CALCRETES
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
CALCRETES ED ITED
BY
V. PAUL WRIGHT Postgraduate Research Institute of Sedimentology (PRIS), University of Reading AND
MAURICE E. TUCKER Department of Geological Sciences, University of Durham
REPRINT SERIES VOLUME 2 OF THE INTERNATIONAL ASSOCIATION OF SEDIMENTOLOGISTS PUBLISHED BY BLACKWELL SCIENTIFIC PUBLICATIONS OXFORD LONDON EDINBURGH BOSTON MELBOURNE PARIS BERLIN VIENNA
© 199I
The International Association
of Sedimentologists Published by Blackwell Scientific Publications Editorial officies:
DISTRIBUTORS
Marston Book Services Ltd PO Box 87 Oxford OX2 ODT
(Orders:
Telex: 837515)
25 John Street, London WC1N 2BL 23 Ainslie Place, Edinburgh EH3 6AJ 3 Cambridge Center, Cambridge Massachusetts 02142, USA 54 University Street, Carlton Victoria 3053, Australia Other Editorial Officies: Arnette SA 2, rue Casimir-Delavigne 75006 Paris France Blackwell Wissenschaft Meinekestrasse 4 D-1000 Berlin 15 Germany
Tel: 0865 791155 Fax: 0865 791927
Osney Mead, Oxford OX2 OEL USA
Blackwell Scientific Publications, Inc. 3 Cambridge Center Cambridge, MA 02142
(Orders:
Tel: 800 759-6102)
Canada Oxford University Press 70 Wynford Drive Don Mills Ontario M3C 1J9
(Orders:
Tel: 416 441-2941)
Australia Blackwell Scientific Publications (Australia) Pty Ltd 54 University Street
Blackwell MZV
Carlton, Victoria 3053
Feldgasse 13
(Orders:
Tel: 03 347-0300)
A-I238 Wien Austria All rights reserved. No part of this publication may be reproduced, stored
British Library Cataloguing in Publication Data Calcretes.
in a retrieval system, or transmitted,
1. Deserts. Sedimentation
in any form or by any means,
I. Wright, V. Paul
electronic, mechanical, photocopying,
II.
recording or otherwise without the
III. Series
prior permission of the copyright owner.
1953-
Tucker, Maurice E.
552.5 ISBN 0-632-03187-5
First published I991 Library of Congress Set by Setrite Typesetters, Hong Kong Printed and bound in Great Britain at the Alden Press, Oxford
Cataloging in Publication Data Calcretes/edited by V. Paul Wright and Maurice E. Tucker. p.
em. - (Reprint series v. 2
of the International Association of Sedimentologists) Includes index. ISBN 0-632-03187-5
l.
Calcretes.
l. Wright,
Il. Tucker, Maurice E. Reprint series
V.
Paul, 1953-
Ill. Series:
of the
International Association of Sedimentologists; v. 2. QE471.15.C27C25 552'.5-dc20
1991
Contents
vu
I77 A rendzina from the Lower Carboniferous of South Wales [pages 159-167 only plus
Preface
references]
I Calcretes: an introduction
V. P.
23
Quaternary Calcretes
WRIGHT
Sedimentology
25 Calcretes of Olduvai Gorge and the Ndolanya
1983
30159-179
I89 The role of fungal biomineralization in the
Beds of northern Tanzania
formation of Early Carboniferous soil fabrics
R. L. HAY & R.J. REEDER
V. P.
Sedimentology
Sedimentology
25
1978
649-673
51 Pellets, ooids, sepiolite and silica in three HAY &
Sedimentology
B.
33
831-838
caliche profiles in a Bahamian Pleistocene dune
WIGGINS
1980
1986
I97 Petrographic and geochemical analysis of
calcretes of the southwestern United States R.L.
WRIGHT
27 559-576
J.A. BEIER
Sedimentology 1987 34
991-998
69 Quaternary pedogenic calcretes from the Kalahari (southern Africa): mineralogy,
205
Biological Activity and Laminar Calcretes
genesis and diagenesis 207 Origin of subaerial Holocene calcareous
N.L. WATTS
Sedimentology
1980
27
crusts: role of algae, fungi and
661-686
sparmicritisation 95
Biological Activity and Calcrete Fabrics
C.F. KAHLE
Sedimentology
1977
24
413-435
97 Caliche profile formation, Saldanha Bay (South Africa)
23 I Calcification in a coccoid cyanobacterium associated with the formation of desert
G. F. KNOX
Sedimentology
1977
24
stromatolites
657-674
W. E. KRUMBEIN &
I15
C.F.
C.
GIELE
Sedimentology 1979 26 593-604
Biolithogenesis of Microcodium: elucidation KLAPPA
Sedimentology 1978 25
243
489-522
Biogenic laminar calcretes: evidence of calcified root-mat horizons in paleosols
I49 Rhizoliths in terrestrial carbonates:
V. P. WRIGHT,
classification, recognition, genesis and
WIMBLEDON
significance
Sedimentology
C.F. KLAPPA
Sedimentology
1980
27
261
613-629
PLATT &
N. H.
1988
35
W. A.
603-620
Aspects of Calcrete Petrography
263 Calcrete conglomerate, case-hardened
I67 Calcrete profiles in the Eyam Limestone (Carboniferous) of Derbyshire: petrology and
conglomerate and cornstone-a comparative
regional significance
account of pedogenic and non-pedogenic
A. E .
carbonates from the continental Siwalik
ADAMS
Sedimentology
1980
27
Group, Punjab, India
651-660
S. K. TANDON & D.
Sedimentology 1981 28
v
NARAYAN
353-367
Contents 321
279 Siliciclastic grain breakage and displacement
Calcretes and Palustrine Carbonates
due to carbonate crystal growth: an example 323
from the Lueders Formation (Permian) of
Lacustrine carbonates and pedogenesis:
north-central Texas, USA
sedimentology and origin of palustrine
C. BUCZYNSKI &
deposits from the Early Cretaceous Rupelo
Sedimentology
1987
H.S.
34
CHAFETZ
Formation, W Cameros Basin, N Spain
837-843
N.H. PLATT
Sedimentology
287 Near-surface shrinkage and carbonate replacement processes, Arran Cornstone 343
Formation, Scotland S.K. TANDON &
Sedimentology
1989
36
349 Index
1113-1126
301 The application of cathodoluminescence to interpreting the diagenesis of an ancient calcrete profile S.T. SOLOMON &
Sedimentology
1985
References
P.F. FRIEND
32
G.M. WALKDEN
877-896
VI
1989
36
665-684
Preface
Calcretes are an important component of many
calcretes from southern and eastern Africa and the
ancient fluvial, lacustrine and shallow-marine car
south-western United States. The examples are all
bonate sequences and they are widely developed in
of calcretes which show few biogenic features. The
many parts of the world at the present time. Cal
second section contains seven papers on modern and
cretes are useful to the earth scientist involved in
ancient calcretes which possess many biogenic fabrics,
reconstructing ancient environments, palaeoclimates
including the enigmatic Microcodium.
and palaeogeographies, and they may also reveal
section,
details of soil biota and chemistry. Papers on cal
laminar calcretes, some of which are the result of
cretes are published in journals of soil science,
calcification of root mats. Some specific textural
geomorphology, sedimentology and general geology,
features of calcretes are illustrated in the fourth
but in the last two decades the journal Sedimentology
section, with four papers describing examples from
with three
reprints,
The next
is concerned
with
has received many on this subject, so that a com
India, the USA, England and Scotland. The book is
pilation of them has been put together to make this
concluded with a case-history of lacustrine sedi
second reprint volume of the International Associa
mentation and pedogenesis, with a description of
tion of Sedimentologists.
palustrine limestones from Spain.
Calcretes have been studied by people from dif
This collection of reprints should illustrate the
ferent backgrounds and with different interests, so
range of calcrete occurrences and the great variety
that this book also provides a review of the work on
of textures and fabrics. It should serve as more than
calcretes as an introduction to the topic and the
an introduction to the subject and be of use to
papers that follow. Eighteen papers are reproduced
geologists, soil scientists and geographers.
here and they have been divided into five groups,
V. Paul Wright
each preceded by a short commentary. The first
Maurice E. Tucker
section has three papers which describe Quaternary
vii
CALCRETES: AN INTRODUCTION*
Calcrete is a near surface , terrestrial, accumulation of predominantly calcium carbonate, which occurs in a variety of forms from powdery to nodular to highly indurated. It results from the cementation and displacive and replacive introduction of calcium carbonate into soil profiles, bedrock and sediments, in areas where vadose and shallow phreatic ground waters become saturated with respect to calcium carbonate. This definition is modified from Goudie ( 1973) and Watts (1980, this volume). The term 'dolocrete' is used where the main carbonate phase is dolomite. Calcretes are not restricted to soil profiles (pedogenic calcretes) but can also occur, for example, below the zone of soil formation but within the vadose zone, or at the capillary fringe and below the water-table to form groundwater calcrete. A very general definition is preferred here be cause the term has been used very loosely in the past. It would serve no purpose to review the ter minological quagmire, but it is more important to identify the processes of formation and hydrological setting of terrestrial carbonate accumulations than to have a post-mortem on the misuse, or supposed misuse of the term. The term is not used to describe tufas, travertines, beachrock and lake carbonates. However, it is a moot point as to whether many types of simple carbonate cementation, such as that seen in aeolianites for example, are not classifiable as calcrete. Most calcretes are finely crystalline and in their more mature forms, consist of a more-or-less continuous secondary matrix of micrite or microspar grade carbonate. Thus the fabric differs from simple cementation which is typically more coarsely cry stalline in a grain/clast supported fabric. Goudie (1973) has provided a detailed review of the various terms used to describe calcrete materials. The term is virtually synonymous with 'caliche' in its current usage by English-speaking workers. Milnes (1991) provided an historical review of the development of ideas on calcrete formation. The most important and widespread calcretes are those which form in soil profiles (e .g. Fig. 1A). These *
Reading University, PRIS Contr. 115.
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
accumulations constitute calcic or petrocalcic hor izons (if continous and indurated) in the terminology of soil scientists. It has been estimated that such soils today cover an estimated 20 million km2 or about 13% of the total land surface (Yaalon, 1988), and it is little wonder that a huge literature exists on such soils. They are a prominent feature in climatic zones where a seasonal moisture deficit occurs, allowing CaC03 to accumulate (Goudie, 1973, 1983 ) . While calcretes are important in landscape development, providing a geomorphic 'threshold' for erosion, they are not as important in this respect as other duricrusts such as laterites and silcretes. They are not as re sistant to erosion as these other forms, since calcretes tend to form nearer the land surface and they are not part of thick saprolite profiles which typically form in more humid climates with deeper weather ing. Calcretes present problems for land use and are associated with serious soil erosion in many regions. Despite this fact, relatively little has been published on catenary relationships, calcrete thicknesses and soil erodability. Calcretes were just as widespread in the past as they are today. They have been widely recognized in ancient sedimentary sequences (e.g. Fig. 1 B , C), even from the Precambrian (Chown & Caty, 1983; Bertrand-Sarfati & Moussine-Pouchkine, 1983 ) . There are numerous records from Phanerozoic se quences where calcretes have been used as palaeo climatic indicators and to assess depositional rates in architectural models of alluvial sequences (e.g. Allen, 1974a; Steel, 1974; Hubert, 1978; Leeder, 1975; McPherson, 1979; Wright, 1982). Palaeo-calcretes have been described from two distinctive settings: alluvial/lacustrine and shallow water carbonate systems. In the latter case, as a result of the widespread development of exposure surfaces within Quaternary shallow-water carbon ates, and because of the realization that important stabilization and cementation take place during exposure to meteoric conditions, there was a surge of interest in identifying such exposure surfaces in ancient sequences. Descriptions of Quaternary limestone-hosted calcretes (Braithwaite, 1975, 1983; James, 1972; Read, 1974; Harrison, 1977) led to the
Calcretes: An Introduction
Fig. 1. (A) Stage 5 Quaternary calcrete, upper La Mesa surface, Rio Grande rift area, New Mexico. This profile is in excess of 400 ka and has a laminated crust capping the petrocalcic horizon. (B) Prismatic Stage 4 calcrete, Lower Devonian, Lydney, England. (C) Stage 3 calcrete, Upper Jurassic, Porto Novo, Portugal. Note overlying channel sandstone.
2
Calcretes: An Introduction
genic calcretes and was developed for geotechnical surveys. Table 1 is largely based on Netterberg's classification. The forms of calcrete recognized in this classi fication relate to stages seen in the development of calcrete profiles (Netterberg, 1980) . For example, scattered nodules with time pass to glaebular cal crete, to honeycomb calcrete , to hardpans, and may later weather to boulder calcrete if soil conditions change . This type of maturity-related classification (a chronosequence) has been offered by a number of workers for both Quaternary and pre-Quaternary calcretes (Gile et al., 1966; Allen, 1974a,b; Steel, 1974; Machette, 1985 ) . Machette ( 1985) has pro vided the most comprehensive sequence and has
discovery of numerous ancient examples (e.g. Walkden, 1974; Walls et al . , 1975; Harrison & Steinen, 1978; Adams, 1980; Adams & Cossey, 1981; Riding & Wright, 198 1 ; Wright, 1983) , and a suite of characteristic and diagnostic petrographic criteria quickly developed, thoroughly reviewed by Esteban & Klappa ( 1983). It is reasonable to state that while most sediment ologists working on calcretes in alluvial/lacustrine sequences (e.g. Allen, 1974a ,b; Freytet & Plaziat, 1982; Wright, 1982) interpreted the calcretes using the terminology and concepts of soil science , many carbonate sedimentologists treated calcretes as diagenetic features. This almost led to a dual ap proach , made worse because many soil scientists were unaware of the extensive literature on soil carbonates to be found in sedimentological journals. Recently these divisions have become blurred with much more 'cross-pollination' of ideas. In order to review calcretes severafbasic questions will be tackled: how can they be classified? where do they form? how do they form? what are the sources of the carbonate? what are the processes responsible for the profile form? what are the microstructures of calcretes? what are the geochemical controls and what isotopic criteria can be used to interpret palaeo-calcretes?
Table 1. Morphological classification of calcretes based on Netterberg (1967, 1980) and Goudie (1983)
CLASSIFICATION
Calcareous soil
Very weakly cemented or uncemented soil with small carbonate accumulations as grain coatings, patches of powdery carbonate including needle-fibre calcite (pseudomycelia), carbonate-filled fractures and small nodules
Calcified soil
A firmly cemented soil, just friable; few nodules. 10-50% carbonate
Powder calcrete
A fine, usually loose powder of calcium carbonate as a continuous body with little or no nodule development
Pedotubule calcrete All, or nearly all, the secondary carbonate forms encrustations around roots or fills root or other tubes (tubules)
Before reviewing how calcretes are classified 1t 1s worth examining the way calcretes fit into soil class ifications. No widely used soil classification includes calcrete (or some synonym) as a soil type (soil order). There are no soils called calcretes. Pedogenic calcretes occur within soil profiles, where they typically constitute several discrete horizons (e.g. calcic horizons or petrocalcic horizons) forming a sub-profile within the main soil profile. The use of the term 'calcrete profile' in this paper refers to a set of related calcic/petrocalcic horizons within a thicker soil profile. Calcretes typically occur within Aridisols, Vertisols and Mollisols (Soil Survey Staff, 1975). Five main types of classification are widely used; all are basically generic, reflecting our still relatively poor understanding of calcrete formation. At the simplest level calcretes can be classified on their morphology, with the system devised by Netterberg ( 1 967, 1980) , which supersedes that of Durand ( 1963 ) , being the most useful. This refers to pedo3
Nodular calcrete
(syn. glaebular calcrete of Netterberg, 1980.) Discrete soft to very hard concretions of carbonate-cemented and/ or replaced soil. Concentrations may occur as laminated coatings to form pisoids
Honeycomb calcrete
Partly coalesced nodules with interstitial areas of less indurated material between
Hardpan calcrete
(syn. petrocalcic horizon.) An indurated horizon, sheet-like. Typically with a complex internal fabric, with sharp upper surface, gradational lower surface
Laminar calcrete
Indurated sheets of carbonate, typically undulose. Usually, but not always, over hardpans or indurated rock substrates
Boulder/cobble calcrete
Disrupted hardpans due to fracturing, dissolution and rhizobrecciation (including tree-heave). Not always boulder grade. (Clasts are rounded due to dissolution)
Calcretes: An Introduction
primarily non-pedogenic carbonates modified by pedogenic processes. Shallow lake , pond or marsh (palustrine) carbonates are easily modified by ex posure and can resemble mature profiles , especially where extensive desiccation-related grainification has occurred (Wright, 1990a) (Fig. 2). This problem is discussed by Platt ( 1989 , this volume) and Esteban & Klappa ( 1983 ) . Very simply, the dense, fine grained carbonate which develops in low energy Jakes and marshes can be confused easily with con tinuous calcrete fabrics of hardpan (Stage 3) profiles, especially when desiccation has overprinted the original micrite . Calcretes developed on unconsolidated carbonate substrates also develop in a series of stages (Arakel, 1982) broadly similar to those on other host materials. Calcretes can also be classified according to their hydrological setting (Carlisle, 1980, 1983) (Fig. 3). A common misconception is that calcretes only form
recognized six stages of development (Table 2; Fig. 2). In this classification the role of parent (host) material is important and the distinction between calcretes developed in gravel-rich substrates , as against those in gravel-poor ones, is critical because the profiles develop much more rapidly in gravel rich substrates (Gile et a! ., 1966; Machette, 1985 ) . The most mature (Stage 4) profiles contain evi dence of polycyclic brecciation and cementation. The Ogallala cap rock of New Mexico and Texas is such an example and contains huge numbers of well rounded peloids and pisoids formed by the fractur ing, including circum-granular fracturing, and trans portation of these grains into fracture and cavity systems in the calcrete. This process of secondary grainification is discussed below, and by Hay & Wiggins ( 1 980, this volume) . A major problem arises in classifying some continental carbonates, which, while resembling mature calcretes (Stages 4-6) are
Table 2. Classification of pedogenic calcretes based on stages of development. From Machette (1985). High gravel content refers to >50% gravel. Low is less than 20% gravel. The per cent CaC03 refers to <2 mm fraction. (K is a carbonate soil horizon; m refers to induration)
Gravel content
Diagnostic features
CaC03 distribution
Maximum CaC03 content
High
Thin discontinous coatings on pebbles, usually on undersides
Coatings sparse to common
Trace to 2%
Low
Few filaments in soil or faint coatings on ped surfaces
Filaments sparse to common
Trace to 4%
High
Continuous, thin to thick coatings on tops and undersides of pebbles
Coatings common, some carbonate in matrix
2-10%
Low
Nodules, soft 5-40 mm in diameter
Nodules common, generally non-calcareous to slightly calcareous
4-20%
High
Massive accumulations between clasts, fully cemented in advanced forms
Continuous in matrix to form K fabric
10-25%
Low
Many coalesced nodules, matrix is firmly to moderately cemented
Continuous in matrix to form K fabric
20-60%
4
Any
Thin (<2 mm) to thick (10 mm) laminae capping hard pan (K,.)
Cemented , platy to tabular structure. K,. horizon is 0.51 m thick
>25% in high gravel content. >60% in low gravel content
5
Any
Thick laminae (>10 mm); small to large pisoids above. Laminated carbonate may coat fracture surfaces
Indurated, dense, strong, platy to tabular. K,. horizon is 1-2 m thick
>50% in high gravel content. >75% in low gravel content
6
Any
Complex fabric of multiple generations of laminae, brecciated and recemented, pisolitic. Typically with abundant peloids and pisoliths in fractures
Indurated, dense, thick, strong tubular structure. K,. horizon is commonly >2 m thick
>75% in all gravel contents
Stage
2
3
4
Calcretes: An Introduction
,,�. .. } 11 � -r.;· . . . 2 secondary lime mudstone
2. (Upper part) Stages in calcrete development in fine-grained sediment. By Stage 4 the calcrete is dense and impermeable (petrocalcic horizon) leading to the pending of soil water and the formation of laminar calcrete. Stage 6 calcretes show evidence of extensive brecciation. Fractures become filled by peloids and cemented and refractured. The calcrete first develops a secondary lime mudstone fabric but later grainification takes place to form peloids. (Lower part) Calcretes can be difficult to distinguish from pedogenically modified shallow lake margin (palustrine) lime mudstones. These can also be affected by grainification (see text).
Fig.
w (f) 0 0
�
u
�w
a: I [l_
Fig.
Soil moisture zone
:-:�
Gravitational wa ter zone
Ca p i l lar y fringe �
�
... ' ' Groundwater
--- G R AVITATIONAL
ZONE NON- PEDOGENIC CALCRETE
local role of phreatophytic plants
• -
•---
- -
,�� .. ... .... _ : _..
.......-----!
I
CAPILL A R Y FRINGE NON-PEDOGENIC CALCRETE
'-¥
., _...:;.. ��--- PHREATIC (VALLEY) OR
J=
c a rbonate movement
3. Classification of calcretes by hydrologic setting (based on Carlisle, 1980, 1983).
5
G R OUNDWATER NON- PEDOGENIC CALCRETE
Calcretes: An Introduction
in very near-surface settings; extensive and _thick calcretes do occur in arid areas due to precipitation within or j ust below the capillary fringe (Arakel & McConchie, 1982), with or without the influence of phreatophytic plants (Semeniuk & Meagher, 198 1), at a depth down to many metres, or tens of metres, below the surface. It should be stressed that such calcretes can be very difficult to differentiate from the more common pedogenic forms, and it is likely that at least some calcretes, interpreted as pedogenic forms in the geological record, are of the phreatic or capillary fringe type (see below, p . 8). A mineralogical division of carbonate duricrusts has been proposed by Netterberg (1980) (Table 3 ) . I t would b e possible t o devise classifications for gypsiferous and siliceous calcretes/dolocretes but there seems little need for this. Dolocretes can also grade into ankeritic forms and these are common in the early Carboniferous marginal marine limestones of Europe (Muchez & Viane, 1987; Searl, 1988) . These ferroan dolocretes form in coastal paludal settings, typically associated with coals (Wright & Robinson, 1988). Another means of classification is microstructure (Fig. 4) and two end-member types have been re cognized (Wright, 1990c) . Alpha calcretes, which correspond to K-fabrics of Gile et al. (1965) and Bal
Table 3. Classification of calcretes and dolocretes based on dolomite content. After Netterberg ( 1980).
Name
o/o dolomite by mass of total carbonates
Approx. equivalent o/o MgC03*
Calcrete Magnesian calcrete Dolomitic calcrete Calcitic dolocrete Dolocrete
<5 5-10 10-50 50-90 >90
<2 2-5 5-25 25-40 >40
* o/o MgCO3
MgC03 MgC03
+
CaC03
X
100.
( 1975 ) , consist of dense, continuous masses of mic ritic to microsparitic groundmasses, typically with such features as crystallaria (including circum granular types); floating, etched or exploded skeleton grains; large euhedral crystals (commonly rhombic); crystal size mottling and displacive growth features. Beta calcretes exhibit microfabrics dominated by biogenic features such as rhizocretions, needle-fibre calcites (lublinite), microbial tubes, alveolar septal fabric, and Microcodium. In such fabrics much of the carbonate was precipitated in association with fungi or other soil micro-organisms. The implications
BETA
ALPHA
1 Dense m i crofabric
1
2 Nodules
2 N eedle fibre calcite
3
3 Calcified tubules
Complex cra cks a nd c r y s tallaria
=
Microbial coati ngs
Microcodium
4 Circum-granular cra cks
4
5 Rhombic cal cite c r y s tals
5 Alveolar s e pt a l fabric
6 Floating s e d i m ent grains
6 Calcified pellets
6
Fig. 4. Micromorphological classification of calcretes (see text). Based on Wright ( 1990c).
Calcretes: An Introduction
of these two types of microfabric are discussed below.
(Atkinson, 1977) while in semi-arid and arid soils it ranges from 0.6-4% to less than 0. 1% respectively (Brook et al., 1983). The relatively low Pco2 in arid and semi-arid soils is a contributory factor leading to carbonate precipitation (Marion et al. , 1985 ) . Pre late Palaeozoic soils would have had lower bio masses than later ones and lacked roots. They would have had lower Pco2 than modern soils. Changes in Pco2 in the atmosphere through time may also have been important in influencing carbonate mobility and precipitation in palaeo-calcretes. Pco2 is now regarded as an important control on carbonate mineralogy in marine settings (see review in Tucker & Wright, 1990) and it is possible that similar in fluences on mineralogy operate or operated in calcretes. The common ion effect is another factor which is important in the precipitation of groundwater cal cretes, for example near playas. It is a contributory factor in some pedogenic calcretes (Reheis, 1987 ) . The role o f organic processes, other than i n in fluencing Pco2 , has been underestimated. Cyano bacteria in soils may induce carbonate precipitation by the uptake of C02 (Krumbein & Giele, 1979, this volume), while bacteria are widely suspected of having the potential to cause extensive precipitation in soils (Krumbein, 1968, 1979; Boquet et al. , 1973 ; Pentecost & Terry, 1989), as a result of chemolitho tropic removal of C02, the production of extra cellular bases such as ammonia, sulphate or nitrate reduction and by the provision of low-energy surface sites for crystal nucleation. Fungi are particularly important in triggering carbonate precipitation (Callot et al. , 1985 ; Phillips & Self, 1987), and this may reflect the dumping of excess Ca2 + by the micro-organism (Phillips et al., 1 987) . I n summary, there is a range of mechanisms causing carbonate precipitation. Evaporation and evapotranspiration, and to a lesser extent , degassing, will be climatically controlled, and, of course, climate will influence the degree of biological ac tivity. Although the use of C and 0 stable isotopes has gone some of the way in assessing the rates of these processes (see below, p. 20) , their effects, and those of microbial activity on profile and micro structure, are not well understood. Microbial mech anisms are clearly important in the formation of beta calcretes but much more work is needed before the links between calcrete morphology/micromorpho logy and climate/biology are clear enough for their use in palaeoenvironmental interpretation.
MECHANISMS OF CARBONATE PRECIPITATION
Relatively little detailed work has been carried out on the mechanisms of carbonate precipitation, and an understanding of these processes and their prod ucts would provide a very powerful tool for inter preting ancient calcretes in the geological record. While some authors have stressed evaporation/ evapotranspiration and degassing as the main mech anisms of precipitation (Salomons & Mook, 1986 ) , other types are also important (Fig. 5 ) . Whatever the source o f the dissolved carbonate (as bicarbonate) in the calcrete (see below, p. 8 ) , its solubility will b e decreased by the removal of H20, C02 and by the addition of Ca2 + (common ion effect) . Water can be removed by direct evaporation or by evapotranspiration (Cerling, 1984; Salomons & Mook, 1986). Evapotranspiration is regarded as a major process in many semi-arid calcretes (Cerling, 1984) and is probably a major cause of rhizocretion formation. C02 loss is another major process (Salomons & Mook, 1986). The partial pressure of C02 in soils is typically much higher than in the atmosphere. Atmospheric Pco2 averages about 0.03%o; in tem perate soils it can reach 1 1 .5%, but averages 0.9%
M EC H A N I S M S OF CaC03 P R ECIPITATION
Co mm on ion effect
Evapotranspiration Microbial activity cyanobacteria, bacteria, fungi etc.
Fig. 5. Mechanisms of precipitation in calcretes (see text for details).
7
Calcretes: An Introduction GROUNDWATER CALCRETES
SOURCES AND MOVEMENT OF CaC03 IN
Calcretes, dolocretes and gypcretes of non-pedogenic origin are common in present-day arid alluvial basins but have not been recognized in the stratigraphic record . Groundwater (syn. phreatic, valley, channel calcrete) calcrete can cement and replace/displace very large volumes of sediment; for example , there are cemented Plio-Pleistocene alluvial fan gravels in Oman (the Wahiba Sands area) which have been diagenetically altered (replaced) to dolomite clays, apparently to depths of over 200 m (Maizels, 1987) . In Australia, groundwater calcretes and dolocretes are commonly kilometres wide (maximum of 10 km) , tens of kilometres long (maximum of 100 km) and have an average thickness of 10 m (Mann & Deutscher, 1978; Mann & Horwitz, 1979; Arakel & McConchie, 1982; Carlisle, 1983 ; Arakel, 1986) . On a local scale these carbonates may be lensoid and locally thickened zones occur as mounds or domes which break surface . They form from carbonate rich, mobile groundwaters which become progres sively concentrated during down-dip flow (Fig. 6). The carbonate is precipitated mainly in the capillary fringe zone , directly above laterally moving subsur face water, but it can also be precipitated below the water-table. The precipitation of carbonate is trig gered by several factors: C02 degassing, evapor ation/evapotranspiration and the common ion effect. Cementation preferentially occurs at 'highs' where basement irregularities bring groundwaters near to the surface, facilitating degassing and ev aporation/evapotranspiration. Where Ca or Mg bi carbonate-bearing waters mix with Ca or Mg sulphate or chlorite-rich playa groundwaters, precipitation due to the common ion effect occurs. Preferential formation of groundwater calcrete/dolocrete also occurs where drainages converge , where flow gradi ents decrease, where saline waters mix, or where permeabilities are low. Groundwater carbonates (calcrete or dolocrete) are typically micritic and densely crystalline although the pore size range of the host sediment is import ant. The carbonate may contain authigenic silica, clays (sepiolite , palygorskite) and gypsum . Figure 6A shows an idealized profile through a groundwater calcrete (see also Jacobson et at., 1988; Arakel et al., 1989). The growth of the carbonate is both dis placive and replacive and, as a result, nodular to massive forms develop. Shrinkage cracks and dis solution features may be abundant. During pro gressive cementation the profile becomes plugged
CALCRETE FORMATION
One still commonly held misconception about cal cretes is that the bicarbonate is sourced from groundwater ('per ascensum' model of Goudie, 1973, 1983) . This certainly applies to groundwater calcretes but in many areas where calcretes are forming near the surface , the water-table may be many tens of metres below the land surface (e.g. Gile & Grossman, 1979 ) . Carbonate cementation related to the capillary fringe is strongly controlled by the grain size of the host material which affects the amount of capillary rise - only a few metres in clays and much less in sands. The sources of CaC03 are varied (Goudie, 1973, 1983) and include rainfall (and seaspray) , surface runoff, groundwater, dust , bioclasts (e .g. terrestrial gastropods) , vegetation litter and rock. The main source of CaC0 3 in pedogenic calcretes is wind blown dust (Machette, 1985 ) . Ca-rich dust accumu lates on the soil surface and is dissolved by rainwater. The carbonate is translocated down into the soil and precipitates, typically at the depth of seasonal wett ing. In the Las Cruces area of New Mexico recent 3 dust fall contains 0.2 g of CaC0 3 per cm2 per 10 yr (Gile & Grossman, 1979), although much higher rates existed in the same area in the Pleistocene (Machette, 1985 ) . In the same area the concentration of Ca2 + in rainfall is also high, perhaps exceeding 5 mg of Ca2 + per litre of water (Machette, 1985 ) . The mechanism of downward movement of dis solved CaC0 3 is referred to as the 'per descensum' model (Goudie, 1983) and most readily explains calcretes developed well above the water-table and on non-calcareous substrates. However, where abundant CaC03 is available from the substrate, calcrete may form from redistribution of carbonate , especially if it contains 'metastable' carbonate (e.g. aragonite) . Rabenhorst & Wilding (1986) have shown that calcretes of the Edwards Plateau, Texas, formed by the in situ dissolution and reprecipitation of host limestones, resulting in stages of develop ment like those of other calcretes . It is likely that such 'redistribution' calcretes form much more rapidly than the dust-dependent type. Palustrine associated calcretes represent a special type of 're distribution' calcrete (Platt, 1989, this volume).
8
A %
POROSITY
CARBONATE
w (/) 0 0 <( >
%
TOP SOIL
.S?
c Q) CJ) 0 "0 Q)
'?-
<J)
� a;
pisolitic on slopes
�������- LAMINAR
BRECCIATED
E
MASSIVE
'1
BRECCIATED
71-
12-
78%
15
t--- --
z
0 f=
() f= <( w a: I c..
"WATER TABLE•
4: ::J f () ::J _J LJ..
._;
�
52
w > (/) ----- (/) <( �
LJ..
0 LU z
t-- t---
MOTT LED
0
2128
48
N
2731
t-- 1--ALLUVIUM
THICKNESS
40-
. ..: . . . :.,; :, - _ :. - :.. - ·0
VARIABLE
0
•
0
•
•
•
•
•
• 0
45
B pediment EVAPORATION playa
calcrete-----'):;.
dolocrete
---7
dol/gypcrete
saline ground water
Ca/Mg decreases
Jl
increase in authigenic (Mg) clays
e_g_ sepiolite, palygorskite ( and corrensite
>
? )
Fig. 6. (A) Groundwater calcrete profile (based on Arakel, 1986 and other sources). Some workers (e.g. Carlisle, 1980) recognize two zones in the massive phreatic unit: an upper earthy zone with remnant soil and alluvium, and a lower 'porcellaneous', dense zone with abundant cracks and cavities. Laminar calcretes may also develop in the capillary fringe related to phreatophytic plants (Semeniuk & Meagher, 1981). (B) Evolution of groundwaters and their precipitates across an arid alluvial basin (see text). Based on Arakel (1986).
9
Calcretes: An Introduction
and groundwater flow will shift laterally, creating the wide, ribbon-like geometries. These cemented channel zones are responsible for inverted relief if erosion of the surrounding, less cemented, deposits occurs (Reeves, 1983). Marked mineralogical changes occur down-dip in these systems as the groundwaters evolve (Fig. 6B) (Arakel, 1986). The pre-Quaternary sedimentary record must contain many palaeo-groundwater calcretes/dolo cretes, but records are few (e.g. Tandon & Narayan, 198 1 , this volume). Simple nodular carbonates in ancient red bed sequences are usually interpreted as pedogenic calcrete/dolocrete without any real assessment of their likely origins. Early spar cements occur in many red bed sequences and these may represent groundwater carbonates. Much more work is needed to define the 'groundwater calcrete facies' and to devise criteria for its recognition. However, a few general points can be made about such carbonates. Groundwater carbonates do not usually form mature profiles comparable to those in pedogenic calcretes but tend to be nodular, massive or brecciated. There are exceptions, apparently, where phreatophytic plants form laminar and rhizo cretionary calcrete at the capillary fringe zone in dunes (Semeniuk & Meagher, 198 1 ) . It is likely, though, that most groundwater calcretes/dolocretes have formed well below the zone of biological activity and will more likely exhibit alpha fabrics (this does not mean that all alpha calcretes are groundwater in origin but reflect similar mechanisms for carbon ate precipitation; see below). Desiccation features and rhizobrecciation will be less common in ground water calcretes but relief inversion and large-scale doming will result in local zones of fracturing and weathering. These carbonates will preferentially occur in the coarser, more permeable lithologies and may exhibit greater thickness than pedogenic forms. Carbonate precipitation will occur over a relatively narrow depth range in pedogenic carbonates and individual horizons will be relatively small. In groundwater types one might expect thicker horizons showing much progressive deformation, such as those described by Mann & Horwitz (1979) from Australia. The characteristic carbonate profiles seen in pedogenic calcretes (Fig. 2) will probably not be produced in groundwater types. In addition many features of pedogenic calcretes, such as peloids and black pebbles, will not develop. In addition, silica replacement can be extensive in some groundwater calcretes/dolocretes. Very little information is avail-
able concerning the range of isotopic values in groundwater calcretes. It might be argued that such groundwaters should exhibit evolved, concentrated (heavy) 018 0 values and water-rock interactions may also lead to heavier values. However, if rapid infiltration of rainwater has occurred during storm events, the groundwater carbonates may exhibit lighter oxygen isotope values than pedogenic forms where strong evaporation may have occurred. It seems reasonable to assume heavier b13 C values in the groundwater situations unless light carbon is introduced from phreatophytic vegetation (see section on stable isotopes). In conclusion, while most calcretes and palaeo calcretes are pedogenic in origin, groundwater forms are, none the less, significant at the present time but they have yet to the recognized in the pre Quaternary record .
PEDOGENIC CALCRETE SOIL PROFILES
Pedogenic calcretes typically have distinctive pro files within the host profile. The accumulations of carbonate can be subdivided into soil horizons , de pending on their vertical arrangement, but they form a distinctive sub-profile in their own right. The typical nodular calcrete profiles (Stages 3-4) are strongly asymmetric with regard to the vertical con centration of carbonate. The upper part exhibits a sharp maximum in carbonate content which tails off down the profile. This general trend, however, is strongly climate dependent (Fig. 7). An important development in understanding the controls on and rates of carbonate profile development has come from numerical models (McFadden & Tinsley, 1985 ; Marion et a/ . , 1985 ; Mayer et a/ . , 1988 ) . These simulations were developed in part to assess the effects of climatic change on the depths of carbonate accumulation in calcrete-bearing soils in the south western United States. The effects of different climates (and Pco2 related to soil activity) on the style of carbonate accumulation are seen very clearly in Fig. 7. Applying this approach to the thousands of Phanerozoic palaeo-calcretes, as a guide to climatic controls, will be challenging because of the difficulty in identifying original soil depths and because many profiles appear to be cumulate in type. However, one of the most exciting aspects of this work is that distinctive polygenetic profiles can be modelled over Quaternary climatic changes. These profiles are 10
Calcretes: An Introduction
a
0
0
E
0. Q.> 0
40
18
100
24
60
0·5
80 100
c
12
�
.r::
0
6
20 u
b
0
0·1
0·5
0·3 CaC03
1·0
1·5 200
0·7 300
(g)
400
0
02
0·4
0·6
Fig. 7. Simulated patterns of pedogenic carbonate accumulation in three different Holocene climatic areas over a period of 13,000 years. In all cases the carbonate influx rate was 1.5 g/cm2 per 104 yr, with different Pco2 and leaching indices. (a) Model for a semi arid, thermic climate; compares well with Holocene soils in southern California. (b) Model for an arid, hyperthermic climate; compares well with profiles in the eastern Mojave Desert of California. (c) Model for a xeric, thermic climate; compares well with soils in the cooler, wetter, Mediterranean-type climates in southern California. This model predicts that small amounts of pedogenic carbonate could accumulate 5 m below the solum but these do not occur in the actual soil profiles. Based on McFadden & Tinsley (1985).
expansion (displacement) of the original soil com ponents. For example, the floating detrital grains commonly seen in Quaternary calcrete matrices re quire a 400-700% expansion of the original grain framework (Machette , 1985) . Displacive growth appears to be the dominant process in the formation of alpha-fabrics. Such features as grain fracturing and 'floating grains' are directly attributable to this process , yet are absent or weakly developed in beta calcrete fabrics. Not only calcite, but also dolomite and gypsum are associated with displacive growth in duricrusts, although silcretes and ferricretes do not exhibit these displacive features. In an attempt to explain this relationship between duricrust minera logy and fabric, Chadwick & Nettleton (1990) have suggested that the differences relate to the nature of adhesive-cohesive bonding types. For example, calcite crystals preferentially form cohesive bonds with other carbonate crystals, reflecting the chemical affinity of the ionic bonding of calcite. The inability of carbonates to form adhesive bonds with non carbonate grains results in the displacement of these (typically silicate) grains. Silica and iron (in silcretes
quite different in form from cumulate profiles (McFadden & Tinsley, 1985 ) , and this may provide a means of recognizing climatic changes in palaeo calcretes. Numerous authors have provided idealized calcrete profiles (see reviews by Klappa, 1983 , and Esteban & Klappa, 1983 ) . Profiles, of course, develop in stages (Fig. 2) which are widely identified in palaeo calcretes. The basic process involved is the progres sive accumulation of carbonate within the host soil. This accumulation involves several processes besides carbonate precipitation and cementation. Replace ment and displacement of the original host is re quired. Many calcretes exhibit clear evidence of replacement of silicates with etched grains but no estimates of the amount of dissolution have been made. The replacement of carbonate grains also takes place, finally culminating in a diagenetic wackestone or lime mudstone fabric (Read, 1974; Arakel, 1982). Displacive growth is a very important process. The amount of carbonate in mature profiles typically exceeds the original pore space of the host, which is best explained as the result of the physical 11
Calcretes: An Introduction
th.;u· marked lateral variability related to topo graphic, even micro-topographic variations (Sobecki & Wilding, 1982; Phillips & Milnes, 1988).
and ferricretes) favour adhesion, being in covalent bonding compounds. Displacive growth also leads to the formation of pseudo-anticlinal features (Watts, 1977) although the shrink-swell behaviour of smectitic clays in cal crete is probably equally important in causing such large-scale deformation . The importance of clay translocation and the for mation of argillic horizons in calcrete-bearing soils, particularly Aridisols, is still debated (Allen, 1985; Nettleton & Peterson, 1983 ) . The high Ca2 + content of such soils causes clays to flocculate, reducing their mobility. Illuviated clay coatings which form may be destroyed by wetting and drying cycles or by the displacive growth of carbonate. Many clay coats which occur in Quaternary carbonate-bearing soils may have formed through stress and are not illuvial in origin (Ranson & Bidwell, 1990) . Silica is an important component in many cal cretes, especially in older forms (Reeves, 1976). The position of the secondary silica concentrations in the profile relative to that of the secondary carbonate is dependent on a number of factors, but especially the composition of the host sediment/rock. Chadwick et al. ( 1987) have provided a detailed discussion of the behaviour of Si in duric soils. Gypsum is also com· monly found in calcrete profiles and its distribution is a useful guide to soil hydrology. Gypsum is more soluble than calcrete such that, in very low rainfall areas, only gypsum is translocated in the soil while carbonates added by dust are simply removed by the wind and no carbonate is moved into the solum. Its high solubility allows gypsum to be moved to lower depths in the solum and gypsic horizons typically underlie calcic ones. However, where updraw from groundwater occurs, gypsum may accumulate above the horizon of carbonate precipitation. In polygenetic profiles gypsum concentrations have strongly in fluenced subsequent carbonate movement because of the common ion effect (Reheis, 1987) . Brecciation i s a striking feature o f many calcretes, reflecting pedoturbation caused by the displacive growth of carbonate, wetting and drying cycles , thermal expansion, swelling clays and root activity. The last may be due to the physical disruption by root growth (Klappa, 1980b) or to tree heave (Semeniuk, 1986). There are, of course, other processes operating during the formation of calcrete profiles but those discussed here are the most important ones. Clay neoformation and recrystallization will be discussed below. One additional feature of calcrete profiles is
Rates of profile development
The rate of development of calcrete chronose quences (Fig. 2) is highly variable (Wright, 1990b ) , depending o n climate, Ca2 + supply, Pco2 and other factors. The well-documented calcretes of the south western United States have taken very long periods of time to form (Shlemon, 1978; Gile et a! . , 1981; Machette , 1985) , an order of magnitude longer than had been thought (see Leeder, 1975) . In contrast, Hay & Reeder ( 1978, this volume) have described calcretes from Olduvai Gorge in East Africa, which developed to Stage 4 in only a few thousand years. Similar calcretes require 100 ka or more in the south-western United States . Based on reliably dated Quaternary calcretes (and allowing for the fact that many of these have been through multiple climatic changes), the estimates for forming a mature (Stage 4) profile range from 3 ka to over 1 Ma, more than two orders of magnitude variation! These calcretes appear to correspond to alpha calcretes (see below, p. 20) , but little is known about the rates of formation of biogenic (beta) calcretes. It seems likely that the rates of formation of these beta forms are much more rapid. In Machette's classification (Fig. 2; Table 2), the thickness of the laminar horizon is used to distinguish Stages 4 and 5. Age differences between these stages can reach over 1 Ma in the south-western United States (Machette, 1985) and the rate of formation, even over shorter periods, is still slow (Robbin & Stipp, 1979) . Care needs to be exercised in extra polating these rates , if only because some laminar calcretes represent calcified root mats (Wright et al . , 1988, this volume ) , which probably formed rapidly. Information on the rates of formation of ground water calcretes is sparse but it appears that they can form much more rapidly than pedogenic forms (Jacobson et a/ . , 1988) .
MACROFEATURES IN CALCRETES
Several macroscopic features are common in cal cretes and warrant discussion. These include rhizo liths, nodules, laminar calcretes, pisoids and 'black pebbles' . 12
Calcretes: An Introduction
develops as mtcnte (commonly chalky at first) (Calvet et al. , 1975 ; Beier, 1987 , this volume) or as crude concentric coatings of micrite or microspar (Klappa, 1980a, this volume; Semeniuk & Meagher, 198 1 ; Mount & Cohen, 1984). Rhizoliths are commonly associated with Micro codium (Klappa, 1978, this volume) and may also exhibit secondary structures reflecting rootlets and roothairs around the main root. Most rhizoliths in calcretes are subvertical but extensive horizontal root mats form over indurated horizons or at water tables to form distinctive laminar calcretes (Wright et al. , 1988, this volume) . Extensive, vertical 'tap' roots are associated with phreatophytic calcretes (Cohen, 1982) . Roots are important in calcrete for mation in other respects. They act as conduits for water and can act as sites of net dissolution, and Jones ( 1 988) has reviewed the wider influences of root activity in calcrete formation. Micritization of carbonate grains is also a common feature around roots (Beier, 1987, this volume; Jones, 1988 ) . The role o f micro-organisms i n rhizolith formation is particularly important (Klappa, 1980a, this volume; Wright, 1986a, this volume; Beier, 1987 , this volume) and much of the carbonate of the
Rhizoliths are organo-sedimentary structures produced by roots and a variety of types has been recognized (Klappa, 1980a, this volume). The type of rhizolith which forms depends on the degree and location of calcification in and around the root and root void (Fig. 8). The simplest case is where the root has decayed leaving a root mould, which may be passively filled with sediment. The preservation of the mould may be enhanced by cementation of the host substrate , and in some cases this cementation results in the formation of a discrete cylinder (en velope) of cemented sediment around the root (Calvet et al . , 1975 ; Jones & Ng, 1988) . This envelope may consist of sparry to micritic calcite (Mount & Cohen, 1984) or of needle-fibre calcite, probably reflecting the site of an ectomycorrhizal sheath of fungi (Wright, 1986a, this volume ) . The actual root zone may become filled by pedogenic carbonate, either during the decay of the root, after decay, or even during the life of the plant. In the latter case the cellular structure of the root can be seen clearly in the resulting carbonate (Jaillard , 1 987; Jaillard & Callot, 1 987) , forming root petrifications in the sense of Klappa ( 1 980a, this volume) . Calcification associated with the dead or decaying root usually
Fig. 8. Styles of root calcification in rhizoliths. A root zone; B root envelope or tubule; C host substrate. In some rhizoliths the central root decays completely and the resulting cavity is filled by sediment. This diagram illustrates four styles of root zone calcification: petrification where cellular structure is preserved, representing calcification of the living root; concentric micrite/microspar calcification in a decaying root; chalky micrite and alveolar septal fabric representing fungal mycelia in root cavity. This pattern may be more complex as individual root moulds may be used by several generations of roots. Several styles of calcification of the root tubule are also known including cementation by microspar or spar cement, micritic cement or by microbial coatings and needle-fibre calcite. Micritized sediment grains may occur and grain packing is typically denser near the root. Grain dissolution may also have occurred. =
=
=
alveolar septal fabric
root petrification
concentnc micrite /microspar
13
Calcretes : An Introduction
micritization') resulted from the dissolution of spar and concomitant precipitation of micrite in the re sulting void space. The distinction between laminar calcretes and stromatolites was considered a problem (Read, 1976), but now it is necessary to distinguish not only these two types, but also biogenic (commonly micro bial) and abiogenic laminar calcretes (Wright, 1989) . Coated grains (pisoids) are a very common feature in calcretes. They can range in size from coated sand grains (Knox, 1977, this volume) to large clasts (Hay & Reeder, 1978, this volume ) . They are typically associated with very mature calcrete profiles where, following hardpan formation, the locus of carbonate precipitation occurs above the laminar layer on suit able nucleii, typically calcrete clasts. They form especially on slopes where the grains are moved downslope by gravity, becoming quite evenly coated en route (Read, 1974; Arakel, 1982). Grains which are not moved within the profile develop asymmetric coatings (Hay & Reeder, 1978, this volume), with preferential growth on the upper, or lower surfaces, the latter forming pendant coatings. The coatings appear to have two main origins . Some are biogenic, the coatings composed of micro bially mediated carbonate, with fungal tubules and needle-fib_re calcite (Knox, 1977, this volume; Calvet, 1982; Calvet & Julia, 1983; Wright, 1986a, this volume; Beier, 1987, this volume ) . Others appear to consist of simple micrite coatings with admixtures of non-carbonate material (Hay & Wiggins, 1980, this volume) . These authors, in de scribing coated grains associated with mature cal cretes from the south-western United States, inter pret the associated opal and sepiolite as having formed from gels on the grains. Hay & Reeder (1978, this volume) have interpreted coated grains in calcretes from Olduvai Gorge, East Africa, as micritic replacements of clay coatings, the latter having apparently formed by accretion at the soil surface and not as argillans, despite their waxy appearance. Black pebbles are a striking feature of many calcretes, subaerial exposure, peritidal and lacustrine deposits (e.g. Platt, 1989, this volume ) . They have been interpreted as reflecting the input of terrestrial organic matter (Strasser & Davaud, 1983; Strasser, 1984), burning in forest fires (Shinn & Lidz, 1988) or to the formation of finely disseminated pyrite (Wright, 1986b ) . Antiformal structures, including tepees, occur in many calcretes (Kendall & Warren, 1987 ) . They
rhizoliths has been produced by fungal mediation to produce the needle-fibre calcite and fungal cal careous tubules. Carbonate nodules (glaebules) are widespread in calcretes but their origins are poorly understood. They can be classified according to shape, or internal structure (e.g. concentric, uniform) (Brewer, 1964), and on their relationship to the soil matrix (Wieder & Yaalon, 1974) . In this latter classification, nodules with gradational margins into the matrix are re ferred to as orthic nodules, while disorthic nodules have sharp margins suggesting that some pedotur bation has occurred. Allorthic nodules are reworked from another horizon or soil. To what extent the abundance of disorthic nodules in calcretes and palaeo-calcretes reflects displacive processes or the inability of calcite to be adsorbed to silicates is unclear. A variety of suggestions has been offered as to why and how nodules form. Diffusion of carbonate to certain sites is a critical factor, followed by pre cipitation and displacive growth, for most nodules contain very little of the original matrix. One of the most common features in calcretes is circum-granular cracks (see below, p. 330), voids which form around a nodule as the matrix contracts during desiccation. A very simple explanation for nodule growth is that once some concentration of carbonate has formed, perhaps around an earlier carbonate-rich area, carbonate precipitation may occur preferentially in the (circum-nodular) voids around the nodule margin. Such relatively large voids will dry more quickly than the smaller ones in the matrix and will have lower Pco2, both factors promoting precipita tion (Chadwick et a/ . , 1987). Laminar calcretes typically develop by accretion on indurated substrates, such as bedrock or hardpan calcrete . They have a variety of origins (Wright, 1989) . Some are purely abiogenic in origin (e.g. Wright & Wilson, 1987) due to precipitation from waters ponded over an impermeable surface . Others are biogenic in origin and lichens (Klappa, 1979a), cyanobacteria (Krumbein & Giele, 1979, this volume) or root mats (Wright et a/. , 1988, this volume) contribute directly to their formation. However, Kahle ( 1977, this volume) has interpreted dense, laminated, micritic calcrete crusts coating the Pleistocene Miami Limestone of Florida as a result of the micritization of boch grains and spar cement in the underlying limestone. Endolithic fungi and algae were responsible for some of the replacement but the replacement of calcite spar (Kahle's 'spar14
Calcretes: An Introduction
may owe their origins to the displacive growth of carbonate within another substrate (Watts, 1977) . Rhizobrecciation (Kiappa, 1980b) is another pro cess, as is argillipedoturbation, whereby the seasonal swelling of smectites can disrupt the profile - a process associated with Vertisols which commonly exhibit calcic horizons .
TEXTURAL
I NVERSION
P E D OG E N I C
IN
CARBONATES
M ic r i t i c parent
G ra i n s t o n e p a r e n t c ir c u m g r a n u la r fractures
CALCRETE MICROMORPHOLO G Y m icritic coatings
The amount of published work relating to the micro morphology (petrography) of calcareous soils and calcretes is now enormous. Calcretes, by virtue of being typically indurated, lend themselves to such study and the identification of a number of diagnostic or characteristic calcrete microfabrics has allowed their recognition in the geological record . Important reviews on calcrete micromorphology have been provided by Braithwaite ( 1983) and Esteban & Klappa ( 1983) . One common feature seen in calcretes is textural inversion (Fig. 9). Finely crystalline parent materials (such as palustrine limestones or even pre-existing calcretes) can become microbrecciated to produce grainstones ('grainification', see above). In grain supported hosts calcretization results in secondary matrices forming diagenetic packstones which can then undergo grain-micritization to form lime mud stone fabrics (Read, 1974; Arakel, 1982) . At the simplest level two end-member microfabric types occur in calcretes (Fig. 4). Beta fabrics exhibit a variety of micro-scale features attributable to the existence and activities of macro- and micro-organ isms. The role of the soil biota in calcrete formation was carefully documented by Knox ( 1977, this volume), in a study of South African Quaternary calcrete. Other workers, particularly James ( 1972) and Harrison (1977), had also noted the importance of organisms in calcrete formation. Communities of fungi and cyanobacteria in par ticular are responsible for inducing carbonate pre cipitation in calcretes. Their role in forming some types of coated grains is now well established (Calvet, 1982; Calvet & Julia, 1983; Wright, 1 986a, this volume) but such precipitates occur throughout some calcrete profiles. A variety of calcified 'filaments' is typically found, coating and bridging grains (Knox, 1977, this volume; Klappa, 1979b; Calvet, 1982; Calvet & Julia, 1983; Beier, 1987, this volume; Phillips et al. , 1987). Such struc tures have even been recorded in Cambrian
peloid
m i critic coatings Diagenetic grainstone
Diagenetic packstone
Fig. 9. Textural inversion in calcretes. Dense lime mudstone fabrics (K-fabrics or alpha fabrics) develop in mature calcretes. This fabric may be fragmented to produce abundant peloids which become coated and can accumulate in fractures and other cavities (see Hay & Wiggins, 1980, this volume). On carbonate grainstones, calcrete formation results in diagenetic packstones (Read, 1974; Arakel, 1982; Knox, 1977, this volume).
phoscrete profiles (Southgate, 1986). Another feature commonly associated with such calcite fila ments is needle-fibre calcite which has been recorded from many Quaternary and fossil calcretes (e .g. Knox, 1977, this volume; Klappa, 1980a, this volume; Solomon & Walkden, 1985, this volume; Wright, 1986a, this volume; Beier, 1987, this volume ) . It consists of elongate needles of low Mg calcite, typi cally up to 10 f!m wide and 50- 100 f!m long, and is one of a wide variety of acicular carbonates occurring in calcretes (Calvet & Julia, 1983 ) . A variety of views has been offered as to their origins (see review in Wright, 1984, 1986a) with the two most popular being that they result from growth during extremely high degrees of supersaturation or that they are related to microbial activity, especially fungi . Callot et al. ( 1985) carefully documented their formation 15
Calcretes: An Introduction
type typically forms outgrowths to coated grains. The septa consist of parallel-oriented needle-fibre calcite (commonly neomorphosed) representing the sites of mycelial bundles. In some cases the septa appear to have resulted from the collapse of con centrically-coated rhizocretions (Wright & Wilson, 1987) . The common occurrence of this structure with root tubules led Wright ( 1986a) to suggest that the mycelial bundles were actually ectomycorrhizal sheaths on the roots. Some fungi form symbiotic associations either on root cells (ectomycorrhizae) or in them (endomycorrhizae). Goldstein ( 1J88) also considered the occurrence of needle-fibre calcite and rhizocretions to indicate a possible mycorrhizal association, but while needle-fibre calcite is rarely found in mycorrhizal sheaths it is commonly found in association with saprophytic basidiomycete fungi (G. Callot, pers. comm. , 1988) . Another biogenic structure abundant in many late Cretaceous- early Tertiary calcretes is the problem atic 'Microcodium' (Klappa, 1978, this volume). It consists of sheets, tubules and connected/uncon nected spheroids consisting of cell-like crystals of calcite. In some paleosols it is the only calcite form present and can constitute the whole calcrete profile (F. Calvet, pers. comm., 1988) . The laminar forms can be found not only coating carbonate substrates but have clearly replaced them via a sharp dissolution front (Bodergat, 1974) . Bodergat (1974) also noted very light 6 1 3 C values in Microcodium. Klappa ( 1978, this volume) interpreted it as a calcification product of a mycorrhizae-cortical root cell asso ciation, a view supported by the discovery of fila mentous structures inside the calcite cells. Freytet &
by the calcification of soil fungal hyphae. Later, Phillips & Self ( 1987) were able to show how the needles formed within fungal mycelial strands but were released by lysis (decomposition) of the strands by bacteria and distributed into the soil. Phillips & Self ( 1987) also showed how the needles were modified by epitaxial overgrowths. Once the fibres become released from the mycelial strands they undergo preferential crystal growth to form serrated-edged rods as a result of epitaxial growth onto ( 1014) faces . Such serrated forms have been widely documented in Quaternary calcretes (e.g. Jones & Ng, 1988) and even in Carboniferous forms (Wright, 1990c) . Even though needle-fibre calcite is ubiquitous in many present-day calcretes it is par ticularly susceptible to dissolution and overgrowth with progressive shape-loss or to neomorphism to micrite (Calvet & Julia, 1983). Solomon & Walkden ( 1985, this volume) found, using cathodolumine scence, that needle-fibre calcite was an important contributor to micrite in early Carboniferous calcretes. Needle-fibre calcite is the main component of alveolar septal structure (Wright, 1986a, this volume) . This was originally termed 'alveolar tex ture' by Esteban (1974) but was renamed in order to avoid confusion with 'alveolar structure', a term used in soil micromorphology to describe fenestral like fabrics. It consists of arcuate septa up to a few hundred microns long and up to 200 [,liD wide, within pore spaces, such as root moulds (Klappa, 1980a, this volume; Wright, 1986a, this volume) or in inter granular pore spaces (Adams, 1980, this volume; Calvet & Julia, 1983) (Fig. 10) . The intergranular
ALVEOLAR S EPTAL STRUCTURE Tubular to i rr e g u l a r p ores contai n i n g arcuate m i c r i t i c septae t y p i c a l l y under 1 0 0 1-J m wide .
