxide production is moisture-limited. Calcite deposition models illustrate how soil and cave air PCO2 can influence stalagmite growth rates, and demonstrate how gradual temperature changes can skew the geochemical proxy signal in stalagmites in favour of certain seasons and eventually can result in total cessation of growth.
Speleothems, such as stalagmites, are critically important archives of climate, particularly in lowto mid-latitude terrestrial locations where glacial ice core palaeoclimatology is not possible. Additionally, records from stalagmites are dateable using high-precision U-series dating and can consequently assist in the dating of previously identified but poorly constrained climatic features (Wang et al. 2001; Spo¨tl & Mangini 2002; Genty et al. 2003), thus helping to determine correlations between different records. However, calcite growth rates may vary both on long and short timescales, consequently complicating the interpretation of stalagmite proxy records. Also, some stalagmites may be fed by drips that experience seasonal, or intermittent, undersaturation with respect to calcite, resulting in either the cessation of calcite deposition or actual dissolution of previously deposited calcite. Stalagmite climate records sampled at multiannual resolution integrate the annual climate signal and therefore potentially bias the climate signal towards the season with the highest deposition rate. Microanalytical techniques are allowing the creation of increasingly better resolved records, but it is critical that high frequency variability in the parameters that control calcite growth are known in order to maximise the value of these important proxy records. Although the basic controls on calcite deposition are well quantified, their spatiotemporal variability in caves is currently not well constrained. Drip rate, drip water temperature, soil temperature, soil moisture conditions, soil and cave air carbon dioxide partial pressure (PCO2), the Ca2þ concentration of drip water, and the thickness of the water film over the stalagmite all control calcite precipitation (Buhmann & Dreybrodt 1985a, b; Dreybrodt
1999), and although the behaviour of many of these is fairly well-constrained, variability in cave air PCO2 is not. Recent studies have demonstrated that intra and interannual variability in stalagmite growth and geochemical climate proxies contained within can be ascribed to rapid and seasonal shifts in cave air PCO2 (Frisia et al. 2000; Spo¨tl et al. 2005; Banner et al. 2007; Baldini et al. 2008; Mattey et al. 2008). Soil PCO2 is the primary control on dissolved CO2 in groundwater in both open and closed system conditions (White 1988). Closed system conditions occur when percolation water loses contact with the CO2 source, thus restricting the amount of limestone dissolution that can occur. Conversely, open system conditions are characterised by continued contact with the CO2 source, thus potentially generating much higher groundwater alkalinity values. However, in practice, most cave percolation waters are likely to have experienced both open and closed system conditions. Generally, soil PCO2 is substantially higher (typically 0.1–10.0% atm) than atmospheric values (c. 0.0385% atm) (Troester & White 1984; White 1988), and is largely responsible for the total dissolved CO2 contained in vadose water. Carbonate dissolution occurs until the dissolved carbon dioxide is completely consumed. The system then remains at equilibrium until the water reaches air-filled voids with lower PCO2 than that dissolved in the water (reflecting soil air PCO2), at which point dissolved CO2 degassing occurs, followed by calcite precipitation. Therefore, a reduction in cave atmosphere PCO2 would increase stalagmite growth rates, assuming all other variables remain unchanged. Previous studies have applied stalagmite growth rate directly as a palaeoclimate proxy (Polyak & Asmerom 2001; Baldini et al. 2002; Spo¨tl et al.
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 283–294. DOI: 10.1144/SP336.15 0305-8719/10/$15.00 # The Geological Society of London 2010.
284
J. U. L. BALDINI
2002; Frisia et al. 2003; Mattey et al. 2008), and recent studies also suggest that the seasonality of calcite deposition can affect stalagmite stable isotope signatures (Baldini et al. 2005; Treble et al. 2005; Baldini et al. 2008). Because stalagmite growth rates affect how other paleoclimate proxy records are recorded, fully understanding how cave atmosphere PCO2 fluctuates and what effects these shifts may have on stalagmite formation is critical. Previous researchers (Dreybrodt 1980; Buhmann & Dreybrodt 1985a, b; Dreybrodt 1996) used rate equations derived from the Plummer– Wigley–Parkhurst (PWP) equation (Plummer et al. 1978; Busenberg & Plummer 1987) to predict calcite deposition on stalagmites. Theoretical calcite accumulation rates on stalagmites can therefore be predicted based on multiple growth determining variables such as thin film thicknesses, drip rate, temperature, drip water [Ca2þ], and cave atmosphere PCO2. However, studies attempting to use stalagmites with well-constrained growth rates to verify these theoretical results were only moderately successful (Baker et al. 1998). One of the more poorly characterized growth determining variables at the studied cave sites was cave air PCO2, and better quantification through high resolution seasonal studies might yield higher correlations. Recent studies suggest that stalagmite growth rates can be modelled more accurately if annual ventilation and cave atmosphere PCO2 variations, along with hydrochemical parameters, are known (Baldini et al. 2008; Mattey et al. 2008). In this study, our current understanding of the origins, distribution, and temporal variability of cave air PCO2 is discussed, and the mechanisms through which cave air PCO2 affects stalagmite climate records are identified.
Previous research and technology Many previous researchers have measured PCO2 in caves but, until recently, very little research existed combining high resolution spatial and/or temporal PCO2 measurements. One difficulty was that the presence of an operator alters the PCO2 values of the measurement. The PCO2 of human respiration is approximately 4% atm, much higher than the ambient PCO2 of most cave sites. Studies illustrate that after five minutes a human presence in a cave raises the PCO2 by c. 30% (assuming an original PCO2 of 0.4% atm) (Ek & Gewelt 1985). If accurate measurements are important, it is therefore imperative to use a breathing apparatus to minimise the effects of respired PCO2, or to log PCO2 automatically without a human operator present. Gewelt & Ek (1983) published a
comparison of the spatial PCO2 variability in two Belgian caves where respired CO2 was absorbed by a breathing apparatus filled with sodium carbonate. A linear relationship existed between the distance from the cave entrances and cave air PCO2. Based on the CO2 distribution in the caves, the soil zone and an underground stream flowing through one of the caves were inferred as two CO2 sources. Another study presented data from Belgium and numerous other countries, and demonstrated that PCO2 is positively correlated with aboveground temperature (Ek & Gewelt 1985) and that PCO2 concentrations are higher near the ceiling of passages. A study of the Aven d’Orgnac in France suggested that air enriched with biogenic CO2 moved through bedrock fissures into the cave (Bourges et al. 2001). Troester & White (1984) demonstrated that groundwater, in this case a cave stream, can act as either a source or a sink for CO2. Several studies have monitored cave air CO2 in tourist caves in order to mitigate damage to either speleothems or to pictographs (Pulido-Bosch et al. 1997; Faimon et al. 2004; Denis et al. 2005), and have shown that human respiration is a significant source of CO2 in these situations. A high-spatial resolution PCO2 survey was conducted for a small cave site in southern Ireland and demonstrated that cave air PCO2 increased not with depth below the ground, but with distance from the entrance (Fig. 2) (Baldini et al. 2006). It also suggested that CO2 at the site was principally derived from soil organic matter decay and root respiration. A five-month time-series PCO2 dataset from the same site confirmed this source, but also demonstrated that ventilation played a key role in modulating the amount of CO2 stored in the subsurface (Baldini et al. 2008). The importance of ventilation was also well illustrated in studies in Austria (Spo¨tl et al. 2005), Texas (Banner et al. 2007) and Gibraltar (Mattey et al. 2008). Cave air PCO2 can be measured by using meters containing a chemical (usually hydrazine hydrochloride) whose colour changes depending on the PCO2, or by using more precise electronic meters that use the absorptivity of infrared radiation by CO2 to calculate the PCO2. Most modern studies typically utilise the IR technology because it allows the logging of PCO2 over long periods of time rather than just obtaining a brief snapshot. They also permit automatic logging without the presence of an operator whose respiration might otherwise affect the accuracy of the measurement.
Cave air sources and circulation Identifying CO2 sources and sinks, and characterizing the dynamics of CO2 in cave air, are critical for modelling how cave air PCO2 might have varied in
CAVE ATMOSPHERES AND STALAGMITES
the past. In general, cave air PCO2 levels are controlled by the relative strengths of sources and sinks characteristic of each individual cave site (Fig. 1). Carbon dioxide in caves is sourced from: (1) vertical movement of gaseous CO2 derived from root respiration and organic matter decay through bedrock permeability; (2) degassing of soilderived CO2 from percolation waters and streams; (3) decomposition of organic material within the cave; (4) respiration of microbes and animals in the cave; and (5) geothermal activity. The mechanisms responsible for the first three sources are all temperature dependent, and therefore suggest that caves in warmer climates should have higher cave air PCO2 values than caves in colder climates, assuming no ventilation. These high cave air PCO2 values should result in lower growth rates in warmer caves; however, the opposite is often the case, and it is clear that soil PCO2 and subsequent bedrock dissolution are more important controls on calcite deposition for stalagmites fed by drips with a certain hydrology (discussed in more detail below). The source also controls the vertical distribution of CO2 in a cave passage. Slow degassing of CO2 from numerous drip sites produces localised high concentrations near the roof of a cave (Atkinson 1977), while more rapidly dripping water may degas preferentially nearer the floor. Vertical movement of soil gas through fissures produces higher concentrations near the roof if the flow is more diffuse through many small pores. Conversely, because CO2 is heavier than ‘average’ air, theoretically if the flow is more concentrated the gas will tend to behave as a fluid and flow downwards spilling over the floor of the passage (Berger 1988). Whether the gas diffuses at the ceiling or falls to the floor is a complex function of diffusion rate, fall height, diameter of the fall column, and the density of the gas (Berger 1988), though more field research is needed to confirm these theoretical predictions. Carbon dioxide derived from either the decomposition of organic matter within the cave or geothermal activity will tend to be found at higher concentrations near the floor. However, regardless of the CO2 source, if the CO2 influx ceases the vibrational energy of the gas molecules will eventually result in diffusion and molecular-mass independent mixing of the gases present, resulting in a homogenous gas mixture. Therefore occasionally elevated PCO2 levels in deeper passages may not be due to the high molecular mass of the CO2 compared to ‘average’ cave air, but rather may be due to a deepseated source or reduced advection in a sheltered location (Fig. 1). The latter is well illustrated in Grotte de Lascaux, where the Mondmilch Gallery has a PCO2 of between 1.0–1.5% atm while the slightly shallower but more sheltered Shaft of the
285
Dead Man has very high values of between 3 –7% atm, although these elevated values could also be due to localised micro-organism metabolism in sediments (Denis et al. 2005). Production of CO2 through the process of aerobic respiration can be identified by a mole-to-mole replacement of O2 by CO2 (e.g. if CO2 increases by 1% and O2 decreases by 1%). By far the most important CO2 sink in caves is the transfer of air out of the cave system by advection. This is usually readily identifiable as a seasonal (or permanent) reduction in cave air PCO2 to values approximating outside atmospheric values (currently c. 0.038% atm), and is commonly observed at many sites (Ek & Gewelt 1985; Baker & Genty 1998; Baldini et al. 2008). Ventilation is often driven by temperature or pressure differences between entrances. A good example of this is the dynamically ventilated Rassl–Bumslucke –O2J System in Austria, where cool cave air exits the cave during the warm season, but this airflow reverses when outside air is colder than cave air (Spo¨tl et al. 2005). While ventilation is more common in caves with multiple entrances, particularly when these are at different altitudes, caves with one entrance can still be ventilated. Ballynamintra Cave (southern Ireland) appears to have active air circulation during the summer, which shuts off during the winter, though the trigger for the ventilation remains unknown (Baldini et al. 2008). Conversely, several caves in Texas appear to ventilate during the winter, when cold winter air displaces warmer cave air, initiating air circulation (Banner et al. 2007). Some cave sites may display very static cave air PCO2 levels, suggesting limited ventilation (Banner et al. 2007). In most caves with no ventilation cave air PCO2 levels can only increase as high as soil air PCO2. Higher levels could theoretically be reached through the consumption of atmospheric O2 and production of CO2 through micro-organism metabolism (Smith 1999), or through the addition of high levels of geothermally-derived CO2. Another possible sink for cave air CO2 is the absorption and dissolution of CO2 into cave streams, percolation water, or water condensed onto cave walls (i.e. cave waters that are not equilibrated with cave air PCO2). While potentially locally important, particularly in the case of a large cave stream composed of water with a very short mean residence time in the aquifer, in general these processes are of secondary importance to ventilation.
Estimating cave air PCO2 Baldini et al. (2008) used long time-series datasets from two cave sites to link cave air PCO2 to primary production and ventilation. Equation 1 is
286
J. U. L. BALDINI
Fig. 1. Schematic illustration of sources (black arrows) and sinks (white arrows) of CO2 in caves. These vary considerably in importance from site to site but generally consist of: (a) CO2 moving as a fluid from the soil zone to the cave through fissures and secondary permeability. The gas can then either diffuse from the roof downwards or sink to the floor, depending on the diameter of the fluid column. (b) Degassing from percolation water with a low percolation rate. This will lead to increased PCO2 near the roof. (c) Aerobic respiration by biota in the cave. (d) Bacterial decomposition of organic matter deposited within the cave. (e) Degassing from percolation water with a rapid flow rate. This will result elevated PCO2 nearer the floor. (f) Flow upward into the cave of geothermally-sourced CO2. (g) Degassing from a river or stream. This can also act as a sink, depending on the dissolved PCO2 of the stream water. (h) Cave air circulation and ventilation through a connection with the outside atmosphere. This is the most important sink of cave air CO2 at most sites.
CAVE ATMOSPHERES AND STALAGMITES
based on the assumption that cave air PCO2 is derived from the soil, where production is modulated by temperature, and that cave air PCO2 (in % atm) is influenced by ventilation: PCO2 ¼ (b) 0:318 b þb Tþb T 2 1 0 1 2 ð283:15 T1 Þ 8:314 0:03
(1)
where b is a ventilation-dependent scaling factor, T is temperature (K), and b0, b1 and b2 are estimated parameters of the quadratic function (1, 2101.1 and 0.1, respectively). The scaling factor (b) accounts for reduced PCO2 in well-ventilated caves or in locations within individual caves that are better ventilated. Based on linear regression between the minimum cave air PCO2 values and b values presented in Baldini et al. (2008), b can be estimated using:
b ¼ 10 (PCO2 min ) 0:1
(2)
This equation assumes that minimum cave air PCO2 value (in % atm) is a measure of total ventilation. Cave air PCO2 measurements from various cave systems that were the focus of various previously published investigations demonstrate that minimum values can approach atmospheric regardless of the total range in values measured (Fig. 2), suggesting that ventilation is actively occurring and producing a minimum b ¼ 0.28. Maximum values generally appear to be controlled by a combination of soil production and ventilation; in many records maxima occur during times of maximum soil CO2 production (usually controlled by temperature) and reduced ventilation. Some sites (e.g. Uamh an Tartair, Scotland) appear to be well ventilated (Fuller 2006), preventing CO2 accumulation and maintaining low cave air PCO2 values. Conversely, poorly ventilated sites (e.g. the Caverns of Sonora) maintain elevated values year round (Banner et al. 2007). The greatest range in cave air PCO2 is found in caves that experience seasonal ventilation combined with high soil temperatures, thereby increasing the soil-to-cave CO2 flux through either drip water degassing or by physical connections between the soil and the cave (Fig. 2). Equations 1 and 2 can be used to estimate cave air PCO2 for a number of sites with published time-series PCO2 data in order to compare predicted to actual values. Overall, Equation 1 predicts values very well, but certain limitations in the model are also apparent (Fig. 2). Of the 20 sites investigated, the model-derived estimates are above the range of observed values for four sites (GB Cave, England; Trou Joney Cave, Belgium; St. Anne Cave,
287
Belgium; Caverns of Sonora, Texas). The model overestimates cave air PCO2 for the well-ventilated site GB Cave by 41% (deviation from maximum measured values), but the PCO2 dataset is limited (n ¼ 3) and intervals with higher PCO2 values could have been missed. The model also slightly overestimates cave air PCO2 values for Trou Joney Cave and St. Anne Cave by 22% and 9%, respectively. The greatest deviation between the modelled and actual values occurs for the Caverns of Sonora, where modelled values are 82% higher than the maximum observed value (Banner et al. 2007). This suggests that maximum soil CO2 production at the site is limited not only by temperature, but also by moisture. The Caverns of Sonora are located in the geographic area with the lowest average annual precipitation (c. 41 cm yr21) of any of the 20 sites investigated here, supporting this interpretation. Equation 1 appears to predict cave air PCO2 values well if temperature and cave ventilation strength can be estimated. The equation was derived from empirical data collected at two temperate sites where soil CO2 production was never moisture limited; consequently the equation appears to breaks down when soil –CO2 production is limited by a factor besides temperature (e.g. moisture limited CO2 production in soils above the Caverns of Sonora). Nonetheless, the cave air PCO2 values predicted using Equation 1 are within the range of previously observed values for 16 out of 20 cave sites, suggesting that with further work these relationships could be used to estimate palaeocave PCO2 values from estimated temperatures.
Cave air control on cave calcite growth Quantifying how cave air PCO2 behaves is important for understanding the systematics of calcite deposition on stalagmites. Drip rate, temperature, drip water [Ca2þ], cave atmosphere PCO2, and the thickness of the thin film of water covering the stalagmite all control calcite deposition rates (Buhmann & Dreybrodt 1985a; Dreybrodt 1999; Genty et al. 2001). The following equation theoretically describes stalagmite growth rates (Baker et al. 1998; Dreybrodt 1999; Baldini et al. 2008): Ro ¼ 1:174103 (CaCaeq ) 1
(d DT 1 )b1 e(aDTd ) c where:
(3)
Ro is the extension rate (mm yr21) 1.174 103 is a conversion constant used to change molecular accumulation rates (mmol mm22 s21) into growth rates (mm yr21)
288
J. U. L. BALDINI
Fig. 2. Measured PCO2 ranges from a selection of caves with available time series measurements, across a variety of temperatures (Ek et al. 1968; Ek 1979; Troester & White 1984; Ek & Gewelt 1985; Baker & Genty 1998; Bourges et al. 2001; Huang et al. 2001; Spo¨tl et al. 2005; Banner et al. 2007; Baldini et al. 2008; Mattey et al. 2008). The approximate mean annual temperature of the cave is shown along the right hand side of the figure, but this ‘y-axis’ is not to scale. The dashed line indicates current outside air mean PCO2 of approximately 0.0038%. The grey filled circles are PCO2 estimates derived by using Equation 1. To estimate the ventilation factor b the minimum PCO2 values measured at each site were used in Equation 2.
CAVE ATMOSPHERES AND STALAGMITES
Caeq is the drip water [Ca2þ] at equilibrium with a given atmospheric PCO2 Ca is the initial water [Ca2þ] (mmol L21) d is the thin film thickness (mm) DT is the time between successive drips (s) a is a ‘kinetic constant’ (mm s21) that is dependent on d and temperature. The only term in Equation 2 affected by ambient PCO2 is Caeq, according to: Caapp ¼ 7:611 P 0:275 CO2 which is derived from datasets and relationships presented in previous research (Dreybrodt 1996; Kaufmann 2003). The thickness of the thin film of water covering stalagmites (d) varies between different stalagmites depending on their surface microtopography and slope, and is an important control on stalagmite growth rate. Estimated thin film thicknesses are believed to typically vary between 0.1–0.4 mm. Consequently, a principal uncertainty in growth rate modelling concerns the ambiguity associated with the d value of individual stalagmite samples. If cave air PCO2 spatial distribution is known throughout an entire cave (Baldini et al. 2006) the spatial calcite deposition rates can be modelled for that site (Fig. 3). Previous work (Baldini et al. 2006) created a snapshot of PCO2 throughout a small cave in Ireland (Ballynamintra Cave), and the values obtained in that work will be used here to model growth rates at the same site. If drip water [Ca2þ] ¼ 2.5 mmol L21, d ¼ 0.1 mm, drip rate ¼ one drip every 60 seconds, and temperature ¼ 10 8C, maximum stalagmite growth rates would occur near the entrance (c. 240 microns per year) and the lowest would occur near the back of the cave (c. 180 microns per year). However, if the drip water [Ca2þ] is lowered to 1.2 mmol L21, the growth rate will be reduced to c. 55 microns per year near the entrance, and the drip water would actually be undersaturated (with respect to calcite) near the back of the cave, potentially resulting in calcite dissolution. This illustrates both the importance of cave air PCO2 and the importance of drip water [Ca2þ]; in fact the major assumption is that drip water [Ca2þ] remains spatially invariable. This is not likely to be valid; although drip water [Ca2þ] spatial variability is not known for the site the overlying soil becomes thicker over the cave further away from the entrance, and the cave is also deepest further away from the entrance (i.e. more overburden). It is therefore predicted that drip water [Ca2þ] increases with distance from the entrance at this site. This is corroborated by the prevalence of calcite deposition in the zone of highest PCO2. Clearly, in this case, cave air PCO2 is
289
only a secondary modulating factor while drip water [Ca2þ] largely controls the location of calcite deposition. Additionally, the area near the entrance that is modelled to have the highest growth rates probably experiences high rates of condensation corrosion, likely offsetting any calcite deposition in this area (Fig. 3). Condensation corrosion is the process where water condenses from warm, humid ‘outside’ air onto colder rock surfaces in the cave. This water then absorbs CO2 from the surrounding air, forming carbonic acid and dissolving the calcite (Tarhule-Lips & Ford 1998). Variability in cave air PCO2 is therefore only a secondary factor in determining the spatial distribution of calcite deposition throughout a cave, but exerts a more important effect at specific sites through time (e.g. an individual stalagmite). Soil air PCO2 controls the amount of calcite dissolution soil water is able to achieve and consequently the [Ca2þ] of the drip water feeding the stalagmite. High soil temperatures (and ideal moisture conditions) increase the amount of soil bioproductivity and consequently the amount of soil air PCO2; waters derived from warmer soils therefore have higher [Ca2þ] than those derived from cooler soils and are therefore able to deposit more calcite on a stalagmite, assuming all other conditions are the a same (e.g. concentrations of growth inhibitors, etc.). However, this is modulated by cave air PCO2, which is also typically elevated (discouraging calcite growth) during warmer periods. These two temperature effects therefore affect calcite deposition in opposite directions, and recent research predicted (Baldini et al. 2008) that different drip types would be affected in different way. A ‘diffuse’ drip site whose drip water has a long residence time in the aquifer tends to have timeaverage hydrochemistry, while a responsive ‘seasonal’ drip reflects the current weather conditions more faithfully. Therefore a stalagmite fed by a diffuse flow drip was predicted to have elevated growth rates during the winter, because drip hydrochemistry is invariant and cave air PCO2 is generally lower in the winter. Conversely a more responsive drip might have elevated growth rates during the summer, when very high drip water [Ca2þ] compensates for higher cave air PCO2 and results in more calcite deposition. Calcite growth resulting from a hypothetical drip that responded to rainfall conditions was modelled here (Figs 4 & 5) for 20 years of deposition. The starting temperature conditions were set to a summer high temperature of 10.9 8C and a winter low temperature of 8.9 8C (mean annual T ¼ 9.9 8C), drip rate was left constant at 1 drip per 100 seconds, and the thin film thickness was 320 microns. The temperature was decreased by 1.7 8C per year, evenly throughout the year
290
J. U. L. BALDINI
Fig. 3. (a) The measured spatial variability in PCO2 in Ballynamintra Cave, Ireland (Baldini et al. 2006). (b) The calcite precipitation rate calculated assuming the PCO2 distribution shown in panel A, [Ca2þ] ¼ 2.5 mmol L21, a drip rate of one drip every 60 seconds, cave temperature ¼ 10 8C, a kinetic constant (a) ¼ 0.0001275 mm s21, and a thin film thickness (d) ¼ 0.1 mm. All growth determining variables were kept constant spatially except for the PCO2 : Growth rates were calculated using the equations presented in Baldini et al. (2008),which were adapted from previous research (e.g. Buhmann & Dreybrodt 1985a; Dreybrodt 1999; Kaufmann 2003). (c) The calcite precipitation rate calculated using all the same values for growth determining variables as for panel B, with the exception of [Ca2þ] which was reduced to 1.25 mmol L21.
CAVE ATMOSPHERES AND STALAGMITES
291
Fig. 4. The drip water d18O (top panel) and temperature (middle panel) for a modelled drip at a hypothetical cave site. The temperature variability shown controls the PCO2 of soil air and consequently bedrock dissolution and drip water [Ca2þ], as well as cave atmosphere PCO2. These two factors influence growth rate in opposite directions for seasonal flow sites (Baldini et al. 2008) but because the [Ca2þ] is the more important control on calcite deposition rate this controls the overall growth rate (lower panel). Growth rate is therefore higher during the summer and lower during the winter. The grey shading indicates times where the model indicates that no calcite deposition is occurring. All calcite deposition ceases after model year 18 when mean annual temperature drops below 7.6 8C.
(Fig. 4). Drip water [Ca2þ] was modelled as reflecting soil temperature using the relationships discussed in Baldini et al. (2008). The drip water d18O was also modelled as varying annually, with high values in the summer of 25‰ SMOW and low values of 29‰ SMOW occurring in the winter. This annual d18O variability was left unchanged throughout the model. Calcite growth rates are highest in the summer due to increased summer drip water [Ca2þ] which overcomes the reduced thermodynamic drive to degas (due to elevated cave air PCO2). Deposition rates on the hypothetical stalagmite are lowest in the winter. Through the course of the model the gradual reduction of winter temperature eventually
causes winter deposition to cease altogether (once temperature drops below 8.4 8C). After 18 years of modelled growth, even summer temperatures become too cold and no growth occurs at any time of the year (Fig. 4). This gradual reduction in growth followed by complete cessation is what is commonly observed in real samples (e.g. Genty et al. 2003).
Effects on palaeoclimate proxies Although cave air PCO2 does not directly affect stalagmite d18O (except in cases of extreme degassing), it does affect the seasonality of calcite deposition; if drip water d18O varies seasonally
292
J. U. L. BALDINI
Fig. 5. The calcite deposition rate plotted against height of the stalagmite (top panel). Numbers represent the model years as in Figure 4. The actual d18O that would result from a combination of the d18O of the drip water used in the model and the growth rate (middle panel) as well as the d18O obtained (lower panel) if the stalagmite were sampled using a conventional microdrill (long lines label with ‘C’, 500 micron integration, 500 micron steps) and a micromill (short lines labelled with ‘m’, 100 micron integration, 100 micron step size). The grey shaded area reflects the actual d18O of the precipitated calcite.
than the annually integrated d18O record could be biased toward the d18O of the season with the most deposition. This is illustrated by the deposition model discussed above, where, although the annual d18O cycle of the percolation water remains invariant through the model (Fig. 4), the d18O of the calcite changes. This effect is particularly strong once winter calcite deposition ceases and an increasing percentage of calcite reflects summer percolation water d18O (Fig. 5). The overall trend is therefore a slowing in growth concomitant with an increase in d18O heading toward summer rainfall values. Microanalytical techniques may only be able to detect a small fraction of the actual annual d18O signal, but this depends on the resolution of the technique and the growth rate of the stalagmite. Palaeoclimate records may therefore
underestimate the actual annual range in d18O values (Fig. 5). Although conservative proxies are only affected by the timing of calcite deposition relative to the timing of the geochemical proxy’s annual cycle, non-conservative proxies are directly affected by either the degassing or more indirectly by growth rate effects. Carbon isotopes are probably the most commonly used non-conservative proxies in stalagmites, and are directly affected by CO2 degassing. Carbon dioxide composed of 12C is preferentially degassed over that composed of 13C, thus increasing the d13C of drip water dissolved inorganic carbon (DIC). This leads to increased d13C of calcite deposited from the water. It should also be noted that growth inhibitors, such as certain colloids, organic molecules, and
CAVE ATMOSPHERES AND STALAGMITES
ions may also attenuate the growth of fast growing crystal faces. For example, free orthophosphate ions may inhibit the growth of certain types of calcite (Borsato et al. 2007; Dove & Hochella 1993). Because these ions and colloids are often preferentially released during certain seasons, crystallographic factors independent of cave air PCO2 may also influence calcite growth timing through the year.
Conclusions/future work Monitoring studies are continuing to clarify how cave air PCO2 changes through time. Production from the soil zone appears to be the most important source of CO2 in many temperate non-geothermal cave sites, but this is often overprinted by ventilation effects. Caves in locations with the highest mean annual temperature also typically have the highest maximum cave air PCO2 values. There is no apparent relationship between the maximum values at individual cave sites and the minimum values, which typically appear to be controlled by ventilation. Cave air PCO2 affects stalagmite growth rate directly by controlling the differential between soil and cave air PCO2 and consequently the percolation water’s thermodynamic drive to degas. For drip sites where the hydrochemistry is responsive to recent weather conditions temperature-dependent soil CO2 production increases percolation water DIC and consequently affects the amount of carbonate rock dissolution. This then raises the [Ca2þ] of the percolation water which increases stalagmite growth rate during seasons where elevated cave atmosphere PCO2 typically discourages calcite deposition. However, annual- to sub-annual scale variability in stalagmite proxy records fed by very time-averaged (‘diffuse’ or ‘seepage’) drips may be almost completely modulated by shifts in cave air PCO2. Further work needs to clarify the relationship between climate and cave air PCO2. More time series records are necessary from caves in different climates to refine existing models to include soil moisture content. Furthermore, more monitoring studies should attempt to verify the relationship between calcite deposition and all the calcite growth determining variables. Because of the importance of cave air PCO2 variability on the emplacement of stalagmite geochemical proxy records, measuring cave air composition at any cave site of palaeoclimatic interest for at least one year would greatly assist with proxy record interpretation. Prof. Dave Mattey and Dr Silvia Frisia are thanked for their reviews and useful suggestions which helped improve the manuscript. Prof. Ian Fairchild and Dr Lisa Baldini are also thanked for useful discussions.
293
References A TKINSON , T. C. 1977. Carbon dioxide in the atmosphere of the unsaturated zone: an important control of groundwater hardness in limestones. Journal of Hydrology, 35, 111– 123. B AKER , A. & G ENTY , A. 1998. Environmental pressures on conserving cave speleothems: effects of changing surface land use and increased cave tourism. Journal of Environmental Management, 53(2), 165–175. B AKER , A., G ENTY , D., D REYBRODT , W., G RAPES , J. & M OCKLER , N. J. 1998. Testing theoretically predicted stalagmite growth rates with recent annually laminated samples: implications for past stalagmite deposition. Geochimica et Cosmochimica Acta, 62, 393– 404. B ALDINI , J. U. L., B ALDINI , L. M., M C D ERMOTT , F. & C LIPSON , N. 2006. Carbon dioxide sources, sinks, and spatial variability in shallow temperate zone caves: evidence from Ballynamintra Cave, Ireland. Journal of Cave and Karst Studies, 68(1), 4 –11. B ALDINI , J. U. L., M C D ERMOTT , F., B AKER , A., B ALDINI , L. M., M ATTEY , D. P. & R AILSBACK , L. B. 2005. Biomass effects on stalagmite growth and isotope ratios: A 20th century analogue from Wiltshire, England. Earth Planetary Science Letters, 240, 486–494. B ALDINI , J. U. L., M C D ERMOTT , F. & F AIRCHILD , I. J. 2002. Structure of the 8200-year cold event revealed by a speleothem trace element record. Science, 296(5576), 2203– 2206. B ALDINI , J. U. L., M C D ERMOTT , F., H OFFMANN , D. L., R ICHARDS , D. A. & C LIPSON , N. 2008. Very highfrequency and seasonal cave atmosphere PCO2 variability: implications for stalagmite growth and oxygen isotope-based paleoclimate records. Earth and Planetary Science Letters, 272(1 –2), 118– 129. B ANNER , J. L., G UILFOYLE , A., J AMES , E. W., S TERN , L. A. & M USGROVE , M. 2007. Seasonal variations in modern speleothem calcite growth in Central Texas, USA. Journal of Sedimentary Research, 77(7–8), 615 –622. B ERGER , J. 1988. Do heavy gases fall? European Journal of Physics, 9, 335. B ORSATO , A., F RISIA , S., F AIRCHILD , I. J., S OMOGYI , A. & S USINI , J. 2007. Trace element distribution in annual stalagmite laminae mapped by micrometerresolution X-ray fluorescence: implications for incorporation of environmentally significant species. Geochimica et Cosmochimica Acta, 71(6), 1494– 1512. B OURGES , F., M ANGIN , A. & D ’H ULST , D. 2001. Le gaz carbonique dans la dynamique de l’atmosphe`re des cavite´s karstique: l’exemple de l’Aven d’Orgnac (Arde`che). Earth and Planetary Sciences, 333, 685– 692. B UHMANN , D. & D REYBRODT , W. 1985a. The kinetics of calcite dissolution and precipitation in geologically relevant situations of karst areas. 1. Open System. Chemical Geology, 48(1 –4), 189–211. B UHMANN , D. & D REYBRODT , W. 1985b. The kinetics of calcite dissolution and precipitation in geologically relevant situations of karst areas. 2. closed system. Chemical Geology, 53(1 –2), 109–124.