•l .
.
. .
.
: :: >> .
-
.
i nt e r g r a n u l a r
Fig. 10. Occurrence and composition of alveolar septal structure (see text)
16
Calcretes: An Introduction
'mottling' reflecting patches with different crystal sizes. Some of these mottles define earlier glaebtJ.les set within, usually, a more coarsely crystalline mass. Many glaebules are defined by circum-granular cracks filled with spar cement. Floating grains , typically of silicates, are common. These may show evidence of grain expansion, such as exploded micas (Watts, 1978) , or fracturing (Buczynski & Chafetz, 1987, this volume) or etching. This 'floating' fabric reflects both grain replacements and displacements (Tandon & Friend, 1989, this volume). Displacive growth is a major process in calcrete formation and results in other features such as multi-directional growth in the crystals (Saigal & Walton, 1988; Tandon & Friend, 1989, this volume; B raithwaite , 1989; Wright & Peeters, 1989), linear growth (Braith waite, 1989 ) , and impeded growth (in the sense of Maliva, 1989 ) , all seen under cathodoluminescence. Alpha fabrics exhibit a variety of calcite-filled cracks (crystallaria) of variable shape, size and orientation. These are mainly due to desiccation and expansive growth, followed by rapid precipitation in larger pores where evaporation and degassing effects would be more enhanced. Alpha fabrics also exhibit intercalary growths, typically large calcite rhombs (Folk, 197 1 ; Lattman & Simonberg, 197 1 ; Chafetz & Butler, 1980; Wright, 1982). Alpha and beta fabrics are end-members and many calcretes exhibit mixtures of the two. How ever, many calcretes exhibit either only alpha or beta fabrics and there appears to be some climatic control on their occurrence (Wright, 1990c). Beta calcretes appear to be best developed in semi-arid to subhumid areas with extensive vegetation cover, the biofabric of the pedogenic carbonate seeming to reflect this relatively high degree of biological activity. Alpha calcretes occur in areas with a more arid climate and less biological activity. The actual dif ference must reflect the mechanism of carbonate precipitation. In beta calcretes biological induction is probably a major factor, but evaporation/evapo transpiration and degassing might be the main mechanism in alpha calcretes. However, these ideas are not easy to substantiate . Present-day alpha cal cretes often lack any obvious evidence of a vegetation cover, such as rhizoliths, even though there might be a 50% vegetation cover above them. Alpha calcretes studied by the author commonly exhibit a light tP C signature suggesting a major input of light carbon in the system even though no biologically-related fabrics occur. The difference between these two
Plaziat ( 1982) have described Microcodium from the late Cretaceous- early Tertiary of Languedoc, France, and found Klappa's interpretation unable to explain the laminar forms where no clear root asso ciation could be seen. To add to the problem Jaillard ( 1987 ) , Jaillard & Callot ( 1987) and Callot & Jaillard ( 1987) have described the intracellular calcification of living root cells producing structures remarkably like 'classic' Microcodium. Jaillard (1987) did not find evidence for microbial tubes inside the calcified cells, arguing against a mycorrhizal origin, a view •) supported by studies by the author and F. Calvet on early Tertiary forms from the Pyrenees. The unusual rock-dissolving ability of fossil Microcodium has been noted in the living roots which form Micro codium-like rhizomorphs (Jaillard & Callot, 1987). The problem remains unresolved but it seems likely that some Microcodium represent a calcification product of roots. The unusual stratigraphic distri bution of Microcodium, with its peak of stratigraphic abundance in the early Tertiary, and later in the Miocene (F. Calvet, pers. comm. , 1988) is signi ficant, and may represent some specific vegetation type. Goldstein ( 1988) has described Microcodium like features from the late Pennsylvanian of New Mexico. However, records of even older material do not resemble classical Microcodium and are invari ably not associated with calcretes. Peloids, sand-silt sized micritic grains , are a very common feature of many calcretes and are closely related to some types of glaebule. Like their marine counterparts they are polygenetic. Some result from desiccation processes whereby fragments of the cal crete matrix are separated and may be transported , dissolved and rounded in the profile (Freytet & Plaziat, 1979; Hay & Wiggins , 1980, this volume; Platt, 1989, this volume; Wright, 1990a). Others may be small glaebules (Harrison , 1977) , but micrit ization is a major process in their formation (e.g. James, 1972). In addition, many coated grains in calcretes have a crudely pelleted fabric and may be microbial in origin (Calvet & Julia, 1983) , perhaps formed by fungi. However, many peloids in calcretes represent faecal pellets (Wright, 1983 , this volume; Jones & Squair, 1989). Micritic calcretes can be totally reworked to form detrital peloids (Sarkar, 1988). Alpha fabrics (Fig. 4) consist of cystalline matrices (groundmasses) having a crystic plasmic fabric in the sense of Brewer (1964) , or K-fabrics in the sense of Gile et al. (1965) and Bal ( 1975). The crystal size ranges from micrite to spar and a common feature is 17
·
Calcretes: An Introduction
fabric types might represent some sort of 'tapho nomic' factor whereby biogenic features have a lower preservation potential in alpha calcrete environ ments . In this respect, recrystallization might be a factor but the nature of the host in which the calcrete develops is probably the most critical factor. Alpha calcretes almost invariably occur in silicate-rich , and beta calcretes in carbonate-rich hosts. In the case of the former the fabric is largely a function of the displacive growth of calcite which overprints other processes. This may be a function of the preference of calcites to form cohesive bonds with other calcites and to displace other grains (Chadwick & Nettleton, 1990) . Beta calcretes do not show such fabrics and reflect the 'passive' cementation of carbonate host grains. This cannot be the whole story and it does not explain the consistent coarseness of crystal size in many alpha-calcretes or the complete absence of biogenically induced carbonates in many such calcretes - other factors must also be operating. Alpha fabrics typically exhibit irregular crystal mosaics resembling classical neomorphic microspars and spars (e.g. see Tandon & Friend , 1989 , this volume). Many authors have interpreted these fabrics as evidence of the replacement of finer crys tals by coarser ones (Sehgal & Stoops, 1972; Tandon & Narayan, 198 1 , this volume; Wieder & Yaalon, 1982; Sobecki & Wilding, 1983 ; Rabenhorst et al., 1984; Rabenhorst & Wilding, 1986; Drees & Wild ing, 1987). However, few studies have accurately documented evidence for this neomorphism. Tandon & Friend ( 1989, this volume) describe mosaics with rhombic calcite which locally show relic luminescence patterns of 'ghost' micrite, especially clear in areas between masses of spar and micrite. This might be inter preted as clear evidence of aggrading neomorphism but many of the rhombs they illustrate lack relic micrite luminescence and it is unclear how such crystals were later able to purge themselves of these relics, while still maintaining their delicate growth zoning. One other possibility is that these inclusion rich areas were trapped during poikilotopic growth . To what extent this neomorphism i s a result o f the replacement of metastable carbonates, such as high magnesian calcite or aragonite by low-magnesian calcite, is unclear. A commonly discussed process is the replacement of micritic grade carbonate by microspar or spar by dissolution/reprecipitation (Sehgal & Stoops, 1972; Sobecki & Wilding, 1983 ) . Small crystals are less stable than larger ones, which have lower free energies. Soil carbonates form in the
zone of wetting and drying and fine crystal sizes will be prone to dissolution. Solomon & Walkden ( 1985 , this volume) and Wright & Peeters ( 1989) noted Carboniferous calcretes which, under cathodo luminescence, showed extensive evidence of dis solution events in calcites and such phases would have preferentially removed the more soluble micrites. The role of this 'solution cannibalization' requires more detailed study. Clay appears to play a significant role in influencing the stability of fine-grained carbonate. Wieder & Yaalon (1974, 1982) noted that the crystal size of the pedogenic carbonate is controlled by the non-car bonate fraction; clays provide smaller pores and possibly more nucleation sites and are associated with micrites, while microspars and spars are asso ciated with coarse materials. They have argued (Wieder & Yaalon, 1974) that high clay contents can retard or prevent recrystallization of the associated micrites. The relationship between recrystallization and clays has also been discussed by Nettleton & Peterson ( 1983) . They pointed out that calcite precipitated in the presence of soil clays has a higher dissolution rate than pure calcite, especially adjacent to clay colloids. They argued that small calcite crystals near clays, because of their dissolution rates, are kept small, while other crystals farther from clays are less easily dissolved and grow larger. The idea that, despite the susceptibility of these small crystals to dissolution, they remain small seems rather odd but the effects of clay in retarding neomorphism in micritic lime stones is well documented (Longman, 1977). Other factors besides neomorphism can create irregular calcite mosaics in calcretes. Multiple phases of growth and dissolution can create irregular crystal shapes (Wright & Peeters, 1989 ) , displacive growth can create non-uniform growth (Saigal & Walton, 1988), as does restricted displacive growth (Maliva, 1989). Many irregular calcite mosaics, under catho doluminescence , reveal phases of fine scale fractur ing and cement-filling, resulting in highly complex micrite to spar relationships . The occurrence of an hedral to subhedral fabrics might also be a response to crystal growth at supersaturations above the critical saturation level, or above the critical rough ening temperature (apparently only 25 oc for calcite; Tucker & Wright, 1990) . Such effects would pro mote the random addition of crystal faces and have been invoked to explain some irregular mosaics in dolomites (Gregg & Sibley, 1984) . In summary, neomorphic-like mosaics are very 18
Calcretes: An Introduction
common in alpha calcretes but they 'may not all be simply the product of the 'cannibalization' of micrite to microspar and spar by dissolutiqn/.precipitation processes. Other possible causes for irregular crystal mosaics must be considered and ca�hodolumine scence microscopy is essential to decipher the growth ' history of the crystals in a mosaic;
carbonate crystal form is often quoted, Chadwick found little evidence for its role in controlling crystal morphology in carbonate coatings from the Mojave Desert region. In their detailed study of crystal forms they noted that equant to parallel prismatic crystals , with irregular interlocking boundaries, were more common in soils with elevated amounts of common salts, small amounts of organic matter and short periods when the soils were moist. In contrast, small amounts of soluble salts, larger amounts of organic matter and longer moist periods were associated with randomly oriented, euhedral prismatic to fibrous crystals. There is a crude re lationship between these results and the idea that alpha calcretes occur in more arid and less biologi cally active soils, while beta calcretes do commonly exhibit prismatic to fibrous crystals. Climate, in con trolling the amount of water moving through the solum and hence the concentration of salts, degree of biological activity and organic matter, must be a control on calcrete microstructure. There is now a real need for experimental work on calcrete micro structures but, as yet, making environmental inter pretations based on carbonate microstructure may be premature; for example Drees & Wilding ( 1987) have speculated that relic microsparite nodules in the soils of the Rolling Plains of Texas reflect wetter periods. One potentially important use of crystal morph ology is to distinguish groundwater from pedogenic calcrete. Raghavan & Courty ( 1987) suggested that sparry calcites (crystal size > 40 f.tm) in Quaternary sediments and soils of the Thar Desert of India et al. ( 1989)
Controls on crystal morphology in calcretes·
The morphology (shape and size) of carbonate crystals is highly variable in calcretes and · a Q \lmber of authors have attempted to id�ntify specifi � factors (Fig. 1 1) . Matrix composition (spe�ifically texture/ grain size) is widely regarded as an impor�a!j� control (Wieder & Yaalon, 1974, 1982). CrystaJ growth can be affected and restricted by various factors· �uch as the incorporation of less favourable ions into the 3 lattice (e.g. Mg, Mn) and by infiu�nces of P04 and N0 3 - (Chadwick et a! . , 1 989) . Coatings on crystals, of clays, organic matter and sesquioxides, may also play a role, as does the rate of precipitation. Ducloux et al. ( 1 984) have emphasized the role of amorphous silicates in influencing crystal form and Nettleton & Peterson ( 1983 , see above , p. 18) have also stressed the role of clays. Biological in�uences may be 'direct' with regard to the calcifica�ion of root cells or microbial filaments, or 'indirect' in the influence of organic coatings and compo.unds in .crys tal growth. Ducloux & Dupuis ( 1 987) have shown, experimentally, how various organic'c6!)1pounds ean influence the composition and stability of pedogenic carbonates. Although the role of Mg2 + in influencing
- - - C LI M AT
�: � i� , 1
//
__.... ..--
duration of
r tal surface interactions with ganic m atter
BIO L IC G I NFLUENCE
�
�
l I
E
wetting phase
of calcite crystals in calcretes.
,
�e of precipitation
\
CRY S TAL S I Z E
/
a MORPHO LO GY
��
icrobial calcification
Fig. 11. Controls on the morphology
__
crystal surface interactions with inorganic particulate matter � MATRIX
19
�
CHEMI S TRY
' lattice poisoning
soil texture (grain-size)
/
/
Calcretes: An Introduction
precipitated from raised groundwaters during pluvial phases. They have argued that more finely crystal line calcites, typically showing dissolution -repre cipitation cycles, reflect pedogenic processes under semi-arid or arid conditions (see also Courty et al. , 1987). However, crystal size alone is not a reliable indicator of the site of precipitation (Wright & Peeters, 1989 ) . In conclusion, calcrete microstructures are diverse and it may be possible to use them in palaeo-calcretes to assess such factors as biological activity and cli mate. However, much more work is required to elu cidate the controls on microstructure and the time for experimentation has now arrived. One avenue of investigation might be to compare pre-late Pal aeozoic calcretes with younger ones to evalute the role of biological activity on microstructure .
Many organisms break down the oxalate, meta·· bolically, particularly bacteria, allowing Ca2 + to form bicarbonate and carbonate (Cromack et al. , 1977) , and it is an important aspect of Ca mobility in soils. Clay minerals are significant components in many calcretes. Smectite is one of the most common and is seemingly detrital in many cases. Many soils in semi·· arid areas contain concentrations of illite in the upper horizons (Nettleton & Peterson, 1983; Singer, 1988) . It can be produced in these upper layers by the alteration of smectites, by K + being fixed in the smectite , initially producing mixed layer illite/sme·· ectite and finally illite. The K+ is derived from the weathering of dust in the surface horizons and from aerosols , particularly near coastal areas. This pedo·· genic illitization can be difficult to recognize in deeply buried paleosols where burial illitization has taken place. However, Robinson & Wright ( 1987) were able to recognize it in Carboniferous calcrete-bearing Vertisols which had only undergone shallow buriaL Mg-rich clays , sepiolite and palygorskite (atta·· pulgite) are common in many calcretes (e.g. Watts , 1980, th_is volume; Hay & Wiggins, 1980, this volume) , and especially dolomitic calcretes or dolo·· cretes (Hu!ton & Dixon, 1981 ) . A range of explan- ations has been offered for their occurrence in soils , tabulated i n McGrath & Hawley ( 1987). Sepiolite is common in· mature calcretes and both minerals are climatically sensitive and useful for palaeoclimatic reconstruction. One unusual feature of some cal· cretes is the inverse relationship between the mole % MgC03 of the calcite and the occurrence of Mg clays (Watts, 1980, this volume) . Calvo et al. ( 1 986) , studying Miocene calcretes and dolocretes in the Madrid. basin, noted pedogenic sepiolite in the cal cretes but none in the associated dolocretes. Watts suggested that the Mg was released from high-mag nesian calcites in the calcrete during the conversion to LMC, enriching the soil waters and leading to sepiolite neoformation. Alternatively, the prefer ential removal of CaC03 as LMC might have led , locally, to high levels of Mg (Calvo et al. , 1986), promoting s·epiolite formation (McGrath & Hawley ,
CALCRETE MINERALO G Y A N D CHEMISTRY
In comparison with many soil mineralogical features, soil carbonates have received less attention. The work of Watts ( 1980, this volume) is particularly important and stands as one of the very few detailed mineralogical studies. While most calcretes consist of low-magnesian calcite (LMC) , the 'stable' form of CaC03 in meteoric water-related settings, Watts has noted high-magnesian calcite and aragonite as well as dolomite in calcretes from southern Africa. High Mg/Ca and C02 concentrations (and rate of supply of C032 -) are likely to be the controlling factors on the mineral phases formed. It is perhaps surprising that present-day calcretes exhibit such predominance of LMC considering the highly com plex processes at work in them. Dolomite is quite common in many calcretes and is a primary pre cipitate in many, if not most, cases (Hay & Reeder, 1978, this volume; Watts, 1980, this volume) . It may occur preferentially in the lower parts of profiles (Hutton & Dixon, 1981 ; Milnes & Hutton, 1983; Phillips & Milnes, 1988) . Calcium oxalate (as weddelite and whewellite) is common in many soils (Cromack et al. , 1977 ) . It is a major reservoir of Ca2 + in the soil and also affects the pH of the soil solution, as well as increases the solubility of Fe and AI (Graustein et al. , 1977 ) . Fungi are capable o f producing calcium oxalate and it is also found in higher plants. It is particularly important as a means of retaining Ca2 + in the rooting zone and it may play a part in rhizolith formation.
1987) .
Analcime occurs in some calcretes (Nettleton & Peterson, 1983) , reflecting locally high water-tables rich in .sodium .carbonate. Stable isotopes in calcretes
Carbon and oxygen isotopes in soil/paleosol carbon ates provide a powerful tool for palaeoenvironmental
20
Calcretes: An Introduction
interpretation, especially for palaeoclimatic and palaeovegetation studies. They are also the main means of assessing the mechanisms of carbonate precipitation. They have been used in palaeoenviron mental studies of Cainozoic (Magaritz, 1986; Cerling & Hay, 1986; Goodfriend & Magaritz, 1988; Quade et al. , 1989a Cerling et al., 1989) and Mesozoic paleosol carbonates (Suchecki et al. , 1988; Naylor et al. , 1989). Interpreting palaeo-calcrete isotope values is particularly difficult because of the many inter related controls on the isotopic composition. In studies of present-day calcretes, rainfall and veg etation composition can be assessed , but in paleo sols even these essential parameters will not be known. Talma & Netterberg (1983) have provided a compilation of soil carbonate isotopic compositions, and reviews of the processes controlling these com positions have been given by Salomons et al. ( 1978), Salomons & Mook ( 1986) and Cerling (1984) . The range of influences on soil carbonate isotopic com position is large , including elevation , latitude, temperature (including the degree of freezing) , the type and amount of vegetation cover, its seasonal activity, rainfall composition, the degree of evap oration, proximity to the sea , monsoonal effects and seasonal variations in rainfall isotopic composition, degree of contamination from pre-existing soil carbonates, and many other factors. Some of the mechanisms of precipitation , such as degassing and evaporation, also control the values, as does the degree of mixing with atmospheric C02 . Many mature calcretes represent development over long periods of time ( 1 05 to 106 years in some cases) and the carbonate in a profile may reflect several different geochemical/palaeoenvironmental regimes (Drever et al. , 1987). There are vertical profile variations in the isotopic composition of carbonate in the soil, reflecting mixing and diffusion with the atmosphere , degassing and evaporation (Drever et al., 1987 ; Quade et al. , 1989b) . This vertical gradient will change in response to changes in evaporation budget and especially as the soil aggrades or degrades. Mature calcretes are likely, therefore , to have highly varied isotopic compositions, and bulk sampling, as in any analysis of carbonate rocks, is likely to yield time-averaged values. In addition , some palaeo-cal cretes, originally containing aragonite and high magnesian calcite , may be converted to low-mag nesian calcite and could have isotopic compositions reflecting a later diagenetic overprint. The ()13 C values may be influenced by contami-
nation from pre-existing carbonate (Amundson et al. , 1989; Rabenhorst et al. , 1984) but most of the car bon is derived from soil C02 , produced by root respiration and microbial oxidation of organic ma terial. The isotopic composition of this soil C02 can be used to assess the type (Quade et al. , 1989a) and degree of vegetation cover (Amundson et al. , 1988). Fractionation occurs during the precipitation of carbonate and it is heavier than soil C02 by as much as +9.6%o (Friedman & O'Neil, 1977; Schlesinger, 1985 ) . In addition, soil C02 is heavier than the associated vegetation (perhaps by as much as 8-9%o; Salomons & Mook, 1986). As a result, negative () 13 C values - as low as - 10 - indicate high input of 1 2C from soil respiration and typically correlate with a vegetation cover dominated by C3 plants (Salomons & Mook, 1986; Quade et al. , 1989a) . Less negative values are more common (Gardner, 1984; Parada et a!. , 1983) and even positive ones occur. Such values can reflect contamination from pre-existing carbon ate in the substrate (although this is usually of minor importance) , a lower rate of net respiration in the soil, a vegetation type with a less negative ()13 C value (e.g. C4 or CAM type vegetation; Salomons & Mook, 1986), or mixing with atmospheric C02 . Kinetic and Rayleigh distillation effects can also result in shifts toward heavier values but these may not be important in all soils (Quade et al. , 1989b) . These are linked to evaporation and degassing but evapotranspiration appears to be the major 'drive' for precipitation in many calcretes, but not the only one recognized using stable isotope geochemistry (Drever et al. , 1987). Mixing of soil respiration C02 and atmospheric C02 is a major cause of more positive ()13 C values. This may reflect the seasonal nature of plant activity during precipitation of carbonate (Schlesinger, 1985 ) . () 18 0 is influenced primarily by the composition of the local meteoric water (Cerling, 1 984) . Evapo transpiration does not cause its fractionation, but evaporation in particular does cause marked changes which can be seen in the heavier values in the upper parts of some profiles (e.g. Drever et at. 1987). In interpretation , it is not always justifiable to take such an obvious trend at face value; Quade et al. ( 1 989b) have suggested that in actively forming cal cretes in the Great Basin in the United States, the trend toward lighter values deeper in the profile reflects the preferential infiltration of isotopically . heavier summer rain. Some workers have attempted to assess palaeo. 21
Calcretes : An Introduction
temperatures using 1'> 1 8 0 values in palaeo-calcretes. Suchecki et al. ( 1 988) and Naylor et al. ( 1989) have derived palaeotemperatures for early Mesozoic calcretes. Such attempts should be made with great care and sets of values from profiles not from single 'spot' samples are needed to determine the range of variation. Stable isotopes might prove useful for differenti ating groundwater from pedogenic calcretes. How ever, little data on the former is available (Manze & Brunnaker, 1977; Jacobson et al. , 1988 ) . One might expect the carbonates to show a greater enrich ment in 1 80 from more 'evolved' groundwaters and less 12C might be expected as the carbonate forms below the main zone of root activity and organic decomposition. The most important use of stable isotopes might be in determining the mechanisms of precipitation in ancient (and present-day) calcretes. Evaporation causes co-variation in C and 0 in soil carbonate, both becoming enriched in the heavier isotope up the profile (Salomons et al. , 1978) . Degassing does not fractionate oxygen (Salomons et al. , 1978) but does cause an increase in the heavier isotope of carbon in the soil carbonate. Evapotranspiration does not cause any fractionation (Salomons & Mook , 1986). These trends can be used to determine pre cipitation mechanisms and changes in those mech anisms through a polygenetic profile (Drever et al. , 1987). However, it is more likely that more than one process operates during the long development of mature profiles. The problem with stable isotope analyses is that the technique generates 'real' num bers, and with so many processes operating in cal cretes, it will usually be possible to find one or two
to explain the data. However, it is often very difficult to resolve the main processes operating in present day calcretes, when so many of the important vari ables can be determined, so what chance do we have in, for example, Palaeozoic calcretes with radically different vegetation types and where rainwater com position will never be known? No matter how peer impressing the technique may seem, the interpret ations derived from such isotopic analyses must be supported by sedimentological, palaeontological or other geochemical data. One gap in our knowledge is on the isotopic compositions of beta calcretes, and what effect bio induction by microbes has on the isotopic values. There is an urgent need for the isotopic compositions of such calcretes to be studied.
CONCLUSIONS
Calcretes, especially pedogenic types, are currently forming over a significant part of the Earth's surface, and were equally abundant in the past. They have the potential to provide diverse information on time resolution , palaeovegetation and palaeoclimates. Their profile forms alone can be used as a means of assessing relative sedimentation rates and possibly palaeoclimates. Microstructure, mineralogy and isotope geochemistry have enormous potential for palaeoenvironmental analysis. The sheer abundance of calcretes in the geological record will eventually provide us with new insights into the Earth's past. This volume contains many of the key studies on calcretes, both Quaternary and pre-Quaternary.
22
QUATERNARY CALCRETES
The following three papers cover a wide range of
rate of calcium carbonate from carbonatite ash.
calcretes formed in continental settings which exhibit
Hay & Wiggins provide one of the few descriptions
alpha-type fabrics. Hay & Reeder describe calcretes
of a very mature calcrete from the south-western
from east Africa deposited on volcanic ash. Besides
United States. It serves to illustrate how mature
the detailed descriptions of mature profiles, including
calcretes, with a dense matrix, can undergo grain
large coated clasts, they provide an example of
ification to produce large volumes of peloids, which
coated grain formation and replacive textures. The
accumulate in fractures and other cavities. These
calcretes they describe also contain dolomite. One
peloids form, in part, as a result of desiccation
of the most remarkable features of these calcretes is
forming circum-granular cracks surrounding sub
their rapid formation, having reached maturity (up
spherical areas of cemented micrite. These peloids
to an early Stage 6 profile ) in only a few thousand
are coated by micrite and clays to form coated
years. This is a consequence of the very high supply
grains. Sepiolite occurs in these calcretes and is
Fig. 12. Quaternary calcrete profile from Morocco showing elongate glaebules in lower part and laminar calcrete from calcification of root mats in upper part, with some brecciation as a result of tree-root heave.
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
23
Quaternary Calcretes pedogenic in origin but the source of Mg2+, as in
It is disappointing that few of the ideas offered by
many calcretes, is unclear. The Mg2+ content of the
Watts have been tested in other calcretes. Watts also stresses the role of displacive growth in calcrete
associated calcretes is low, 3-5 mole% MgC03. Watts provides an important study of the min
formation.
eralogy of Quaternary calcretes from the Kalahari.
A striking feature of these three studies is the
The calcrete contains not only LMC but high-Mg
absence of biogenic features. Watts notes the rarity
calcite (HMC), aragonite and dolomite. Watts sug
of organic structures such as rhizocretions, root
gests that slow evaporation and/or C02 loss causes
hairs and fungaUalgal filaments. However, the case
LMC to precipitate, whereas rapid evaporation and/
studies illustrate the major processes and features of
or C02 loss results in HMC, and minor aragonite. Mg concentrated during LMC precipitation, and
alpha-type
released during the conversion of HMC to LMC,
peloids and coated grains derived from the matrix,
leads to the formation of Mg-rich clays and dolomite.
and authigenic silicates.
replacive
24
calcretes: growth,
cementation,
dense
micritic
displacive/
groundmasses,
Reprinted from Sedimentology (1978) 25 649-673
·Calcretes of Olduvai Gorge and the Ndolanya Beds of northern Tanzania
R . L . H AY
and
R.
J.
REEDER
Department of Geology and Geophysics, University of California, Berkeley, California, U.S.A.
ABSTRACT Pedogenic calcretes are clo sely a ssociated with Pliocene to Holocene w ind worked deposits of volca nic ash in the Olduvai and Ndolanya Beds of northern Tanzania . The typical profile with calcrete consists of a n unconsolidated sediment layer, an underly ing laminar calcrete, and a lowermost massive calcrete. The laminar calcrete is a relatively pure limesto ne, w hereas massive calcrete is aeolian tuff cemented and replaced by calc ite. An Olduva i calcrete profile can develo p to a mature stage in only a few tho usand years . Car bonatite ash was the dominant source for mo st of the calcite in the calcretes. Replacement was a major process in for mation of the massive calcretes, a nd oolitic textures have resulted fro m micr ite r eplacing pelleto id clay coatings around sand grains-. Phillipsite and possible o ther zeolites were extensively r eplaced i n the massive calcretes. Replacement of c lay by micrite in the Olduvai calcretes is accom panied by dissolutio n or leaching of phengitic illite and the formatio n of clay a pproaching the compositio n of halloysite or kaolinite. In the upper calcrete of the Ndolanya Beds, mo ntmor illonite was a ltered to a kaolinite-type mineral and to dioctahedral chlor ite. Authigenic dolomite, zeolite, a nd dawsonite i n the Olduvai calcretes probably received a t least some of their components fro m r eplaced materials.
INTRODUCTION Scope and purpose Until recently, calcretes (or caliches) have received only limited attention as sedimentary rocks and have typically been treated as local, special occurrences. They are, however, widespread in semi-arid climates, particularly in tropical and sub tropical latitudes (e.g. Kalahari, Africa; Pampas, Argentina; lowlands of Venezuela; and much of Australia). Their distribution is discussed by Goudie (1972, 1973). In East Africa, calcrete is well developed in the drier areas in and adjacent to the Eastern Rift Valley, as for example, Magadi, Kisumu, Senmgeti Plain, Northern Frontier District of Kenya, and the Wajir Basin (Goudie, 1973). Much is now known about the field occurrence and origin of calcrete (Goudie, 1972, 1973; Reeves, 1976), but relatively little has been written about the petrographic, Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
25
R. L. Hay and R. J. Reeder
mineralogic, and chemical aspects. Most of the recent, more detailed petrographic descriptions are of calcretes developed in pre-existing carbonate deposits in coastal areas (e.g. Multer & Hoffmeister, 1968; James, 1972; .Siesser, 1973; Read, 1974; Scholle & Kinsman, 1974; Ward, 1975; Kahle, 1977). The purpose of this contribution is to provide petrographic, mineralogic, and chemical information about the calcretes of Laetolil and the Olduvai region of the eastern Serengeti Plain, in northern Tanzania, which are inland savannah areas with only small amounts of carbonate bedrock. Our emphasis is on the origin of oolitic textures by micrite replacement of pre-existing clay coatings on sand grains, and on mineralogical and chemical changes in the clay fraction associated with replacement by micrite. Terminology and methods We use the term 'calcrete' as defined by Netterberg ( I 969, p. 88): 'any terrestrial material which has been cemented and/or replaced by dominantly CaC03 . . . cave deposits excluded . . . the mechanism of calcification is not restricted and calcretes may be of pedogenic or non-pedogenic origin, or both'. The calcretes we describe herein are developed in lithified deposits of aeolian tuff, and with a few exceptions they are purely pedogenic in origin. 'Calcite' refers to low-Mg calcite ( �4 mol percent MgC03), whereas micrite, microspar, and spar are used descriptively to denote calcite grain sizes of 1-5 11m, �7-15j.lm, and> 15j.lm, respectively. We realize that this purely descriptive use of microspar differs from that of Folk (1965). The term 'pellet' will refer to a rounded aggregate of clay or micrite, or both, generally of sand size. 'Pelletal' will refer to a texture characterized by pellets. 'Ooids' and 'oolitic' refer to rounded particles of sand size consisting of concentrically banded micrite, usually enclosing a detrital grain which served as a nucleus. 'Pseudo-ooids' and 'pseudo-oolitic' denote rounded particles in which a single layer of micrite coats a detrital grain. 'Pelletoid' is used to describe sand-size grains coated with clay, silt, or ash. 'Pisolith' will denote a lithic fragment of nodule coated with laminar calcrete and generally rounded, with a minimum mean diameter of 0·5 em. The aeolian tuffs described herein are lithified deposits of tephra reworked by wind action prior to burial and consolidation. The materials used in this study consisted of roughly fifty samples of calcrete, calcareous tuff, pisoliths, and mode�n pelletal sediments collected by the senior author in East Africa. Hand samples were slabbed, polished, and etched with dilute HCl for observation with the binocular microscope. Etching proved quite useful in differentiat ing carbonate and non-carbonate material. Thin sections and acetate peels were used to study the textural relationships. Mineralogy was determined by selective staining, X-ray diffraction methods, and in some cases optically by refractive indices. An A.R. L. electron microprobe was used to determine fabric-related chemical variations. Bulk chemical analyses are reported for several of the calcretes.
GEOLOGIC AND CLIMATIC SUMMARY Pedogenic calcretes are widespread in northern Tanzania, and they are parti cularly well developed in the vicinity of Olduvai Gorge, at the eastern margin of the 26
Calcretes of northern Tanzania
35°
3•s
3•s
�
Basement Rocks
rTTTTTT1 Major Fault
o
10
20 km
Fig. 1. Regional map showing the loca tio n of Olduva i Gorge, Laeto lil, and the carbonatite volcanoes Oldoinyo Lenga i, Kerima si and Sa diman.
Serengeti Plain (Fig. 1). Here they occur at many levels in a stratigraphic section representing the past 2· 1 m.y. They are found at several horizons in the vicinity of Laetolil, which lies on the Eyasi Plateau only a few kilometres south of the divide separating drainage into Olduvai Gorge from that into Lake Eyasi. In the Laetolil region are two prominent calcretes between 3·0 and 3·5 m.y. old. The eastern part of Olduvai Gorge exposes a sequence of fossiliferous deposits about 100 m thick, termed the Olduvai Beds, which were deposited in a basin about 25 km in diameter. This basin lies to the east and southeast of several large volcanoes which have been intermittently active over the past 4 m.y. The Olduvai Beds are subdivided into seven formations that are from oldest to youngest as follows: Bed I, Bed II, Bed III, Bed IV, the Masek Beds, the Ndutu Beds and the Naisiusiu Beds (Hay, 1976). Sediments of Holocene age are also present in and near the gorge. Beds I through IV span the interval from about 2·1 to 0·6 m.y.B.P. Calcretes are found locally in Beds I and II, but were not studied in detail and are only considered to a minor extent in this report. Calcretes are widespread in the stratigraphic units younger than Bed IV, which comprise the Masek, Ndutu, and Naisiusiu Beds and unnamed deposits of Holocene age. The present study of Olduvai calcretes has focused primarily on these younger calcretes (Fig. 2), and unless otherwise specified, 'Olduvai calcretes' and 'calcretes of the Olduvai Beds' will refer to the calcretes formed over the past 600,000 years. The stratigraphic section representing the past 600,000 years is as much as 30 m thick and consists largely of aeolian tuff (Hay, 1976). Calcretes are found at many levels in this sequence and are widespread at disconformities. The Masek Beds (0·4-0·6 m.y.B.P.) are as much as 25 m thick and comprise an upper member, the Norkilili 27
R. L. Hay and R. J. Reeder HOLOCENE SEDIMENT
0
UNC ONSOLIDATED WIND-WORKED ASH
NAISIUSIU BEDS
rzzJ
INDURATED AEOLIAN TUFF
[I]]
MASSIVE CALCRETE
B
LAMINAR
�
PISOLITH
�
CLAYSTONE
[3
NONVOLCANIC SANDSTONE
NDUTU BEDS
(upper unit}
Norkilili Mbr.
(f) 0 w (l) �
bJ <1: ::;;:
CALCRETE
(lower unil)
Fig. 2. Generalized stratigraphic column show ing the pr incipal calcrete horizons a nd types of occur r ence in the Masek Beds and y ounger deposits of Olduva i Gorge.
Member, and an unnamed lower member. Within the lower member, calcretes are widespread both within and over the top of beds of aeolian tuffs (Fig. 2). The upper 60-90 em of the Norkilili Member is generally a calcrete. The Ndutu Beds ( 40,000400,000 year B.P.) were deposited during a period of intermittent erosion and partial filling of the gorge and are subdivided into two units, the upper and more widespread of which is probably 40,000-70,000 years old. The upper unit is generally 2-3 m thick where exposed in the rim of the gorge and over the eastern Serengeti Plain. The lower unit is generally 30-60 em thick. Calcretes occur at several levels in the Ndutu Beds but are best developed and most widespread at the base of the lower and upper units, and over the top of the upper unit. The Naisiusiu Beds ( � 15,000-23,000 years B.P.) are generally between 1 and 3 m thick. Calcrete is well developed at the base and top, and thin calcretes occur within the Naisiusiu Beds. Holocene sediments of the plain comprise a calcrete and an overlying unconsolidated deposit of wind-worked ash 30 em to 1 m thick. This ash is a mixture of material discharged in a major eruption 1300 years ago that was both produced by earlier eruptions of Holocene age and eroded from aeolian tuffs of late Pleistocene age. Volcanic ash of all the deposits listed above is nephelinite or related to nephelinite in composition (Hay, 1976). Kerimasi volcano (Fig. 1) discharged the ash of the Masek Beds, and Oldoinyo Lengai supplied most or all of the ash in the younger deposits. The Laetolil area contains a stratigraphic sequence at least 180 m thick, most of which is older than the Olduvai Beds (Leakey et al., 1976). The Laetolil Beds, at least 130 m thick, form the lowest stratigraphic unit, which ranges in age from 3·6 to about 4 m.y.B.P. This unit consists chiefly of aeolian tuff. The Laetolil Beds are disconform ably overlain by a 14-20 m thickness of aeolian tuffs, termed the Ndolanya Beds, which are probably only slightly younger than the Laetolil Beds. The Ndolanya Beds are subdivided into upper and lower units, the lower of which is phonolitic, the upper being nephelinitic. Well developed calcrete profiles are widespread at the top of both units. �
28
Calcretes of northern Tanzania
The present climate at Olduvai Gorge is arid, with a mean annual rainfall of 55-60 em, a mean annual temperature of 23°C, and evaporation slightly more than 1 em per day (Hay, 1976; Norton-Griffiths, Herlocker & Pennyevick, 1975). The Serengeti Plain in this area is a grassland savannah with Commiphora scrub and scattered Acacia (thorn trees). Vegetation is insufficient to prevent movement of sediment by wind, particularly during the dry season. Climate of the Olduvai and Laetolil regions was dominantly semi-arid, with brief periods of aridity, from about 4 to 0·6 m.y. B.P. (Hay, 1976). A dominantly arid climate has prevailed for the past 600,000 years (Hay, 1976; Cerling, Hay & O'Neil, 1977).
Table 1. Principal calcre te deposits of the Olduvai and Ndolanya Beds
Stratigraphic position
Age of calcre te*
Nature and thickness of calcrete
Beneath Holocene se dime nt
2,000--9,000 years B.P.
D iscontinuous laminar calcre te as much as 4 em thick over massive calcretes 5-30 e m thick
Be neath Naisiusiu Beds
16,000--22,000 years B.P.
D iscontinuous laminar calcre te as much as 4 mm thick over massive oolitic calcre te 30-45 em thick
Be neath Ndutu Beds
0·05-0-4 m. y . B .P .
One or two lami nar calcretes totalling 1-5 e m over massive, oolitic l ocally pisolithic calcre te 0·5-1·0 m thick (Figs 3 and 4)
Within lower unit of Mase k Beds
0·5-0·6 m.y.B.P .
Lami nar calcre te 1-2 em thick is wide spread over and within ae olian tuffs at different levels; massive calcrete, ge nerally a fe w ce ntime tres thick. Thick pisolithic coatings common ar ou nd tuff clasts in conglomerates
Bed II, Lemuta Member
1·6 m .y.B.P.
D iscontinuous massive calcretes ge ner ally 10 em to I m thick overlain rarely by lami nar calcrete
Bed I
1·7-2·0 m.y.B.P .
Massive calcretes ge ner ally 0·6-1·0 m thick in we stern part of basin
Top of Ndolanya Beds
�3-4 m.y.B.P .
D iscontinuous oolitic and pelletal laminar calcre te 1-2 em thick over dense oolitic massive calcre te 60 em to 1·5 m thick. Pe lletal calcite ve ins are widespread in upper 30 em of massive calcre te
Ndolanya Be ds, top of lower unit
�3·5 m.y.B.P .
Laminar calcrete 1-30 e m thick over relative ly porous massive calcrete 30 e m t o 1 m thick. Pelle tal and oolitic textures common in many layers of the laminar calcretes
* Ages for the Olduvai Beds are taken fr om Hay (1976 ). 29
R. L. Hay and R. J. Reeder
DE SCRIPTION AND ORIGIN OF CALCRETE S
Calcrete types and profiles Calcretes are of two types: laminar calcrete, which is a relatively pure dense limestone, and massive calcrete, which consists largely of aeolian tuff cemented and replaced by calcite. The two types can occur separately, but a fully developed, typical profile consists of laminar overlying massive calcrete. Laminar calcretes range in thickness from less than 1 mm to 5 em and are most commonly about 1 em thick. They tend to even out irregularities on the underlying surface (Fig. 3). They are dense, relatively impermeable, and range in colour from pale to dark yellowish-brown. Laminar calcrete may directly overlie bedrock such as quartzite and gneiss, and it can be developed on porous, micritic aeolian tuff rubble without an intervening cemented layer. Pelletal textures occur locally in small pockets of this micritic rubble. Massive calcretes are aeolian tuff cemented and replaced by calcite, and they grade downward into calcareous tuff. They range from a few centimetres to a metre or more in thickness. The upper part of a well developed massive calcrete is hard, well cemented, and non-porous, and it grades downward through softer. calcrete into porous, calcareous tuff. The lower part of the massive calcrete commonly has a 'wormy' appearance as a result of non-uniform cementation which accentuates root holes and small burrows. The hard, well cemented calcretes are dark yellowish-brown. Oolitic and pseudo-oolitic textures are widespread in the massive calcretes, and pisolithic structures are locally common. Pelletal textures occur locally in soft, poorly cemented areas of massive calcrete. Calcretes of the Olduvai Beds are overlain by wind-deposited tuffaceous sediment which now supports, or previously supported, vegetation. Thus, the complete profile consists of a sediment layer representing a soil or palaeosol, an underlying laminar calcrete, and a lowermost massive calcrete (Fig. 6). This profile is similar to calcrete profiles found elsewhere such as in New Mexico (Gile, Peterson & Grossman, 1966) and Shark Bay, Australia (Read, 1974). The massive calcrete represents the plugged horizon of Gile et a!. (1966). The fully developed calcrete profile of Olduvai Gorge compares morphologically with Stage 4 pedogenic profiles of Gile et a!. ( 1966). Three of the calcretes are atypical in one or more respects and will be briefly described. The calcrete at the base of the Ndutu Beds is the thickest and most highly developed of the Olduvai calcretes, and it formed intermittently during the deposition of the Ndutu Beds over a period of perhaps a few hundred thousand years. This calcrete forms the rim of the gorge in most places, and it underlies the eastern Serengeti Plain at shallow depths, preventing its dissection by streams. The massive calcrete is commonly about 1 m thick and is widely pisolithic. The overlying laminar calcrete is commonly composed of a lower, pale unit and an upper, darker unit consisting of alternating darker and lighter laminae. In many places the lower unit is fractured or eroded beneath the upper unit, and clasts of the lower unit may have pisolithic coat ings similar to the upper unit of calcrete. These two units of laminar calcrete are genetically related to the upper and lower units of the Ndutu Beds. Calcrete profiles in the lower unit of the Masek Beds lie at varied horizons and are not widespread at any single horizon. Laminated calcrete is the dominant form, and massive calcrete is generally only a few centimetres thick and can be porous and 30
Calcretes of northern Tanzania
0 I
t
l
em I '
Fig. 3. Laminar c alcrete o ver lying the upper par t of the massive c alcrete at the base of the Ndutu Beds,
Olduvai Gorge. Pisolith s with thi n coatings are scattered through the massive c alcrete. Arro ws indicate white micr ite laminae wh ich fill shr inkage cracks. Circle at A indicates locatio n of the photo micrograph of Fig. 9 a, and B indic ates the area of F ig . 9b.
poorly defined. The laminated calcretes occur both within aeolian tuff sequences and over the upper surface of aeolian tuff buried by claystone or sandstone. The lower calcrete of the Ndolanya Beds is unusual in having a laminar horizon, as much as 30 em thick, which consists of massive layers alternating with layers of pellets, ooids, and pseudo-ooids. The massive calcrete is commonly porous, chalky, and exhibits pelletal textures. Although atypical of the Olduvai region, laminar calcrete with pellets, ooids, and pseudo-ooids, has been widely noted in calcretes developed on carbonate rocks in coastal regions (James, 1972; Siesser, 1973; Scholle & Kinsman 1974; Ward, 1975). Moreover, we observed pellets and ooids within or directly overlying most of the well developed laminar calcretes we studied on a reconnaissance trip through the south western United States. These textures were found in calcretes of the Spring Mountains (Lattman, 1973) and Mormon Mesa (Gardner, 1972) of southern Nevada and un described calcretes near Tombstone and Wickieup, Arizona, and in the Argus Mountains, between Owens and Panamint lakes, in southeastern California. Thus, the Olduvai laminar calcretes seem to be unusual in their scarcity of pellets and ooids. Pisoliths
Pisoliths characteristically have a nucleus of aeolian tuff or calcrete, which 31
IS
R. L. Hay and R. J. Reeder
I
em
I
I
5 I
Fig. 4. Large pisolith wh ich is a coated clast overlying the eroded surface of the Masek Beds at
Olduva i Gorge . Two ge nerations of calcre te ( light a nd variegated dark) coat a fragment of massive calcre te with several pisoliths. Sma l l cavities, ind ica ted by arrows, are l i ned with crysta ls of d ol omite . At the bottom of photo is some of the aeolian tuff in wh ich the pisolith was e mbedded . The dark, conspicuously laminated calcrete a t top is the same calcre te which coats the pisoliths of F ig. 5 .
coated by micrite laminae similar to the laminar calcrete horizon. They are widespread and occur in massive calcrete, in sediments overlying a laminar calcrete horizon, and in conglomerates of the lower unit of the Masek Beds. Pisoliths in sediments above laminar calcretes are commonly large and characteristically have thick coatings, similar in appearance to underlying laminar calcrete. The rubble on aeolian deflation surfaces is commonly coated to form prominent pisolithic horizons, in which pisoliths are commonly 20-30 em in length (Fig. 4). Clasts in unconsolidated sediment of the Serengeti Plain are coated to a thickness of no more than 1 mm and represent an early stage in the formation of pisoliths. A mature stage is represented by pisoliths in the upper unit of the Ndutu Beds where coatings are commonly on the order of 1 em thick (Fig. 5). These pisoliths, like those in the modern sediment, grew while the matrix sediment was still unconsolidated. The evidence for this is pisoliths which were coated to a thickness of 1 em, then broken and. the halves considerably displaced, after which the broken pisoliths were coated to a thi<;kness of 1-2 mm (Fig. 5b). Pisoliths in the massive calcrete rarely exceed 2 em in diameter, and the coatings are generally oniy a few millimetres thick. The pisoliths of massive calcretes are most abundant and varied in the thick, widespread calcrete at the base of the Ndutu Beds Clasts of aeolian tuff are coated to form pisoliths in many conglomerates of the lower unit of the Masek Beds. Tuff clasts as much as 30 em in diameter are coated by calcrete 1-3 em thick. Clasts of lava and metamorphic rock in the same conglomerates are uncoated or only thinly coated with calcrete. 32
Calcretes of northern Tanzania
0 I
I
em
Fig. 5. Pisoliths of ae olian tuff in the upper unit of the Ndutu Beds, Olduvai Gorge . Pisolith s are orie nte d as they were in the tuff matrix. (a) Pisolith has a clast of lami nar calcrete as a nucle us; (b) massive calcrete forms the core of pisol ith wh ich was broken after all but the outermost pisolith ic coatings h ad bee n deposited.
ii . .
D m DID B
UNCONSOLIDATED PELLETOlD ASH INDURATED AEOLIAN MASSIVE CALCRETE LAMINAR
STAGE 3
STAGE 2
STAGE I
TUFF
@] � �
INDURATED CLAST PISOLITH VEGETATION
CALCRETE
FORMATION OF CALCRETE
:r
Fig. 6. Diagram illustrating the formation of calcrete at Olduvai Gorge .
Coatings are almost invariably thickest on the top and thinnest on the bottoms of those pisoliths found in sediments above a laminar calcrete (Fig. 5). This pattern is unique, as far as we are aware, for coatings are typically thicker on the bottom (Swineford, Leonard & Frye, 1958; Reeves, 1976). Coatings are thickest on the tops of some pisoliths within the Olduvai massive calcretes but are roughly uniform in others. Coatings are more or less uniform in conglomerates of the lower unit of the Masek Beds. Curiously, pisoliths rarely have nuclei of metamorphic rock or lava, even though clasts of this composition commonly occur in deposits with thickly coated pisoliths 33
R . L. Hay and R . J. Reeder
having nuclei of aeolian tuff and calcrete. Clasts of quartzite, for example, are rarely coated to a thickness of 1 mm, even where they lie adjacent to pisoliths with aeolian tuff nuclei of similar size and pisolithic coatings 1-3 em thick. Fractures and veinlets Fractures, both open and filled, are widespread in the calcretes and pisoliths. They are, however, much less common than in other well developed calcretes that have been described (Reeves, 1976). The largest fractures are vertical and cut the entire thickness of the massive calcrete. Most of these are no more than 2 mm wide and are thinly coated by calcite and other minerals. A few, wider fractures in massive calcrete are coated or filled by laminar calcrete continuous with the horizontal laminar calcrete horizon above. Smaller fractures are found in pisoliths and in calcretes, both massive and laminar. These are generally widest in the middle and pinch out at either end Most are roughly perpendicular to the surface of the pisolith or the top of the calcrete. Thin cracks are present between layers in some laminar calcretes and pisolithic coatings. Most of these are filled, but some are not. Veinlets of pelletal, fine-grained calcite as much as 2 em wide characterize the upper 30-50 em of the calcrete at the top of the Ndolanya Beds. They are dominantly subhorizontal and locally connected to form a network. Origin of calcretes These calcretes were formed chiefly at shallow depth ( 30-100 em) beneath wind deposited sediment at weathered horizons or disconformities (Fig. 6). Calcareous material in the aeolian sediment layer supplied the CaC03, and the horizons of low permeability reduced the infiltration of rain water and aided in precipitation of calcite at this horizon as the water evaporated. This mechanism is the same as that of Gile et al. (1966) except for the role of a weathered horizon or disconformity in determining the horizon of calcite accumulation. Some if not most of the calcretes of the lower member of the Masek Beds were very likely precipitated from vadose water at depths of at least a few metres, below the level of pedogenesis. As evidence, these calcretes can coat the surface of beds of aeolian tuff overlain by as much as 5 m of claystone and sandstone of fluvial origin and lacking development of palaeosols. A ground water origin seems required for the thick pisolithic coatings around tuff clasts in conglomerates. A striking feature of the Olduvai calcretes is the rapid rate at which they have formed. The youngest horizon, which overlies the Naisiusiu Beds, has yielded 14C dates ranging from 2190 ± 105 to 9130 ± 130 years B.P., and the average of seven dates is 5070 years B.P. Calcrete at the base of the Naisiusiu has given dates of 15,850 ± 270 to 22,470 ± 420, averaging 19,530 years B.P. Dates of 15,000-17,500 years B.P. were obtained from organic materials in the Naisiusiu Beds, suggesting that the calcrete was formed over several thousand years while the Naisiusiu Beds were deposited. These calcrete profiles would be classed as Stage 4 in the system of Gile et al. (1966), and at least 100,000 years were required to form the Stage 4 calcretes of the Jornada surface in New Mexico (Hawley, Bachman & Manley, 1976). Natrocarbonatite ash, a Na-Ca carbonate, was probably the major source of calcium carbonate in the calcretes of the Olduvai and Ndolanya Beds. Relatively little bedrock limestone is available as a source of aeolian calcite, either within, or �
34
Calcretes of northern Tanzania
upwind to the east of the Olduvai-Laetolil region. Two major carbonatite-producing volcanoes, and several smaller ones were active, however, during the deposition of the Masek Beds and younger formations of the Olduvai Basin. Dawson (1964) has suggested that the Olduvai calcretes are carbonatite tuffs from these volcanoes. Oldoinyo Lengai has erupted carbonatite ash and lavas in historic time (Dawson, 1962), and prehistoric carbonatite deposits have been documented for Kerimasi (Dawson, 1964; Hay, 1976). Carbonatite tuffs in the Laetolil Beds were erupted from Sadiman volcano, which could have supplied carbonatite ash to the Ndolanya Beds. Analysed natrocarbonatite lava of Oldoinyo Lengai (Dawson, 1962) is almost entirely composed of Na2C03 (51%), CaC03 (25%) and K2C03 (10%). Prehistoric carbonatite eruptive deposits of this and other volcanoes are now represented only by calcite, but very likely contained Na2C03 and K2C03• This ash would dissolve in congruently in rain water to give calcite, the least soluble of these carbonates, and Na +, 2 K+, Ca +, HC03-, and C03 - ions in solution. 14C should be absent in magmatic carbonatite, hence the radiocarbon age sequence 2 of calcretes must reflect equilibrium or near-equilibrium between C03 - ions from the dissolved carbonatite ash with atmospheric C02 before deposition of CaC03 in the calcrete. The scatter of dates (e.g. 2190-9130 years B.P.) in different samples from the same calcrete horizon can be attributed to several causes and will not be discussed further here (see Hay, 1976). Silicate reactions are another source of calcium, particularly for calcretes of the Masek Beds. Glass of the aeolian tuffs is typically altered to phyllosilicates and zeolites which contain less Ca than the original glass. As an example, a mafic melilitic vitric tuff of the Masek Beds has been altered to a mixture of iron-rich phengitic illite and phillipsite having 1·73% CaO (Hay, 1976, table 26, No. 7). Unaltered glass of this type should contain 10-16% CaO (Nockolds, 1954), hence the alteration process released considerable calcium. In many tuffs, particularly in the Ndutu and Naisiusiu Beds, the calcium is precipitated as CaC03 in the groundmass. In some other beds, particularly in the lower unit of the Masek Beds, calcium is extensively leached and may well have been precipitated elsewhere as calcrete and pisolithic coatings. Alter ation of glass to smectite may have contributed CaO to the calcrete at the top of the lower unit of the Ndolanya Beds.
PETROGRAPHY AND MINERALOGY Aeolian sedimentary deposits The modern sediment layer of the plain consists largely of nephelinite ash, much of which originated in a major eruption of Oldoinyo Lengai volcano about 1300 years B.P. (Hay, 1976). The tuffaceous sediment occurs in the form of dunes, sand ridges, and a widespread blanket-like deposit. The blanket-like deposit consists of air-fall tephra of the major eruption mixed by wind with older ash and varied other materials. The deposit is generally between 15 em and 1 m in thickness and underlies most of the grassland of the eastern Serengeti. The surface of this deposit is reworked and re deposited by winds during the dry season. The nature and origin of the dunes and sand ridges are described elsewhere (Hay, 1976). The blanket-like deposit is of special interest here because it provided a source of CaC03 for the Holocene calcrete, which it overlies. The deposit is unstratified and 35
R. L. Hay and R. J. Reeder
relatively fine-grained. The median diameter is close to 3
Calcretes of northern Tanzania
CLAY
ACCRETION
REPLACEMENT AND CEMENTATION
Fig. 7. D i agram illustrating the fo rmatio n of clay co ats by accretio n around grains and the replace ment and cementation of pelleto id grains to for m oolitic massive calcrete.
iron is presumably in the oxide form rather than within the illite structure. Analysed clay coatings in the Ndolanya Beds are montmorillonite, notably lower in Fe and K and more siliceous than the phengitic illite (Table 2). Laminar calcrete and pisoliths Laminar calcrete is a dense accumulation of multiple layers of micrite of varying purity. The individual laminae are generally less than 0·5 mm thick and may be slightly wavy. Micro-erosional surfaces or disconformities are present in the laminations, indicating discontinuous accretion and erosion of micrite (see Fig. 3). The laminated Table 2. E lectro n microprobe percentage analysis of relatively unaltered phyllosilicates in calcretes and aeo lian tuffs*
Olduvai Beds Clay Coatings
Clay Clasts (1)
(2)
(3)
(4)
(5)
Dark Lamina (6)
(7)
Ndolanya Beds Clay Coatings (8)
(9)
(10)
Si02 At.o. Ti02 FeO t MnO MgO CaO Na.o K.o
50·81 18 ·14 n . d. 12·03 0·01 4·04 1·33 0·28 7·02
46·62 44·72 20·06 16·75 1·94 1·21 9·79 9·14 n.d. n.d. 4·15 3·18 1·09 3·28 0·22 0·73 6·54 5·51
52·82 19·51 n.d. 6·34 0·25 1·83 5·40 0·89 4·93
45·11 20·43 n . d. 13·20 0·17 3·10 4·62 0·26 4·31
48·39 16·49 n.d. 11·94 0·06 3·78 0·88 0·37 6·54
50·86 20·66 n.d. 11·10 0·37 3·05 1·19 0·68 6·84
55·30 18·06 0·32 3·35 n. d . 3·13 2·66 0·31 0·41
53·63 14·80 1 . 51 5·35 n.d. 1·30 8 . 30 0·31 0·68
48·70 18·51 0·72 5·42 n.d. 2·20 7·77 0·54 0·85
Total**
93·66
89·54
85·39
91·97
91·20
88·45
94·75
83·54
85·88
84·71
* Nos 2 and 3 are from aeolian tuffs, No . 7 is fro m laminar calcrete and the o thers are from massive calcrete. The Olduvai samples are fro m the Masek Beds and yo unger deposits; the Ndolanya samples are from the upper calcrete. ** To tals are less than 100'/;; because of H20, OH- and epoxy. t To tal F e given as FeO. 37
R. L. Hay and R. J. Reeder
or layered appearance in the laminar calcrete is due to the alternation of light, pure micrite laminae with dark, impure laminae, which may be largely clay. Calcium carbonate forms 77-93"/o of bulk samples (Fig. 8), which average the light and dark laminae. Most of the laminae were formed by accretion of calcium carbonate and detrital materials at the top of the laminar calcrete. Some of the pure micrite laminae fill shrinkage fractures (Fig. 3), which are particularly common along clay-rich layers. Pisolithic coatings of clasts in the sediment layers above the laminar calcrete have the same patterns of laminae found in the underlying laminar calcrete, indicating a unique origin. The thickness of the coating varies between 0·5 and 30 mm, and is notably greater on the top surface and less on the sides and bottom (see Figs 4 and 5). Closer observation shows that the bulk of this top-side thickening is due to a thicken ing of only a small percentage of the total number of laminae (usually the pure laminae). In addition, the filling of shrinkage(?) fractures aids expansion of the top laminae. Pisoliths within the massive calcrete have relatively thin ( �0·5 em) coatings which generally lack the distinctive lamination found in the laminar calcites.
D �
V> "'
c. E
� 0 ci z
5
Lam1nor calcrete Mass1ve calcrete
4
3 2
40
50
60
70
80
90
100
Weight per cenl CaC03
Fig. 8. Histogram showing weight per cent calcium c arbonate in c alcretes of the Olduvai and Ndo l anya Beds. The percentage of calcium carbonate is taken as the acid-so luble fractio n of samples dissolved in lOYo HCl (six samples) and in acetic acid buffered to a pH of 4·5 (ten samples). No s 1, 2, 12, and 13 are from the Ndolanya Beds. These data show that laminar c alcretes contain more calcium c ar bonate th an massive calcretes, and the massive c alcretes vary greatly in content o f calcium carbo nate.
Massive calcrete Massive calcrete in all but the lower of the Ndolanya calcretes consists of variably cemented and replaced aeolian tuff. Clay and zeolites are extensively replaced, and oolitic textures have resulted from micrite replacement of clay coatings. Micrite pellets have been formed by replacement of clay pellets. Pellets, ooids, and pseudo oolites of accretionary origin are both common and widespread in the lower Ndolanya calcrete. The calcium carbonate content ranges from 47-74"/o in the seven analysed oolitic samples (Fig. 8). The oolitic texture is grain-supported, hence these per centages represent substantial replacement in the Olduvai samples which were cemented by zeolites before being replaced by calcrete. Assuming a 25"/o porosity for the Olduvai samples, which were cemented by zeolites prior to replacement, then 25"/o CaC03 could represent simply cementation, 50"/o would represent cementation plus 33"/o replacement and 70"/o would represent 60"/o replacement. Thus, replacement is an 38
Calcretes of northern Tanzania
Fig. 9. Oo litic textures i n the massive calcretes of Fig . 3: (a) is a polished surface i n reflected light, and centres of ooids appear dark. The fabric is grain suppo rted by the ooids ; (b) is a n acet ate peel i n transmitted light , and a n ooid wit h co ncentric layers of micrite i s near centre of photo .
39
R. L. Hay and R. J. Reeder
Fig. 10. Pseudo-oo litic and oolitic textures illus trated by acetate peels of the massive calcrete at the base of the Ndutu Beds : (a) depic ts an early stage in the formatio n of pseudo-ooids in which micrite (light) is rep l acing clay (dark) at margins (m) and in concentric zo nes (c) with in pelleto id clay co ats; (b) depicts a late s tage of replacement. Oo id with concentric layers at lower left is also shown in Fig. 9 . Original grain outlines are indicated b y inked l ines.
40
Calcretes of northern Tanzania
Fig. 11. Pel letal m icrite of accretionary origin in ve ins of the upper calcre te of the Ndola nya Beds .
Cement is largely m icrospar, and fabric appears to be gra in s upporte d . This fabric is identical to the structure grumeleuse of Cayeaux (1935). Ph otographs are of ace tate peels .
41
R. L. Hay and R. J. Reeder
important process in forming the massive calcretes. Swineford et al. (1958) are among the few workers who have demonstrated the major role of replacement in forming calcrete. Mineral grains, such as quartz, nepheline and augite, are etched and replaced to varying extents by calcite. These minerals are etched to about the same extent in the underlying tuffs, and most of the replacement in the calcrete represents the precipit ation of calcite around previously etched grains. Ooids, pseudo-ooids and pellets of replacement origin
These ooids and pseudo-ooids range in size from 70-500 J.!m in mean dimension, but generally fall between 90 and 300 J.!m (Figs 9 and 10). Pellets are usually smaller, most commonly ranging from 60-125 J.!m in mean dimension (see Fig. 11). Clay coatings of pelletoid grains have distinctive features which can be traced through various stages of replacement. Clay particles of the unaltered cortexes are often tangentially oriented, giving the cortex a radial extinction under crossed polars. Two distinct generations of clay coatings are commonly present, the older of which is typically a densely packed, dark clay, exhibiting a prominent radial extinction pattern. The outer, second generation is much lighter in colour, less dense, and somewhat silty, exhibiting a diffuse radial extinction. Remnant clay in ooids may preserve both the tangential orientation and evidence of two generations of coatings. Normally, the inner coating is much less replaced than the outer coating (Fig. lOa). Varying degrees of micrite replacement occur in coatings of the ooids and in pellets, and slightly replaced coatings or pellets lie adjacent to extensively replaced ones. These features clearly indicate a primary control, such as porosity, on micrite distribution. Ooids, pseudo-ooids, and pellets of accretionary origin
Accretion was important in forming pellets, ooids, and pseudo-ooids in the calcretes of the Ndolanya Beds, where they form layers and fill fractures. Pellets are more common than ooids and pseudo-ooids, and they form a grain-supported fabric. There is markedly less clay in the accretionary pellets and ooid coatings than in pellets and ooids of replacement origin, and its distribution is uniform, suggesting that accretion of micrite is the dominant process. These pellets, ooids, and pseudo ooids are similar to those of calcretes in coastal regions described by Multer & Hoffmeister, 1968; James, 1972; Siesser, 1973; Scholle & Kinsman, 1974; Ward, 1975. An accretionary origin is accepted by all of these but Ward (1975, p. 563), who suggests that the spherical shapes originate by desiccation fracturing of micrite and its sub sequent solution, deposition, and slight transportation. Calcite cement fabrics
The massive calcretes are cemented to varying degrees by micrite, microspar and, less commonly, spar. Both X-ray diffraction and electron microprobe analysis indicate the calcite is low in Mg ( 1 mol per cent MgC03). Microspar is most common and typically occurs as a pore-filling, interlocking mosaic of crystals. Grain sizes commonly increase toward Ihe centre of a pore (Fig. 11b), as often observed in cement fabrics. In many of the massive calcretes, microspar cementation of micrite pellets produces a 'clotted texture' or 'structure grumeleuse' identical to those described for many ancient, massive limestones (Cayeux, 1935; Carozzi, 1960; Bathurst, 1971). Micrite �
42
Calcretes of northern Tanzania
of the pellets is randomly embayed by microspar cement in this delicate interplay of calcite grains. This is clearly not a neomorphic fabric in these calcretes, but rather a primary pedogenic fabric. Non-calcite components The non-calcite fraction of the calcretes is a mixture of inherited and authigenic materials. Primary mineral grains, rock fragments, and vitroclasts generally form the largest amount of the acid-insoluble residue. Generally less common are clay minerals and zeolites, both primary and authigenic. Other minerals, including dolomite and opal, are additional authigenic constituents. Iron and manganese oxides occur widely, at least in the laminar calcretes, and silica-sesquioxide 'gel' is present in all of the analysed calcretes. Olduvai calcretes
Silt- and sand-size detritus in the laminar calcretes is similar in composition to that in the overlying sediments from which it has been incorporated. The clay fraction in two samples of laminar calcrete is composed of illite and dioctahedral chlorite with interlayered illite. The non-calcite fraction of the massive calcretes is composed largely of materials inherited from the parent aeolian tuff. Thus, it comprises mineral grains, rock fragments, and zeolites. Chabazite is the only zeolite represented in diffractograms of four of the five residues analysed, and phillipsite is the sole zeolite in the fifth. The dominance of chabazite in the calcretes contrasts with the over whelming predominance of phillipsite in the parent tuffs (Hay, 1976,table 25),suggest ing that phillipsite was selectively replaced by calcite. Pelletoid clay coats formerly unreplaced by calcite are waxy and appear unaltered, whereas partly replaced coats are porous and appear chalklike. The fraction finer than 2 f.tm in four of the five samples of massive calcrete dissolved in buffered acetic acid is wholly or largely dioctahedral chlorite with interlayered iiiite. Illite is the only clay mineral in one sample, and minor illite occurs in three others. A very weak 7·1 A peak, presumably of a kaolinite-type mineral, was noted in one sample. Dolomite is a common authigenic mineral which occurs principally as a late filling of fractures and other cavities in both laminar and massive calcrete,including pisoliths. Crystals are coarsest,10-50 f.tm in diameter,in the lower of the calcretes,including that at the base of the Ndutu Beds. The dolomite is disordered, as determined by lack of ordering reflections in Debye-Scherrer films. Dawsonite is found in all calcretes,but is common only in the calcrete at the base of the Ndutu Beds. Crystals of natrolite and phillipsite line some of the fractures in calcretes of the Masek Beds and at the base of the Ndutu Beds. Manganese and iron oxides have been deposited in some of the laminar calcretes. Both are concentrated in the darker layers,and probe analyses give maximum values of 50·5% MnO and 28·1% FeO,respectively. The manganese occurs largely in the form of dendrites, probably of hydrated Mn02• Silica-sesquioxide 'gel' was obtained from all eleven samples in which the acid solution of dissolved calcrete was neutralized with NH40H. The percentage of gel, weighed after drying overnight at about 90°C,ranges from 3·5 to 1 1·8,averaging 7·1% in samples dissolved in HCl; it averaged 1·3-3·0, averaging 2· 1 /';;,in samples dissolved in acetic acid buffered at a pH of 4·5. The gels were reduced to only 10-15% of their original weight on ignition, demonstrating that they are quite hydrous. One sample, 43
R. L. Hay and R. J. Reeder
from an HCl solution, has the following composition, on a water-free basis, as analysed by X-ray fluorescence: Si02, 34·3% ; Al203, 38·9%; Ti02, 0·3%, Fe203, 4·9% ; MnO, 0·1%; MgO, 1·5% ; CaO, 18·0% ; Na20, 1·8%. Ndo/anya calcretes
Pumice, nepheline and other minerals in the upper Ndolanya calcrete are surpris ingly little altered, indicating that the calcrete was formed before the tuffs had been appreciably weathered. Partly replaced pelletoid clay coatings in the massive calcrete are white, porous and chalklike. Montmorillonite constitutes the fraction finer than 2 Jlm in the laminar calcrete. The same size fraction of one massive calcrete sample is principally dioctahedral chlorite with interstratified illite. Another sample exhibits only the 7·1 A peak of a kaolinite-type mineral. The dioctahedral chlorite and a 7·1 A clay mineral are considered authigenic, formed from montmorillonite, \-V hich is the clay mineral in beds below the calcrete. Opal coats root cavities in the upper calcrete. The lower Ndolanya calcrete contains partly altered mineral grains and montmoril lonite, and it lacks fresh glass. No authigenic minerals other than calcite have been identified in this lower calcrete. Thus, authigenic minerals are both less varied and less abundant in the Ndolanya calcretes than in those of Olduvai Gorge.
CHEMICAL ASPECTS OF CALCRETE Chemical analyses for six laminar calcretes from Olduvai Gorge (Table 3) are similar to the mean composition for world calcretes computed by Goudie (1972). The Olduvai calcretes average 79·8% CaC03 and 7·2% of Si02, compared to Goudie's (1972) averages of 79·28% of CaC03 and 12·30% Si02• The Olduvai calcretes also compare rather closely with dense laminar calcretes from the High Plains, New Mexico (Aristarain, 1970). Bulk compositions of massive calcrete were not determined because of the highly variable nature of the replacement. Small-scale texturally related chemical variation within five calcite samples was studied using the electron microprobe. The elements analysed were Si, AI, Fe, Mg, Ca, Na, K and either Ti or Mn. We studied both laminar and massive calcretes of the Olduvai Beds, but only massive calcrete of the Ndolanya Beds. Within the laminar calcrete, the light laminae are essentially pure micrite with about 1 mole % MgC03 in solid solution. Dark laminae are potassic, presumably iilitic clay, diluted to varying extents with CaC03 (Fig. 12b) and with varying amounts of Mn and Fe, at least partly in the form of oxides. A few laminae are essentially pure clay, similar in com position to unaltered pelletoid clay coatings in the aeolian tuffs (Table 2). The composition of the clay in pseudo-ooids and pellets in the Olduvai calcretes varies considerably as a function of calcite replacement, as measured by increasing values of CaO (Fig. 13). The unaltered clay coatings and pellets, as noted earlier, have the composition of a phengitic illite with. an atomic Si/Al ratio averaging 2·5 and Al/K ratios generally between 2 and 4 (Table 2 : Fig. 13). In the most thoroughly studied sample, increasing replacement is accompanied by decreasing Si/AI ratios (Fig. 12) and increasing Al/K ratios (Fig. 13b). The Si/Al ratio drops to about I in highly replaced coatings and pellets. We interpret these data to indicate that increasing replacement of clay by calcite is accompanied by dissolution or leaching of phengitic illite clay and formation of clay approaching the composition of kaolinite or halloysite. 44
Calcretes of northern Tanzania
Table 3. Chemical analyses in percentage of laminar calcretes fro m Olduvai Gorge*
(I) Per cent s io . Al203 Ti02 Fe203 M nO MgO CaO NazO K .o Cl
s
ppm Nb Zr y Sr Rb Ba �C02 for MgO for CaO
7·14 2 ·85 0·43 2 ·62 0·09 I ·71 44·15 0·66 0·69 0·001 0 · 065
70 160 20 4120 40 400
(2)
6 28 2·42 0 - 41 2 · 04 0 · 06 1·84 4 3 · 54 1 ·95 0·70 0· 552 0 · 072
40 10 20 60 50 1350
(3)
6 · 30 2·15 0 · 37 1 . 88 0·08 1·94 45 ·12 0·89 0·88 0·006 0·029
40 170 30 5200 50 680
(4)
9·50 3 · 45 0 · 50 2 ·91 0·52 1 .1 1 44 ·12 0 · 45 0·80 0·039
90 1 30 20 2500 50 790
(5)
8·53 2 · 54 0·48 2 ·68 0 ·19 1 · 12 44 · 37 0· 58 0·95 0 · 018 0·022
60 120 20 2060 50 990
(6)
5 · 33 1 · 41 0·47 1 ·90 0 ·11 1 ·81 47 ·00 1 · 02 0 ·25 0 · 005 0 ·110
30 80 20 1040 30 220
I ·87 34·65
2·01 34· 17
2 · 12 35 ·41
1 ·21 34 ·63
1 · 22 34 ·82
1 · 98 36 · 89
To tal
97 · 39
95 · 58
97·83
99 · 59
97·83
98 · 31
% acidinsoluble
10·1
15 ·0
n. d .
14·2
22·7
11 · 1
* Analyses are b y R . N . Jack using X-ray fluorescence. A l l F e is given as Fe203. (1) Holocene calcrete ; U.C. M useum No . 478-2 34; (2 ) Calcrete at the base of the Naisiusiu Beds ; U.C. No . 478-249; (3) Calcrete at the base of the Naisiusiu Beds ; U.C. No . 478-254; (4) Calcrete at the base o f th e upper unit o f the Ndutu Beds ; U.C. No . 4 78-275; (5) Calcrete o f th e lower u n i t o f the M asek Beds ; U.C. No . 478-266 ; (6) Pisolithic coating of th e tuff clast in a co nglomerate of th e lower unit o f the M asek Beds ; U.S. No . 478-272 .
This inference is supported by the presence of a 7 · I A clay-mineral peak in a con centrate of altered clay coatings. The diffraction pattern is weak, however, and non crystalline aluminosilicate materials such as allophane may have resulted from the replacement process. The silica-sesquioxide gel that flocculated from solutions of dissolved calcrete may represent material formed in this way. The few analyses in two other samples scatter widely and show no clear trend. Reaction of illite to kaolinite and allophane(?) would yield Si, Fe, K and Mg. Aluminium (or aluminate) ions could also be released in the solution of illite. Phillipsite and perhaps other zeolites were dissolved in replacement by calcite. Thus, silicate alterations in the massive calcrete could have supplied components for the authigenic dolomite, dawsonite and zeolites. Although X-ray data indicate that the clay-mineral fraction in the upper Ndolanya calcrete has been changed mineralogically, microprobe analyses show no clear evidence 45
-
3 2
R . L. Hay and R. J. Reeder
(a)
---- �-
-
-
-
_
:_
KAOL INITE
.
.
-· _
20
40 60 80 Percent CoO of Tota l Analysis
0
3
��
•.
..: _
_
_
100
( b)
2
.
.
· .
.� (f)
..
·.
.
.
KAOLINITE
20
0
4
60
40
80
100
80
100
Percent CoO of Total Analysis
.
.
(c)
.
.. .
.
.
.
2o
20
40
.
60
Percent CoO of Total Analysis
Fig. 12. Atomic Si/Al ratio plotted against perce nt CaO in (a) pelletoid coatings re place d to varying e xte nts by calcite in Olduvai m assive calcre te at the base of the Ndutu Beds, (b) Olduvai lam inar cal cre te overlying the m assive calcrete of (a), and (c) pe lletoid coatings re place d to varying e xtents i n Ndolanya massive calcrete at the t o p of the Ndolanya Beds . Chemical data were obtained b y e lectron m icroprobe .
of chemical change in the replacement process. The Si/Al ratio, for example, scatters rather widely, but shows no trend as a function oheplacement (Fig. 12c). Mineralogic data, however, suggest that montmorillonite in the calcrete has been altered to diocta hedral chlorite and to a kaolinite-type mineral. Reaction of montmorillonite to form kaolinite or dioctahedral (aluminous) chlorite releases silica, but the microprobe results suggest that any silica released in clay-mineral reactions remained in the pelletoid coatings. The scatter of Si/Al ratios in analysed coatings (Fig. 12c) may indeed reflect small-scale movement of silica in the coatings. Finally, we wish to emphasize that this is only an introductory mineralogical and chemical study of silicate reactions, based on few samples. Despite their limitations, our data clearly show that zeolites and clay minerals have been altered and dissolved, and not simply displaced, in the development of these East African calcretes. 46
Calcretes of northern Tanzania
2 D- -
MUSCOVITE
2
0
- - - - - - - - - - - - - - - -7 KAOLINITE
10
4
12
� (b) __-Unaltered Clayclasts and Coats D
2
0-
.
- - - - - :...· - .!.. ....: ·_: - :- � - - . - - :;...
MUSCOVITE
2
0
100
•
0
....
z
12
10
4
14
(c)
0 80 0
14
'
.
60
11.1 40
� �
20 0
2
10
4
12
14
Fig. 13. Atomic Al/K ratio plotte d a gainst (a) Si/Al ratio in unreplace d clay c lasts and pelle toid c oa ts
of the Olduvai Beds (e, in calcre te ; _.. , in aeolian tuff), (b) Si/Al ratio in c lay coats re placed to varying extents ·in a sample of Olduvai mass ive calcre te at the base of the Ndutu Beds, and (c) Ca O in the ana lyses of (b). Chemical data were obtained by e lectron microprobe .
CON CL U SION S (I) Pedogenic calcretes of the Olduvai and Ndolanya Beds are closely associated with wind-worked volcanic ash of nephelinite or closely related composition. (2) A complete profile with calcrete typically consists of an unconsolidated sedi ment overlying laminar calcrete, which coats and overlies a massive calcrete. These profiles correspond closely to the mature, Stage 4 pedogonic calcretes of southern New Mexico (Gile et a!., 1966). Laminar calcretes of one Olduvai stratigraphic unit were deposited by vadose water below the level of pedogenesis. (3) An Olduvai calcrete profile can develop to a mature stage in only a few thousand years, probably reflecting the periodic fall of large quantitites of carbonatite ash, which is a source of readily available calcium carbonate. ·
47
R. L. Hay and R. J. Reeder
(4) Replacement was a major process in the formation of massive calcrete. Oolitic textures were formed chiefly by micrite replacing pelletoid clay coatings around sand-sized grains. (5) Increasing replacement of clay by micrite in the Olduvai massive calcretes was accompanied by decreasing Si/Al and K/Al ratios, reflecting dissolution or leach ing of phengitic illite and formation of clay approaching the composition of halloysite or kaolinite. Si, AI, K and Mg were lost in the replacement and transformation of clay and zeolites and were probably precipitated in the form of dolomite, zeolites and dawsonite. In the upper Ndolanya calcrete, montmorillonite was altered to a kaolinite type mineral and to dioctahedral chlorite.
ACKNO WLEDGMENT S This study was supported by the National Science Foundation (grants EAR 72-0 1523 and EAR 76-84583). The electron microprobe used in this study was pur chased with NSF grant GA 38086. We are indebted to L. H. Gile and J. W. Hawley for showing us the calcrete sequence near Las Cruces, New Mexico, which gave us a standard of comparison for the East African calcretes. We profited from discussion with T. E. Cerling on clay-mineral analysis and calcrete pedogenesis. We thank J. Hampel for the photomicrographs and for assistance in X-ray diffraction analysis. Microprobe mounts were made by Len Leudke and Sharon Hudson, and thin sections are by S. J. Chebul. X-ray fluorescence analyses were made by R. N. Jack. The manu script was reviewed by L. H. Gile, N. P. James, and A. Goudie.
REFERENCES L.F. (1 970) Chemical analyses of caliche profiles from the High .Plains New Mexico. J. Ceo!. 78, 201-2 1 2 .
ARISTARAIN,
BATHURST, R . G . C . ( 1 9 7 1 ) Carbonate Sediments and Their Diagenesis. Developments in Sedimentology,
12, pp. 505-5 1 3 . Elsevier Publishing Co . , Amsterdam. BRINDLEY, G .W. (1961)
The chlorite minerals. The X-ray Identification and Crystal Structures o.f Clay
Minerals (Ed. by G . Brown), pp. 242-296. Mineralogical Society, London.
A . (1 960) Microscopic Sedimentary Petrography. Wiley and Sons, New York. CAYEUX, L . ( 1 935) Les Roches Sedimentaires de France ; Roches Carbonatees. Masson, Paris. CERLING, T.E., HAY, R.L. & O NEIL , J. R. (1 977) Isotopic evidence for dramatic climatic changes in East Africa during the Pleistocene. Nature, 267, 1 37-1 3 8 . D AWSON, J . B . ( 1 962) The geology o f Oldoinyo Lengai. Bull. volcan. 24, 349-3 87. DAWSON, J.B. (1 964) Carbonatitic volcanic ashes i n northern Tanganyika. Bull. volcan. 27, 1-J l . FOLK, R.L. (1 965) Some aspects of recrystallization i n ancient limestones. In : Dolomite and Limestone Diagenesis : a Symposium (Ed. by L. C. Pray and R. C. M urray), Spec. Pubis Soc. econ.
CAROZZI,
'
Paleont. Miner. , Tulsa, 13, 14-48 . GARDNER,
L . R . ( 1 972) Origin of the Mormon· Mesa caliche, Clark County, Nevada. Bull. geol. Soc.
Am. 83, 1 43 - 1 5 6 .
L.H. & GROSSMAN, R.B. ( 1 967) Morphology of the argillic horizon in desert soils of southern New Mexico. Soil Sci. 106, 6-1 5 . GILE, L . H . , PETERSON, F . F . & GROSSMAN, R . B . (1 966) Morphological a n d genetic sequences of carbonate accumulation in desert soils. Soil Sci. 101, 347-360. GouDIE, A. (1 972) The chemistry of world calcrete deposits. J. Ceo!. 80, 449--463 . GouDIE, A. ( 1 973) Duricrusts in Tropical Landscapes. Clarendon Press, Oxford.
GILE,
48
Calcretes of northern Tanzania
HAWLEY, J.H., BACHMAN, G.O. & MANLEY, K. ( 1 976) Quaternary stratigraphy in the Basin and Range and Great Plains provinces, New Mexico and western Texas. Quaternary Stratigraphy of North America (Ed. by W. C. Mahaney), pp. 23 5-274. Dowden, Hutchinson, and Rose, Stroudsburg, Pa . , U.S.A. HAY, R.L. ( 1 963) Zeolitic weathering i n Olduvai Gorge, Tanganyika. Bull. geol. Soc. Am. 74, 1 28 1 1 286.
HAY, R.L. ( 1 976) Geology of the Olduvai Gorge. University of California Press, Berkeley. JAMES, N.P. ( 1 972) Holocene and Pleistocene calcareous crust (caliche) profiles : criteria for subaerial exposure. J. sedim. Petrol. 42, 8 1 7-83 6 . KAHLE, C. F. ( 1 977) Origin of subaerial Holocene calcareous crusts : Role of algae, fungi, and sparmi critization . Sedimentology, 24, 4 1 3-43 5 . LACROIX, A . ( 1 904) L a Montagne Petee et ses Eruptions. Masson, Paris. LATTMAN, L.H. ( 1 973) Calcium carbonate cementation of alluvial fans in southern Nevada. Bull. geol. Soc. Am. 84, 301 3-3028.
LEAKEY, M . D . , HAY, R.L., CuRTIS, G . H . , DRAKE, R.E., JACKES, M . K . & WHITE, T.D. (1 976) Fossil hominids i n the Laetolil Beds. Nature, 262, 460-466. MuLTER, H.G. & HOFFMEISTER, J.E. (1 968) Subaerial laminated crusts of the Florida Keys. Bull. geol. Soc. Am. 79, 1 83-192.
F . (1 969) Ages of calcretes in southern Africa. S. Afr. archaeol. Bull. 24, 347-374. S.R. ( 1 9 54) Average chemical compositions of some igneous rocks . Bull. geol. Soc. A m .
NETTERBERG, NocKOLDS,
65, 1007-1032.
M., HERLOCKER, D . & PENNYEVICK, L. ( 1 975) The patterns of rainfall i n the Serengeti ecosystem. E. Afr. wild!. J. 13, 347-374. READ, J.F. (1 974) Calcrete deposits and Quaternary sediments, Edel Province, Shark Bay, Western Australia. Mem. Am. Ass. Petrol. Ceo!. 12, 250-282. REEVES, C.C. (1 976) Caliche; Origin, Classification, Morphology and Uses. Estacado Books, Lubbock, Texas. ScHOLLE, P.A. & KINSMAN, D.J.J. ( 1 974) Aragonitic and high-Mg calcite caliche from the Persian Gulf-A modern analog for the Permian of Texas and New Mexico. J. sedim. Petrol. 44, 904NORTON-GRIFFITHS,
916. SIESSER,
W.G. ( 1 973) Diagenetically formed ooids a n d intraclasts i n South African calcretes. Sedi
mentology, 20, 539-55 1 .
A . , LEONARD, A.B. & FRYE, J . C. (1 958) Petrology o f the Pliocene pisolitic l imestone i n t h e Great Plains. Bull. Kans. geol. Surv. 130, 97-1 1 6 . WARD, W.C. (1 975) Petrology a n d diagenesis of carbonate eolianites o f northeastern Yucatan Peninsula, Mexico. I n : Belize Shelf: Carbonate Sediments, Clastic Sediments and Ecology (Ed. by K . F. Wantland and W. C. Pusey III). A m . Ass. Petrol. Ceo!. Stud. Ceo!. 2, 500-571 . WEAVER, C.E. & POLLARD, L.D. ( 1 973) The Chemistry of Clay Minerals. Developments in Sedi mentology, 15, Elsevier Publishing Co . , Amsterdam.
SwiNEFORD,
(Manuscript received
4
November 1 977 ; revision received 7 February 1 978)
49
Reprinted from Sedimentology ( 1980) 27 559-576
Pellets, ooids, sepiolite and silica in three calcretes of the southwestern United States
R. L. H AY & B R I A N W IGGI N S Department o f Geology and Geophysics, University o f California, Berkeley, Cal((ornia 94720, U.S.A.
ABS T R A C T Pellets and ooids are widespread and locally abundant i n mature calcrete profiles i n the Argus Range, California ; near Wickieup, Arizona ; and in Kyle Canyon, Nevada. Most concentrations of pellets and ooids either overlie laminar calcrete at various levels in the calcrete profile or fill subhorizontal fractures in the petrocalcic horizon. In all three profiles the petrocalcic horizon has been thickened by the pelletal, chemically deposited fracture fillings. Pellets range from 0·02 to 8·0 mm in diameter and consist principally of micritic calcite and sepiolite. Ooid coatings are chiefly calcite and opal or calcite and sepiolite. The pellets represent small concretions, some of which grew by accretion, either in void space or by displacing adjacent sediment, and the others of which were formed by cementation of pellet-shaped bodies of porous micrite. Ooid coatings with opal or sepiolite may have been deposited as a gel with sufficient strength for surface tension to thin the coatings over angular corners of nuclei so as to increase the roundness and sphericity of the particles. Major problems in calcrete genesis are (1) the cause of subhorizontal fractures and the mechanism for widening a fracture as sediment accumulates in it and (2) what determines the deposition of calcite, sepiolite, and opal as pellets and ooid coatings or as laminar layers.
INTRODUCTION
coatings over edges and corners of angular nuclei. Some is known, but much remains to be learned, about the authigenic clay minerals in calcretes of inland regions. Palygorskite or sepiolite in calcretes has been documented by Vanden Heuvel ( 1 966); G ardner ( 1972); Frye et a!. ( 1 974); Goolsby ( 1 975); Bachman & Machette ( 1 977); and Millot et a!. ( 1 977). Most of these writers attributed the formation of palygorskite and sepiolite to alteration of detrital montmorillonite and mixed-layer montmorillonite illite because the latter clays are commonly present in sediments below the calcrete but are generally absent or rare in the calcrete horizon with paly gorskite or sepiolite (see Gardner, 1 972). Bachman& Machette consider sepiolite a late-stage product formed in soils where palygorskite is dominant, and they did not find sepiolite in soils younger than middle-Pleistocene.
Micritic pellets and oolites o f pedogenic origin are widespread in the calcretes developed on carbonate sediments in coastal regions. Examples have been described by Multer & Hoffmeister ( 1 968), James ( 1 972), Siesser ( 1 973), Braithwaite ( 1 975) and Ward ( 1 975). The origin of most pellets is as yet poorly understood and no satisfactory explanation has been offered for pedogenic ooid coatings which thin over corners and edges so as to develop spherical particles . Calcretes o f inland semi-arid and arid regions have been studied much less, but oolites were noted and pellets were figured by Swineford, Frye & Leonard ( 1 958) in their study of the Ogallala calcrete of the High Plains. Bachman & Machette ( 1 977) report the occurrence of ooids and pellets in calcretes of the southwestern United States . These reports add little to the understanding of the origin of pellets and do not offer a hypothesis for the thinning of ooid Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
51
R. L.
Hay and B. Wiggins horizon of Gile, Peterson & Grossman ( 1 965). Pellet refers to a spherical or ovoid particle of authigenic origin Jacking a nucleus and generally consisting of micritic calcite, with or without clay. Ooid refers to a spherical or ovoid particle Jess than 2 mm in diameter comprising a nucleus enclosed by one or more laminae. The majority of ooid coatings are concentrically banded. Pisolites are structurally similar to ooids but greater than 2 mm in diameter. Crumb aggregates are irregular micritic aggregates separated by voids or by cement (Braith waite, 1975). Clotted texture refers to a flocculent structure similar to that produced by coalescing pellets (Bathurst, 1 975, p. 86). Micrite, microspar, and spar are used to denote calcite grain sizes of 1 -5, -7- 1 5, and 15 fLm, respectively. An aqueous solution of methyl orange was used to aid in identifying sepiolite, both in the field and the laboratory. In at least nine-tenths of the samples, sepiolite stains pink to red either if pure or mixed with micritic calcite. Some of the rarer varieties of trioctahedral smectite (e.g., hectorite) can give a similar stain. Thin sections were used for studying textures, and sawed and smoothed slabs etched with 1 0% HCI were particularly useful in showing distribu tions of carbonate and non-carbonate minerals . The distribution of sepiolite on the slab is readily shown by methyl orange stain. The non-carbonate fraction of most samples was obtained by crushing a sample to pass through a 1 20-mesh sieve and dissolving the carbonate in a 1 0% solution of acetic
The present study is primarily a petrographic and mineralogical analysis emphasizing the distri bution, nature, and origin of pellets, ooids, sepiolite, and silica in three calcrete profiles of the south western United States . Structure of the calcretes was studied in the field, the main result being to show that the petrocalcic horizon of calcrete profiles can be substantially thickened by chemical precipitation in subhorizontal fractures . This study began in 1977 as a two-week recon naissance search for pelletal and oolitic calcretes . Pellets and ooids proved to be widespread and locally abundant in mature calcretes, and laboratory study was concentrated on samples from three calcretes developed in deposits of different mineral composition : ( 1 ) basalt and basaltic detritus in the Argus Range of southwestern California (Fig. 1 ) ; (2) silicic volcanic alluvial deposits near Wickieup, in west-central Arizona ; and (3) carbonate alluvial fan deposits in Kyle Canyon of the Spring Mountains of southwestern Nevada. Additional field study was made of the Argus Range and Kyle Canyon cal cretes .
T E R M I N O L O GY A N D M ETHO D S Calcrete is used here to designate a carbonate-rich, dominantly indurated profile or horizon of pedogenic origin. A petrocalcic horizon is a continuously cemented or indurated calcic horizon (Soil Survey Staff, 1975), usually synonymous with the K2m
·,·
Argus
�
Range
i I
::o\
�� Z l>
Mojave Desert • Barstow
Needles
N
..._Kingman -.
\ I Wickieup-_
35°
�...... 0
50
WK:
100 km 115°
Fig. 1. Index map showing the calcrete localities studied. AR indicates the Argus Range calcrete, WK indicates the
calcrete near Wickieup, and KC indicates the Kyle Canyon calcretes. 52
Three calcretes of SW U.S.A. 1 979). The general geology of this arn is shown on the Death Valley Sheet of the Geologic Map of California (I :250,000) published by the California Division of Mines and Geology ( 1 958) . The calcrete profi le exposed in the roadcuts com prises three units totalling about 2 m in t hickness (Fig. 3) . Unit 1, at the top, is 30-60 em t hick and is
acid buffered to a pH of 4·5-5· 0 with Li2C03• An hour's treatment in a 5% solution of acet ic acid was used in the later stages of the study, as this solution dissolves calcite much more rapidly, and comparative X-ray analyses showed no differ ence in clay minerals obtained by t he two methods. Diffractometer analyses were made of oriented and unoriented samples of most clay concentrates, and glycolation and heating were used to aid in some identifications. An A.R.L. e lectron microprobe was used to determine fabric-re lated chemical variations. Beam diameter was about 5-10 [Lm, and the sample current was 0·018 [Lamps.
LITTER OF BASALT BOULDERS DOMINANTLY SEMICONSOLIDATED CALCAREOUS CLASTIC SEDIMENT
LAYERS AND TABULAR BLOCKS OF DENSE LAMINAR AND PELLETAL LIMESTONE
1--'-"-
DE S C R IPTION A N D O R I G I N OF C A L C RETES
1.0 m
o. 5
Calcrete of the Argus Range, California
CEMENTED BASALTIC
This calcrete is notable for its high content of pellets and for a thick petrocalcic horizon develope d on relatively steep slopes by accretion of calcite along fractures. It is exposed best and was studied in most detail in discontinuous roadcuts over a distance of about 1 km along Highway 190 near the crest of the Argus Range (Fig. I) at an elevation of about 1 430 m, and centring at approximately 36°22'33"N and l i 7°37'37"W (Darwin 15-minute Quad., 1 95 1 ) . This calcrete overlies olivine basalt and i s developed within and upon basaltic colluvium; and, at lower elevations to the east of t he roadcuts, it is developed on alluvial-fan deposits (Fig. 2) . Equivalent basalt nearby has been dated at 5·3-6·0 m.y. BP, (G iovanetti,
0
Fig. 3. Generalized columnar section of the calcrete
profile exposed in roadcuts at the north end of the Argus Range. Small circles represent pellets.
a massive, poorly sorted micritic deposit, containing detrital silt, sand, and fragme nts of basalt and cal crete as much as 2 em in diamete r. This unit is chiefly unconsolidated t o se mi-consolidated but includes cemented massive layers and thin, dis continuous hard laminar layers. Lenticular layers of pellets as much as 2 em thick overlie the laminar layer at the top of Unit 2 and the laminar layers within U nit 1. In a few places, Unit 1 is represented by as much as 1 m of calcareous basaltic colluvium filling channels eroded into Unit 2. Unit 1 is overlain by a litter of basalt boulders and blocks. Unit 2, the upper part of the petrocalcic horizon, is 1 ·0- 1·5 m thick and consists largely of l ayers and tabular blocks of dense limestone 5-15 em thick (Fig. 4) and capped by discontinuous laminar layers 1-2 em thick. The limestone layers and blocks of Unit 2 consists of alternating laminar and comparatively massive, generally pelletal deposits. The layers and blocks are separated by fractures, most of which parallel the 10-25° slope of the land surface . Steeply inclined to vertical fractures are spaced at intervals of about 1-5 m and may penetrate the entire thick ness of Units 2 and 3. In a few places Unit 2 directly overlies basaltic lava. Unit 3, at the base , is as much as a metre thick and consists of calcite-ceme nted
X
X
BASALT X
X
X
X
X
X
X
X
0
COLLUVIUM
500m
Fig. 2. Cross-section eastward from the crest of the Argus
Range, California, showing the calcrete, indicated by hachured line, which overlies basalt and basaltic volcani clastic sediments. Locality (I) represents the roadcut exposures of calcrete, and (2) represents the sampled locality on the fan surface. Vertical exaggeration is
3 ·5 X . 53
R. L.
Hay and B. Wiggins
Fig. 4. Roadcut exposure of Unit 2 and the lower part of Unit 1 of the Argus Range calcrete. Unit 2 consists of layers and tabular blocks of dense limestone separated by fractures. Arrow indicates pellet-filled fracture at the base of Unit 1. Vertical scale bar is 50 em long.
sediment. Fractures filled with soft pelletal sediment are most common in the northernmost roadcut, where seven fractures filled with pelletal sediment and totalling 15 em in thickness were noted in a single vertical section. These fracture fillings of pelletal sediment are typically u ncemented at the base of th� fracture, and become progres sively more cemented upward. The topmost pelletal sedi ment may be thoroughly cemented onto the bottom layer of the overlying block of dense limestone. Uncemented sed iment at the base of the fracture filling represents the latest deposit of the fracture, which was either precipitated in place from solution or washed down from a higher level along vertical fractures. Clearly the limestone blocks were some how lifted as they were thickened i n order for the fractures to remain open channelways for water and sediment. Cracks are a common feature of calcrete profiles (Reeves, 1976, p. 48), but their origin is not understood. Once formed, however, the fractures of Unit 2 may have been kept open by downslope creep of the limestone layers and blocks, thus accounting at least in part for the development .of this thick unit of dense limestone. Laminar layers and uncemented to moderately
poorly sorted basaltic colluvium with an admixture of quartzo-feldspathic detritus of granitic origin. This cemented unit forms the lower part of the petro calcic horizon and constitutes the 'plugged horizon' of Gile, Peterson & Grossman ( 1 966). Unit 2 is an unusual deposit, not found in the mature, Stage IV calcrete profiles of Gile et al ( 1 965), and will be described in more detail. The blocks consist of alternating layers of laminar and pelletal texture (Fig. 5). The layering parallels, at least crudely, the fractures bounding the blocks, and the dominant layering parallels the land surface. This layering shows that the blocks grew by accretion along the fractures, and the layers are thickest on the lower part of the blocks showing that the blocks grew chiefly by addition of material to their u ndersides. Laminar layers coating uppe r surfaces of blocks are generally even, whereas those coating the bottoms are generally bumpy, or pustulose. Truncation of horizontal layers by vertical layers (Fig. 5) represents the development and coat ing of a vertical fracture. Subhorizontal fractures are as much as 4 em wide, and fractures more than about 1 em wide are filled with unconsolidated to weakly cemented pellet-rich 54
Three calcretes of S W U.S.A. that of calcite in one fracture filling of Unit 2 (Table 1 , No. 5). The dense limestone of Unit 2 consists principally of calcite and opal (Table 1 , No. 6). Illite and chlorite of detrital origin were identified in a silty sample from Unit 1 . The calcrete profile developed on the alluvial fan is 1·2-1·5 m thick and consists largely of cemented basaltic alluvium of gravelly texture. Here the surface of the fan has a slope of 3-4°. Discontinuous laminar layers as much as 2 e m thick occur at various levels but are most common in the upper 60 em of the profile. Concentrations of pellets overlie most the laminar horizons, and less commonly fill sub horizontal fractures in the cemented alluvium. Pellets also occur in small lenticles within laminar layers coating the bottoms of pebbles and cobbles. Both laminar and pelletal layers consist of calcite and either opal or sepiolite, or both. This profile is similar to the Stage IV calcrete of Gile eta!., ( 1 966) except that laminar layers lie at various levels within the cemented alluvium rather than as a coating upon it. The calcrete on the alluvial-fan deposits contains by volume, only about a third as much calcite and other chemical precipitates as calcrete exposed at higher elevations on steeper slopes. Where exposed in roadcuts, the Argus Range calcrete is a composite profile, in which accumulation of calcium carbonate was interrupted at least locally
THIN, E V E N LAYERS UNEVEN, PUSTULOSE LAYERS PISOLITE PELLETS AND OOIDS DETRITAL SAND FRAGMENT OF BASALT Fig. 5. Diagram showing the internal structure of the right-hand part of a tabular limestone block of Unit 2 of the Argus Range calcrete. Fractures border the top, bottom, and right-hand margins of the block. Layers are numbered 1 to 6, from oldest to youngest.
cemented pellets of Units 1 and 2 consist of micritic calcite with subordinate sepiolite (Table 1 , Nos 1 , 2 , and 3). The amount o f sepiolite approximately equals
Table 1. Composition of samples from calcretes
Sample no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14
%Acidinsoluble
Locality* and sample type AR-laminar layer at top of Unit 2 AR-uncemented pellets at base of Unit 1 AR-turtleskin pellets in fracture of Unit 2 AR-semiconsolidated pelletal limestone within Unit 1 AR-semiconsolidated, colloform, semipelletal fracture filling in Unit 2 AR-hard siliceous pelletal limestone Wk-topmost laminar layer Wk-semiconsolidated oolitic limestone above laminar calcrete of no. 7 Wk-semiconsolidated pelletal fracture filling Wk-0·5-1 mm fraction of uncemented pellets in solution cavity KC-iaminar layer, top of calcrete of Surface I KC-<:emented matrix of conglomerate KC-pelletal limestone filling fracture in calcrete KC-pelletal pisolitic limestone filling fracture in calcrete
Composition of acid-insoluble fractiont
24 32 18 50
sep., �5%sand+silt sep. �pal., trace sand + silt sep. sep. + about 30%silt + sand
54
sep. + about 5% silt+ sand
50 45 55
opal + 5-1 0% silt + sand opal+ trace silt + sand opal + about 50%detrital silt+sand
60 54
sep. + trace of silt+sand sep. �pal; trace opal; ,;;;5% silt+ sand
4
silt, sand, + trace clay (unident.)
18 9
illite, chlorite, + about 50% silt + sand sep.
20
sep. + 1 0-20% sand + silt
* AR refers to Argus Range, Wk to Wickieup, and KC to Kyle Canyon. t Sep. refers to sepiolite, and pal. to palygorskite. 55
R. L.
Hay and B. Wiggins
by channel erosion at the top of Unit 2. The upper most laminar layer of Unit 2 yielded a 14C date of 1 6,450 ± 250 year BP (Teledyne Isotopes No. I-10,489). This layer is not eroded, and the date presumably falls within the post-erosional period of calcrete formation represented by Unit 1. Con tamination by older or younger carbonate cannot be ruled out for this single date, but the determined age is compatible with field relations. Units 2 and 3 of the Argus Range were probably formed over at least a few hundred thousand years, by comparison with other mature profiles whose age is either known or estimated (see Bachman & Machette, 1 977).
KYLE CANYON
WICKIEUP WEAKLY CEMENTED PELLETAL
•}
•
SEDIMENT FILLS CRAC K AND SOLUTION CAVITY
CM
[
Calcrete near Wickieup, Arizona
This calcrete, developed on silicic volcanic fanglomer ate, is exposed in north-south roadcuts along High way 93 about 26 km south of Wickieup, in west central Arizona. These exposures lie at an elevation of about 915 m and are located at about lat. 34° 3 1 '49"N and long. 1 1 3°22'22"W ( 1 :250,000 Prescott Quadrangle, 1 954). The fanglomerate is dissected, and the calcrete conforms roughly to the topography although it is thicker over hillcrests than on hillslopes. The calcrete has a maximum thickness of about a metre (Fig. 6) and consists largely of well cemented conglomerates, sandstones, and reworked tuffs of silicic (dacitic?) composition. Discontinuous laminar layers, 2-1 0 mm thick, are present at various levels in the cemented zone, and as many as four were noted in a single vertical sec tion. Most of the laminar layers are overlain by a layer of ooids between 1 and 5 em thick. The laminar and overlying oolitic layers parallel the hillslopes, cutting across the near-horizontal bedding in the alluvium. Pellets and semiconsolidated pelletal limestone fill a subhorizontal fracture 1 ·5-3 em wide which lies about 25 em below the top of the calcrete exposed in the eastern roadcut. At one place the fracture has been enlarged by solution to form a cavity 60 em wide and 15 em deep (Fig. 6). The bottom surface of the fracture and the solution cavity are coated with a laminar layer as much as 1 em thick. The solution pocket is filled with uncemented pellets, f ragments of calcrete, and rare pisolites. The ooids and associated laminar layers are siliceous (Table 1 , Nos 7, 8), whereas pellets fi lling the widespread fracture and the solution cavity are sepiolitic (Table 1, Nos 9, 1 0). The laminar layer coating the lower surface of the fracture
WEAKLY C�MENTED
t'ZZl
PETROCALCIC HORIZON
1°001
CONGLOMERATE
r:::;-::;1 �
l•••••l B
CONGLOMERATE WITH
PARTLY
DISSOLVED PEBBLES PELLETS AND OOIDS
L AMINAR CALCRETE
Fig. 6. Generalized profiles with calcrete in Kyle Canyon,
Nevada, and near Wickieup, Arizona.
contains both opal and sepiolite. Sepiolite and silica are generally lacking in the calcite cement of the conglomerate in the petrocalcic horizon. The layers of ooids and associated laminar calcrete appear to have developed in tuffaceous sandstone of which sand-sized grains became nuclei of ooids. The ooid coatings represent a substantial increase in volume and consequent displacement of nuclei, hence these oolitic deposits must have been formed before the sandstone had been cemented. The f rac ture was formed and locally enlarged by solution after cementation of the alluvium and it then was filled by materials precipitated from solution. The content of calcium carbonate and complex history of this calcrete suggest that its development began at least 1 00,000 years ago (see Bachman & Machette, 1 977). Calcrete of Kyle Canyon, Nevada
Kyle Canyon lies on the northern side of the Spring Mountains of southern Nevada. In the canyon are three widespread fan surfaces underlain by carbonate alluvium and preserved by calcretes, which are most fully developed on the older surface and least developed on the youngest, sub-modern surface 56
Three calcretes of SW U.S.A. micrite. Opal replaces perhaps 1-2% of the fracture filling pelletal limestone in the calcretes older than that of surface 2. The opal forms chertlike lenticles and layers, which preserve the pelletal texture of the limestone. A thin layer of opal coats the bottom of a few pebbles. Illite and chlorite of detrital origin are the clay minerals in a sample of calcite-cemented conglomerate matrix (Table 1, No. I2).
(Lattman, I 973). Surface I, the highest and oldest, has a petrocalcic horizon as much as 2 m thick; the petrocalcic horizon of surface 2 is 0· 3-1· 5 m thick; and pebble coatings and thin calcic horizons charac terize surface 3 . One well developed calcrete profile and several weakly developed ones are interbedded in fan deposits below the calcrete of surface 2. The well developed calcrete, generally I ·O-I ·5m thick, is l:: elieved to represent the buried calcrete of surface I . The present study is largely confined to the cal crete of surface 2 and the buried calcrete of surface I . These were examined and sampled principally in sections 23, 24, and 27, T. I 9S., R.58E. (see U.S. Geol. Surv. Corn Creek Springs I S-minute Quadrangle, 1952). Here surface 2 has an elevation of about I 400 m and a slope of 2°45'. By extrapolating data given by Lattman ( I 973), the average annual rainfall in this area is on the order of 35-40 em, and the mean annual temperature is about I 4°C. The petrocalcic horizons of the two calcretes are . chiefly cemented alluvium, generally with an over lying laminar layer 0·5-I ·O em thick and with dis continuous laminar layers having about the same thickness at various levels within the profile (Fig. 6). Pelletal limestone and weakly cemented aggregates of calcareous pellets overlie many, if not most, of the laminar layers. Subhorizontal fracture fillings of the laminar and pelletal limestone are widespread in the well developed petrocalcic horizons. These are generally between I and 8 em in width and several metres in lateral extent. Laminar layers coat the bottom surface of most fractures and are overlain by massive, generally pelletal micritic limestone. The fractures were formed and filled after forma tion of the petrocalcic horizon, but some and perhaps most of the laminar layers and associated con centrations of pellets within the petrocalcic horizon were deposited before the horizon had been fully cemented. Evidence for the latter is provided by the calcrete of surface 2 where it is weakly developed and consists of a series of thin, discontinuous cemented horizons within uncemented to weakly cemented alluvium. The cemented horizons are capped by thin laminar layers overlain by concentrations of pellets. A few pellet concentrations have a matrix of silty clay. Sepiolite is present in most of the pellets (Table I, Nos 13 and I 4) and has been identified in some laminar layers. However some of the pelletal lime stone fi lling fractures consists of pure micritic calcite pellets dispersed in a matrix of sepiolitic
AUTHIG ENIC M I N E R A L S I N C A L C R E TES Calcite Calcite is the principal authigenic mineral. The calcite of pellets and most of the laminar layers is micritic, whereas microspar cements pelletal lime stone and may form thin layers in the undercoatings of limestone blocks in Unit 2. Spar occurs both as a cement and as layers in undercoatings. Based on lattice constants deter mined by X-ray diffraction (Goldsmith & Graf, 1958), the calcite in calcretes of the Argus Range and Wickieup contains no more than 2-3 mol % MgC03• One microprobe analysis of calcite spar from the Argus Range indicates 2 ·5% MgC03. Calcite of the Kyle Canyon calcretes commonly contains as much as 5% MgC03 as inferred from X-ray measurements.
Sepiolite
Sepiolite is by far the most common clay mineral' and its authigenic origin is demonstrated by its association with chemically precipitated calcite and silica in deposits Jacking detrital sediment. It also occurs in forms characteristic of chemical pre cipitates such as thin layers, veinlets, cavity fillings, pellets, and ooid coatings. Relatively pure sepiolite is birefringent and has a mean refractive index of about I ·500 in the few samples for which it was determined. Most of the sepiolite gives the major d spacings of well crystallized sepiolite (see Caillere & Henin, 196 I ), and diffractograms typically resemble those given by Vanden Heuvel (1966) for pedogenic sepiolite of New Mexico. Some of the Argus Range sepiolite gives d spacings both different from .and smaller in number than those of well crystallized sepiolite. These anomalous samples are probably of disordered, poorly crystalline sepiolite. A few samples give a ( I IO) d spacing of I2·6 A. which is asymmetric to lower angles of 2 6, and others exhibit a broad peak extending from I2·6 to 14 A. 57
R. L. Hay and
These anomalous samples give only the following d spacings, in addition to that at 12·6-14 A: 4·5 1 4·27, 2·58, and 1·52 A. The ( 1 1 0) diffractometer peak does not expand in glycolated samples, and i t remains almost u nchanged by heating a t 300°C overnight and is only slightly reduced by heating at 500°C for 1 h. By contrast, the ( 1 10) peak at 1 2· 1 A of well crystallized sepiolite from playa deposits of southwest Nevada (Papke, 1 972) i s greatly reduced b y heating overnight a t 300°C, and i s almost eliminated by heating at 500°C. Still more anomalous, poorly crystalline (?) sepiolite is the clay mineral in one pelletal sample from Unit 1 in the Argus Range. Its diffractometer pattern, though poorly defined, only exhibits peaks at 1·52 and 1 4 A. The 1 4 A peak does not expand on glycola tion, and its size is reduced but otherwise unchanged by heating at 500°C. The clay has a delicate fibrous structure as shown by TEM, supporting its i dentifica tion as sepiolite. Sepiolite in the Argus Range calcrete is a relatively pure Mg-silicate, u nlike the aluminous sepiolite described by Vanden Heuvel ( 1 966). This is i ndicated by electron microprobe analysis of sepiolitic pellets
B. Wiggins and their coatings (Table 2, Nos. 2, 5, 6, 7). Using the i deal formula for sepiolite [ Mg2Si30n (OH) . 3 H20], some analyses shown an excess of silica, relative to Mg, and others show a deficiency. The excess silica probably represents opal, which is commonly associated with sepiolite. The silica deficiency may reflect inaccurate analyses resulting from i nappropriate microprobe standards.
Opal
Most opal occurs together wit h either calcite or sepiolite, and they can form a cement, layers, vein lets, pellets, and ooid coatings. Opal may also occur as localized, relatively pure, chertlike layers and nodules in the Argus Range and Kyle Canyon calcretes and as nodules in sepiolitic calcite veins in basalt beneath the Argus Range calcrete. The more abundant, non-cherty opal gives a broad diffractometer hump centring at about 4·0 A and is thus a type of amorphous silica. Dense, chertlike opal gives a strong, well defined 4· 1 A diffractometer peak for o:-cristobalite, and most of the analysed samples have a shoulder at 4·3 A representing
Table 2. Microprobe analyses* and calculated mineral compositiont of pellets and ooid coatings
Ooid coatings
Pellets
sio. Al203 FeO MgO CaO Na20 K.o
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8 )
(9)
9·3 0· 1 0·0 6·3
22·3 0·2 0·0 1 2·0 3 1 ·6 0·2 0· 1
23·2 0·2 0·0 2·1 36·5 0·2 0·1
45·7 0·2 0·0 2·8 2 1 ·2 0·2 0·1
22·9 0· 5 0·1 1 0·7 27·6 0·3 0·2
37·1 0·9 0·2
18·0 1 ·8 0· 1 0· 1
83·6 0·4 0·1 1 0·2 1 ·4 0·1 0·0
1 6·0 0·7 0·2 0·5 43·5 0·0 0·0
35·3 0·4 0· 1 0·3 3 1 ·9 0·0 0·0
70·2 37·8 0· 9
60·9 77-6 1· 1
68·0
0 1 6·0
0 35·3
0·9 95·6
0· 5 93·4
38·7 0· 4
0·2
Total
55·0
66·4
62·3
CaC03 MgCO Sep. Si02
69· 0
1· 5 23·0 -3·2
56·4 1 ·3 46· 9 -3·2
65·1
0·7 9 1 ·0
0·5 101·9
0· 5 92·9
Anal. Other Total
1 ·5 5·7 20·1
62·3
58·2
95·8
3·2 0·1 74·1 -3·2
1· 1
9·5 40·5
40·9 0·9 42·0 0·1
0 42·0 60·8
0·5 89·2
1 ·1 85·0
1·3 75·5
1 04·5
( 1 ), (2) Pellet in fracture filling of Unit 2 of calcrete in Argus Range. (3), (4) Pellet in semi-consolidated limestone of Unit 2. (5) Ooid coating of pellet in semi-consolidated limestone of Unit 2. (6) , (7) Ooid coating of pellet from fracture filling in Unit 2. (8), (9) Ooid coating from the calcrete near Wickieup.