294
J. U. L. BALDINI
B USENBERG , E. & P LUMMER , L. N. 1987. A comparative study of the dissolution and crystal growth kinetics of calcite and aragonite. U.S. Geological Survey Bulletin, 1578, 139– 168. D ENIS , A., L ASTENNET , R., H UNEAU , F. & M ALAURENT , P. 2005. Identification of functional relationships between atmospheric pressure and CO2 in the cave of Lascaux using the concept of entropy of curves. Geophysical Research Letters, 32(L05810), doi: 10.1029/2004GL022226. D OVE , P. M. & H OCHELLA , M. F. 1993. Calcite precipitation mechanisms and inhibition by orthophosphate in situ observations by scanning force microscopy. Geochimica et Cosmochimica Acta, 57(3), 705– 714. D REYBRODT , W. 1980. Deposition of calcite from thin films of natural calcareous solutions and the growth of speleothems. Chemical Geology, 29, 89–105. D REYBRODT , W. 1996. Keynote Lecture: Chemical kinetics, speleothem growth, and climate. Climate Change – The Karst Record symposium, University of Bergen, Norway. D REYBRODT , W. 1999. Chemical kinetics, speleothem growth and climate. Boreas, 28, 347– 356. E K , C. 1979. Variations saisonnieres des teneurs en CO2 d’une grotte Belge: Le Trou Joney a Comblainau-Pont. Annales de la Socie´te´ Geologique de Belgique, 102, 71– 75. E K , C., D ELECOUR , F. & W EISSEN , F. 1968. Teneur en CO2 de l’air de quelques grottes Belges. Technique employee et premiers resultats. Annales de Speleologie, 23(1), 243– 257. E K , C. & G EWELT , M. 1985. Carbon dioxide in cave atmospheres – new results in Belgium and comparison with some other countries. Earth Surface Processes and Landforms, 10(2), 173– 187. F AIMON , J., Z IMA´ K , J., Z AJI´ Cˇ EK , P., S CHWARZOVA´ , M. & Sˇ TELCL , J. 2004. The study of recent destructive processes in the Moravian karst caves (Czech Republic). Geographica, 38, 9– 13. F RISIA , S., B ORSATO , A., F AIRCHILD , I. J. & M C D ERMOTT , F. 2000. Calcite fabrics, growth mechanisms, and environments of formation in speleothems from the Italian Alps and southwestern Ireland. Journal of Sedimentary Research, 70(5), 1183–1196. F RISIA , S., B ORSATO , A., P RETO , N. & M C D ERMOTT , F. 2003. Late Holocene annual growth in three Alpine stalagmites records the influence of solar activity and the North Atlantic Oscillation on winter climate. Earth and Planetary Science Letters, 216, 411–424. F ULLER , L. 2006. High-resolution multiproxy geochemical Holocene climate records from 1000-year old Scottish stalagmites. PhD Thesis, The University of Birmingham. G ENTY , D., B AKER , A. & V OKAL , B. 2001. Intra- and inter-annual growth rate of modern stalagmites. Chemical Geology, 176, 191 –212. G ENTY , D. ET AL . 2003. Precise dating of DansgaardOeschger climate oscillations in western Europe from stalagmite data. Nature, 421, 833–837. G EWELT , M. & E K , C. (eds) 1983. L’evolution saissonie`rre de la teneur en CO2 de l’air de deux grottes
Belges: Ste-Ann et Brialmont, Tilff. New Directions in Karst. Geo Books, Norwich. H UANG , Y. ET AL . 2001. Seasonal variations in Sr, Mg and P in modern speleothems (Grotta di Ernesto, Italy). Chemical Geology, 175(3 –4), 429– 448. K AUFMANN , G. 2003. Stalagmite growth and palaeoclimate: the numerical perspective. Earth and Planetary Science Letters, 214, 251–266. M ATTEY , D. ET AL . 2008. A 53 year seasonally resolved oxygen and carbon isotope record from a modern Gibraltar speleothem: reconstructed drip water and relationship to local precipitation. Earth and Planetary Science Letters, doi: 10.1016/j.epsl.2008.01.051. P LUMMER , L. N., W IGLEY , T. M. L. & P ARKHURST , D. L. 1978. The kinetics of calcite dissolution in CO2 – water systems at 58 to 60 8C and 0.0 to 1.0 ATM CO2. American Journal of Science, 278, 179–216. P OLYAK , V. J. & A SMEROM , Y. 2001. Late Holocene climate and cultural changes in the southwestern United States. Science, 294, 148 –151. P ULIDO -B OSCH , A., M ARTIN -R OSALES , W., L OPEZ C HICANO , M., R ODRIGUEZ -N AVARRO , C. M. & V ALLEJOS , A. 1997. Human impact in a tourist karstic cave (Aracena, Spain). Environmental Geology, 31(3– 4), 142 –149. S MITH , G. K. 1999. Foul air in limestone caves and its effects on cavers. Australian Speleological Federation 22nd Biennial Conference. S PO¨ TL , C., F AIRCHILD , I. J. & T OOTH , A. F. 2005. Cave air control on dripwater geochemistry, Obir Caves (Austria): implications for speleothem deposition in dynamically ventilated caves. Geochimica et Cosmochimica Acta, 69(10), 2451– 2468. S PO¨ TL , C. & M ANGINI , A. 2002. Stalagmite from Austrian Alps reveals Dansgaard-Oeschger events during isotope stage 3: implications for the absolute chronology of Greenland ice cores. Earth and Planetary Science Letters, 203, 507– 518. S PO¨ TL , C., M ANGINI , A., N ORBERT , F., E ICHSTADTER , R. & B URNS , S. J. 2002. Start of the last interglacial period at 135 ka: Evidence from a high Alpine speleothem. Geology, 30(9), 815– 818. T ARHULE -L IPS , R. F. A. & F ORD , D. C. 1998. Condensation corrosion in caves on Cayman Brac and Isla de Mona. Journal of Cave and Karst Studies, 60(2), 84–95. T REBLE , P. C., C HAPPELL , J., G AGAN , M. K., M C K EEGAN , K. D. & H ARRISON , T. M. 2005. In situ measurement of seasonal d18O variations and analysis of isotopic trends in a modem speleothem from southwest Australia. Earth and Planetary Science Letters, 233(1– 2), 17– 32. T ROESTER , J. W. & W HITE , W. B. 1984. Seasonal fluctuations in the carbon dioxide partial pressure in a cave atmosphere. Water Resources Research, 20(1), 153–156. W ANG , Y. J. ET AL . 2001. A high-resolution absolutedated late Pleistocene monsoon record from Hulu Cave, China. Science, 294, 2345– 2348. W HITE , W. B. 1988. Geomorphology and Hydrology of Karst Terrains. Oxford University Press, New York.
Petrology and geochemistry of annually laminated stalagmites from an Alpine cave (Obir, Austria): seasonal cave physiology ¨ TL2, SILVIA FRISIA3, ANDREA BORSATO4, IAN J. FAIRCHILD1*, CHRISTOPH SPO 5 JEAN SUSINI , PETER M. WYNN6, JEAN CAUZID5,7 & EIMF8 1
School of Geography, Earth and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
2
Institut fu¨r Geologie und Pala¨ontologie, Leopold-Franzens-Universita¨t Innsbruck, Innrain 52, 6020 Innsbruck, Austria 3
School of Environmental and Life Sciences, University of Newcastle, Callaghan, New South Wales, Australia 4
Museo Tridentino di Scienze Naturali, Via Calepina 14, Trento 38100, Italy
5
European Synchrotron Radiation Facility, F-38043 Grenoble Cedex, France
6
Department of Geography, University of Lancaster, Lancaster, LA1 4YQ, UK 7
8
G2R, Universite´ Henri Poincare´, 54506 Vandœuvre-le`s-Nancy, France
Edinburgh Ion Microprobe Facility (EIMF), Institute of Earth Science, School of Geosciences, University of Edinburgh EH9 3JW, UK *Corresponding author (e-mail:
[email protected]) Abstract: Seasonality is encoded in palaeoproxies of secondary cave mineral deposits (speleothems) and the code is becoming cracked. The petrology of calcite stalagmites from Obir, an Alpine (1100 m altitude), perennially wet cave, was characterized by optical and electron backscatter diffraction, and their chemistry by bulk ICP-MS analysis, ion microprobe and synchrotronbased micro-X-ray fluorescence. Vadose water penetrates 70 m through Triassic limestones (with some Pb– Zn mineralization) to the chamber Sa¨ulenhalle where the stalagmites were collected. Strong seasonal ventilation in the cave leads to low PCO2 in winter associated with falls in speleothem sulphate S and increase in d13C values. All samples display autumnal event lamination defined by a narrow, optically visible zone with increases in trace element concentrations, within which synchrotron studies have resolved mm-scale enrichments of Pb and Zn. Small-scale (10 mm) lateral trace element variations reflect alternate flat faces and rough crystal edges, influenced by high Zn content. The elemental covariations are consistent with the transport of Pb, Zn, P, F, Br and I adsorbed onto organic colloids in dripwater, but the final deposition may have been from aerosols and we propose this as a new mechanism requiring further investigation. This study represents the most complete demonstration of how chemical variations are powerful expressions of seasonal cave physiology in humid temperate caves, including the contrast between summer and winter conditions, and the preservation of sub-weekly events during the autumn season.
The current surge of interest in speleothems as recorders of environmental change (McDermott 2004; Fairchild et al. 2006a) has mostly focused on multimillennial stable isotope time series with typically scant description of speleothem petrology. Nevertheless, it is known that certain disequilibrium speleothem fabrics are associated with nonequilibrium geochemistry (Frisia et al. 2000) and more generally that there are wide variations in the trace element composition of speleothems at
different spatial scales (Roberts et al. 1998; Borsato et al. 2007). In particular, a series of advances over the past decade have taken us to the point where we have the outline of a process understanding of how the behaviour of a cave over an annual cycle can lead to diagnostic characteristics in these cave deposits. A distinct type of behaviour is displayed by caves that have a strong seasonal drought (Fairchild & McMillan 2007), whereas in this article we focus on phenomena which, based
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 295–321. DOI: 10.1144/SP336.16 0305-8719/10/$15.00 # The Geological Society of London 2010.
296
I. J. FAIRCHILD ET AL.
on studies in Scotland, northern China and Alpine settings for example, are typical for humid temperate environments (Fairchild et al. 2001; Frisia et al. 2005; Tan et al. 2006; Borsato et al. 2007; Baker et al. 2008). In these environments, there is a significant seasonality both of external temperature, which strongly influences cave circulation, and of infiltrating precipitation. This fluid transfer constitutes an important aspect of what can be termed the physiology of cave environments, and we examine how this physiology impacts on the formation of stalagmites. When comparing the behaviour of the cave systems with that of the resulting speleothems, it is necessary to be aware of the processes that fractionate isotopes and partition elements between solution and crystal. Specifically for trace elements, Fairchild et al. (2006a) and Fairchild & Treble (2009) considered three types of pattern. The first of these, the temperature-controlled pattern, does not concern us here as the cave under discussion does not display significant temperature variation on the annual scale. The fluid-dominated pattern refers to stalagmite patterns of change that directly reflect analogous changes in trace element composition (absolute abundance, flux, or ratio to calcium or carbonate) in the fluid. The crystal-dominated pattern refers to compositional changes dictated by crystallographic properties. Such properties could be functions of growth rate (and hence solution supersaturation), or more specifically pH, and so there can still be a link to the cave physiology. Here, we describe and discuss the petrography and geochemistry of stalagmites from Obir cave in Austria (Spo¨tl et al. 2005). Obir lies beneath an alpine forest at 1100 m altitude and the climatic context and ecological setting are comparable to that of the more extensively described Grotta di Ernesto in NE Italy (Fairchild et al. 2000, 2001; Frisia et al. 2000, 2003, 2005; Huang et al. 2001; Borsato et al. 2007; Wynn et al. 2010; Miorandi et al. in press) which lies 220 km to the WSW. Both caves contain stalagmites with distinct annual laminae as viewed in thin section. However, whereas Ernesto consists of a simple tube mostly 15 –30 m below the sloping ground surface with which it intersects, the Obir cave system is much more extensive, including a number of chambers, and our study site is 70 m below the surface. The caves display similar seasonal patterns of variation in PCO2 (Frisia et al. 2000; Spo¨tl et al. 2005) although the magnitude of air movement required to maintain steady values (that is by exchanging more CO2-rich internal air with outside air) is much lower at Ernesto because of its near-surface position. Spo¨tl et al. (2005) detailed the distinctive pattern of seasonal variation of air circulation at Obir based
on three years monitoring. The result was the formulation of a new model for d13C variations in which: (a) enhanced winter circulation leads to low PCO2; (b) dripwaters develop high d13C values by degassing isotopically light CO2; and (c) supersaturations for calcite are increased. The expectation would be that, during the winter, stalagmites should grow faster and display higher d13C values. Another parameter which might be expected to respond to changes in air circulation is sulphate since Busenberg & Plummer (1985) identified solution pH as an important control on its incorporation and dripwater pH is limited by the PCO2 of cave air, which varies seasonally as stated above. In a synchrotron study of stalagmite calcite from Ernesto cave, Frisia et al. (2005) interpreted annualscale variations of sulphate in this way, corresponding to an example of a crystal-dominated trace element pattern as introduced above. In this paper we test these ideas using data of d13C and sulphate variation within the annual scale. A second mode of chemical variation within the Obir stalagmites is also present, coinciding with the presence of optically visible laminae. Smith et al. (2009), in introducing a new statistical model for testing for annual variation of trace element properties, used 14C evidence to demonstrate that such laminae in stalagmite Obi84 from the chamber Sa¨ulenhalle were annual. They also showed that the laminae in this and two other similar stalagmites (Obi12 and Obi55) coincided with depletions in Sr and enrichments in several other trace elements. In this paper, we extend the range of determinands and note similarities in the type of elemental enrichments with annual laminae at Ernesto cave described by Borsato et al. (2007). These authors proposed that the enrichments, associated with the development of non-equilibrium crystal faces, were the result of element transport in association with soil-derived humic substances by high rates of water infiltration during the autumn period when vegetation is dying back. This is a waterdominated pattern, requiring a time-limited high flux of water to the stalagmite top, or a higher concentration of trace elements in the dripwater, or both, to explain the pattern. We look critically at the application of this model to Obir cave, making use of quantitative comparisons of dripwater and speleothem chemistry.
Methods Thin sections for petrography and ion microprobe analysis for stalagmite samples Obi12, 55 and 84 were polished and c. 150 mm thick. Lamina thickness measurement was carried out on a polarizing microscope using a graticule in which the smallest
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
division was 5 mm, making frequent use of photomosaics, and varying the optical conditions to optimize lamina visibility. In Obi84, the thickness of the last 140 laminae was measured in at least three traverses and the mean value determined. A representative piece of the top cm of sample Obi84 was ground into a powder in a pestle and mortar and analysed using a Siemens D5000 X-ray diffractometer in transmission geometry. Data were collected in 0.028 steps with a counting time of 11 s/step over the range 2u 108 to 808 using Cu Ka radiation. Rietveld refinement was carried out in the software GSAS (Larson & Von Dreele 2000). The electron backscatter diffraction (EBSD) analyses were acquired using the HKL Channel 5 analysis system on the Philips XL30CP scanning electron microscope (SEM) in the School of GeoSciences, University of Edinburgh. The SEM was operated at an accelerating voltage of 20 kV and a c. 2 nA beam at a working distance of 20 mm and a tilt angle of c. 708. The sample was given a final colloidal silica polish to remove surface crystal damage after mechanical polishing and the sample was used uncoated with the SEM operating at a controlled pressure of 0.1 mbar to minimise any charging effects. For each point the diffraction pattern was collected and solved by the software for the calcite crystal structure to within ,18. Maps for the orientation of the crystal structure of the sample were generated from point analyses collected every 1 mm over a 2 2 grid of maps each c. 500 mm wide and which were subsequently tiled together to form the final map. The whole map was then processed to highlight any variation in orientation of the crystal structure within the sample. Samples for stable oxygen and carbon isotope analysis were micromilled at 100 mm intervals for all three stalagmites (using methods as in Spo¨tl & Mattey 2006). Measurements were performed using an on-line, automated carbonate preparation system linked to a triple collector gas source isotope ratio mass spectrometer at Innsbruck University. Values are reported relative to VPDB standard. Long-term precision of the d13C and d18O values, estimated as the 1s-standard deviation, is 0.06 and 0.08‰, respectively (Spo¨tl & Vennemann 2003). Methods for water analyses are given in Spo¨tl et al. (2005). Elemental analysis by inductively coupled plasma mass spectrometry (ICP-MS) analysis was conducted at the NERC ICP-MS facility at Kingston University UK, on drilled sub-samples from stalagmites and on selected water samples. We needed to use water samples that had been collected within a 24 hour period because colloidally transported trace elements tend to bind to the surfaces of tubing and collection vessels over longer periods
297
(Fairchild & Treble 2009). We also required samples from different seasons. There are significant logistical difficulties in accessing the cave chambers regularly and this had only been possible (monthly) during a previous intense period of monitoring (Spo¨tl et al. 2005). We checked the collection of (unacidified) waters which had been collected up to three years previously and, where the samples were intact, acidified them to release all trace metals into solution. An Agilent single-collector instrument was used for dissolved powders and waters, supplemented by a high mass-resolution Axiom instrument for S and P analysis of dissolved speleothem powders, as reported in Frisia et al. (2005). Both solutions from speleothem powders (50 mg in 100 mls solution) and acidified waters were analyzed in a matrix of 2% Aristar-grade HNO3. Standardization procedures are as reported in Borsato et al. (2007) and all reported results are well above calculated detection limits with analytical precision of 5%. Ion microprobe analysis at the Edinburgh Ion Microprobe Facility (EIMF) was carried out under a variety of experimental conditions using polished gold-coated samples. One sub-set of analytical work was summarized by Smith et al. (2009), with more examples being given here. This employed similar methods to those of Fairchild et al. (2001) and Borsato et al. (2007), using a Cameca ims-4f instrument, with a primary 20 nA beam of diameter c. 30 mm. Each of the three stalagmite samples was bombarded with a primary O2 beam leading to sputtering of secondary ions of 1Hþ, 23Naþ, 26 Mgþ, 31Pþ, 44Caþ, 88Srþ and 138Baþ which were measured. The step scan mode was used to analyse the sample at approximate 5 mm intervals, with ions being drawn from an area with an approximate diameter of 8 mm. Measurements were taken along linear traverses at 5 or 10 mm intervals. Precision of results for most elements is typically 1 –5% (but worse for H and Na whose analyses are regarded as semi-quantitative because of variable surface effects); Mg, Sr and Ba were referred to a carbonate standard (Oka) and other elements to an apatite standard (Durango) with agreement within 10% of mean bulk analyses as determined by Borsato et al. (2007). Sample Obi84 was also analysed on the Cameca ims-1270 instrument at the EIMF under three different experimental conditions using a primary Csþ beam to generate negative secondary ions. Precise analysis of 13C/12C ratios was carried out in two sessions in February and May 2005. The primary beam current was 30 nA (February) or 40–45 nA (May), and a 50 –60 mm image field, to yield an elliptical analytical spot. The area of gold removed is typically 35 30 mm, but SEM analysis of February pits showed a pit 13–20 20 mm with a narrow
298
I. J. FAIRCHILD ET AL.
core about 6 6 mm. Secondary carbon ions of masses 13 and 12 were simultaneously collected using two Faraday cups. A combination of high transmission and a high beam current was necessary to obtain sufficient counts (1.5*106 cps on 13C). Charge neutralization was achieved using a normal incidence electron gun. Each analysis consisted of 20 blocks of 5 seconds measurement time following a period of pre-sputtering. Analysis of a group of 20– 50 sample points alternated with blocks of 10 analyses of standard UWC (University of Wisconsin calcite, which has a composition of 22.14‰ (relative to VSMOW) determined by multiple bulk analyses by J. W. Valley). Analytical points were spaced at 30 –60 mm to avoid surface charging effects and their X –Y coordinates were recorded automatically. The 1-sigma precision on 10 successive spots on the UWC standard was normally ,0.6‰, but accuracy is more limited by variable instrumental drift and is best judged by the reproducibility of parallel scans on the sample as presented later. For S and P analysis in May 2005, a primary beam current of 4.5 nA were used with Ko¨hler illumination to generate an analytical spot around 20 –25 mm in size. The instrument was used in line scan mode (10 or 15 mm step) with a pre-sputter reduced to 10 seconds. Ratios of the secondary ions of 31P and 32S (at a mass resolution of 2500 to distinguish from O2) to 13C were counted using an electron multiplier. Results have a precision of ,2% from counting statistics, but were not standardized. Analysis of 31P2 and halogen elements (19F2, 35Cl2, 37Cl2, 79Br2, 81Br2 and 127 2 I ) was undertaken in November 2008 with a 1 nA primary beam focused to a diameter of c. 10 mm. Analysis was in line scan mode, stepping at 5 mm between analyses, which also had the effect of cleaning the sample by pre-ablation. Analytical precision is limited by counting statistics and is ,1% for Cl, 2–3% for P and 3–4% for Br and I. Halogen concentrations were standardized by direct comparison with a lead –silicate glass K1053. The absolute ion yields for the halogens were confirmed as being very high, varying from 5.0 (for F2) to 0.86 (for I2) times that of O2. The effects of sample matrix on ion yields on the silicate glass compared with carbonate are currently unknown. Sample Obi84 was analyzed at the European Synchrotron Research Facility using soft X-rays at beamline ID21 in May 2005 and using hard X-rays at beamline ID22 in December 2006. For beamline ID21, experimental details are similar to those given in detail in Frisia et al. (2005). Samples were prepared as thin (,1 mm thick) wafers up to around 22 19 mm, with polished upper surfaces and were mounted with double-sided tape to a circular 3 cm-diameter holder. The samples
were excited with monochromatic synchrotron radiation of 2.9 keV in order to stimulate Ka radiation from light elements (up to Cl). Ca was not analysed, but the consistent beam conditions permit the assumption that excitation conditions were consistent; minor reductions in beam intensity of up to 10% were corrected for. Specific energy levels within the resultant X-ray fluorescence spectrum characteristic of particular elements were then selected for study and scans and maps were generated. The X-ray fluorescence was predominantly generated at depths of just a few microns in the sample and maximum penetration of the exciting radiation was 20–30 mm. The beam was focused to either 5 mm (scans and maps) or, for a high-resolution map using 1 mm pixels, to 0.6 0.3 mm, but resolution is limited by the slightly larger excitation volume. At beamline ID22, the conditions are mostly similar to those described in detail by Borsato et al. (2007), but some differences are noted here. The sample was prepared as a doubly-polished wafer of thickness c. 150 mm, mounted over a hole in the sample holder. Excitation energy used was 23 keV, enabling in principle the detection of the K-lines of all the elements up to atomic number 47 and of the L-lines of other elements. The spectrum was deconvolved by using PyMCA software (Sole´ et al. 2007) and detection limits were found to be sub-ppm level for Ca, Sr, Zn and Pb. Detection of other elements was limited by spectral interferences. Elemental mapping was carried out with a 2 mm resolution over a length of 1.1 mm and 5 mm resolution over a width of 0.1 mm. The fluorescence signal was integrated for 2 seconds in each pixel. The information depth is defined at the depth to which 63% of the characteristic X-ray emission line intensity of an element is being collected and depends on the X-ray energy. It varies from around 17 mm for Ca to 110 mm for Sr. Generation of X-ray fluorescence over a range of depths for heavier elements results in the collection of data from within the crystals as well as from the surface, and hence a more complex pattern is observed than in a strictly two-dimensional image. Unlike tomographic images, it is not possible to distinguish which are shallower and which are deeper sites of X-ray generation.
Study site and dripwater characteristics The local context of Obir is only briefly described here, being summarized from Spo¨tl et al. (2005). The cave is on the eastern flank of the Hochobir massif (46.518N, 14.548E), 22 km SE of Klagenfurt, Austria. One distinct group of cave passages within the cave system is accessible only through
269.4 29.91 1.1 1.15 27 29 267.6 28.43 1.6 2.45 27 29 268.8 210.27 0.8 0.92 22 26 269.1 210.26 0.8 0.75 13 17 210.25 0.15 45 210.11 0.18 32 210.23 0.10 26 210.25 0.10 16 23.05 0.15 44 23.11 0.15 31 23.07 0.14 25 23.15 0.09 7 10
2.70
2.20
0.03
December 2002 Obi84 SH4
February 2002
May 2001 Obi55 SH3
May 2001
Not collected SH2
November 2000
August 1998 Obi12 SH1
September 1998
Collection date of stalagmite
Start date of observations
mean stdev n mean stdev n mean stdev n mean stdev n
2.06 1.82 44 0.42 0.39 31 4.42 1.02 27 9.08 3.82 17
0.11
292 11 45 252 38 29 296 17 26 299 13 14
8.19 0.15 44 8.19 0.12 31 8.23 0.13 25 8.24 0.10 13
186 10 45 158 24 32 191 9 26 190 12 16
5.29 4.01 0.49 0.32 0.44 1.25 0.12 0.03 45 44 44 44 5.11 2.61 0.50 0.36 0.28 0.55 0.13 0.08 32 31 32 32 4.76 2.38 0.45 0.32 0.21 0.29 0.10 0.06 26 26 26 26 4.44 2.64 0.40 0.33 0.11 0.80 0.20 0.03 7 7 7 10
0.23 0.03 45 0.25 0.04 31 0.23 0.03 26 0.23 0.02 10
0.15 0.02 45 0.17 0.03 32 0.13 0.02 26 0.14 0.02 9
51.51 7.74 0.03 1.52 3.47 0.62 0.01 0.25 45 45 29 40 42.89 7.19 0.03 1.50 8.78 0.42 0.01 0.12 32 32 18 30 52.97 7.07 0.04 1.53 5.26 0.43 0.01 0.20 26 26 13 26 51.17 7.03 0.03 1.64 4.65 0.23 0.01 0.15 9 9 4 5
0.44 0.15 44 0.29 0.16 31 0.51 0.13 25 0.58 0.07 7
d18O Discharge ml/sec
Minimum driprate ml/sec
EC mS/ cm
pH
HCO3
SO4
NO3
Cl
Na
K
Mg/ Ca
Ca
Mg
Sr
SiO2
Calcite saturation index
PCO2
dD
299
Drip site
Table 1. Summary of drip hydrochemistry including samples collected to the end of 2003
an adit from an abandoned Pb –Zn mine at an altitude of 1090 m. The adit was opened in 1870, partly closed in the second half of the 20th century, and then re-opened before a door, with slits to allow movement of bats, was installed in 1999. In this area several chambers around 200 m into the hillside have been monitored, displaying a remarkably congruent pattern of variation in PCO2 through the year. Of these, Sa¨ulenhalle lies at a depth of c. 70 m from the surface and is accessible via a narrow horizontal passage opening to a chamber several metres in size in each dimension. The overlying vegetation is a spruce-dominated forest underlain by a brown forest soil. Annual precipitation lies within the range 1100–1600 mm, mostly in the second half of the year and the waterexcess peaks in October and November. The temperature is below freezing normally between November and March and the ground is snowcovered for most of this period. The air temperature in Sa¨ulenhalle, which should be close to the annual mean, is 5.8 + 0.1 8C throughout the year and dripwaters in the same chamber are 5.8 + 0.2 8C. Spo¨tl et al. (2005) presented several years of hydrochemical data from three representative drips in the cave system which varied from near-invariant in drip rate to displaying modest seasonal changes through to more extreme and inter-annual variability. Despite differences in hydrological behaviour they showed a consistent pattern of seasonal changes in the carbonate system, but no seasonal variations in other chemical parameters. In the winter: (1) pH rises, in equilibrium with changing PCO2 of cave air and calcite saturation rises similarly; (2) d13C of dissolved inorganic carbon rises from a base of 211.2 to 211.9‰ by 2 to 7‰ as isotopically light CO2 is degassed; and (3) in some drips, there is a modest decrease in Ca and alkalinity (most precisely monitored by changes in the electroconductivity of the solution) indicative of some CaCO3 precipitation; winter precipitation is also seen on the surface of a pool (Silbersee) in Sa¨ulenhalle. The three studied stalagmites from Sa¨ulenhalle (Obi12, Obi55 and Obi84) are all approximately cylindrical structures 3–4 cm in diameter fed by stalactites a few cm to dm above. The elemental composition of four drip water sites, including the three feeding the three stalagmites, is summarized in Table 1. The drips feeding the studied stalagmites are very similar in: (1) major and minor element compositions; (2) the consistent Ca concentrations of 51–53 + 5 mg l21; and (3) calcite saturation indices (around 0.44–0.58 + 0.1) controlled by seasonally changing ventilation. Generally, no relationship between instantaneous measurements of drip rate and EC (electrical conductivity, a proxy for total ion content) was
d13C
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
300
I. J. FAIRCHILD ET AL.
found. SH2, which feeds a stalagmite that has not been collected, was an exception in that at the slowest drip rates (to 0.36 ml hr21), EC was low and correlated with low Ca and hence high Mg/Ca; d13C was also high. This is indicative of prior calcite precipitation stimulated by degassing (Fairchild et al. 2000; Fairchild & McMillan 2007). Data on drip rates (from a tipping bucket drip logger), water and air temperature is illustrated in Figure 1. None display a simple seasonal pattern of variation; in particular there is no sign of an increased drip rate during peak infiltration during the autumn season. Drip SH4, feeding Obi84, was not continuously logged, but over a period of two years maintained a relatively high drip rate that would have ensured steady growth. Drips SH1 to SH3 all displayed inter-annual growth variations; the most pronounced relative change was SH1 (Obi12) which displayed a pronounced decline from 2000–2003 (confirming instantaneous drip rates presented in Spo¨tl et al. 2005) when monthly volume collections indicated drip rates as low as 0.36 ml hr21. Figure 1 illustrates a more regular pattern of variation of temperature of drip water (using SH3 – Obi55 as a typical example), displaying an upward deviation compared with summer values of around 0.1 8C in the first part of winter followed by a larger drop and gradual recovery later in the winter. Air temperature in Sa¨ulenhalle displays a similar pattern, although the upward displacement is more typically 0.2–0.3 8C. Since the air circulation is so extensive (Spo¨tl et al. 2005), we argue that the deviations in the water temperature are controlled by the seasonal air circulation. Some spot measurements of air velocity have been made at the narrow squeeze at the entrance to Sa¨ulenhalle as well at the next one which leads to subsequent chambers: maximum air velocities of c. 0.5 m sec21 are reached during winter.