0·6
56·9
0·7
* Analyses have been corrected for the inferred contents of C02 and H20 in the sepiolite and calcite. t Mineral composition was calculated as follows. All of the CaO and an appropriate amount of MgO (if available) were used to form-calcite with 2·5 mol :Yo MgC03• The remaining MgO was combined with Si02 and H20 to form sepiolite with the formula Mg2Si307•5(0H).3H20. Any remaining Si02 is interpreted as opal, and negative values for Si02 represent the deficiency of Si02 needed to combine with MgO to form sepiolite. Al203, FeO, K20, and Na20 were not used in the calculated minerals and are lumped together as Anal. Other. 58
Three calcretes of SW U.S.A.
tridymite. This type of opal has been termed opal CT. Soft porous opal widely replaces sepiolite in Unit 2 of the Argus Range, and dense opal replaces both sepiolite and calcite in the Argus Range and Kyle Canyon calcretes . The soft porous opal may have resulted from leaching of Mg, but the dense opal represents a gain of Si as well as a l oss of Mg and CaC03• Replacement of sepiolite and calcite by opal is evidence of a change in the chemical environment from s uper-saturated to under-saturated with respect to sepiolite and calcite. It is not clear whether this is a result of short- or l ong-term changes in water chemistry. Short-term changes might result from periods of unusually heavy precipitation. A l ong-term change may have been caused by more abundant vegetation during a Pleistocene glacial stage by increasing the pC02 and lowering the p H o f water in the vadose zone.
evidence of a substantial aeolian contribution from the west. Much of the Ca in the Wickieup calcrete is probably derived from aeolian calcite as weather ing of the associated s ilici�S alluvial deposits would yield a low i:atio of Ca to Si, yet calcite is much more abundant than opal and sepiolite. The sources for Mg are difficult to establish inasmuch as it is much less abundant than Ca, and aeolian sediment h as been supplied to all three profiles . Magnesium in the Argus Range and Wickieup calcretes may h ave been supplied both by weathering of volcanic rocks and of aeolian detritus . Magnesium in the Kyle Canyon calcretes· may well have been derived by solution of both alluvial and aeolian carbonate sediment.
Relation of sepiolite to palygorskite
Palygorskite, as identified in diffractograms, is a minor constituent in one sample of sepiolitic pellets from the Argus Range calcrete and in one from the Wickieup calcrete. The rarity of palygorskite in the three calcrete profiles requires comment inasmuch as i t is more common than sepiolite in most of the calcretes that have been described. Palygorskite differs in composition from sepiolite in a higher content of AI and a lower ratio of Si to Mg. The availability of AI must determine to a l arge extent whether palygorskite or sepiolite is formed i n chemical environments with varying activity ratios of Mg and Si. Very little AI can be dissolved i n fl uids with a pH o f 7-8·5, a common range for water associated with carbonate deposits . Authigenic minerals in the three calcretes were deposited from solutions, and sepiolite is thus to be expected rather than palygorskite. The common occurrence of X ray-amorphous opal -in the W ickieup and Argus Range calcretes points to h igh activities of silica, at least intermittently, which would favour the form ation of sepiolite. Palygorskite would be expected to form by reaction and replacement of aluminous detritus s uch as montmorillonite or aluminous bedrock (e.g. Gardner, 1 972; Millot et a!., 1 977).
Sources of silica, calcium, and magnesium
The Si, Mg, and Ca of the authigenic minerals were · derived to varying extents from the primary materials in the calcrete profile and from the aeolian sediment derived elsewhere. Most of the Si was probably provided by weathering (hydrolysis) of the primary silicate minerals associated with the calcrete profiles . This explains why opal and sepiolite are much more common in the calcretes developed on volcanic rocks than on dominantly carbonate alluvium. Calcium was supplied both by solution and hydro lysis of primary materials and by solution of wind transported calcite. Although Lauman ( 1 973) has stressed the important contributions of wind transported calcite to the calcretes of southwest Nevada, a substantial part of the authigenic calcite in the Kyle Canyon calcretes was derived by solution of the associated carbonate alluvium. As evidence, the upper one-fourth to one-half of the limestone pebbles are dissolved in the well developed petro calcic h orizons . Calcium in the Argus Range calcrete was undoubtedly supplied by weathering of basaltic bedrock, but aeolian calcite very l ikely also contributed Ca to the calcrete in view of the large excess of authigenic calcite over silica and sepiolite in the calcrete. Aeolian calcite may have been derived from the intermittently dry lake margin zone of Owens Lake, which l ies west and upwind, from the Argus Range. M oreover, the quartzose silt and sand of the Argus Range calcrete and associated basaltic volcaniclastic rocks are
D E S C RIPTI O N A ND O R I G I N O F PELL ETS, O O IDS A ND P I S O L ITES Summary of occurrence
Pellets and ooid coatings form an es timated 1 5-20% of the authigenic materials in Unit 2 of the Argus 59
R. L.
Hay and B. Wiggins
Range profile (Figs 3 and 4) and 5-10/;; of the authigenic constituents of the petrocalcic horizon of the profile developed on alluvium at the foot of the range. Pellets and ooids form about 5-10% of the authigenic materials in the petrocalcic horizon of the Wickieup profile and probably on the order of 1-3/;; of the authigenic materials in the petrocalcic horizons of Kyle Canyon. The largest amount of pellets is found in sub horizontal fracture fillings and other cavities in the petrocalcic horizons . Layers and pockets of pellets have also developed widely in unconsolidated sedi ments above laminar layers which lie upon the petrocalcic horizon or were formed within i t before it was fully cemented. Pellets may also form thin l ayers and small lenses within laminar layers , particu larly that capping the petrocalcic horizon. pellets commonly have a grain-s upported texture (Figs 7a, 8a and b), and interstices can either be void or filled by microspar cement or a matrix of micritic crumb aggregates and small pellets (Fig. 8a and b). Pellets may, however, be dispersed in structureless micrite filling fractures in the Kyle Canyon calcretes . Pellets commonly occur together with sand in some fracture fi lli ng deposits of the Argus Range calcrete. It should be emphasized that pellets are absent or rare in some mature calcretes , particularly those lacking subhorizontal fractures . Fractures and pellets are, for example, extremely rare in calcretes of the Olduvai Beds , Tanzania, whereas both are common in calcretes of the nearby Ndolanya Beds (Hay & Reeder, 1978). Ooids are overall much less common than pellets , but their occurrence is generally similar, and at least a few ooids are present in most fracture-filling concentrations of pellets . Concentrations of ooids are thickest and most abundant in the Wickieup calcretes where they overlie laminar layers within the petrocalcic horizon. Here they generally have a grain-s upported texture and a matrix of crumb aggregates and small pellets (Fig. 8c) . Locally the ooids are dispersed in the matrix. Ooids with thin coati ngs predominate over pellets in many layers of the Argus Range calcrete, where the nuclei may either be pellets or sand grains. Ooids with thick coati ngs have sand grains as nuclei and are generally restricted to opal-rich layers . Concentrations of ooids are generally cemented by calcite or opal or by both. Pisolites are of two types, one of which occurs in concentrations of ooids and is similar to the associ ated ooids except for a diameter greater than 2 mm. Pisolites of the other type are generally larger and
occur pri ncipally in fracture fillings, where they are typically dispersed in a matrix of pellets.
Pellets·
Pellets range in diameter fro m about 0·02 to 8·0 mm, and the grain size of samples may be unimodal, bimodal or, rarely, trimodal. Concentrations of pellets vary considerably in degree of sorting (Table 3, Nos 2, 3, 5), and with few excepti ons the bulk of pellets fall within the size ranges of 0·5-1 ·5 mm and 0· 02-0·10 mm. In one unusually coarse sample, the bulk of pellets is about evenly dist ributed between 1·0 and 6·0 mm (Table 3, No. 3). Pellets between 0· 5 and 2·0 mm i n diameter are dominantly ovoid (Fig. 7a and b), although spherical pellets predominate in a few samples (e.g. Fig. 8b). Surfaces of pellets are generally smooth, but a reticulate pattern of s urface fractures characterizes most of the pellets larger than 2 mm i n an unusually coarse sample (Fig. 7c). These pellets, termed turtleskin pellets, become more irregular in shape and in fracture pattern with i ncreasing size. With de creasing size, the turtleskin pellets grade into typical smooth-surfaced ovoid pellets . Pellets in the range of 0·02 to 0·10 mm can form s mall concentrations or they may occur together with crumb aggregates , bordering or filling small rootholes or forming a matrix for the larger pellets . These small pellets are l ess uniform in shape than the larger pellets , and with decreasing roundness they grade into crumb aggregates . In the Argus Range calcrete, concentrations of small pellets grade i nto areas with clotted texture. Pellets are chiefly a mixture of micritic calcite and sepiolite except in the Argus Range calcrete, where opal has replaced much of the sepiolite and some calcite. The content of sepiolite in analysed samples ranges from 9/;; i n pellets of Kyle Canyon (Table 1 , N o . 1 3 ) t o about 60% i n pellets of Wickieup (Table 1, No. 9). Sepiolitic pellets may also contain detrital silt and fine sand. Purely micritic pellets are found i n fracture fillings i n the Kyle Canyon calcretes. More than nine-tenths of the pellets appear structureless in thin sections (Fig. 8a and b) but a small percentage show a faint to well marked concentric zoning in content of detritus or calcite. Smal l pellets forming the matrix for larger pellets generally, but not always, appear texturally and mineralogically similar to the adjacent larger pelle.ts . Most pellets remain where they were formed i n all o f the calcretes except for Unit 2 o f the Argus 60
Three calcretes of SW
U.S.A.
Table 3. Size of pellets
Locality* 1. AR
2. AR 3. ARt
4. Wk 5. Wk 6. KC
Occurrence In fractures, fillings, and '-overlying laminar layers
Diameter of pellets Most are between 0·5 and 2 mm ; pellets 0·02-0· 1 0 mm form small concentrations and occur in larger pellets Sieve analysis : median diameter= 1 · 1 5 mm ; crcp= 1 ·5; modal concentration at 1 ·0-1·5 mm
Uncemented pellets with ooids filling fracture 2·5 em wide in upper part of Unit 2 Uncemented pellets filling fracture Sieve analysis: median diameter= 1·85 mm; 2·5 em wide in lower part of crcp =1·8; no modal concentration, and pellets Unit 2 are evenly distributed between 1 and 6 mm. Pellets larger than 2 mm generally have 'turtleskin' surface textures. In fracture fillings 0·02-2 m m ; modes of 0·02-0·03, 0·05-0· 1 0, and at about 0·50 mm Uncemented pe1lets in solution Sieve analysis: median diameter=0·37 mm ; cavity crcp= 1-1 In fracture fillings and overlying 0·02-2 mm; modes of 0·02-0·05, 0·25-0·50, and 1-2 mm laminar layers
* Abbreviations are AR=Argus Range, Wk=Wickieup, and KC=Kyle Canyon. t Estimated percentages of pisolites, oolites, and angular fragments of calcrete are excluded from sieved fractions to give size data for pellets alone.
Range calcrete. A small percentage of pellets in the fracture fillings of these profiles differ in compositi on fr om adjacent pellets and were very likely either washed into or redeposited within the fracture. However, many and perhaps most of the pellets in Unit 2 of the Argus Range calcrete were trans ported at least a short distance to their present position. As evidence, many layers contain broken pellets and adjacent pellets of different composition. Field relationships indicate that some of the pellets which had been formed over the top of Unit 2 were transported, together with detrital silt and sand, down along vertical fractures into horizontal fractures, where they became cemented onto the base of the limestone blocks above the fractures. Some fracture-filling concentrati ons of pellets i n Unit 2 have not been redeposited, however. These generally contain little detrital silt and sand, and the pellets are unbroken and relatively uniform in composition. The abundant, distinctive turtleskin pellets in one layer are additional evidence for growth in place, as similar pellets have not been found elsewhere in the calcrete. The large pellets and most of the smaller ones are small concretions which grew in some places by accretion, either in void space or by displacing adjacent sediment, and elsewhere by cementation of pellet-shaped bodies of porous sepiolitic micrite. U ncemented pellets of uniform composition i n fractures and other cavities must initially have grown in void space and later displaced each other as they
grew so as to fill the fractures. Examples are the pellets filling a solution cavity in the Wickieup calcrete, and the layer with turtleskin pellets in the Argus Range. As noted earlier, pellets in fracture fi llings are formed directly above an i mpermeable horizon, generally a laminar calcrete, and each new layer of pellets must displace the overlyi ng pellets upward, and the topmost pellets become cemented to the overlying limestone. A small percentage of pellets very likely displaced fine-grained matrix during growth. Micritic pellets in a matrix of detrital silty clay in a Kyle Canyon exposure presumably displaced the clay, as the clay had been deposited before the calcrete was formed. In some fracture-filling deposits of Kyle Canyon, pellets of pure micrite are rather uniformly dis persed in a matrix of homogeneous sepiolitic micrite. The pellets are unbroken and most likely grew in place by displacement of the sepi olitic micrite before it was fully cemented. A substantial proportion of pellets appears to have been formed by cementation of pellet-shaped bodies of porous sepiolitic micrite. As evidence, i n all three calcrete profiles are concentrations of pellets in a matrix of crumb aggregates which grade into structureless sepiolitic micrite through a zone in which outli nes of pellets are incompletely devel oped, and blocky fragments of matrix material may fit together with the adjacent pellets (Fig. 8a). We . i nterpret this transition zone as one in which pellets are in the process of being separated from the 61
Fig. 7. Sepiolitic micrite pellets and pelletal limestone from Unit 2 of the Argus Range calcrete. (a) Surface view of weakly
cemented pelletal limestone, which overlies a thin laminar layer near the top of Unit 2. Some of the pellets are coated by sepiolite and calcite and are classed as ooids. (b) Pellets of the 1·0-1·5 mm fraction sieved from uncemented pelietal sediment filling a subhorizontal fracture. Note the ovoid shape of many pellets. (c) Turtleskin pellets from an uncemented pelletal sediment filling a subhorizontal fracture. All scale bars are 2 mm.
Three calcretes of SW softer matrix along curved fractures (the 'circum granular cracking' of VVard, 1 975) as a result of bioturbation, gravitational shear, or by expansion and contraction on wetting and drying. Circum granular fracturing takes place along the contact of cemented and uncemented micrite and should be considered a result rather than the cause of the pellets. The clearest evidence for this mechanism is in the few examples where, in the transitional zone, the pellet margins appear darker and more highly cemented than the adjacent structureless micrite and clasts of micritic matrix. The matrix was further fractured, and clasts of the matrix material in some samples were rounded by accretion or solution to form small pellets. At some stage in this process, pore space in the matrix clasts was filled by calcite so that they became as thoroughly cemented as the larger, earlier formed pellets. Most and perhaps the vast majority of the small pellets, 0·02-0· 1 0 mm in size, are simply very small concretions formed by the same processes as the larger pellets. The uncemented pellets that grew in open cavities span a wide range in size (e.g. Table 3, Nos 2 and 3), and no basis exists for postu lating different origins for pellets of different size. Moreover, the large pellets must have been small pellets at an early stage of development. Small pellets were also formed by spotty cementation of porous micrite, as evidenced by concentrations of small pellets which grade into structureless micrite through a transitional zone similar to that described for the larger pellets. It is commonly difficult to assess whether a pellet grew as a separate entity, either in void space or by displacing adjacent particles, or by cementation of porous micritic sediment. Similar pellets can be produced by both mechanisms, and on the basis of field and petrographic data substantial volumes of pellets are attributed to both mechanisms. Pellets with detrital silt and sand deserve comment as they are unlike most calcareous concretions, which typically exclude silt and sand during growth. Pellets formed by cementation of ovoid bodies of porous micrite would be expected to retain the particles of silt or sand originally present in the micrite, and this mechanism may account for pellets with the highest contents of silt and sand. However, some other mechanism is required to account for the silt and sand in sepiolitic micrite pellets formed by accretion. Sepiolite fibres may have become attached to adjacent particles of silt and sand and held them in place during growth of the pellet.
U.S.A.
The ovoid pellet shapes are suggestive of faecal pellets, but this possibility is excluded by the inorganic composition of the pellets and the lack of a suitable soil-dwelling organism in the arid climates under which these calcretes were formed. The cause of the ovoid shape is unknown, but one possibility, which could be tested by SEM study, is that the long axis of the pellets reflects a preferred orientation of sepiolite fibres. Hydrological conditions may largely determine whether calcite-sepiolite mixtures are deposited as thin layers or as accretionary pellets. Accretionary pellets are typically formed directly above a laminar calcrete or other impervious horizon, possibly because they grow at least partly in the thin, tem porary perched water table after rains. Slighf differ ences in the hydrological environment or water chemistry may well drastically change the form of the precipitation, for pelletal layers may alternate with non-pelletal layers in laminar calcretes. It should be stressed that the proportions of sepiolite and calcite bear no relation to the morphology of the chemical deposit in samples which have been ana lysed (Table 1).
Ooids and pisolites
Ooids may have sepiolitic micrite pellets or detrital grains as nuclei (Fig. 8c and d), and they may have single or multiple coatings. Almost without excep tion, the coatings thin over corners and edges of angular nuclei so as to increase the roundness and sphericity of the particle. Average total thickness of the ooid coatings generally ranges from 0·02 to 0·10 mm. Coatings are generally concentric with regard to the nucleus except for opal-rich coatings of ooids in siliceous laminae of the Argus Range calcrete, which are commonly thicker on the bottom than on the tnp of the ooids (cf. Swineford et a/. , 1958). Ooid coatings most commonly consist of micritic calcite with either sepiolite or opal (Table 4) . In the Argus Range calcrete, however, they may lack calcite, whereas in the Kyle Canyon calcretes they may consist of micritic calcite alone. Individual laminae in the calcite-opal coatings are typically thinner and more numerous than in the calcite-sepiolite coatings. Thinly laminated calcite-opal coatings are attributed to primary deposition of silica together with calcite, rather than to replacement of sepiolite in calcite-sepiolite coatings. As evidence, the original thicker lamination is preserved unchanged where 63
.
Fig. 8. Pelletal and oolitic sediments as seen in thin section, using plane light. (a) Pelletal limestone from a filled fracture in calcrete of Kyle Canyon. Pellets were formed by cementation of pellet-shaped bodies of porous micrite, and this photograph shows an early stage in the separation of pellets from less-cemented matrix. Solid lines are drawn over clearly defined pellet margins, and dashed lines indicate poorly defined or incompletely developed pellet margins. (b) Pelletall sediment from the interstices of uncemented gravel overlying a thin laminar layer in the calcrete of surface 2 in Kyle:
64
Canyon. Pellets lie in a matrix of silt, sand, and micritic crumb aggregates. (c) Ooids from the calcrete near Wickieup. Ooid coatings consist of calcite and opal and thin over corners of angular nucleus in central ooid. Ooids lie in a patchy matrix of crumb aggregates. (d) Ooid with a pelletal nucleus from the Argus Range calcrete. The nucleus is sandy, sepiolitic micrite, and ooid coating consists largely of sepiolite. All bar scales are 0·5 mm.
65
R. L. Hay and
B.
Wiggins
Table 4. Composition and size of ooids
Locality*
Nucleus
Ooid coatings
Diameter
( l ) AR
Detrital grain
Calcite and sepiolite or opal ; thicker coatings have opal
Dom. 0 · 1 2-0-40 mm
(2) AR
Pellet
Calcite + sepiolite and/or opal
Dom. 1 ·0-1 ·5 mm
(3) Wk
Detrital grain
Calcite + opal
0·05-3 mm ; Dom. 0·5-2 mm
(4) Wk
Pellet
Calcite + sepiolite
Dom. 0· 5-0·7 mm
(5) KC
Pellet or detrital grain
Calcite, with or without clay
Dom. 0· 5-1 ·0 mm
* Abbreviations are the same as in Table 3.
sepiolite has been replaced by dense opal i n ooid coatings of the Argus Range calcrete, and no relict sepiolite has been identified in well developed ooids with multiple thin laminae of calcite and silica. Four types of ooid coating were analysed by the electron microprobe : ( 1 ) calcite-sepiolite (Table 2, No. 5) ; (2) relatively pure sepiolite (No. 6) ; (3) sepiolite-opal (No. 7) ; and (4) calcite-opal (Nos 8, 9). Types 1-3 illustrate the range i n chemical composition of ooid coatings in the Argus Range calcretes. The calcite-opal coatings are from the Wickieup calcrete, and the percentage of silica ranges from 1 6 to 35% in 1 4 analyses, the extremes of which are given in Table 2 (Nos 8 and 9). Analytical totals in the analyses range from 75·5 to 1 04·5% by inclu ding the C02 and H20 necessary for calcite and sepiolite. The low totals may reflect pore space or epoxy, and the one total greater than 100% very likely reflects the effect of microprobe standards differing from the minerals analysed. The smaller, well developed pisolites found i n concentrations o f ooids are similar to t h e ooids in all respects except size, and they rarely exceed 3 mm in diameter. Pisolites of the larger type are mostly between 0·5 and 2·0 em in diameter, irregular to subrounded in shape, and generally contain angular fragments of calcrete as nuclei. Coatings are typically thickest on the bottoms and thinnest on the tops of these pisolites, a feature observed in pisolites else where (e.g. Swineford et al., 1 958). Bottomside coatings commonly have an aggregate thickness of 4-6 mm. Most coatings consist of calcite with subordinate sepiolite, but layers of relatively pure sepiolite and opal are not rare in bottomside coatings of the Argus Range and Wickieup calcretes. Some of the topside coatings of Wickieup pisolites are pelletal, and pellets may be embedded in opal coating the bottom of pisolites in uncemented pelletal fracture fillings in the Argus Range calcrete.
The major difference between the ooids and the larger pisolites is in the different thickness and composition of the coatings between the top and bottom of the pisolites. This difference between the topside and bottomside coatings of the pisolite is probably largely an effect of gravity on the water film surrounding the pisolite. A major problem in ooid growth is why the coatings deposited around an angular nucleus should thin over corners and edges to create spherical particles. This feature was noted but not explained by James ( 1 972), Siesser (1973) and Braithwaite (1975). The ooids can be considered a type of 'quiet-water' ooid, yet the quiet-water ooids of marine environments are rarely spherical (Freeman, 1962). Coatings with opal or sepiolite may well have been deposited as a gel with sufficient strength for surface tension to stretch the coating to a more spherical shape. The purely micritic coatings at Kyle Canyon are more difficult to explain. In labora tory experiments some naturally occurring organic compounds including humic acid can cause aragonite to precipitate as ooids, with or without agitation (Suess & Fi.itterer, 1972 ; Davies, Bubela & Ferguson, 1978). The organic compound forms a membrane, which takes the shape of a sphere and may indeed be essential to the formation of marine ooids (Davies et a!., 1978). This mechanism does not seem appli cable to Kyle Canyon, in view of the sparse vegetation and generally oxidizing conditions that characterize soils of such regions. Alternatively, the structure of the ooids, either pure micrite or of mixed composi tion, is perhaps attributable to movements within the soil which cause the developing ooids to rub against each other, abrading the ooid coatings over the more exposed corners and edges. Another possi bility for thinning of the ooid coatings over corners and edges is intermittent solution during ooid growth. Precipitation of micrite should produce 66
Three calcretes of SW
U.S. A .
layers o f uniform thickness, b u t solution would be expected to remove a greater thickness over corners and edges because of the greater surface area per unit mass. Another problem is why ooid coatings as in the Wickieup calcrete should push grains apart rather than join them together with a meniscus cement into a rigid framework. This feature may be a result either of movements within the soil, which keep the ooids separated, or to a surface property, possible electrostatic, of the layers being deposited. Still another problem, most clearly exhibited in the Wickieup calcrete, is what determines whether calcite and silica will be deposited as ooid coatings or as thin layers of laminar calcrete. No satisfactory explanation is provided for this problem, which is essentially the same as what causes sepiolite and calcite to be deposited as pellets or as thin even laminae. A final problem is what determines whether pellets or ooids will be formed. Ooids consistently overlie laminar layers near Wickieup, whereas pellets overlie laminar layers in the equivalent part of the profiles in the Argus Range and Kyle Canyon calcretes. Composition of the materials being pre cipitated may well be the major control in these examples, for the Wickieup ooids are siliceous, whereas the pellets in the other localities are sepiolitic.
calcite with as much as 5% MgC03, is found in the Kyle Canyon calcrete. The sepiolite is a relatively pure Mg silicate clay which varies considerably in its d spacings and presumably in structure. It occurs as a matrix material and in pellets, ooids, layers, and veins. Opal can be either X-ray amorphous or cristobalitic, and it can occur as a cement and in layers, nodules, veins, and ooid coatings. It replaces sepiolite and less commonly calcite. (3) The vast majority of pellets consist of sepiolite and calcite and represent small concretions which grew in some places by accretion, either in void space or displacing adjacent sediment, and elsewhere by cementation of pellet-shaped bodies of porous micritic sediment. Pellets are typically ovoid for reasons not yet known. (4) Ooid coatings are chiefly mixtures of calcite and opal or calcite and sepiolite. The coatings thin over edges and corners of angular nuclei so as to increase the roundness and sphericity of the particles. Coatings with opal or sepiolite may have been deposited as a gel with sufficient strength for surface tension to stretch the coating to a more spherical shape. (5) As yet unknown are the factors which deter mine whether calcite-sepiolite mixtures are pre cipitated as laminar layers or as accretionary pellets, and whether calcite-opal mixtures are precipitated as laminar layers or as ooid coatings.
CONCLUSIONS
A C K N O W L E D GM E N T S
(1) Pellets and ooids are widespread and locally abundant in three mature calcrete profiles developed on materials of different composition : basaltic rocks in the Argus Range, California ; silicic volcanic rocks near Wickieup, Arizona ; and carbonate alluvial deposits in Kyle Canyon, Nevada. Pellets and ooids are most common within subhorizontal fractures and above laminar layers. In all these profiles the petrocalcic horizon has been thickened by addition of chemically deposited pelletal sediment in subhorizontal fractures. Indeed, · on steeper slopes in the Argus Range, a dense petrocalcic horizon 1 m thick was formed largely by chemical precipitation along fractures. (2) Calcite and sepiolite are common authigenic materials in all three calcretes, and opal is common in the calcretes developed on volcanic rocks. Most of the calcite has Jess than 3 mol % MgC03, but
This study was supported b y the National Science Foundation (Grants EAR 76-84583 and EAR 78-01 776) and the University of California. We are indebted to Lung S. Chan and Mark Rivers for microprobe analyses, to Don Harvey and Joachim Hampel for photomicrographs, and to Peter Daugherty for thin sections and probe mounts. R. J. Reeder assisted in the reconnaisance study in 1 977 and stimulated and clarified our understanding of calcrete pedogenesis throughout the subsequent phases of the work. Janet Sowers discovered the silicified pelletal sediment in Kyle Canyon, and her work there, still under way, has increased our understanding of the calcrete stratigraphy. We thank T. E. Cerling, P. A. Sandberg and R. L. Gresens for fruitful suggestions. The manuscript was greatly improved by the rigorous reviews of M. Talbot and C. J. R. Braithwaite. 67
R. L. Hay
and B.
R.L. & REEDER, R.J. (1978) Calcretes of Olduvai Gorge and the Ndolanya Beds of Northern Tanzania. Sedimentology, 25, 649-673. JAMES, N.P. (1972) Holocene and Pleistocene calcareous crust (caliche) profiles : criteria for subaerial exposure:. J. sedim. Petrol. 42, 8 1 7-836. LATTMAN, L.H. (1 973) Calcium carbonate cementation of alluvial fans in southern Nevada. Bull. geol. Soc. Am. 84, 301 3-3028. MILLOT, G., NAHON, D., PAQUET, H., RUELLAN, A. & TARDY, Y. ( 1 977) L'Epigenie calcaire des roches silicatees dans les encroutements carbonates en pays subaride Antiatlas, Maroc. Sci. geol. Bull. 30, 1 29-152. MuLTER, H.G. & HOFFMEISTER , J.E. ( 1 968) Subaerial laminated crusts of the Florida Keys. Bull. geol. Soc. Am. 79, 1 83-192. PAPKE, K.G. ( 1 972) A sepiolite-rich playa deposit in southern Nevada. Clays Clay Miner. 20, 21 1-2 1 5. REEVES, C.C. (1 976) Caliche ; Origin, Classification, Morphology, and Uses. Estacado Books, Lubbock, Texas. SIESSER, W.G. ( 1 973) Diagenetically formed ooids and intraclasts in South African calcretes. Sedimentology, 20, 539-5 5 1 . SUESS, E . & FDTTERER, D. ( 1972) Aragonitic ooids : experimental precipitation from sea water in the presence of humic acids. Sedimentology, 19, 1 29-1 39. SoiL SuRVEY STAFF (1 975) Soil Taxonomy. Soil Conserv. Serv., U.S. Dept. Agr., Agricultural Handbook 436, U.S. Government Printing Office, Washington, D.C. SwiNEFORD, A., FRYE, J.C. & LEONARD, A.B. ( 1958) Petrology of the Pliocene pisolitic limestone in the Great Plains. Bull. Kans. geol. Surv. 130, 97-1 16. VANDEN HEUVEL, R.C. (1 966) The occurrence of sepiolite and attapulgite in the calcareous zone of a soil near Las Cruces, New Mexico. Proc. 13th Nat. Conf on Clays and Clay Min . , pp. 1 93-207. WARD, W.C. (1975) Petrology and diagenesis of carbonate eolianites of northeastern Yucatan Peninsula, Mexico. In : Belize Shelf: Clastic Sediments, and Ecology (Ed. by K.F. Wantland and W. C. Pusey III). Am. Ass. Petrol. Geol. Stud. Geol. 2, 500-571 .
REFERENCES
HAY,
G . O . & MACHETTE, M.N. (1 977) Calcic soils and calcretes in the southwestern United States.
BACHMAN,
Open-file Report U.S. geol. Surv. 77-794. R.G.C. (1975) Carbonate Sediments and Their Diagenesis, 2nd edn. Developments in Sedimentology,
BATHURST,
12. Elsevier Publishing Co., Amsterdam.
C.J.R. (1975) Petrology of palaeosols and other terrestrial sediments on Aldabra, western Indian Ocean. Phil. Trans. R. Soc. Lond. 273, 1-32. CAILLERE, S. & HENI N, S. (1961) Sepiolite. In : The X-Ray
BRAITHWAITE,
Identification and Crystal Structure of Clay Minerals
(Ed. by G. Brown), pp. 325-342. The Mineralogical Society, London. DAVIES, P.J., BUBELA, B. & FERGUSON, J. (1978) The formation of ooids. Sedimentology, 25, 703-730. FREEMAN, T. (1962) Quiet water oolites from Laguna Madre, Texas. J. sedim. Petrol. 32, 475-483. FRYE, J.C., GLASS, H.D., LEONARD, A.B. & COLEMAN, D.D. (1 974) Caliche and clay-mineral zonation of Ogallala Formation, central-eastern New Mexico. Circular N. Mex. Bur. Mines Miner. Resour. 144,
1-1 6 .
L . R . ( 1972) Origin o f the Mormon Mesa Caliche, Clark County, Nevada. Bull. geol. Soc. Am. 83, 143-1 56. GILE, L.H., PETERSON, F.F. & GROSSMAN, R.B. (1965) The K horizon : a master soil horizon of carbonate accumulation. Soil Sci. 99, 74-82. GILE, L.H., PETERSON, F.F. & GROSSMAN, R.B. (1966) Morphological and genetic sequences of carbonate accumulation in desert soils. Soil. Sci. 101, 347-360. GIOVANETTI, D.M. ( 1 979) Volcanism and sedimentation associated with the formation of southern Owens Valley, California (abstr.) Abstr. Progm. geol. Soc. Am. 11, 79. GoLDSMITH, J.F. & GRAF, D.L. (1 958) Relation between lattice constants and composition of the Ca-Mg carbonates. Am. Mineral. 43, 84-10 1 . GooLSBY, J.E. (1975) Cenozoic stratigraphy and geo GARDNER,
morphology of Lynn
and
Terry
Wiggins
counties,
Texas.
Unpublished Ph.D. Thesis, Texas Techno!. Univer sity.
(Manuscript received 1 3 August 1 979 ; revision received 28 October 1 979)
68
Reprinted from Sedimentology (1980) 27 661-686
Quaternary pedogenic calcretes from the Kalahari (southern Africa): mineralogy, genesis and diagenesis
N. L. W A T T S Konink/ijke/Shell Exploratie en Produktie Laboratorium, Vo/mer/aan 6, Rijswijk (ZH), The Netherlands
A BS T R A C T
The calcretes o f the Kalahari are amongst the thickest i n the world representing pedogenic episodes i n a semi-arid climate during Pliocene t o Recent times. The descriptive morphological terminology of Netterberg is used to describe the calcrete types and a differentiation into simple and composite profiles is made. A pedogenic/diagenetic scheme has been constructed using all available data. Early calcite cementation is induced by two mechanisms. Slow evaporation and/or C02 loss causes the precipitation of low-Mg calcite, whereas rapid evaporation and/or C02 loss precipitates predomi nantly high-Mg calcite, in thermodynamic disequilibrium with the low Mg/Ca ratio vadose water, and minor aragonite (which transforms rapidly to low-Mg calcite). High-Mg calcite is also precipitated from high Mg/Ca ratio vadose waters in calcretes developed on Mg-rich host lithologies and by capillary rise from shallow groundwaters in saline depressions. Calcite precipitation may be passive (cement), displacive or replacive, in the latter released silica migrating down-profile to precipitate length-slow chalcedony, clinoptilolite (saline conditions), length-fast chalcedony and megaquartz (non saline conditions). Displacive introduction of calcite takes place from highly supersaturated solutions due to rapid evaporation (with C02 loss) of vadose waters. During low-Mg calcite precipitation (in a 'closed' system) the Mg concentration of the resulting solution increases. This, combined with Mg released during high-Mg to low-Mg calcite transformation, induces precipitation of authigenic palygorskite, sepiolite and minor dolomite. Vadose dolomite is often present whilst some dolomite may be precipitated at the mixing-zone of vadose and phreatic waters. The proposed model applies to the Kalahari calcretes, although it may perhaps be extrapolated to other areas. Further detailed studies, involving analyses of pore water chemistry, soil microclimate, and trace element and iso topic analyses of individual cements are necessary.
INTRODU C T ION AND GENE R A L
Calcretes have long been recognized in South Africa (Du Toit, 1954; Mountain, 1967 ; Netterberg, 1 967, 1969a, b, c, 1971), but those of Botswana have received little attention. The region was chosen for study because of the wide variety of well exposed calcrete types. Although the best sections are found in river gorges (e.g. Molopo, see Fig. 1) and terraces, numerous ' borrow pits' are present, calcrete usually being used as a road surfacing material. These pits (mainly in Botswana) are spaced at regular intervals along the roadsides exposing sections up to 5 m deep. Recently, Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
SETTING
further excavations in calcrete have been made for diamonds, but access to these is limited for security reasons. Botswana is an area of inland drainage and is dominated by the Kalahari Desert which extends from the Northern Cape Province in South Africa, through Botswana and into the Congo Basin. The climate is semi-arid with annual rainfall nearly al ways less than 500 mm. Although this is a high precipitation for calcrete-forming areas, extremely high rates of evaporation cause a large rainfall deficit. Temperature ranges are large, from near 0 to 69
N. L. Watts 4°C at night to approximately 25°C during the day. Generally, the southeast is the wettest area with the southwest and northwest becoming quite arid. The Kalahari generally lies 850-1000 m above sea level and is drained towards its own interior by the ephemeral Botletle, Nata and Okwa rivers. Inselbergs are rare but occur in the extreme northwest and around Dibete in the southeast. With the exception of salt pans, occasionally dry river valleys and vague undulose ridges (old dune fields), the Kalahari is remarkably flat. The geomorphology of the area has been shown by Grove ( 1 969) to retain evidence of Tertiary climate and physiography, the climate being at different periods both wetter and drier than at present. Studies by Cooke ( 1 975) in northwestern Bots wana have shown that a simple dry-wet-dry-present sequence of climatic history may only be an approxi mation. His analyses of cave sediments, tufa and calcretes have suggested that rapid alternations of climate have occurred over the last tens of thousands of years. In a well reasoned account he has proposed a relationship between climatic (and tectonic) events and sedimentary processes and their deposits ; he correlates his cave tufas and sinters with 'wet' periods, calcretes with semi-arid episodes, and sil crete formation and dune-migration with arid times. Although not in total agreement, there is a similarity between Cooke's findings and those of Netterberg ( 1 969c), who studied calcrete age in southern Africa (see below). The present climate, although not 'wet', is more akin to a 'pluvial' than a dry period. Active dune migration is only seen in the extreme southwest, and most of the rivers flow for at least part of the year. Degradation of calcretes by solution is taking place, particularly in the southeast and the region is covered by a sparse vegetation dominated by shrubs and grasses. The southern and central Kalahari is characterized by Thornveld vegetation (Wellington, 1 955), whilst in the north Mopaniveld is dominant. The distribution of vegetation in southern Africa, although related to climate, aspect and drainage, is particularly sensitive to host sediment type and has enabled differentiation of calcrete areas from aerial photography (e.g. Mountain, 1 967). This paper examines the mineralogy of a number of calcrete profiles from the Kalahari (see Fig. 1 ). Most are of Middle Pleistocene to Recent age, and provide important information relevant to the understanding of early diagenesis in calcretes. Additional details are given by Watts ( 1 977b). Sample numbers pre-
fixed 'S' in the text and on Figs 4-1 0 refer to speci mens deposited in the Archive collection of the Department of Geology, University of Reading.
AGE OF CALCRETES
Although Goudie ( 1 973) and Watts ( 1 977a) record Quaternary calcretes in Botswana, it is possible that calcrete formation has been taking place episodically since at least the Pliocene (King, 1 963). Mabbut (1 952) recorded calcrete development on schists which were overlain by Pliocene sediments. This led Netterberg ( 1 969c) to speculate that calcrete older than Pliocene must exist in southern Africa. The latter author has provided a detailed evaluation of the late Cainozoic climatic history of southern Africa, and has suggested that four major episodes of calcrete formation have occurred during ( 1 ) Pliocene times and includes the Mid-Pliocene 'Kalahari Limestone' ; (2) Acheulian or Middle Pleistocene times; (3) Upper Pleistocene (or Fauresmithian) times. The precise ages of these episodes are, how ever, not well known. Some of the calcretes may correspond to the oldest forms of Cooke ( 1 975), whilst Netterberg's 4th episode (Late Pleistocene Recent) seems equivalent to Cooke's younger two calcretes. The Recent calcretes include <;>nly two known hardpans, suggesting that widespread hard pan formation has not occurred since the Pleistocene (Netterberg, 1 969c, p. 92). As Netterberg has sug gested, calcretes must be forming at present in southern Africa ; the problem is to prove it.
T E R MINOLO GY
Considerable confusion exists over the use of the term 'calcrete'. The calcretes described here are entirely pedogenic, and although some non-pedo genic types do occur, they have been omitted from the present discussion. The term calcrete is used here in a similar sense to that of Netterberg (1969b) in that various types of calcrete occur (nodular, honeycomb, hardpan) related to stages of calcrete formation and hardpan calcrete (the 'true' calcrete of some workers) is merely an advanced stage of this process. As some calcretes are uncemented (e.g. powder calcret\!s ; Netterberg, 1 969b, 1 97 1 ) whilst not all vadose cemented sands are calcrete, the definition of Goudie 70
Kalahari calcretes: origin and diagenesis common i n the relatively high rainfall zone of eastern Botswana, such as at Artesia, Mookane and east of Francistown, whilst silica-toughened calcretes are common in the highly saline Makgadikgadi de pression (Fig. 1 ) . The Kalahari calcretes are probably amongst the thickest in the world (see Goudie, 1 973, p. 33 and Fig. 2) with calcrete sequences often being in excess of 50 m thick (oral communication, Botswana Geological Survey ; Goudie, 1 973, table 1 5) . Simple profiles are generally < 2·5 m thick but may be thicker where sediment accretion has been constant and sufficiently slow. Borehole data indicate com posite calcretes at least 3·0 m thick, with most being greater than 6·0 m. Some thick composite calcretes in the south (some are non-pedoge n"ic) may be Pliocene but most of the intermediate thickness forms (6- 1 5 m) are probably Quaternary. A variety of logged calcretes are shown in Fig. 3(a). Numerous authors have suggested a relationship between calcrete thickness and geomorphological position, calcrete generally thinning with increased slope or over positive areas (e.g. McFarlane, 1 975). Although here laminar calcrete thickness generally increases adjacent to salt pans (e.g. at Mopipi, Fig. I), there appears to be no systematic variation with surface relief. This may reflect poor exposure and difficulties in correlation, but detailed work on this problem is needed and would prove invaluable in environmental interpretations of ancient calcretes. In some cases there may be an increase in calcrete thickness adjacent to river channels, reflecting the importance of ground-water recharge through valley sides during calcrete formation. Goudie ( 1 973, pp. 35-36) has shown that calcrete thickness in the southern Kalahari may be related either to zones of calcareous dust deposition or to the valleys of the Kuruman and Molopo rivers which drain carbonate rich areas. Many of these calcretes, however, may be non-pedogenic and analogous to the 'valley calcrete' of Carlisle ( 1 978). Although some calcretes are developed in and on bedrock (see below), the majority occur in superficial Kalahari sand. The thickness of sand overlying the calcretes is rarely greater than 30 m and is usually less than 3 m (Grove, 1 969, p. 1 95). The sands are mostly fine to medium in grain size, are usually well sorted, and consist of sub-angular to sub-rounded grains. The dominant mineral is quartz with subsidiary feldspar, rock fragments and biotite. Occasional silcrete fragments and grains of calcite and dolomite occur with, on a larger scale, sporadic blebs and
( 1 973) has here been rewritten to read, 'pedogenic calcretes are terrestrial materials composed domi nantly, but not exclusively, of CaC03, which occur in states ranging from nodular a nd powdery to highly indurated and result mainly from the displacive and/or replacive introduction of vadose carbonate into greater or lesser quantities of soil, rock or sediment within a soil profile'. The terminology used is largely geological with specific soil science micro-morphological nomen clature (e.g. Brewer, 1 964) being given where neces sary. The distinction between high- and low-Mg calcites is taken as 4·0 mol% MgC03 (after Chave, 1 952) and not 8 ·0% as proposed later by Winland (1 969). Finally, in order to discuss the interrelationships between various calcrete types, calcrete profiles have been divided into simple and composite forms (in the sense of Allen, 1 974, p. 1 1 3). Although the division is intended to be descriptive, aspects of calcrete genesis in terms of maturity (e.g. Gile, Peterson & Grossman, 1 966 ; Goudie, 1 973) must be considered. Simple profiles represent a single episode of cal crete formation, show a variety of stages of maturity and are unaffected by later periods of calcrete formation. They are therefore discrete units and are separated from overlying and underlying profiles by uncalcretized material. Composite profiles represent recurring episodes of calcrete formation. They may be identified by a stacking, with overlap, of individual calcrete profiles. Where sediment accretion takes place above a simple profile and renewed calcrete formation affects the underlying profile an 'upward composite profile' is produced. Renewed calcrete development, either after no net sediment accretion (rare) or net defla tion produces a 'downward-composite profile'.
C A L C R E T E T Y P E S , DIS T RIBU T I O N , HO S T M A T E R IA L S A N D THI C K N E SS
The calcretes of the Kalahari possess a wide range of physical characteristics, ranging from powdery states, through nodules, to a honeycomb form, to dense hardpans with lamellar rinds, to solutionally disrupted and extremely dense boulders. In compo site profiles several types of calcrete or, indeed, several examples of the same type of calcrete, may be found (Figs 2 and 3). There are some regional differences : thus, boulder calcretes are particularly 71
N. L. Watts z:zo
.-·-
-·
+
RJ/Of)ESIA '·
·"·
I ,4PA NILLJ' I
+2o0
I I
SE-'?oWE LELJC/NGE
BOTcJ'W'ANA
• •PA . LA/'YE
•
• MtffAMABlli.A
MANALAP.Y£
(i}::+ f.·,
,' . SERt/.Rt/M£ YAi.i..E)" -..._,'OIBETE
•;
MACJ./.ANEN6
MOOK.dNE
+24°
GABERONE
.KHAKEA
;·'L08ATJ'E P�ETO.(!IA
I
A;fAFEK!NG
260 KA'U6£.NStlOA'P.+•JoJ.IANNESBl/1?6
IJEi..4JlEY¥1Li.£ •KLE.RKSOO�P
' I
r \ '
OF SOVTII AFI?/C.4 IJLOEMHO.F
OLO COVRS£
(..,...-OF MOJ..OPO
+28°
0
+
20°
50
100
2ookm.
+
24°
Fig. 1. Location map of the field area showing the main towns and rivers. The central part of the area is largely inacces sible without suitable vehicles. Numbers 1-4 refer to the localities of the sampled profiles shown in Figs 7-10.
stringers of brown and green clay. In general, these are sub-litharenites (Pettijohn, Potter & Siever, 1 973), but examination by XRD and cathode lumi nescence (J.A.D. Dickson, personal communication) suggests a greater proportion of feldspar (up to 20% in some cases, but usually around 1 0%), which would place the sands into the sub-arkose class. The red colour of the sands is due to haematitic
and limonitic coats (ferrans-terminology of Brewer, 1 964) around the detrital grains. This iron is attributed to in situ weathering of ferro-magnesian minerals (e.g. biotite, hornblende), the released iron migrating to grain boundaries in the manner outlined by Walker (1 967), Walker & Honea (1 969) and others (see review by Van Houten, 1 973). Sands within calcretes generally lack iron coatings suggesting that either the 72
Kalahari calcretes: origin and diagenesis
Fig. 2. Extremely thick, composite calcrete profile, with multiple hardpan development, overlying Dwyka Tillite. A vertical log through the profile is shown in Fig. 3 (sample numbers 26-34b), as is a laterally equivalent section 200 m to the west (sample numbers 54-67). Most of the individual profiles are pedogenic with some non-pedogenic groundwater-influenced calcretes near the base. Locality: Molopo River, south side of Gorge, near Bogogobe (Fig. 1 ) . Author (arrowed) i n foreground for scale.
iron has been removed during the introduction of carbonate or that iron migration post-dates calcite crystallization ; petrographic evidence is inconclusive here. The sands are treated in more detail by Baillieul (1 975), most of the sands examined being from within his Type 2 forms. Other host lithologies examined included tholeiitic dolerite, kimberlite, silcrete, ostracod-rich silice:ms pan sediment and various red and green clayey siltstones. With calcrete formation, host rock texture and mineralogy is often lost. Calcrete distribution is thus not merely a reflection of geomorphological factors. Climate, host material, sediment accretion rate, carbonate source (and its availability), vegetation and time all seem to play important roles in calcrete formation and may
consequently greatly influence the form, type and thickness of the calcrete produced. A N A LY T I C A L P R O C E D U R E S
A large number of calcrete profiles were logged,
sampled and subsequently analysed using stained thin sections, scanning electron microscopy (SEM) and X-ray diffraction (XRD). Tables listing all the data and procedures of analysis are given in Watts ( 1 977b). A description of the procedures involved in the quantification of the clay mineralogy is given below. Quant ificat ion o f clay m ineralogy
Accurate quantification of the clay data was found 73
N. L. Watts (a )
- 34A 39-9*
-348 15
-67
15
-66 - 65
-64
- 33A
- 63 -338 -25
---=�
- 38
479
*
* 2s.9 '
-62 -51 -50
'
"' "' "' .... "' :1
� "' .... "
"' u
<:Jo<S'--,() (0
_: a\ -:----..:._
-37
-24
10
-61
-49
-60
-48
0
59
0
0
\
0
-58
-23
�s·
5
,.J o� .:--- ) . -.. -0
0
0
-46
-22
-57
-36 -56 -21 -45
�6 50
100
50
-44 -43 -42 -41
-35
671* 100
50
100
50
100
- 55 - 54 50
100
PERCENTAGE CARBONATE (FIELD ESTIMATE) Fig. 3. (a) Selection of logged calcrete profiles (simple and composite) from the southern Kalahari. Key to all logs is shown in Fig. 3(b). Localities : profile 21-25: borrow pit on west side of road, J'hephane Valley, just above confluence with Molopo, 21 km from Yorsterdorp; profile 26-34B (see Fig. 2); profile 35-38: cliff exposure on Precambrian quartzite, south side of Molopo Gorge, near Bogogobe; profile 4 1 -5 1 : cliff exposure, south side of Molopo Gorge, 8 km west o f Bogogobe; profile 54-67 (see Fig. 2 ) . Numbers with asterix are Schmidt Hammer readings taken at outcrop, a n d other numbers refer to field sampling locations.
14 A heated chlorite has twice the intensity of illite, whilst kaolinite has an intensity x 2·5_ Percentage mixed-layer clays were calculated using the techniqw� of Weir et at. (1 975, P- 380), and palygorskite and sepiolite were assumed to have relative intensitie!) of 3 : 1 respectively [although Weaver & Beck ( 1 977) have shown how the scattering power of montmoril-
to be extremely difficult in most cases. A number of methods and intensity conversion factors for various clays have been described in the literature (e.g. Schultz, 1 964 ; Collins, 1 976), often involving de tailed measurements, numerous assumptions and complex calculations. In this study the method of Weir, Ormerod & Mansey ( 1 975) was followed ; e.g. 74
Kalahari calcretes: origin and diagenesis
KEY TO FIGURES
(b)
CALCRETE LOGS
Nodular calcrete
Brecciated nodules
Hardpan calcrete
Honeycomb calcrete
� � ��(91 jUJ I
� ��0 j � ' ' ""'"
�
Laminar calcrete
Silcrete
Pseudo anticlines
Ill
CLAY MINERALOGY
� � D
Rootlets
Palygorskite
Burrows etc
Sepiolite
Lithoclasts
Montmorillonite
Brecciated horizons
Chlorite
Soluble salts
Mixed-layer clays
Crystallaria
Illite
Replacement by
Kaolinite
II
Clinoptilolite
M �
carbonate
�
I===I D
Fig. 3. (b) Calcrete and clay mineralogical symbols used in the logs in this paper.
lonite to palygorskite varies with differing propor tions of the two minerals]. Where mixed-layer clays and palygorskite were present together, the relative intensity of glycolated to heat-treated 1 0 A pure palygorskite was used to counteract the heated palygorskite 1 0 A influence on the heat collapsed mixed-layer peak ; the mixed-layer percentage was then calculated using the method of Weir et al. ( 1 975). Although the results are plotted as distinct per centages (the intensity values were standardised and expressed as percentages), the errors involved in the calculations are such that mainly trends of clay dis tribution are obtained. Even using the method of Weir et al. (1 975), in the absence of palygorskite, montmorillonite and chlorite, a considerable error is obtained: e.g. Austin & Leininger (1976) have point ed out that, when using oriented samples of sedi mented clays, calculations based on heat treatments
should be considered with caution. As the sample preparation and treatments were kept constant for each profile, clay mineral trends are thought to be fairly accurate. Finally, where palygorskite was absent, but chlorite, kaolinite, illite and mixed-layer clays present, considerable difficulties arose. To find the relative amounts of chlorite and kaolinite, acid treatment was employed, the intention being to calculate the amount of mixed-layer clays using kaolinite as a standard, and the chlorite percentage obtained by measurement and recalculation of the heated 1 4 A peak (after Weir et al., 1 975). However, because with acid treatment some chlorites are less easily removed than others, and therefore the exact proportions of kaolinite and chlorite could not be calculated, clay mineral distributions in such profiles were determined by eye, and plotted as trends.
�
75
N. L. Watts D ET A I L E D M I N E R A L O GY
Palygorskite Palygorskite (also known as attapulgite) has a fibrous habit. Bradley ( 1 940) suggested that the mineral consists of a 2: 1 layer structure with five octahedral positions, four of which are filled. On either side of the octahedral sheet are four silica tetrahedra, the structural units alternating in a regular pattern producing a series of channels be tween the units. The ideal formula for the half-unit cell is Si8Mg5020(0H2)4-4H20 (Weaver & Beck, 1 977). Although palygorskite is relatively rare, it forms in a variety of environments. For example, it has been ascribed to a hydrothermal origin (e.g. Furbish & Sands, 1 976), sometimes with marine influences (Bonatti & Joensuu, 1 968), and to .deposition in lagoons (and various brackish environments, e.g. Weaver & Beck, 1 977), lakes and playas (Parry & Reeves, 1 968a) and true marine environments (e.g. Millot, 1 970 ; Couture, 1 977). Weaver & Beck ( 1 977), however, believe that many of the so-called 'marine' palygorskites are detrital (but see Couture, 1 978). Palygorskite is common to many modern and ancient calcretes (e.g. Vanden Heuvel, 1 966 ; Strakhov, 1 970, p. 1 3 ; Lang & Pias, 1 97 1 ; Netter berg, 1 97 1 ; Gardner, 1 972 ; Menillet, 1 975 ; Nahon & Ruellan, 1 975 ; Nahon, 1 976 ; Reeves, 1 976) and in other semi-arid soils (e.g. Beattie & Haldane, 1 95 8 ; BJorn, 1 970 ; Gupta & Raychauduri, 1 972 ; Dobrovol'skiy, 1 973). Its origin has been ascribed to neoformation (e.g. Singer & Norrish, 1 974 ; Eswaran & Barzanji, 1 974 ; Watts, 1 976), alteration of mont morillonite (e.g. Yaalon & Wieder, 1 976 ; Weaver & Beck, 1 977) and to inheritance fro m palygorskite rich materials (e.g. Singer, 1 97 1 ; Aba-Husayn & Sayegh, 1 977). lsphording ( 1 973) and Singer ( 1 979) provide a summary of ideas related to the origin of palygorskite-rich deposits. Palygorskite was found in all the Kalahari calcrete profiles analysed. Although occasionally absent, in some cases the mineral constituted over 90% of the less than 2(-lm acid-insoluble fraction and up to 6% of the bulk samples. Generally speaking, it would seem that the palygorskite content of the whole rock insoluble fraction is much greater than the equivalent < 2 !-lffi fraction would suggest, confirm ing the dominantly large crystal dimensions of the palygorskite fibres and their relatively high settling velocities (in sample preparation) over platey minerals. The mineral occurs as dense mats of interwoven fibres (up to 1 !-lffi wide, and 50 !-lffi long) lining
Authigenic s ilicates
The Kalahari calcretes are dominated by the clays palygorskite and sepiolite, with subsidiary mont morillonite, mixed-layer clays, illite, glauconite, chlorite and kaolinite. A series of X-ray diffraction traces from samples within a composite calcrete pro file is shown in Fig. 4. In this sub-section a discussion of each authigenic silicate mineral identified is pre sented as an aid to later interpretation of the results. K
S24813 S24812 S24811
S24806
7 6 Degrees 20
5
4
3
2
Fig. 4. X-ray diffraction traces (glycolated) of oriented
< 2 (-lffi mounts of samples from profile S24805-S248 1 3 (see Fig. 7 ) . Range o f major reflections o f each clay shown at top of figure. K, kaolinite; I, i11ite; P, paly gorskite; S, sepiolite; M, montmorillonite. All samples run under the same conditions. Note absence of paly gorskite and sepiolite from sample S248 1 3 (uncalcretized Kalahari Sand) and increasing 'raggedness' of mont morillonite peak with increased palygorskite and/or sepiolite content.
76
Kalahari calcretes: origin and diagenesis !J.ffi fraction) (samples
S24807, S2481 8, S24843, S24844), although values between 20 and 50% are more common in mature simple profiles.
solutional channels, ped surfaces and sometimes skeletal grains dominantly in mature calcretes. Longer, coarser fibres form loosely woven mats, whilst smaller fibres form composite sheets similar to those described by Yaalon & Wieder (1 976) and Nahon (1 976, fig. 1 9). Palygorskite is rarely associa ted with calcite, although dolomite rhombs are commonly coated with fibres (see also Weaver & Beck, 1 977, figs 65, 66). In rare cases (e.g. Fig. 5d) the palygorskite fibres are coated with a thin sheath of micrite (note similarity to some 'calcite fibres' described from some South African calcretes by Knox, 1 977), but in the majority of cases palygorskite is later than the calcite matrix (e.g. Fig. 5c) and no etching by calcite has been observed (cf. Millot et al., 1 977).
Montmorillonite Ideas on the genesis, chemistry and structure of montmorillonite (and additionally on illite, chlorite, kaolinite and mixed-layer clays) are discussed i n Millot (1 970). Montmorillonite in weathering pro files may be detrital, or produced by weathering of primary minerals or neoformation (e.g. Singer & Arnie!, 1 974), migrating during pedogenesis by dispersion, physico-chemical fractionation and leac hing (Millot, 1 970, p. 1 08). In Kalahari calcretes abundant montmorillonite occurs mainly in immature calcretes, lower portions of calcrete profiles and unaltered Kalahari sand (sample S248 1 3 , Fig. 7). Where montmorillonite is the dominant clay it is particularly well crystallized, but becomes less crystalline and 'ragged' (on XRD traces) with increasing proportions of palygorskite and sepiolite. In SEM photomicrographs and thin section i t appears as diffuse amorphous to fluffy coats on skeletal grains and in some early channels (illuviation argillans), its early nature being revealed by disrupted montmorillonite grain-free argillans and papules (Fig. 5b).
Sepiolite Sepiolite is similar to palygorskite in structure, but is composed of three pyroxene-like chains rather than two. This led Nagy & Bradley (1 955) to suggest that it was amphibole-like in character but represented a new type of silicate structure. It conforms to the general chemical formula Mg3(Si4011)H20·1 1 H20 (Theodorovitch, 1 9 6 1 ), although higher Mg contents are possible (see Caillere & Henin, 1961). Like _p_alygorskite, with which it is so often asso ciated, s�piolite is characteristic of Mg-rich alkaline environments and has been described from lacustrine or playa-lake deposits (e.g. Parry & Reeves, 1 968b) having formed by detrital and chemical sedimenta tion (i.e. Millot, 1 970, pp. 1 72-1 74), marine environ ments (see Millot, 1 970, pp. 1 99-200 for review) and arid-zone soils (Vanden Heuvel, 1 966 ; Reeves, 1 976), mainly as a result of neoformation (e.g. Millot, 1 970 ; Weaver & Beck, 1 977), although Lacroix ( 1 941) has described sepiolite forming by dissolution of calcite. Sepiolite is often associated with palygorskite in the Kalahari calcretes, both minerals appearing as shiny films on fresh fracture surfaces. Sepiolite, however, is more restricted and predominates in mature to very mature calcrete, often associated with dolomite. In SEM photomicrographs it occurs as loosely interwoven fibres (usually thinner and longer than palygorskite) blanketing skeletal grains, chan nel walls or ped surfaces (Fig. Sa), often associated with small, randomly distributed, authigenic dolo mite rhombs. Dense composite sheets of sepiolite (like those of palygorskite) and calcification of fibres were not seen. In composite profiles individual samples may contain over 80'%; sepiolite (of the <2
Mixed-layer clays Mixed-layer clays are composed of regular to random interlayered mixtures of a variety of primary clays and are common in weathering profiles. In Kalahari samples mixed-layer clays are dominantly illite/ montmorillonites, and are common in immature profiles and in the upper part of profiles in which montmorillonite is abundant. They seem to occur in identical situations to montmorillonite (and illite) and this, plus their gradation in some XRD traces into illites, suggests they have formed by partial weathering of illite (see Robert et al., 1 974). Illite Illite has been described from virtually every sedi mentary environment. However, although authigenic i llite does occur and has been synthesized at surface temperatures and pressures (e.g. Harder, 1 974 ; Eberl & Hower, 1 976), most illites in soils are detrital (see Millot, 1 970). Illite is present in some Kalahari calcrete profiles in both well crystallized and degraded (to mixed layer clay) forms. Textural evidence (SEM studies) seems to indicate that illite is dominantly detrital, 77
N. L. Watts
Fig. 5. Clay minerals. (A) SEM photograph showing a loose mat of interwoven sepiolite fibres coating an irregular void wall composed of calcite. Sample S2481l. Bar is 40 f.l.m long. (B) Thin section photomicrograph (plane light) showing three quartz grains (Q) 'floating' in a calcite matri�. Note grain-free montmorillonans (M) and disrupted montmorillonitic papule lP) illustrating early nature of montmorillonite. Bar is 1 00 fJ.m long. (C) Thin section photo micrograph (crossed polars) showing a circumglaebular palygorskan. A dense clotted calcite glaebule (G 1) is coated with well oriented palygorskite (lower arrow) which connects with a channel or ped palygorskan (upper arrow). Relation ship suggests palygorskite formation after calcite. Sample S24828 . Bar is 0·25 mm long. (D) SEM photograph showing an irregular mass of palygorskite fibres coating and filling a void within dense micrite. Note slight calcification of the fibres bv micrite. Sample S24839. Bar is 32 fJ.m long.
but percentages are low and the presence of other clays prevents the determination of the illite poly typlsm (see Velde & Hower, 1 963).
vlei Pan suggest it is due to the presence of glauconite (as earlier proposed by Netterberg, 1 969a). Boocock & van Straten ( 1 962) described glauconite from Sekhuma Pan in Botswana, and Netterberg (1 969a) has pointed out its occurrence and shown its signifi cance in continental pan sediments over much of the Kalahari, glauconite being generally attributed to marine environments. Restriction of the green coloLJr ation to pan sediments over much of the Kalahari
Glauconite Many Kalahari calcretes are developed in 'green pan' sediments. In most cases the colouration cannot be related to any obvious green detrital mineral, but studies on carbonate-free green clay from Heunings78
Kalahari calcretes: origin and diagenesis calcite textures suggestive of original high-Mg calcite are found (Fig 6d, e). These occur as ' brush-like' bladed aggregates and splays of long acicular crystals similar, in many respects, to some high-Mg calcite beach rock and submarine cements. In addition, SEM work revealed the presence of distinct 'rice grain' calcite (Fig. 6c) similar to that described by Weaver & Beck (1 977) and J0rgensen ( 1 976), and identified as high-Mg calcite. Additional chemical/ mineralogical evidence (Watts, 1 977b) suggests that some of the micrite within the Kalahari calcretes was originally high-magnesian which has now transformed to a low-magnesian form. Aragonite is occasionally found within the calcrete profiles and textures suggestive of low-Mg calcite after aragonite are also apparent (Fig. 6b). For example, crusts of strongly fibrous calcite occur in which linear crystal inclusions and square-ended crystal terminations are common (Fig. 6b). These bear a close similarity to calcitized aragonite des cribed by Folk & Assereto ( 1 976) and Loucks & Folk ( 1 976), and are reminiscent of aragonitic speleo thems (R. L. Folk, personal communication, 1 977). Botryoidal low-Mg calcite aggregates, almost identical with those described by Davies ( 1 977) from ancient limestones and interpreted as neomorphic after aragonite, were also identified (sample S24805). Dolomite occurs in two forms in most mature profiles. Firstly, as early limpid, sub-euhedral to euhedral rhombs lining solutional channels (Fig. 6g) and sometimes coated by later calcite (sample S24834). Secondly, as small (5-20 fJ.m), cloudy rhombs, either concentrated in pods within the matrix or regularly scattered through a sepiolite rich plasma (see later). XRD reveals that the dolo mite is well ordered and no protodolomite is present.
suggests an authigenic origin but because of difficul ties of distinction fro m illite in polymineralic samples, glauconite has been included under 'illite' in the clay results. It is probably lost during calcrete development as authigenic glauconite and calcite are incompatible (Krumbein & Garrels, 1 952). Chlorite and kaolinite
These clays are of little significance in Kalahari cal cretes and never exceed 1 2% of the <2 f.l. m fraction (sample S24806). The chlorite is thought to be domi nantly detrital, its paucity reflecting the clay-poor nature both of the Kalahari sand and its source materials. Kaolinite, is slightly more abundant and, although dominantly detrital, in some cases results from early in situ weathering of feldspars (e.g. plagio clase feldspars in tholeiitic dolerite, samples S24805S24807). Clinoptilolite
In two profiles (e.g. Fig. 8) the sodium calcium zeolite 'clinoptilolite' was found. Clinoptilolite is a zeolite conforming to the general formula (CaNa2) [Al2Si7018]-6H20 (Deer, Howie & Zussman, 1 963). It is known to form both in saline lakes (e.g. Surdam & Eugster, 1 976) and marine environments (especi ally associated with volcanic materials, e.g. Kashkai & Babaev, 1 976). It is often associated with authi genic silica (see review by Nathan & Flexer, 1 977). Both Netterberg (1 969a) and Smale ( 1 968) have described clinoptilolite from pan sediments of the Kalahari, the latter pointing out the intimate association of the mineral with silcrete development.
Carbonates
By far the dominant mineral of the calcretes is low Mg calcite, occurring as micrite, microspar and spar cements (Fig. 6a), the last being preferentially preserved by late-stage clay or silica cutans. Within some profiles XRD revealed the presence of high-Mg calcites (up to 1 7 mol% MgC03 see Fig. 8), occurring in three distinct settings: (I) in calcrete developed on, or in, Mg-rich host materials (e.g. dolerites) ; (2) at the bases of some profiles associated with high groundwater levels (Fig. 8) ; (3) in immature profiles in fluvial (?) overbank fine sands. The diagenetic implications of high-Mg calcite, which will be dis cussed later, provide evidence of present day calcrete formation in the Kalahari. In some samples composed entirely of low-Mg
Auth igenic silica
The most common occurrence of silica is as a cement (Fig. 6f). Four consecutive cement types have been recognized: (I) a thin (usually < 5 f.l.m) diffuse layer of opaline silica ; (2) a moderately thick (40-90 f.l.m) zone of semi-opaque, brown, length slow chalcedony (Fig. 6f) which often grows in discrete to coalesced botryoidal aggregates ; (3) a sequence of rapid alternations of length-fast and length-slow chalcedony (possibly with a fine opaline coat) (4) a crystalline quartz fill. Most of the length slow chalcedony seen was quartzine (Folk & Pittman,. 1971), except for rare ' lutecite' occurring as large blade-shaped fibrous rays, commonly in place of the 79
N. L. Watts
80
Kalahari calcretes: origin and diagenesis 7
% DOLOMITE
6
en "' a: .... "' ::E
�
"' ...J
50 50
100
100
% CARBONATE (FIELD ESTIMATE)
Fig. 7. Clay mineral, mol% MgC03 (of calcites) and dolomite variations within a composite calcrete profile developed on kimberlite. Upper levels of calcrete occur in Kalahari Sand. Locality : No. I on Fig. 1. Key to figures in Fig. 3(b). megaquartz fill. This ideal sequence is rarely com plete and usually one or more stages are missing. Occasional calcite spar forms a final cement. Where opaline/amorphous silica is the only cement genera tion, good vadose (meniscus and gravitational) cements are seen, and where the megaquartz fill is absent lepispheres and fluffy coats of opal-CT (Jones & Segnit, 1 97 1 ) are common. Quartz is only seen replacing carbonate in ad vanced silcretes. In one example small isolated dolo mite rhombs have been replaced by single crystals of quartz whose c-axis is parallel to the long diagonal of the pseudomorphed crystal ; this r(}lationship is identical with that described by Folk & Pittman
( 1 97 1 , p. 1 049). Quartz overgrowths, jarosite and re entrant corrosion pits (see Taylor & Smith, 1 975) were not observed . MIN E R A L DIS TRIBUTI O N I N P R O FI L E S Clay minerals an d clino ptilolite
In all, eight logged sequences were analysed for clay minerals and carbonates, of which four have been selected for detailed discussion. The detailed results of the XRD analyses of all examined samples are listed in Watts ( 1 977b ; tables 3 A- l , 3A-2).
Fig. 6. Carbonates and silica. (A) SEM photograph showing low-Mg calcite cement crystals growing into a void. A thin layer of late-stage clay (palygorskite) originally coated the calcite but was removed by ultrasonic disaggregation prior to the photograph being taken. Sample S248 1 6. Bar is 50 [Lm long. (B) Thin section photomicrograph (plane light) showing possible original aragonite texture now inverted ·to calcite. Note strongly fibrous appearance (due to abundant linear inclusions) and similarity to aragonite speleothems. Sample S24804. Bar is 50 [Lm long. (C) SEM photograph showing 'rice grain' calcite. Note composite nature of some 'crystals', irregularity of size and anhedral form. This sample contains high-Mg calcite, S24823. Bar is 25 [Lm long. (D) Thin section photomicrograph (crossed polars) showing possible original high-Mg calcite (now low-Mg) occurring as long acicular crystals arranged in fan-like aggregates (middle-left). Note weak development of botryoidal fibrous calcite (extreme upper-right) (after aragonite?). Sample S24805. Bar is 0·25 mm long. (E) Thin section photomicrograph (plane light) showing possible original high-Mg calcite (now low-Mg) occurring as small brush-like blades extending outwards at the margin of an orthic glaebule. Note fluted terminations of some blades. Sample S24803 . Bar is 0·25 mm long. (F) Thin section photomicrograph (crossed polars) showing stages in authigenic silica void-fill. a, thin opaline silica coat; b, length-slow chalcedony (quartzine); c, length slow/length-fast chalcedony; d, megaquartz. Sample S24838. Bar is 0·25 mm long. (G) Thin section photomicrograph showing neoformed dolomite lining a solution channel. Note rhombic overgrowths on large crystal (left-centre). Sample S24834. Bar is 0·1 mm long.
81
N. L. Watts All the profiles analysed contained palygorskite and/or sepiolite, these minerals being generally more abundant in more mature calcretes. Netterberg (1969a) found that sepiolite was characteristic of hardpan calcretes. The present results, however, suggest that sepiolite distribution is related not to a particular calcrete type, but to the maturity of a given profile. This is seen in Fig. 9, where sepiolite is almost totally restricted to the lower mature calcrete profile. Palygorskite generally increases in abun dance upwards (Figs 7 and 8), but in composite profiles its distribution is more irregular. Because of the generally low amounts of chlorite, kaolinite and illite in the samples, variations of mixed-layer clays and montmorillonite tend to show inverse relation ships to palygorskite and/or sepiolite. This may be due to alteration of smectitic clays to palygorskite and/or sepiolite, or to the closed number-system effect of the standardized results. In immature profiles montmorillonite, with minor amounts of palygors kite and traces of other clays (illite, chlorite, kaoli nite etc.), is the dominant clay mineral. Occasional examples of sepiolite in profiles with little or no palygorskite were seen although such occurrences were rare. The crystallinity of the palygorskite and sepiolite was generally high (estimated by XRD peak sharpness), whilst that of montmorillonite decreased with increasing amounts of palygorskite, mixed layer clays being always poorly crystalline. Clinoptilolite was only identified in pan sediments, usually intimately associated with silicification features. In one profile (Fig. 8) clinoptilolite was absent in silcrete at the base but gradually increased upwards to 20% of the < 2 [Lm fraction.
total carbonate (see Figs 8-10), but this may be obscured by either composite profile development or the presence of neoformed dolomite. Authigenic s il i ca
Authigenic silica predominantly occurs in the lower levels of mature calcrete profiles and is preferentially concentrated in those developed in pan sediments. This distribution is consistent with ideas of etching of silicate mineral� by calcite in higher calcrete levels and downward migration and precipitation from silica-rich solutions (see Flach et a!., 1 969; Reeves, 1976).
DISCUSSION Origin o f sili ca tes
The clays palygorskite and sepiolite are thought to be intimately related to calcrete development and are thus discussed in some detail. The distribution of most other clays present is attributed to pedogenic influences involving both authigenesis and reorgani zation of primary detrital components prior to, or even later than, calcrete formation and are thus not considered. Illite and chlorite are thought to be dominantly detrital, whereas kaolinite may also be authigenic after feldspar. On the basis of XRD, mixed-layer illite-montmorillonites are attributed to degradation of illite. Montmorillonite may be authigenic after illite (or glauconite, see Robert eta!., 1 974) or detrital, whereas glauconite is authigenic within pan sediments. No vermiculite was found, suggesting limited or no chemical weathering of chlorite [consistent with the chlorites being Fe-rich ; e.g. Ross (1975) has found that Fe-chlorites suffer little or no alteration to vermiculite] or glauconite (potassium depletion of glauconite may produce an iron-rich vermiculite with an infra-red spectrum similar to that of nontronite; Abudelgawad, Page & Lund, 1975) within the profiles. The origin and significance of palygorskite and sepiolite can now be treated in more detail. Palygorskite, often associated with sepiolite, is characteristic of many arid-zone soils, and with the exception of work by Aristarain (1971, confirmed personal communication 1974) who found neither clay, most published clay analyses of calcretes re cord these minerals. Their almost ubiquitous presence suggests that they play an integral role in calcrete formation.
Carbonates
In Figs 7-10 variations in dolomite percentage (of the total carbonate fraction) and mol% MgC03 of the calcites have been plotted alongside the clay mineral data. Examination of these distributions reveals some interesting points. In general, the mol% MgC03 of the calcite shows an inverse relationship to paly gorskite and/or sepiolite concentration (Figs 9 and 1 0). Confirmation of this is shown in profile 824797824801 (Fig. 8). High-Mg calcites predominate in the lower part of the profile (note sample 824797 has high- and low-Mg calcites present) and no palygor skite is present ; only where high-Mg calcite is absent does palygorskite occur (e.g. sample 824801). There is an inverse relationship between the mol % MgC03 of the calcites and the dolomite % of the 82
Kalahari calcretes: origin and diagenesis .. 59· 5
- S24801
%
% DOLOMITE ( of Carbonate frocllon)
C L AYS
.. 50·6
1 5
�0� 0
. .
. "' �
Ul w a: f- w :::E
. 824800
• :: .
..
0
. ..ij)
,., : o
. S24798
�
. S24 799
w ..J
' ' ' '
. 524 797
4
• MEAN
50
8
12
16
20
2
3
4
5
SCHMIDT HAMMER VALVE
100
% CARBONATE ( F I E L D E S T I MATE )
Fig. 8. Clay mineral, mol% MgC03 (of calcites) and dolomite variations within a calcrete profile developed in (partially silcretized) pan sediments. Locality 2 on Fig. 1 . Asterisk numbers are Schmidt Hammer readings taken at outcrop.
In Kalahari calcretes alteration of montmorillon ite to palygorskite (with addition of Mg) may have been involved in the development of the dense, matted sheets of small palygorskite fibres. Weaver & Beck ( 1 977) have suggested that such alteration should take place where Si0 2 > 2·7 p.p.m., AI > 0 · 1 p.p.m., with increasing Mg 2 +. They drew atten tion, however, to the kinetic difficulties involved in transforming a sheet to a chain silicate, since energy requirements are high in the solid state. To overcome this problem they suggested that montmorillonite with the structure proposed by Edelman & Favejee (1 940), with alternate tetrahedra inverted, would be more easily transformed, particularly as the activity of Mg is one of the controlling factors in mont morillonite dissolution (Yaalon, 1 974). Under hypersaline conditions (high pH) Mg hydroxide apparently pre::: i pitates in the interlayer position forming chloritic minerals, but where there is less tendency to form hydroxides (lower pH) the Mg ions migrate to the octahedral sheet, increasing layer strain and forcing the tetrahedra to invert and form palygorskite (Weaver & Beck, 1 977, p. 2 1 0).
The occurrence of palygorskite in matted form discounts any possibilities of detrital origin (Singer & Norrish, 1 974). The mineral is considered authi genic, formed either from alteration of montmoril lonite (e.g. Yaalon & Wieder, 1 976) or secondly, by neoformation (e.g. Millot, Paquet & Ruellan, 1 969 ; Singer & Norrish, 1 974). Singer & Norrish (1 974) and Weaver & Beck (1 977) considered palygorskite for mation to be favoured by water-logged or brackish conditions, respectively, the primary requirement being alkaline Mg-rich conditions. Although Nahon & Ruellan (1 975) considered decalcification impor tant, Yaalon et al. ( 1 966), Mathiew, Thorez & Ek (1 975) and Yaalon & Wieder ( 1 976) thought the mineral was related to a calcification stage, the last authors showing that palygorskite disappeared on decalcification. Millot et al. ( 1977), in a study of calcrete developed on schist, identified palygorskite as an early stage mineral (their 'argillization' stage) suggesting that epigenesis of carbonates occurs at the expense of authigenic days. In the present study, however, palygorskite neoformation is demonstrably after or during calcite precipitation. 83
N L. Watts
- 52 4 7 8 1
•
. 524780
47·6·
. 524779
f--T'-'-+4.�-Y 49·1
• 00 CJOO . 46· 2 () 0 Q
(f) w a:
....
w :::<
�
' 0
w
"' u (f) ..J
o. o . o ao · · o o· o
% DOLOMITE ( of Carbonate fraction )
CLAYS
. 524 7 7 8
. 524 7 7 7
() 6 0 · 0 _o 0 0 0 0 ()
0
0
,..._. i
%
0
•
. 5 2 47 7 6
•
. 5247 7 5
3 5 · 9·
. 5 2 47 7 4
d· 0
0
100 50 % CARBONATE ( FIELD ESTIMAT E )
. 524 7 7 3
50
2
100
•
MEAN
3
5
50
100
SCHMIDT HAMMER VALUE
Fig. 9. Clay mineral, mol% MgC03 (of calcites) and dolomite variations within a composite calcrete profile developed in Kalahari Sands. Locality 3 on Fig. 1 . Asterisk numbers are Schmidt Hammer readings taken at outcrop.
It is possible that such a smectite to palygorskite transformation occurs in some Kalahari calcretes and results in the development of densely matted sheets of palygorskite, although petrographic evi dence is not conclusive. It is unlikely, however, that such a simple transformation process could take place in the solid-state and some intermediate (dissolution?) phase must be involved as the mineral structures of montmorillonite and palygorskite are completely different. Most palygorskite is thought to have been pre cipitated directly from solution (neoformed, e.g. the loose mats of long palygorskite fibres) as possibly confirmed by two recent studies. Singer (1 979) has recently discussed the origin of palygorskite in sedi ments and favoured a neoformation, over transfor mation, mode of authigenesis. In addition, Hassouba & Shaw (1 980) have described palygorskite in Quaternary sediments from Egypt in which no smec tite precursor was present. They attributed the min eral to a neoformational origin.
Weaver & Beck ( 1 977, p. 1 08) found inter relationships between montmorillonite, palygorskite and corrensite. No corrensite was found in the Kala hari samples. This is consistent with fairly high silica levels (discussed below) and moderate salinities [although recently Almon, Fullerton & Davies (1 976) have questioned the common belief that corrensite formation demands hypersaline conditions (see Kubler, 1 973)]. Paquet & Millot (1 972) have stated that palygor skite in soils is unstable and weathers to montmoril lonite when the mean annual rainfall exceeds 300 mmfannum. Whilst this is a broad generalization, it is consistent with observations of palygorskite in arid-zone soils and may be of use in palaeoenviron mental analyses of ancient palygorskite-rich se quences (e.g. Watts, 1 976). Sepiolite is common in Kalahari calcretes and SEM evidence suggests a dominantly neoformational origin. Weaver & Beck ( 1 977) suggested that sepio lite forms by direct precipitation under more alkaline 84
Kalahari calcretes: origin and diagenesis 1 ·6 % DOLOMITE (of toto I Corbonote)
% CLAYS
\
- 524784
- 524783 c ..!!
:PA+.-':-r-'-'---r-'-:'-1. - 524782
50
100
0
2
4
6
B
0
50
100
% nodules
Fig. 10. Clay mineral, mol% MgC03 (of calcites) and dolomite variations within two simple profiles developed i n Kalahari Sands. Locality 4 on Fig. J .
of the Kalahari, the Na and Ca originating from the pan sediments and the lower pH inhibiting calcite precipitation.
conditions than palygorskite (where the Si and Mg levels are high and AI low), and usually in the ab sence of smectite. This could explain the dominance of sepiolite as a late-stage mineral. Early palygorskite formation reduces available aluminium to such a point that, with increasing Mg (possible sources discussed below) sepiolite precipitates. This would also account for the common association of dolomite with sepiolite in Kalahari calcretes, dolomite for mation being enhanced by raised Mg levels. A similar association has been documented recently by Post ( 1 978) who described sepiolite neoformation within the lower, dolomitic levels of a calcrete profile from Nevada, U.S.A. Finally, the occurrence of clinoptilolite within silicified pan sediments must be discussed. Although Couture ( 1 977) described palygorskite and clinop tilolite occurring together in marine sediments of the Pacific Ocean, the two minerals are usually thought to be mutually exclusive, palygorskite being favoured by high pH, high [H4Si04 °] and high Mg/Na ratios of the precipitating fluids (Weaver & Beck, 1 977 ; cf. Couture, 1 978). Kashkai & Babaev ( 1 976) have suggested that clinoptilolite preferentially precipi tates where the Si/AI ratio is greater than 4 : 1 . This is consistent with its occurrence in silcrete-rich material
Origin of authigenic silica
In discussing the origin of silica in calcretes and its environment of precipitation, ideas relevant to the process of silcrete formation are first outlined. The source of silica in silcretes has been attributed to a number of sources. For example, fluctuating ground water (Du Toit, 1 954), kaolinitization of feldspars and serpentinization (discussed by Smale, 1 973) and proximity to basalts (silica release and/or baking effects, Browne, 1 972). In addition, upward-moving, silica-rich solutions, possible mixing with downward percolating alkaline water (Smale, 1 973) or sealed in by an impermeable basalt caprock (Taylor & Smith, 1 975 ; see discussion by Leeder, 1 976) have also been proposed. None of these hypotheses satisfies the widespread nature of silicification features in the Kalahari. The intimate relationship between silicifica tion and calcretes suggest that the major source is the release of silica on replacement of silicates by calcite · (Walker, 1 960). It is well known that silica shows an inverse 85
N. L. Watts solubility relationship with calcite (e.g. Correns, 1 950), high pHs (over 9, see Alexander, Hester & Iler, 1 954) favouring calcite precipitation and silica solution, and at low pHs the reverse. Therefore, as calcrete formation proceeds (under fairly high pH conditions), silicate replacement (as described above) would be expected to occur, the silica content of the pore fluids gradually increasing. Migration of these fluids into areas of high salinity or high free C02, and corresponding slight decrease in pH, may cause silica precipitation (see Lovering & Patten, 1 962). The latter authors have shown that Si02-rich solu tions are stable in the presence of carbonates but become unstable, with consequent rapid precipita tion of silica, in saline or hi g h-C02 environments. Such a process, acting in calcrete would account for the frequently observed early opaline silica cement (due to precipitation from a colloidal gel?) followed by the length-slow chalcedony. Folk & Pittman ( 1 97 1 ) attributed length-slow chalcedony to pre cipitation from highly concentrated solutions of fairly high pH (thus inhibiting Si02 polymerization and crystallization as a length-fast variety). Such a hypothesis agrees with observations and chemical considerations of calcrete, although the suggestion that length-slow chalcedony indicates that the former existence of evaporites is questionable. Length� slow chalcedony has been frequently described from calcretes (e.g. Folk & Pittman, 1 971 ; Stalder, 1 975) and silcretes (e.g. Smale, 1 973 ; Taylor & Smith, 1 975) and in most cases original evaporites have not been observed. Indeed Milner ( 1 976) states 'the presence of length-slow chalcedony in a rock unit is an ambiguous criterion to infer an evaporitic de positional history if used alone, and Folk ( 1 977, personal communication) agrees that high salinity (with accompanying fairly high pH) may be the important factor. Consequently, the rapid alter nations of length-fast and -slow chalcedony in the Kalahari samples (Fig. 6f) may be related to fluc tuations in pH of the precipitating fluid (see dis cussion above on clinoptilolite). The coarse crystal line quartz is attributed to slow crystallization of Si02 as a void-fill.
carbon is probably important in South African calcretes (Salomons, Goudie & Mook, 1 978), al though micromorphological evidence of organic processes on the scale of those described by Klappa ( 1 978) from less arid pedogenic calcretes were not observed ; rhizocretions, root hairs and some calcified algal/fungal filaments are seen only rarely. Stalder ( 1 975) has demonstrated the importance of sheetfloods in transporting ions from weathered bedrock to sites of calcrete formation and a similar process may supply carbonate to the non-pedogenic calcretes of the Kalahari (particularly in the south). Finally, in and around saline depressions (e.g. Makgadikgadi, see Fig. 1) shallow groundwaters probably contribute to calcrete formation by capillary rise (see later). Origin of low-Mg calcite
Detailed measurements of vadose waters of the Kalahari have not been made but information (Botswana Geological Survey, oral communication) suggests fairly low Mg/Ca ratios. It is not surprising, therefore, that most samples analysed are dominated by low-Mg calcites. These are ascribed to precipi tation from low Mg/Ca ratio solutions either as a result of mild evaporation, C02 loss, or both. Neomorphic calcite results from diagenetic recrystal lization of earlier carbonates within mature calcretes. Origin of high-Mg calcites
Although most evidence of high-Mg calcites is found in neC>morphic low-Mg calcites, originally high-Mg calcites have been identified in three distinct settings : ( I ) In calcretes developed on Mg-rich host materials including dolerite, kimberlite and some pan sediments. The coarse crystal size of these high-Mg calcites suggest fairly slow precipitation from vadose waters locally enriched in magnesium. (2) In the lower levels of calcretes developed in and around saline depressions (e.g. Fig. 8). At the locality of the profile shown in Fig. 8 a groundwater level was seen at a depth of 3 m below the topo graphic surface. These groundwaters are probably preferentially enriched in magnesium compared to vadose waters and may account for the high-Mg calcites, precipitation being stimulated by C02 loss during capillary rise from the shallow ground waters. (3) The third occurrence of high-Mg calcites is somewhat problematic. They are found in the upp�r levels of immature calcretes developed on, or in, Mg-poor Kalahari sands (e.g. at Nata, Fig. 1 ; see
Origin of carbonates Source of calcium and carbonate ions
Aeolian dust is considered to be the major source of ions for calcrete formation (see Goudie, 1 973 ; Reeves, 1 976), but contributions from rainwater may be locally important (e.g. Gardner, 1 972). Biogenic 86
Kalahari calcretes: origin and diagenesis rapid precipitation accounting for the anhedral nature of the micritic crystals.
Watts, 1 977b). The carbonate is totally micritic and the crystals, as seen in SEM, are anhedral and re semble miniature rice-grains (Fig. 6c). It is difficult, at first, to account for the precipitation of these high-Mg calcites from vadose solutions with low Mg/Ca ratios. Glover & Sippel ( 1967) and Marschner (1 968) have shown that the Mg content of a given calcite is proportional to the Mg/Ca ratio of the solution fro m which it precipitates. Why then should high-Mg calcite precipitate from waters of low Mg/Ca ratios? It is known that magnesium (and other) ions affect and sometimes inhibit calcite nucleation (e.g. Bischoff 1 968 ; Bischoff & Fyfe, 1 968). De Boer (1 977) has stated that low-Mg calcite is the most stable poly morph at Mg/Ca ratios of 0-100 (or more), but in evaporitic lakes and in seawater high-Mg calcite forms, but never in true equilibrium with the pre cipitating water. Calcite precipitation may be induced by either C02 loss (e.g. Matthews, 1 969 ; Land, 1 970), increased temperature or evaporation (e.g. Thortensen, McKenzie & Ristvet, 1 972 ; Lippman, 1 973). In order to understand the origin of this micritic high-Mg calcite let us consider water of a low Mg/ Ca ratio (say around 1 ·0) within a vadose pore in calcrete. To produce rapid precipitation, strong evaporation and perhaps corresponding rapid Joss of C02 are needed, conditions to be expected within semi-arid calcretes. If evaporation is more rapid than calcite nucleation, the Mg/Ca ratio of the solution would remain constant but the salinity would increase. Such a solution, as shown by Folk & Land (1 975, Fig. 2), would thus tend to precipitate high-Mg calcite (and/or aragonite) and not the expected low-Mg variety. A similar situation would be expected within a kettle on almost complete evaporation of the water. Butler (1 975) studied the carbonate crust of a tea kettle into which water of a low Mg/Ca ratio had been introduced. Although the precipitated carbonates were to some extent affected by the fairly high temperatures within the kettle Butler found that from water with a Mg/Ca ratio of only 0· 15 high-Mg calcite (up to 12 mol% Mg C03) and ar< gonite precipitated. He attributed this to 'literally forced crystallization' and speculated that such a process should occur where fresh-water is trapped and subjected to evaporation. Butler's kettle is considered analogous to the conditions within a calcrete. The high-Mg calcite of the calcretes, precipitated by 'forced crystallization', will thus not be in equilibrium with the precipitating fluids, such
overall
The implications of such 'non-equilibrium' high Mg calcite precipitation are thus considerable, as it is possible that rapid evaporation (as opposed to C02 Joss) deviates from predicted thermodynamic con straints. The precipitation of high-Mg calcite micrite in calcretes from low Mg/Ca ratio waters leads one to conclude that thermodynamic equilibrium has not been reached. It is therefore possible that some pro cess similar to the 'stoichiometric saturation' of Thortensen & Plummer (1 977) occurs during rapid evaporation of pore water within the Kalahari calcretes in calcites above 4·0 mol% Mg C03• The results of Thortensen & Plummer (1 977) not only suggest that metastable thermodynamic equilibrium can never be achieved between high-Mg calcites and natural waters, but that with a constant Mg/Ca ratio of water the Mg content of calcite> (at 25°C) should be solely determined by the degree of supersaturation of the water with respect to calcite. High supersatura tion would occur with rapid evaporation and has already been shown to be a requirement for dis placive calcite crystallization within calcretes (Watts, 1 978). While it may be argued that Mg-enrichment in calcites can take place at elevated temperatures (e.g. Fuchtbauer & Hardie, 1 980), the concentration of magnesium in calcites within some Kalahari profiles cannot be explained by this factor alone. It is suggested here that rapid precipitation of calcite, from highly supersaturated solutions, is essentially kinetically controlled, and precipitation is too rapid for equilibrium to be attained. This implies that nucleation rate, in addition to temperature, should be considered when evaluating the partitioning of magnesium during calcite precipitation. Barnes & O'Neil ( 1 97 1 ) have described high-Mg calcite cements in a Holocene conglomerate from California forming in thermodynamic disequilibrium but isotopic equilibrium. Because such high-Mg calcite> are metastable, and rapidly transform to Jow-Mg calcite (see later), isotopic studies of calcretes should be undertaken with some care. Moreover, Usdowski , Menschel & Hoefs (1 980) have recently described isotopic disequilibrium in calcites resulting from rapid precipitation from highly supersaturated solutions. In view of the possible kinetic controls on calcite precipitation in calcretes outlined above, the work of Usdowski et a!. may have important impli cations in future isotopic studies of calcretes. 87
N. L. Watts and schizohaline (Folk & Siedlecka, I 974) dolomiti·· zation models. An intriguing aspect of Kalahari calcretes is the presence of features suggestive of a schizohalim� environment (see criteria outlined by Folk & Sied·· lecka, 1 974) ; for example, length-slow chalcedony, finely crystalline dolomite, euhedral dolomite, microspar and sparry calcite. This is not entirely unexpected in an environment with high evaporation and periodic flushing by rain water. It is, therefore, possible to draw certain analogies between the schizohaline and Kalahari calcrete diagenetic: environments. Finely crystalline, often cloudy, dolomite crystals are found in many pedogenic calcrete profiles. Their close association with pedogenic sepiolite (see above} suggests an origin from waters of similar composition to those required for sepiolite neoformation. It is proposed, therefore, that this finely crystalline: dolomite formed from solutions with high Mg/Ca ratios induced by evaporation ; as discussed below,. Mg-enrichment of the pore fluids is attributed predominantly to concomitant precipitation of low-Mg calcites within the profile. The moderately coarse, limpid dolomite, however, presents a different problem. It is predominant in older, thicker calcrete profiles and often occurs in solution channels. In places, high concentrations of limpid dolomite (and clear rims on finely crystalline: dolomite) are associated with the lower portions of calcrete profiles and appear to be related to (past?) groundwater influences. This dolomite is attributed to precipitation from low Mg/Ca ratio waters. Such waters may originate during periodic flushing of the profiles by rainwater which explains the association of limpid rhombs with solution channels. Higher concentrations of dolomite in older cal-· cretes are tentatively ascribed to mixing of vadose: and phreatic waters (possibly at a capillary front) at, and around, the groundwater table. The 'mixing zone' within Kalahari calcretes differs, however,. from the classical Dorag model in that, due t o evaporation, the overlying vadose waters are the more: saline. If such a model is universally applicable i n arid-zone continental environments, i t would pro-· vide an interesting alternative to the various dolo-· mitization models listed above.
Origin of aragonite One of the major problems encountered throughout this study has been the lack of data on carbonate minerals in 'fresh water' or continental environments. Although the major occurrences of aragonite cements are in marine or peritidal environments (see Bathurst, I 975), they have been described in vadose conditions, such as calcretes (e.g. Panos & Stele!, I 968 ; Scholle & Kinsman, I 974), in cave deposits (e.g. Fishbeck & Muller, I 97 1 ; Thraikill, I 97 1 ; Reams, I 974), and in fresh-water sediments (e.g. Konishi & Sakai, I 972) and soils (e.g. Veen & Arndt, I 973) where precipitation seems to be controlled by high Mg/Ca ratios. Preferential crystallization of aragonite is also favoured by elevated temperatures (e.g. Kinsman, I 965) and rapid loss of C02• Reams ( I 974) has d emonstrated experimentally that water with a low Mg content will precipitate aragonite, if stirred to induce C02 loss. Calcite nucleation has been shown to be slower than its rate of crystal growth (Matthews I 969) and kinetic factors usually favour aragonite crystallization over calcite (Reams, I 974). Further more, Butler ( 1 975) observed 25% aragonite, asso ciated with high-Mg calcite, in the upper crust of his kettle. It is suggested here that aragonite precipi tation in calcretes is possible, and could take place by rapid C02 loss (associated with evaporation) of low Mg/Ca ratio vadose waters. Origin of dolomite Because of the obvious importance of dolomites as major hydrocarbon reservoirs throughout the world, great efforts have been made to understand the origin of dolomite. This has resulted in the development of a number of dolomitization models which include evaporative reflux, evaporative pumping, compaction and dewatering of shales and the mixed-water ('Dorag') and schizohaline models (see Davies, I 979, for review). Opinions are generally divided over the need for low Mg/Ca ratios of the precipitating pore fluids (Morrow, I 978). Folk & Land (1 975) reviewed the problem and suggested that dolomite may pre cipitate from solutions of high Mg/Ca ratios if sufficient time for their ordering is available. Because of this ordering problem, they believed that at high salinities the Mg/Ca ratio must exceed 5-10 : I , but with reduced salinities (and thus slower crystalliza tion) Mg/Ca ratios of as low as I : I are sufficient. Such low Mg/Ca ratios can be attained by mixing of saline (marine) and fresh waters which is an inherent process in the Dorag (Badiozamani, 1 973)
Sources of magnesium for silicate authigenesis
�mll
dolomitization
In the above account details of silicate and carbonate� 88
Kalahari calcretes: origin and diagenesis loss and/or evaporation within the vadose pore fluids and subsequent inversion to low-Mg calcite. The primary calcite precipitation mechanisms are also thought to involve either rapid or slow evapora tion and/or C02 loss. Slow evaporation predomin antly gives low-Mg calcite with a consequent gradual i ncrease of magnesium concentration in the resulting solutions. Rapid evaporation may precipitate high Mg calcites which are in thermodynamic disequilib rium with the low Mg/Ca ratio vadose waters. Both passive (void-fill) and replacive (of detrital silicates) calcite occurs, silica being released in the latter. Under 'saline' conditions (in the broad sense) length-slow chalcedony and/or clinoptilolite precipitates, where as length-fast chalcedony and/or megaquartz are formed in 'non-saline' micro-environments. The silica is derived fro m replacive calcitisation and migrates down-profile to accumulate in the lower calcrete horizons. Finally, rapid evaporation may result in solutions highly supersaturated with respect to calcite, and displacive growth of calcite may occur (Watts, 1 978). Neomorphism of high-Mg to low-Mg calcite takes place fairly rapidly and magnesium is released. This, combined with the increased Mg/Ca ratio due to low-Mg calcite precipitation, increases the mag nesium concentration of the pore fluids to such a level that authigenic Mg-rich silicates may precipi tate. Clay authigenesis is an integral part of calcrete formation in the Kalahari, most profiles containing appreciable amounts of palygorskite and sepiolite. Palygorski te may form by reaction of Mg with mont morillonite or may be precipitated direct from solu tion in association with neoformed sepiolite and dolomite. Magnesium may also 'react' with (cement or neomorphic) low-Mg calcite to form replacive dolomite. The above pedogenic and diagenetic scheme com plies with most of the observations and results of the Kalahari calcretes. The model implies a dominantly 'closed' system but episodes of extremely high rainfall may obscure this and flush the system. Local precipitation of high-Mg calcite is thought to occur by capillary rise of groundwaters in, and around, saline depressions (e.g. Makgadikgadi). Microcrystalline dolomite, often associated with neoformed sepiolite, precipitates under evaporitic conditions from Mg-enriched pore fluids. Coarse, limpid dolomite is attributed to crystallization fro m low Mg/Ca ratio waters. These result either fro m flushing o f the calcrete profile during periods o f
authigenesis have been presented without detailed discussion of magnesium sources. The major cause of magnesium enrichment is believed to be the selective removal of calcium by low-Mg calcite precipitation raising the Mg/Ca ratio of the vadose solutions to conditions favourable for palygorskite, sepiolite and dolomite formation. In the description of mineral distributions in profiles, however, the inverse relationships between the mol% MgC03 of the calcites and palygorskite, sepiolite and dolomite content was shown. If some low-Mg calcite now present in the samples is a transformation product of original high-Mg calcites, magnesium released during alteration could have locally enriched the vadose waters and aided clay and dolomite formation. Land & Epstein (1 970) have suggested that loss of Mg from calcites may cause an increase in the Mg/Ca ratio of the fluids and induce dolomite precipitation. Folk & Land ( 1 975), whilst agreeing that this process is possible, thought that only minor amounts of dolomite would result. It is thought here, however, that such quantities of dolomite as seen in simple Quaternary Kalahari calcrete profiles may be explained by this mechanism. The removal of Mg from high-Mg calcites cannot take place rapidly by solid-state diffusion, certainly not at low temperatures, and it is probable that some dissolution-reprecipitation process is active. Such a process would be fairly rapid and preserve in detail original high-Mg calcite morphologies and textures.
CONCLUSIONS
The Kalahari calcretes described in this paper result from dominantly pedogenic processes occurring episodically throughout the Pliocene to Recent in a semi-arid climate. Varying degrees of calcrete maturity are related to a number of interdependent factors : time, climate, host materials, carbonate source, geomorphological position, organic in fluences, sedimentation (or erosion) rate and various localized conditions. The interplay of such a number of parameters over an area as large as the Kalahari obviously results in a highly diverse suite of calcrete types, but such variations are frequently observed even on a single outcrop. Consequently, broad con clusions must be circumspect. Fig. 1 1 is a schematic flow diagram of the main processes operating i n the Quaternary calcretes. For simplicity, aragonite has been separated into a minor category, its precipi tation resulting from rapid C02 89
N. L. Watts
�
PRIMARY CARBONAT E PRECIPITAT ION
4
�
� / " ' " I • < ' " ? CO, ""'
,A
D I S P L C I VE
.
?.
l
PAS S I V E
� � � ;lo• '' '
( m inor aragonite)
'"
l
G E N E T I C CAW T E l"ES
- ------v---+-R E P L ACIVE
high-Mg cte
:r
high - Mg cte
S1
�------
release
........_.._ ........__
SALINE
_
0� 1
REP ACI V E
(
PAS I VE
Mg cone. o f ppting solutions
?I
1
g r o d u l l y increasing NON - SA L I N E
msm
,
law-Mg cte
low-Mg c!c
original cement or c l inoptilol i te
megaquartz
� Mgf
" ""
+ M g --+----.,
sepiolite do lomite
_ _ _ _ _ _ neoformed
_ _ _ _ _ _ _
f--- + montmorillonite
I'"T'"I
-4
palygor s k i te
B
pa l y gorsk i t e
Fig. 1 1 . Schematic flow diagram summarizing the major pedogenic a n d diagenetic processes within the Kalahari calcretes. Note aragonite is not shown in detail. See text for discussion.
concentration of magnesium (and the bulk (Mg/Ca ratios) of the Kalahari calcretes is similar to that of calcretes elsewhere (Watts, 1 977b). The model presented above and in Fig. 1 1 applies to the pedogenic calcretes of the Kalahari. It remains to be seen whether it can be applied to calcretes elsewhere. If not, then we must explain the domin ance of palygorskite and sepiolite in many calcrete profiles throughout the world and further research on this problem is essential. Detailed measurements of pore water chemistry, calcrete microclimate, and trace element and isotopic analyses of individual cement generations within calcretes would be of considerable use, and would greatly aid our under standing of this obviously major process in semi arid, continental environments.
rainfall or, in older, thicker calcretes, by mixing of saline vadose and fresh phreatic waters. The evidence presented indicates that high-Mg calcite is a significant component of the Kalahari calcretes contrasting sharply with calcretes from most other areas. It may be that original high-Mg calcite textures are present in other calcretes, but have not been identified. Textures suggestive of high-Mg calcite or aragonite precursor cements have been seen in thin sections of calcretes from Australia, North Africa (courte3y of M. G . Talbot) and central Africa (courtesy of J. Beauchamp) suggesting that, perhaps, the Kalahari material is fairly representative as far as carbonate cementation is concerned. In addition, the number of authentic Recent calcretes is remark able low (Goudie, 1 973), and thus if high- to low-Mg calcite transformation is rapid (it is too slow to be observed in the laboratory, R.B. de Boer, personal communication, 1 977) all of the calcite in relic calcretes should now be low-Mg calcite. Finally, it could be that the Kalahari material is exceptional. While it is true that some of the rocks surrounding the Kalahari are dolomites, and that some of the calcrete host-materials are quite Mg-rich, the overall
A CKNOWLEDGME N T S
This work was performed whilst the author was in receipt of a NERC postgraduate studentship at the Sedimentology Research Laboratory, University of Reading. I should like to thank all my friends and 90
Kalahari calcretes: origin and diagenesis colleagues at Reading for their help, and in particular my supervisors, Prof. J. R. L. Allen, Dr A. Parker and Dr R. Till for their efforts and invaluable advice. Prof. P. Allen, Mr G. Brown and Dr J. A. D. Dickson contributed greatly to the refinement of ideas and numerous colleagues and co-workers gave stimulat ing discussions. I am deeply indebted to Dr C. J. R . Braithwaite for his constructive comments on an earlier version of this paper. I am, however, wholly responsible for the interpretations and any errors included within this paper. I thank Shell I nter nationale Research Mij B.V. for their permission to publish this work.
BISCHOFF, J.L. & FYFE, W.S. ( 1 968) Catalysis inhibition,
and the calcite aragonite problem. I. The aragonite calcite transformation. Am. J. Sci. 266, 65-79. BLOM, G.I. ( 1 970) Buried palygorskite soils in the Lower Triassic of the Moscow syneclise. Dokl. (Proc.) A cad. Sci. U. S.S.R. Earth Sci. Sect. 194, 52-54. BoER, R.B. DE ( 1 977) Stability of Mg-Ca carbonates. Geochim. cosmochim. Acta, 41, 265-270. BoNATTI, E. & JOENsuu, 0. ( 1 968) Palygorskite from Atlantic deep sea sediments. Am. Miner. 53, 975-985. BOOCOCK, C. & STRATEN, O.J. VAN (1 962) Notes on the geology and hydrogeology of the central Kalahari region, Bechuanaland Protectorate. Trans. geol. Soc. S. Afr. 65, 1 25-1 7 1 . BRADLEY, W.F. ( 1 940) The structural scheme o f atta pulgite. Am. Miner. 25, 405-41 0. BREWER, R. (1964) Fabric and Mineral Analysis of Soils. Wiley & Sons, New York. BROWNE, W.R. (1 972) Grey-billy and its associates in Eastern Australia. Proc. Linn. Soc. N.S. W. 97, 98-129. BuTLER, J.C. ( 1 975) Tea kettle carbonates. J. sedim. Petrol. 45, 891-893. CAILLERE, S. & HENIN, S . ( 1 96 1 ) Sepiolite. In : The X-ray
REFERENCES ABA-HusAYN, M . M . & SAYEGH, A.H. ( 1 977) Mineralogy
of Al-Hasa desert soils (Saudi Arabia). Clays Clay Miner. 25, 1 38-148. ABUDELGAWAD, G., PAGE, A.L. & LUND , L.J. ( 1 975)
identification and Crystal Structure of Clay Minerals,
2nd edn. (Ed. by G. Brown), pp. 325-342. Minera logical Society, London. CARLISLE, D. ( 1 978) The distribution of calcretes and
Chemical weathering of glauconite. Proc. Soil Sci. Soc. A m . 39, 567-5 7 1 . ALEXANDER, G.B., HESTON, G . M . & li.ER, H . K . ( 1 954) The solubility of amorphous silica in water. J. phys. Chem. 58, 453-455. ALLEN, J.R.L. (1 974) Sedimentology of the Old Red Sandstone (Siluro-Devonian) in the Clee Hills area, Shropshire, England. Sedim. Ceo!. 12, 73-167. ALMON, W.R., FULLERTON, L.B. & DAVIES, D.K. (1 976) Pore space reduction in Cretaceous sandstones through chemical precipitation of clay minerals. J. sedim. Petrol. 46, 89-96. ARISTARAIN, L.F. ( 1 97 1 ) Clay minerals in caliche deposits of Eastern New Mexico. J. Ceo!. 79, 75-90. AUSTIN, G.S. & LEININGER, R.K. (1 976) The effect of heat-treating sedimented mixed-layer illite-smectite as related to quantitative clay mineral determinations. J. sedim. Petrol. 46, 206-2 1 5 . BADIOZAMANI, K. ( 1 973) The Dorag dolomitisation model : application to the Middle Ordovician of Wisconsin. J. sedim. Petrol. 43, 965-985. BAILLIEUL, T.A. 0 975) A reconnaisance survey of the cover sands in the Republic of Botswana. J. sedim. Petrol. 45, 494-503. BARNES, I. & O NEIL , J.R. (1971) Calcium-magnesium carbonate solid solutions from Holocene con glomerate cements and travertines in the Coast Range of California. Geochim. cosmochim. A cta, 35, 699-7 1 8 . BATHURST, R.G.C. (Ed.) (1 975) Carbonate Sediments and their Diagenesis (2nd edn). Developments in Sedi mentology, 12. Elsevier Publishing Co ., Amsterdam. BEATTIE, J.A. & HALDANE, A.D. (1 958) The occurrence of palygorskite and barytes in certain parna soils of the Murrumbidgee region, New South Wales. Aust. J. Sci. 20, 274-275. BISCHOFF, J.L. ( 1 968) Kinetics of calcite nucleation: magnesium ion inhibition and ionic strength cata lysis. J. geophys. Res. 73, 3 3 1 5-3322.
gypcretes in Southwestern
United States and their
uranium favorability. Based on a study of deposits in Western Australia and South West Africa (Namibia).
Open File Report, University of California. CHAVE, K.E. ( 1 952) Solid solution between calcite and
dolomite. J. Ceo!. 60, 1 90-192. COLLINS, R.J. (1 976) A method for measuring the mineralogy variation of spoils from British Columbia. Clay Miner. 1 1 , 3 1-50. CooKE, H.J. ( 1 975) The palaeoclimatic significance of caves and adjacent landforms in Western Ngamiland, Botswana. Geogrl J. 141, 430-444. CoRRENS, C.W. ( 1 950) Zur Geochemie der Diagenese. Geochim. cosmochim. Acta, 1, 49-54. CouTURE, R.A. ( 1 977) Composition and origin of paly gorskite-rich and monotmorillonite-rich zeolite-con taining sediments from the Pacific Ocean. Chem. Ceo!. 19, 1 1 3-1 30. CouTURE, R.A. ( 1 978) Miocene of the SE United States : a model for chemical sedimentation in a peri-marine environment-comments. Sedim. Ceo/. 21, 149-154. DAVIES, G.R. ( 1 977) Form�r magnesium calcite and aragonite submarine cements i n Upper Palaeozoic reefs of the Canadian Arctic: a summary. Geology, 5, 1 1- 1 5 . DAVIES, G.R. ( 1 979) Dolomite reservoir rocks: processes, controls, porosity development. Am. Ass. Petrol. Ceo/. Continuing Education Course Note Ser., No. I , Part C, 1 7 pp. DEER, W.A., HOWIE, R.A. & ZUSSMAN, J. (1963) Rock Forming Minerals, vol. 4. Framework Silicates. Longmans, London. DOBROVOL'SKIY, V.V. ( 1 973) Mineralogical and geo chemical characteristics of the Black Soils of Kenya, Uganda and Tanzania. Pochvovedeniye, 8, 1 4-1 5 .
'
91
N. L. Watts Du Ton, A.L. (1 954) The Geology of South Africa, 3rd edn. Oliver & Boyd, Edinburgh. EBERL, D. & HOWER, J. ( 1 976) Kinetics of illite formation. Bull. geol. Soc. Am. 87, 1 326-1 330. EDELMAN, C.H. & FAVEJEE, J.C.L. (1 940) On the crystal structure of montmorillonite and halloysite. Z. Krystallogr. 102, 4 1 7-43 1 . ESWARAN, H . & BARZANJI, A.F. (1 974) Evidence for neoformation of attapulgite in some soils of Iraq. Trans. lOth Int. Congr. Soil Sci., Moscow, 1 54-1 6 1 . FISHBECK, R . & MULLER, G . (1971) Monohydrocalcite, hydromagnesite, nesquehonite, -dol'omite and calcite in speleothems of the Fraenkiscne Scheweiz, W . Germany. Contr. Miner. Petrol. 33, 87-92. FLACH, K.W., NETTLETON, W.D., GILE, L.H. & CADY, J.G. (1969) Pedocementation : induration by silica, carbonates, and sesquioxides in the Quaternary. Soil Sci. 107, 442-453. FoLK, R.L. & AssERETO, R . (1 976) Comparative fabrics of length-slow and length-fast calcite and calcitized aragonite in a Holocene speleothem, Carlsbad Caverns, New Mexico. J. sedim. Petrol. 46, 486-496. FoLK, R.L. & LAND, L.S. (1975) M g/Ca ratio and salinity ; two controls over crystallisation of dolomite. Bull. Am. Ass. Petrol. Geol. 59, 60-68. FOLK, R.L. & PITTMAN, J.S. (1971) Length-slow chalce dony : a new testament for vanished evaporites. J. sedim. Petrol. 41, 1045-105 8 . FOLK, R . L . & SIEDLECKA, A. (1 974) The 'schizohaline' environment : its sedimentary and diagenetic fabrics as exemplified by Late Paleozoic rocks of Bear Island, Svalbard. Sedim. Geol. 1 1 , 1 - 1 5 . FOCHTBAUER, H . & HARDIE, L . A . ( 1 980) Comparison of experimental and natural magnesian calcites (abstr.).