Petrology Stalagmite fabrics The successive growth layers of the stalagmite are made visible by the presence of internal laminae which are parallel to the external surface of each stalagmite (Fig. 2a). A slight depression toward the centre of the stalagmites likely reflects the impact point of the feeding droplets. Figure 2b highlights the characteristics of the columnar fabrics which compose Obir stalagmites. These, as for other speleothems, are aggregates of crystals that precipitated synchronously from the same medium to form single growth layers. The ‘synchronous’ crystals precipitated in each growth layer (where the unit of ‘synchronous’ time could be a year or a season)
are commonly referred to as ‘crystallites’, following Kendall & Broughton (1978), to distinguish them from the composite columnar individuals which are typically recognized by their systematic extinction. The fabrics in the most intensively studied example (Obi84) are transitional from columnar to microcrystalline types as defined by Frisia et al. (2000). Columnar fabric is commonly characterized by the parallel arrangement of prismatic crystals (Onac 1997), where the equal orientation of the crystals is related to a common direction of most rapid growth. In microcrystalline fabric, some individuals within the aggregate are not parallel or sub-parallel to the adjacent individuals and their direction of most rapid growth is not the same as that of the majority of the individuals within the aggregate. In the centre of Obi84 (Fig. 2b), extinction sweeps regularly as a wave through the crystals (with a total variation of up to 208), similar to the variation in orientation of growth layers but crystal boundaries are typical of columnar calcite. However, the extinction sweeps counter to the crystal elongation (see crystal S in Fig. 2b) in the manner of radiaxial calcite, which has recently been recognized in a Mg-bearing calcitic speleothem sample by Neuser & Richter (2007). In speleothems, radiaxial fibrous calcite aggregates have been interpreted as being the product of crystal splitting, that is, where individual crystals split through various mechanisms, among which because ions with a ‘poisoning’ effect on growth sites are present in the parent solution or because ions with a larger ionic radius substitute for ions in the mineral structure. The most extreme form of crystal splitting found elsewhere is a spherulite (Onac 1997). Further from the growth axis, the composite crystal aggregates are closer to the microcrystalline fabrics described by Frisia et al. (2000), the latter representing a type of columnar fabric forming when growth inhibitors or drip rate variability characterize the system (Frisia et al. 2000). Highresolution analysis of this microcrystalline calcite has been achieved using electron backscatter diffraction in an area c. 8 mm below the top of sample Obi84 (Fig. 2b) and is illustrated in Figure 2c. Figure 2c shows a mosaic of small crystallites, with grain boundaries which range from well defined to more ‘diffuse’, typically around 20 10 mm. There is a clear preferred orientation of the individuals roughly from the left bottom corner to the right upper corner of Figure 2c. This direction coincides with the overall orientation of the aggregate crystal which has been analysed (see the area of the map in Fig. 2b). The relative crystallographic orientation of each small crystal (or crystallite) within the aggregate is indicated by the false colours of Figure 2c and correspond to the orientations shown in the stereographic projections of
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
301
Fig. 1. Characteristics of Saulenhalle drips, mostly using continuously logged data. Features of all four drips, including the three feeding the studied stalagmites, are a lack of clear seasonality and large inter-annual variations in mean drip rate. Water temperature is given for drip SH3 (Obi55) as an example for comparison with air temperature. The two follow a similar pattern of a rise followed by a fall during the winter season; this is attributed to the forcing effects of winter air ventilation.
302
I. J. FAIRCHILD ET AL.
Fig. 2. Petrology of stalagmite sample Obi84. (a) image of ion microprobe section with calendar ages of annual laminae superimposed. In the central part of the speleothem the laminae are approximately tabular and horizontal, but curve and merge on the lateral flanks. (b) Photomicrograph under crossed polars of thin section made from synchrotron sample (viewed to a similar sample depth as in photograph (a) illustrating quasi-columnar crystal morphology approximately normal to growth surface. ‘S’ indicates example of crystal with undulose (specifically, radiaxial) extinction (see text). Small red rectangle indicates area of electron backscatter diffraction (EBSD) map. (c) EBSD map of a grid of 931 by 701 points at 1 mm resolution. The map has been false-coloured to highlight the relative crystallographic orientations of the calcite across the map. The colour range blue to red indicates a 08– 158 deviation in orientation with respect to the crystallographic orientation of the calcite at the reference point indicated by the red cross near the centre. The large crystals to the right of the map are grey and have no colour as the orientation of these crystals differ by more than 158 from that of the reference point (see also Fig. 1d). Scattered grey pixels in the upper and left areas of the map were the results of systematic mis-indexing of the calcite diffraction patterns while others (mostly in the central areas of the map) represent locations where the quality of the diffraction pattern was too poor to give a solution for the crystallographic orientation of the calcite and were assigned a zero solution. The map shows that the large crystals are divided into crystallites around 10 mm wide. (d) Upper-hemisphere stereographic projections of the poles to potential indicated crystal forms for the crystals illustrated in the EBSD map in Figure 1c. The XY plane of the projections is that of the surface of the sample section. The colours shown correspond to those in Figure 1c. The reference point (red cross) shown in Figure 1c was chosen as having a crystallographic orientation close to the centre of the brightly coloured clusters. The grey clusters indicate the orientation of the large (grey) crystals to the right of Figure 1c. The f0001g diagram also corresponds to the zone axis and orientation of the z crystallographic axis which can be seen to be sub-vertical in the diagram, corresponding to sub-perpendicular to the elongation and growth directions of crystals at the top of the stalagmite, that is, the crystals are optically length-slow. The f10–11g diagram displays a pole to one face approximately in the equatorial plane: this is thought to be equivalent to the flat face seen on crystallite tops in Figure 3b. Colour references in the caption correspond to the online colour version.
Figure 2d. The crystallites show a range in orientation by up to around 158 within each crystal, but adjacent crystallites are often mismatched by only 1–28. The pole to f0001g in Figure 1d (left), which
also corresponds to the ,0001. zone axis cluster near the centre of the projection, indicating that the z crystallographic axes are near-vertical with respect to the surface of the sample section shown
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
in Figure 1c and sub-parallel to the growth surface at the top of the speleothem. SEM analysis at the sample top (Fig. 3b) reveals that the morphology of the calcite crystals is characterized by sub-micrometre-scale macrosteps. Some flat faces also developed which are apparently all parallel to each other. This contrasts with the more usual situation in stalagmites (Frisia et al. 2000), where SEM observations reveal that speleothem surfaces are characterized by a variety of faces in different orientations. The scaling indicates that each flat face seen by SEM may correspond to a single crystallite on the EBSD image. If the orientation analysed by EBSD on a portion of the stalagmite holds true for the rest of the specimen, as is supported by the extinction patterns as seen by optical microscopy, possible candidates for the flat faces which developed at the top of the Obir stalagmites are those that lie in the equatorial plane of the stereographic projection and pointing towards the top of the sample as in Figure 2d. Plots for various prismatic, rhombohedral and scalenohedral forms are shown and the most likely example is a face in the f10 –11g
303
rhombohedral form (Fig. 2d; which differs from the cleavage-parallel rhombohedron f10–14g). Later in the paper, we interpret the presence of macrosteps and the overall morphology of the calcite crystals of Obir stalagmites as being controlled by the presence of impurities in the parent solution.
Visible annual lamination The top few centimetres of stalagmite Obi84 display a consistent structure in which narrow visible laminae, shown to be annual by Smith et al. (2009), are spaced at c. 100–200 mm intervals. Similar laminae are also developed in the other two samples, but their development is less consistent. The stalactite feeding Obi55 was collected and also displays such laminae, but very closespaced. Where most distinctly developed in stalagmites, the laminae are a few microns wide and may be nearly flat in geometry, but more commonly display a zig-zag shape. Usually the zig-zags have a relief of a few micrometres, but this can exceptionally be up to 200 mm where it is followed by the
Fig. 3. Petrology of sample Obi84. (a) ion probe thin section, transmitted light, illustrating a series of annual event laminae with the last one being close to the top of the sample prior to collection in December 2002. Zig-zag crystallite shapes are visible, and are of much higher relief in the 1998 layer, apparently corresponding to a pit on the crystal surface which subsequently evolves into a (black) air inclusion. A within-year hiatus prior to the 2000 infiltration lamina is marked by a re-nucleation horizon (hi). (b) SEM of top surface of sample illustrating crystallites which display a mixture of flat smooth surfaces and rough surfaces representing stacked edges.
304
I. J. FAIRCHILD ET AL.
development of a large air inclusion (Fig. 3a). Some years show a more complex structure in which two or three such laminae are found close together, although the optical imaging of these is complicated by internal reflections (Fig. 4a). The zig-zag shapes have dimensions consistent with those of the crystallites in Figures 2c and 3b. Laterally, some laminae may become diffuse and apparently are composed of a zone of fluid inclusions, locally inclusions forming vertical trains at crystallite margins. These areas are interpreted as representing places where crystallites coalesced just below the growth surface (as in Kendall & Broughton 1978). In Obi84, which was collected in December 2002, the latest visible lamina lies very close to the top surface (Fig. 3a), implying that it formed in the autumn. In the review of Fairchild et al. (2007), we presented a similar figure to Figure 3a and identified these features as infiltration laminae, implying a formation by infiltration of water excess (Fairchild et al. 2006a), but we now regard this interpretation as not proven, and here
use the term ‘event laminae’ instead. By event lamina we mean a layer deposited in relation to a specific time-limited set of processes. Re-nucleation horizons are present in each of the three samples and are sufficiently prominent as to suggest they each represent a hiatus in growth. However, it is notable that the hiatuses do not contain enrichments in insoluble residues or signs of corrosion, and so are distinct from those that represent significant time gaps in speleothems (Tan et al. 2006). Indeed, close study of the prominent event shown in Figure 4c reveals that, locally, growth continues through the event, and that it is no more than two or three years in duration (Fig. 4d). The prominent hiatus in Obi84 occurs at the level of the 1834–1835 A.D. laminae around 23 mm below the top, whereas in Obi55 there is one just 9 mm from the top and in Obi12 at 12 mm from the top. Since the statistical analyses of Smith et al. (2009) show that the growth rate of each stalagmite is similar, the hiatuses are clearly of different ages in each sample.
Fig. 4. Stalagmite petrology. (a) Sample Obi 84, transmitted light, illustrating two annual infiltration laminae. The lower one illustrates three (arrowed) sub-laminae plus internal reflections, whereas the upper one is a single lamina. The fibrous sub-vertical structure of crystallites is seen to correspond to the zig-zag lamina structure. (b) Sample Obi84. Arrow illustrates hiatus between infiltration laminae for the years 1889 and 1890. To the left of the arrow the hiatus disappears. To the right (the stalagmite flank) various laminae converge to give more prominent hiatus (i.e. renucleation) surface. (c, d) Sample Obi12. (c) Prominent hiatus surface with box (0.9 mm long) showing location of photo d. (d) enlargement of hiatus, illustrating local area where growth continued throughout (arrows indicate individual annual infiltration laminae). The hiatus itself, although very prominent in photo c, is only of three or four years duration.
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
An example of an incipient hiatus, representing probably of the order of a month’s gap in growth, is shown between the 1999 and 2000 annual layers in Figure 3a. Essentially it represents a re-nucleation horizon where some new crystallites of diverse optical orientation are found and which can persist upwards (e.g. the upwards-expanding crystallite under the word ‘top’ in Fig. 3a). Such an intra-year hiatus can be seen to pass laterally into a more profound pause in growth in Figure 4b and this is typically of the lamina geometry as they are traced into the flanks of the stalagmites. In summary, it is notable that hiatuses are developed at different times in the three specimens and that they do not represent significant time gaps. The presence of inter-annual slowdowns in drip rate mentioned earlier in the paper (and found elsewhere, e.g. by Baldini et al. 2006) provides a possible mechanism for hiatus development, growth being limited by lack of ion supply through dripwater. Where the monitored drip rates (Fig. 1) go well below 0.36 ml hr21 (1 ml sec21), significantly slower growth would be expected from the quantitative model of Dreybrodt (1988) and Baker et al. (1998). However, since the different drips in Sa¨ulenhalle do not vary in discharge synchronously, this phenomenon is not of climatic significance. At least several hundred laminae of similar type and spacing are present at the top of the sample. Specifically a lamina chronology for Obi84 for the past 140 years is shown in Figure 5b. The mean annual growth thickness is 141 + 25 mm and the robustness of the series is shown by the low mean relative standard deviation of measurement of any individual year, which is 19%. Note that there is an overall reduction in annual growth thickness with time. This contrasts with the record in Grotta di Ernesto stalagmites where there is a strong increase, correlated with Northern Hemisphere temperature trends (Frisia et al. 2003; Smith et al. 2006). An increase would also have been expected had the opening of the mine adit at Obir had any appreciable influence on studied cave chamber. Hence we regard the circulation as essentially a natural feature of the cave system, mainly utilizing small openings in the rock.
Trends in chemical composition and comparison with dripwaters Trace elements Figure 5 summarizes the variations in chemistry found in stalagmites Obi84 (Fig. 5a, c) and Obi12 (Fig. 5d) over the past 200 years. Trace elements, including some additional elements not showing temporal variation, are also summarized in Table 2. The most unusual feature about both
305
samples is their extremely high content of Zn (typically c. 5000 ppm) and Pb (400–1400 ppm). These contents tend to reduce over time, albeit somewhat irregularly from the early 19th century to the late 20th century. Cu shows a similar trend in Obi84, but concentrations are low. Other trace elements (Mg, P, Mn, Y, Sr and Ba) show steady compositions, but S (present in the calcite in the form of sulphate) increases strongly with time. This parallels the impact of late 20th century pollution recorded in speleothem ER78 from Ernesto cave (Frisia et al. 2005). In work presented elsewhere (Wynn et al. 2010), we confirm this using the first in-situ micro-measurements of d34S in carbonateassociated sulphate. The trace element content explains X-ray diffraction data which demonstrate that there is a small ˚, reduction in (hexagonal) unit cell size (a ¼ 4.985 A ˚ ) compared with pure calcite (a ¼ c ¼ 17.042 A ˚ , c ¼ 17.06 A ˚ ; Reeder 1983). The mean 4.9896 A composition of the top cm of the sample in molar terms is Ca0.989 Zn0.0070Mg0.0042Pb0.0003CO3. For ions that form isomorphous carbonates with calcite, there is a linear change in unit cell size with substitution of trace species and both Mg and Zn have the effect of decreasing cell size. From data in Reeder (1983) and Mackenzie (1983), and assuming an additive effect of the two ions, the observed Mg and Zn concentrations predict parameters of ˚ and 17.038 A ˚ , very close to the observed a ¼ 4.986 A values. Therefore, it is interpreted that these ions are dominantly substituted for Ca. For Mg and Sr, it is appropriate to calculate a distribution coefficient to express the fractionation in the ratio of the ion to calcium where: (Tr=Ca)CaCO3 ¼ KTr (Tr=Ca)solution
(1)
where Tr is the trace ion and KTr is the distribution coefficient, which may vary to a greater or lesser extent with temperature, precipitation rate, crystal morphology, or other aspects of solution chemistry (Fairchild & Treble 2009). The resulting values in Table 3 are reasonably consistent with those observed by slow-growth, low ionic-strength experiments and in Ernesto cave speleothems (Huang & Fairchild 2001), but kinetic and competition factors are known to be important for Sr (Borsato et al. 2007). The behaviour of sulphate is being studied in much more detail and will be presented elsewhere, but it should be noted that the concentrations in drip water and in the speleothems are similar to the Ernesto stalagmite ER78 (Frisia et al. 2005). In both cases sulphate partitions into calcite to a lesser extent than would be expected from the experiments of Busenberg & Plummer (1985), which
306
I. J. FAIRCHILD ET AL.
Fig. 5. Depth trends in stalagmite Obi84 (a–c) and Obi12 (d). (a) stable isotope data derived by micromilling at 0.1 mm intervals; thick lines are 5-point moving averages. (b) Variations in lamina thickness; calendar years are shown at 20-year intervals. (c, d) Trends in composition with depth from ICP analyses of samples drilled at approximately 1.5 mm intervals. Data for 10 elements is summarized in Table 1. S increases during the 20th century whereas the base metals tend to decrease, especially in Obi84.
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
307
Table 2. Bulk compositions of stalagmite samples (ppm in calcite) mm below top Obi84 0 –10 11–20 .20 Obi12 0 –10 11–20 .20
Mg
P
S
Mn
Cu
Zn
Sr
Y
Ba
Pb
Number of analyses
903 976 1040
66 52 67
42.4 6.7 10.5
0.99 1.01 0.87
2.2 2.7 3.4
3410 4590 6420
32.4 35.3 30.9
0.020 0.040 0.039
174 196 191
500 770 1360
8 6 7
1160 1280 1250
56 67 93
30.6 12.8 17.3
1.02 0.67 1.3
4.5 3.8 5.4
4010 5970 5690
32.6 29.2 30.9
0.015 0.014 0.020
244 234 237
370 630 510
8 6 5
were conducted at relatively fast growth rates using high ionic strength solutions. For other solutes, the partition coefficient concept is not so applicable and it is also more difficult to obtain water analyses for many species. Borsato et al. (2007) found that at Ernesto metals such as Pb and Zn, as well as P, strongly adsorbed to collection vessels because of being transported in the form of unstable colloids. Obir water samples are normally collected monthly, or less frequently from cumulative collection vessels, although some direct samples were taken from drip SH4 from 2002 to early 2004. On realizing the importance of colloidally transported elements, in November 2005, we re-analyzed remaining previously collected samples from drips SH3 and SH4, acidifying samples to release adsorbed elements, and carried out ICP-MS analysis. Data from SH3 (feeding Obi55) and SH4 (feeding Obi84) are presented in Table 4 and results are generally consistent between the two. The Fe, Al and Si probably represent colloidal particles, but organic colloids could be even more important (Fairchild & Treble 2009). Pb, Zn and Y (all elements inferred to be colloidally transported at Ernesto Cave) have, as expected, high variations in abundance in contrast with Sr and Ba which are expected to be very largely present as free ions. There were no obvious seasonal variations in abundance. Empirical distribution coefficients, as in equation (1), between the dripwater data of Tables 1 and 4 and the composition of the top 10 mm of the stalagmite are shown in Table 5. The calculated values for Zn and Pb (1) are consistent with experimental and theoretical data summarized by Rimstidt et al. (1998) and Curti (1999) and with the evidence of strong sorption of these elements to calcite surfaces (Zachara et al. 1991; Godelitsas et al. 2003; Chada et al. 2005). The values of 1 for Cu and Y are quite inconsistent with data in Rimstidt et al. (1998) and Curti (1999) where values 1 can be predicted, using rare earth elements as a model for Y. This discrepancy can be accounted for if: (1) the partition coefficients
for organic colloidal substances are ,1; and (2) Cu and Y are more tightly bound to colloids than are Zn and Pb. The net effect would be that Y and Cu are not so readily available for incorporation into calcite.
Stable isotopes The mean stable isotope composition of the top 2 mm (c. 20 years) of growth of sample Obi84 is 26.41 + 0.36 and 27.83 + 0.23‰ for d13C and d18O, respectively (n ¼ 200). In the case of d13C there is a direct comparison with the aqueous d13C composition since the fractionation between CaCO3 and dissolved inorganic carbon at the pH values of interest is very small (Mu¨hlinghaus et al. 2007). The mean d13C value for drip SH4 (which is very close to those from SH3) is 210.25 + 0.74‰ (n ¼ 19) Table 1. However, the pattern of seasonal variation of aqueous d13C (with a maximum of 28.36‰ representing as one of only two points higher than 29‰) is less strongly developed than for the more slowly dripping points SH1 and SH2, which display increases of around 3 and 8‰ from the low summer levels of 211‰ (Spo¨tl et al. 2005). Hence the mean difference of 4‰ between SH4 drip water and the stalagmite is likely to reflect continued degassing perhaps with some kinetic effects during precipitation. The d18O composition of dripwater feeding SH4 is 210.23 + 0.10‰ (n ¼ 18). Using the experimentally determined fractionation factors of Kim & O’Neil (1997), such water should precipitate calcite at equilibrium at a temperature of 5.8 8C with a composition of 28.5‰. This is 0.7‰ lower than the observed composition and might indicate a small kinetic effect, but such a discrepancy is very commonly observed in natural speleothems (McDermott et al. 2005) and is not apparent using the older experimental values of Friedman & O’Neil (1977). No detailed treatment of the variation in stable isotopes with depth is attempted here. However, we can note that over time the isotopes show rapid
9.2*10 (8.329.8*10 ) 8.6*1025 (8.029.4*1024) 5.5*1025 (3.927.9*1024)
24 25
Stalagmite Sr/Ca (top mm)
shifts from covariations to antipathetic variation and there are no strong correlations with mean temperature, winter temperature or rainfall in the instrumental period. Longer-term d18O trends in Obi84 and other stalagmites from this cave show amplitudes of about 1‰ over the Holocene which correlate with isotopic changes identified in speleothems from Spannagel Cave in the Central Alps of Austria (Vollweiler et al. 2006; Vollweiler, Mangini & Spo¨tl, unpublished data).
Seasonal variations in chemical composition
0.15 (0.1320.21) 0.13 (0.1220.20) 0.14 (0.1320.17)
0.003 (0.002520.004) 0.0024 (0.001120.0031) 0.0021 (0.00120.0032)
0.020 0.018 0.015
24
Obi12 Obi55 Obi84
Stalagmite Mg/Ca (top 0.5 mm)
KMg
Drip water Sr/Ca
24
6.9*10 (5.129.0*10 ) 7.4*1024 (4.728.8*1024) 7.0*1024 (5.329.0*1024)
Carbon isotopes
Dripwater Mg/Ca
Table 3. Mean values (and range in parentheses) of dripwater and stalagmite Mg and Sr compositions with the calculated value of the distribution coefficient K
0.13 0.12 0.079
I. J. FAIRCHILD ET AL.
KSr
308
The monitoring data show that enhanced ventilation during the winter season is a key process at Obir. The winter season is associated with low PCO2 which is found to control the pH of dripwater by enhancing the degassing towards equilibrium (Spo¨tl et al. 2005). The high d13C values found in dripwater should be expected to be seen in winter speleothem growth. In faster-growing stalagmites, it has recently proved possible to micromill subsamples and demonstrate seasonal variations in d13C (Frappier et al. 2002; Johnson et al. 2006; Mattey et al. 2008), but the Obir laminae are too thin for this to be done. Hence, we made the first attempt to determine such variations in terrestrial samples by ion microprobe. The work was focused primarily around a group of relatively thick annual laminae which formed between 1958–1966. Interlaced forward-and-back trains of ablation pits were created during the analysis which was repeated in some parallel traverses (Fig. 6b). The work was technically extremely demanding because of the high beam current that was required leading to frequent electrical discharges that terminated the analysis. Enough data was collected however to demonstrate that considerable intra-annual variability is present with an amplitude of up to 6‰ and always at least 2‰, although the magnitude could not be reproduced between analytical tracks. The expectation from cave monitoring would be that the winter growth of calcite should be isotopically heavier and would occur earlier in the ‘hydrological’ year that starts with the visible lamina. Heavier mean values in the first half of the growth in a particular year was found (Fig. 6b) in 1959–1960 (4 replicates), 1960–1961 (3 of 4 replicates), 1961 –1962 (2 clear, 2 unclear replicates) and 1962–1963 (both replicates). The results can be regarded as providing indicative support for the conceptual understanding. At other cave sites, where faster growth permits conventional isotopic analysis, clear annual d13C
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
309
Table 4. Indicative ICP-MS analyses (in ppb) of drip water trace element compositions for SH3 (Obi55, month-long collection) and SH4 (Obi84, instantaneous collection)
SH3 (n ¼ 12) SH4 (n ¼ 6)
mean stdev mean stdev
Al
Si
P
Fe
Cu
Zn
Sr
Y
Ba
Pb
31 31 33 29
781 295 707 240
48 76 38 14
287 36 270 38
6.7 5.5 4.8 0.8
48 24 48 25
38 2 34 2
0.020 0.023 0.021 0.021
128 4 115 8
3.0 2.3 2.9 2.1
Table 5. Summary of water-calcite partitioning calculations for the top 10 mm of Obi84 (SH4) from data in Tables 1 and 4
Distribution coefficient, calculated % efficiency of removal
Ca
Mg
Cu
Zn
Y
Pb
– 0.63
0.017 0.010
0.059 0.04
9.1 5.7
0.122 0.076
22 14
variability has been demonstrated (e.g. Johnson et al. 2006; Mattey et al. 2008).
Sulphate Sulphate can be expected to vary in abundance with seasonal ventilation because of the influence of pH on its incorporation into calcite. Frisia et al. (2005) noted a significant annual variation in sulphate in stalagmite ER78 from Ernesto Cave which was attributed to pH changes, based on the model of Busenberg & Plummer (1985). These authors proposed that sulphate incorporation could be modelled by a distribution coefficient approach where the ratio of interest [cf. equation (1)] was 22 SO22 4 /CO3 . Since aqueous sulphate does not vary through the year, and the abundance of CO22 3 is primarily controlled by pH in terms of its overriding control on the ratio of HCO2 3 (the more abundant species) to CO22 3 , a low value for sulphate in the speleothem each winter would be expected. As was done by Frisia et al. (2005), synchrotron radiation was used to confirm that S in Obi84 was present as sulphate using the position of the dominant X-ray absorption peak. However, quantitative analysis of S content was impossible by this method because the Ka emission line was swamped by Pb. Instead, the ion microprobe was used to analyze S, utilizing negative secondary ions, which permitted the simultaneous determination of P. Results for the 1958–1966 laminae illustrate that P peaks lie close to or precisely on the visible event lamina. In comparison with this, S also shows annual-scale variability, but it is offset from P. Figure 6a illustrates that in most cases the low values also occur predominantly from around the time of deposition of the visible lamina through to the first half of the
next year’s growth. A similar approach was taken to analyzing the last few years of growth at the top of the sample (Fig. 7). Here laminae were not visible in the precise area analyzed, but their thicknesses on a lateral part of the section are in agreement with the lamina positions that can be deduced from the P peaks (although it should be noted that the 1997 P peak is very small and the 2000 P peak is double). Again the low-S analyses are found to be overlapping with and following the P peaks, consistent with their formation during the winter period. An additional feature of Figure 7 is that an overall decline in mean S/C is found which is consistent with the fall in aqueous sulphate observed from dripwater monitoring data from 2002– 2004 (Spo¨tl et al. 2005) as the catchment recovers from the effects of acid deposition in the late 20th century.
Event laminae Smith et al. (2009) reported ion probe analyses of positive secondary ions from the Obir stalagmites which demonstrated a strong annual covariation of H, P, Na and Mg with enrichments centred around the event laminae, where Sr decreases. The Sr pattern was found to be most reliable for using trace element layers to determine the rate of stalagmite growth. For Obi84, a principal component analysis found that this explained 55% of the variation in elemental chemistry, with a further 23% being explained by an independent mode of Sr and Ba variability (these results were obtained on data from which long-term trends had been removed). This pattern of variability is illustrated for the years 1958–1966 in Figure 8a. A discrepancy with Grotta di Ernesto stalagmites is that Mg increases
310
I. J. FAIRCHILD ET AL.
Fig. 6. Chemical trends in sample Obi84 across an interval of wide laminae corresponding to the years 1958–1966 (year marker indicates position of visible infiltration lamina for that year). Compare with Figure 8. A relatively large spot size of 20– 25 mm was used, so laminae were crossed obliquely to compensate for this. No S standard was available at the time the analyses were done, so results are presented as ratios to 13C. (a) ion microprobe traverse of negative secondary ions (15 mm step interval) illustrating the offset of S variability from P. Horizontal bars are low S intervals which tend to overlap with or follow P peaks. (b) Ion microprobe traverse of d13C; analytical spots are around 25 mm in diameter and spaced at 50 mm intervals. Inset photograph illustrates that the points are arranged in interwoven forward-and-back arrays which can be identified on dataplot by numbered points. Significant within-year variation in d13C is demonstrated.
in event laminae, whereas it does not usually do so in Ernesto stalagmites. It also varies annually in the Obir stalagmites typically by a factor of 2, much more than observed from the equivalent dripwater chemistry. These observations either imply that there are growth factor influences on Mg incorporation (Fairchild & Treble 2009), or that there was an independent mode of enrichment of Mg. Further ion probe analyses of negative secondary ions were later undertaken in order to study the pattern of halogen elements. The iodine analyses are the first such data from a speleothem. Figure 8b shows a strong pattern of enrichment in F, Br and
I coinciding with P. Cl, some of which may be a surface contamination, displays little variability. These elements, like the positive ions, show more than one peak per year in the period 1962–1965; such noise in some years was noted in Smith et al. (2009), but the annual pattern dominates overall. A more specific imaging of elemental variations in event laminae proved to be possible using synchrotron radiation. The soft X-ray studies showed a particularly strong response from the M-shell of Pb. Lead was seen to be periodically strongly enriched with a spatial pattern matching the distribution of event laminae (Fig. 9).
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
311
Fig. 7. Trends in S and P in the last few years of growth of sample Obi84. The horizontal scale is distance rather than time since infiltration laminae were not clearly visible in this section, but the P peaks match the lamina widths in a laterally adjacent profile (but note poor P peak in 1996 and double peak in 2000). S displays low intervals partly overlapping and partly following P peaks, and also displays a distinct downward inter-annual trend; likewise dripwater sulphate declined by 25% between 1999– 2003 in Sa¨ulenhalle (Spo¨tl et al. 2005).
Mapping of elemental distribution across event laminae was carried out within a zone in the sample that was also subsequently used for the EBSD analysis of Figure 2. Figure 10a illustrates a map of several annual layers with Pb enrichments, and the boxed area was analyzed at ultra-high resolution (Fig. 10b). The results obtained require some knowledge of the physics of X-ray fluorescence generation for their interpretation. The spectrum (Fig. 10c) reflects the excitation of various elements by radiation with energy of 2.9 keV. The sensitivity of excitation decreases rapidly at lower energies and is not high enough to excite Ca, so the consistent total ion flux is used to justify an assumption that the element records are concentrations. Normally, only K-shell excitation is expected, but if the concentration of an element is high, L- and even M-shell excitation can occur. Since the Pb is abundant in the sample and the M-shell signal lies fairly close to the excitation energy, this technique proved especially sensitive for Pb. Zn L-shell excitation also proved much stronger than Na K-shell excitation (Fig. 10c), but was much less responsive than the Pb peak. P and Mg Ka radiation could also be detected. A key point is that the use of soft X-rays causes excitation very close to the sample surface and so generates a sharply defined truly 2-dimensional map, so the sample offers exceptional clarity of the distribution of Pb.
Figure 10b illustrates two features. Running horizontally across the maps is the position of the infiltration lamina. Within a broad zone in the centre of the map enrichment in Pb, Zn and P can be seen. It is hard to discern a similar enrichment in Mg, but this peak is weaker and Mg in any case does not normally show the annual pattern quite as distinctly as P (Fig. 8a). This broad zone is 15 – 20 mm high compared with the 80–150 mm thickness of the entire annual layers in Figure 10a and therefore only forms during a subordinate part of the year (perhaps 1–3 months). Within the enriched zone, two continuous peaks in Pb (and less distinctly in Zn and P) are visible 1–2 pixels (i.e. 1–2 mm) across. Based on the annual growth rate, the narrow metal peaks must represent individual events of at most a few days duration when Pb and other elements were flushed onto the stalagmite. Also visible in Figure 10b are structural features running NE– SW across the maps. These have the same dimensions as the crystallites of Figure 2c and Figure 3b and so demonstrate a different chemical composition associated with the flat crystal faces of Figure 3b compared with the rough surfaces that surround them (corresponding also to the zigs and the zags of laminae in Figs 3a & 4a). Obi84 was also studied using hard X-ray synchrotron excitation, with energy of 23 keV, on a portion of the sample just below the very top
312
I. J. FAIRCHILD ET AL.
Fig. 8. Chemical trends in sample Obi84 for the years 1958 to 1966 (cf. with Fig. 6). (a) ion microprobe analyses of positive secondary ions (5-point moving average of 5 mm steps); note Sr is on right-hand axis and its scale is linear and reversed. H, Mg, Na and P tend to peak at visible laminae, inverse to Sr. (b) ion microprobe analysis of negative secondary ions (5 mm steps). P, F, Br and I enrichments coincide with visible laminae, but Cl does not display distinctive variation.
surface and representing up to eight years of growth (Fig. 11). This yields a spectrum with a much higher continuum background (Fig. 11a) and in which there are many peak overlaps. The strongest peaks are for Zn (L), Ca (K), Pb (L) and Sr (L). Y may be just quantifiable from its Kb peak, but its Ka peak overlaps with a Pb peak. The broad annual patterns of enrichment in Zn and Pb are matched by small reductions in Sr, whilst Ca variation is limited which confirms that the other elements can be interpreted as concentrations. Under these high-energy conditions, excitation takes place at much greater depths and over a
larger range of depths than under the conditions of imaging of Figure 10, hence providing images integrated over a larger thickness. Thus it would normally be expected that a less sharp image would be obtained unless chemical zones were precisely perpendicular to the excitation path which is 458 from the surface. This explains the details in the Pb and Zn maps which reveal fine bands of enrichment within individual crystallites in slightly different positions. Hence one time-band forms a series of sweeping curves in the image and it is difficult to reconstruct just how many are present in one annual layer.
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
313
Fig. 9. Line scan of Pb (three-point moving average of analyses at 5 mm intervals in sample Obi84) using X-ray fluorescence induced by relatively low-energy synchrotron radiation. The concentrations have been normalized to ppm using the mean composition from ICP-MS analysis in the same interval. The depth beneath the top surface of visible annual laminae in a similar, but different slice of the sample is shown as is the magnitude of autumn (September to December) rainfall since the year 1922. The highest Pb peak is closest to 1946 in the chronology of the other slice, but there is a registration uncertainty of +3 years because of the curvature of the laminae.
Fig. 10. Obir84, level within the area of the EBSD map of Figure 2. X-ray elemental mapping at ultra-high resolution, sample Obi 84, induced by relatively low-energy synchrotron radiation. (a) a low-resolution map (400 100 mm), showing the development of four annual laminae rich in Pb. (b) high-resolution maps, 30 30 mm with 1 mm pixels of Pb, Zn, P and Mg. The Pb map displays a 15 mm-wide zone of enrichment with two 1– 2 mm high-Pb zones within it. The enrichment zone is also displayed, but less prominently in the Zn and P maps. All maps show an oblique structure running from upper right to lower left, seen to correspond to crystallite development by the zig-zag pattern of the Pb zones. The grey shade linear scale-bar below the image has the following upper and lower limits in counts: Pb (0– 154), Zn (0–12), P (0– 31) and Mg (0– 22). (c) X-ray emission energy spectrum showing the location of peaks used to produce the maps. Elements in small type are too insignificant to quantify. The excitation energy was 2.9 keV and the peak at around this energy is due to scattering.
314
I. J. FAIRCHILD ET AL.
Fig. 11. X-ray fluorescence mapping using high-energy synchrotron radiation very close to the top of the Obi 84 sample. (a) Energy spectrum resulting from excitation with 23 keV radiation. Only Zn, Pb, Ca and Sr can be reliably quantified using a deconvolution procedure. (b) scale bars in maps in counts (logarithmic for Zn and Pb and linear for Ca and Sr). (c) Elemental maps of an area 1.1 mm by 0.04 mm at 2 mm resolution (top of section at left side). Probably eight annual layers are shown by Zn and Pb with a complex pattern of enriched micro-laminae because of excitation of crystallites at different depths below the surface: hence this is a pseudo-3D image. Levels of Zn and Pb are more uniformly high in two annual cycles on the far right.
Discussion Chemical control of crystal fabrics The Obir stalagmites, and specifically Obi84, which has been studied in most detail, display a series of striking fabric characteristics, some of which have not been previously recorded. However, their preferred optical orientation is a common characteristic of speleothems. In freely growing columnar crystals this preferred orientation arises through geometrical selection, whereby crystals orientated perpendicular to the growth surface grow fastest and become predominant, occluding the individuals that are not favourably orientated (Dickson 1993; Onac 1997). The resulting aggregates can be optically length-fast or length-slow dependent on crystal morphology (Dickson 1993). However speleothem growth is confined within a mm-scale solution film (Dreybrodt 1988) that limits competition either to periods whilst crystals are still small (at
the beginning of growth), or following a hiatus (when heterogeneous re-nucleation is often observed). Hence, the similar optical orientation of adjacent composite crystals at the top of Figure 2b implies that competition has occurred. Speleothems often are formed of composite (rarely of single) crystals with uniform extinction (Frisia et al. 2000). Commonly, the columnar crystals will be characterized by an overall orientation of the z axis (the ,0001. direction) roughly perpendicular to the speleothem surface. If the system is not disturbed by the presence of growth inhibitors, such as organic compounds or trace metals, or by changes in flow at the top of the speleothem, the crystallites covering the top of the stalagmites will be characterized by flat rhombohedral faces. In the case of columnar crystals (sensu Frisia et al. 2000), the crystallites would be expected to be strictly parallel in lattice orientation, whereas in microcrystalline calcite (sensu Frisia et al. 2000), some mismatches in orientation of crystallites are
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
found. Microcrystalline calcite is typically associated with perturbations of the system with selective adsorption of species from solution on particular surfaces. The result is a change in the morphology of calcite crystals, with the formation of macrosteps which indicate the presence of crystal structural defects (Frisia et al. 2000; Jimenez-Lopez et al. 2003; Fernandez-Diaz et al. 2006). In Obi84, XRD data demonstrates that there is a distortion of the lattice induced by substitution for Ca of the smaller ions Mg and Zn and the presence of nondivalent ions such as phosphate and halides implies significant substitution in defect sites. These impurities may cause the development of sweeping extinction patterns, even in the quasicolumnar composite crystals near the growth axis of the stalagmite. In the off-axis area studied by EBSD, adjacent crystallites are seen to differ in lattice orientation. The SEM image of the top of Obi84 shows the predominance of unstable surfaces, that is, those characterized by macrosteps, with only subordinate flat, defect-free faces. Also, we did not observe calcite rhombohedra which, so far, have been typically illustrated as characteristic at the surface of speleothems (Frisia et al. 2000). Cave precipitation experiments carried out in other caves, and in particular at Grotta di Ernesto, highlighted the development of stepped crystal faces with forms deviating from the equilibrium form (i.e. f10 –14g rhombohedron) when supersaturation or presence of impurities perturbed the system (Frisia et al. 2000). In Obi84 each crystallite displays a single, poorly developed flat face possibly pertaining to f10–11g. The appearance of macrosteps and the deviation from the equilibrium rhombohedra morphology have been documented by Braybrook et al. (2002) who grew calcite crystals doped with cobalt. By increasing the concentration of this trace metal in the growth medium, the morphology of calcite crystals became progressively anisotropic, probably because of the preferential localization of the dopant at the crystal surface (Reeder 1996) and the preferential development of a face in the f10 – 11g form was found. Zn has a very similar ion size to Co, and similarly adsorbs strongly to calcite surfaces, forming distinct phases at higher concentrations (Zachara et al. 1991). Zn prefers the opposite surface site to Co (Reeder 1996) and yet despite this, Reeder et al. (1999) showed from X-ray absorption fine-structure studies that Co, Pb, Zn and Ba in calcites all adopted octahedrallycoordinated Ca-substituted lattice positions with varying degrees of local distortion around the ions. Many trace species are present in Obi84, of which Zn and Mg are the most abundant, but the peculiar morphology is not typical of Mg-calcite. Given that the ionic radii of zinc and cobalt are nearly
315
identical, a homology with the morphologies identified by Braybrook et al. (2002) seems plausible. The presence of trace metal ions with diverse ionic radii with respect to Ca may explain also the observed radiaxial fibrous fabric in some aggregates in Obi84. Finally, Figure 10b illustrates differences in composition between alternating zones across crystallites; arguably these could be sector zones, reflecting differential trace metal incorporation in different crystallographic forms (e.g. Reeder & Grams 1987). It is of interest that despite the strong disequilibrium effects noted above, there is a consistency of mean Mg data with published literature (Huang et al. 2001) implying that deposition on multiannual timescales was not strongly out of chemical equilibrium, consistent also with the d18O data. Also d13C data in the speleothem can be regarded as equilibrium values, assuming some further degassing occurs in the field beyond that typical of the dripwaters sampled.
Phenomena responsible for trace element patterns in event laminae The coincidence of narrow optically visible laminae with trace element anomalies is striking in the Obir stalagmites. In Smith et al. (2009) we had interpreted these event laminae as forming in the same way as the infiltration laminae from Grotta di Ernesto (Borsato et al. 2007). Our understanding of the Ernesto laminae has progressed through the work of Huang et al. (2001), Fairchild et al. (2001) and Treble et al. (2003) who established a link between P enrichments in stalagmites and patterns of seasonal infiltration. Borsato et al. (2007) went much further in demonstrating a very specific hierarchy of association of trace element enrichments with infiltration laminae which they related to the relative importance of colloidal transport. The most specifically associated element with UV-fluorescent laminae was Y (a proxy also for behaviour of the heavy rare earths), followed by Cu, Pb and Zn, whereas P and Br were found to display a rather broader enrichment, centred on the infiltration lamina. Sr was depressed, which was interpreted as due to being out-competed for lattice defect sites. Ion microprobe analysis confirmed the role of Y and also showed that F, H and Na were focused around these laminae, although the latter two might reflect the presence of fluid inclusions (Borsato et al. 2007). In the infiltration lamina hypothesis, multiple pulses of trace elements associated with infiltration laminae (Figs 10 & 11) could be consistent with early work on UVfluorescence in which multiple laminae were recognized (Shopov et al. 1994) and with monitoring work in modern caves, such as the Shihua Cave
316
I. J. FAIRCHILD ET AL.
near Beijing where a series of pulses of organic carbon are found during a wet season (Tan et al. 2006; Ban et al. 2008). However, a review of all the evidence concerning the Obir laminae indicates that we need also to consider an alternative origin as aerosol laminae. The first line of evidence concerns the patterns of chemical change associated with the laminae. Of the colloidally-associated species identified from Ernesto, interferences with more abundant species prevented microanalysis of Y in Obi84, but all the other species are found enriched together in event laminae together with iodine, which has been analyzed for the first time. Our unpublished UV-fluorescence and Fourier-Transform infrared analyses indicated that the infiltration laminae at Ernesto host organic compounds (previously observed by standard UV-lamp excitation) in addition to trace metals, and literature documents that most trace metals detected could be complexed, especially by humic substances. This fits readily with our identification of iodine enrichments since there is also a large body of evidence, summarized by Steinberg et al. (2008), that a large fraction of soluble iodine is associated with colloidal humic substances in terrestrial environments. It is significant also that Cl, which shows minimal complexing effects, shows no variation in the stalagmite. Hence, this pattern of element enrichments clearly supports a similar mechanism of formation of trace element laminae at both cave sites. UV-fluorescence is not readily detectable from direct excitation of the Obir samples, but fluorescent organic matter is recovered by dissolution of the speleothem calcite in weak acid (A. Hartland, pers. comm. 2009). This is consistent with the much deeper (70 m v. 10– 15 m) position of Sa¨ulenhalle relative to the ground surface compared with Ernesto. Secondly, we need to consider the delivery of the elements to the speleothem. A mass balance can be calculated for Obi84 given the mean discharge of dripwater at this site (which varies less than the other drips studied, Table 1), the mean composition of dripwaters (Tables 1 and 4), and the mean composition of the top 10 mm of the stalagmite (Table 2). The annual discharge is 287 litres and the annual volume of calcite precipitated on the top surface of the stalagmite is c. 0.23 g (given a thickness of 0.12 mm and diameter of the top surface of 30 mm). The results, in terms of % removal of the species from the solution, are shown in Table 5 along with the distribution coefficients previously discussed. It can be seen that only a small fraction (0.63%) of Ca is removed from the dripwater and much smaller percentages of Mg, Cu and Y. However the calculated figures for Zn (5.7%) and Pb (14%) are high, but still well below 100%. Hence, on an annual basis, dripwaters could
supply sufficient trace elements. However, as has been shown, the event laminae display concentrations 3–10 times higher than outside these zones, but there is no indication of a seasonal variability in concentration, even though three of the water samples were collected in November when infiltration might be expect to be at a maximum. It appears that much higher rates of delivery of ions is required in the autumn season and yet the dripwater data do not clearly indicate higher rates of discharge at this time. One can construct a hydrological mechanism by which this could occur, using the plumbing model concepts in Tooth & Fairchild (2003) and Fairchild et al. (2006b). If a proportion of the dripwater during the autumn season derived from a fracture-fed source with high colloidal content, it could result in significant changes in dripwater chemistry. The fracture-fed flow might enter a reservoir with an overflow such that the outflow to the drip had an approximately constant pressure head, or the volume of fracture-fed water might be small. However, the weakness of the argument is the lack of clear evidence for enhanced trace element concentrations in the archived samples of autumn dripwater, and so an alternative needs to be explored. Given the uncertainty as to whether dripwater could supply sufficient elements for the event laminae, we now consider as an alternative explanation that the trace element enrichments represent deposition from aerosol in the cave system. This hypothesis only emerged after our analyses and specifically after the period of more intense cave monitoring (2001–2003) when instantaneous water samples were available at this logistically difficult site. As a result, we currently lack specific observations that could be used to support this hypothesis, so we will argue in more general terms. One point that emerged from the 14C study in Smith et al. (2009) was the result that the 14C concentration of the speleothems increased in parallel with that of the external atmosphere during the era of atmospheric bomb testing, in contrast with the normal case where the radiocarbon bomb peak is assumed to be purely related to aqueous transport from the soil and hence is delayed via soil recycling. Smith et al. (2009) concluded that there must, in addition to soil transport, be isotopic exchange of carbon between air and solution, despite the net degassing of the latter. This result was unexpected, but draws attention to the air as a potential source of chemical species for speleothem deposition. Aerosols (liquid droplets and fine particles) are generated within the cave system or brought from the outer atmosphere within the cave by air currents (Cigna & Hill 1997). Deposition from aerosols contributes to the accretion of speleothem deposits at
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
certain sites. However, the main issue in the literature has been whether speleothems could be entirely explained from aerosol deposition (Cigna & Hill 1997). At Obir, it is accepted that the main crystal growth originates from solution: the only question is whether aerosols might add to the solutes supplied from dripwater during the formation of event laminae in particular. In an extensive study on the physics of cave passages, Badino (1995) suggested that aerosols deposit onto speleothems in zones where air currents are feeble to stagnant through the phenomenon of ‘evaporation’, that is, when the dimensions of the aerosol particles decrease below a critical size (the pressure within the aerosol droplet drops and the surrounding cave air equilibrates with a pressure which is similar to that of the disappearing aerosol droplet). If these aerosols carry trace metals as particulate they may deposit their load onto the growing speleothem surface. An instructive case study is provided by the work of Jeong et al. (2003) and Chang et al. (2008) on pervasive black carbon deposits on speleothems from touristic caves in Korea. Microscopic and 14C studies were used to identify the sources of carbon as being both from vehicle exhausts and biomass burning (the latter potentially being both external and internal to the cave). Elevated levels of Mn, N, S and Pb were attributed to airborne anthropogenic sources. The organic carbon was of colloidal dimensions (20– 50 nm) and so clearly would be much more mobile than pollen grains which are known to be aerially transported only close to the entrance of cave systems (McGarry & Caseldine 2004). We must stress that although the circulation at Obir is efficient, strong draughts are not felt in the cave interior (flow of up to 0.5 m s21 at constrictions was mentioned earlier) and flows within chambers could even be laminar. Hence we do not anticipate that micron-scale particulates (e.g. rock flour) are transported, but smaller colloidal (organic-hosted) entities could be. A suggestive piece of evidence is the observed and irregular increase in air temperature at the beginning of the winter circulation regime (Fig. 1). This may imply a more vigorous circulation regime at the time, which could be equated with the event laminae that overlap with and are partly followed by the low-S winter growth. For example in autumn 2003, four distinct pulses of higher temperature can be recognized (Fig. 1), which are likely to reflect changes in the intensity of air circulation. Such pulses would be responsible for the distinct micron-scale enrichments in metal content observed in Figures 10 and 11. The chemical properties of the elements found enriched in the event laminae are distinctive and are precisely those expected from binding to organic matter, supporting their interpretation as
317
originally bound to soil humic substances. Therefore an overall two-stage hypothesis for the development of the event laminae can be tentatively proposed as: 1. Delivery of elements bound to colloidal humic substances in solutions percolating through fracture-fed components of the karstic aquifer into the cave system; and 2. Suspension of part of the colloidal material as aerosol and redistribution within the cave system. These and other models will be tested in future observations at the study site by attempting to sample air currents as well as the composition of wall deposits.
Expressions of seasonality and proxies for climate This study has been based on samples that show strongly developed seasonality. The most obvious effect observed during monitoring is the development of a stronger winter circulation, once the external temperature falls below that of the interior, associated with reduced PCO2 in cave air. The most obvious annual phenomena in the speleothems are the event laminae (of the order of 1021 years), many of which are composite of individual events representing only a few days growth. The enrichment in trace elements expected to be associated with organic colloids could be interpreted as reflected a seasonal hydrological flushing as in the model set up for Grotta di Ernesto (Smith et al. 2009). Borsato et al. (2007) found the most convincing relationship between abundance of such elements and surface conditions for the Ernesto cave stalagmite arose when a period of deforestation around World War I led to enhanced metal contents. However, in sample Obi84, although year-to-year variability in element abundances is strong, it is not related to the inter-annual variability in the intensity of autumnal rainfall (Fig. 9). Neither has cave monitoring detected changes in the flux of elements in dripwater, whereas there is evidence for unusually high and fluctuating airflows at the start of the winter season. For this reason, an aerosol origin for event laminae is proposed as an alternative model for testing in future work. The ‘event lamina’ type of seasonality appears to be characteristic of many cave sites in humid temperate environments and it will be interesting to explore the extent to which the phenomena are modified across a wider range of climatic settings. The reduction in CO2 content of cave air in the winter, as observed by monitoring, in itself leaves no physical trace in the speleothem, unlike the changes in saturation state in caves in seasonally dry climates where clear and cloudy calcites alternate
318
I. J. FAIRCHILD ET AL.
seasonally (Genty & Quinif 1996; Baker et al. 2008). However, we have demonstrated from d13C and sulphate evidence that the effects of low PCO2 in winter do have a distinct chemical expression, which accords with theory and previous experimental evidence. Since seasonal changes in PCO2 are normally found in cave monitoring studies, these effects could be recognized widely in speleothems. Unlike the Ernesto cave, the rate of speleothem growth is not proportional to temperature, but instead has decreased since the mid 19th century. It is plausible that this may relate to a reduction in the severity of winters, leading to less influence of winter ventilation, associated with high supersaturations. We do not present here our work on statistical comparison with climatic indices, but it is clear that any correlations are weak. Obir speleothems, like the physiological variation of the cave system through the year, are remarkable, and should yet yield more specific information that can be interpreted in terms of seasonality under past conditions. We are currently working on the premise that it is the length of the winter season during which strong ventilation applies (currently around 4 months) which is the key factor and have recently obtained more high-resolution S profiles to try to test S variation as a proxy for the duration of the winter season.
Conclusions The results of a comprehensive range of geochemical and petrological studies of stalagmites from a chamber in the interior of the humid alpine Obir cave site, 70 m below the surface, show that: (a) The speleothems have a distinct Zn- and Pbenriched chemistry, related to local host-rock mineralization. This has resulted in changed unit-cell size, and a distinctive microstructure is present of sub-parallel crystallites with alternating rough and smooth surfaces, the latter representing the preferential development of one specific face in the f10 –11g form. The distinctive chemistry has not significantly affected fractionation of stable carbon and oxygen isotopes or the bulk partitioning of Mg. (b) A pronounced visible lamination, forming during the autumn, takes the form of annual packages or individual ‘event laminae’ separated by dominant inclusion-free calcite. In structure and spacing, they are similar to laminae associated with fluorescent organic matter in other humid temperate sites; fluorescent organic matter is present, but less abundant at the deep Obir site. (c) The event laminae are associated with distinct trace element enrichments of Zn, Pb and
P. Micrometre-scale zones of trace element enrichment within broader event laminae and the thin zones must represent pulses of trace-element incorporation during not more than a few days growth. Such short events have only rarely been imaged before, and only by UV-fluorescence imaging, not by chemical analysis. They may be present widely in speleothems from humid temperate environments and are particularly clearly shown here because of fortuitous analytical conditions. (d) The enrichments in Zn and Pb result in X-ray interferences that prevent quantitative determination of some other elements, but there is sufficient evidence to draw a close parallel in element behaviour with event laminae at Ernesto cave where a spatial and interpreted genetic association with colloidal humic substances is found. This comparison is reinforced by enrichments also in the halides F, Br and I since these elements also display organic associations in near-surface environments. This is the first demonstration of I enrichments in speleothems. There is a clear contrast with the uniform stalagmite concentrations of Cl, which exists as a free ion in solution. The evidence from Obir strengthens the idea of a characteristic suite of elements, transported in colloidal form, which tend to covary in speleothems, and be associated with a particular season in mid-latitudes. (e) A quantitative comparison between chemistries of archived dripwaters and speleothems draws attention to very different selectivities of element partitioning into calcite, but an origin of seasonally high Pb and Zn purely from local dripwater is only possible if virtually all of these metals is sequestered into calcite. Because of this observation, and a lack of evidence for significantly higher drip discharge during the autumn, it is proposed that chemical enrichments and the formation of event laminae might arise by redistribution of colloids as aerosols during particularly intense air flows at the start of the winter season. This draws attention to aerosol analysis as an under-researched area in cave studies. (f) The strongly developed seasonal variation in airflow at Obir results in low PCO2 in cave air in winter, which is associated with higher d13C and pH of dripwaters. Enrichments in 13 C were indeed found in winter growths of most laminae traversed during ion microprobe studies, but the observations were limited by the extreme technical difficulties in analysis. Sulphate also displays a seasonal variation, offset from the distribution of P, and which is
ANNUALLY LAMINATED STALAGMITES, AUSTRIA
consistent with pH controlling its incorporation. Sulphate analyses could more generally provide a means of examining the impact of seasonal changes in cave circulation on speleothems. (g) The effects of hydrology and air circulation, aspects of cave physiology, each have roles to play in imparting a record of seasonality of deposition. Seasonality might be recorded in speleothems generally and could be recognized whenever they grow sufficiently quickly for it to be determined. This work was initiated during a Royal Society joint project grant between IJF and CS with assistance from Anna Tooth, and funding continued under the UK Natural Environment Research Council grants NER/T/ S/2002/00448 and NE/C511805/1 (IJF principal investigator) and funding from the European Synchrotron Facility (allocations ME1103 and EC104, SF principal investigator). CS acknowledges laboratory assistance by Manuela Wimmer, logistic support in Obir cave by Harald Langer, and partial support from FWF grant Y122-GEO. EIMF staff associated with this work are Nicola Cayzer, John Craven and Richard Hinton. We thank staff at the former NERC ICP-MS facility at Kingston University for their assistance and Dr Louise Male (University of Birmingham) for the X-ray diffraction data.
References B ADINO , G. 1995. Fisica del clima sotterraneo. Memorie dell’Istituto Italiano di Speleologia, 7, Serie II. Bologna. B AKER , A., G ENTY , D., D REYBRODT , W., B ARNES , W. L., M OCKLER , N. J. & G RAPES , J. 1998. Testing theoretically predicted stalagmite growth rate with recent annually laminated samples: implications for past stalagmite deposition. Geochimica et Cosmochimica Acta, 62, 393 –404. B AKER , A., S MITH , C. L., J EX , C., F AIRCHILD , I. J., G ENTY , D. & F ULLER , L. 2008. Annually laminated speleothems: a review. International Journal of Speleology, 37, 193– 206. B ALDINI , J. U. L., M C D ERMOTT , F. & F AIRCHILD , I. J. 2006. Spatial variability in cave drip water hydrochemistry: implications for stalagmite paleoclimate records. Chemical Geology, 235, 290–304. B AN , F., P AN , G., Z HU , J., C AI , B. & T AN , M. 2008. Temporal and spatial variations in the discharge and dissolved organic carbon of drip waters in Beijing Shihua Cave, China. Hydrological Processes, 22, 3749–3758. B ORSATO , A., F RISIA , S., F AIRCHILD , I. J., S OMOGIY , A. & S USINI , J. 2007. Trace element distribution patterns in annual stalagmite laminae mapped by micrometerresolution X-ray fluorescence: incorporation of colloidally-transported species. Geochimica et Cosmochimica Acta, 71, 1494–1512. B RAYBROOK , A. L., H EYWOOD , B. R., J ACKSON , R. A. & P ITT , K. 2002. Parallel computational and experimental studies of the morphological modification
319
of calcium carbonate by cobalt. Journal of Crystal Growth, 243, 336–344. B USENBERG , E. & P LUMMER , L. N. 1985. Kinetic and thermodynamic factors controlling the distribution of and Naþ in calcites and selected aragonites. SO22 4 Geochimica et Cosmochimica Acta, 49, 713–725. C HADA , V. G. R., H AUSNER , D. B., S TRONGIN , D. R., R OUFF , A. A. & R EEDER , R. J. 2005. Divalent Cd and Pb uptake on calcite f10–14g cleavage faces: an XPS and AFM study. Journal of Colloid and Interface Science, 288, 350–360. C HANG , S. J., J EONG , G. Y. & K IM , S. J. 2008. The origin of balck carbon on speleothems in tourist caves in South Korea: chemical characterization and source discrimination by radiocarbon measurement. Atmospheric Environment, 42, 1790–1800. C IGNA , A. A. & H ILL , C. A. 1997. Aerosols: are they a mechanism of speleothem growth? In: H ILL , C. & F ORTI , P. (eds) Cave Minerals of the World. National Speleological Association, Huntsville, Alabama, 255– 258. C URTI , E. 1999. Coprecipitation of radionuclides with calcite: estimation of partition coefficients based on a review of laboratory investigations and geochemical data. Applied Geochemistry, 14, 433– 445. D ICKSON , J. A. D. 1993. Crystal-growth diagrams as an aid to interpreting the fabrics of calcite aggregates. Journal of Sedimentary Petrology, 63, 1 –17. D REYBRODT , W. 1988. Processes in Karst Systems. Springer, Berlin. F AIRCHILD , I. J. & M C M ILLAN , E. A. 2007. Speleothems as indicators of wet and dry periods. International Journal of Speleology, 36, 79–84. F AIRCHILD , I. J. & T REBLE , P. C. 2009. Trace elements in speleothems as recorders of environmental change. Quaternary Science Reviews, 28, 449–468. F AIRCHILD , I. J., B ORSATO , A. ET AL . 2000. Controls on trace element (Sr– Mg) compositions of carbonate cave waters: implications for speleothem climatic records. Chemical Geology, 166, 255– 269. F AIRCHILD , I. J., B AKER , A., B ORSATO , A., F RISIA , S., H INTON , R. W., M C D ERMOTT , F. & T OOTH , A. F. 2001. Annual to sub-annual resolution of multiple trace-element trends in speleothems. Journal of the Geological Society, London, 158, 831–841. F AIRCHILD , I. J., S MITH , C. L. ET AL . 2006a. Modification and preservation of environmental signals in speleothems. Earth-Science Reviews, 75, 105– 153. F AIRCHILD , I. J., T UCKWELL , G. W., B AKER , A. & T OOTH , A. F. 2006b. Modelling of dripwater hydrology and hydrogeochemistry in a weakly karstified aquifer (Bath, UK): implications for climate change studies. Journal of Hydrology, 321, 213–231. F AIRCHILD , I. J., F RISIA , S., B ORSATO , A. & T OOTH , A. F. 2007. Speleothems. In: N ASH , D. J. & M C L AREN , S. J. (eds) Geochemical Sediments and Landscapes. Blackwell, Oxford, 200– 245. F ERNA´ NDEZ -D IAZ , L., A STILLEROS , J. M. & P INA , C. M. 2006. The morphology of calcite crystals grown in a porous medium doped with divalent cations. Chemical Geology, 225, 314– 321. F RAPPIER , A., S AHAGIAN , D., G ONZALEZ , L. A. & C ARPENTER , S. J. 2002. El Nin˜o events recorded by stalagmite carbon isotopes. Science, 298, 565.
320
I. J. FAIRCHILD ET AL.
F RIEDMAN , I. & O’N EIL , J. R. 1977. Compilation of stable isotope fractionation factors of geochemical interest. US Geological Survey Professional paper, 440-KK, 1– 49. F RISIA , S., B ORSATO , A., F AIRCHILD , I. J. & M C D ERMOTT , F. I. J. 2000. Calcite fabrics, growth mechanisms and environments of formation in speleothems (Italian Alps and SW Ireland). Journal of Sedimentary Research, 70, 1183– 1196. F RISIA , S., B ORSATO , A., P RETO , N. & M C D ERMOTT , F. 2003. Late Holocene annual growth in three Alpine stalagmites records the influence of solar activity and the North Atlantic Oscillation on winter climate. Earth and Planetary Science Letters, 216, 411–424. F RISIA , S., B ORSATO , A., F AIRCHILD , I. J. & S USINI , J. 2005. Variations in atmospheric sulphate recorded in stalagmites Variations in atmospheric sulphate recorded in stalagmites by synchrotron micro-XRF and XANES analyses. Earth and Planetary Science Letters, 235, 729–740. G ENTY , D. & Q UINIF , Y. 1996. Annually laminated sequences in the internal structure of some Belgian stalagmites – importance for paleoclimatology. Journal of Sedimentary Research, 66, 275– 288. G ODELITSAS , A., A STILLEROS , J. M., H ALLAM , K., H ARISSOPOULOS , S. & P UTNIS , A. 2003. Interaction of calcium carbonates with lead in aqueous solutions. Environmental Science and Technology, 37, 3351–3360. H UANG , Y. & F AIRCHILD , I. J. 2001. Partitioning of Sr2þ and Mg2þ into calcite under karst-analogue experimental conditions. Geochimica et Cosmochimica Acta, 65, 47– 62. H UANG , H. M., F AIRCHILD , I. J., B ORSATO , A., F RISIA , S., C ASSIDY , N. J., M C D ERMOTT , F. & H AWKESWORTH , C. J. 2001. Seasonal variations in Sr, Mg and P in modern speleothems (Grotta di Ernesto, Italy). Chemical Geology, 175, 429 –448. J EONG , G. Y., K IM , S. J. & C HANG , S. J. 2003. Black carbon pollution of speleothems by fine urban aerosols in tourist caves. American Mineralogist, 88, 1872–1878. J IMENEZ -L OPEZ , C. C., R ODRIGUEZ -N AVARRO , A., D OMINGUEZ -V ERA , J. M. & G ARCIA -R UIZ , J. M. 2003. Influence of lysozym on the precipitation of calcium carbonate: a kinetic and morphologic study. Geochimica et Cosmochimica Acta, 67, 1667– 1676. J OHNSON , K. R., H U , C., B ELSHAW , N. S. & H ENDERSON , G. M. 2006. Seasonal trace-element and stable isotope variations in a Chinese speleothem: the potential for high resolution paleomonsoon reconstruction. Earth and Planetary Science Letters, 244, 394–407. K ENDALL , A. C. & B ROUGHTON , P. L. 1978. Origin of fabric in speleothems of columnar calcite crystals. Journal of Sedimentary Petrology, 48, 550 –552. K IM , S. T. & O’N EIL , J. R. 1997. Equilibrium and nonequilibrium oxygen isotope effects in synthetic carbonates. Geochimica et Cosmochimica Acta, 61, 3461–3475. L ARSON , A. C. & V ON D REELE , R. B. 2000. General Structure Analysis System (GSAS). Los Alamos National Laboratory Report, LAUR 86– 748. M ACKENZIE , F. T., B ISCHOFF , W. D., B ISHOP , F. C., L OIJENS , M., S CHOONMAKER , J. & W OLLAST , R.