JoNES, J.B. & SEGNIT, E.R. (1971) The nature of opal. I . Nomenclature and constituent phases. J. geol. Soc. A ust. 18, 57-68. J0RGENSEN, N.O. (1 976) Recent high magnesium calcite/
aragonite cementation of beach and submarine: sediments from Denmark. J. sedim. Petrol. 46, 940-· 95 1 . KASHKAI, M .A. & BABAEV, I.A. (1 976) Clinoptilolite from zeolitized tuffs of Azerbaidzham. Mineralog. Mag. 40, 501-5 1 1 . KING, L.C. ( 1 963) South African Scenery : a Textbook of Geomorpholog� , 3rd edn. Oliver & Boyd, Edinburgh. KINSMAN, D.J.J. (1965) Co-precipitation of Sr +2 with aragonite from sea water at 1 5-95°C. Prog. 1 965 Ann. Mtg, geol. Soc. A m . , 87. KLAPPA, C.F. ( 1 978) Morphology, composition and genesis of Quaternary calcretes from the western Mediterranean : a petrographic approach. Unpublished
Ph.D. Thesis, University of Liverpool. KNOX, G.J. (1977) Caliche profile formation, Saldana Bay (South Africa). Sedimentology, 24, 657-674. KoNISHI, K. & SAKAI, H. (1 972) Fibrous aragonite of fresh-water origin in sealed Pliocene Glycymeris yessoensis. Jap. J. Geol. Geogr. 42, 1 9-30. KRUMBEIN, W.C. & GARRELS, R . M . ( 1 952) Origin and
classification of chemical sediments in terms of pH and oxidation-reduction potentials. J. Geol. 60, 1 -3 3 . KuBLER, B. (1973) L a corrensite, indicateur possible de milieux de sedimentation et du degne de tranS·· formation d'un sediment. Bull. Cent. Rech. Pau. 7, 543-556. LACROIX, A. ( 1 941) Les gisements de phlogopite des Madagascar et les renferment. Ann/. geol. Serv. Mines Madagascar, 11, 1-1 19. LAND, L.S. (1 970) Phreatic versus meteoric diagenesis in limestone : evidence from a fossil water table. Sedi-· mentology, 14, 1 75-1 8 5 . LAND, L . S . & EPSTEIN, S . (1 970) Late Pleistocene dia·· genesis and dolomitization, North Jamaica. Sedi mentology, 14, 1 8 7-200. LANG, J. & PIAS, J. ( 1 97 1 ) Morphogenese 'dunaire' et pedogenese dans le bassin intramontagneux de Bamian (Afghanistan Central). Revue Geogr. phys. Geol. dyn. 13, 359-368. LEEDER, M. ( 1 976) Discussion : the genesis of sub-basaltic silcretes fro m the Monaro, New South Wales. J. geol. Soc. A ust. 23, 1 1 3-1 14. LIPPMANN, F. (1973) Sedimentary Carbonate Minerals. Springer, Berlin. LoucKs, R .G . & FoLK R.L. (1 976) Fanlike rays of wrmer aragonite in Permian Capitan Reef pisolites. J. sedim . Petrol. 46, 483-485. LovERING, T.G. & PATTEN, I..E. (1 962) The effect of C02 at low temperature and pressure on solutions super saturated with silica in the absence of limestone and dolomite. Geochim. cosmochim. Acta, 26, 787-796. McFARLANE, M .J. (1975) A calcrete from the Namanga·· Bissel area of Kenya. J. geogr/. Soc. Kenya, 1, 3 1-43 . MABBUTT, J.A. (1 952) The evolution of the middle Ugab Valley, Damaraland, South West Africa. Trans. R. Soc. S. Afr. 33, 333-365. MARSCHNER, H . (1968) Ca-Mg distribution in carbonates fro m the Lower Keuper in N.W. Germany. In :
Int. Ass. Sediment. , 1 st Europ. Reg. Mtg, Bochum,
pp. 1 67-169. FURBISH, W.J. & SANDS, T.W. (1 976) Palygorskite : by
direct precipitation from a hydrothermal solution. Clay Miner. 11, 147-152. GARDNER, L.R. (1 972) Origin of the Mormon Mesa caliche, Clark County, Nevada. Bull. geol. Soc. Am. 83, 143-1 56. GILE, L.H., PETERSON, F.F. & GROSSMAN, R.B. ( 1966) Morphological and genetic sequences of carbonate accumulation in desert soils. Soil Sci. 101, 347-360. GLOVER, E.D. & S IPPEL, R.F. (1 967) Synthesis of mag nesium calcite. Geochim. cosmoschim. A cta, 31, 603613. GOUDIE, A . S . (1973) Duricrusts in Tropical and Sub tropical Landscapes. Clarendon Press, Oxford. GROVE, A.T. (1969) Landforms and climatic changes in the Kalahari and Namib. Geogrl J. 135, 1 9 1-2 1 2. GUPTA, R.D. & RAYCHAUDHURI, S.P. (1973) Clay minerals in different groups of soils in India and factors affecting their genesis. Technology, India, 10, 333-33 8 . HARDER, H . (1974) Illite mineral synthesis a t surface temperature. Chern. Geol. 14, 241-253. HASSOUBA, H. & SHAW, H.F. ( 1 980) The occurrence of palygorskite i n Quaternary sediments of the coastal plain of north-west Egypt. Clay Miner. 15, 77-83. ISPHORDING, W.C. ( 1 973) Discussion of the occurrence and origin of sedimentary palygorskite-sepiolite deposits. Clays Clay Miner. 21, 391-40 1 .
,
Recent Developments in Carbonate Sedimentology in
92
Kalahari calcretes: origin and diagenesis Central Europe (Ed. by G. M uller and G. M. Fried
NETTERBERG, F. ( 1 969a) The geology and engineering properties of South African calcretes. Unpublished
man), pp. 1 28- 1 35 . Springer-Verlag, New York.
Ph.D. Thesis, University of Witwatersrand.
MATTHEWS, R.K.
(1 969) Diagenetic environments of possible importance to the explanation of cementation fabric in subaerially exposed carbonate sediments. In : Carbonate Cements (Ed. by 0. P. Bricker), pp. 1 701 82. The Johns Hopkins Press, Baltimore. MATH IEW , L., THOREZ, J. & EK, C. ( 1 975) Contribution a !'etude des encroutements calcaires dans le cadre de Ia morphogenese et de Ia pedogenese en climat Medi terranean semi-aride et aride : application a Ia region de Taza (Maroc). I n : Co!loque: Types de Croiites Calcaires et Leur Repartition Regionale (Ed. by T. Vogt), pp. 1 14-120. Strasbourg.
NETTERBERG, F. ( 1 969b) The interpretation of some basic
calcrete types. Bull. S. Afr. Archeo/. 24, 1 1 7-1 22. NETTERBERG, F. (1 969c) Ages of calcretes in Southern Africa. Bull. S. Afr. Archeol. 24, 8 8-92. NETTERBERG, F. ( 1 97 1 ) Calcrete in road construction. R.N T.R.R., Pretoria, C.S.l.R. Res. Report, 286, 73 pp.
PANOS, V. & STELCL, 0 . ( 1 968) Physiographic and geologic control i n the development of Cuban mogotes. Z. Geomorph. 12, 1 1 7- 1 64. PAQUET, H. & MILLOT, G. (1972) Geochemical evolution of clay minerals in the weathered products and soils of Mediterranean climates. Proc. int. Clay Conf Madrid, 1 99-203 . PARRY, W.T. & REEVES, C.C. Jr (1968a) Clay m ineralogy of pluvial lake sediments, southern High Plains, Texas. J. sedim. Petrol. 38, 5 1 6-529. PARRY, W.T. & REEVES, C.C. Jr (1 968b) Sepiolite from pluvial Mound Lake, Lyn and Terry Counties, Texas. Am. Miner. 53, 984--993 . PETTIJOHN, F.T., PoTTER, P . E . & SIEVER, R. ( 1 973) Sand and Sandstone: Springer-Verlag, New York. PosT, J.L. ( 1 978) Sepiolite deposits of the Las Vegas, Nevada area. Clays Clay Miner. 26, 58-64. REAMS, M .W. (1 974) Rate of carbon dioxide loss and calcite-aragonite deposition in caves. I n : Proc. 4th Conf Karst. Geol. Hydro!. (Ed. by H. W. Rauch and E. Werner), pp. 1 79-182. West Virginia Geological Survey. REEVES, C.C. Jr ( 1 976) Caliche: Origin, Classification, Morphology and Uses. Estacado Books, Texas. ROBERT, M . , TESSIER, D., ISAMBERT, M . & BAIZE, D. ( 1 974) Evolution of glauconite and illites. Contribu tion to the problem of smectites in soils. Trans. l Oth Int. Congr. Soil Sci. , Moscow, 7, 97-106. Ross, G.J. (1 975) Experimental alteration of chlorite into vermiculites by chemical oxidation. Nature, 255, 1 3 31 34. SALOMONS, W., GOUDIE, A.S. & MOOK, W.G. ( 1 978) Isotopic composition of calcrete deposits from Europe, Africa and India. Earth Surf Process. 3, 43-57. SCHOLLE, P.A. & KINSMAN, D.J.J. ( 1 974) Aragonitic and high Mg-calcite caliche from the Persian Gulf: a modern analogue for the Permian of Texas and New Mexico. J. sedim. Petrol. 44, 904-9 16. ScHULTZ, L.S. (1 964) Quantitative interpretation of mi neral composition from X-ray and chemical data for the Pierre Shale. Prof Pap. U.S. geol. Surv. 391-C, 31 pp. SINGER, A. (1971) Clay minerals in the soils of the Southern Ramat, Hagolan. Israel J. Earth Sci. 20, 1 05-1 1 2 . SINGER, A. ( 1 979) Palygorskite in sediments : detrital , d iagenetic or neoformed : a critical review. Geol. Rdsch. 68, 996-1008 . SINGER, A . & AMIEL, A.J. ( 1 974) Characteristics of Nubian Sandstone-derived soils. J . Soil Sci. 25, 3 1 0-3 1 9 . SINGER, A. & NORRISH K. (1 974) Pedogenic palygorskite occurrences in Australia. Am. Miner. 59, 508-5 1 7 .
MENILLET, F. (1 975) Niveaux calcaires finement rubanes
en milieu continental, hydromorphe et confine a palaeogeographie simple : l'exemple des calcaires de Beauce (Stampien Sup.-aquitanien de Bassin du Paris). I n : Colloque: Types de Croiites Calcaires et Leur Repartition Regiona/e (Ed. by T. Vogt), p. 1 46. Strasbourg. M ILLOT, G. ( 1 970) Geology of Clays. Springer-Verlag, New York. M ILLOT, G . , NAHON, D . , PAQUET, H., RuELLAN, A. & TARDY, Y. ( 1 977) L'epigenie calcaire des roches silicatees dans les encroutements carbonates en pays Subaride Antiatlas, Maroc. Sci. Geol. Bull. 30, 1 291 52. MILLOT, G . , PAQUET, H. & RuELLAN, A . (1 969) Neo formation de l'attapulgite dans les sols a carapaces calcaires de Ia basse Moulouya (Maroc Oriental). C.r. hebd. Seanc. Acad. Sci., Paris, 268, 2771-2774. MILNER, S. ( 1 976) Carbonate petrology and syndeposi tional facies of the Lower San Andres Formation (Middle Permian), Lincoln County, New Mexico. J. sedim. Petrol. 46, 463-482. MORROW, D.W. ( 1 978) The influence of the Mg/Ca ratio and salinity on dolomitization in evaporite basins. Bull. Can. Petrol. Geol. 26, 389-392. MouNTAIN, M.J. (1 967) The location of pedogenic materials using aerial photographs, with some examples from South Africa. Proc. 4th Reg. Conf Africa Soil Mech . Fndn Engng, Cape To wn, I , 35-40. NAGY, G. & BRADLEY, W.F. (1 955) The structural scheme of sepiolite. Am. Miner. 40, 885-892. NAHON, D. ( 1 976) Cuirasses ferrugineuses et encroute ments calcaires au Senegal occidental et en Mauritanie systemes evolutifs : geochemie, structures, relais et coexistence. Unpublished Ph.D. Thesis, Universite de
Droit, Marseille. NAHON, D. & RuELLAN, A. (1 975) Les accumulations de
calcaires sur les marnes Eocene de Ia Falaise de Thies (Senegal). Mise en evidence des phenomenes d'epi genie. I n : Col!oque: Types de Cl·oiites Calcaires et Leur Repartition Regiona/e (Ed. by T . Vogt), pp. 7-1 1 . Strasbourg. NATHAN, Y. & FLEXER, A. (1 977) Clinoptilolite, para genesis and stratigraphy. Sedimentology, 24, 845-
855. NETTERBERG, F. ( 1 967) Some roadmaking properties of South African calcretes. Proc. 4th Reg. Conf Africa
,
Soil Mech. Fndn Engng, Cape To wn, 1, 77-8 1 .
93
N. L. Watts SMALE, D . (1 968) The occurrence of clinoptilolite i n pan
WALKER, T.R. (1 960) Carbonate replacement of detrital
sediments i n the Nata area, Northern Botswana.
crystalline silicate minerals as a source of authigenic minerals in sedimentary rocks. Bull. geol. Soc. Am. 71,
Trans. geol. Soc. Afr. 71, 147-1 52. SMALE, D. ( 1 973) Silcretes and associated silica diagenesis in Southern Africa and Australia. J. sedim. Petrol. 43, 1 077-1 089. STALDER, P.J. (1975) Cementation of Pliocene-Quaternary
145-152. WALKER, T. R. ( 1 967) Colour of recent sediments in
tropical Mexico : a contribution to the origin of red beds. Bull. geol. Soc. Am. 78, 9 1 7-920. WALKER, T.R. & HoNEA, R . M . (1 969) Iron content of modern deposits in the Sonoran Desert : a contri bution to the origin of red beds. Bull. geol. Soc. Am. 80, 535-594. WATTS, N.L. ( 1 976) Paleopedogenic palygorskite from the basal Permo-Triassic of Northwest Scotland. Am.
fluviatile clastic deposits in and along the Oman Mountains. Geologie Mijnb. 54, 148- 1 56. STRAKHOv, N.H. ( 1 970) Princples i of Lithogenesis, vol. 3. Plenum Publ. Co., New York. SURDAM, R . C. & EUGSTER, H.P. (1 976) Mineral reactions i n the sedimentary deposits of the Lake Magadi region, Kenya. Bull. geol. Soc. Am. 87, 1 739-1752. TAYLOR, G . & S MITH , I.E. ( 1 975) The genesis of sub basaltic silcretes fro m the Monaro, New South Wales, J. geol. Soc. Aust. 22, 377-385. THEODOROYITCH,
G.
( 1 96 1 )
A uthigenic
Minerals
Miner. 61, 299-302. WATTS, N.L. ( 1 977a) Pseudoanticlines and other struc
tures in some calcretes of Botswana and South Africa. Earth Surf Process. 2, 63-74. WATTS, N.L. (1 977b) A comparative study
in
of some
Quaternary, Permo- Triassic and Siluro-Devonian calcretes. Unpublished Ph.D. Thesis, University of
Sedimenta'ry Rocks (translation). Consultants Bureau,
New York.
Reading.
THORTENSEN, D.C., M cKENZIE, F.T. & RISTYET, B.L. (1 972) Experimental vadose and phreatic cementation of skeletal carbonate sand. J. sedim. Petrol. 42, 1 62-
WATTS, N . L. ( 1 978) Displacive calcite : evidence from
recent and ancient calcretes. Geology, 6, 699-703. WEAVER, C.E. & B ECK , K.C. (1 977) Miocene of the S.E.
1 67.
United States : a model for chemical sedimentation i n a peri-marine environment. Sedim. Geol. 17, 1-234. WEIR, A . H . , ORMEROD, E.C. & MANSEY, I.M.J. EL. (1975) Clay mineralogy of sediments of the Western Nile Delta. Clay Miner. 10, 369-386. WELLINGTON, J.H. ( 1 955) Southern Africa: a Geographical Study. Vol . I . Physical Geography. Cambridge University Press, London. WINLAND, H.D. ( 1 969) Stability of calcium carbonate polymorphs in warm, shallow seawater. J. sedim.
THORTENSEN, D.C. & PLUMMER, L.N. (1 977) Equilibrium
criteria for two-component solids reacting with fixed composition i n an aqueous phase-example: the magnesium calcites. Am. J. Sci. 277, 1 203-1223. THRAIKILL, J. ( 1 9 7 1 ) Carbonate deposition in Carlsbad Caverns. J. Geol. 79, 683-695. USDOWSKI, E., MENSCHEL, G . & HOEFS, J. ( 1 980) Some kinetic aspects of calcite precipitation (abstr.). Int. Ass. Sediment. 1st Europ. Reg. Mtg, Bochum, pp. 1 6 1 1 63 .
Petrol. 39, 1 579- 1 5 87.
VANDEN HEUvEL, R . V . (1 966) The occurrence of sepiolite
YAALON, D.H. (1974) The significance of Mg for the montmorillonite-kaolinite stability relationships in natural solutions and soils (abstr.). Int. Symp. Water-Rock Interactions, Geol. Surv., Prague, 25. YAALON, D.H., NATHAN, Y., KOYUMDJISKY, H. & D AN, J. ( 1 966) Weathering and catenary differentiation of clay minerals i n soils on various parent materials in Israel. Proc. Int. Clay Conf , Jerusalem, 1 , 1 87-198 ;
and attapulgite in the calcareous zone of a soil near Las Cruces, New Mexico. Proc. 1 3th natn. Conf Clays Clay Miner. 1964, 193-207. VAN HouTEN, F.B. (1973) Origin of red beds : a review, 1961-1972. Ann. Rev. Earth Planet. Sci. 1, 39-6 1 . VEEN, A.W.L. & ARNDT, W . (1973) Huntite and aragonite
nodules in a vertisol near Katherine, North Territory, Australia. Nature Phys. Sci. 241, 37-40. VELDE, B. & HQWER, J. ( 1 963) Petrological significance of illite polymorphism in Palaeozoic sedimentary rocks.
2, 1 3 9-144. YAALON, D . H. & WIEDER, M. (1 976) Pedogenic paly
gorskite in some arid brown (calciorthid) soils i n Israel. Clay Miner. 1 1 , 73-80.
Ain. Miner. 48, 1 239-1254.
(Manuscript received I October I 978 ; revision received I July 1 980)
94
BIOLOGICAL ACTIVITY AND CALCRETE FABRICS
These papers illustrate the range of biogenic features
(1983) describes another early Carboniferous cal
found in calcretes (mainly beta calcretes). Knox
crete which contains horizons composed of small,
describes calcretes from South Africa containing
well-sorted,
abundant evidence of microbial carbonates in tubules
faecal pellets.
well-rounded
peloids
interpreted
as
and needle-fibre calcite. The formation of pedogenic
Carboniferous alveolar septal fabric is also de
packstones and micritization by boring is also de
scribed by Wright (1986a) and this material also
scribed. K.lappa, in two papers, describes root-related
exhibits needle-fibre calcite. Wright speculates that
carbonates in Spanish calcretes. The Microcodium
some ASF represents calcification within mycelial
study favours an endomycorrhizal origin with the
strands forming symbiotic ectomycorrhizal sheaths.
intracellular calcification of root cells. In the second
The role of fungi in forming coated grains in these
paper a wide range of rhizolith types is carefully
palaeo-calcretes is also described. Beier (1987) provides descriptions of microbial
documented.
fabrics in Bahamian Pleistocene calcretes. The stable
Adams describes 'alveolar texture' from early Carboniferous limestones.
Alveolar septal fabric
isotopic compositions are also reviewed.
may occur in any type of pore in a calcrete, either
All of these biogenic (beta-type) calcretes de
within root moulds or in intergranular pores. They
veloped on carbonate substrates, and in the case of
represent the sites of calcification within mycelial
Quaternary forms, in areas which have or had a
strands.
semi-arid to sub-humid climate. The alpha-type cal cretes described by Hay & Reeder, Hay & Wiggins
Peloids not only form by physical processes (Hay & Wiggins, 1980, this volume) but are commonly
and Watts in the previous section developed on
calcified faecal pellets. The extract from Wright
mainly silicate-rich substrates in more arid settings.
Fig. 13. Needle-fibre calcite from an Eocene calcrete near Montserrat, Barcelona, Spain. Note the mainly random distribution pattern but with a bundle of tightly-arranged needles crossing the field of view. From studies of Quaternary calcretes (see text) it appears that all these needles of low-Mg calcite formed within mycelial bundles, as parallel masses. Surprisingly these needles show no evidence of extensive overgrowths.
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
95
Reprinted from
Sedimentology (1977) 24 657-674
Caliche profile formation, Saldanha Bay (South Africa)
GORDONJ. KNOX KoninklijkejShell Exploratie en Produktie Laboratorium, Rijswijk (ZH), The Netherlands
ABSTRACT A sequence of gradational lithification events can be observed in caliche profiles, in the Saldanha Bay area (South Africa), from friable lightly cemented aeolian calcarenites or littoral shelly deposits through an intermediate semi-indurated zone to an upper strongly indurated zone (calcrete). Lightly cemented sediment fabrics exhibit bridge and meniscus cements, micritic druses and vadose compaction phenomena. The middle semi-indurated zones exhibit coated grains in which irregular borings and/or tubules with tangential acicular fibres contribute to coated grains. Random networks of acicular fibres also occur in void spaces. In fully indurated upper layers of the caliche profiles, fabrics of micrite and microspar (in voids) occur in complex brecciated macro-fabrics. The features represent changes in a sequence from the friable primary sedi ments to the calcretes. Fresh-water vadose flushing leaches grains and causes for mation of meniscus and bridge cements and uneven druses. In the middle zone, inorganic processes are aided by the action of micro-organisms; fungi, bacteria or algae which produce tubules and irregular borings; the overall effect of which is to break down original detrital carbonate particles and enclose them in a crypto crystalline micrite. The acicular fibres probably result from evaporation of super saturated solution. Mechanical processes cause fracturing, which repeated many times gives complicated brecciated fabrics within the upper indurated zone.
INTRODUCTION
Forming outcrops around Saldanha Bay, South Africa, are dominantly cal careous coastal deposits (Fig. 1), known as the Dorcasia Limestone (Du Toit, 1917), Langebaan Limestone (Visser & Schoch, 1973) or 'Coastal Limestone' (Siesser, 1972). These deposits consist of mixed aeolian sand and littoral shelly deposits. Depositional age is probably Middle to Late Pleistocene. Commonly, the deposits have developed a caliche profile, which were best developed as a hard, strongly indurated surface of calcrete. Similar profiles, more or less developed are present within sections of the coastal deposits (Fig. 2). Siesser (1973), has carried out studies on the diagenesis of calcretes from South Africa, including samples from Saldanha Bay. He described diagenetically-formed ooids and intraclasts which he interpreted as being the result of carbonate-mud Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
97
G. J. Knox
srud1Cd R
-"
5
"
0
E
U
P
T
U
H
B
L
A
I
F
C
R
I
C
A
.""" """ """ """ """'-::----.,-J 0�---k---�IQ m
'J
r
0
G
N
CJso�rs � Modern dune sands and eol1oniles s tT: :IJ �����; :�c���ew��ei���;ce�: ����,j���e , 6lli]} Basement Fig. 1. Local geology of Saldanha Bay (modified from Visser & Schoch, 1973).
98
� >Om
1 1
� �
:]
above
305 . 0m
2980- 3050m
:] : 1: :1 V""1 l_ _j:
1� £�
(f) __J (f) (f)
_
25m
0 lL 23 20-2980m
t:=L=d _ · _:: 8 PTI
-{}-
lCD
20m
�
o o o o o<:Do
Fr. able
oeolionite calcarenites without
Plont root tubes
Aeolian calcarenite
V�rtebrotes bones
Littoral shelly dep osit
&, r}i
Foromin"it�ra
� J!
Gostropods Echinoid!
J5
Lomellibronchs
_L_ Cross-bedding
Land gastropods common and Foromrnlfero
-Q-
Differential cementol!on has preserved the
T l I :1
!Om
j
21 50· 23 20m
Ba!ement conglomfro1e Basement
"' ""
pale cream to chocolate brown 20.20.21 5 0.m
AeoiKll'l 'll e sand wdh forams. Traced loterolly,the
2 0m
..,
bed has o convell. upper surface
19.tW ·2020m t8.40-1980m
.g,
Irregular coltche bond
Aeolianite sand with land gastropods
ond
1.60- 2.10m
root
I
n l :1- 1:
Hard col1che formot10n of seve� I cenlimttru thick, decreases down word to Qlve d1scontinuous
traces preserved
concret1onory layers and nodules of col1che .
17.90-t840m
COIJChe bond passing downwards into
t 5 5 0-1790m
oeollcmte, m wh1ch cross· bedd1ng may be I
0 -
l 60m
Gra1nstones composed of shelly lroqments lomlnlliJOn common w1fh much voflol•on in
!.Om A•oltonrte porllolly cem•nted
14 30-14 GOm
lrrequlor bond o f coiJche nodules
1000-14.30m
AtoiJonJt• more or less cemented w1th
�..ff -o-
orom sue. More or tess cemented . I 0m
J.Om local
preservol10n of root marks and burrows.(?) Land qosrropods, foroms and bioclastiC groins
Approumotety
a
10m
Aeolionite stron.qly Cemented by caliche.
lor
Caliche
a Tew "Centime tres downward
Degree of cementation decreases downwards, where homogeneous friable
eo lionile occurs .
Land gastropods common
fairly common.
l______1 :l.._: ___l i;Q0_0
weolhered surface w1!h solution channels.
has been bored
() ..
�
'Cj>
� ;:,
5· ;:s
poriiCul o r l y from tome l ltbronchs. Shelly
14 60. 15.50m
Q
�
to powdery rimy material. Colora! ron varies from
_bi:
T :1:
IJnconsolidotod blollln sand
Zone of caliche form•ng drscontinuous toyers
.
�?'c:q
Aeolian colcor�nite in whith
cross-bedding is portly dutroyed by caliche formation.
and concretions.Locolly,cotrche has broken down
�a;:;_.: ..:o:·.
� ..
( brokon)
> (9 0 __J 0 :r: f__J
...L
""1 -{}-
us
Indurated calicho nod ules of va r i o sizes
r
b o k en
Burrows
cciJcheloyers
l _:j j_ 1
Indurated caliche poveml!nt
broken
complete and
recognisable sed1mentory structure
outlrne of roots and on1mol burrows l?l
_
complet� and
{
L&
porliolly preserved by for motJOr> of thin
15m
{
CD
::� 1 �rj£ E2=23 �
--{
Col1che surface, rubbly d1scontmuous masses of concretion
J: :
T
Key
Loose blown sand
above mean sea-level
b
Fig. 2. Examples of aeolianite and littoral shelly deposit sections on and within which caliche profiles are present. (a) Hoedjies Point. (b) Blauwaterbaai.
G. J. Knox
precipitation around detrital particles associated with slight expansion of the sedi ment. James (1972) in a study of surface calcareous crust profiles on Barbados has observed similar textures, but on a small scale also recognized tubule and needle-like structures. Recent examination of Saldanha Bay caliche samples collected on land and from beneath the bay shows that similar tubule structures to those described by James (1972) occur in these sediment fabrics. A sequence of fabric changes can be recognized from unlithified sediment to indurated calcrete caused by both inorganic and organic activities.
GEOLOGICAL SETTING
Saldanha Bay lies about 160 km NW of Cape Town and is set in an area of gentle relief without prominent drainage lines. Low hills may be topped by granite basement while lower areas consist of dunes. Flat-lying scrub-covered areas a few metres above sea-level may be covered by a caliche or calcrete. Within the granite basement are localized pockets of Miocene/Pliocene phosphatic sands. The coastal deposits are described by Siesser (1972), who relates them to various 'Coastal Sandstones' occurring in South Africa. The maximum thickness of the deposits recorded by Visser & Schoch (1973) in the Saldanha Bay area is 88 m. The average thickness is probably approximately 30 m. The aeolianites can be subdivided into fine-grained homogenous and coarser cross-bedded calcarenites. Fine-grained sands contain 'dikaba' root structures (Fig. 5b) similar to those defined by Glennie & Evamy (1968) and appear to have been deposited rather slowly, allowing plants to establish themselves from time to time. Coarser cross-bedded fossil dunes may have been established more rapidly under stronger wind conditions and may only rarely have sustained plant growth. The littoral shelly deposits consist of varying thicknesses of calcareous sands and shelly grainstones, which cover planation surfaces within dune sequences and then in turn may also be covered by aeolian sands. Interfingering of these deposits was related to their close association near to a Pleistocene fluctuating sea-level (Siesser, 1972). Equivalent environments occur today with development of coastal-associated recent dune and beach deposits. (Fig. 1). Comparison of thin sections of both aeolianites and shelly littoral sediments shows that the detrital constituents are essentially the same in both consolidated and present-day sediments. They consist dominantly of carbonate allochems and silici clastic detritals. The latter consist mainly of quartz and feldspar with lesser amounts of granite/gneiss lithoclasts, glauconite and rare heavy minerals. Carbonate allochems consist mainly of recognizable bioclastic fragments of echinoderms, foraminifera, gastropods and lamellibranchs. Some fragments show evidence of boring, which in the aeolian sands occurred before incorporation into a wind-blown environment. These bored cavities within allochems are of 50-100 l.l. dimensions and are not to be confused with the much smaller irregular borings described below. Siesser (1972) 1973) has determined by X-ray diffraction and atomic adsorption spectrophotometry that mineralogically the carbonate consists of low Mg-calcite. Cathode luminescence properties of the carbonate allochem fragments showed them to be non-luminescent, as were also the caliches. Stalder (1975) suggest this to be characteristic of fresh water influence. Disappearance of aragonite and Mg-calcite as original mineralogical 100
Caliche profile formation
constituents of shells of certain marine organisms is in agreement with this. Some aragonite, however, is present in land gastropod shells (Siesser, 1972).
CALICHE PROFILES
Caliche profiles have commonly developed in S Africa both at the present day and in the past (Netterberg, 1969a). They are not restricted to formation on calcareous substrates but are also formed on igneous basement or on soils developing from the latter. Usually, the top surface is indurated and therefore termed 'calcrete' (Lamplugh, 1907) or the 'K master soil horizon' of pedologists (Gile, Peterson & Grossmann, 1965). In South Africa Netterberg (1967, 1969b) has called the upper surface 'hardpan calcrete'. Morphologically, the Saldanha Bay examples vary (Fig. 3). Continuous pavements of indurated caliche are developed on calcareous sediments and cover widespread low-lying areas with considerable lateral variation in thickness. The indurated upper surface passes down into a less indurated caliche, which eventually grades into the substrate in the case of the coastal calcareous deposits. In addition continuous pavement ( ± 2 m thickness) may grade laterally into lenses, nodules, bifurcating sheets (10 em thickness) or irregular aggregates of indurated nodules (10-15 em diameter). Funnel-shaped structures 1-2 m high have filled solution hollows and cut through fossil dune profiles, which rest on the calcrete surface of a fossil caliche pro file. The downward development of the funnel has been arrested by the hard indu rated caliche layer (Fig. 5b). Thicknesses of the upper indurated caliche or calcrete on the present-day surface or interstratified within the coastal deposits may vary from a few centimetres to greater than a metre. In a few outcrops, caliche were observed lying directly over a non-calcareous substratum of fresh or weathered basement. Those resting on fresh basement are very similar in composition to examples developed upon aeolian calcareous deposits. Weathered basement becomes totally altered, as can be seen in a good quarry outcrop by Saldanha Municipality Workshop. Kaolinized basement passes upwards into clay residue with quartz wash bands. Above these is an 80 em caliche profile consisting of two parts,. (1) a semi-indurated zone containing irregular broken a-quartz crystals (derived directly from weathered basement) and some rounded quartz grains, cemented by micritic calcite with meniscus druse and (2) an upper part consisting of a calcrete rind which is extremely hard. Locally rinds may form directly on weathered basement. The macroscopic internal structure is rather complex (Fig. 4a, b) and similar to calcretes from New Mexico (Bretz & Horberg, 1949, Fig. 4; Ruhe, 1967, Fig. 3). They have an overall brecciated appearance in which variously sized fragments of calcrete, rare detrital quartz and feldspar, and terrestrial gastropods are enclosed in concentric crusts in which slight differences in texture and fabric are visible owing to colour and etching. Hollows are present which have been filled by blown sand or fragmented material from the same layer. Some infills have become lithified by recent cementation processes. Numerous fractured surfaces are present which penetrate into the calcrete; some have become closed by precipitation of carbonate cement. 10 1
G. J. Knox POSITION
IN SECTION S
THICKNESS
TYPICAL FORM (Indurated caliche)
aSb, Fig.2.
a) 30,20 and 18.2 metres
I
b) 2.0 qnd 0.9metres
0
lm
2m
al
Massive, laterally continuous thickness. 22.5 and 14.4 metres
n
pavement with
variations in
0
lm
b)± 1.7 metres
2m
m
a l 22.5 and 16.0
Irregular masses and nodules of var ying d imensions .
0
0 .2m
0.4 m
0.6 m
0 Ill a n d ;z: were nol observed in tht preclst sections illustrated
0.2 m
m Fig.2 however, both were commonly observed m outcrops around
Saldanha
Boy
. : . . .. . . . . . . �
0.4 m
.
.
.
.
.
Aggregates of nodules in len ses more or less continuous. 0
Massive vertical funnels with numerous cavities. Often restino on a continuous pavement.
l m Note : Different combinations of J,n,m,.&,and 'Z may commonly occur.
Fig. 3. Different morphologies of indurated caliche. The calcareous sediment surrounding these indurated forms is usually partially indurated.
After formation, some profiles have been destroyed by weathering along slopes. Others have been covered by aeolian littoral deposits. Excellent examples of buried caliche (Fig. Sa, b) occur in a recent road and railway cutting on the approaches to Hoedjies point, where all types are present. At Blauwaterbaai a caliche profile has been partly eroded and covered by a littoral deposit. The caliche has formed on aeolian sand and is strongly indurated on its upper surface. The upper few centimetres 102
Caliche profile formation
Fig.4a
Surface of a caliche layer near Olifantskop. Concretionary surfaces are
visi b l e within which brecciated fragments of caliche occur. A hollow
(h)
has been filled
by fragments and blown sand and is showing cementation. The complexity of this surface is shown by polished s a mples in Fig.4b
-
/� I
I I
/· I
!A
Aeolianite sand cemented by calcite
I
Calcrete formed from shelly deposit
IB
Original sand recrystalled so that original texture no longer recognisable
II
Concretionary calcrete plus sand grains
ill
Concentric calcite deposits,sometimes showing irregular patterns
II
Sand-filled hollow cemented by calcite
ill
Concretionary calcite layer with sand grains
l'il
Concretionary calcite-filled fractures and surfaces
I , II ,ill .... etc represent calcrete generations
FIG.4b INTERNAL STRUCTURE OF TWO CALCRETE SAMPLES 103
G. J. Knox
Fig. Sa. A buried indurated part of a caliche profile on Hoedjies point. The layer consists of a calcrete breccia over which a concretionary rind has formed. Underneath, semi-indurated caliche contains the remains of terrestrial gastropods, Above is a later aeolian sand.
Fig. Sb. Funnel-shaped indurated caliche (f). Notice that the structure widens towards the base, where it rests on a fossil brecciated calcrete (c). This calcrete passes downward into friable, loosely cemented aeolianite (a). Both above and below the calcrete are 'dikaba' root structures (d) preserved by differential cementation. Location railway cutting between Saldanha and Hoedjies Point.
104
Caliche profile formation
are pitted by solution channels and possible borings which are truncated by the contact with the shelly deposit. The shelly deposit passes upwards into fine-grained sand and a caliche profile at the present surface, which is covered by recent dunes or a sparse loamy soil. In all cases primary structures such as bedding, roots, and most fossils are eradicated in the caliche profile development. Petrography of the caliche profiles
Within the profiles the caliches can be traced downwards from an indurated upper surface through a softer zone in which the primary rock becomes recognizable. The degree of alteration and cementation increases upward such that the original fabric and sedimentary structures are eventually no longer recognizable. The sequence of alterations described below are broadly in order of increasing calichification. Figs 6, 7 and 8 also represent the change from original sediments to hard indurated caliche. Fig. 9 summarizes these changes in relation to a generalized caliche profile. Textural features of the caliche profiles
Examination of caliche profile samples with the optical petrographic microscope and the scanning electron microscope (SEM) shows the following textural features: Leached particles
Numerous bioclastic fragments exhibit internal voids (Fig. 6b) lined by replacive microspar crystals ± 20 1-1m length. These voids are chiefly observable in the primary sediment grainstones which are extremely lightly cemented (Fig. 6a). Bridge and point-contact cement
Localized at grain contacts or as fine bridges with maximum lengths of 300 )lm (Fig. 6c) are patches of cement. Point-contact cement consists of a fine microspar; while bridges appear to have a micritic appearance. Examination with the SEM showed that some of the bridges consist of regular to irregular build-ups of sym metrical bundles of rods or fibres (Fig. 6d), between which considerable interparticle porosity occurs. Associated with these cementation features are the presence of long and slightly concavo-convex contacts between grains, indicating vadose compaction. This is a process by which fresh-water films around grains dissolve carbonate to allow original point contacts between grains to become flatter. Uneven druses
Numerous fine-bladed crystal druses (Fig. 6c) have developed on detrital grains and allochems. They point outwards into interparticle pore space and typically form broad crystals approximately 10-20 )lm length. Thickness varies considerably with local formation of microstalactitic dripstone features. Tubules
Semi-indurated zones exhibit meandering tubular structures (Fig. 6e, f ). Typically they may follow interparticle pore space or form coatings on· detrital grains. Branching structures are present, but often the tubules occur as broken aggregates. The tubules are generally hollow with walls consisting of normally orientated calcite crystals, 105
Fig. 6. (a) Calcarenite in which contacts owing to vadose compaction occur (c). Micritic bridge (b) and meniscus structures (m) are present between grains. Considerable primary pore space (0) is visible. (b) Aeolian calcarenite in which the centres of some bioclastic allochems are leached leaving ap1icritic margin. A drusy calcite (c) lines these intraparticle voids (v). A thin micritic coating (m) covers most of the grains. Primary pore space is considerable (0). (c) SEM photograph of a bridge structure (b) extending between grains (g). The latter are partly covered by a drusy calcite (c). (d) Close-up SEM photograph of bridge structure (see square 6c), which consists of regular bundles of acicular fibres orientated in two planes. The size of individual fibres is approximately 10 J..lm. (e) SEM photograph showing the outline of coated grains (g) in which the coatings consist of aggre gates of tubules (t) and some fibres (f ) and micritic calcite. (f) SEM photograph exhibiting coated grains in which the coatings consist mainly of tubules and micrite. Some of the tubules may continue for some distance, following a winding course between coated grains.
Fig. 7. (a) Composite-coated grain with other coated grains. The composite grain contains quartz (q) and a bioclastic allochem (a). The latter with the coating (c) shows an irregular fibre riddled with borings. A second coated grain also exhibits borings (b) in the micritic coating. Micrite has filled the intraparticle space and this too has a few borings. (b) Micrite from within a funnel-shaped body riddled with irregular borings (b) or build up of tubules. Some pore space occurs (0). (c) SEM photograph showing a void filled with an interlocking network of acicular fibres. A quartz grain (q) with an irregular micritic coating (c) occurs in the foreground. Tubules (t) are present to the right. (d) SEM photograph exhibiting coated grains (g) in which the coatings are made up of tubules (t) and tangential networks of acicular fibres. (e) SEM view of interlocking acicular fibres bound together· by membraneous coatings. Note the furrow running down the fibres-. (f) SEM view of the surface of an indurated caliche (calcrete) which consists of a microcrystalline aggregate of micritic calcite. A few voids (v) occur which may be lined by microspar calcite.
107
G. J. Knox
Fig. 8. SEM photograph of microspar calcite (c) filling a void within a microcrystalline micrite containing detrital quartz grains (q) and carbonate allochems (a).
TERMINOLOGY
MICROSCOPIC TEXTURES
CALICHE PROFILE
Netterberg This paper
{ 1969) -T-�--i----;;;-;;;�� -------t-----;:::::;: : =;------: :;::;:::;::; w > w a:: u -'
<'i
z
0:
0 a::
I
0 z
0 w >-
0 w "-
u
UNEVEN
-'
::!'
w
I
(f)
(f) ' Z W >-
Oo
"'w 0::>-
I I
Zo
w
jcn
>--
z w
::!'
0 w
(f) >-
a::
�
a:: a.
w
/
-�/// / / �/ /.
-;'w
!DW �::!' a::w u..u
/
/
/
/// . .
/
::!'
-o w>...JZ
/
/
DRUSE
Yh .
. ·
.
>. . ·
.
· :- .
.
.
Fig. 9. Schematic relationship of the position of observed textural features in the caliche profile. 108
Caliche profile formation
which give the tubules a drusy exterior appearance. The central cavity is approx. 1 f.Lm in diameter. Irregular borings
Associated with the tubules and of similar dimensions are intensely bored allo chems and coating (Fig. 7a). Within the allochems they may be true borings, but within the coatings they may, in fact, be build-ups of tubules (Fig. 7b) around detrital grains or borings formed subsequent to the formation of the coatings. Acicular networks
Randomly patterned networks of calcite fibres (Fig. 7c) fill some intergranular spaces. Individual fibres average lengths of 30 f.Lm and thicknesses of slightly < 1 f.Lm. Contacts between 'fibres' appear to be marked by membraneous sheaths which give the networks rigidity (Fig. 7e). Networks may also form tangentially (Fig. 7d) t"o detrital particle surfaces to form coatings. Acicular networks are often associated with tubules and, like the latter, are present in semi-indurated caliches. Micritic fabrics
Strongly indurated caliches have an overall cryptocrystalline calcite fabric (Fig. 7f ) in which it is sometimes difficult to recognize original carbonate allochems which have faded out. Where allochems are still recognizable or where quartz grains are common the texture is not grain supported though some grain contacts occur. Some times, laminated micrite textures are present which lack any siliciclastic or carbonate grains. These represent surface rinds which are similar to rind-like fabrics, which Multer & Hoffmeister (1968) described from the Florida Keys. Completely indurated caliche has a mottled earthy appearance in thin section, in which a few quartz grains and ghosts of original allochems and chitinous exoskeletons of insects are visible. Calcite is totally cryptocrystalline. Tubules, borings and acicular networks are no longer recognizable and the rock is extremely hard and has low permeabilities. Similar fabrics are known in calcretes from Shark Bay, Australia from equivalent outcrop locations (Logan, Read & Davis, 1970, Figs 5-3, 9 and 8-6). Sparry fabrics
(1) Partially or completely filled vugs with equant blocky calcite (individual crystals approx. 50 �tm) occur in consolidated caliche (Fig. 8). (2) Concentric layers of blocky calcite occur within fractures in the indurated caliches. Relationship of the textures to caliche profiles
The textural features described above can be related in a general way to zones in the caliche profile (Fig. 9). Intraparticle voids, vadose compaction phenomena, bridge and meniscus structures and uneven druses are found in the lowest part of the profile and occur in the recognizable primary sediment. Tubules, ramifying borings and acicular fibres are present in the middle semi-indurated zone, while micritic fabrics and sparry calcite fillings of vugs and fractures occur in the uppermost indurated zone of calcrete. 109
G. J. Knox
In the middle and upper zones some overlap and combination of textural types occurs. Micritic rinds and fabrics may grade into coated grains in which the coating consists of acicular fibres and tubules. In other cases the coating is purely micritic or may be a combination of borings and micrite.
Origin of the fabrics
Progressive textural changes eventually result in the upper hard indurated surface . Some expansion of the original fabric appears to have taken place during these changes which are the complex result of mechanical, inorganic and organic processes in a vadose environment. Mechanical processes are represented by expansion in the surface layer of calcrete with the formation of fractures: creep along slopes, settlement of underlying sediment and the wedging action of plants have all led to the formation of fractures. These have subsequently been infilled by overlying unconsolidated wind-blown detritus or encrusting rinds of calcite. Repeated fracturing and infill eventually give complex brecciated fabrics (Fig. 4b). At lower levels, intraparticle voids, vadose compaction phenomena, bridge and meniscus cements and uneven druses are characteristic of freshwater vadose diagenesis. Similar fractures have been described by Land, MacKenzie & Gould, 1967), who devised a textural scheme for diagenetic evolution in Bermudan Pleistocene limestones. The tubules and irregular borings are sometimes part of composite textural structures, which Siesser (1973) has called diagenetically formed ooids and intra clasts. He describes their origin as being due to precipitation of carbonate mud around incipient ooids and intraclasts, which at the same time involved some expansion of the sediment and dissolution of original grains. Similar features described by Siesser (1973) mainly occur in the semi-indurated zone. Thin section examination shows that coated grains may commonly show point or tangential contacts which interrupt the coating. These contacts are probably original and alleviate the need for expansion of the whole sediment to incorporate the coatings. Although some expansion may occur, original pore space is high and allows coatings of micrite to build-up. On the SEM photographs some of these coatings are seen to consist of tubules and tangential needle fibres. The tubules record the presence of micro-organisms. In addition irregular borings, present in the semi indurated zones, were not present in the original sediments. As they also occur in coatings, they have formed during development of the caliche profile. They may have been caused by subaerial equivalents of algal or fungal boring organisms, which have caused gradual centripetal replacement of allochems by micrite (Bathurst, 1971, p. 383). The irregular borings were probably caused by the same organisms which produced the tubules.James (1972) has observed similar tubules from weathered zones. The tubules described byJames penetrate grains and thus may be similar in part to the ramifying borings from SaJdanha Bay caliches. In fact, James (1972) comments that such tubules are similar to blue-green algae tubes. In a microl biological investigation of two samples from the Nari-lime-crust, Israel, Krumbein (1968) found that a well developed microfl.ora, dominated by algae including bacteria, fungi and actinomycetae, was present. In addition he observed, in culture experi ments, that this microfl.ora could produce a large number of calcite crystals. Thus, the tubules may be the remains of fungal, algal or even bacterial activity, which 110
Caliche profile formation
enclosed and coated allochems. It is not clear how the ramifying borings were made but it is likely that similar organisms were able to produce them. Later recrystal lization destroys these fabrics and gives the micritic coatings of the indurated caliche. James (1972) suggests that such needle fibres are the result of crystallization from highly supersaturated solutions caused by strong near-surface evaporation of void solutions. Salt linings of fractures/voids indicate that locally concentrated solu tions have precipitated halite. Salt spray blown inland by prevailing SSW winds, after incorporation and evaporation in caliche zones, would become very saline. Such spray could also carry Mg2+ ions, which would locally augment the Mg/Ca ratio. In non-marine speleothem environments, Miiller, Irion & Forstner (1972) observed that increased evaporation tends to increase the Mg/Ca ratio. According to Folk (1974), increasing the latter ratio will tend to poison the sideward growth of CaC03, so that fibrous crystals or elongated rhombs develop. A second possibility of origin is through the activities of micro-organisms. The needle fibres are of similar size to crystals produced in cultured solutions by micro flora from the Nari-lime-crust (Krumbein, 1 968). Gleason & SpackmaQ. (1973) have observed that blue-green algae produce fresh-water lime mats in the southern Ever glades. Different species can produce equant, acicular and encrusting structures, which become lithified over short distances. Ward (1970, inJames, 1972) observing similar needles in fossil weathering surfaces in calcarenites in Mexico, suggested that they are related in origin to fungi. The upper indurated zone has largely lost the above textural features. Two dominant fabric types remain: an overall cryptocrystalline texture, which tends to enmesh original grains, and destroy caliche textures of the semi-indurated zone and pockets of sparry calcite, which fill vugs and fractures. The micritization appears to have been a constant process which eventually changes original allochems and earlier cement structures. Chilingar, Bissell and Wolf (1967, p. 194) report the findings of Wolf(l963) that algae can cause the precipitation of crusts on detrital particles and cavity walls in a beach-rock environment. In addition, deposits and crusts of algal origin were found changed to a dense cryptocrystalline micrite. Thus, the absence of tubule structures and fibres in the upper caliche zone of Saldanha Bay samples does not exclude their previous existence, because overall micrite formation has destroyed them. As permeability is low within the indurated caliche, micritization can take place both within the sediment and on the surface to form rinds. In addition, pockets of solution may be held for longer periods in vugs and fractures where a slower crystal lization rate can take place to develop the sparry calcite. The final fabric of the indurated surface zones is very complex and is a result of the processes above. However, this is not a static situation; fractures, breccias and solution hollows are continuously forming. They may be infilled by rubble from the calcrete, wind-blown material or calcite rinds. Eventually, they become as tightly cemented as the rest of the horizon.
Source of the calcium carbonate
Detrital carbonate allochems are an obvious source for the calcium carbonate. Siesser (1 973) has suggested that solution and reprecipitation of the latter takes place in similar coastal sediments. Saldanha Bay samples show clear examples of solution 111
G. J. Knox
at grain contacts and within particles to produce voids. Bridge and meniscus cements indicate the beginning of calcium carbonate dispersal into pore spaces. Dispersal is augmented at a later stage by the activities of micro-organisms. Centripetal replacement of allochems was accompanied by the development of needle fibres and tubule structures which further clogged pore spaces. At the surface of these caliche profiles aggradation was dominant. A continuous supply of aeolian carbonate detritals is blown onshore by prevailing south to south westerly winds. Fractures, hollows and sinks in the indurated caliche profiles become filled with mixed siliciclastic-carbonate detritals. These younger infills exhibit varying degrees of calicification. Repeatedly, as the section on Hoedjies Point illustrates (Fig. 2a) individual caliche profiles (Fig. 5a, c) have been overwhelmed and covered by dunes. After stabilization these deposits have in turn developed caliche profiles. The area has a semi-arid climate with an average rainfall of 25 em per year (Mountjoy et al., 1970, p. 471). Thus, airborne dust may carry considerable amounts of calcium carbonate. The profiles developed on basement rocks probably derived some calcite by this means. In Texas and New Mexico, Brown (1956) and Ruhe (1967) consider that carbonate dust contributes to caliche formation especially in areas where a carbonate substrate is lacking. James (1972) notes that salt spray adds calcium carbonate to profile surfaces. The existence of halite on the surfaces of fractures indicates that evaporation of seawater trapped in the profiles has been common. The Saldanha Bay profiles clearly exhibit aggradation and association with coastal sediments. Solution, alteration and reprecipitation of original carbonate grains, par ticularly in pore spaces would ultimately reduce the thickness of the caliche profile compared to the original coastal deposit. This tendency is counteracted by the con tinual supply of material on aggrading profiles. Calcium carbonate is carried within the profiles in solution. At different levels in the profile solution has taken place. Periodic rainwater and seawater spray will per colate downwards. Aristarain (1970) from a study of the geo-chemistry of New Mexico caliches concluded that chemical elements were transported by water moving from top to bottom, thereforer educing the thickness of the profile. Multer & Hoff meister (1968) also invoke movement of calcium carbonate downwards by rainwater. Around Saldanha Bay, downward percolation clearly occurs freely in newly stabilized coastal deposits with high permeabilities. During this flushing bridge and meniscus cements develop concurrently with vadose compaction. During extended drought, ascending capillary water, supplied from a brackish water table, and salt water spray evaporate and precipitate micrite. Micro-organisms aid this process by developing characteristic textural types. Eventually, permeability is effectively destroyed so that capillary suction ceases and rinds develop on the surface. Mechanical fracturing allows restarting of processes. In some cases caliche profile development may end when the caliche profile becomes impermeable (Netterberg, 1969b). Around Saldanha Bay, it is more likely that the developing profiles are overwhelmed by aggrading coastal deposits.
CONCLUSIONS
Widespread caliche profiles occur in other parts of South Africa (Netterberg, 112
·
Caliche profile formation
1969a, b; Viljoen et a!., 1975, Fig. 6). Morphologically, the examples described by Netterberg (1967, 1969b) are similar to the Saldanha Bay types and others described in the United States (Bretz & Horberg, 1949; Brown, 1956; Gile, 1961; Gile et a/., 1965; Ruhe, 1967, and Lattman, 1973), Barbados J ( ames, 1972) and Western Australia (Logan et a/., 1970). Caliche profiles have developed on both calcareous and non-calcareous substrates. A generalized profile consists of an upper indurated zone of calcrete passing down wards into a zone of lesser induration, which in turn passes into the original substrate. The indurated caliche can exhibit many forms such as laterally continuous pavements, nodules, lenses, bifurcating sheets, funnels and rinds. In all cases, the process. alters and eradicates original substrate fabrics and structures by dissolution, mobilization and precipitation of calcium carbonate. The internal fabrics and textures are very similar to those described byJames (1972) on Barbados. Siesser (1973) observed the basic fabrics but did not recognize needle fibres or tubule structures. These and centripetally micritized allochems which show irregular borings indicate the activities of micro-organisms during caliche profile development. Multer & Hoffmeister (1968) and Lattman (1973, Fig. 4b) also indicate that specifically algae may have been present, while Krumbein (1968) recognized a well developed microflora in the Nari-lime-crust, Israel. Conversely, many caliche profile developments are related solely to inorganic and mechanical processes (Aristarain, 1970; Bretz & Horberg, 1949; Netterberg, 1969b, and Siesser 1972). A simple answer to this may be that biogenic structures are destroyed as calichification proceeds. The Saldanha Bay profiles show that early bridge and meniscus cements, intraparticle voids and clearly discernible biogenic structures, in coastal calcareous deposits, are later destroyed by (a) aerobic decay of the organic matter of micro-organisms and (b) secondary destruction of biogenic . carbonate structures and fabrics by recrystallization. The biogenic structures may have been preserved by smothering of the profiles by later coastal deposits. On alluvial plain or pediment surfaces the tendency will also be for biogenic structures to be destroyed. Nevertheless, careful study of semi-indurated parts of profiles may bring biogenic structures to light in areas where caliche formation is considered only due to mechanical and inorganic processes. The Saldanha Bay examples illustrate mechanical, inorganic processes and also an influence owing to micro-organisms. ACKNOWLEDGMENTS
Permission to publish this paper was given by Salcon (a joint venture of the Amsterdam Ballast Dredging and the Royal Netherlands Harbour Works Company) and Shell Research B.V., The Hague. Logistical support for field work was given by the staff of Salcon and C. Hartman. Early drafts of the manuscripts were read by A.J. Keij, M. Epting and W. Schollnberger, who made constructive suggestions for its improvement. The SEM photographs were taken at the Central Laboratory, T. N. 0. Delft, The Netherlands. REFERENCES ARISTARAIN, L.F. (1970) Chemical analyses of caliche profiles from the high planes, New Mexico.
J. Geol. 78, 201-212.
113
G. J. Knox BATHURST, R.G.C. (1971) Carbonate Sediments and their Diagenesis. Developments in Sedimentology,
12, Elsevier Publishing Co., Amsterdam. BROWN, C.N. (1956) The origin of caliche on the north-eastern Llano Estacado, Texas. J. Geol. 64,
1-15. BRETZ, J.H. & HoRBERG, L. (1949) Caliche in south-eastern New Mexico. J. Geol. 57, 491-511. CHILINGAR, G.V., BISSELL, H.J. & WOLF, K.H. (1967) Diagenesis of carbonate rocks. In: Diagenesis
in Sediments (Ed. by G. Larsen and G.V. Chilingar), Developments in Sedimentology, 8, pp. 179-322. Elsevier Publishing Co., Amsterdam. Du ToiT, A.L. (1917), Report on the phosphates of Saldanha Bay. Mem. Geol. Soc. S Afr. 10, 38 p. FOLK, R.L. (1974) The natural history of crystalline calcium carbonate: effect of magnesium content and salinity. J. sedim. Petrol. 44, 40-53. GILE, L.H. (1961) A classification of Ca horizons in soils of a desert region, Dona Ana County, New Mexico. Proc. Soil Sci. Soc. Am. 25, 52-61. GILE, L.H., PETERSON, F.F. & GROSSMAN, R.B. (1965) The K Horizon: A master soil horizon of carbonate accumulation. Soil Sci. 99, 74-82. GLEASON, P.J. & SPACKMAN, W. (1973) The algal origin of a fresh-water lime mud associated with peats in the Southern Everglades. Abs. Prog. Geol. Soc. Am., Ann. Meetings SE Sect. 5, 5, 398. GLENNIE, K.W. & EvAMY, B.D. (1968) Dikaka: Plants and plant-root structures associated with aeolian sand. Palaeogeogr. Palaeoclim. Palaeoecol. 4, 77-87. JAMES, N.P. (1972) Holocene and Pleistocene calcareous crust (caliche) profiles: criteria for sub aerial exposure. J. sedim. Petrol. 42, 817-836. KRUMBEIN, W.E. (1968) Geomicrobiology and geochemistry of the 'Nari-Lime-Crust' (Israel). In: Recent Developments in Carbonate Sedimentology in Central Europe (Ed. by G. Muller and G.M. Friedman), pp. 138-147. Springer-Verlag, Berlin. LAND, L.S., MAcKENZIE, F.T. & GouLD, S.J. (1967) Pleistocene history of Bermuda. Bull. geol. Soc. Am. 78, 993-1006. LAMPLUGH, G.H., (1907) Geology of the Zambezi Basin around Batoka Gorge. Q. J. geol. Soc. Land. 63, 162-216. LATTMAN, L.H. (1973) Calcium carbonate cementation of alluvial fans in southern Nevada. Bull. geol. Soc. Am. 84, 3013-3028. LoGAN, B.W., READ, J.F. & DAVIES, G.R. (1970) History of carbonate sedimentation, Quaternary Epoch, Shark Bay, Western Australia. Am. Ass. Petrol. Geol. Mem. 13, 38-84. MOUNTJOY, A.B., EMBLETON, C. & MORGAN, W.B. (1970) Africa a geographical study. Hutchison, London (3rd impression). MOLLER, G., IRION, G. & FoRSTNER, U. (1972) Formation and diagenesis of inorganic Ca-Mg car bonates in the lacustrine environment. Natunvissenschaften, 59 Jg. 4, 158-164. MuLTER, H.G. & HoFFMEISTER, J.E. (1968) Subaerial laminated crusts of the Florida Keys, Bull. geol. Soc. Am. 79, 183-192. NETTERBERG, F. (1967) Some roadmaking properties of South African calcretes. Reg. Co11f Afr. Soil Mech. Foundation Eng. 4, 77-81. NETTERBERG, F. (1969a) Ages of calcretes in southern Africa. Bull. S. Afr. Archaeol. Soc. 24, 88-92. NETTERBERG, F. (1969b) The interpretation of some basic calcrete types. Bull. S. Afr. Archaeol. Soc. 24, 117-122. RUHE, R.V. (1967) Geomorphic surfaces and surficial deposits in southern New Mexico. State Bur. Mines Miner. Res. N. Mexico, Inst. Min. Techn. Campus Station, Mem. 18, 66 pp. SIESSER, W.G. (1972) Petrology of the Cainozoic coastal limestones of the Cape Province, South Africa. Trans. geol. Soc. S. Afr. 75, 177-185. SIESSER, W.G. (1973) Diagenetically formed ooids and intraclasts in South African calcretes. Sedi mentology 20, 539-551. STALDER, P.J. (1975) Cementation of Pliocene-Quaternary fluviatile clastic deposits in and along the Oman Mountains. Geologie Mijnb. , 54, 148-156. VILJOEN, R.P., VILJOEN, M., GROOTENBOER, J. & LoNGSHAW, T.G. (1975) ERTS Imagery: An apraisal of applications in geology and mineral exploration. Minerals Sci. Engng, 7, 132-168. VISSER, H.N. & SCHOCH , A.E. (1973) The geology and mineral resources of the Saldanha Bay area. Rep. S. Afr. Dept Mines Geol. Surv. Mem. 63, 150 pp. WoLF, K.H. (1963) Syngenetic to epigenetic processes, paleoecology and classification of limestones in particular reference to Devonian algal limestones of central New South Wales. Thesis, University of Sydney (unpublished).
(Manuscript received 9 May 1975; revision received 21 October 1976)
Reprinted from Sedimentology (1978) 25 489-522
Biolithogenesis of Microcodium: elucidation
COLIN F . KLAPPA* Jane Herdman Laboratories of Geology, University of Liverpool, Liverpool, U.K.
AB STRACT
Petrographic studies of Tertiary and Pleistocene caliche from the western Mediterranean show some unusual calcite structures. These structures were desig nated Microcodium elegans GlUck 1912. New data are presented which question earlier interpretations with regard to the origin of this structure. The new discovery of Microcodium in Recent soils extends its stratigraphic range into the Holocene. Retention of fine detail in Recent samples, revealed by light microscopy and SEM, has suggested an origin hitherto unconsidered, calcification of mycorrhizal asso ciations. Ancient and Recent Microcodium fabrics are compared; sufficient preservation of ultrastructure in the Ancient indicates a homologous origin. Environmental, stratigraphic and palaeoecological significance of Microcodium is discussed; correct recognition indicates existence of a palaeoso( and hence is a valuable criterion for recognition of continental conditions, cessation of sedimenta tion, subaerial exposure, and time-equivalent horizons. In particular, Microcodium is a characteristic component of caliche in the western Mediterranean. A review of the literature suggests that its presence may have been overlooked ot misinterpreted in other parts of the world and, thus, may be more widespr�ad than' hitherto suspected. This study, in its embryonic stage of development, illumines the potential importance of biolithogenesis within terrestial carbonates.
INTRODUCTION
The term Microcodium was used by GlUck (1912) to describe unusual calcite crystals from the marine Miocene of southern Germany (Baden). From their shape and arrangement, evoking 'cells' in palisades around small nuclei, GlUck created the genus Microcodium elegans to designate these calcite crystals, which he considered to be organic in origin. He attributed them to siphonaceous algae and, hence, placed Microcodium in the Codiaceae of the Chlorophyta. Later workers (e.g. Jodot, 1935; Moret, 1952a), doubted their organic origin, considering them to be purely physicochemical precipitates. Interest in Microcodium was directed mostly towards an organic versus inorganic debate until Johnston (1953) argued convincingly in favour of an organic origin. * Present address : Memorial University of Newfoundland, Department of Geology, St John's, Newfoundland, Canada AlB 3X7.
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
115
Colin F. Klappa
Curved faces and presence of certain internal structures are criteria commonly put forward to argue against a purely inorganic crystallization. In addition, Bodergat (1974) has demonstrated, using isotopic ratios, that the carbon within the calcite is organic in origin, having been metabolised by Microcodium in the meteoric environ ment. Thus, the analytical results of Bodergat (1974) corroborate the morphological evidence. Further details on ultrastructure, presented in this paper (p.499), establishes the organic nature of Microcodium beyond doubt. Apart from the mere noting of its presence, and recording its facies associations, our understanding of Microcodium was not advanced until Lucas & Montenat (1967) described and interpreted its internal structure. More recently, Bodergat (1974) has given a detailed petrographic and geochemical analysis of Microcodium. This work, together with that of Esteban (1972, 1974), also provides an excellent literature review on this hitherto enigmatic organism. Esteban states that the genus Microcodium refers to elongate, petal-like calcite prisms about 1 mm long and hexagonal in basal section. The prisms are grouped in spherical, elliptical, sheet or bell-like clusters (Esteban, 1974). This description refers to his Microcodium (a) form and appears to embrace types 1 (epis de mai's) and 2 (colonies en laminae) described by Bodergat (1974). Esteban (1972) also defines a new form of Microcodium (b), which is distin guished by its smaller size and quadrangular section of its prisms. Nevertheless, because a complete range of sizes exists between (a) and (b) forms, Esteban (1972, 1974) considered it unnecessary to define another genus. Furthermore, he gives additional ground for not establishing taxonomic refinement (and/or confusion!) as follows . . .'Another reason is that we do not know what Microcodium is.' Microcodium has been discussed in several studies concerned with caliche (Esteban, 1972, 1974; True, 1975a, b). Similarly, my interest in Microcodium was initiated from an examination of caliche samples of Pleistocene age collected in the western Mediter ranean (mainland Spain and the Balearic Islands). In addition, the new discovery of Microcodium within soil samples collected from the province of Murcia (south eastern Spain) extends its stratigraphic range into the Holocene. True (1975b, p. 51) states'... Or les microcodiums, qui jouent un role primordial dans ces systemes, n'ont jamais ete signales encore dans les encrofitements actifs actuels.' The discovery of Microcodium in the Recent necessitates a revision of this quotation. Because Microcodium forms a significant component of many of these samples it was considered unscientific to pass off Microcodium as a 'whim of nature' (an attitude mentioned, although not necessarily followed, by Cuvillier & Sacal, 1961). As a consequence, a detailed examination of this material has led to new findings on Microcodium which are presented herein. Purpose
The intention of this paper is: (1) to document, for the first time, the presence of Microcodium in samples collected from the Pleistocene of Ibiza (Balearic Islands); (2) to demonstrate the existence of Microcodium in the Recent; (3) to provide further details on Hie inicrofabrics of Microcodium ; (4) to demonstrate that Ancient and Recent Microcodium structures are the product of the same phenomenon. Sufficient details are retained in Ancient Microcodium sampled from the Eocene of northeast Spain and southern France to allow such direct comparisons; (5) to point out the environmental significance of Microcodium ; (6) to stimulate a search for Microcodium
116
Biolithogenesis of Microcodium
whose presence may have been overlooked or misinterpreted; and (7) to suggest an hitherto unconsidered origin for Microcodium. Previous investigations
Well documented specimens of Microcodium have been recorded by French geologists in lacustrine, paludal (swamp), alluvial, and fluvial deposits, at palaeokarstic horizons, and as 'contaminants' in marine facies. Microcodium occurs dominantly in Tertiary rocks, being particularly conspicuous at the Cretaceous-Eocene boundary. The stratigraphic levels at which Microcodium occurs share several common charac teristics: a carbonate-rich substratum, indications of subaerial exposure and fre quently the presence of an overlying palaeosol. Microcodium appears to be associated with continental conditions during periods of negligible sedimentation, i.e. with sedi mentation rates being insufficient to preclude pedogenetic processes. The Garumnian and Vitrollian of French stratigraphy are examples of such facies. Most recorded in situ colonies of Microcodium are from rocks of Eocene age. In younger rocks they may be detrital but evidence for in situ growth within Pleistocene rocks has been documented (Bodergat, 1974; Ward, 1975). Visual evidence that Microcodium has grown in situ is indicated by grain truncations within the host rock. This relationship with the enclosing substrate has led to the suggestion that Micro codium actively dissolves carbonate in search of trapped organic matter (Lucas & Montenat, 1967; Esteban, 1972, 1974; Bodergat, 1974; Bodergat, Triat & True, 1975; True, 1975a, 1975b). Insoluble residues (e.g. layer lattice silicates, quartz grains) are pushed aside and the dissolved carbonate is reprecipitated to form the calcite prisms of Microcodium. This scheme of development, however, does not explain fully the observed micro fabrics and raises questions of actual processes. For instance, Bodergat (1974) has pointed out that although there may be many micro-organisms capable of dissolving carbonate (e.g. certain algae, fungi, lichens, actinomycetes, bacteria), the same organ ism cannot also bring about a reprecipitation. This line of reasoning led Bodergat (1974) to the hypothesis that two organisms (at least) have played a role in producing the structure of Microcodium. She suggested that actinomycetes were responsible for the destructive component of her 'biocorrosion-biosynthesis' system but the mechan isms of biosynthesis and the type of organism responsible for such a process are not discussed. Although the study of Microcodium has been largely confined to France, its presence has not been overlooked completely in other parts of the world. For example, Microcodium has been reported from the Miocene in Germany (Gllick, 1912), the Eocene and Pleistocene in Spain (Esteban, 1972, 1974), the Permian in Russia (Maslov, 1956), the Eocene in Switzerland (Kamptner, 1960), the Devonian and Carboniferous in North America (Wood & Basson, 1972), the Eocene of Turkey (Richard, 1967), the Pleistocene in Mexico (Ward, 1975) and in the Miocene from the Pacific Islands of Bikini and Saipan (Johnson, 1953, 1957). The single reference found (from an extensive, though not exhaustive literature survey) of Microcodium occurring in the Recent is by Marie (1957).According to this worker, Microcodium is present in littoral deposits from the Bay of Along, Indochina, although its presence is explained as being due to reworking rather than in situ growth. Table 1 summarizes pertinent citations on Microcodium, giving the geographical
1 17
Table 1. Summary of published citations and undocumented reports on Microcodium, giving geographical location, geological age of Microcodium and/or age of
substrate(s), environment and interpretation with respect of origin Author(s) and year
...... ......
00
Sacco (I 886)* Capeder (1904)* Gli.ick (19 1 2) Edwards (1932) Jodot (1935) Rech-Frollo (1948) Moret (1952a, b) Johnson (1953, 1957, 1961a, b) Rutte ( 1954) Faure-Muret & Fallot (1954) Cuvillier (1955) Gubler (1955) Maslov (1956) Demangeon (1956) Cuvillier & Sacal (1961)
Geological age
Geographical location
Interpretation
Envi; onment (lithofacies)
Lithothamnian alga Coral Codiacean alga
Marine Marine Marine Freshwater limestones Marine Lacustrine Marine Marine Marine Marine Marine; continental
Inorganic origin Complex diagenetic (inorganic) I norganic origin Alga Green alga Alga Alga
Conglomerates Epicontinental
Blue-green alga (Demorarpales) Vegetation Alga (?)
Maritime Alps, France
Eocene Miocene L. Eocene Miocene Danian U. Cretaceous Miocene Miocene Maastrichtian-Lutetian Cretaceous-Lutetian Eocene Permian; Palaeogene Montain Maastrichtian-Danian
Fran9ois & Sigal (1957) Marie (1957) Boulanger & Cros (1957) Allard et a!. (I 959) Kamptner (1960, 1962) Guillaume (1961) Paquet (1961) Moret & Flandrin (1961) Durand (1962)
Baden, Germany Trieste, Italy Baden, Germany N. Pyrenees, France French Alps & Pyrenees Pacific Islands Baden, Germany Maritime Alps, France Mouries, France Maritime Alps Russia Languedoc Meaux (Seine-et-Marne, France) Landes, France France; Indochina Limoux (Aude), France Bresse, France Jura, Switzerland Doubs, France Ardeche, France Alps, France Provence, France
Landenian Cretaceous; Tertiary Montain Tertiary Eocene Turonian-Coniacian Sannoisian U. Cretaceous U. Cretaceous; Eocene
Marine Lacustrine Marine; brackish; lacustrine Fluvio-lacustrine
Gottis (1963) Bourrouilh & Magne (1963) Lapparent (1966)
Mouries, France Minorca, Balearics Alps, France
Neocomian(s); Eocene Pliocene(s); Pleistocene Eocene
Karst Marine Continental
Lacustrine; Fluviatile Lacustrine; Continental Limestones; sandstones Lacustrine
Red alga; liverworts Organic
Organic (recrystallized) Analogous to stromatolites Dasycladacean algae CMizzia; Macroporella)
Obligatory heterotrophs
Table 1 continued o n page 493
� � �
(S � � �
Table 1 continued
Author(s) and year
...... ...... \0
Lucas & Montenat ( 1967) Richard ( 1967) Misik ( 1968) Bodelle & Campredon ( 1968) Freytet ( 1969) Roux ( 1970) Nury, Rey & Roux ( 1970) Freytet ( 1971a, 1971b, 1973) Plaziat ( 1971) Masse, Triat & True ( 1972) Esteban ( 1972, 1974) Wood & Basson ( 1972)t Bodergat ( 1974)
Geographical location
Drome, France Gocek, Turkey Brezova, Yugoslavia Alps, France and Italy Languedoc, France Castellane, Basal Alps Rouet, France Languedoc, France Languedoc, France S.E. France N.E. Spain Missouri, U.S.A. France; Spain, incl. Minorca (Balearics) Digne-Valensole, France Gigot ( 1974) Paris, Basin, France Freytet ( 1975) Yucatan, Mexico Ward ( 1975) Bodergat, Triat & True ( 1975) S.E. France S.E. France True ( 1975a, 1975b) Mallorca (Balearics) Calvet et al. ( 1975) Montenat & Echallier ( 1977) S.E. Spain Spain, incl. Ibiza (Balearics) Klappa (this study)
Geological age Eocene Cretaceous-Tertiary Maastrichtian Eocene U. Jurassic(s); Maastrichtian Eocene-Oligocene Oligocene-Miocene Eocene Thanetian-Sparnacian Eocene Eocene; Pleistocene Devonian-Carboniferous(s) Mesozoic-Pleistocene M iocene Eocene Pleistocene Paleogene Pleistocene Pleistocene Eocene-Recent
Environment (lithofacies)
Interpretation
Continental
Filamentous bacteria
Marine Brackish-laguna] · Karst
Alga to (3'
:::-:
So c
Fluvio-lacustrine, palustrine Fluviatile conglomerates Caliche Non-carbonate shales Various Conglomerates Fluvio-laucustrine Aeolianites Continental Calcareous crusts; palaeosols Aeolianites Caliche Caliche
� ;:s "' "'
Bacteria; blue-green alga Fungus Micro-organisms, actinomyctees ( ?)
Roots ( ?) Micro-organisms Roots ( ?) Bacteria ( ?) Mycorrhiza (root+ fungus)
* Cited in Sturani ( 1963); t probably not Microcodium (Klappa, personal observations; Dr J. M. Wood, written communication).
1:;•
� �
;:;·
.... c (") c >:>...
$2' ;::!
Colin F. Klappa
location, geological age and stratigraphic relationships, together with interpretations (if attempted) on origin. Problems of terminology
Early attempts to describe Microcodium directed me to problems of terminology. The term 'Microcodium' has been used in various ways. Some workers refer to Micro codium as the actual (hypothetical) organism, while others have used it to describe the observed calcite structure. Two important considerations may have contributed to this semantic problem: firstly, whether Microcodium is organic or inorganic; and secondly, whether the calcite is part of the skeleton (as in calcareous algae) or a later precipitate within a vacated or original chamber. Terms frequently used such as 'thallus', 'cell', 'vacuole', 'filament' have genetic overtones. Because of potential confusion in terminology, the morphological term 'Microcodium grain' is used in this paper to describe individual prisms or units. In instances where a number of Micro codium grains form an organized arrangement the term 'aggregate' is employed.
GEOGRAPHIC LOCATION AND GEOLOGICAL SETTING Ibiza-PI eistocene
The island of Ibiza, situated between the latitudes of 39° 6' N and 38° 5' N and longitudes 4° 45' E and 5° 1' E, is an emergent part of the Balearic Platform in the western Mediterranean (Fig. I b and b). Carbonates (limestones, dolomites, marls) are the dominant lithologies, ranging in age from Muschelkalk (Trias) to Recent. Except for the more elevated areas, the solid geology is covered by Quaternary sedi ments and calcareous crusts (including caliche sense stricto, i.e. having a pedogenetic origin). Microcodium has been recorded in calichified bedrock (dominantly carbonates but also in profiles with igneous substrates) of Tertiary and Mesozoic age, and in aeolianites and colluvial silts of Pleistocene age (Klappa, unpublished data). Southeastern Spain - occurrence of Recent Microcodium
Recent Microcodium was discovered in southeastern Spain, 8 km south of Cieza (30 km NNW of Murcia). The site occurs along an unmetalled forest track (38° 11' N, 2° 18' W) on the east side of the road Mula-Cieza (Fig. 1c). The bedrock of upland areas is composed of Upper Triassic dolomites, whereas the lower slopes and valleys consist of Lower Eocene lime muds and unconsolidated or poorly consolidated marls. Conglomeratic slope deposits, which overlie the solid geology, also lack consolidation apart from their uppermost layers. At or near the surface, subaerial vadose pedogenetic and diagenetic processes have led to the form ation of caliche profiles in various stages of development, from nodular to thin laminated crusts. Present-day pedogenetic processes, however, are causing modi fication and/or destruction of these indurated layers, mainly by the mechanical penetration of root systems. Samples containing Microcodium were located 25 em below the present-day surface in a rubbly calcareous soil. Anastomosing channels, 1·0 mm wide, were noticed on the surfaces of many pebble-sized clasts (Fig. 3a). The arrangement of the 120
Biolithogenesis of Microcodium
c
lbiza
... Eocene
Fig.
• Pleistocene
+ Recent
1. Recorded localities of caliche profiles containing Microcodium sampled in this study. Age of
is denoted by symbol. (a) General map of western Mediterranean ; (b) Ibiza, Balearics, (c) location of recorded Recent Microcodium, southeast Spain.
Microcodium
channel networks suggests that roots have caused peripheral dissolution of these pebbles. Evenari, Shanan & Tadmor (1971) have noted that the surfaces of pebble sized grains, in otherwise fine-grained soils, tend to be sites of greater moisture content and, therefore, provide readily available water for the indigenous flora. While many channels were devoid of any material, some contained rows of white to translucent, ellipsoidal grains (Fig. 3b). Their morphology and ordered arrangement initiated the idea that they may be the same as Microcodium grains recorded from the Pleistocene of lbiza. Subsequent laboratory examination substantiated this preliminary field observation (Figs 2a-e, 3b-d, 4a and b, 6a-d). The possibility that these Microcodium grains are detrital, having been reworked from older geological successions, can be discounted for several reasons. Firstly, organized aggregates occur within the present-day soil profile. Secondly, the soil matrix surrounding and supporting undisturbed aggregates is friable; the aggregates
121
Fig. 2. SEM photomicrographs of Recent Microcodium aggregates and grains. Loose sediment mounts, Cieza, S.E. Spain. (a) Ellipsoidal Microcodium aggregate with grains showing interference· growth boundaries; (b) enlargement of (a), showing surface detail. Note pore pattern and presence of subsurface channels (arrow); (c) elongate Microcodium aggregate with grains in concentric layers; (d) enlargement of (c), showing grain surface concavities and naturally etched Microcodium grains; (e) single Microcodium grain with concave faces. Note subsurface tubular networks with tube dia meters of 1·0 �tm or less (arrow); (f) detail of (e), showing protuberances on surface of naturally etched Microcodium grain. Scale in ).lm.
122
Fig. 3. SEM photomicrographs of Recent Microcodium gra ins. Untreated surface samples, Cieza, S.E. Spain. (a) Surface of dolomite lithoclast fragment with pitted channels due to the corroding action of Microcodium; (b) linear aggregate of Microcodium grains within surface channel of dolomite lithoclast; (c) Enlargement of (b), showing intragranular protuberances (arrow) within a completely dissolved (natural) Microcodium grain. (d) detail of (c). Lower surface is partially etched wall. Aerial fungal hypha is probably post-cavity formation . Residual intragranular structures are arrowed; (e) surface of 'bored' wall showing 1·0 J.!m diameter cylindrical pores surrounded by fibrillar mat (arrow). The latter is interpreted as disaggregated plant cell wall material; (f) naturally etched Micro· codium grain showing anastomosing subsurface (originally) tubes, 1· 0 J.!m diameter. Scale in J.!m . 123
Colin F.
Klappa
Fig.
4. Loose sediment mounts of Recent Microcodium grains, Cieza, S.E. Spain. (a) Microcodium grains surrounded and penetrated by fungal hyphae. Grains have been mounted i n Polyric immersion oil and viewed under the petrographic microscope, P . P.L.; (b) 'floating' elongate Microcodium grains supported by fungal hyphae. Note rhizomorph on right-hand side (arrow) . SEM natural (untreated) surface; (c) detail of Microcodium grain with pitted surface. Note that several pits are surrounded by raised borders (arrow). SEM photomicrograph . Scale in J.lm.
could not maintain form if mechanically churned or transported. Thirdly, aggregates unaffected by chemical corrosion, are extremely delicate; even the slightest pressure of a steel needle is sufficient to cause disaggregation. Fourthly, moribund fungal mycelia (masses of hyphae) surround and penetrate Microcodium grains (Fig. 4a and b). Fifthly, partially decayed vascular plant material surrounds Microcodium aggre gates. Similar plant debris, and also fungal hyphae, are present throughout the soil matrix. Finally, some of the fine details of Microcodium grain ultrastructure, as revealed by SEM (Figs 2e, 3e and f ), are considered unlikely to be preserved completely in the Ancient.
124
Biolithogenesis of Microcodium
Fig. 5. SEM photomicrographs of polished and etched rock samples from a calcified terra-rossa soi l .
Pleistocene, Ibiza. (a) Isodiametric Microcodium grains with- convex a n d concave boundaries. Indi vidual grains have been outlined in ink; (b) detail of (a), showing linear pore pattern (accentuated by etching in dilute HCl acid) indicated by arrow; (c) single Microcodium grain showing cracks and tubular pores; (d) detail of (c), indicating presence of filamentous structures (tips just visible) within tubular pores.- Scale in 1-1m.
LABORATORY
ANALY SI S
OF
MICROCODIUM
Sample preparation for petrographic analysis
The following procedures were adopted for samples containing Microcodium from lbiza, mainland Spain and southern France. (A) Standard petrographic thin sections. (i) Unstained; (ii) etched (1·5% hydro chloric acid for I 0 sec) and stained: (a) combined Alizarin red S and potassium ferricyanide (Dickson, 1966), (b) Feigl's solution (Feigl, 1943), (c) Clayton yellow (Winland, 1971); (iii) decalcified (10% hydrochloric acid until all carbonate was removed); and (iv) as (iii) plus staining with Gentian Violet dissolved in 90% methanol (Gurr, 1965). 125
Colin F. Klappa
Fig. 6. Petrographic details of internal structures of Recent Microcodium grains from Cieza, south eastern Spain. Grains have been mounted in Polyric oil and viewed under the petrographic micro scope. (a) Elongate Microcodium grain displaying a radial-fibrous calcite fabric. Note discrete peripheral nucleus for part of radial-fibrous calcite (arrow); (b) Microcodium grain consisting of bundles of radial-fibrous calcite. (c) central tubular structure (10 J.!m diameter) and smaller tubular pores . ( 1·0 J.!m diameter) within single Microcodium grain; (d) enlargement of (c). Scale in J.lm.
(B) Loose sediment mounts. (i) Binocular examination; and (ii) temporary slide mounts using polyric immersion oil to coat Microcodium grains. (C) Scanning electron microscopy. (i) Freshly fractured rock surfaces; (ii) polished and etched (1·5/o hydrochloric acid for 30 sec) chips of Microcodium ; and (iii) Micro codium grains, hand-picked under a binocular microscope from friable, calcareous sediments. Prepared samples for SEM were mounted on a 1·0 em diameter specimen stub using Durofix adhesive mixed with acetone (50:50). Acetone was used to ensure an even spread of adhesive to minimize charging effects. Prior to coating, the specimens were oriented and features recorded under a 40 x magnification Nikon zoom binocular microscope. Using a MGN SG-2 12" coating unit, the mounted samples were coated under vacuum with 60/o gold-palladium. Oriented specimens were viewed employing a Cambridge Stereoscan, Mark IIA, operated at an accelerating voltage of 20 kV, with a beam angle of 45° and a working distance of 9-11 mm. Petrography
Abundance. Point-count analysis of fifty thin sections contammg Microcodium of Pleistocene age indicated that this component ranges from less than 0·5/o to 43/o of the total rock by volume, with an average of 17/o. 126
Biolithogenesis of Microcodium
Size.The sizes of individual Microcodium grains were measured for loose sediment mounts and thin sections which contained them. Apparent dimensions (measurement on grains with various orientations in thin section) showed a range from 100 J..lm to 200 J..lm, averaging 120 J..lm, for long axes, and from 30 J..lm to 100 J..lm, averaging 70 J..l m for grain widths (diameters of isodiametric grains and/or transverse sections). The maximum observed length for loose grain mounts was 375 J..lm, considerably less than the typical 1·0 mm prisms of Eocene Microcodium described by Esteban (1974). Shape. The shape of individual Microcodium grains varies from well-defined prisms with length:width ratios of 2:1 to 3:1 (Fig. 7a), to vague ellipsoidal or subspherical outlines. Transverse sections show hexagonal, quadrangular or subspherical outlines (Fig. 8c). Curved faces, both convex and concave,lI tend to be commoner than straight (Fig. 2a). Many Microcodium grains display re-entrants or embayments (Fig. 7b) giving shapes that cannot be attributed simply to mechanical abrasion or fracture during transport.
Fig. 7. Photomicrographs of petrographic thin sections from the Pleistocene of Ibiza. P.P.L. (a) Microcodium aggregate composed of apparently overlapping prismatic grains (due to interference growth) with linear pore patterns (dark areas); (b) single detrital Microcodium grain with irregular ('wiggly') tubular pores radiating from periphery. Note : (i) re-entrant at top; and (ii) partial micri tization at base. Grain to lower left is a partially micritized coralline algal fragment; (c) part of Microcodium aggregate stained with Alizarin red S and potassium ferricyanide. Radial-fibrous fabric has taken up stain (red, indicating non-ferroan calcite), whereas grains or parts of grains with u niform extinction have remained unstained; (d) detail of (c), showing stained radial-fibrous calcite (lower half) and u nstained monocrystalline calcite (upper half). Scale in 11m .
127
Colin F.
Klappa
Arrangement. The arrangement of individual grains with each other and the enclosing substrate varies from single, isolated crystals, to organized aggregates The latter may occur as rosettes ('epis de ma:is' of Franr;:ois & Sigal, 1957) (Fig. lOa), with prisms radiating from a central nucleus (Fig. 9a), or as groups of prismatic or isodiametric grains lacking any discernible radial or concentric pattern (Fig. 8a). Another type of arrangement, although rare in Pleistocene samples from Ibiza but common in the Eocene of northeastern Spain and southern France, is a laminar or sheet-like layering of prismatic grains (long axes normal to layers). In general, how ever, because of reworking isolated detrital grains of Microcodium or fragmented Microcodium aggregates are far more numerous than in situ growth forms. Optical properties. Optical properties suggest a calcite mineralogy for the Micro codium grains. In plane-polarized light, grains are colourless or pale brown, but contain dark radial-fibrous inclusions. The filamentous inclusions appear white using reflected light. SEM examination of Microcodium grains reveals that non carbonate filaments may be present (Figs 5d and lOb) but in many cases, selective leaching has produced pits which give an optical effect analogous to the porous'shell residue micrite' described by Alexandersson (1972). Some Microcodium grains have a rhombohedral cleavage but many show only an irregular pattern of cracks (Figs 5a, c, 7c and d). Extinction patterns vary between grains and within the same grain. Uniform (relatively uncommon) to aggregate (suggesting a number of sub-crystals) extinction may be observed in one part, while a sweeping extinction may be present in the remainder of an individual grain. Some grains show a complete radial-fibrous structure with fibres radiating from a point on the perimeter of an individual grain and not from the centre as in typical spherulites (Figs 6a, 7c and d). Adjacent grains show also a radial-fibrous structure radiating from the same point (Fig. 8c). As a result, a pseudo-uniaxial cross is formed when the juxtaposed grains are seen between crossed nicols. Because of this extinction pattern, combined with overlapping sub-crystals, optical interference figures were not readily obtained. In cases where interference figures could be recognized a uniaxial-negative figure corroborated the evidence in favour of calcite. Ultrastructure. In several Microcodium aggregates the calcite was removed by etching in dilute hydrochloric acid (1·5/';;). Total dissolution revealed the presence of a network of branching filaments 1·0-2·0 !liD in diameter. The filaments may have been originally transparent (common in fungi, for example) but the presence of iron and/or natural organic staining rendered them visible in reflected and transmitted light. SEM examination of Microcodium from the Pleistocene of Ibiza shows several interesting features that have not, to my knowledge, been documented elsewhere. Linear patterns of tubular cavities 0·5-2·0 !liD cross-section can be seen in grains (Fig. 5b). These may radiate from a larger tubular cavity 5·0 !liD in diameter (Fig. 5a) or from a point on the grain perimeter. Similar tubular patterns can be seen within Recent Microcodium grains mounted in immersion oil and viewed under the petro graphic microscope (Fig. 6d). Some tubes have a prismatic cross-section which per haps suggest that these cavities are moulds of aragonite needles but two points of observation do not favour such an interpretation. Firstly, the tubes are not necessarily straight but show curved or wiggly shapes (Figs 2e, 6d and 7b), whereas aragonite needles have planar crystal faces. Secondly, gentle etching of the calcite prior to SEM coating, revealed the presence of residual structures less than 1·0 !liD diameter within
128
Biolithogenesis of Microcodium
Fig. 8. Photomicrographs of petrographic thin sections from Pleistocene deposits, Ibiza, P . P . L .
(a) Fragmented, detrital aggregate consisting o f isodiametric Microcodium grains 'floating' in a calcisiltite matrix; (b) in situ Microcodium grains with dark inclusions or pores (cf. 'shell-residue micrite' of Alexandersson, 1972); (c) detail of (b), showing dark inclusions radiating from discrete nuclei (arrows). Scale in J.Lm.
129
Fig. 9. SEM photomicrographs of fractured samples from the Eocene of northeastern Spain and
southern France. (a) Transverse section of Microcodium cylinder composed of radiating petal-like prisms; (b) oblique section of cylinder with constrictions or disc-like structures. Cylinder axis lies NE-SW in photomicrograph; (c) contact between Microcodium grains (right) and calcisiltite matrix (left). Concentric bands within Microcodium grains may be trapped insoluble residues at successive growth fronts; (d) detail of (c) at junction. Note micro-honeycombed structure (arrow) at periphery of Microcodium grain; (e) walls between and within Microcodium grains consisting of porous clay sized aggregates; (f) enlargement of (a) . Scale in J.!m.
130
Biolithogenesis of Microcodium
Fig. 10. SEM photomicrographs of P leistocene Microcodium, Ibiza. (a) Partially collapsed rod of Microcodium, resembling a 'corn on the cob'. Isodiametric grains which constitute the rod similarly show signs of deflation; (b) detail of surface features in a, revealing presence of filamentous structures within, and traversing across, Microcodium grains. Scale in J.tm.
131
Colin F.
Klappa
some of the tubes (Figs 5c and d). These structures seem to have been unaffected by acid etching, a point which further argues against them being aragonite.Because of the fineness of scale it has not been possible to isolate these filaments for further tests. Viewed with SEM, the tubular pores can be equated in petrographic thin sections with the so-called dark, filamentous inclusions that show a radial pattern (Figs 7b and 8c). It is significant that these filaments radiate from a point on the periphery of the Microcodium grain and not from the centre. Furthermore, the calcite within the Micocodium grain also shows a radial-fibrous fabric (Figs 6a, b, 7c and d). The crystallites that make up this radial-fibrous fabric similarly radiate from the grain perimeter, unlike spherulites which radiate from a central point. This indicates growth from a 'wall' into a_cavity. Moreover, the radial arrangement for the vacated or tubular pores, likewise suggests penetration from the walls of Microcodium grains. The question now arises as to what constitutes this 'wall' and its relationship, if any, to the tubular pores. In thin section, Microcodium grains from the Eocene of northern Spain show walls between, and cross-walls within, petal-shaped grains. The walls consist of dusty, pale to dark brown microcrystalline calcite (staining and etching in dilute hydrochloric acid indicates that at least part of wall has a calcite mineralogy). SEM observations of fractured surfaces show that these walls are made up of equant, subhedral to anhedral clay-sized grains with a high intergranular porosity (Figs 9e and f ). The fine grain size and high porosity is probably responsible for the dusty appearance of the walls. No organized fabric was recognized within these walls; the micrite has an apparently random fabric. SEM studies of Recent Microcodium grains from southern Spain provide further surface and internal features that have relevance to the above outlined details, and perhaps more fundamentally, provide evidence for overall interpretation (p. 5 10). The surfaces of Recent Microcodium grains show a granular appearance with bordered pits and cylindrical pores 0·4-1·0 J.lm in diameter (Figs 2b, f and 4c). The pores appear to be connected to a subsurface anastomosing network of tubes (Figs 2b and c). Indeed, some Microcodium grains with the surface layer removed by fracturing or natural flaking, reveal the presence of a fine network of tightly bound, branching or coiled tubes 0·4-0·8 J.lm in diameter (Fig. 3f). This network does not appear to be continuous throughout the grain but restricted to a thin superficial layer less than 5·0 J.lm thick. This layer corresponds to the 'wall' of Ancient Microcodium grains. The ultrastructure of the remainder of the grain, both in the Recent and Ancient, consists of foliate calcite (Figs 2c, d, 9e and f). Bodergat ( 1974) stated that the 'platy' calcite was oriented perpendicular to the long axes of elongate Microcodium grains. Although this orientation was recorded in many grains examined in this study, calcite plates were observed also with a sub-parallel alignment with respect to grain long axes. The complex ultrastructure of Microcodium grains, as revealed by SEM examin ation, perhaps helps to explain the reasons for anomalous optical properties and staining patterns of the same grains when viewed under a petrographic microscope. Mineralogy
In order to substantiate the petrographic data several microchemical tests were carried out to evaluate the mineralogy of Microcodium grains. Staining thin sections with a filtered solution containing Alizarin red S and potassium ferricyanide combined 132
Biolithogenesis of Microcodium
(Dickson, 1966), revealed certain anomalous features. Staining colours were weak or absent in many Microcodium grains, or present within irregular cracks (possibly be cause of the difficulty of thorough washing within micro-pores). In grains with com posite extinction patterns a red stain (non-ferroan calcite or aragonite) was commonly taken up only by the fibrous part of the grain (Figs 7c and d) whereas the clear spar, with uniform extinction, remained unstained. According to Dickson (1966) the intensity of the combined stain depends on the amount of iron present, the orientation of the c axis with respect to the plane of the thin section, and the concentration of the acid in solution. Additional factors considered in the course of this work that may have affected the staining pattern and colour intensity include: the presence of organic matter, particularly mucilaginous films; the presence of non-carbonate clay-sized grains; the micro-porosity of individual crystals (p. 502 and Figs 2b and Sa-d); and the presence of other foreign ions (in addition to iron). Because of the atypical staining patterns as outlined above, some Microcodium grains were suspected of having a mineralogical composition other than pure calcite. Further microchemical tests were carried out for the presence of aragonite (Feigl's solution, Feigl, 1943), high magnesium calcite (Clayton yellow, Winland, 1971) and dolomite (Alizarine cyanide green, Davies & Till, 1968). The presence of these minerals was not detected using these methods. Because Microcodium is considered to be organic in origin, several microbiological tests were carried out. It is well known that certain plants secrete crystals (cystoliths) within their cells (Cutter, 1969). Most contain calcium; calcium oxalate is the commonest organic compound found within plant tissues, although calcium carbonate also occurs.Such crystalline deposits are generally considered to constitute deposits of waste products. Examples of oxalate crystals-whewellite (CaC204• H20), and weddellite (CaC204• 2H20)-are organic salts of inorganic cations (calcium) and organic acids. Following the procedure of Gurr ( 1965), a test was carried out to detect calcium oxalate. Observation of the chemical reactions, both under the binocular and petro graphic light microscopes, suggested the presence of calcium oxalate within part of the Microcodium grains tested, but not throughout. The reliability of the method is unknown, however, and further geochemical tests were undertaken. Hand-picked Microcodium grains were ground to a fine powder and prepared for X-ray diffraction analysis. A Guinier camera was employed, using CuKa radiation, for mineral identification. The results showed that only calcite was present in these samples. If other compounds were present, as earlier tests seemed to indicate, then their trace amounts were masked by calcite.
CO M PARATIVE A P PROACH- MICROCODIU M AND TERRESTIAL VEGETATION Preliminary observations
This section examines cumulative data gained from field and laboratory studies on Eocene to Recent Microcodium, giving due-and in my opinion, long overdue consideration to present-day soil systems. Until now, Microcodium has disguised itself so well that it has not been recognized
133
Colin
F.
Klappa
in the Recent. After studying Pleistocene samples from Ibiza containing abundant Microcodium grains, this discrepancy seemed somewhat puzzling. Did Microcodium really become extinct at the end of Pleistocene, or has the environment of its formation not been studied by geologists? The presence of Microcodium near or within rhizocretions has been recorded by Calvet et al. (1975) in Pleistocene aeolianites from Mallorca. Dr W.C.Ward (written communication) suggests that Microcodium from the Pleistocene of Yucatan may, in some way, be related to roots. Moreover, the conviction that roots are of paramount importance in determining many textures in caliche, and are related to the formation of Microcodium, has been amplified in valuable discussions with Dr M. Esteban, of the University of Barcelona. From my field observations, cumulative evidence indicated that roots play an important role in determining macromorphological features of caliche profiles (author's unpublished data). This realization led to a search for rhizospheres (Gk. rhiza root) on a microscopical scale. =
Biogenic aspects
In earlier sections (Petrography, Mineralogy), several features were described but left unexplained, namely: the occurrence of walls, radial-fibrous calcite and filaments radiating from discrete points on the grain perimeter; preservation of grain mor phology; and possible presence of calcium oxalate. In order to account for these observed features a new model is proposed in this paper (p.510) which contends that Microcodium is the result of calcification of mycorrhizae. Before the morphological observations can be interpreted in the light of this proposal, it is deemed necessary to clarify terminology and introduce basic concepts of plant anatomy and physiology. Great variability exists in the shape and structure of roots (Fahn, 1974). This is related mainly to root function, i.e. whether they are storage roots, succulent roots, aerial roots, pneumatophores, prop roots, or whether they contain symbiotic fungi (to form mycorrhizae). Nevertheless, the general anatomy of young roots shows several common characteristics which can be conveniently divided into the following zones: (I) the root cap (situated at the tip of the root); (2) the epidermis (outermost layer of cells), including root hairs which are projections or tubular outgrowths of the epi dermal cells; (3) the root cortex (parenchyma tissue surrounding the vascular cylinder and bounded on the outskirts by the epidermis); and (4) the vascular or central cylinder (consisting of xylem and phloem). It is well known that roots provide habitats for many soil micro-organisms (Burges, 1958). Fungi may form a union with roots to make composite structures known as mycorrhizae. It should be made clear, however, that not all fungal attacks on roots are necessarily mycorrhizal; many fungi are parasitic or saprophytic, whereas a mycor rhiza is defined as 'a symbiotic association between a non-pathogenic (or weakly pathogenic) fungus and living, primary cortical cells of a root' (Marks & Kozlowski, 1973). Mycorrhizae can be divided into two main categories (ecto- and endo-), although a transitional stage has also been recognized (ectendomycorrhizae). Marks & Kozlowski (1973) define these as follows: Ectomycorrhiza fungus is confined exclusively to the intercellular spaces of cortical cells of the host root; Endomycorrhiza . fungus is confined exclusively to the intracellular spaces of cortical cells of the .
134
.
.
Biolithogenesis of Microcodium
host root; and Ectendomycorrhiza . . fungus occupies the intercellular spaces of the root and penetrates some (or all) of the adjacent cortical cells. As a result of fungal infection, the plant cell may be structurally modified in a characteristic way. A particular type of mycorrhiza, known as vesicular-arbuscular, usually applies to an endomycorrhiza where the fungal hyphae inside the cortical cells are either coiled or divided into haustoria! branches (Figs 3c and d). Haustoria (Latin haustor drinker) are generally regarded as specialized absorbing organs. They may be knob-like in shape, elongate, or branched like a miniature root system (Alexopoulous, 1 962). Although the role fungi play in mycorrhizal associations is unclear, the general consensus of opinion is that mycorrhizal infections assist in the absorption of mineral salts, particularly in soils where the levels of available minerals are low (Burges, 1958). This point may be of particular relevance to caliche profiles which are poor in many essential mineral salts for normal plant growth. The actual mechanisms for uptake of nutrients into mycorrhizae are imperfectly understood as the following statement by Kelley (1950, p. 12) indicates '. . . whether materials get from the soil into the plant by mechanical means or black magic is left to the imagination of the reader'. Recent work on mechanisms subsequent to Kelley's cynicism is reviewed by Bowen (1973). His discussion may have some relevance to the problem of calcification of Microcodium, but much remains to be learned before actual processes are understood. For the time being, the above quotation from the work of Kelley (1950) can also be applied to this problem, even though several possible mechanisms for calcification have been outlined (p. 512). The possibility that Microcodium is related to a fungus-root association is con sistent with previously noted associations recorded on a macroscale. For example, Ancient Microcodium has been found within palaeosols, particularly calcareous soils (Bodergat, 1974), at unconformities indicating subaerial exposure of marine succes sions (Lapparent, 1966; Esteban, 1972), and at palaeokarstic horizons (Freytet, 1969). With respect to the latter, the presence of Microcodium within deep fractures and solution hollows has raised doubt as regards an algal origin for Microcodium ; algae generally require light for their vital life processes. Because of this factor, Gottis (1963) suggested that Microcodium was an obligatory heterotroph. Lucas & Montenat (1967) overcame this problem by considering Microcodium to be the result of the activities of colonial bacteria. Wood & Basson (1972) state that the occurrence of their specimens and presence of chitin suggest that the organism could be a fungus. They note (p. 212) that '. . . if this organism is a fungus, the question as to how M. elegans was able to live in the absence of light . . . would be answered. ' Similarly, mycorrhizae (fungus-root symbiosis) occur generally in a subterranean environment and, thus, do not require direct light. Several pertinent general comments regarding plant roots may help to convey the reasons for emphasizing their importance with respect to the occurrence of Micro codium. Roots are responsible for acid reactions that may stimulate rock decom position. Roots add C02 to soil-air and soil-water, thus increasing the production of carbonic acid which lowers the pH of circulating waters. This may lead to dissolution of carbonate minerals. Roots provide channels which allow easier circulation of water and air. Roots penetrate joints and cracks, causing mechanical disintegration. Roots are surrounded by a concentration of micro-organisms within the rhizosphere, or may provide habitats for micro-organisms on the root surface (rhizoplane), or actually .
=
135
Colin
F.
Klappa
within root tissues (intracellular infection). Such micro-organisms contribute to changes in the chemical micro-environment by respiration, secretion of acids, organic decomposition and other complex interacting processes. Because of the biogeochemical complexity of soil formation and modification, it is not always possible to substantiate such generalities as outlined above. For example, the assumption that roots increase acidity in soil around them by excretion of carbon dioxide, and possibly H+ ions, may not be valid. It has been suggested that roots take up on average more anions than cations and would therefore tend to pass out HC03ions, rather than H + ions to preserve electrical neutrality (Gray & Williams, 1971). This would increase the pH of the soil around the roots and possibly counteract the effect of carbon dioxide which would diffuse away from the root region more rapidly than the carbonate ions. Nevertheless, whatever the complex processes may be, field observations indicate that roots can both dissolve solid rock and act as nuclei for cementation. Cementation around roots leads to the formation of rhizocretions (Kindle, 1925). Dissolution by roots, on the other hand, provides the biochemical-corrosion component for Micro codium, i.e. the attack by Microcodium on the substrate may be a function of rhizo solution. The preservation of plant form (biosynthesis component) still requires explanation; it is suggested herein that penetrative fungal hyphae within plant cells and walls provide a template for such preservation. Synthesis of model
Following the above rudimentary introduction on plant physiology and anatomy, it is now possible to present evidence to substantiate the claim that observed morpho logical features of Pleistocene and Recent Microcodium are the result of mycorrhizal associations. This will be presented in two sections, (i) evidence for the presence of roots, and (ii) evidence for fungal presence and modification. Unfortunately, because the biological systems are no longer living it is impossible to demonstrate that the root-fungal association is definitely a symbiotic one. Neverthe less, the morphological similarity between the material studied here and actual mycorrhizal associations is considered sufficient to suggest that this was the case. (i ) Evidence for the presence of plant roots
(I) Size of Microcodium grains. Grains (Microcodium) from lbiza and southeastern Spain have similar dimensions to plant cells. (2) Shape of Microcodium grains. Grain shapes are similar to parenchymatous and collenchymatous plant cells. (3) Non-planar grain boundaries (Figs 2d, e, 5a, b, and 7a-c). Convex, concave and re-entrant faces are common for Microcodium grains; curved faces are not typical of inorganically precipitated calcite. (4) Arrangement of Microcodium aggregates. The cylindrical arrangement (Fig. 2c) in many aggregates is like that of the cortical layer of a plant root. (5) Occurrence of in situ Microcodium aggregates within channels (Fig. 3b). Channels in caliche (millimetre scale) are common. Many owe their origin to root channels. (6) Insoluble residues from indurated caliche profiles. Xylem vessel members (specialized vascular plant cells used for transporting water) and other plant remains 136
Biolithogenesis of Microcodium
have been extracted from Pleistocene caliche. Microchemical tests indicate that these residues contain lignin. (7) Presence of calcium oxalate. Plants may secrete crystals within their cells. According to Kelley (1950), however, a general opinion held maintains that fungal hyphae do not penetrate raphide cells (cells containing bundles of acicular needles). On the other hand, Kelley (1950) also mentioned the work of Busich (1913) who said that a fungus is not warded off by calcium oxalate but on the contrary forms it.Thus the minor amounts of calcium oxalate detected in Microcodium grains may be the product of fungal activity rather than of vascular plant secretion. (8) Surface features of Microcodium grains. Bordered pits on grain surfaces (Fig. 2f) correspond to positions of connection between cells of thin cytoplasmic strands (plasmodesmata). (9) Presence offine strands or fibrils (Fig. 3e). Fungal (?) bores (1 0 J.lm diameter) which penetrate Microcodium walls are surrounded by fine fibres (less than 0·05 J.lm diameter). The latter are interpreted as macrofibrils that constitute plant cells walls. ( 1 0) Subterranean habitat of Microcodium. Microcodium is encountered dominantly within a subterranean environment that shows pedological features, i.e. within a soil which by definition supports a biological ecosystem. ·
(ii) Evidence for the presence offungi (1) Filaments (1·0 J.lm diameter) radiating from the grain perimeter (Figs 7b and 8c). In thin section, radiating filaments show a radial arrangement with respect to Micro codium walls. These are interpreted as intracellular fungal hyphae, possibly haustoria (special absorptive hyphae that invade living cells). In endo- and ectendomycorrhizae these hyphae radiate out from the cell wall into the cell lumen. (2) Presence of filamentous networks. Following acid etching of Microcodium grains, rod- or needle-shaped structures (0·4--1·0 J.lm diameter) were revealed by examination under the SEM (Fig. 5d). Their size and shape are compatible with them being of fungal origin. In some Microcodium aggregates the arrangement of filaments forms an organized pattern. The total structure resembles a sclerotium which is a firm, rounded, often hard, mass of hyphae devoid of spores that forms a resting stage (Marks & Kozlowski, 1973).Trappe (1971) shows a photomicrograph (his fig.5, p.25) of a microtomed section of the surface of a sclerotium of Cenococcum graniforme. The hypha! arrangement in Microcodium resembles that of the mycorrhizal mantles formed by this fungus. Trappe (1971) illustrates such a mycorrhizal mantle (his fig. 4, p. 24) which is similar in shape and size to Microcodium aggregates sampled from the Eocene of northern Spain. (3) Coiled filaments or finely divided branches within Microcodium walls (Fig. 3f). These are considered to be fine networks of closely packed fungal hyphae within cell walls. Fungi are composed dominantly of chitin, whereas cellulose constitutes most of the cell wall in higher plants with minor amounts of lignin, tannins and pectic sub stances.Cellulose is rapidly broken down by microbial decomposition but chitin, when associated with polyphenols contained within the hypha! wall, resists decay for much longer (Potgieter & Alexander, 1966).Therefore, it is possible that the presence of fungi within cell walls preserves the cell form.A tentative proposal made here is that such a template is the reason for preservation of plant morphology. (4) Bores within Microcodium grains (Fig.3e).Tubular pores (0·4-0·1 J.lm diameter)
137
Colin F.
Klappa
within Microcodium grains are thought to be moulds of intracellular fungal hyphae. The hyphae have been later embedded in calcite so that the 'bores' are, strictly speaking, embedment structures (Bromley, 1970). (5) Protuberances in naturally etched Microcodium grains (Figs 2f and 3c). Rod shaped protuberances (0·5 J..Lm diameter) within naturally etched grains have been observed by SEM. Such intragranular (intracellular) structures, whose dimensions and forms are similar to haustoria, are interpreted as the fungal component of vesi cular-arbuscular mycorrhizae. (6) Hypertrophied Microcodium grains (Figs 2a and c). Hypertrophy (Gk. hyper= over + trophe food) is an excessive enlargment of cell size without an increase in cell number (Alexopoulous, 1962). Hypha! infection (intracellular) exerts a physical pressure on adjacent cortical cells (fungi forming mycorrhizal associations do not generally penetrate beyond the cortical layer). Enlargement of cells, because of this fungal infection, leads to interference during growth between adjacent cells. Con cavities and bizarre cell shapes (Figs 2e, 6a and b) are considered to be the result of this phenomenon. (7) Presence ofperitrophic (surrounding surfaces of roots) mycelia. Viewed under a binocular or petrographic microscope, Recent Microcodium grains from southeastern Spain were observed to be surrounded by networks of fungal hyphae (Fig. 4a). These form a mycelium (mass of fungal hyphae) or a rhizomorph (densely packed mass of fungal hyphae that resembles a tree root). SEM examination shows that these hyphae are slightly larger than the previously mentioned filaments. Hence, a dimorphism exists in the diameter of relatively thick aerial hyphae (2·0-5·0 J..Lm diameter) around Recent Microcodium grains, and finer hyphae ( 1·0 J..Lm or less in diameter) on and within these grains (Fig. 4b). A similar example of hypha! dimorphism has been recorded by Nicolson (1967). Thus, by documenting the fabric (size, shape and arrangement) of Microcodium grains by light and scanning electron microscopical observations and comparing the results with known biological features, Microcodium is interpreted as a product of mycorrhizal activity. =
MECHANI S M S FOR CALCITE PRECIPITATION WITHIN PLANT CELL S-CALCI FICATION
This section is an attempt to indicate possible mechanisms that may lead to the accumulation of calcite within plant cells, and conclusions reached are to be regarded as highly tentative. The term 'calcification' is used here to include preservation of plant form by calcite precipitation within vacated or original pore spaces (vacuoles) and metasomatic replacements of organic compounds by calcium carbonate. Treat ment of this subject can be conventionally divided into several categories as follows. I. Phenomena associated with plant growth
(i) Direct biochemical (a) Metabolic products of plants during normal growth, (b) metabolic products of symbiotic or parasitic micro-organisms, (c) secretion of substances in an attempt to flush out foreign intruders, and (d) selective uptake or rejection of ions by sorption (Lovering, 1959), ion exchange, or contact exchange (Keller & Frederickson, 1952).
138
Biolithogenesis of Microcodium
(ii) Indirect biochemical (a) Change in partial pressure of C02 because of synthesis and respiration of plant and associated microflora, (b) change in solute concentration because of evapo transpiration, (c) change in pH because of plant and microbial activity, and (d) exudation of organic acids and the sloughing off of plant debris.leading to changes in the chemical micro-environment of the rhizosphere. (iii) Physico-chemical (a) Accumulation of H 20 in the proximity of the root because of surface tension effects, and (b) diurnal and seasonal changes in temperature which affect the solubility of calcium carbonate. II. Phenomena associated with plant decay
(i) Microbial decomposition (a) Decay of plant organic material, and (b) decay of micro-organisms associated with plants during life (symbiotic or parasitic) or after death (saprophytic). (ii) Metabolic products of saprophytic organisms Documented isotopic analyses on Ancient (Eocene) Microcodium are given by Bodergat ( 1974).With respect of 1 3C/12C ratios, the conclusion reached by J. C. Fontes (who carried out the isotopic work) is quoted by Bodergat ( 1974, p. 209) as follows: '. . . pratiquement tout le carbone qui s'integre au carbonate a ete prealablement metabolise. ' Bodergat points out that this analytical evidence confirms the organic nature of Microcodium and that the carbon is photosynthetic in origin. By tacitly assuming that a 13C/12C isotopic analysis on Recent Microcodium would give similar results, only the above mentioned mechanisms for inducing calcification that involve metabolic processes will be considered in greater detail. Metabolic products ofplants during normal growth. As a result of metabolic activity in the cell, some plants form ergastic substances as cell inclusions (Cutter, 1969). Such substances include proteins, starch, fats, oils and crystals. Some of these may be waste products, other are stored food material. Crystalline deposits in various forms occur in the cells of certain plants and are generally considered to be waste products (Cutter, 1969). The size, shape and arrangement of isodiametric Microcodium grains are consistent with those of certain types of plant cells (particularly the parenchymatous cells of vascular plants). At first, this led to the idea that Microcodium grains were secretory crystals within plant cells. The observation, however, that Microcodium has destroyed or at least modified the substrate (Fig.3a) by dissolution and reprecipitation indicates that the product cannot be attributed simply to passive cavity filling of cell lumina. Metabolic products of symbiotic, parasitic and saprophytic soil-plant micro organisms. Little is known regarding the ability of microbes to precipitate calcium carbonate (Alexandersson, 1974) but the culture experiments of Krumbein (1968) may have some relevance in this context. He demonstrated that the microflora from a nari lime-crust (caliche) could produce large quantities of calcite. Likewise, Adolphe & Billy ( 1974) have observed the precipitation of calcite by bacteria in vitro, but it is unclear whether the phenomenon is the result of a direct or indirect biochemical control.
139
Colin
F.
Klappa
Metabolic secretion in response to foreign bodies. This idea may be treated with the greatest scepticism as I am unable to substantiate whether such mechanisms occur in nature for this given situation. Reasoning by analogy with pearl formation in oysters, the suggestion is put forward here that the penetration of plant cells by micro organisms may induce the secretion of substances in order to flush out intruders. Although mycorrhizae involve symbiotic rather than parasitic associations, the physical presence of fungal hyphae within the cytoplasm, or even restricted to the cell wall, may cause the release of certain substances. Alternatively, there may be a chemotactic response, stimulated by substances diffusing from the fungal hyphae. Whatever the exact mechanisms are, it is evident from consideration of the above possibilities that higher and lower plants have the ability to exert direct or indirect biochemical controls which may culminate in calcification of plant tissues. It is not denied that the above comments are speculative and should be treated with caution. The chemical complexities of the soil micro-environment are beyond the scope of this study. Nevertheless, as a concluding remark in this section, it can be stated unequivocally from direct observation that calcification has taken place.