1983. Magnesian calcites: low-temperature occurrence, solubility and solid-solution behavior. In: R EEDER , R. J. (ed.) Carbonates: Mineralogy and Geochemistry. Reviews in Mineralogy, 11, 97–144. M ATTEY , D., L OWRY , D., D UFFET , J., F ISHER , R., H ODGE , E. & F RISIA , S. 2008. A 53 year seasonally resolved oxygen and carbon isotope record from a modern Gibraltar speleothem: reconstructed drip water and relationship to local precipitation. Earth and Planetary Science Letters, 269, 80–95. M C D ERMOTT , F. 2004. Palaeo-climate reconstruction from stable isotope variations in speleothems: a review. Quaternary Science Reviews, 23, 901– 918. M C D ERMOTT , F., S CHWARCZ , H. P. & R OWE , P. J. 2005. Isotopes in speleothems. In: L ENG , M. J. (ed.) Isotopes in Palaeoenvironmental Research. Springer, Dordrecht, 185–225. M C G ARRY , S. F. & C ASELDINE , C. 2004. Speleothem palynology: an undervalued tool in Quaternary studies. Quaternary Science Reviews, 23, 2389–2404. M IORANDI , R., B ORSATO , A., F RISIA , S., F AIRCHILD , I. J. & R ICHTER , D. K. in press. Epikarst hydrology and implications for stalagmite capture of climate changes at Grotta di Ernesto (N.E. Italy): results from long-term monitoring. Hydrological Processes. M U¨ HLINGHAUS , C., S CHOLZ , D. & M ANGINI , A. 2007. Modelling stalagmite growth and d13C as a function of drip interval and temperature. Geochimica et Cosmochimica Acta, 71, 2780–2790. N EUSER , R. D. & R ICHTER , D. K. 2007. Non-marine radiaxial fibrous calcites – examples of speleothems proved by electron backscatter diffraction. Sedimentary Geology, 194, 149 –154. O NAC , B. P. 1997. Crystallography of Speleothems. In: H ILL , C. & F ORTI , P. (eds) Cave Minerals of the World. National Speleological Association, Huntsville, Alabama, 230– 235. R EEDER , R. J. 1983. Crystal chemistry of the rhombohedral carbonates. In: R EEDER , R. J. (ed.) Carbonates: Mineralogy and Geochemistry. Reviews in Mineralogy, 11, 1– 47. R EEDER , R. J. 1996. Interaction of divalent cobalt, zinc, cadmium, and barium with the calcite surface during layer growth. Geochimica et Cosmochimica Acta, 60, 1543– 1552. R EEDER , R. J. & G RAMS , J. C. 1987. Sector zoning in calcite cement crystals: Implications for trace element distributions in carbonates. Geochimica et Cosmochimica Acta, 51, 187– 194. R EEDER , R. J., L AMBLE , G. M. & N ORTHRUP , P. A. 1999. XAFS study of the coordination and local relaxation around Co2þ, Zn2þ, Pb2þ, and Ba2þ trace elements in calcite. American Mineralogist, 84, 1049– 1060. R IMSTIDT , J. D., B ALOG , A. & W EBB , J. 1998. Distribution of trace elements between carbonate minerals and aqueous solutions. Geochimica et Cosmochimica Acta, 62, 1851–1863. R OBERTS , M. S., S MART , P. & B AKER , A. 1998. Annual trace element variations in a Holocene speleothem. Earth and Planetary Science Letters, 154, 237– 246. S HOPOV , Y. Y., F ORD , D. C. & S CHWARCZ , H. P. 1994. Luminescent microbanding in speleothems: highresolution chronology and palaeoclimate. Geology, 22, 407– 410.
ANNUALLY LAMINATED STALAGMITES, AUSTRIA S MITH , C. L., B AKER , A., F AIRCHILD , I. J., F RISIA , S. & B ORSATO , A. 2006. Reconstructing hemispheric-scale climates from multiple stalagmite records. International Journal of Climatology, 26, 1417– 1424. S MITH , C. L., F AIRCHILD , I. J., S PO¨ TL , C., F RISIA , S., B ORSATO , A., M ORETON , S. G. & W YNN , P. M. 2009. Chronology-building using objective identification of annual signals in trace element profiles of stalagmites. Quaternary Geochronology, 4, 11–21. S OLE´ , V. A., P APILLON , E., C OTTE , M., W ALTER , P. & S USINI , J. 2007. PyMCA: a multiplatform code for the analysis of energy-dispersive X-ray fluorescence spectra. Spectrochimica Acta B, 62, 63–68. S PO¨ TL , C. & M ATTEY , D. P. 2006. Stable isotope microsampling of speleothems: a comparison of manual drill, micromill and laser ablation techniques. Chemical Geology, 235, 48–58. S PO¨ TL , C. & V ENNEMANN , T. W. 2003. Continuous-flow isotope ratio mass spectrometric analysis of carbonate minerals. Rapid Communications in Mass Spectrometry, 17, 1004–1006. S PO¨ TL , C., F AIRCHILD , I. J. & T OOTH , A. F. 2005. Cave air control on dripwater geochemistry, Obir caves (Austria): implications for speleothem deposition in dynamically ventilated caves. Geochimica et Cosmochimica Acta, 69, 2451–2468. S TEINBERG , S. M., K IMBLE , G. M., S CHMETT , G. T., E MERSON , D. W., T URNER , M. F. & R UDIN , M. 2008. Abiotic reaction of iodate with sphagnum peat
321
and other natural organic matter. Journal of Radioanalytical and Nuclear Chemistry, 277, 185–191. T AN , M., B AKER , A., G ENTY , D., S MITH , C., E SPER , J. & C AI , B. 2006. Applications of stalagmite laminae to paleoclimate reconstructions: comparison with dendrochronology/climatology. Quaternary Science Reviews, 25, 2103–2117. T OOTH , A. F. & F AIRCHILD , I. J. 2003 Soil and karst aquifer hydrological controls on the geochemical evolution of speleothem-forming drip waters, Crag Cave, southwest Ireland. Journal of Hydrology, 273, 51–68. T REBLE , P., S HELLEY , J. M. G. & C HAPPELL , J. 2003. Comparison of high resolution sub-annual records of trace elements in a modern (1911– 1992) speleothem with instrumental cliate data from southwest Australia. Earth and Planetary Science Letters, 216, 141–153. V OLLWEILER , N., S CHOLZ , D., M U¨ HLINGAHUS , C., M ANGINI , A. & S PO¨ TL , C. 2006. A precisely dated climate record for the last 9 kyr from high alpine stalagmites. Geophysical Research Letters, 33, L20703, doi:10.1029/2006GL027662. W YNN , P. M., F AIRCHILD , I. J., S PO¨ TL , C. EIMF. 2010. High resolution sulphur isotope analysis of speleothem carbonate by ion microprobe. Chemical Geology, 271, 101– 107. Z ACHARA , J. M., C OWAN , C. E. & R ESCH , C. T. 1991. Sorption of divalent metals on calcite. Geochimica et Cosmochimica Acta, 55, 1549–1562.
Seasonal microclimate control of calcite fabrics, stable isotopes and trace elements in modern speleothem from St Michaels Cave, Gibraltar DAVID P. MATTEY1,*, IAN J. FAIRCHILD2, TIM C. ATKINSON3, JEAN-PAUL LATIN4, MARK AINSWORTH4 & RICHARD DURELL4 1
Department of Earth Sciences, Royal Holloway University of London, TW20 0EX, UK
2
School of Geography, Earth & Environmental Sciences, University of Birmingham B15 2TT, UK 3
Department of Earth Sciences, University College London, WC1E 6BT, UK
4
Cliffs and Caves Section, Gibraltar Ornithological and Natural History Society, Gibraltar *Corresponding author (e-mail:
[email protected]) Abstract: Detailed monitoring of three drip sites in New St Michael’s Cave, Gibraltar, reveals a strongly coherent seasonal pattern of dripwater chemistry despite each site having significantly different flow paths and discharge patterns. Calcite saturation is closely linked to regular seasonal variations in cave air pCO2 caused by seasonally reversing ventilation driven by temperature difference between the cave interior and the air outside. A coupled model of CO2 degassing and calcite precipitation links seasonal d13C variations in coexisting dripwater, cave air CO2 and speleothem calcite to large variations in pCO2 that are driven by cave ventilation. The relationships between stable isotope ratios, Sr/Ca and speleothem fabrics across annually formed calcite laminae are consistent with a degassing– calcite precipitation process in which rapid degassing controls the d13C of both drip water DIC and calcite whereas a much slower rate of calcite precipitation causes seasonal cycles of Sr in a more complex manner. By demonstrating the causes of laminated speleothem fabrics plus trace element and isotope cycles in modern speleothem from a closely monitored cave, this study provides clear links between the local microclimate and the proxy record provided by speleothem geochemistry. In Gibraltar, low cave air pCO2 in summer is unusual compared to what has been revealed by cave monitoring carried out elsewhere and shows that caution is needed when linking paired speleothem fabrics to specific seasons without knowledge of local processes operating in the cave.
Speleothems provide continuous and precisely dated records of past environmental change which have advanced understanding of climate variability on timescales from glacial –interglacial cycles (Wang et al. 2008) down to seasonal patterns of precipitation (Borsato et al. 2007). Speleothem deposition in stable cave environments can record changes in surface climate as variations in properties such as extension rate, trace element abundances and stable isotopes (McDermott 2004; Fairchild et al. 2006a) but the causal relationships between these proxies and climate are not always fully understood. For some proxies they appear to be straightforward, for example the dependence of extension rate on the amount of rainfall (Baker et al. 2008), but for others such as stable isotopes interpretations have often been based on assumptions and guesswork regarding the aspects of climate that are most closely reflected. Ideally, any proxy-climate transfer function used should be based on a full understanding of the physico-chemical workings of the local
climate –karst-cave system and its influences on the recording process. Careful, multi-annual monitoring of the cave microclimate, dripwater chemistry and calcite growth mechanisms reveals some of the local effects that may modify the climate recording process. Some of the important issues include the relationships between precipitation, recharge, drip rates and solute chemistry (Bottrell & Atkinson 1991; Genty & Deflandre 1998; Baker & Brunsdon 2003; Tooth & Fairchild 2003; Cruz et al. 2005; Baldini et al. 2006; Genty 2008), the role of seasonal ventilation and degassing (Ek & Gewelt 1985; Bar-Matthews et al. 1996; Spo¨tl et al. 2005; Banner et al. 2007; Baldini et al. 2008) and the impact of kinetic factors such as fast degassing or crystal chemical effects on CaCO3 growth (Hendy 1971; Mickler et al. 2004, 2006). Knowledge of these local processes and their stability through time are a critical step in the derivation of reliable climate –proxy transfer functions that can be used for quantitative climate reconstruction.
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 323–344. DOI: 10.1144/SP336.17 0305-8719/10/$15.00 # The Geological Society of London 2010.
324
D. P. MATTEY ET AL.
Several recent studies of speleothem records at high resolutions have revealed climate features on seasonal (Treble et al. 2003; Johnson et al. 2006; Banner et al. 2007; Mattey et al. 2008) or even synoptic time scales (Frappier et al. 2002) which provide the critical direct link between the local weather and how it is recorded during the speleothem deposition process. Speleothem calcite deposition in caves is commonly seen to be cyclical, resulting in the development of laminae defined by alternating pairs of fabrics (Baker et al. 2008; Genty & Quinif 1996). Using constraints from growth on dated artefacts and 14C analyses (Baldini et al. 2005; Genty et al. 2001; Tan et al. 2006; Mattey et al. 2008) cyclic laminae have sometimes been shown to be annual features and related to strong seasonality of the local climate. Annual growth laminae provide a means of deriving a chronology at the best possible precision, and may also preserve trace element and stable isotope patterns that can be related to the local climate and hydrological cycle. Our work in Gibraltar combines comprehensive multi-annual cave monitoring with high resolution analyses of fabric, trace elements and stable isotopes in modern speleothem. A recent study of a modern stalagmite from New St. Michaels Cave (Mattey et al. 2008) revealed annual growth laminae which preserve exceptionally well-defined seasonal d13C and d18O cycles linked to ventilation. We were able to identify the d18O of winter dripwater from the complex seasonally resolved speleothem record and show excellent inter-annual correspondence with the d18O of winter precipitation. In the present paper we present a more detailed overview of the results of the first 4 years of cave environment monitoring which includes local meteorology, cave and soil temperature, humidity and pCO2, and of drip discharge and monthly analysis of drip water for trace element and isotopic analysis. The monitoring data enable the seasonally resolved speleothem fabric, trace element and isotope record to be precisely linked to the nature and timing of local processes in the soil, cave air and local climate. We propose a coupled CO2 degassing–calcite precipitation model which links the development of annual cycles in d13C and Sr with the effects of seasonal cave ventilation.
Regional setting, monitoring techniques and analytical methods Old and New St Michaels Caves The Rock of Gibraltar, located where the Atlantic meets the Mediterranean at the junction of Europe and Africa, forms a North –South trending ridge
2.5 km long with a maximum elevation of 423 m (Fig. 1). The ridge is asymmetric, having a steep to near-vertical eastern slope which is partly banked by Pleistocene sand-dune deposits, and a western slope falling more steadily at 358 towards the town of Gibraltar near sea level. Above 100 m altitude the western slopes are covered in Mediterranean scrub forest with low rock outcrops of the Gibraltar limestone. The peninsula of Gibraltar links the Betic and Rif mountain chains, which form the southwestern end of the Mediterranean Alpine belt and is mainly composed of early Jurassic age limestone and dolomite which form the lower limb of an overturned nappe (Rose & Rosenbaum 1991). These beds dip steeply to the west and although there are no surface streams, swallets or resurgence features, the dolomites and limestones contain many solution caves located at altitudes ranging from below present sea level to near the summit ridge at over 400 m. Many caves have natural entrances exposed by erosion, but other significant caves have also been revealed though tunneling (Rosenbaum & Rose 1991). The location and a plan of St Michaels Cave is shown on Figures 1 and 2. Old St Michaels Cave (OSM) (Shaw 1955) has been known since Roman times and is open to tourists as a show cave. The cave has developed in faulted dolomitic limestone creating a large main chamber. Dissolution has also followed bedding planes, creating minor caves linked to OSM and forming natural entrances to the system. During World War II, a new access tunnel was driven into the lowest part of the show cave, known as the Hospital, exposing a lower series of solution rifts leading southwards along the strike of the Gibraltar limestone at an altitude of around 325 m (Fig. 2). This system, New St Michaels Cave (NSM) (Shaw 1954), rivals the old show cave system in terms of the scale of speleothem decoration, and also contains a 6 m deep lake which accumulates water from seepages and drips entering the southern part of NSM. Gibraltar caves such as the St Michaels system preserve evidence of phreatic origins and have since undergone several phases of draining and decoration with secondary speleothem deposits (Tratman 1971). Because the present altitude of the large St Michaels system is over 300 m asl, the phreatic features indicate that these caves have undergone significant uplift to their present position (Tratman 1971; Rose & Rosenbaum 1991; Rodrigues-Vidal et al. 2004). Tunnelling near sea level in the 19th century revealed more large natural caves such as the Ragged Staff system with similar overall morphology but with far less speleothem deposition. Ragged Staff Cave contains brackish lakes with water filled passages extending
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
325
Fig. 1. Location of Gibraltar, St. Michaels cave and other features described in the text. Figure adapted from Mattey et al. (2008).
below sea level suggesting that initial phases of dissolution may have taken place by mixing near the freshwater-seawater interface via processes similar to flank margin cave formation (Mylroie & Carew 1988; Romanov & Dreybrodt 2006). Wave-cut platforms (90 and 130 m asl) are a prominent feature of the geomorphology of Gibraltar (Rodriguez-Vidal et al. 2004) recording higher stands of sea level in the past, and although the age and development history of Gibraltar caves are not yet well understood, solution voids and their subsequent decoration by speleothem deposition may be controlled by neotectonic uplift in conjunction with sea level fluctuations between glacial and interglacial periods (Rodriguez-Vidal et al. 2004). OSM has a number of natural entrances as well as artificial high and low level tunnels used for tourist access. The showcave develops strong
natural chimney ventilation (Wigley & Brown 1976; Atkinson et al. 1983) between high- and low-level entrances and is known to show seasonal reversals in flow direction. In winter, warm cave air flows upwards drawing cool outside air into the lower entrances whereas in summer cave air flows out of lower entrances, drawing atmosphere into upper entrances. Before modern development for tourism the natural entrances would all have been at a higher level and chimney ventilation may not have been as strong a feature as it is at the present time. The NSM cave system has no known natural entrances and remained unconnected to OSM until the link was made in 1942 with the lowest part of main show cavern. Seasonally reversing chimney ventilation can also be detected though the 1 m2 trapdoor link between OSM and NSM, although sealing this connection for one month in August 2006 was found
326
D. P. MATTEY ET AL.
Fig. 2. Plan of the St. Michaels Cave systems based on the survey by Shaw (1954) and figures from Rose & Rosenbaum (1991). Locations of monitoring sites and features discussed in the text are shown.
to have no impact on CO2 levels and ventilation of NSM as a whole (our unpublished data). Deeper within the NSM system very weak air
flow patterns can be detected using smoke tracers, but the dramatic seasonal fluctuations in CO2 levels described by Mattey et al. 2008 and in
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
this work suggest the existence of connections to the surface or other caves which allows large-scale natural advective transport of air through the entire system.
Cave environmental monitoring All meteorological data were recorded at the RAF Meteorological Office located 3 km away from the cave (Fig. 1). The cave monitoring, sampling and analysis program obtained data via continuous logging (l) and spot measurements or sample(s) taken during regular monthly visits from June 2004–April 2008, for the following parameters: 1.
2. 3.
cave air: pressure (l), temperature (l) and relative humidity (l,s), mixing ratios of CO2 (l,s) and CH4 (s), analysis of d13CCO2 and d13CCH4 (s) soil: temperature (l) and soil air mixing ratios of CO2 and CH4 (s), analysis of d13CCO2 and d13CCH4 (s) drip water: discharge (l); pH, total alkalinity, electrical conductivity, cations, d18O and dD (s)
The locations of environmental monitoring sites are shown on the cave plan in Figure 2. Seasonal variations in temperature and relative humidity in ambient air at the cave entrance, within the main chambers of the cave system and within the soil zone above the cave (temperature only) were measured by continuous logging supplemented by spot measurements made using hand held instruments during monthly visits. Temperature and humidity were measured by continuous logging using single- (temperature) and dual-channel (temperature and humidity) loggers from Gemini Data loggers deployed outside above the cave entrance tunnel, and in the Hospital, Dark Rift and Gib04a sites within the cave (Fig. 2). Soil temperature at 50 cm depth was logged at the soil CO2 sampling site (see below). Temperature-logging at a resolution of 0.4 8C early in the project was improved to 0.18C resolution at some sites from 2007. Spot measurements were made at the time of sampling at air and drip water sampling sites and all temperature measurements were corrected relative to a BS certified mercury thermometer to an accuracy of +0.05 8C.
Surveying and dye tracing A line survey was carried out to accurately position the map of NSM made by Shaw (1954) relative to surface topography. This showed that the soil monitoring site is located almost directly over the cave (Fig. 2) and allowed us to identify sites for injecting
327
dyes with the aim of tracing connections between infiltrating rainwater and points where seepage enters the cave. Following methodology described by Smart (1976) and Bottrell & Atkinson (1992), two aliquots of the fluorescent dye Photine CU were simultaneously released into excavated fissures in bedrock at nearby points above the Gib04a site in the cave. A second dye, Diphenyl Brilliant Flavine DY96, was washed into scree at a third point at a higher altitude, located where the dip of the Gibraltar limestone projected upwards from Gibo4a in the cave intersects the surface near the summit ridge (Fig. 3). The injections were made in March 2007 during a period of high rainfall, and the dyes were washed into the ground by runoff from sheets of polythene placed for this purpose. Within the cave, a network of cotton detectors were deployed at 27 stations. These provided regular monitoring for both tracers for 85 days after the date of injection, as well as a period of background monitoring for 59 days beforehand.
Drip water discharge rates, sampling and analysis Drip water was collected in a HDPE beaker containing an acoustic drip rate logger (Collister & Mattey 2008) and fitted with an outlet tube leading to a 1.5 l storage reservoir open to cave air. The drip logger counted the total drips falling in 30 minute intervals and counts were converted to discharge in l/d based on a mean drop volume of 0.15 ml (Collister & Mattey 2008) for all sites. Aliquots of water were stored in 100 ml HDPE doubly sealed bottles and pH, conductivity and alkalinity were measured within two hours of sampling. pH was determined using a Jenway pH and conductivity probe which was calibrated at pH 7 and 10 before the sampling session. Alkalinity was determined on site by titration against bromocresol green using a Hach digital titrator. Dripwater cation analyses were determined by ICP-AES in the Department of Earth Sciences, Royal Holloway with a total precision of less than 5%. All isotopic analyses in this study were carried out using GV Instruments Multiflow –Isoprime systems at Royal Holloway. Drip water d13C DIC was determined by acidification of 0.5 ml of water with orthophosphoric acid and equilibrating for 7 h at 40 8C. d13C values were normalized to the V-PDB scale via a calibrated sodium bicarbonate internal standard and have an external reproducibility of better than 0.08‰.
Air sampling and analysis Spot samples of background atmosphere, cave air and soil gas were collected in 1 l or 3 l Tedlar
328
D. P. MATTEY ET AL.
Fig. 3. Geological setting of St. Michaels Cave showing injection sites for the yellow dye DY96 (yellow) and blue dye Photine CU (blue) together with schematic flow pathways. Bedrock geology adapted from the map compiled by Rosenbaum & Rose (1991). Cave plan located relative to 1000 m UTM grid co-ordinates using data from a new line survey carried out in 2007.
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
bags using a small hand-held pump for analysis of mixing ratios and the d13C of CO2. Soil gas samples were collected initially using a steel tube inserted into the soil and from June 2007 using porous PTFE gas sampling cells permanently buried at 25 cm and 50 cm depth. Mixing-ratio analysis of CO2 was carried out using a LiCor 6252 NDIR analyser calibrated against NOAA standards. Isotope analyses of CO2 were determined using a GV Instruments Trace Gas – Isoprime system and precisions (1s) for 10 consecutive analyses of the secondary standard tank were between 0.03– 0.05‰ for carbon dioxide d13C analysis.
Results Cave temperature and humidity The variation of temperature and humidity within the OSM and NSM cave systems generally decrease as a function of distance from entrances, and temperatures approach the local mean annual values in the deeper parts of NSM. In detail, though, these variations are quite complex both in terms of spatial position within the cave, and as a function of the time of year. Variations in cave air temperature and humidity at two of the monitoring sites, the Hospital (the lowest part of the show cave, Fig. 2) and at the GIb04a site (representative of the distal areas of the NSM system, Fig. 2) measured by continuous logging and monthly using hand held instruments are summarized in Figure 4. Figure 4a compares three NSM cave air temperatures measured in 1948, 1954 and 1958 (Shaw 1955; Tratman 1971), the mean annual temperature (MAT) at the cave entrance measured in this study between 2004– 2008 and the local MAT recorded by the Meteorological Office since 1940. The MAT plotted on Figure 4 are sea level data corrected to the corresponding temperature at 325 m asl using a lapse rate calculated from over 1300 measurements of the difference in mean daily temperature measured at the Hospital tunnel entrance (Fig. 2) and at sea level by the Meteorological Office station (1.22 + 0.25 8C). The three historic spot measurements show a 1.6 8C range and cannot be distinguished from modern temperatures which are within error of the MAT corrected for altitude. Unfortunately there are insufficient data to assess whether cave temperatures tracked the decreasing and then rising trend in MAT observed since 1948 (Fig. 4). Temperatures monitored by continuous logging appear to be relatively constant in the deep cave (, +0.4 8C) but show marked seasonal variation of around 3 8C in the Hospital area of the show cave (Fig. 4c, d). Relative humidity measurements average 95 + 3% in the deep cave but are again
329
are much more variable in the Hospital area (Fig. 4b). Here temperatures are significantly lower than the MAT of 17.5 8C all year round and fail to reach MAT even by the end of the summer (Fig. 4d). Humidity in the Hospital area is also lowest in winter when chimney ventilation is vigorously pulling cold outside air through the show cave (Fig. 4b). The permanently below-MAT temperatures of the showcave indicate that winter ventilation removes sufficient heat from cave wall rock such that the mean annual temperature of the Hospital area (16.3 8C) remains over 1 8C cooler than the annual outside MAT (17.5 8C) (cf. Atkinson et al. 1983). Conversely, the slightly higher MAT of the deeper NSM (17.9 8C) may be indicative of advected warmer air rising from lower levels in the rock, a hypothesis developed in more detail below to explain the cave air CO2 data. Spot measurements taken over the four-year period are more variable than data returned by temperature logging recording spot measurements at 30-minute intervals and suggest that there may some local short term variation of up to +0.5 8C at the Gib04a and other deeper sites in NSM. Although the 0.4 8C resolution of the early logging devices would not have fully resolved such small variations in air temperature, logging since June 2008 at an improved 0.1 8C resolution fails to record regular seasonality (Fig. 4) indicating that small fluctuations in air temperature may occur even in the more distal regions of OSM and are too rapid to be clearly resolved by temperature loggers recording at 30-minute intervals.
Drip-rate variations and recharge pathways Following the release of tracers near the end of the period of winter recharge Photine CU was detected at sites dispersed along the length of the cave (Fig. 3). The drip falling onto the Gib04a stalagmite itself was not monitored but the Photine dye was detected at seepages nearby. The position of the two Photine injection sites almost directly above the cave requires that the detected dye must have followed a nearly vertical pathway through 68 m of rock to reach these. Dye was also detected in seepages up to 60 m horizontally NW of the nearest injection, and 50 m away S –SE. These pathways have overall deviations of 418 and 368 from vertical, and indicate that percolation through a network of interlinked voids spread the tracer laterally between paths that diverge vertically by up to 778. The second dye, DY96, was detected at the flowstone site, which lies almost directly down dip from the injection point (Fig. 3). This dye was found at only a few of the other sites monitored, and the fluorescence at these was much weaker and less prolonged than at the flowstone site.
330
D. P. MATTEY ET AL.
Fig. 4. Temperature and humidity variations from 2004– 2009. (a) Mean annual temperatures (MAT) at the airport (open squares), adjusted by 1.22 8C to be comparable with the cave data as described in the text, compared with historic OSM spot data and MAT from logging at cave sites Gib04a (closed circles) and Hospital (open circles). (b) spot relative humidity measurements at the Gib04a and Hospital sites. (c) Comparison of daily mean temperature measurements measured at seal level by the Met Office (grey) and at the cave entrance (black) with daily average cave air temperatures measured at the Hospital (weak seasonality) and Gib04a sites (flat line). (d) Spot temperature measurements compared with logged temperatures at the Gib04a and Hospital sites. The Hospital site displays much larger variations than Gib04a, which lies deeper in the cave interior. Meteorological data # The Met Office, UK.
However these subsidiary detections were spread along the length of the cave, indicating that percolation of the DY96 took place along a principal channel feeding the flowstone site, but was also spread laterally through a network of subsidiary
channels. The horizontal angle across which this spread occurred was 708, which is similar to the vertical angle of spread displayed by the Photine tracer. The angle between vertical and the pathway to the flowstone site is 308. Taking the patterns of
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
detection of the two dyes together, it seems probable that infiltrating waters follow bedding-parallel fractures for distances up to c. 100 m, but also migrate vertically down intersecting joints and small faults. The properties of the 3-D network of fractures are such that each dye probably spreads through a volume shaped approximately as an upright half-cone with its apex at the point of injection, its axis vertical, and the vertical plane of bisection of the apical angle lying parallel to the strike of the limestone beds. The half-angle at the apex of this cone is about 358. Such strong lateral dispersion of the tracers implies that infiltrating water derived from any single point on the surface must mix extensively with water from other points. However the fact that DY96 was detected strongly at only one site indicates that some parts of the flow network are effectively channeled parallel to the bedding and that these channeled pathways encounter relatively few distributary junctions. Such pathways may be expected to show less mixing, and seepages fed by them may therefore possess distinctive chemistry and/or hydrological characteristics compared with others that are derived from better mixed parts of the network. Such contrasts do in fact occur between the three seepage sites that were monitored as part of this work. Variations in drip discharge rates at three sites measured at 30-minute intervals across three annual hydrological cycles are shown in Figure 5 and show that drip responses to the seasonal pattern of winter rain differ dramatically at each site. The Gib04a speleothem site (Mattey et al. 2008) is fed at a low discharge of 0.04 to 0.06 l/d which shows no direct relationship to the winter–summer cycle of recharge or to individual high-rainfall events. Dye-tracing suggests it is fed by mixed water from a recharge zone that lies upslope and directly above the cave chamber where the roof thickness is 64 m (Fig. 3). The drip rate pattern is consistent with flow through roof rock which has a capacity for significant storage and mixing before emerging at the drip site. The drip monitoring site in the Dark Rift, situated at a higher level in the cave system (Fig. 2) shows a more complex drip response that is highly correlated with periods of heavy rainfall. The enlarged scale in Figure 5 shows that the drip rate has a base flow component of a similar magnitude to that seen at a lower level at the Gib04a site, but this is punctuated by short-lived but intense discharge events which may peak at over 8 l/d for a few hours. At the beginning of the winter season there is a six-hour lag between the start of a rainfall event and the sharp rise in drip rate which subsequently declines usually to a new, higher level of base flow. The sharp response may be termed quick-flow and is suggestive of a channeled
331
pathway. Each winter season begins with a high degree of correlation between rain and quick-flow events but as the wet season develops this relationship becomes less clear, presumably as recharge fills the fracture network and flow switching begins to contribute water from other reservoirs. Neither dye was detected at this site, but Photine CU was found at two drips a few metres away, suggesting that the the main recharge zone for base flow lies directly above the cave chamber where the roof thickness is around 38 m. Analogy with the spread of DY96 tracer suggests that the recharge for the quick-flow component may lie somewhat upslope and feed the same reservoir as the baseflow. The pattern of water chemistry at this site suggests a source that is quite distinct from both Gib04a and the flowstone site. The flowstone site is located at the same level and less than 4 m away from the Gib04a site (Fig. 3). Here the drip response shows remarkable annual regularity with peak discharges around midJanuary exceeding 80 l/d. The discharge rate then steadily declines for the remaining winter season and by the start of the following autumn the site is close to drying up. In winter 2005/6 the flowstone discharge ceased completely until mid-January but during 2006/7 and 2007/8, although flows increased erratically at the start of the winter rainfall season, the peak discharge was delivered in each case at the same time in mid-January. Dye-tracing reveals that this water is fed from a recharge zone near the summit of the Rock, and groundwater follows bedding places down to the flowstone site. The unusual cyclical pattern of delayed peak discharge followed by an exponential decline suggests the operation of a siphon which carries over accumulated early winter rainfall into a second reservoir that directly feeds the flowstone. During heavy periods of rainfall the siphon can be by-passed creating irregular pulses observed before the main discharge event in January. Water reaches this site via a longer flow path (c. 120 m) and the drip water, whose composition is also quite different with higher levels of Ca and total alkalinity (see below), is depositing flowstone on the cave floor.
Cave water hydrochemistry The main features of compositional variation in cave waters are illustrated in Figures 6 and 7. The Mg/Ca compositions of lake and drip water are compared on Figure 6a and show coherent trends consistent with extensive prior calcite precipitation (PCP) from parental waters having Ca values at or above the upper limits of the water analyzed (Fairchild et al. 2000). The rationale for this is that calcites have much lower Mg/Ca than the waters
332
D. P. MATTEY ET AL.
Fig. 5. Comparison of drip discharge rates at the Flowstone, Dark Rift and Gib04a sites compared with daily rainfall and recharge expressed as monthly P-ET. Daily water discharge rates were calculated using data from continuously logged using acoustic drip counters and a drop volume of 0.15 ml (Collister & Mattey 2008). Daily rainfall amounts were measured at the Gibraltar Meteorological Office, 3 km from the entrance to St. Michaels Cave and P-ET calculated using the Thornthwaite method (Thornthwaite 1948). Meteorological data # The Met Office, UK.
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
333
Fig. 6. (a) Plot of Ca versus Mg/Ca ratios for NSM waters compared with modeled prior calcite precipitation (PCP) lines. The distribution coefficient for Mg in calcite (KMg) used for modeling is 0.02, but the lines are not sensitive to the exact value. The pCO2 values correspond to those at equilibrium with calcite for waters of appropriate Mg/Ca using MIX4 modeling (Fairchild et al. 2000). (b) Similar plot for Sr/Ca; value of KSr used is 0.3 (Huang & Fairchild 2001). See text for discussion.
334
Fig. 7.
D. P. MATTEY ET AL.
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
from which they form such that the distribution coefficient (K) is small: KMg ¼ (Mg=CaÞwater =(Mg=Ca)calcite ¼ 0:02 (at 17 8C; Huang & Fairchild 2001)
(1)
Hence as calcite is precipitated along the flowline, Ca changes much more than Mg and drip water compositions evolve following curves on Figure 6 such as the two PCP model lines illustrated. The highest-Ca waters along a trendline represent waters that are least modified and their pCO2 value constrains the minimum value encountered by the waters along their flow route. The highest-Ca points are those that feed the flowstone, which has a recharge zone at a higher altitude and is delivered to the drip site along bedding planes via a siphon. The Ca concentrations reach over 140 mg/l Ca, corresponding to equilibrium at a pCO2 of .1021 ( 7% CO2 by volume), and since the cave air has a much lower pCO2 values, degassing leads to a high degree of supersaturation and vigorous precipitation of calcite as flowstone. Lake water samples collected monthly over the monitoring period are generally more constant in composition relative to drip water in terms of Ca (around 80 mg/l) and with initial Mg/Ca ratios similar to drip water from the Gib04a site (Fig. 6a). The Gib04a drip water shows a large seasonal decrease in Ca to below 40 mg/l accompanied by a shift in Mg/Ca to .1500 as a result of seasonal calcite precipitation during the summer months (Fig. 7). Drip waters from the Dark Rift are offset from the other data samples. Although this could reflect a source with a lower ratio of dolomite to calcite, it is notable that the lowest Mg/Ca ratios are very similar for the Dark Rift waters and the flowstone waters and both may represent the original composition of bedrock being dissolved. These minimum 1000*Mg/Ca values are around 400, corresponding to a molar Mg/Ca ratio of around 1/3 suggesting dissolution of 2/3 dolomite and 1/3 limestone. Note that the proportion of limestone bedrock would be expected to be much lower than this because of faster calcite dissolution (Fairchild et al. 2000). The reason for the lower Ca values of the Dark Rift could be that its water acquires its dissolved CO2 from soil air rather than the more CO2-rich source from which Gib04a is evidently
335
derived. Soil air has lower pCO2 values (no more than 1021.4 ( 4% by volume) according to Fig. 6a) and indeed variability in soil pCO2 is expected from its highly varied thickness and moisture content. Figure 6b illustrates Sr data which also illustrate the PCP effects. These data are more scattered at each site which is probably at least partly analytical because of the very low Sr concentrations (typically 0.1 ppm) encountered. There is a less marked increase in Sr/Ca as Ca falls compared with Figure 6a. This is due to a relatively high value for KSr (a value of 0.3 was used to construct the lines in Figure 6b; Huang & Fairchild 2001). Figure 7 shows monthly variations of cave water pH, total alkalinity, Ca, Mg/Ca, and d13CDIC, in relation to drip discharge and cave air CO2 levels from late 2004 to late 2008/early 2009. An obvious feature of the data is the strong seasonal pattern of variation where drip water collected in winter months has lower pH, highest total alkalinity, highest Ca, lowest Mg/Ca, and lowest d13CDIC. Another striking feature is the coherent pattern of variation displayed by water collected at drip sites which are each significantly different from each other in terms of flow paths (Fig. 3) and discharge patterns (Fig. 4). The flow stone drip water has a distinctive composition having higher Ca and total alkalinity levels yet shows the same seasonal patterns of variation; the lake water reservoir also displays similar, but more attenuated seasonal variations in hydrochemistry. This commonality in seasonal behaviour suggests a common control that must be located within the cave, because it affects waters with such diverse hydrology and chemistry.
Cave air CO2 and ventilation regimes CO2 mixing ratios in cave air sampled monthly from NSM at the GIb04a sites and from the lake area are shown on the lowest graph in Figure 7. As for the temperature and humidity variations discussed above, the distribution of cave air CO2 also shows complex patterns of spatial as well as strong temporal variation which are clearly related to regular diurnal and seasonal cycles modified by local synoptic scale weather conditions. Overall, cave air CO2 mixing ratios show very regular seasonal variations with highest concentrations in winter
Fig. 7. (Continued) Relationships between rainfall amount, drip discharge, dripwater composition and cave air pCO2 measured between 2004– 2009. Grey bands mark the period between mid-November and mid-April when cave air pCO2 is highest. From top to bottom: daily rainfall; drip discharge rates from Figure 5 plotted together on a log scale; monthly dripwater pH, total alkalinity, Ca concentration, 1000*Mg/Ca and d13CDIC. Lower plot is the concentration of CO2 in deep cave air, measured on monthly spot samples at the Gib04a (black circles) and Lake (grey circles) sites along with data (unpublished) obtained by continuous logging at 2 hr intervals measured at the Gib04a site. Meteorological data # The Met Office, UK.
336
D. P. MATTEY ET AL.
(3000–6000 ppmv) falling to around 500 ppmv in the summer. Continuous monitoring of CO2 levels at seven locations has been carried out since October 2006 and shows that CO2 concentrations can exceed 10 000 ppmv for brief periods and are subject to rapid fluctuations which are closely related to wind strength and direction rather than barometric pressure (our unpublished data). These data also clearly show that the seasonal switch from high to low CO2 modes is closely correlated with the temperature difference between the exterior and interior of the cave which seasonally reverses chimney ventilation driven by density contrast of cave air. Thus the high winter levels of CO2 in cave air are diluted in the summer by penetration of outside atmosphere. The switches between winter and summer modes take place in mid-April and mid-November and correspond to the external temperature respectively rising above and falling below the MAT of the deep cave (17.9 8C). The former allows dilution by the atmosphere and the latter permits CO2 levels to rise by flux from the interior of the Rock. This chimney ventilation is clearly revealed by seasonal temperature and humidity variations in OSM, but ventilation dilution of CO2 is observed in even the most distal regions of NSM despite having no
known natural entrances and only linked to the show cave via a small hatchway. Furthermore, the transition from high winter levels to low summer levels is very rapid and occurs throughout the cave system within hours. Sealing of the hatchway between OSM and NSM for a month in August 2007 had no effect on the CO2 behavior in NSM and dilution of cave air must take place via natural pathways to other caves or directly to the surface. These observations, supported by data for CH4 and d13C data for both CO2 (see below) and CH4 (to be presented elsewhere), provide compelling evidence for seasonal advective transport of CO2-rich ‘ground air’ through macroporous Gibraltar limestone via enlarged bedding planes and joints. Whether this process is local to the elevated cave systems located near the summit of the Rock or part of a larger scale advective system affecting cave systems at other levels in the rock is not yet known. The carbon isotopic composition of cave and soil air CO2 is shown on a Keeling plot of 1/CO2 versus d13C in Figure 8. The d13C of cave air CO2 shows wide variation between 210‰ to 223‰ which is a result of mixing between local atmosphere (d13C ¼ 29.6‰) and an isotopically light CO2-rich component, a processes also seen in Obir cave
Fig. 8. Abundance and carbon isotopic composition of CO2 in local atmosphere, cave air (all sites) and soil air. Line is a best fit mixing vector between atmospheric CO2 and a ‘ground air’ CO2 endmember with a d13C of 222‰. See text for discussion.
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
(Spo¨tl et al. 2005). Sources of the CO2-rich ‘ground air’ component may include CO2 degassed from groundwater, CO2 from the soil zone that has penetrated the epikarst as a gas phase, CO2 respired from plant roots that may penetrate deeply into fractures, and CO2 generated from decomposition of colloidal or dissolved organic matter in infiltrating water (Atkinson 1977; Wood & Petraitis 1984; Wood 1985). The tight mixing array on Figure 8 constrains the ground air d13C as 222.0 + 1.5‰ with no evidence of seasonal variation. Soil CO2 levels measured from a site vertically above the cave (Fig. 2) show seasonal variation with lowest values in the dry summer months and maximum values reaching 7300 ppmv in the winter when soil moisture is greatest and bioproductivity is still active. The isotopic composition of soil air is shown on the Keeling plot on Figure 8 and forms a more scattered array that shown by cave air. The soil air CO2 data are also a resulting of mixing, this time between a respired CO2 end member and ambient atmospheric CO2. The d13C of soil respired CO2 is much more variable, ranging from 212 to 222‰ but does not show evidence of regular seasonality. The heavier values are most likely a result of diffusive loss of CO2 to the atmosphere which is greatest during dry periods. Thus the lightest end member values approach the true value of respired soil CO2 which, at 220‰ (Fig. 8), is slightly different (i.e. heavier) than the composition of CO2-rich ground air inferred as the end-member of the mixed air observed in caves. Calculated pCO2 in equilibrium with undegassed cave drip waters (.1021 or 7% by volume, see above) are much higher than measured soil CO2 concentrations and it is speculated that sites of karstic CO2 production may lie below the soil zone in pockets where organic material has accumulated (cf. Atkinson 1977; Wood & Petraitis 1984; Wood 1985). In this environment CO2 generation may take place at a more constant rate throughout the seasonal year to become the main source of ‘ground air’ percolating into the epikarst.
Discussion Shifts in drip water chemistry may be a result of a variety of processes including changes in the water source composition, dilution of groundwater by rain, calcite precipitation elsewhere in the aquifer or changes in the degree of CO2 degassing (Baldini et al. 2006; Banner et al. 2007; Spo¨tl et al. 2005; Tooth & Fairchild 2003). In Gibraltar, the drip water compositions are largely independent of discharge rates and simultaneously have lowest pH (around 7) lowest d13C and highest Ca in winter and lowest Ca with highest pH and d13C DIC in
337
summer. This seasonal variation is synchronized with changes in cave air CO2 levels (Fig. 7) related to reversing chimney ventilation patterns. Covariance among non-conservative parameters such as pH, total alkalinity, d13C, Mg/Ca and cave air pCO2 are consistent with variable degrees of calcite precipitation coupled to CO2 degassing which is externally controlled by the switch from winter ‘high’ to summer ‘low’ levels of cave air pCO2. Modern speleothem carbonates forming in this environment are characterized by annual laminae composed of paired columnar and dark compact calcite bands which preserve well developed cycles in trace elements and stable isotopes (Mattey et al. 2008).
Evolution of calcite and drip water compositions in a seasonally ventilated cave The composition and carbon isotopic evolution of drip water undergoing coupled degassing–calcite precipitation has been modeled as a Rayleigh fractionation process using PHREEQC and code adapted from Appelo & Postma (2007) in Figures 9a, b. The monthly compositions of coexisting cave air and drip water sampled between 2004 and 2009 are plotted on Figure 9a as cave air ppmv CO2 versus [Ca]drip water and on Figure 9b as cave air ppmv CO2 versus the d13C of cave air CO2 and drip water HCO2 3 . Figure 9a also shows the curve for the compositions of solutions in equilibrium with calcite and CO2(g) at 18 8C and curves showing the compositions of supersaturated solutions with calcite saturation indices (log of the ratio of ionic activity product to calcite solubility product) of 0.5, 1.0 and 1.5. Drip water compositions at both the flowstone and Gib04a sites define trends which are a result of variable amounts of coupled degassing–calcite precipitation from groundwater that is initially in equilibrium with water in very CO2-rich environments. On emerging into ventilated cave spaces these waters are then subjected to degassing cycles driven by the winter–summer switches in cave air CO2 levels. The flowstone drip waters carry the highest solute load (Figs 6 & 7) and the elevated Ca levels require that the groundwater originally equilibrated with carbonate in a very CO2-rich environment where pCO2 was in the order of 7% (Fig. 9a). This far exceeds the observed CO2 levels in the soil zone suggesting that the site of CO2 accumulation may be deeper, perhaps within voids along the inclined bedding planes forming the feeder network for the drip (Fig. 3). The Gib04a drip water, fed vertically downwards, equilibrated in an environment with a maximum pCO2 of around 1% which is closer to the actual measured pCO2 levels in the soil zone directly above.
338
D. P. MATTEY ET AL.
Fig. 9. Observed and modeled variations in drip water Ca and coexisting calcite, drip water and cave air d13C as a function of fluctuating cave air pCO2 responding to seasonal reversals in cave ventilation. (a) Calcite solubility curve and curves representing the compositions of solutions with calcite saturation indices (log of the ratio of ionic activity product to calcite solubility product) of 0.5, 1.0 and 1.5 calculated for pure water at 18 8C using PHREEQC (Appelo & Postma 2007). The loci of solution compositions evolving via coupled degassing–calcite precipitation from an initial solution in equilibrium with calcite and air containing 70 000 ppmv CO2 (pCO2 1021) containing 4.5 mmole/l Ca, of are shown for ‘summer’ degassing where final equilibrium would be attained under low cave air levels of 500 ppmv CO2; the four curves represent cases where degassing is 10, 102, 103 (heavy line) and 104 greater than the rate of calcite precipitation. The cumulative amount of calcite precipitated is shown for the 103 curve which mirrors the evolution of dripwater at the flowstone site (triangles). The compositions of dripwater (open circles) for the Gib04a site along with a hypothetical initial solution in equilibrium with calcite and air containing 10 000 ppmv CO2 are shown for
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
The two variables that exert greatest control over the solute and isotopic evolution of carbonate saturated solutions during coupled degassing–calcite precipitation are: (1) the contrast between the initial equilibrium pCO2 and that of the new lower pCO2 environment; and (2) the rate of CO2 degassing relative to calcite precipitation. The contrast between the initial and final pCO2 is well constrained by field measurements but the kinetics and rate of degassing and calcite precipitation may depend on local factors such as the discharge rate, drip size, water film thickness (for degassing) and the nucleation and crystal growth mechanisms during calcite precipitation. Figure 9a shows modeled degassing curves for the flowstone drip water compositions using an initial solution in equilibrium with pCO2 set at 7% containing 4.5 mmol/l Ca. If the pCO2 contrast remains small and CO2 degassing and calcite precipitation keep pace with each other the solution will evolve following the equilibrium calcite solubility curve as pCO2 decreases to a new lower level. For CO2 saturated drip water emerging into a well ventilated space the CO2 degassing rate will be much greater than the calcite precipitation rate rapidly forming calcite supersaturated solutions as shown by the drip water compositions at both the flowstone and Gib04a sites in Figure 9a. The drip water data plotted in Figure 9a represent three complete ventilation cycles and are the end product of degassing under two regimes: winter and summer. To illustrate the effects of different degassing–calcite precipitation rates, four degassing curves are calculated for ‘summer’ degassing where final equilibrium would be attained under low cave air levels of 500ppmv CO2. The curves represent degassing rates 10, 102, 103 and 104 greater than the rate of calcite precipitation and show in all cases that solutions rapidly become strongly supersaturated until delayed calcite precipitation restores equilibrium at the new lower pCO2. As the degassing rate increases the degree of maximum supersaturation also rises and the form of these curves mirror the observed trends in drip water evolution (Fig. 9a). Flowstone drip water compositions suggest that degassing from at this site was around 103 times faster than calcite precipitation, and this is also the case for
339
Gib04a drip water although at lower overall degrees of calcite supersaturation (Fig. 9a, open circles). For degassing under winter conditions the pCO2 contrast is lower and degassing vectors will return to the equilibrium curve more quickly. This is illustrated by the dashed curve on Figure 9a which is the 103 rate curve (solid line) recalculated for winter degassing to a cave air pCO2 of 5000 ppmv and embrace data at relatively low Ca levels sampled under high pCO2 conditions on Figure 9a. Thus the spread of measured drip water compositions fit within a dynamic degassing model where degassing vectors are continuously responding not only to seasonal ventilation but possibly also to the rapid (day-week) synoptic time scale fluctuations revealed by continuous monitoring of cave air pCO2. The evolution of d13C in bicarbonate, cave air and calcite has been calculated as a Rayleigh model using PHREEQ as a complete system that accounts for apportionment of 13C among all coexisting carbon species as a function of changing pH during degassing (Appelo & Postma 2007). Monthly values for d13C of coexisting dissolved inorganic carbon in drip water from the flowstone and Gib04a sites, and for coexisting cave air CO2 (calculated as the end-member composition ‘added’ as ground air, see Fig. 8) can be compared with the modeled isotopic evolution of the composition of least degassed drip water (sampled from within roof straws at 216‰) under the same summer conditions as the solid degassing curve in Figure 9a, where the rate of CO2 degassing is set 103 times faster than calcite precipitation. Starting with an initial solution having a d13C of 216‰ (representing the bulk composition of all carbon components, measured as total DIC) measured and calculated data d13CDIC are closely comparable and d13CDIC values rise as a result of degassing of isotopically light carbon under decreasing cave air pCO2. The modeled bulk isotopic composition of carbonate is initially remains fairly constant at around 212.5‰ and rises to around 210‰ during the final stages of degassing. Although the instantaneous isotopic composition of calcite will track that of dissolved bicarbonate and would rise to isotopically heavy values during the final stages of degassing, the modeled range in
Fig. 9. (Continued) comparison. The dashed curve represents the case where degassing takes place under winter high pCO2 conditions. (b) Monthly values for d13C of dissolved inorganic carbon in drip water from the flowstone (and Gib04a) sites coexisting with end-member cave air CO2 compared with the isotopic evolution of dripwater total DIC, 13 dripwater HCO2 3 degassed CO2 and precipitated calcite. The parent drip water has an initial d CDIC of 216‰ and degasses under the same summer conditions plotted in Figure 9a as the solid curve. The amplitude of calcite d13C annual cycles in Gib04a are also shown (from Fig. 10 and Mattey et al. 2008) which closely match the modeled evolution of calcite d13C responding to degassing. See text for discussion. [Degassing curves are calculated using PHREEQC and code adapted from Appelo & Postma (2007) with calcite precipitation rate constants from Plummer et al. (1978) and relevant isotope fractionation factors from Clark & Fritz (1997).]
340
D. P. MATTEY ET AL.
Fig. 10. Relationship between petrographic, trace element and stable isotope variations for laminae pairs growing over four years from 2000–2004 in Gib04a. (a) Macroscopic appearance of the uppermost portion of Gib04a. Active growth surface on left; width of image 11 mm and adapted from Mattey et al. (2008). (b) Area outlined in (a) prepared from facing slice as a thin section and viewed in plane polarized light. DCC, Dark compact calcite; LCC, light columnar
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
bulk calcite d13C resulting from winter–summer ventilation cycles of 2– 2.5‰ is very similar in magnitude to the d13C cycles preserved in Gib04a (Mattey et al. 2008) and there is no evidence of kinetically enhanced d13C values as found at Obir cave by Spo¨tl et al. (2005). The isotopic composition of degassed CO2 shows a similar pattern, the initial stages of degassing show little change around 223‰, rising only in the final stages of degassing to around 222‰. The calculated d13C trend for degassed CO2 is only slightly lower than the d13C of end-member cave air CO2 but a close match would only be expected if the source of cave-air CO2 was solely from locally degassed dripwater. This cannot be the case as the quantity of drip water percolating into NSM is very low and the behaviour of CO2 in the cave suggests that advective processes introduce ‘ground air’ CO2 from elsewhere.
Paired laminae in Gib04a: relationships between cave microclimate and calcite fabric, stable isotope and trace element cycles Modern calcite deposition in NSM results in paired laminae composed of light columnar calcite (LCC) and dark compact calcite (DCC) which preserved regular cycles in trace elements, d13C and d18O (Mattey et al. 2008). Mattey et al. (2008) assigned each lamina pair to a calendar year by counting back from the time of collection (June 2004) and the age model was confirmed by locating the radiocarbon ‘bomb’ peak in its correct position. The cycles in d13C and d18O reported in Mattey et al. (2008) were defined by samples taken at 100-mm resolution which produced quasi-regular sinusoidal variation in d13C, but variations in d18O and Sr were more complex. The d13C minima were found to be located in the LCC fabric and the switch to DCC fabrics occurs when d13C and d18O are reaching highest values of the annual cycle but the precise position of the LCC-DCC fabric transition relative to stable and trace element cycles, and their timing within the annual cycle were more difficult to locate. The cave air monitoring results shows that the switch between winter high-CO2 and summer low-CO2 and back again is very rapid (Fig. 7) and provide time markers for mid-April and midNovember in the annual cycle of d13C in calcite, which can be used to obtain a better understanding of environmental and kinetic controls on fabric
341
type and trace element uptake during calcite precipitation. The correlation between paired fabrics, calcite Sr abundance, and stable isotope composition across four growth cycles – representing calcite deposited between 2000–2004 – are shown in Figure 10. Figure 10a, taken from Mattey et al. (2008), shows the macroscopic appearance of the uppermost portion of Gib04a in a polished surface where the paired LCC and DCC can be seen. Figures 10b and 10c show the structure of the final four cycles in plane polarized light and as an electron backscattered diffraction grain boundary map which clearly shows the alternations of columnar macroporous LCC calcite with the microporous compact DCC calcite which forms the darker bands visible on cut surfaces. The ESBD image in Figure 8c shows that the boundary surface between the LCC and DCC fabrics is irregular and is marked by a sharper transition from coarse- to fine-grained. The DCC fabric increases in grain size and the transition back to LCC is more diffuse (Fig. 8b). A new stable isotope profile obtained by micromilling at 50-mm resolution was performed adjacent to the high resolution trace element profile by obtained by synchrotron analysis reported in Mattey et al. (2008) to examine the topology of the isotopic transitions between the LCC and DCC fabrics. The stable isotope, trace element and fabric maps can now be correlated with a confidence of +100 mm. Each analysis in the stable isotope profile in Figure 10 nominally represents 2–3 weeks of growth. They show that the cyclical pattern in d13C rises to values that remain fairly constant before falling more sharply to the lower ‘winter’ value. The abrupt switching of pCO2 in the cave atmosphere implies that the annual cycle of d13C in the calcite should be a square wave. The high resolution record in Figure 10 may represent a somewhat rounded version of this, rather than the sinusoidal patterns seen in the lower resolution profile reported in Mattey et al. (2008). Here the rounding is introduced because the boundary surface between the LCC and DCC fabrics is irregular with amplitude of up to 500 mm. Micromill sampling in 50 mm increments along a 2 mm wide face aligned parallel to the layering of the laminae cannot fully resolve a sharp isotopic change corresponding to the jump in cave air pCO2 that occurs in November and April. A Sr profile across the four annual cycles was measured by synchrotron micro-XRF (Mattey
Fig. 10. (Continued) calcite (adapted from Mattey et al. (2008). (c) Grain boundary map of (b) obtained by electron backscatter diffraction. (d) Sr profile measured by synchrotron m-XRF using a beam 1 micron across in 10 micron steps (data from Mattey et al. 2008). (e) d13C and d18O profile measured on samples removed from slab (a) in 50 micron steps by micromilling. Timelines for mid-November (N, grey dashed) and mid-April (A, grey dotted) are shown for reference. The faint dark band is radiation damage cause by synchrotron microbeam analysis. See text for discussion.
342
D. P. MATTEY ET AL.
et al. 2008). The Sr/Ca ratios varies by around a factor of 2 in dripwater during the monitoring period at Gib04a (Fig. 6b) which is comparable with the range of calcite Sr in Figure 10, making it unnecessary to invoke kinetic effects to account for annual Sr variability at this site (cf. Fairchild & Treble 2009). The data in Figure 10 were obtained using a 1 micron beam in 10 micron steps. They are not affected by mixing of different calcite types caused by sampling parallel to irregular boundaries, but suffer from a different problem in that the precise position of the irregular LCC – DCC boundary cannot precisely be mapped along the Sr line traverse. However it can be seen that Sr variations closely follow the pattern of variation in d13C, but Sr values are more constant during winter deposition of LCC and in summer, where d13C is more uniformly high (i.e. a plateau with rounded shoulders caused by micromill sampling), Sr values peak then gradually decline. Time lines corresponding to mid-April and midNovember are shown on Figure 10 and it is apparent that the Sr cycle begins its rise at about the time of the April transition, but reaches a short-lived maximum well before the November transition in July or August. The Sr values then decline through the autumn and the minimum is reached after the November transition. Neither the poor ‘square wave’ resolution of the Micromill nor the uncertainty of the fabric boundaries with respect to Sr can influence these phase relations, which must have a different explanation. One possibility is that the timing reflects the influence of drop interval on degassing and PCP onto the straw stalactite above Gib04a. The discharge log in Figure 5 shows that discharge generally increases for a month or two in mid-summer to autumn, but because the logged years do not cover the four years of speleothem growth shown in Figure 10a precise comparison is not possible. Nevertheless, if discharge regularly increases in late summer, as it seems to, there would be less time between drops for degassing and therefore less PCP, so Sr/ Ca would fall. We suggest timing of the Sr peak represents the onset of increased summer flow at Gib04a; flow is then slowly reduced again resulting in increased Sr/Ca values before the November CO2 rise reduces degassing more completely. After the sharp winter switch in pCO2, PCP slows down and Sr/Ca falls sharply to the winter minimum plateau. Thus the details of the d13C and Sr cycles appear to be subtly decoupled from each other with respect to fabric development. This is consistent with the results of the degassing–calcite precipitation process modeled above in which degassing rates are highly responsive to seasonally variable cave air pCO2 whereas calcite precipitation rates (around 103 slower, see above) drive the PCP enrichment
of Sr at a far slower rate. Banner et al. (2007) also showed that calcite precipitation can temporally cease under high cave air pCO2 conditions and the sharper junction between LCC back to DCC may represent a temporary cessation of growth, also marked by steps in Sr concentration in the calcite. Clearly, even at this resolution and high level of confidence in correlating fabrics and chemical profiles, the fabric –isotope–trace element relations still remain rather ambiguous. However the jumps in d13C clearly mark the suppression and restoration of degassing as cave air pCO2 levels rise and fall sharply in mid-November and mid-April accompanied with synchronous changes in d18O. Because no clear evidence exists of kinetic enhancement of d13C during seasonal cycles several other processes may control the irregular annual d18O excursions to heavier values seen in Figure 10, including changes in the composition of the drip water, and will be discussed further elsewhere. The minimum d18O values, as discussed in Mattey et al. (2008), represent the dripwater compositions recorded when the cave environment is most conducive to equilibrium precipitation (i.e. lowest d13C and highest cave air pCO2) and these values correlate well with the weighted mean d18O of winter precipitation over a 54-year period (Mattey et al. 2008).
Conclusions Detailed monitoring of three drip sites in NSM reveals a strongly coherent seasonal pattern of dripwater compositions despite each site having significantly different flow paths and discharge patterns. Calcite saturation is closely linked to regular seasonal variations in cave air pCO2 which is highest between November–April. The seasonal switch to low pCO2 in the summer is caused by chimney ventilation linked to temperature differences between the exterior and interior of the cave. Advection of isotopically homogenous CO2-rich ground air derived from deeper levels in the Rock maintains high cave air pCO2 levels resulting in winter speleothem deposition of columnar calcite having the lowest d13C and d18O values. Flushing by the outside atmosphere lowers cave air pCO2 levels in summer leading to higher degrees of dripwater degassing rates precipitation a dark compact calcite having elevated d13C values. A coupled CO2 degassing–calcite precipitation model links the development of annual cycles in d13C and dripwater evolution to switching pCO2 driven by seasonal cave ventilation. The model shows that drip water supersaturation is consistent with degassing rates 104 greater than the rate of calcite precipitation and also accounts for the
SEASONAL MICROCLIMATE CONTROLS ON SPELEOTHEM GROWTH
observed seasonal d13C variations in coexisting dripwater, cave air CO2 and speleothem calcite. The relationships between stable isotope, Sr and speleothem fabrics across calcite laminae deposited between 2000 and 2004 have been examined at the highest possible resolution using micromilling, synchrotron m-XRF and electron backscatter diffraction techniques and shows that d13C and Sr are subtly decoupled from each other with respect to fabric development. The Sr- d13C topology is consistent with a degassing–calcite precipitation process where the d13C compositions of drip water and calcite are highly responsive to seasonally switching cave air pCO2 whereas calcite precipitation rates (up to 104 slower, see above) drive the PCP enrichment of Sr at a slower and more complex manner. High resolution speleothem fabric, trace element and isotope records from caves that have been closely monitored provide crucial links between the local climate and the way that climate is recorded during speleothem deposition and remove some of the assumptions sometimes necessary in climate reconstruction from speleothem records. In Gibraltar the low cave air pCO2 in summer is unusual compared to cave monitoring studies carried out elsewhere and shows that caution is needed when linking paired speleothem fabrics to specific seasons without knowledge of local processes operating in the cave. Furthermore, monitoring shows that the climate recording process may vary among sites in the same cave resulting in differing bias to winter and summer growth depending on the interplay between ventilation and hydrological cycles. We are indebted to John Balestrino, Joanne McCarthy and Hela de la Paz for help with the cave monitoring program, to John Cortes (GONHS) for logistic support, and Tito Vallejo and the cave guides for their co-operation. We also thank Jacqui Duffet, Dave Lowry and Rebecca Fisher for assistance in the laboratory and field during the project. The manuscript was greatly improved with the help of comments made by two anonymous reviewers. This work was supported by the NERC under grant NE/D005280.
References A PPELO , C. A. J. & P OSTMA , D. 2007. Geochemistry, groundwater and pollution. Balkema Publishers, Leiden, The Netherlands. A TKINSON , T. C. 1977. Carbon dioxide in the atmosphere of the unsaturated zone: an important control of groundwater hardness in limestones. Journal of Hydrology, 35, 111–123. A TKINSON , T. C., S MART , P. L. & W IGLEY , T. M. L. 1983. Climate and natural radon levels in Castleguard Cave, Columbia Icefields, Alberta, Canada. Arctic and Alpine Research, 15, 487–502.
343
B AKER , A. & B RUNSDON , C. 2003. Non-linearities in drip water hydrology: an example from Stump Cross Caverns, Yorkshire. Journal of Hydrology, 277, 151– 163. B AKER , A., S MITH , C. L., J EX , C., F AIRCHILD , I. J., G ENTY , D. & F ULLER , L. 2008. Annually laminated speleothems: a review. International Journal of Speleology, 37, 193–206. B ALDINI , J. U. L., M C D ERMOTT , F., B AKER , A., B ALDINI , L. M., M ATTEY , D. P. & R AILSBACK , L. B. 2005. Biomass effects on stalagmite growth and isotope ratios: a 20th century analogue from Wiltshire, England. Earth and Planetary Science Letters, 240, 486–494. B ALDINI , J. U. L., M C D ERMOTT , F. & F AIRCHILD , I. J. 2006. Spatial variability in cave drip water hydrochemistry: implications for stalagmite paleoclimate records. Chemical Geology, 235, 290– 304. B ALDINI , J. U. L., M C D ERMOTT , F., H OFFMANN , D. L., R ICHARDS , D. A. & C LIPSON , N. 2008. Very highfrequency and seasonal cave atmosphere PCO2 variability: Implications for stalagmite growth and oxygen isotope-based paleoclimate records. Earth and Planetary Science Letters, 272, 118– 129. B ANNER , J. L., G UILFOYLE , A., J AMES , E. W., S TERN , L. A. & M USGROVE , M. 2007. Seasonal Variations in Modern Speleothem Calcite Growth in Central Texas, U.S.A. Journal of Sedimentary Research, 77, 615–622. B AR -M ATTHEWS , M., A YALON , A., M ATTHEWS , A., S ASS , E. & H ALICZ , L. 1996. Carbon and oxygen isotope study of the active water– carbonate system in a karstic Mediterranean cave: implications for palaeoclimate research in semiarid regions. Geochimica et Cosmochimica Acta, 60, 337– 347. B ORSATO , A., F RISIA , S., F AIRCHILD , I. J., S OMOGIY , A. & S USINI , J. 2007. Trace element distribution patterns in annual stalagmite laminae mapped by micrometerresolution X-ray fluorescence: incorporation of colloidally-transported species. Geochimica et Cosmochimica Acta, 71, 1494–1512. B OTTRELL , S. H. & A TKINSON , T. C. 1992. Tracer study of flow and storage in the unsaturated zone of a karstic limestone aquifer. In: H O¨ TZL , H. & W ERNER , A. (eds) Tracer Hydrology. Balkema, Rotterdam, 207– 211. C LARK , I. D. & F RITZ , P. 1997. Environmental Isotopes in Hydrogeology. CRC Lewis Publishers, New York. C OLLISTER , C. J. & M ATTEY , D. P. 2008. Controls on water drop volume at speleothem drip sites: an experimental study. Journal of Hydrology, 358, 259–267. C RUZ , F. W., K ARMANN , I. ET AL . 2005. Stable isotope study of cave percolation waters in subtropical Brazil: Implications for paleoclimate inferences from speleothems. Chemical Geology, 220, 245 –252. E K , C. & G EWELT , M. 1985. Carbon dioxide in cave atmospheres. New results in Belgium and comparison with some other countries. Earth Surface Processes and Landforms, 10, 173–187. F AIRCHILD , I. J. & T REBLE , P. C. 2009. Trace elements in speleothems as recorders of environmental change. Quaternary Science Reviews, 28, 449–468. F AIRCHILD , I. J., B ORSATO , A. ET AL . 2000. Controls on trace element (Sr– Mg) compositions of carbonate cave waters: implications for speleothem climatic records. Chemical Geology, 166, 255– 269.
344
D. P. MATTEY ET AL.
F AIRCHILD , I. J., S MITH , C. L. ET AL . 2006a. Modification and preservation of environmental signals in speleothems. Earth-Science Reviews, 75, 105 –153. F RAPPIER , A., S AHAGIAN , D., G ONZALEZ , L. A. & C ARPENTER , S. J. 2002. El Nin˜o events recorded by stalagmite carbon isotopes. Science, 298, 565. G ENTY , D. & Q UINIF , Y. 1996. Annually laminated sequences in the internal structure of some Belgian stalagmites – importance for paleoclimatology. Journal of Sedimentary Research, 66, 275– 288. G ENTY , D. & D EFLANDRE , G. 1998. Drip flow variations under a stalactite of the Pe`re Noe¨l cave (Belgium). Evidence of seasonal variations and air pressure constraints. Journal of Hydrology, 211, 208–232. G ENTY , D., B AKER , A., M ASSAULT , M., P ROCTOR , C., G ILMOUR , M., P ONS , E. & H AMELIN , B. 2001. Stalagmite dead carbon proportion variation: paleodissolution process and soil organic matter dynamics recorder - implications for 13C variations in stalagmites. Geochimica et Cosmochimica Acta, 65, 3443–3457. G ENTY , D. 2008. Palaeoclimate research in Villars Cave (Dordogne, SW-France). International Journal of Speoleology, 37(3), 173–191. H ENDY , C. H. 1971. The isotopic geochemistry of speleothems - I. The calculation of the effects of different modes of formation on the isotopic composition of speleothems and their applicability as palaeoclimatic indicators. Geochimica Cosmochimica Acta, 35, 801–824. H UANG , Y. & F AIRCHILD , I. J. 2001. Partitioning of Sr2þ and Mg2þ into calcite under karst-analogue experimental conditions. Geochimica et Cosmochimica Acta, 65, 47– 62. J OHNSON , K. R., H U , C., B ELSHAW , N. S. & H ENDERSON , G. M. 2006. Seasonal trace-element and stable isotope variations in a Chinese speleothem: the potential for high resolution paleomonsoon reconstruction. Earth and Planetary Science Letters, 244, 394– 407. M ATTEY , D., L OWRY , D., D UFFET , J., F ISHER , R., H ODGE , E. & F RISIA , S. 2008. A 53 year seasonally resolved oxygen and carbon isotope record from a modern Gibraltar speleothem: reconstructed drip water and relationship to local precipitation. Earth and Planetary Science Letters, 269, 80– 95. M C D ERMOTT , F. 2004. Palaeo-climate reconstruction from stable isotope variations in speleothems: a review. Quaternary Science Reviews, 23, 901–918. M ICKLER , P. J., B ANNER , J. L., S TERN , L., A SMEROM , Y., E DWARDS , R. L. & I TO , E. 2004. Stable isotope variations in modern tropical speleothems: evaluating equilibrium vs. kinetic effects. Geochimica et Cosmochimica Acta, 68, 4381–4393. M ICKLER , P. J., S TERN , L. A. & B ANNER , J. L. 2006. Large kinetic isotope effects in modern speleothems. Geological Society of America Bulletin, 118, 65– 81. M YLROIE , J. E. & C AREW , J. L. 1988. Solution conduits as indicators of late Quaternary sea level position. Quaternary Science Reviews, 7, 55– 64. P LUMMER , L. N., W IGLEY , T. M. L. & P ARKHURST , D. L. 1978. The kinetics of calcite dissolution in CO2-water systems at 58 to 608C and 0.0 to 1.0 atm CO2. American Journal of Science, 278, 179–216.
R ODRIGUEZ -V IDAL , J., C A´ CERES , L. M., F INLAYSON , J. C., G RACIA , F. J. & M ARTI´ INEZ -A GUIRRE , A. 2004. Neotectonics and shoreline history of the Rock of Gibraltar, southern Iberia. Quaternary Science Reviews, 23, 2017– 2029. R OMANOV , D. & D REYBRODT , W. 2006. Evolution of porosity in the saltwater–freshwater mixing zone of coastal carbonate aquifers: an alternative modelling approach. Journal of Hydrology, 329, 661– 673. R OSE , E. P. F. & R OSENBAUM , M. S. 1991. A field guide to the geology of Gibraltar. The Gibraltar Museum, Bomb House Lane, Gibraltar. R OSENBAUM , M. S. & R OSE , E. P. F. 1991. Map: The Geology of Gibraltar. The Gibraltar Museum, Bomb House Lane, Gibraltar. S HAW , T. R. 1954. New St. Michaels Cave. Cave Science, 3, No. 22. S HAW , T. R. 1955. Old St. Michaels Cave. Cave Science, 3, No. 24. S MART , P. L. 1976. The use of optical brighteners for water tracing. Transactions British Cave Research Association, 3(2), 62– 76. S PO¨ TL , C., F AIRCHILD , I. J. & T OOTH , A. F. 2005. Cave air control on dripwater geochemistry, Obir caves (Austria): implications for speleothem deposition in dynamically ventilated caves. Geochimica et Cosmochimica Acta, 69, 2451–2468. T AN , M., B AKER , A., G ENTY , D., S MITH , C., E SPER , J. & C AI , B. 2006. Applications of stalagmite laminae to paleoclimate reconstructions: comparison with dendrochronology/climatology. Quaternary Science Reviews, 25, 2103– 2117. T HORNTHWAITE , C. W. 1948. An approach toward a rational classification of climate. Geographic Review, 38, 55–94. T OOTH , A. F. & F AIRCHILD , I. J. 2003. Soil and karst aquifer hydrological controls on the geochemical evolution of speleothem-forming drip waters, Crag Cave, southwest Ireland. Journal of Hydrology, 273, 51–68. T RATMAN , E. K. 1971. The formation of the Gibraltar Caves. Trans Cave research Group of Great Britain, 13, 135– 143. T REBLE , P., S HELLEY , J. M. G. & C HAPPELL , J. 2003. Comparison of high resolution sub-annual records of trace elements in a modern (1911–1992) speleothem with instrumental cliate data from southwest Australia. Earth and Planetary Science Letters, 216, 141–153. W ANG , Y., C HENG , H. ET AL . 2008. Millennial- and orbital-scale changes in the East Asian monsoon over the past 224,000 years. Nature, 451, 1090–1093. W IGLEY , T. M. L. & B ROWN , M. C. 1976. The Physics of Caves. Ch. 9, p. 329–358. In: F ORD , T. D. & C ULLINGFORD , C. H. D. (eds) The Science of Speleology. Academic Press, London. W OOD , W. W. & P ETRAITIS , M. J. 1984. Origin and distribution of carbon dioxide in the unsaturated zone of the southern High Plains. Water Resources Research, 20, 1193–1208. W OOD , W. W. 1985. Origin of caves and other solution openings in the unsaturated (vadose) zone of carbonate rocks: A model for CO2 generation. Geology, 13, 822–824.
Travertine terracing: patterns and mechanisms ØYVIND HAMMER*, DAG K. DYSTHE & BJØRN JAMTVEIT Physics of Geological Processes, University of Oslo, PO Box 1048 Blindern, 0316 Oslo, Norway *Corresponding author (e-mail:
[email protected]) Abstract: Travertine terracing is one of the most eye-catching phenomena in limestone caves and around hydrothermal springs, but remains fairly poorly understood. The interactions between water chemistry, precipitation kinetics, topography, hydrodynamics, carbon dioxide degassing, biology, erosion and sedimentation constitute a complex, dynamic pattern formation process. The processes can be described and modeled at a range of abstraction levels. At the detailed level concerning the physical and chemical mechanisms responsible for precipitation localization at rims, a single explanation is probably insufficient. Instead, a multitude of effects are likely to contribute, of varying importance depending on scale, flux and other parameters.
Travertine terracing is undoubtedly among the most spectacular geological phenomena on Earth. Ranging in vertical scale from millimetres to tens of metres, travertine terraces form intricate, delicate patterns as well as imposing waterfalls. Such terraces are common not only in limestone caves and around hot springs, but also in streams and rivers in limestone terrain (Pentecost 2005, 59– 66). In addition, analogous patterns are found in completely different systems, such as silica sinter deposits and water ice. Some of the more spectacular examples of travertine terracing are found in Yellowstone National Park (Bargar 1978; Fouke et al. 2000), Pamukkale in Turkey (Altunel & Hancock 1993), Huanglong Scenic District in China (Lu et al. 2000) and northern Spitsbergen, Norway (Hammer et al. 2005). A list of travertine occurrences in Europe and Asia Minor with notes on terracing was given by Pentecost (1995). Travertine terraces formed an inspiration for Renaissance ornamental water cascades (Berger 1974). However, in spite of their great interest, both aesthetically and scientifically, the formation of travertine terraces has until recently received only scattered scientific attention, especially on the theoretical side (Wooding 1991 provides one notable exception). When trying to understand this pattern formation system, a series of fundamental questions arise. What processes are responsible for the enhanced precipitation at the terrace rim? On a higher level of abstraction, how do these local processes lead to global self-organization? How do the terraces evolve in time, and what are their statistical properties under different conditions? Answering such questions requires a cross-disciplinary approach, involving field work, laboratory experiments, computer modeling and mathematical theory in order to study the complicated feedbacks between hydrodynamics, water chemistry, calcium carbonate
precipitation and possibly particle transport and biology.
Morphology and hydrodynamics The terminology of travertine terrace morphology is confusing. Speleologists often use the roughly equivalent terms ‘rimstone’ and ‘gours’, with ‘microgours’ referring to cm-scale terraces. For open-air travertine systems, morphological classification schemes have been proposed by several authors. Pentecost & Viles (1994) define ‘barrages’ as terraces that are filled with water, forming pools and lakes. ‘Cascades’ are smoother forms on steep slopes, often with smaller terrace-like structures superimposed on them (Fig. 1). Pentecost (2005) uses ‘dams’ instead of barrages. Fouke et al. (2000) and Bargar (1978) use three size categories of barrages, namely ‘terraces’ with areas of tens of square metres, ‘terracettes’ of a few square metres and ‘microterracettes’ of a few square centimetres or less (Fig. 2). Microterracettes are therefore analogous to microgours. Pentecost (2005) suggests the term ‘minidam’ for a dam with an interdam distance (IDD) ranging from 1 cm to 1 m. In this paper we will use the classification of Fouke et al. (2000), extending it to the analogous speleothems. In addition, we recognize that on steep slopes, microterracettes will not form pools (and are therefore not barrages) but grade into subdued ‘microridges’ normal to flow. We also use ‘terrace’ as a general term regardless of size. We use ‘rim’ for the top of the outer wall of the terrace, and ‘pool’ for the water body behind it. The rim is usually convex outwards (downstream), or consists of a series of near-parabolic lobes pointing in the downstream direction (Fig. 2). Inside pools, flow velocity is small due to the larger
From: PEDLEY , H. M. & ROGERSON , M. (eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society, London, Special Publications, 336, 345–355. DOI: 10.1144/SP336.18 0305-8719/10/$15.00 # The Geological Society of London 2010.
346
Ø. HAMMER ET AL.
Fig. 1. Travertine terracettes (left) and cascades (right). Mammoth Hot Springs, Yellowstone National Park, USA.
Fig. 2. Terraces, terracettes and microterracettes at Mammoth Hot Springs, Yellowstone National Park, USA.
TRAVERTINE TERRACING
depth, and the travertine morphologies are often botryoidal or ‘stromatolitic’ with porous textures, reflecting a diffusion-dominated regime often with substantial biological activity. These ‘shrubs’ are often assumed to have a bacterial origin (e.g. Chafetz & Folk 1984) but can probably form from purely inorganic processes (Jettestuen et al. 2006). Over the rim and on the steep outside wall of the terrace, water flows in a thin sheet and flow velocity increases. Here, the travertine is usually more compact. A basic observation is that travertine terraces seem to display scaling properties, with microterracettes often having similar morphologies as the largest terraces. As a working hypothesis we might postulate that a similar mechanism is responsible for pattern formation at all scales, and that a continuous coarsening process could lead from smaller to larger terraces. However, at the smallest scales surface tension exerts an important influence on the hydrodynamics, producing ‘bags’ of water overhanging the rim and a meniscus at the base of the exposed drop wall (Fig. 3). In most localities it can be observed that the height of terracettes
347
and microterracettes is fairly constant, on the order of 4 –5 mm, regardless of slope, thus producing pools of much larger area in regions of small slope (Fig. 4). In fact, given a constant height h and a minimum inter-dam distance k, we can geometrically predict the IDD as a function of slope angle a: IDD ¼
h þk tan a
Data given by Pentecost (2005) seem to fit this simple model well (Fig. 5). Conversely, if the underlying slope is constant, the IDD will also be nearly constant. This effect may contribute to the possible characteristic wavelength which will be discussed below. The ‘unit-step phenomenon’ can be explained by the fact that the surface-tension meniscus lapping onto the rear wall of the pool can only reach to a certain constant height, allowing free flow from the upstream rim over the wetted terrace wall. Given the temperature control on surface tension, terracette height might possibly vary with temperature. At larger scales, surface
Fig. 3. Due to surface tension, ‘bags’ of water hang over the rims of these microterracettes at the Jotun hydrothermal spring, northern Spitsbergen. Message board peg for scale.
348
Ø. HAMMER ET AL.
Fig. 4. Dried-up microterracettes at the Troll hydrothermal springs, Spitsbergen. Lens cap for scale. The almost constant height of terracettes implies larger pool areas in regions of small slope. (a) Oblique view. (b) Horizontal view from the same position.
tension loses importance and unit steps are not observed. Travertine terraces owe some of their beauty to the almost perfectly horizontal orientation of the rims. Any artificially added protuberance will reduce water flow and precipitation locally, allowing the surrounding rim to catch up, while any incised notch will increase flow locally but divert flow away from the rest of the rim, causing relatively faster precipitation in the notch. Clearly, a flat rim will have a tendency to ‘self-repair’, restoring to a stable horizontal line after perturbation. In fact, water flow over the rim is rarely uniform, but localized to a number of narrow sites (Fig. 6). Clearly, these sites are continuously repositioned, producing a statistically uniform growth rate along the rim over time.
Terrace dynamics The emergence and dynamics of travertine terraces can be studied with a number of techniques. Cross sections of travertines allow direct observation of relative growth rates and depositional sequences. The travertine quarries of Rapolano Terme in Italy
offer spectacular exposures (Guo & Riding 1998; Hammer et al. 2007), demonstrating general increase in scale and steepness with time (upwards coarsening), higher precipitation rates at rims and terrace walls, and upstream or usually downstream migration (Fig. 7). Several authors have mapped travertine precipitation rates in natural terraced systems (Liu et al. 1995; Lu et al. 2000; Bono et al. 2001; Hammer et al. 2005). These studies demonstrate substantially higher precipitation rates in areas of high flow velocity near terrace rims and walls, causing relative upwards and outwards growth of the rims. The most spectacular method for the study of travertine terrace dynamics is time lapse photography. Veysey & Goldenfeld (2008) produced a movie based on a year-long data series from the Mammoth Hot Spring complex in Yellowstone National Park, showing progressive coarsening by pond inundation where the rim of a pool grows faster than the rim of the upstream pool, causing drowning of the upstream pool and the formation of a single large pool. Their movie also shows downstream migration of terraces. The formation and expansion of downstreampointing lobes is seen both in time lapse movies and in simulations. Hammer et al. (2007) attempted to explain this ‘fingering instability’ by observing that the regional, underlying terrain slope implies a higher and steeper terrace wall at the downslope tip of the lobe than at the sides. As discussed below, this leads to faster precipitation rates at the tip of the lobe and therefore differential downslope migration rates of the rim, causing a downslope stretching of the pool. It is commonly observed that terrace systems are partly dry (Fig. 6a). This is not always due primarily to reduction in overall flux. As a result of travertine build-up, water continuously finds new routes. Old pathways may be abandoned and dry up, but become active again at a later date (Chafetz & Folk 1984).
Pattern formation mechanisms Travertine terracing is a self-organizing pattern formation process involving a number of coupled physical processes. Several authors have recently modeled the process at different levels of abstraction, highlighting different aspects of the problem. At the purely geometric level, it is clear that a simple relationship between slope steepness and growth rate is sufficient to produce a terraced pattern. Starting from a rough slope with small random perturbations, Jettestuen et al. (2006) used such a growth rule to computer simulate the emergence and coarsening of steps. Interestingly, the
TRAVERTINE TERRACING
349
Fig. 5. Dots, Slopes and inter-dam distances (IDD) measured in minidams by Pentecost (2005). Line, Theoretical IDD with constant terrace height h ¼ 4 mm, minimal IDD k ¼ 6 mm (see text).
resulting morphology was reminiscent of the scalloped microterracettes seen on steep travertine surfaces with thin film flow, where flow velocity is fairly well predicted by slope alone. For a film thickness h, gravity g, slope a, kinematic viscosity n and an empirical constant K, we have u ¼
8gh2 sin a Kn
(Horton et al. 1934). However, in order to understand the mechanism for the slope-precipitation relationship, a more detailed model is required. Hammer et al. (2007) used a computer simulation of shallow water flow over a surface (computational fluid dynamics), coupled with a simple precipitation model where growth rate was proportional to flow velocity. This produced a proximal mound and distal apron covered with terracettes, coarsening
Fig. 6. Visible-light (a) and infrared (b) images of the same area in Mammoth Hot Springs, Yellowstone National Park, USA. Hot water flow from left to right. Lower right area (bright, Fig. 6a) is dry. Plumes of hot water (arrows, Fig. 6b) in the largest pool demonstrate localized flow over the upstream rim.
350
Ø. HAMMER ET AL.
Fig. 7. Cross section of travertine at Rapolano Terme, Italy. Flow was from right to left. Note the initiation and upwards coarsening of rim A, leading to the upstream drowning and termination of rim B.
with time. Rims folded into lobes and migrated downstream with differential rates. In such a model, high flow rates are found not only on steep slopes, but also in shallow regions over rims where velocity must increase to maintain the flux. This provides a positive feedback mechanism between hydrodynamics and precipitation on the rim, leading to localization of precipitation and the stabilization and growth of rims. Goldenfeld et al. (2006) and Veysey & Goldenfeld (2008) developed a much more detailed cellular model, including simplified rules for water flow, surface tension, water chemistry and outgassing of CO2. Their precipitation rule included idealized terms incorporating oversaturation level, flux normal to the surface, and flow velocity. The simulated travertine terraces were natural-looking and reproduced statistical properties observed in the field (see below). Considering the work of Hammer et al. (2007) it is likely that the flow rate term in the precipitation rule is the most important
cause of terrace formation in the simulations of Veysey & Goldenfeld (2008).
Mechanisms for precipitation localization The models described above assumed a relationship between flow rate and precipitation, but did not address in detail the mechanism for such a relationship. In addition, it is conceivable that the relationship is not causal, but that both quantities reflect other, underlying causes. An extraordinary variety of mechanisms have been proposed to explain the higher precipitation rates at terrace rims. A popular explanation has been increased outgassing of CO2 at rims due to agitation and shallowing, the latter increasing the local surface to volume ratio (Varnedoe 1965; Chen et al. 2004). Considering the travertine system as a whole, under normal (e.g. not hyperalkaline) conditions calcite precipitation depends on loss of CO2 from aqueous solution to the atmosphere (e.g. Dreybrodt et al. 1992). An
TRAVERTINE TERRACING
extremely simplified, cartoon-like equation, summarizing many partial reactions that are variously important at different pH, can be written as Ca2þ þ 2HCO 3 O CO2 þ CaCO3 þ H2 O In the classical model for calcite precipitation and dissolution in karst settings by Buhmann & Dreybrodt (1985), it is therefore assumed that precipitation/dissolution rate is stoichiometrically equal to outgassing/ingassing. However, this holds true only under equilibrium conditions and at large scales. In the presence of non-homogeneous advection and diffusion, local, small-scale variations in outgassing at the water-atmosphere interface are not echoed by corresponding local variations in precipitation rate at the bottom, at least in deep water relative to flow rate (Hammer et al. 2008). One promising application of the outgassing hypothesis is the case of thin, laminar film flow, forming microterracettes on steep surfaces and ridges on stalactites. Ogawa & Furukawa (2002) studied the problem of ridge formation on icicles, which are analogous to stalactites but with precipitation being dependent on heat loss rather than CO2 loss. They explained the characteristic ridge wavelength as a result of two competing processes. Thermal diffusion in air involves steeper temperature gradients and therefore faster heat transport around protrusions, encouraging the formation of structure at small wavelengths (so-called Laplace instability). However, thermal transport in the flowing water film makes the temperature distribution more uniform, suppressing small wavelengths. Ueno (2003) carried out a similar analysis, but emphasized the role of gravitational and surface tension forces on the water film to explain the suppression of short wavelengths. Several authors have invoked the Bernoulli effect as a mechanism for increased outgassing under rapid flow, leading to faster precipitation (Chen et al. 2004; Veysey & Goldenfeld 2008). This effect refers to the lower fluid pressure associated with higher flow velocities. Hammer et al. (2008) showed that even at a very high flow rate of 1 m/s, this effect would lead to only about 0.5% pressure drop at the water-air interface. The corresponding 0.5% decrease in dissolved gas concentration under equilibrium conditions, according to Henry’s law, is unlikely to have a major effect on precipitation rates. Mixing in the water column will strongly increase precipitation rates by efficiently bringing solutes to and from the calcite surface. In the case of turbulent flow, we can assume almost complete mixing in the turbulent core away from the travertine surface. Precipitation rates will then be limited by diffusion of solutes through the laminar
351
boundary layer. Since the thickness of the boundary layer decreases with flow rate, this provides a mechanism for a causal link between flow rate and precipitation rate (Buhmann & Dreybrodt 1985). This idea has been confirmed experimentally (Liu & Dreybrodt 1997; Dreybrodt et al. 1997) and by comparison with field measurements (Dreybrodt et al. 1992; Liu et al. 1995). Wooding (1991) presents a particularly interesting analysis of the growth of individual travertine and ice terraces, using a similar conceptual model. However, this model primarily applies to turbulent flow, and can not explain pattern formation at the smallest scales under shallow, slow laminar flow. Hammer et al. (2008) developed a detailed, mechanistic model of carbonate precipitation on an imposed obstruction in shallow 2D laminar flow, with the aim of understanding the localization of precipitation on terrace rims (Fig. 8). This model included hydrodynamics, diffusion of solutes, solute carbonate chemistry, precipitation kinetics and outgassing, and compared well with laboratory experiments. A Laplace instability (Ogawa & Furukawa 2002) caused enhanced precipitation in regions of high convex curvature. In addition, precipitation was high in the shallow, high-velocity region on top of the obstruction, but also in a position downstream from the obstruction. Dramatic experimental increase of outgassing in local positions in the model had no effect on precipitation patterns. Hammer et al. (2008) concluded that advection is of central importance, by bringing ions to and away from the calcite surface. This was also implied in the model of Goldenfeld et al. (2006), which includes a term in the precipitation model for flow rate normal to the surface. In addition, solute concentration gradients set up in a pool are geometrically compressed when advected through the shallow region over the rim. This results in steeper gradients and therefore faster vertical diffusion, leading to enhanced precipitation. Additional possible mechanisms include ballistic deposition of carbonate particles onto the rim from suspension (cf. Eddy et al. 1999) and biological effects (Chafetz & Folk 1984).
Statistical properties The quantitative morphology (morphometrics) and size distribution of travertine terraces as a function of parametres such as slope, flux and water chemistry have not yet been studied in detail. It is possible that shape or size descriptors could be useful proxies for the reconstruction of, for example, paleoflux. However, statistical properties of terraces in individual settings have recently been subject to investigation.
352
Ø. HAMMER ET AL.
Fig. 8. Computer simulation of calcite precipitation under shallow water flow over a protuberance (flow from left to right). Top: Map of CO2 concentration. The upstream concentration gradient is compressed over the obstruction. Downstream there is thorough mixing until a gradient slowly reappears at the far right as CO2 is released by precipitation. Bottom: Precipitation rate, showing enhanced precipitation on the top of the obstruction and also downstream. Small peaks are due to the computational grid, causing small corners of the obstruction to stick out from the boundary layer. The model includes hydrodynamics, advection, diffusion, degassing, solute carbonate chemistry and precipitation kinetics. Details of simulation in a somewhat different geometry are given by Hammer et al. (2008).
Pentecost (2005) discusses terrace wavelength in the downstream direction and the relationships between slope, discharge, terrace wavelength, depth and height. In general, pools are shorter on steep slopes, but height can be larger. Inter-dam distance (IDD) seems to increase with larger discharge. The IDD sometimes displays a characteristic wavelength. In the study by Viles & Pentecost (1999) IDD was found to be random, but in this case terraces were possibly initiated by large woody debris. For microterracettes, the ratio between IDD and depth is higher on gentler slopes. Hammer et al. (2007) studied terracette topography in downstream cross sections, both using their simulation results and field observations in Rapolano Terme, Italy. In spite of the problems of such cross sections cutting the pools at random transversal positions, and not always at their widest points, a regular spacing (characteristic wavelength) was observed. Microterracette pool width distributions were studied in great detail by Veysey & Goldenfeld (2008), again using both simulation and field data. They defined their pool width as the largest width
in a direction normal to the maximum chord. Comparing with a statistical null model of Brownianmotion (random walk) rim shape, they found good accordance except for small terrace widths. Although their distribution is unimodal, it has large variance, and is somewhat difficult to reconcile with the more regular spacing observed at larger scales by Hammer et al. (2007). Veysey & Goldenfeld (2008) interpreted their results as indicating no interaction between rims for large terrace widths, but an attraction effect for small widths. Veysey & Goldenfeld (2008) also present an interesting analysis of pool areas. Using a simple null model of initially small-sized pools merging randomly, they show that the expected steady-state pool area distribution is inverse-square, in accordance with their field observations.
General aspects of travertine terrace pattern formation Pattern formation in the travertine terrace system results from the interaction between two opposing
TRAVERTINE TERRACING
353
Fig. 9. Microterracettes in Pamukkale, Turkey, seemingly displaying a characteristic spacing.
processes: enhanced growth at the rim constituting local positive feedback, and long-range negative feedback involving both reduced growth in the pools and upstream inundation. In a very general sense, travertine terrace patterning can therefore be included in the local self-activation/lateral inhibition class of pattern formation systems (Gierer & Meinhardt 1972, 2000). Such systems typically produce patterns of regularly spaced points, or parallel or labyrinthic stripes, as observed in many biological and geological settings. In the travertine terrace system, the lateral inhibition proceeds predominantly upstream, by the pool dammed by the rim. The possible regular spacing between rims in the downslope direction (Fig. 9, and Hammer et al. 2007; but see Viles & Pentecost 1999 and Veysey & Goldenfeld 2008) may be understood in terms of this theoretical framework. However, the mechanism for lateral inhibition involves advection driven by complex hydrodynamics rather than simple isotropic diffusion, producing intricate morphologies. High travertine precipitation rates are generally found in positions of high flow velocity, whether or not there is a causal relationship between the two. In this respect, the travertine terrace system is precisely the opposite of more familiar systems of erosional flow (Hammer 2008). In high-velocity locations in rivers and streams the substrate is typically removed, rather than accreted, leading to localization of the flow channel. In contrast, water flow in the travertine terrace system is generally diverging, and the pools sprawl out laterally in
mutual competition. This unusual pattern formation regime may be responsible for the surreal impression invoked by travertine terrace landscapes.
Conclusions There is probably not a single mechanism responsible for localization of precipitation at the rim in all circumstances. At small scales with slow, laminar shallow water flow, we suggest that the Laplace instability effects suggested for ice by Ogawa & Furugawa (2002) and Ueno (2003) and confirmed in a travertine terrace setting by Hammer et al. (2008) can initiate microterracing. When the ridge begins to affect bathymetry and hydrodynamics, advective effects come into play by dampening small-wavelength features (Ogawa & Furugawa 2002; Ueno 2003), by bringing ions to and from the calcite surface (Veysey & Goldenfeld 2008; Hammer et al. 2008) and by compression of concentration gradients over the rim (Hammer et al. 2008). At slightly larger scales and for faster flow, turbulence sets in and diffusionlimited precipitation becomes dependent on flow rate by thinning of the laminar boundary layer (Buhmann & Dreybrodt 1985). At even larger scales relative to water depth and flow rate, spatial patterns of outgassing across the water-air interface can reach the water –calcite interface and open up a positive feedback loop involving loss of CO2. Across all scales and flow rates, mechanical sticking of particles on the rim may also play a role. Surface
354
Ø. HAMMER ET AL.
tension affects the hydrodynamics especially at smaller scales, influencing terrace morphology. In spite of recent results, travertine terracing remains an intriguing problem. Although there are many candidate mechanisms, the precise nature of precipitation localization under different conditions is not known. Do terrace size and spacing stabilize under certain conditions, or do they always continuously coarsen? When do we see terrace size distribution following a random null model (Veysey & Goldenfeld 2008) and when do we see a characteristic wavelength (Ogawa & Furukawa 2002)? What effects do parameters such as flux, oversaturation and slope have on patterning, and why? Such questions can probably be answered through a combination of computer simulations, theoretical studies, field studies and laboratory experiments, but the complexity and diversity of the system should not be underestimated. As noted by Veysey & Goldenfeld (2008), travertine precipitation is a convenient geological process to study because of the relatively fast processes. In addition, the accessibility of this earth-surface system, the availability of analytical and computational techniques, the cross-disciplinary nature of the problem and the sheer beauty and mystique of travertine terraces all make travertine terracing an attractive area of research. We would like to thank Paul Meakin for valuable discussions, and Allan Pentecost for permission to use his data for figure 5. This study was supported by a Center of Excellence grant to PGP from the Norwegian Research Council.
References A LTUNEL , E. & H ANCOCK , P. L. 1993. Morphology and structural setting of Quaternary travertines at Pamukkale, Turkey. Geological Journal, 28, 335–346. B ARGAR , K. E. 1978. Geology and thermal history of Mammoth hot springs, Yellowstone National Park, Wyoming. U. S. Geological Survey Bulletin, 1444, 1 –55. B ERGER , R. E. 1974. Garden cascades in Italy and France, 1565–1665. The Journal of the Society of Architectural Historians, 33, 304– 322. B ONO , P., D REYBRODT , W., E RCOLE , S., P ERCOPO , C. & V OSBECK , K. 2001. Inorganic calcite precipitation in Tartare karstic spring (Lazio, central Italy): field measurements and theoretical prediction on depositional rates. Environmental Geology, 41, 305–313. B UHMANN , D. & D REYBRODT , W. 1985. The kinetics of calcite dissolution and precipitation in geologically relevant situations of karst areas. 1. Open system. Chemical Geology, 48, 189– 211. C HAFETZ , H. S. & F OLK , R. L. 1984. Travertines: depositional morphology and the bacterially constructed constituents. Journal of Sedimentary Petrology, 54, 289– 316.
C HEN , J., Z HANG , D. D., W ANG , S., X IAO , T. & H UANG , R. 2004. Factors controlling tufa deposition in natural waters at waterfall sites. Sedimentary Geology, 166, 353–366. D REYBRODT , W., B UHMANN , D., M ICHAELIS , J. & U SDOWSKI , E. 1992. Geochemically controlled calcite precipitation by CO2 outgassing: Field measurements of precipitation rates in comparison to theoretical predictions. Chemical Geology, 97, 285–294. D REYBRODT , W., E ISENLOHR , L., M ADRY , B. & R INGER , S. 1997. Precipitation kinetics of calcite in the system CaCO3-H2O-CO2: the conversion to CO2 by the slow process Hþ þ HCO2 3 ,-. CO2 þ H2O as a rate limiting step. Geochimica et Cosmochimica Acta, 61, 3897–3904. E DDY , J., H UMPHREYS , G. S., H ART , D. M., M ITCHELL , P. B. & F ANNING , P. 1999. Vegetation arcs and litter dams: similarities and differences. Catena, 37, 57–73. F OUKE , B. W., F ARMER , J. D., DES M ARAIS , D. D., P RATT , L., S TURCHIO , N. C., B URNS , P. C. & D ISCIPULO , M. K. 2000. Depositional facies and aqueous-solid geochemistry of travertine-depositing hot springs (Angel Terrace, Mammoth Hot Springs, Yellowstone National Park, U.S.A.). Journal of Sedimentary Research, 70, 565–585. G IERER , A. & M EINHARDT , H. 1972. A theory of biological pattern formation. Kybernetik, 12, 30– 39. G IERER , A. & M EINHARDT , H. 2000. Pattern formation by local self-activation and lateral inhibition. BioEssays, 22, 753–760. G OLDENFELD , N., C HAN , P. Y. & V EYSEY , J. 2006. Dynamics of precipitation pattern formation at geothermal hot springs. Physical Review Letters, 96, article number 254501. G UO , L. & R IDING , R. 1998. Hot-spring travertine facies and sequences, Late Pleistocene, Rapolano Terme, Italy. Sedimentology, 45, 163–180. H AMMER , Ø. 2008. Pattern formation: watch your step. Nature Physics, 4, 265– 266. H AMMER , Ø., J AMTVEIT , B., B ENNING , L. & D YSTHE , D. K. 2005. Evolution of fluid chemistry during travertine formation in the Troll thermal springs, Svalbard, Norway. Geofluids, 5, 140– 150. H AMMER , Ø., D YSTHE , D. K. & J AMTVEIT , B. 2007. The dynamics of travertine dams. Earth and Planetary Science Letters, 256, 258– 263. H AMMER , Ø., D YSTHE , D. K., L ELU , B., L UND , H., M EAKIN , P. & J AMTVEIT , B. 2008. Calcite precipitation instability under laminar, open-channel flow. Geochimica et Cosmochimica Acta, 72, 5009–5021. H ORTON , R. E., L EACH , H. R. & VAN V LIET , R. 1934. Laminar sheet flow. Transactions of the American Geophysical Union, 2, 393 –494. J ETTESTUEN , E., J AMTVEIT , B., P ODLADCHIKOV , Y. Y., DE V ILLIERS , S., A MUNDSEN , H. E. F. & M EAKIN , P. 2006. Growth and characterization of complex mineral surfaces. Earth and Planetary Science Letters, 249, 108–118. L IU , Z., S VENSSON , U., D REYBRODT , W., D AOXIAN , Y. & B UHMANN , D. 1995. Hydrodynamic control of inorganic calcite precipitation in Huanglong Ravine, China: Field measurments and theoretical prediction
TRAVERTINE TERRACING of deposition rates. Geochimica et Cosmochimica Acta, 59, 3087– 3097. L IU , Z. & D REYBRODT , W. 1997. Dissolution kinetics of calcium carbonate minerals in H2O-CO2 solutions in turbulent flow: The role of the diffusion boundary layer and the slow reaction H2O þ CO2 ,-. H þ HCO2 3 . Geochimica et Cosmochimica Acta, 61, 2879–2889. L U , G., Z HENG , C., D ONAHOE , R. J. & B ERRY L YONS , W. 2000. Controlling processes in a CaCO3 precipitating stream in Huanglong Natural Scenic District, Sichuan, China. Journal of Hydrology, 230, 34–54. O GAWA , N. & F URUKAWA , Y. 2002. Surface instability on icicles. Physical Review E, 66, article number 41202. P ENTECOST , A. 1995. The Quaternary travertine deposits of Europe and Asia Minor. Quaternary Science Reviews, 14, 1005–1028. P ENTECOST , A. 2005. Travertine. Springer, Berlin.
355
P ENTECOST , A. & V ILES , H. 1994. A review and reassessment of travertine classification. Ge´ographie physique et Quaternaire, 48, 305– 314. U ENO , K. 2003. Pattern formation in crystal growth under parabolic shear flow. Physical Review E, 68, article number 21603. V ARNEDOE , W. W. JR 1965. A hypothesis for the formation of rimstone dams and gours. Bulletin of the National Speleological Society, 27, 151– 152. V EYSEY , J. & G OLDENFELD , N. 2008. Watching rocks grow. Nature Physics, 4, 310–313. V ILES , H. & P ENTECOST , A. 1999. Geomorphological controls on tufa deposition at Nash Brook, South Wales, United Kingdom. Cave and Karst Science, 26, 61– 68. W OODING , R. A. 1991. Growth of natural dams by deposition from steady supersaturated shallow flow. Journal of Geophysical Research, 96, 667– 682.
Index Note: Page numbers denoted in italics indicate figures and those in bold indicate tables. abiogenic processes 7 absolute dating methods 266 active springs 270 air circulation see circulation air sampling methods 327– 329 akaganeite 33 algae calcification 178 –186 colonization 181 desmid 155– 157 see also microalgae algal tufas 239 alkaline streams, tufa 65–81 aphotic zone, microbial processes 18–20 aragonite 18, 19, 33 chemical composition 211 crystal orientation 215– 218 laminae 215 morphologies 219 nanotextures 211–224 striations 221 scanning transmission x-ray microscopy (STXM) mapping 220– 221 textures 218 Austria Obir cave 295 –321 study site 298– 299 bacteria active/passive role 39 ion concentration 35 oxidizing 31 bacterial mineralization 32–36 active/passive 43– 44 models 31, 35 barite precipitation 42–43, 42 barrage tufa 193, 194 barrages 345 biofilm 7, 41, 52 anchorage 201, 205, 205 annual lamination 105– 107, 106 biodiversity 85 calcification pattern 105–107, 106 calcite precipitates 197– 198 colonization 193, 195 colonization rates 201, 202 development 75 erosion 207 growth 208 humidity 17–18 light gradients 17–18 nutrient availability 208 photic 197– 198 photosynthesis induced calcification 83 species composition 85
structure 105– 107, 106, 195– 197, 196–197, 202–203 substrate 205, 205 substrate destruction 11 temperature gradients 17–18 biogenic minerals, sequestration of radionuclides 32 biogenic processes 7 biogenicity 20– 21 biogeochemical cycles 44 biological activity 183 light intensity 180– 181 biomarker signals, lipid groups 23 biomineralization 31– 50 control mechanisms 44 myxococcus induced 37–43 organic macromolecules 35–36 strain specific 42 biosorption 38– 39 borings 10– 11, 11, 24 bryophyte tufas 239 buffering capacity 51, 52, 65, 68, 68, 75 calcarenite 39 calcification 83–118 calcite 18, 19, 33, 40–41, 40, 136 basal layer 199, 206 concentrations 236 constructive processes 12–17 corrosion 7 –30, 289 dissolution 11– 12, 24 mass balance 105, 112 nanofibres 225– 238 nanospherulites 197, 198 partition coefficient 305– 307 seasonal composition 337–339 solubility curve 338 spatiotemporal variability 283 sump grown 206– 207 see also rhombic calcite calcite crystals 13, 14– 15, 19, 201 microcrystalline calcite 300, 314– 315 morphology 15, 303, 315 type 12– 15 calcite fabric 339– 342 seasonal microclimate 323 –344 calcite growth rate 181–182 control parameters 283 calcite nucleation rate, algal growth 182 calcite precipitation 1, 2, 7 –30 aphotic extracellular polymeric substances (EPS) 198– 200 computer simulation 352 crystal morphology 17 microbial control 7, 17, 23
calcite saturation 83, 104, 104, 110, 110, 111, 323 calcite saturation index 15, 92, 143, 151–152, 172 –174, 178, 178, 183, 186, 255, 338, 342 tufa texture 179 water temperature 186 calcite-water partitioning data 309 calcium bicarbonate, Ion Activity Product 76 calcium carbonate polymorphs 33 saturation state 52 supersaturation 245 calcium carbonate growth, crystal effects 323 calcium carbonate precipitation 51–63 abiotic 42 by polysaccharides 59– 60 experiments 54– 55, 59, 60 calcium hardness 196–197 calcium ion concentration 194 –195, 289–291 carbon 14 dating 245, 258, 324 carbon dioxide see CO2 carbonate annual growth laminae 23 depositional climate 272–273 Henderson Hasselbach behaviour 69 mineralogy 51 morphology 51 nucleation 52, 144 production 39– 42 stable isotopecomposition107,108 textures 160 –167 carbonate precipitation 3, 5 biologically induced 3, 15 chemical processes 3 fungal contribution 225 –238 photosynthesis 60 rate 1 Carpathian Range 143 –191 study sites 145–149, 145, 146, 148 topography 147 water chemistry 151– 152, 153 cascades 345, 346 cave air 336 advection 285 calcite growth 287– 291 carbon isotopic concentrations 336, 338, 339 circulation 284–285, 317 CO2 partial pressure 283, 285–287, 288, 289, 293, 323, 334– 335, 335– 337, 337–339, 338, 342 humidity results 329, 330
358 cave air (Continued) sources and sinks 284–285, 286 stalagmite growth rate 283– 294 temperature 293 ventilation 293 cave microbes chemotaxonomic characters 9– 11 colonization 11 diversity 9 –11 migration 11 substrate 10–11, 12 twilight zone 17–18 cave microclimate 339–342 cave pearls see cave pistoliths cave pistoliths 19, 22–23 cave temperature 324, 329, 330 cave ventilation 285, 287, 338, 342 regimes 335–337 seasonal 295, 337–339 cave water hydrochemistry 331–335 percolation 283 caves aerosols 316– 317 aphotic zone 7 environmental monitoring methods 327 fungal activity 226 geomicrobiology 226–227 manganese deposits 33– 34 microbes 7– 30 microbial diversity 24 modern speleothems 323– 344 seasonal physiology 295– 321 trace element enrichment 316–317 twilight zone 7, 10–11, 12, 14– 15, 17–18 Cayman Islands, speleothems 7, 8 chalcocite 34 chlorides 35 circulation 296, 317 CO2 partial pressure 317–318 winter 317– 318 climate changes 24 cyclic 263 deposition/erosion 273–275, 276 climate proxies 317–318 climate reconstruction, speleothem fabrics 342 CO2 degassing 1, 83, 85, 88– 92, 119, 198, 206, 255, 283, 285, 323, 337, 350 –351, 353 carbon isotopes 292 surface tension 351 temperature gradients 351 CO2 degassing rate 338–339, 338, 342 CO2 degassing-calcite precipitation model 324, 338, 341, 342 CO2 partial pressure 283, 295, 296, 318, 324 circulation 317–318 seasonal variation 308 spatial distribution 289, 290 spatial measurement 284
INDEX temporal measurement 284 vertical distribution 285 colloid filtration 209 colloidal transport 307, 315, 318 columnar calcite 314, 339, 340–341, 342 compact calcite 339, 340– 341 computational fluid dynamics 349– 350 conductivity 122, 150, 196– 197, 258 Ion Activity Product 76 Kohlrausch’s Law 76 photosynthesis 78– 79 precipitation 76–77 variation 66, 70, 70 continental carbonates, tectonics 273 copper substrate 171 crystal fabrics, chemical control 314– 315 crystal orientation 205, 207, 219, 220– 221, 221, 300–303, 302, 314 crystal splitting 300 crystals columnar 314 rhombohedral 315 stepped faces 315 cyanobacteria 31, 93–99, 184 diversity 83, 94–95, 97, 111 extracellular polymeric substance (EPS) producers 52 filaments 155, 162, 163, 167, 171 light intensity 181 morphotypes 93, 93, 94, 95–97, 96–97, 111 phylogenetic analysis 96–97 picocyanobacteria 52 polysaccharides 51–63 rDNA lineages 95– 97 sequence analysis 113 desmid algae 155–157 detritic components 167 –168, 184 diatom frustules 172 diatom moulds, light intensity 181 diatoms 93–99, 157–159, 164, 166, 184, 202 diversity 83 phylogentic analysis 98 sequence analysis 113 species 98 tufa biofilms 98, 99 distribution coefficient 307, 316, 331, 333 drip water 295 calcite partitioning 318 calcium 338 characteristics 298–300 chemical composition 305– 308 chemistry 299, 323, 337 colloidal content 316 composition 316, 334–335, 337– 339, 342 discharge rates analysis 327 mean discharge 316
oxygen isotope ratio 291–292, 291 recharge pathways 329– 331, 332 seasonal composition 316, 337– 339, 342 stable isotope composition 307– 308 stalagmites distribution coefficient 307– 308 trace element composition 309 drip water rate 300, 301, 323 ion supply 305 variations 329–331, 332, 334–335 event laminae 304, 309– 312, 318 aerosols 317 circulation regime 317 elemental variation 309– 312 strontium pattern 309, 312 trace element patterns 315–317 exopolymers 83–118 extracellular polymeric substances (EPS) 35– 36, 44, 51, 85, 101–104, 193, 218 binding sites 51–52, 56, 57, 58, 58 biofilms 83, 219 calcite precipitation 85, 207 calcium binding capacity 52 crystal growth 203–206 degradation 85, 103, 110, 110, 112, 206 extraction protocol 55 Fourier transform infrared (FTIR) spectra 57 functional groups 55–56 heterotroph dominated 208 infrared spectroscopy 54, 55–56 ion transport 206 microbes 197 nucleation sites 85, 206 polygonal structures 200, 202– 203, 206, 207 polysaccharide isolation 52–53 potentiometric titration 53–54, 56, 57 secretions 157, 159, 164, 166 structural domains 83, 101–103, 102, 111 total buffering capacity 56– 59 extremophiles 32, 34, 44 ferrihydrite 33 filamentous microbes 13, 14–15, 225 algae 184 characteristics 227 phototrophs 199 trapping & binding 15 filaments 153–155 flume systems experimental methods 194–195 pH 67– 70, 68 seasonal data 71–74 fluvial facies 124, 125, 126, 132 fluvial tufa deposits 245, 263, 276 focused ion beam milling 212
INDEX framboidal pyrite 34 freshwater stromatolites 111, 203 lamination 207– 208 fungal hyphae 227– 228, 231 breakdown of 225, 236 crystal nucleation enhancer 236 scanning electron microscopy (SEM) photomicrographs 233 size 227 fungal hyphae cell wall 228, 233 calcite precipitation template 236 mineralization 235 structure and composition 228–229 fungi 225, 232 carbonate precipitation induction 236 caves 226 growth 236 soil 226 geomorphological survey methods 266 Gibraltar St Michaels cave 323– 344 study setting 324–327, 325, 326, 328 gours 345 groundwater 23 growth inhibitors 314 gypsum 19 halite 35 halogen elements 310 heterogeneous nucleation 205, 314 heterotrophic activity 75, 85, 112 prokaryotes 110 heterotrophic digestion 206 heterotrophic exopolymer degradation 110 heterotrophic organisms 1, 193, 201, 225 hunite 19 hydrochemistry 83–118, 111, 252 experimental methods 112 –113 karst steam water 87–92, 89–90, 91 River Piedra 124 hydrological cycle 183 stable isotope pattern 324 trace elements 324 hydrothermal springs 34, 267, 269 tectonics 273 Indonesia, Satonda crater lake 211– 224 inductively coupled plasma mass spectroscopy 297 infiltration laminae 304, 304, 315 infrared spectroscopy, extracellular polymeric substances (EPS) 54, 55– 56 inorganic carbon, dissolved concentrations 109– 111
inorganic precipitation 83 copper ions 171, 172 –174 ion microprobe analysis 297– 298, 308 data 310, 312 iron formations 31 iron sulphides 34 irradiance 183 isotope analysis 85, 143 see also stable isotope analysis isotopic geochemistry 23 tufa carbonate 65 Italy, Tuscany 263– 281 karst 1, 2 calcite precipitation model 351 dissolution 275 karst water 83– 118 cycles 88 experimental methods 66– 67 Germany 84, 86–87, 87 hydrochemistry 87–92, 89–90, 91 kerolite 18 lacustrine stromatolites 211 lacustrine tufa deposits 245 lake water levels, terracing 272– 273, 274 laminae 168–170, 184, 318 annual lead concentration 313 chemical trends 310 drivers for development 203 intercrystalline porosity 174 orientation 205 paired 339– 342, 340 –341 palaeoclimate 107– 109 seasonal 24 see also event laminae; infiltration laminae laminated speleothem fabrics 323 laminated tufa 119 laminated tufa biofilm 203 larval housings 160, 169, 184 lead 295, 313, 314, 318 light conditions 65, 67 irradiance 108 pH 77 photoperiod length 70–75, 71–74 seasonal changes 83 light gradients, biofilm 17– 18 light intensity 78– 79 biological activity 180– 181 limestone substrate 171 magnesium 315 calcium compositions 331, 333 calcium substitution 305 composition 308 magnesium calcite 41–42, 41 magnetite 33–34 magnetotactic bacteria 33– 34 manganese deposits, caves 33– 34 marine ecosystem 111 marine stromatolites 111 biofilms 202–203
359 massive tufa 119, 134–135, 140 metabolic activity 39 micrite 15, 16, 19, 22–23, 136, 170, 184 clotted 160–162, 174 fibrous 165– 166 hemispherically layered 162, 170 and water energy 179 micro erosion meter 123 microalgae 159 microbes calcification 12–15 caves 7 –30 deep sea floor 44 deep sub sea 32 extracellular polymeric substances (EPS) 197 extremophiles 32, 34, 44 fossilization 33 preservation 21 microbial activity 68 microbial assemblage, changes 79 microbial biofilms 1 microbial colonization 186 microbial composition flow rate 201 –202 temperature 201– 202 microbial diversity, tufa biofilms 100 microbial laminated carbonate 276 microbial precipitation 193 active or passive role 23 microbial processes, aphotic zone 18–20 microbialite fabrics 109– 111 microcrystalline calcite 300, 314–315 microorganisms 83– 118 micropeloids 197, 202, 204 microspar crystals 197, 201, 203, 206, 207 mineral precipitation 1, 15– 17 abiotic processes 32 biotic processes 32 genetic control 32 metabolic activity 32, 36 myxococcus induced 31–50 nucleation sites 101 types 33 water saturation index 77 mineral saturation indices 248 mineralized microbes fabrics 21 identification 21, 24 preservation 21–23 trapping and binding 24 monetite 242 monohydrocalcite 19 moonmilk 19–20, 227, 230 moss stems 159–160, 172 myxobacteria 36–37 Myxococcus Xanthus 43 myxospores 37 nanobacteria 44 nanofibres energy dispersive spectrometry (EDS) spectra 234
360 nanofibres (Continued) literature on 226 mineral 225 organic origin 225, 233–236 origin 232, 236 scanning electron microscopy (SEM) photomicrographs 231, 232, 234 structures 229 transmission electron microscopy (TEM) photomicrograph 232 vadose environments 232 nanoSIMS 213 nanospherulite 203, 204 newberyite 38 nitrates 34 non phototrophic prokaryotes 85, 99–101, 110, 111, 113 affiliation of isolates 100 nucleation 35, 219 feedback mechanisms 182 heterogeneous 205, 314 process 174 rate 172 role of organics 44 substrate 144 surfaces 176 see also renucleation horizons oncoids 2 Oocardium stratum 155–157, 170, 176 oolites see cave pistoliths organic colloids 295, 307 organic compounds 314 organic nanofibres 225, 233– 236 oxygen tufa biofilms 104, 104 see also stable isotopes palaeoclimate indicators 107–109 tufa stromatolites 112 palaeoclimate proxies, stable isotope concentration 291–293 palaeoclimate reconstruction, lake levels 263 palaeoclimate records, stalagmite growth rate 283–294 palaeoclimate studies 23– 24 palaeoenvironmental indicators 1– 3, 5 proxy 5 tufas 65 paludal depositional systems 245 paludal tufa systems 193, 276 palustrine tufa 263 pathological precipitates 44–45 perched springline tufa deposits 245– 262 depositional features 251– 252, 259 isotope composition 254 lobe top terrace 251 lower slope 252
INDEX regional setting 260 scanning electron microscopy (SEM) images 256 stable isotope data 260, 260 study site 249, 250 trace elements 254 water chemistry 253 waterfall zone 251–252 petrographic analysis methods 266, 296–297 pH 65, 122, 150, 205– 206, 255, 296, 318 diurnal cycle 67–70, 69 flume systems 67– 70, 68 light conditions 77 photosynthesis 75, 78–79 respiration 75 seasonal variation 308 temperature 77 tufa biofilms 104, 104 variability 66, 75, 308 phosphates 34–35 production 37–39, 39 phosphorus 313 acid fraction 241, 243 alkali fraction 239, 241, 243 carbonate fraction 242 co-precipitation 242 concentration 311 dithionite fraction 239, 240, 243 ethylene diamine tetraacetic acid (EDTA) fraction 240, 243 form of 242 fractionation 239– 244, 241, 241 fractionation methods 239– 240 sodium carbonate fraction 242 uptake ratios 239 water soluble 240, 242 photosynthesis 85, 105 carbonate precipitation 60 CO2 assimilation 110 pH 75 respiration 75, 78–79 water chemistry 75– 76 phototrophic metabolism 75, 206 phototrophic microbes 193 picocyanobacteria 52 polysaccharides, active sites 51, 52– 53 pool fingers 18– 19, 20 potentiometric titration, extracellular polymeric substances (EPS) 53– 54, 56, 57 precipitation activation energy knick point 2 conductivity 76–77 mechanisms 205 –206 microbial influence 77– 79 photosynthesis induced 109, 112 products 207 see also calcite precipitation; mineral precipitation prokaryotes 85, 99– 101, 100, 110, 111, 113
quartz formation 44 radiometric data 263 rainfall 108, 296, 334 –335 data 121 tufa deposition 263 renucleation horizons 304–305, 304 respiration 205– 206 pH 75 photosynthesis 75, 78–79 rhombic calcite 206 rimstone 345 river valley, geomorphological evolution 263– 264 scanning electron microscopy 248, 303, 303 methods 212, 297 results 229– 232 scanning transmission x-ray microscopy 213 schertelite 38 seawater, composition 109, 111 sediment isotopic composition 140 monitoring methods 122–124, 266 stable isotope analysis 122– 124 thickness 126 sedimentation rates 122, 127–128 measurements 126–129 periodic patterns 126 seasonal pattern 119 seasonal variations 124– 126 sedimentological data 119 –142 silicates 34 soil air 336 carbon cycle 225 CO2 partial pressure 283, 289 fungal activity 226 mineralized nanofibres 225 samples 229 temperatures 283, 287, 289– 291, 291, 324 Spain climate 121–122 geology 121–122 hydrology 121– 122 tufa study area 120 sparite 162– 163, 174, 176 sparite bushes 163– 165, 176 sparry calcite 165–167, 172, 178, 183 speleothems 1 –5, 2, 243, 283 calcite partitioning 318 Cayman Islands 7, 8 climate reconstruction 342 cyclical calcite deposition 324 environmental change record 295 fabrics 1, 7, 323, 342 formation processes 7, 20 geochemistry 24, 323 growth rate 7, 318 microbes 20– 23 microclimate proxy record 323
INDEX water supply 1 zinc 318 sphalerite 34 spring water, chemical composition 258 stable isotopes analysis 245, 297 annual variations 296 composition 107– 109, 108, 136, 137, 138 –139, 295, 307–308 concentrations 83, 248, 251 data 119–142 flow conditions 138–139 hydrological cycle 324 ratios 143, 323 seasonal microclimate 323– 344 seasonal pattern 138–139 tufa growth rates 143 variations 339–342, 340–341 ventilation 324 stalactites 1, 15, 16, 18 abiogenically precipitated 24 stalagmite growth rate 287–289, 291, 309 cave atmosphere 283– 294 palaeoclimate records 283–294 stalagmites 18 annual lamination 295– 321, 303–305 composition 305– 308, 306, 307, 307, 308–313, 308, 312, 316 depth trends 306 drip water 289–291, 291, 307 fabrics 300–303 geochemistry 295– 321 growth inhibitors 292 –293 height data 292 ion microprobe 302 lamina thickness 306 lead concentration 313 magnesium concentrations 309–310 petrology 295–321, 300–305, 302, 303, 304 phosphorus variation 311 seasonality 291, 308–313, 317–318 stable isotopes 306, 307– 308, 308–309 sulphur variation 311 water film thickness 289 stromatolites 31, 41 biogenicity 211 compositional analysis 214 coral texture 219 cyanobacteria 96–97 data 213–218 formation 111, 211 freshwater and marine 111, 193 Indonesia 211–224 laminated 15, 16, 83, 96– 97 mineralogical textures 211
neomorphism 112 samples 212 scanning electron microscopy (SEM) images 215, 217 transmission electron microscopy (TEM) images 216, 218 see also freshwater stromatolites; lacustrine stromatolites; marine stromatolites; tufa stromatolites stromatolitic tufa 119, 132–134, 140 strontium 309, 312, 323 abundance 339, 336, 341 composition 308 struvite 37, 38 substrate 171, 206 sulphate 34, 295 annual variations 296, 309 calcite partitioning 305 pH variation 309 production 42–43 sulphate reducing bacteria 52, 111 sulphur concentration 311 surface tension 347, 347, 353–354 CO2 degassing 351 surveying and dye tracing methods 327 taylorite 43 temperature 65, 67, 70– 75, 140, 183, 296, 300, 301 calcite precipitation 207 calculations 139–140 data 121 pH 77 seasonal changes 83 stable isotope data 119 tufa microfabric 193–210 ventilation 323 water chemistry 183, 185 temperature changes 70– 75, 71–74, 126 temperature gradients biofilm 17– 18 CO2 degassing 351 terraced carbonate deposits 267– 271, 346–352 terraces 345 terrestrial carbonates 263–264 thermal springs see hydrothermal springs topography, travertine terraces 348, 349, 351– 352 total buffering capacity, extracellular polymeric substances (EPS) 56– 59 trace elements 314 concentrations 24, 296 cycles 323, 339–342, 340–341 seasonal microclimate 323 –344 transmission electron microscopy 213 travertine 143 travertine terraces 345–355, 346 bacterial origin 347 characteristic spacing 353, 353
361 dry 348 dynamics 348 flow velocity 351–352, 352, 353 formation 348– 350, 352 –353 growth rate 348, 353 hydrodynamics 345 –348 morphology 345– 348, 351– 352 precipitation localization 350–351 rim growth rate 353 rim orientation 348 rim self repair 348 rims 347, 347, 350– 351 topography 348, 351–352 water chemistry 351– 352 travertines 2, 181, 245, 273 cross section 350 deposition and tectonics 273 growth 172 tritium 255 tufa 1– 5, 2, 119, 143 active deposition 145 alkaline streams 65–81 annual laminations 144 biogenic components 153–170 biological composition 255 biomediated precipitates 78 cascade model 245 cave deposits 140 characteristics 122 chemical conditions 65–81 chemistry 255 climatic indicators 263–281 cold climate 275, 276 composition 240 cool humid models 193, 194 cross sections 129– 132, 129, 130, 131, 132 crystalline texture 143 crystals 173 cyclic deposition 272 dense laminated 132– 134 depositional properties 245– 262 fabrics 1, 52 facies 132– 136 Fe:P ratios 243 fibrous texture 181 fluvial 119– 142 formation processes 65, 85 geochemistry 245– 262 geomorphological evolution 272 Germany 85 high energy settings 177 internal structure 151–170, 267 isotopic geochemistry 65, 79 laminae 65, 79, 132– 134, 144 lobe age relations 252, 258 mass exchange 78 micritic texture 143 microbial components SEM images 257 microbial influence 65–81, 144 microfabric 193–210 mineralogy 122, 136, 255 organically bound phosphorus 243
362 tufa (Continued) origin 144 palaeoclimatic indicators 65, 245 palaeoenvironmental archive 144–145 palaeohydrological indicators 245 phosphorus content 240 phosphorus fractionation 241, 241 porous spongy 134–135, 135, 140 precipitation processes 65, 85 profile 208 radiocarbon 255–258 seasonal record 119– 142 sediment photomicrographs 133, 134, 135 semi arid models 193, 194, 275 Spain 119–142 stable isotope composition 136, 137, 138, 255–258 steep banded deposits 135– 136, 140 structure 270 temperature 139, 193–210 thickness 144, 158–160 water hydrochemistry 258 water supply 1, 193 –210 Yorkshire 239– 244 see also barrage tufa; lacustrine tufa deposits; massive tufa; paludal tufa systems; palustrine tufa; perched springline tufa deposits tufa biofilms 83– 118, 85 affiliation of isolates 100 calcification 83, 112 cyanobacteria 96–97 diatoms 98, 99 microbial diversity 100 microgradients 104–105, 104 oxygen 104, 104 pH 104, 104 seasonal irradiance 109 tufa growth rate 143 –191, 152 –153, 154, 155, 156 annual 185 climate conditions 184– 186, 275–276 copper ions 171, 172–174 data 158–160 environmental conditions 119 facies 143 isotope ratios 143 measurement methods 150–151 microenvironment 174–178 microorganisms 170– 174 photosynthesis 119 rainfall 263 seasonal 176, 182–184, 186 substrates 171 temperature 184– 186, 185 texture 181 variation 182– 183 water chemistry 185, 186 water energy 179 –180
INDEX water flow 143 water origin 184–186 tufa samples laboratory methods 247– 248, 248– 251 thin sections 162 Yorkshire 239 tufa stromatolites 105– 107, 106 annual laminae 112 biofilms 86, 98 formation 83 geobiology 84– 85 neomorphism 109 palaeoclimatic indicators 112 stable isotopes 112 tufa texture 143– 191, 178, 208 algae density 181 algae diversity 181 calcite saturation index 179 data 158–160 environmental factors 179– 184 growth rate 181 microenvironment 174– 178 seasonality 174– 178, 182– 184 study methods 151 water flow 186 Turkey active waterfall 248, 251 field photos 249 geological setting 245–246, 246, 247 Guney waterfall 245– 262 perched springline tufa deposits 245– 262 uranium phosphate 38– 39 uranium series dating 267, 283 vadose water 295 Valdelsa succession 263, 276 carbon 14 data 272 chronological dating 271 episodes of tufa development 272 facies associations 267– 271, 267, 268, 269, 270 geological setting 264–266, 265 palaeoclimate 272, 275 stratigraphic units 266 tectonic regime 272 U/Th data 271 vaterite 17, 19, 33, 40– 41, 40, 207 Vaucheria 184 filaments 153– 154, 161, 172 light intensity 180– 181 ventilation 318, 323 cycles 342 seasonal variation 308 stable isotope pattern 324 temperature 323 see also cave air water calcite partitioning data 309 calcite saturation index 180
depth 122 monitoring methods 122 –124 physiochemistry 252– 255 saturation index and mineral propitiation 77 temperature 108, 122, 180, 252, 258, 300, 301 see also drip water; groundwater; hydrothermal springs; rainfall; seawater; spring water water chemistry 65, 248 Carpathians 151– 152, 153 data 155, 158 measurement methods 150 –151 oxygen isotopes 136 perched springline tufa deposits 253 photosynthesis 75–76 temperature 184 travertine terraces 351– 352 tufa growth rate 178–186, 178, 185 variations 5 water energy micrite 180 porous tufa 179 sparry crystals 179 tufa growth 179– 180 water flow 183, 314 caves 131–132 fast 129– 130 precipitation rates 1, 200– 201, 207 rate 194 –195, 196– 197, 201, 351 rate thresholds 202 sedimentary record 129– 130, 131– 132 solute concentration gradients 351, 353 spray areas 131–132 stable isotope composition 138– 139 standing to slow 131 stepped waterfalls 131 travertine terraces 351–352, 352, 353 tufa growth rate 174, 349– 350 tufa microfabric 193–210 tufa texture 186 velocity 1, 122, 174, 193–210, 345– 347, 349– 350, 349, 351– 352, 352, 353 weddelite 43, 43 x-ray diffraction 248 data 305 methods 297 results 230 x-ray fluorescence data 313, 314 Yorkshire, tufas 239–244 zinc 295, 305, 313, 314, 315
Our understanding of calcium carbonate precipitation within freshwater carbonate systems is being revolutionized by new quantitative approaches at both field and laboratory scale. These systems cover a diverse range of topical research areas including tufas, speleothems, stromatolites and microbial processes. Progress by various international research groups has been impressive, with major contributions to such areas as climate change, absolute dating, carbon sequestration, and biofilm construction and precipitation. A diverse sample of interrelated research is presented that provides a tantalizing glimpse of the interplay between microbial, geochemical and physical processes that control the development of tufas and speleothems. This volume will provide a cross-disciplinary platform that will stimulate further exchanges about new concepts, methodologies and interpretations associated with freshwater carbonates. In particular, it will help reinforce the importance of cross-discipline research: the driving force behind the new field of Geobiology.