ANCIENT MICROCODIU M
It is not the intention of this paper to give a complete petrographic analysis of Ancient Microcodium. This has been covered by the study of Bodergat (1974). After reviewing this work, together with the studies of Esteban (1972, 1974), Freytet (1969, 197 l a, 1971b, 1973), Lucas & Montenat (1967) and True (1975a, 1975b), and using my own findings on the same material, it has been possible to make comparisons between Ancient and Recent Microcodium. Description
The illustrations and descriptions of Bodergat (1974) and Esteban (1972, 1974), generally agree with observations made in the course of this work. Viewed with the light and scanning electron microscopes, Microcodium grains include the following features. The calcite that makes up the grains is non-limpid. Grains are dominantly prismatic in shape. Inclusions are commonly aligned parallel or sub-parallel to the length of the prisms. Grain boundaries typically display curved faces, both concave and convex, and show re-entrant outlines. Lines of insoluble residues occur between prisms and at the boundary between the Microcodium aggregate and the attacked rock. Some transverse sections perpendicular to the long axes of prisms have a central tubular hollow. Fractured grains show that the ultrastructure of the calcite that makes up the solid part of the prism is composed of a pile of thin plates, commonly although not invariably, oriented perpendicularly to the long axes of the prism. Some differences and additional features have been noted in the material examined in this study. For example, some samples show that the calcite plates that constitute the individual prisms are aligned sub-parallel to the long axes of elongate Microcodium grains. Another point at variance with previous studies concerns the insoluble residues. The presence of insoluble clay residues at the contact between the Microcodium grain and the enclosing substrate was not recognized in all samples (Figs 9a and c).Instead, in some cases, a micro-honeycombed structure (hitherto undocumented) with pore
140
Biolithogenesis of Microcodium
diameters of 1·0 Jlm was observed (Fig. 9d). An additional feature, not previously recorded in Ancient (Eocene) Microcodium, is the presence of disc-shaped constrictions along rod-shaped aggregates (Fig. 9b). Interpretation of new features recorded in Ancient Microcodium
A micro-honeycombed structure within Microcodium grains has apparently been found in Eocene samples studied by Dr L. Pomar (Esteban, written communication). Esteban thinks the units may be bacterial precipitates. In line with observations of Recent and Ancient Microcodium made during this study, however, it is considered that the walls forming the honeycombed structure are the result of calcite precipitation on intracellular fungal hyphae. Following decay of the hyphae, the resulting pores mark their former positions. The origin of the constrictions is not clearly understood, but as a tentative proposal it is suggested that they may represent the nodes of a plant (parts of a plant stem where one or more leaves arise). If this proves correct, then calcification of plant cells penetrated by fungal hyphae may not be restricted only to the root zone as emphasized earlier. Thus, preservation may be simply a function of environment; structures already enclosed within a sediment (roots) have a better chance of survival than aerial parts of a plant which tend to be rapidly oxidized (Barghoorn, 1952). The sample illustrated in Fig. 9b may reflect a case of unusual preservation.
COMPARI SONS BETWEEN ANCIENT AND RECENT
MICR O C O D I UM
On a megascopic scale, the occurrence of Microcodium at specific lithostratigraphic levels, namely at surfaces indicating subaerial exposure and within dominantly continental facies, is perpetuated throughout the geological record. The presence of Microcodium within calcareous sediments affected by pedogenesis suggests that a genetic relationship exists between Microcodium and the rhizosphere of lime-rich soils. Observation at microscopic level indicates the presence of radiating filamentous structures within single Microcodium grains. Pitted surfaces, and tubular pores of 1·0, Jlm diameter or less, are characteristic of Ancient and Recent Microcodium (Figs 2f, 4c 9c and d). Shapes of grains typically show curved faces (Figs 2a, e, 3b, 6a, b, 9a and b) a feature which militates against a purely inorganic origin. Some Ancient Microcodium aggregates, however, contain prisms with straight faces. Monocrystalline calcite with uniform extinction is characteristic of Ancient Microcodium, whereas Pleistocene and Recent Microcodium grains tend to be composed of a number of sub-crystals, commonly displaying a radial-fibrous fabric. The latter gives aggregate or sweeping extinction patterns.These differences are thought to result from subsequent diagenetic modification of Ancient Microcodium. Following the classification of Bodergat (1974), type I ('epis de mals' = corn on the cob) and type 2 ('colonies en laminae' = laminar colonies) are common habits shown by Ancient Microcodium. Type 3 (isodiametric grains forming a cortical layer that surrounds a central canal, Fig. 8b) is apparently rare in the Ancient, whereas samples of Pleistocene and Recent age are dominantly of this form. The size-range for type 3 grains is similar for Ancient and Recent Microcodium but types 1 and 2 tend to be 141
Colin F. Klappa
composed of somewhat larger, elongate prisms up to 1·0 mm in length. Such mono crystalline, elongate grains of Ancient Microcodium, commonly have cross-walls normal to the length of the prism (Figs 9e and f). Cross-walls within single Micro codium grains have not been recorded in Quaternary samples. The widths of Ancient and Recent Microcodium grains are similar; the greater lengths of the former may be due simply to recrystallization of a number of isodiametric grains. Thus, the 'cross walls' may, in reality, mark the sites of former outer walls of juxtaposed, more or less isodiametric, single grains. With regard to the smaller Microcodium (b) forms of Esteban (1972) which are possibly equivalent to the 'seed-plots' of Lucas & Montenat (1967) and Microcodium described by Montenat & Echallier (1977), their occurrence at the perimeter of rhizocretions or subjacent to root channel walls, displaying either a concentric arrangement in transverse sections or palisade rows in longitudinal sections, suggests an intimate relationship with plant root systems. They occur both in the Ancient and Recent. Their origin may be the result of calcite encrustation on fungal hyphae. Fungi commonly form a peritrophic mantle around root surfaces, utilizing sloughed-off debris and exudates of the root as sources of food. Both calcite encrusted fungal hyphae and detached calcified root hairs have been recorded by SEM around rhizo cretions and root moulds (Klappa, 1978). Microcodium (b) form grains typically have a central tube (unfilled) or channel, or a central rod (filled), which is surrounded by a layer of calcite 5-20 J.tm thick. This arrangement gives Microcodium (b) grains an overall diameter of between 10-40 J.lffi. The diameter and shape of the central tube or rod allows distinction between the two suggested origins; root hairs have diameters of 5-17 J.tm according to Dittmer ( 1949) and tend to be straight unless penetrated by fungi, whereas fungal hyphae tend to be somewhat narrower (0·5- 10 J.tm) depending on species (personal observations) and are commonly sinuous. This morphology, combined with the cut-effect gives a spaghetti-like appearance in thin section. Microcodium is considered to be, on the basis of this study, the product of calci fication of plant cells whose forms have been maintained by fungi which show mycor rhizal associations. Since mycorrhizae are not restricted to a particular species of plant, variations in form may exist between different plant species. Moreover, given a time-span from the Eocene to the present-day, morphological differences between Ancient and Recent Microcodium are to be expected. As well as the involvement of completely different species, sufficient time is also availabl� for variation to result from evolutionary change within a single plant species. As a concluding remark in this section, it is pointed out that variations between Ancient and Recent Microcodium may also be explained (away), in part, by the 'cinderella' of carbonate sedimentology, that is, by diagenesis.
CO N SE QUENCE S AND I MPLICATIO N S Misinterpretations, oversights
( ?)
In reviewing the literature, in an attempt to elucidate the origin of Microcodium, several petrographic descriptions were found that show affinities with those of Microcodium. For example Seghal & Stoops (1972, pp. 67-68) state that '. . . a puzzling form of calcite accumulation is the occasional occurrence of sand-sized, single,
142
Biolithogenesis of Microcodium
rounded calcite grains with wavy extinction.' They suggest that abrasion during trans port was responsible for the rounded shapes, and the wavy extinction is attributed to their derivation from metamorphic rocks. They also note that similar kinds of calcite crystals have been observed by other authors (AI Rawi, Sys & Laruelle, I 968 ; Altaie, Sys & Stoops, I 969) who could not categorically state their origin. Schlanger ( I 964, p. D I I ) states '. . . small radiating and sheaf-like clusters of stubby, acicular and rhombohedral crystals have formed in some mosaics. These crystals are coffee coloured and show weak pleochroism or absorption in shades of light yellow brown. Absorption is greater paraliel to the long axis of the crystals. These crystals show high birefringence and indices of refraction greater than calcite; they have not been identified.' Unfortunately, neither of the above descriptions are furnished with any illus trations, thus, direct comparisons are difficult to make. None the less, some points in their descriptions may be applied to Microcodium and perhaps should be interpreted with this consideration in mind. Folk ( 1971), in discussing unusual neomorphic fabrics illustrates an example of 'neomorphic bladed calcite forming a very crudely oriented (N .B4.C) crust on an intraclast' (his fig. 84B, p. I 65). He mentions a '. . . microspar matrix of blades L/W 2: I to 4: 1, often circular in cross-section. Some of these particles taper at their end, others splay out like a worri toothbrush. They have slightly undulose extinction, and appear to be made of poorly defined fibres.' Folk's illustration (his fig. 84B) shows a remarkable similarity with Plate V, figs 2 1, 22 and 23 of Calvet, Pomar & Esteban ( 1975). The latter figure illustrates rhizocretions surrounded by structures considered analogous to Microcodium (b). Harbaugh ( I96 I ), in a discussion on calcite fabrics in late Paleozoic limestones from Kansas, Texas and New Mexico, established four specific types of visibly crystalline calcite. He suggests that his 'blade calcite', characterized by tapered, blade shaped crystals bunched in flower-like aggregates, probably formed by recrystal lization under mild shearing stresses. The photomicrograph illustrating blade calcite (his Plate I B, p. 99), is alarmingly similar to Ancient Microcodium. No plausible explanation is given as to how mild shearing stresses could produce this form ; perhaps a closer examination at the stratigraphic horizon from which this sample was taken might lead to the discovery of a former Microcodium-attacked subaerial exposure surface. The observation that Harbaugh's ( I 96 I ) blade calcite shows a close resem blance to Microcodium was noted independently by Misik ( 1968). With such limited data it is not the intention here to reinterpret the above citations. They are pointed out only to give Microcodium a hearing before sentence is passed or deferred regarding unusual calcite fabrics. Geological importance
The chronological and stratigraphical importance of Microcodium is well estab lished in the French geological literature. Correct recognition allows precision of correlation and is equally effective in application to sedimentological studies. As already indicated, Microcodium is intimately related to a land surface, which, by definition is a disconformity and represents a time-equivalent horizon marker. Thus, the presence of in situ Microcodium indicates terrestrial conditions, and may provide evidence for subaerial exposure in otherwise marine lithofacies. 143
Colin F. Klappa
An origin associated with root systems necessitates the presence of a soil cover. Therefore, Microcodium may be used as a criterion for recognition of palaeosols. Soil formation, itself, requires a cessation or pause in sedimentation sufficiently long to allow pedogenetic processes to act on a given substrate. A prerequisite for the development of Microcodium appears to be a lime-rich soil. Petrographic analysis demonstrates that Microcodium not only modifies or obliter ates pre-existing textures but provides a potential source for sand-sized detrital calcite grains in reworked sediments. Thus, the cumulative effect of Microcodium on sub strate may provide valuable clues that assist in palaeoenvironmental reconstructions.
SUM MARY AND CONCLUSIONS
From samples of caliche collected in the western Mediterranean, calcite grains showing atypical fabrics have been recorded. A review of the literature indicates that such grains have been designated Microcodium, a hypothetical organism considered by early investigators to be possibly algal in origin. More recently, several workers have presented models which, although considering Microcodium to be the result of microbial activity, favour fungal, actinomycete, or bacterial interference. This study presents a new model based on petrographic examination of Eocene to Recent samples of Microcodium. Cumulative evidence, interpreted in the light of modern plant-soil ecosystems, suggests a mycorrhizal origin for Microcodium. The significant points resulting from this study are summarized as follows. ( 1 ) Occurrence. Caliche samples collected from mainland Spain and the island of lbiza, Balearics, reveal the presence of Microcodium grains constituting up to 43% of the rock by volume. The occurrence of Microcodium on Ibiza has not been pre viously documented. (2) Age. Stratigraphic, palaeontological and lithological relationships indicate the presence of significant quantities of Microcodium in the Eocene and Pleistocene. The existence of Microcodium in the Recent is reported here for the first time. (3) Unusual fabrics. Preservation of fine detail in Pleistocene and Recent samples, as revealed by SEM, portrays an exceedingly complex ultrastructure. The presence and subsequent calcification of microtubules, filamentous structures, radiating pore systems within grains, and protuberances, pits and raised borders on grain surfaces give a somewhat bewildering array of calcite fabrics when viewed in thin section. (4) Ancient versus Recent Microcodium. Sufficient details are retained in Ancient Microcodium to allow fruitful comparisons with Recent samples. Their origins are considered to be homologous. (5) Origin. Previous investigations regarding the origin of Microcodium are outlined. New field and petrographic data are at variance with earlier studies and have led to the formulation of a new model of formation. Microcodium is reinterpreted as being the result of calcification of mycorrhizae, a symbiotic association between fungi and cortical cells of roots. (6) Geological importance. Correct recognition of Microcodium has wide appli cation in terms of environmental, stratigraphic and palaeoecological studies. Emphasis is placed on Microcodium being a pedological feature and, thus, a valuable criterion for the recognition of the existence of a palaeosol.
144
Biolithogenesis of Microcodium
(7) Scope and limitations. It is hoped that this paper will at least provide an in troduction to those unfamiliar with this potentially useful diagenetic 'whim of nature' and lead to a search for a better understanding of distribution, environmental par ameters and diagenetic processes. Refinement is required to clarify many poorly understood but fundamental details resulting from this study. Such progress can be achieved only by an interdisciplinary approach.
AC K N 0 WLED G MENTS
It is a pleasure to acknowledge my debt to Prof. R.G.C.Bathurst for his continual guidance, astute criticism and intangible encouragement during the course of this work.To Dr M.Esteban, I am sincerely grateful for his advice, teaching and stimulat ing discussion during field studies in Spain and the friendship and genial hospitality shown by him and his 'Grup d'Estudi de Calcaries' at the University of Barcelona. SEM instrument time was provided by the Department of Botany, University of Liverpool, and operation was aided by the superlative technical assistance of C. J. Veltkamp. Special thanks are due also to many members of the Botany Department at Liverpool, especially to Dr J. C. Collins, Dr H. A. McAllister, Dr G. Russell and Dr S. T. Williams whose fruitful discussions helped to clarify biogenic aspects presented in this study. I greatly appreciate the comments of Prof. R. G. C. Bathurst, Dr P. J. Brenchley and Dr S.T.Williams which gave constructive criticism to an earlier draft of this work. Dr P. Enos provided useful suggestions during preparation. I am greatly indebted to J. Lynch for his cartographical assistance. Dr J. W. Wood kindly provided specimens from Palaeozoic shales of Missouri. I thank also the referees, Dr P. Freytet and Dr W. E. Krumbein, for their helpful suggestions on the original manuscript. Financial support provided by a NERC Research Studentship (Grant No. GT4/ 75/GS/131) is gratefully acknowledged.
REFERENCE S ADOLPHE, J.P . . & BILLY,
C. ( 1974) Biosynthese de calcite par une association bacterienne aerobe. 2873-2875. ALEXANDERSSON, E.T. (1972) Micritization of carbonate particles : processes of precipitation and dissolution in modern shallow-marine sediments. Bull. geol. Inst. Univ. Uppsala, 3, 201-236 . ALEXANDERSSON, E.T. ( 1974) Carbonate cementation i n coralline algal nodules i n the Skagerrak, North Sea : biochemical precipitation in undersaturated waters. J. sedim. Petrol. 44, 7-26. ALEXOPOULOl)S, C .J. ( 1962) Introductory Mycology, 2nd Ed. Wiley and Sons, New York. ALLARD, P., GANNAT, E . , LAPAICHE, N . , LEFAVRAIS-RAYMoi-m, A. & MARIE, P. ( 1959) Sur un niveau a Microcodium a Ia base du Tertiaire de Bresse. C. r. somm. Seanc. Soc. geol. Fr. 6, 150-151. AL RAWI, G .J., Svs, C. & LARUELLE, J. ( 1968) Pedogenetic evolution of the soils of the Mesopotamian Flood Plain. Pidologie, Gent, 18, 63-109. ALTAIE, F.H., Svs, C. & STOOPS, G. ( 1969) Soil groups of Iraq-their classification and characteriza tion. Pidologie, Gent, 19, 65-148. BARGHOORN, E.S. (1952) Degradation of plant tissues in organic sediments. J. sedim. Petrol. 22, 34--41. BoDELLE, J. & CAMPREDON, R. (1968) Les formations a Microcodium dans les Alpes-Maritimes franco-italiennes et Jes Basses-Alpes. Leur importance paleogeographique. Mem. Bur. Rech. Geol. Min. 58, 453-471 . C. r. hebd. Seanc. Acad. Sci. , Paris, 278,
145
Colin F. Klappa BoDERGAT,
A . M . ( 1974) Les microcodiums, milieux et modes de developpement. Docum. Lab. Geol. 137-235 . BODERGAT, A . M . , TRIAT, J . M . & TRuc, G. ( 1 975) L'origine organique des microcodiums : exemple du role des microorganismes dans Ia biocorrosion des roches carbonatees et Ia biosynthese de Ia calcite en milieu continental. Int. Sediment. Congr. 9th, Nice, 1 975. Theme, 2, 7-10. BOULANGER, D. & CROS, P. ( 1 957) Presence de Microcodium dans Ia region de Limoux (Aude). Bull. Soc. geol. Fr. Ser . 6, 7, 353-354. BouRROUILH, R. & MAGNE, J. ( 1 963) A propos de depots du Pliocene superieur et du Quaternaire sur Ia cote nord de l'ile de Minorque (Baleares). Bull. Soc. geol. Fr. Sir. 7, 5, 298-302. BowEN, G. D . ( 1 973) Mineral nutrition of Ectomycorrhizae. I n : Ectomycorrhizae, Their Ecology and Physiology ( Ed . by G . C. Marks & T. T. Kozlowski), pp. 1 5 1-205 . Academic Press, New York. BROMLEY, R.G. ( 1970) Borings as trace fossils and Entobia cretacea Portlock, as an example. In : Trace Fossils (Ed. by T. P. Crimes & J. C. Harper), pp. 49-90. Geol. J. Special Issue, 3. BuRGES, A. ( 1 958) Micro-Organisms in the Soil. Hutchinson, London. BusiCH, E. ( 1913) Die endotrophen Mykorhiza der Asclepidaceae. Verh . zoo!. bot. Ges. Wien, 63, 240264. CALVET, F., PoMAR, L. & EsTEBAN, M. ( 1 975) Las Rizocreciones del Pleistoceno de Mallorca. lnst. Invest. Geol. Univ. de Barcelona, 30, 3 5-60. CUTTER, E.G. ( 1 969) Plant Anatomy : Experiment and Interpretation Part I. Cells and Tissues. Edward Arnold Ltd, London. CuviLLIER, J. ( 1955) Sur l'origine de Microcodium . Bull. Soc. geol. Fr. Ser. 6, 5, 295-297. CuviLLIER, J. & SACAL, V. ( 1 961) Stratigraphic correlations by microfacies in Western Acquitaine. Int. sedim. petrogr. Ser. 2, E. J. Brill, Leiden . DAVIES, P.J. & TILL, R. ( 1 968) Stained dry cellulose peels of ancient and Recent impregnated sedi ments. J. sedim . Petrol. 38, 234-237. DEMANGEON, P. ( 1956) Importantes formations dues a des algues calcaires dans le Montien rouge (Vitrollien) du Midi de Ia France. C. r. hebd. Seanc. Acad. Sci. , Paris, 242, 1905-1 907. DICKSON, J.A.D. ( 1 966) Carbonate identification and genesis as revealed by staining. J. sedim. Petrol. 36, 491-505. DITTMER, H.J. ( 1 949) Root hair variations in plant species. Am. J. Bot. 36, 1 52-1 55 . DURAND, J.P. ( 1962) Role e t repartition des Microcodium dans les formations fiuviolacustres pro venc;ales du Cretace superieur et de !'Eocene. C. r. somm. Seanc . Soc. geol. Fr. 9, 263-265. EDWARDS, W.N. ( 1 932) Lower Eocene plants from !stria. Ann. Mag. nat. Hist. 56, 21 3-216. EsTEBAN, M . ( 1 972) Una nueva forma de prismas de Microcodium elegans Gluck 1912 y su relaci6n con el caliche del Eoceno Inferior, Marmella, provincia de Tarragona (Espana). Rev. Inst. inv. Geol. Dipt. Prov. Barcelona, 27, 65-81. ESTEBAN. M. ( 1 974) Caliche textures and Microcodium. Boll. Soc. geol. ita!. 92, Suppl. 1 973, 105-1 25. EVENARI, M., SHANAN, L. & TADMOR, N. ( 1 971) The Negev: the Challenge of a Desert, Harvard University Press, Cambridge, Mass. FAHN, A. ( 1974) Plant Anatomy, 2nd Ed. Pergamon, Oxford. FAURE-MURET, A. & FALLOT, P. ( 1954) La formation a Microcodium au pourtour de I'Argentera Mercantour. Bull. Soc. geol. Fr . Sir. 6, 4, 1 1 1-138. FEIGL, F. ( 1 943) Laboratory Manual of Spot Tests, Academic Press, New York. FOLK, R.L. ( 1971) Unusual neomorphism of micrite. In : Carbonate Cements (Ed . by 0. P. Bricker), pp. 163-168. Johns Hopkins Press, Baltimore. FRANCOIS, S. & SIGAL, J. ( 1957) Quelques donnees nouvelles sur Ia morphologie et Ia repartition stratigraphique des Microcodium Gluck 1 9 1 2 . C. r. somm. Seanc. Soc. geol. Fr. 10, 1 68-171 . Fac. Sci. Lyons, 62,
FREYTET, P.
( 1969) Un nouveau gisement de Microcodium cavernicoles : le paleokarst de St Beauzille (Herault). C. r. somm. sianc . Soc. geol. Fr. 5, 1 66.
FREYTET,
P. ( 1971 a) Les depots continentaux et marins du Cretace superieur et des couches de passage a l'E ocene en Languedoc. Bull. Bur. Rech. Geol. Min. Sir. 2 , Sect. 1, 4, 1-54.
FREYTET,
P. ( 1 971 b) Paleosols residuels et paleosols alluviaux hydromorphes associes aux depots fiuviatiles dan� le Cretace superieur et l' Eocene basal du Languedoc. Revue Geogr. phys. geol. dyn. Ser. 2, 13, 245-268. FREYTET, P. ( 1 973) Petrography and paleo-environment of continental carbonate deposits with particular reference to the Upper Cretaceous and Lower Eocene of Languedoc (Southern France). Sedim. Geol. 10, 25-60. 146
Biolithogenesis of Microcodium FREYTET,
P. ( 1 975) Quelques observations petrographiques sur les calcaires continentaux rencontres
a !'excursion de mai 1 974 de I' A.G .B.P. : facies lacustres, modifications pedologiques, Micro
1 5-23. P. ( 1974) Presence de couches inframiocenes a Microcodium dans Ia marge orientale du Bassin Tertiaire de Digne-Valensole. C. r. hebd. Seanc. Acad. Sci. , Paris, Ser D, 278, 2087-2090. GLUCK, H. ( 1 9 1 2) Eine neue gesteinbildende Siphonee (Codiacee) aus dem marinen Tertiiir von SUddeutschland. Mitt. bad. geol. Landesanst. Bd. 7, 3-24. GoTns, M . ( 1 963) Sur un cas d'heterotrophie de Microcodium. Bull. Soc. geol. Fr. 5, 838-843. GRAY, T.R .G . & WILLIAMS, S .T. (1971) Soil micro-organisms. Oliver & Boyd, Edinburgh. GUBLER, Y. (1 955) L' E ocene subbrian<;onnais au N.E. du massif d' Argentera. C. r. somm. Seanc. Soc. geol. Fr. 6, 82-85. GUILLAUME, S. ( 1 961) Presence du Turonien dans Ia vallee de I'Ognon (Doubs). C . r. hebd. Seanc . Acad. Sci. , Paris, 258, 3006-3007. GuRR, E. (1 965) The Rational Use of Dyes in Biology and General Staining Methods. Leonard Hill, London. HARBAUGH, J.W. (1961) Relative ages of visibly crystalline calcite in late Paleozoic limestones. State geol. Surv. Kansas, Bull., 152, 9 1 - 1 26. JoooT, P. ( 1 935) Microcodium elegans GlUck du Miocene de Bade ne semble pas etre une Algue. C. r. somm. Seanc. Soc. geol. Fr. 4, 5 1-52. JoHNSON, J.H. (1 953) Microcodium GlUck est-il un organisme fossile? C. r. hebd. Seanc. Acad. Sci., Paris, 237, 84-86. JOHNSON, J.H. (1957) Paleontology. Calcareous algae of Saipan. Prof Pap . U.S. geol. Surv. 280-E, 209-246. JOHNSON, J. H . (1961a) Fossil algae from Eniwetok, Funafuti and Kito-Daito-Jima. Prof Pap. U.S. geol. Surv. 260-Z, 907-950. JoHNSON, J. H . ( 1 961b) Limestone-Building Algae and A lgal Limestones. Colorado School Mines, Boulder. KAMPTNER, E. ( 1960) Microcodium aus dem Eoziin des Basler Tafeljura. £clog. geol. Helv. 53, 843845. KAMPTNER, E. (1962) Neues Uber das problematische Fossil Microcodium. Paliiont. Z. 36, 1 1 . KELLER, W.D. & FREDERICKSON, A. F. (1952) Role of plants and colloidal acids in the mechanism of weathering. Am. J. Sci. 250, 594-608. KELLEY, A.P. (1950) Mycotrophy in Plants, Chronica Botanica Co. , Waltham, Mass., U.S.A. KINDLE, E.M. ( 1 925) A note on Rhizocretions. J. Geol. 33, 744-746. codium. Bull. d'Inform. Geol. Bassin de Paris, 12,
GIGOT,
KLAPPA,
C. F. (1978) Morphology, composition and genesis of Quaternary calcretes from the western Unpublished Ph .D. Thesis, University of Liverpool .
Mediterranean : a petrographic approach. KRUMBEIN,
W.E. (1968) Geomicrobiology and geochemistry o f the 'Nari-Lime-Crust' (Israel). In : (Ed. by G . MUller and G. M . Friedman), p p . 1 3 8-147. Springer-Verlag, Berlin. Recent Developments in Carbonate Sedimentology in Central Europe
LAPPARENT,
A.F. (1 966) A propos des conglomerats antenummulitiques des Alpes de Provence. Bull · 454-457.
Soc. geol. Fr. Ser. 7, 8,
T.S. ( 1 959) Significance of accuuJulator plants in rock weathering. Bull. geol. Soc. Am. 70, 781-800.
LOVERING,
0
LucAs, G. & MoNTENAT,
C. (1967) Observations sur les structures internes et le developpment des 909-9 1 8 .
Microcodium. Bull. Soc. geol. Fr. Ser. 7, 9, MARIE, 10,
P. ( 1957) Sur quelques gisements fran<;ais a Microcodium . c. r . somm. Seanc. Soc. geol. Fr. 1 71-175.
MARKS,
G.C. & KozLOWSKI, T.T. (1973) Ectomycorrhizae : Their Ecology and Physiology. Academic Press, New York.
MASLOV,
V.P. (1 956) Iskopaemye sinezelnye vodorosli SSSR. Trudy. !nsf. geol. Nauk, Mask. 160, 1-301 . MASSE, J .P , TRIAT, J.M. & T RUC , G . (1972) Surfaces a Microcodium affectant de Cretace de Ia partie occidentale des Monts de Vaucluse (sud-est de Ia France) : mise en evidence de leur age E ocene. C. r. hebd. Seanc. Acad. Sci. Paris, Ser. D, 275, 325-328. Miil iK, M. (1 968) Some aspects of diagenetic recrystallization in limestones. Intern. Geol. Congr. 23rd, Prague, 1 968. Rept. Session, 8, 1 29-1 36. .
147
Colin F. Klappa MoNTENAT, codium
C. & ECHALLIER, J.C. ( 1977) L'importance des structures organiques du type Micro dans Ia constitution des croutes calcaires pleistocenes . C. r. hebd. Seanc. Acad. Sci.,
Paris, Ser D, 285, 3-6. MoRET,
L. ( 1 952a) Curieuses cristallisations de calcite, attribuees a tort a des Algues (Microcodium), dans Ia partie terminale du Cretace superieur Alpin et Pyreneen. C. r. hebd. Seanc. Acad. Sci. , Paris, 235, 762-764. MORET, L. ( 1952b) Curieux encroutements de calcite attribuees a des Algues ( Microcodium) dans Ia partie terrninale du Cretace superieur Alpin et Pyreneen. Trav. Lab. Geol. Univ. Grenoble, 30, 55-59. MoRET, L. & FLANDRIN, J. ( 1 961) Nouvelles observations de 'Microcodium' dans les Alpes fran<;;a ises. Trav. Lab. Geol. Univ. Grenoble, 37, 1 9-24. NICOLSON, T.H. (1967) Vesicular-arbuscular mycorrhiza-a universal plant symbiosis. Sci. Prog. Oxford, 55, 561-581 . NURY, D . , REY, R . & Roux, R.M.N. ( 1970) Les gasteropodes des series detriques d u Rouet (Bouches du-Rh6ne). C. r. hebd. Seanc. Acad. Sci. Paris, Ser. D, 271, 2283-2285. PAQUET, J. ( 1961) Observation sur 'l'echaille de Virac' et sur Ia presence de recifs a Microcodium au N du sillon d'Ales (Ardeche). C. r. somm. Seanc. Soc. geol. Fr. Ser. 7, 3, 106-107. PLAZIAT, J.C. ( 1971) Observations paleolimnologiques sur les lacs eocenes situes entre le massif de Mouthoumet et La Montagne Noire (detroit de Carcassonne, Aude-Herault). Remarques sur les impressions marines a prelude lacustre. Colloque de Pateolimnologie. 96th Congr. Soc. Sav . . Toulouse, 1971, Sci. , 2 , 71-93. PoTGIETER, H.J. & ALEXANDER, M. ( 1966) Susceptibility and resistance of several fungi to microbial lysis. J. Bact. 91, 1 526-1 532. RECH-FROLLO, M. ( 1 948) Conditions lacustres en milieu marin . Le Danien de Ia partie orientale du sillon nord-pyreneen . Bull. Soc. geol. Fr. Ser. 5, 18, 115-122. RICHARD, F. ( 1967) Decouverte d'un horizon a Microcodium dans Ia serie Cretaceo-Tertiaire de Gocek (prov. de Mugla, Turquie). C. r. hebd. Seanc. Acad. Sci. , Paris, 264, 1 1 33-1 1 36 . Roux, M . ( 1970) Le synclinal de Taulanne (Basses-Alpes) et les consequences du chevauchement de Castellane sur sa bordure nord-est. Geol. Alpine, 46, 177-188. RuTTE, E . ( 1954) Zwei neue Vorkommen von Microcodium elegans (Chlorophyceae) im Tertiar Siidwestdeutschlands. Paliiont, Z. 28, 1 45-154. SCHLANGER, S .O . ( 1 964) Petrology of the limestones of Guam. Prof Pap. U.S. Geol. Surv. 403-D, 1-52. SEGHAL, J.W. & STOOPS, G. ( 1 972) Pedogenic calcite accumulation in arid and semi-arid regions of the Indo-Gangetic alluvial plain of erstwhile Punjab (India)-their morphology and origin . Geoderma, 8, 59-72. STURANI, C. ( 1 963) La couverture sedimentaire de l'Argentera-Mercantour dans le secteur compris entre les Barricate et Vinadio. Trav. Lab. Geol. Univ. Grenoble, 39, 83-124. TRAPPE, J.M . ( 1971) Mycorrhiza-forming Ascomycetes. In : Mycorrhizae (Ed . by E. Hacskaylo), pp. 19-37. Proc. 1st N. Amer. Conf on Mycorrhizae, 1969. TRuc, G . (1975a) Sols a profil calcaire differencie et pellicules rubanees dans Le Paleogene du Sud Est de Ia France. In : Colloque ' Types de croutes calcaires et leur repartition regionale', Strasbourg, 1975, pp. 108-113 . TRUC, G. ( 1 975b) Les encroutements carbonates lies a Ia p6dogenese ; role important des micro organismes : biocorrosion et biosynthese de Ia calcite en milieu pedologique confine. Int. sedim . Congr. 9th, Nice, 1975. Excursion Guide A2, pp. 47-55. WARD, W . C . ( 1975) Petrology and diagenesis of carbonate Eolianites of northeastern Yucatan, Mexico. In : Studies in Geology No. 2-Belize Shelf-Carbonate Sediments, Clastic Sediments and Ecology, pp. 500-571. Am. Ass. Petrol. Geol. WINLAND, H.D. ( 1971) Nonskeletal deposition of high-Mg calcite in the marine environment and its role in the retention of textures. I n : Carbonate Cements (Ed . by 0. P. Bricker), pp. 278-284. Johns Hopkins Press, Baltimore. WooD, J.M . & BAssoN, P.W. ( 1972) Specimens resembling Microcodium elegans Gliick from Paleozoic Shale(of Missouri. Am. Mid!. Nat. 87, 207-214.
(Manuscript received 25 November 1 977 ; revision received 3 February 1 978)
148
Reprinted from Sedimentology (1980) 27 613-629
Rhizoliths in terrestrial carbonates: classification, recognition, genesis and significance
C O L I N F. KL A P P A*
Jane Herdman Laboratories of Geology, University of Liverpool, Liverpool, U.K.
ABSTRACT
Rhizoliths are defined as organosedimentary structures resulting in the preservation of roots of higher plants, or remains thereof, in mineral matter. They are abundant and characteristic features of Quaternary terrestrial carbonates (calcretes and aeolianites) from coastal regions of the western Mediterranean. Field and petrographic observations indicate that five basic types of rhizoliths can be recognized: (I) root moulds, which are tubular voids that outline positions of former, now decayed roots; (2) root casts, which are sediment- and/or cement-filled root moulds; (3) root tubules, which are cemented cylinders around root moulds; (4) rhizocretions s.s., which are pedodiagenetic mineral accumulations (here low magnesian calcite) around living or dead plant roots; and (5) root petri factions, which are mineral impregnations or mineral replacements of organic matter whereby anatomical features of roots have been preserved partially or totally. Apart from rhizoliths them selves, roots of higher plants are responsible for the formation of numerous and characteristic features of pedogenetically affected terrestrial carbonates. Plant roots are responsible for, or contri bute to, the formation of alveolar textures, in situ brecciation (rhizobrecciation) textures, horizontal sheet cracks, vertically elongate glaebules (concretionary soil structures) and micritization (rhizo micritization) within terrestrial carbonates. Rhizoliths, together with the above features, are products of pedodiagenesis. More significantly, rhizoliths and related features are indicators of palaeosols and hence of subaerial vadose environments in ancient (post-Silurian) successions.
INTRODUCTION
preserved records, fossil root marks and casts have received little attention from geologists, and their palaeoecological potential has scarcely even begun to be examined'. Boyd (1975), in referring to the problems of dis tinguishing between burrow and root traces stated (p. 77): 'The time seems propitious for a major study, from a geological point of view, of modern root systems and their characteristics. . . . Until such work provides trustworthy criteria for recognition of fossil roots, authors should include descriptions of the features they interpret as roots . . . rather than be content merely to announce their presence'. In response to the comments quoted above, the purpose of this paper is: (1) to propose a terminology and classification for rhizoliths; (2) to document and
Rhizoliths, herein newly defined as organosedi mentary structures produced by roots, provide evidence of higher plant colonization of subaerially exposed sediments and rocks in the post-Silurian rock record. These structures, termed in various ways by other workers (Table 1), have been recog nized and reported widely. However, few accounts have documented their microscopic fabrics and textures as pioneered by Calvet, Pomar & Esteban (1975). Indeed, most citations go no further than merely listing their presence in the field. Sarjeant (1975, p. 173) remarked: 'Despite the abundance of
*Present address: Gulf Canada Resources Inc., P.O. Box 130, Calgary, Alberta, Canada T2P 2H7. Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
149
Colin F. Klappa
T
Term Rhizomorph Rhizoconcretion; rhizocretion; root-shaped concretion Petrified wood (replacement and impregnation) Root structures Racines petrifiees Travertinized roots Calcareous concretions; rhizoconcretions; calcareous rhizo-concretions Pedotubule Fossil mangrove Rhizoconcretions Dikaka Akkyrshi Tubular fossils Root casts Rhizocretions Carbonate crystal tube (carbonate root pseudomorph) Rhizocretions Pedotubules Rhizocretions; root hair sheaths Root-like structures Rhizocretions Rhizocretions Land plant root structures Rhizoliths
illustrate their characteristic features; (3) to distin guish them from animal burrowing systems; (4) to show how they form; and (5) to point out their environmental significance.
TERMINOLOGY AND CLASSIFICATION
Northrop (1890) used the term 'rhizomorph' to describe root-like structures from the Bahamas. Derived from the Greek rhizo root + morphe form, this term would be useful as a basic field description but, unfortunately, has a strict biological densely packed mass of meaning (rhizomorph fungal hyphae that resembles a tree-root). Thus, its use in this context for higher plant root structure is not advocated here. The term 'rhizoconcretion', considered synony mous with Northrop's 'rhizomorph' (American Geological Institute, 1972), was used by Kindle (1923) to describe cylindrical concretions of calcium carbonate and quartz sand which had formed around decayed roots. In a later communication, Kindle (1925) reported that 'rhizocretions' or root shaped concretions occurred also around living roots and suggested (1925, p, 744) that 'some, and prob ably all, rhizocretions must therefore be recognized as phenomena associated with growth, instead 9f the decay of roots'. The term rhizocretion has been used increasingly =
MATERIALS AND METHODS
=
Samples examined during this study were collected from coastal regions of mainland Spain and the Balearic Islands of Ibiza and Mallorca in the western Mediterranean. The samples were taken from terrestrial carbonates, mainly calcretes (caliche) and aeolianites, of Quaternary age. A deliberate attempt was made to collect both Pleistocene and Recent rhizoliths so that their fabrics and textures could be compared and their respective diagenetic histories followed. This study is based mainly on direct observation at various levels, ranging from field to microscopic (light and scanning electron microscopy). These studies were supplemented by microchemical stain ing and X-ray diffraction methods, details of which are summarized in Klappa (1978a, b). 150
=
Rhizoliths in terrestrial carbonates
in recent years to describe various forms of root structures that occur in calcretes and subaerially exposed marine carbonates (e.g. Johnson, 1 967; Steinen, 1 974; Ward, 1975; Calvet et a!., 1 975; Esteban, 1 976; Read, 1 974; Harrison, 1 977). The use of the term rhizocretion to describe pedodia genetic accumulations around roots is reasonable. However, when the root itself has been replaced partly or totally by calcite, the concept of con cretionary growth is inappropriate for description of plant root preservation by mineral replacement. Replacement of organic matter by mineral matter is better termed petrifaction or petrification (American Geological Institute, 1 972) and to be more spe:::ific, calcification and silicification for replacement by calcite and silica respectively. In the field, it is rarely possible to discern petri faction, especially in samples where concretionary growth also has taken place around roots. Thus, designation of all plant root structures as rhizo cretions or rhizoconcretions (preferential cemen tation around roots) does not strictly consider preservation of plant anatomical features. Because of this conceptual omission, I propose the general term rhizolith to include accumulation and/or cementation around, cementation within, or replacement of, higher plant roots by mineral matter. In addition, and at the risk of neologistic excessiveness, I propose also the term rhizolite for a rock showing structural, textural and fabric details determined largely by the activity, or former activity, of plant roots. The term rhizolite would appear to be equivalent to 'root rock', an informal term used by Perkins (1977) to describe Pleistocene rocks of south Florida com posed almost exclusively of 'root tube calcifications'. It is pointed out here that the terms rhizolith and rhizolite are necessarily genetic; their use and usefulness depend on availability and reliability of criteria for their recognition and, more fundamentally on the identification of such criteria which allow reasonable interpretations to be made. To ensure that prevalent confusion and ambiguity in rhizolith terms are not perpetuated (see Table 1 for existing terminology), the following definitions are given for five basic types of rhizoliths which can be recognized: (1 ) root moulds; (2) root casts; (3) root tubules; (4) rhizocretions s.s.; and (5) root petrifactions. (1 ) Root moulds are tubular voids which mark the positions of now decayed roots; (2) root casts are sediment- and/or cement-filled root moulds; (3) root tubules are cemented cylinders around root moulds;
(4) rhizocretions s.s. are pedodiagenetic mineral accumulations (here low magnesian calcite) around plant roots; and (5) root petrifactions are mineral impregnations or mineral replacements of organic matter which have preserved anatomical features of roots partly or totally. These genetic forms may occur individually or in various combinations. In addition, unaltered or decayed plant root material may occur essentially intact within sediment and be cemented to form composite rhizoliths. Fig. 1 summarizes the common forms of rhizoliths recorded in Quaternary calcretes and aeolianites from the western Mediterranean. The possible diagenetic pathways of the basic types of rhizoliths and their relationships to each other are also indicated in Fig. 1. Figure 2 compares root systems with burrow systems and Figs 3-9 illustrate various types of rhizoliths.
FIELD CHARACTERISTICS OF RHIZOLITHS AND THEIR DISTINCTION FROM BURROWS
Shape
Rhizoliths, li.ke animal burrows, are commonly circular in cross-section and cylindrical in shape (Figs 2, 3c, f and 4c). Size
Lengths of rhizoliths vary from a few centimetres (Fig. 4c) to several metres (Fig. 3e), the longest recorded in this study being 4 m long. Burrows, on the other hand, are generally less than a metre long but much larger burrows are known; Bromley et a!. (1 975) have re:::orded vertical dwelling burrows up to 9 m long (but only 0·5 em in diameter). The diameters of rhizoliths range from 0·1 mm to about 20 em. Branching
Downward bifurcations with decreasing diameters of second, third and fourth order branches (Fig. 2c) distinguish rhizoliths from animal burrows (Plaziat, 1 971). Burrowing systems usually have uniform diameters (Fig. 2d-f). Orientation
Rhizoliths show many orientations although the commonest forms occur· as vertical, isolated or 151
Colin F. Klappa e
b Root systems
d
d
e
20 0 L..-.----1 em
g
0 10 L..-.----1 em
Fig. 2. Root systems versus burrow systems (a)-(c) root systems; note downward tapering diameters. (a) Lateral roots; (b) taproot; (c) generalized root system with well developed taproot and laterals. (d)-(g) Common burrows· (d) Skolithos; (e) U-shaped burrow with spreite� Diplocraterion; (f) branching burrow network Chon drites; (g) part of boxwork burrow system Ophiomorpha. h Substrate
Although roots prefer to grow in unconsolidated sediments, they can, if necessary, bore through indurated rock (Klappa, 1 978a). Roots are con cerned primarily with the activities of absorbing water and mineral nutrients from soil solutions storage of reserve food supplies and anchorag� (Russell, 1 973). There is no reason why roots cannot
Fig. 1. Schematic representation of types of rhizoliths and their relationships. (a) Hairless part of root in lime sand. (b) Root mould in lime sand. Vadose fringe cement (low magnesian calcite) prevents mould from collapse. (c) Root tubule. Intergranular pore space in proximity of root mould has been filled with low magnesian calcite cement. Remainder of lime sand is uncemented or weakly cerr.ented. Cementation took place either during lifespan c) or from wall of of root which has since decayed (a a root mould because of greater porosity and permea bility and hence greater availability of percolating c). CaCO:�-bearing solutions through the mould (b (d) Root cast formed by filling root mould or tubule with sediment and/or calcite cement. (e) Root with root hairs in lime sand. (f) Rhizocretion. Etching by root hairs of lime sand and localized reprecipitation of dissolved calcite as isolated micritic grains produces a chalky rhizocretion (compare with e). (g) Root petrifaction. Partly decayed root and root hairs impregnated with calcite (see text and Figs 7 and 8 for details). (h) Rhizo cretion. Partly decayed root and sloughed off root hairs encased in partly etched lime sand. Root materials· are not calcified (contrast with g).
Fig. 3. (a) Road section of glaebular calcrete profile. Thin hardpan (H) has been disrupted in three places (X). Glaebules are better developed below these disrupted areas than below undisturbed hardpan. Closer inspection reveals that many glaebules are rhizoliths (see b, c). Disrupted areas underlain by rod-shaped glaebules are interpreted as fossilized root systems (root casts); each X marks the position of a former tree. Cuenca, E Spain. Hammer in centre above hardpan is 28 em long. (b) Ver tically elongate glaebules developed at hardpan horizon shown in a. Petrographic examination shows that rod shaped glaebules (R) are root casts. (c) Well cemented root casts (R) in glaebular horizon below hardpan. Less well cemented surrounding matrix has been eroded away. (d) Living root surrounded by chalky calcareous sheath to form an incipient rhizocretion. Constanti, NE Spain. (e) Vertical taproot rhizocretion (root still present but not visible in photograph) composed of cemented aeolianite. Note second order rhizocretions form lateral extensions. Pleistocene, Cala Conta, lbiza. (f) Root tubule consisting of cemented aeolianite. Arrow points to root mould (2 mm diameter). Note difference between size of tubule and genetically related root mould. Pleistocene, Cala Conta, Ibiza.
-->
-->
branching structures (Figs 3b-f, 4b and c) and as ramifying horizontal networks. Burrow systems are commonly vertical (e.g. Skolithos), horizontal (e.g. Fucl(sopsis) or inclined (e.g. Chondrites, Zoophycos).
1 52
Rhizoliths in terrestrial carbonates
153
Colin F. Klappa
man & lkawa, 1 958), gibbsite (Valeton, 1 971 ) and gypsum (Calvet et al., 1 975). Rhizoliths of White Sands, New Mexico are made entirely of gypsum (M. Esteban, personal communication).
live within indurated rock if these activities can be carried out normally in such substrates. Thus, the occurrence of borings in rocks does not necessarily militate against an origin related to former plant activity. On the contrary, branching borings with decreasing bore diameters downward strongly sug gest an origin associated with root systems. Animal traces, like plant root traces, are found in both unconsolidated sediments (burrows) and rocks (borings).
PETROGRAPHIC CHARACTERISTICS OF RHIZOLITHS
Whilst modern rhizoliths can be recognized easily in the field, especially when surrounding, less well consolidated materials have been removed by erosion and root materials are still present, they are more difficult to recognize in older rocks which have a long history of diagenesis. Apart from several notable exceptions (Kindle, 1 923, 1925; Ruth, 1 927; Johnson, 1 967; Calvet et a!., 1 975), rhizoliths have not received the attention they deserve. This section documents petrographic characteristics of Pleistocene and Recent rhizoliths to help identification of ancient equivalents in the geological record, and presents new information on the morphology and genesis of these structures.
Location
Rhizoliths and associated features of the rhizosphere (zone of soil around root which is affected by root activity) such as soil fungi indicate terrestrial con ditions and subaerial exposure of marine or lacustrine sediments. Most burrows, however, are found in marine sequences although terrestrial burrows are not uncommon (Brewer, 1964). Rhizoliths can be used to help recognize pedo genetically modified substrates, an example of which is illustrated in Fig. 4. Figure 4a shows an unaltered, cross-bedded aeolianite. Figure 4b and c show aeolianites of the same age (Upper Pleistocene) which have been subjected to pedoturbation (churn ing by soil organisms) and subsequent cementation. The formation of rhizoliths has destroyed pre existing bedding structures of the aeolianites shown in Fig. 4b and c whereas cross-bedding of the aeolianite shown in Fig. 4a has not been affected by pedogenesis.
Root moulds
Root moulds are simply tubular voids left after roots have decayed (compare Fig. 1 a with b). Preservation of the morphology of root networks indicates that soil materials must have had sufficient rigidity to prevent collapse of root mouldic porosity (Fig. 9a). This can be accomplished if soil materials are partly or completely lithified, the latter condition indicating that roots must have actively bored into rock or occupied pre-existing tubular voids within indurated substrates. In thin section (Fig. 9b, see also Esteban, 1 974, figs 2 and 5; Steinen, 1 974, fig. 4; Harrison, 1 977, plate 9), closely spaced sub-millimetre-sized root moulds give what is referred to as an alveolar texture.
Internal structures
The presence of contained root material (Figs 3d and 5b) provides an unequivocal criterion for the recognition of rhizoliths. In samples lacking plant material, the anatomical features of roots can some times be recognized in the field but are more easily recognized in thin section (Figs 5e, f, 6e and f). Although some animal burrows contain internal structures (e.g. spreiten), they are usually distinguish able from root anatomical features.
Root tubules
MINERALOGY
Rhizoliths examined in this study are composed dominantly of low magnesian calcite although silicified roots have been recorded. Other workers have reported rhizoliths containing dolomite (Sher1 54
Root tubules are cement or cemented sediment cylinders around root moulds (Fig. lc). The cement, usually consisting of low magnesian calcite, is responsible for preservation of root morphology in otherwise unconsolidated sediments (Fig. 3f). Cementation in the rhizosphere may take place during life or during decay of the root. Precipitation of low magnesian calcite needles tangentially around roots also produced root tubules after roots have
Rhizoliths in terrestrial carbonates
Fig. 4. (a) Cross-bedded aeolianite of Pleistocene age. Coli d'en Rabassa, Mallorca. (b) Pleistocene aeolianite containing tuberose rhizoliths. Note absence of primary bedding structures. Compare with a which shows no evidence of former plant activity. Punta Chinch6, Ibiza. (c) Ramifying network of branching rhizoliths within Pleistocene aeolianite. Two types of rhizolith present: tuberose rhizocretions (Tr) and calcified (petrified) roots (Cr). Location: Campo de Tiro, Mallorca. 155
Colin F. Klappa
decayed completely (Fig. 9e and f). In thin section, these calcite needles (encrusted on now decayed roots) form micritic walls and, if sufficiently abun dant, form alveolar textures (Fig. 9c and d). Root casts
Filling of a root mould, either by sediment or cement, or both, produces a root cast (Fig. I d). For example, Fig. 3b and c illustrates sediment-filled root moulds which have been cemented with cryptocrystalline calcite to form rigid root casts.
Rhizocretions
The term rhizocretion is used here to describe pedodiagenetic accumulations of mineral matter around roots. Accumulation, usually accompanied by cementation, may occur during life or after death of plant roots. Figure 5b shows a fibrous root system surrounded by a cemented sheath of aeolianite material. The aeolianite host material is capped by an extensive horizontal micritic crust. Figure 5a is a general view of the surrounding area. Rhizoliths are abundant in this material which, at first glance, is surprising since there are no visible signs of vegetation on the land surface (Fig. 5a). Yet the rhizocretion shown in Fig. 5b contains intact or only partially decayed roots. Thus, the aerial roots of former dune vege tation have decomposed or have been eroded, whereas their subterranean root systems remain. Figure 5c is a transverse section through the rhizocretion shown in Fig. 5b. The calcareous sheath or envelope (cf Calvet et a!., 1975) is about I mm thick and consists of micritized (rhizomicritized) and micrite-cemented carbonate grains. Fibrous roots, together with weakly cemented sand-sized grains of forams, molluscs, coralline algae and lithoclasts, occupy the central or axial part of the rhizocretion. The inner wall of the calcareous sheath in Fig. 5c is irregular in form, the embayments marking the positions of earlier roots which have now decayed. The younger roots, which are still preserved, have occupied root moulds of earlier roots.
Root petrifactions
The American Geological Institute (1972) defined petrifaction as 'a process of fossilization whereby organic matter is converted into a stony substance
by the infiltration of water containing dissolved inorganic matter (e.g. calcium carbonate, silica) which replace3 the original organic materials, some times retaining the original structure'. For the purpose of this section, the term petrifaction will be extended here to include mineral precipitation within voids of organic structures and deposition of minerals on (encrustation) and within (impregnation) living and dead tissues. Thus root petrifaction is a process which involves replacement, impregnation, encrustation and void-filling of organic matter by mineral matter without total loss of root anatomical features. In calcrete deposits, calcite is the dominant mineral which preserves plant morphology. Fig. 5. (a) White surface-exposed laminar calcrete hard pan (micritic crust) with irregular microtopography. Underlying Pleistocene aeolianite contains rhizocretions (arrow). Note absence of vegetation on surface of hard pan. Location: Punta de Sa Cals, Ibiza. (b) Detail of arrowed rhizocretion shown in a. Fibrous root sur rounded by cemented cylinder of Pleistocene aeolianite. (c) Transverse section through rhizocretion shown in b. Cemented cylinder is composed of calcareous wind deposited sand grains; cement is cryptocrystalline calcite. Roots (outlined in ink) and weakly cemented sand grains within cylinder. Note: rhizocretion will become a root tubule if roots decay completely. Thin section, PPL. (d) Rhizolith lacking fine cellular details. Note vague concentric arrangement of micritic layers. Black areas are voids. Thin section, crossed polarized light. Location: Cala Bassa, Ibiza. (e) Oblique sections through bifur cating calcified roots. Epidermis and cortical cells have been calcified. Clotted micrite matrix is composed of calcified faecal pellets. Thin section, PPL. Calcrete hard pan of Pleistocene age. Location: as d. (f) Detail of upper rhizolith shown in e. Voids (V) mark positions of former vascular systems. Micritic matrix, which has a vermicular texture, contains calcified root hairs (arrow). Thin section, PPL. Fig. 6. (a) Partially decomposed Recent root showing collapsed, parenchymatous cells of cortex. Cell contents have been lost. Fine filaments on surface are fungal hyphae; spheres are conidia. SEM. Location: Vandellos, NE Spain; chalky horizon of Pleistocene calcrete. (b) Epidermis (lower part) and partly decayed cortical cells of Recent root. Note that cell walls have decom posed. SEM. Location: as a. (c) Calcified cortex of root. Isodiametric calcite grains pseudomorph parenchymatous cells. SEM. Location: Falset, NE Spain; sheet calcrete horizon of Pleistocene calcrete. (d) Detail of c. Isodia metric calcite grains. Note curved boundaries and intergranular voids. Intergranular voids mark positions of cell walls. SEM. (e) Oblique sections through calcified root. Parenchymatous cells of root cortex have been replaced by calcite. Anatomy of central vascular system has not been preserved. Thin section, PPL. Location: Es Cana, lbiza; Hardpan horizon of Pleistocene (or older) calcrete. (f) Detail of e. Individual calcite grains· of calcified root cortex extinguish as single crystals. Voids (V) of channel shown in e. Thin section, PPL.
156
157
158
Fig. 7. (a) Calcified epidermis of root. Tubular extensions are root hairs. Calcified fungal hyphae and sloughed off calcified root hairs are present in surrounding micritic matrix. SEM. Location: Albacete, SE Spain. (b) Detail of a. Note ridges between cells. SEM. (c) Detail of b. Ridges consist of micron-sized calcite grains. Arrow points to interior of root hair tube. SEM. (d) Detail of central part of c. Fibrillar material (arrow) is cell wall material; composition not determined, possibly cellulose. Micron-sized calcite plates are located between cells in position of middle lamella. SEM. (e) Outer surface of epidermis, including root hair (R). Composition not determined. SEM. Location: top right of a. (f) Detail off Note microporosity resulting from selective loss of cell wall material. SEM.
Fig. 8. (a) Calcified root with long axis lying NE-SW in photomicrograph. SEM. Location (and for Fig. 8b-f): Salou, NE Spain; hardpan of Pleistoce:1e calcrete. (b) Calcified walls of (?) xylem vessels composed of rhombic micrite (micron sized calcite grains). SEM. (c) Outer wall of rhizolith shown in a. Wall is composed of micron-sized calcite grains. Hyphantic nedle-fibre calcite occurs in voids within and around rhizolith. SEM. (d) Calcified parenchymatous cells of root. Note cell contents have not been calcified; cells are empty. SEM. (e) Detail of top right cell in d showing empty cell. Calcite needles (N) show a random distribution. SEM. (f) Detail of inner surface of calcified cell wall shown in e. Calcite consists of micron-sized plates oriented perpendicular to the wall surface. Plates have been overlain partly by a second generation equant micron-sized epitaxial calcite. SEM.
Rhizoliths in terrestrial carbonates
Figure 6a shows collapsed cells of a root cortex. The sample was collected from the transitional horizon of a Pleistocene calcrete developed in an alluvial fan from northeastern Spain. Cell contents of the root have decomposed but the cell walls, which have been attacked by saprophytic fungi, are still present. Figure 6b is part of a root from the same sample but showing a different form of preservation. Here, lignified ground tissues of the outer part of the cortex and epidermis have remained intact, whereas the cell walls have decayed. The state of root preservation at the time of petrifaction determines the morphology of the resulting rhizolith. In Fig. 6c-f the cortical cells have been replaced by calcite. Viewed by light and scanning electron microscopy, these rhizoliths consist of equant calcite crystals, 20-30 f.I.m in diameter. The vascular system in these example> has not been preserved (Figs 5e, f, 6e and f). Petrifaction of cortical cells by calcite is a common, but by no means the commone>t, form of preservation. Petrifaction of cell walls is of greater importance in samples studied here (Figs 7 and 8). In Fig. 7 the rhizolith consists of a petrified epi dermis. Root hairs radiate outwards from the root (Fig. 7a). Calcified root hairs occur as tube> in the surrounding micritic matrix, giving a vermicular or spaghetti-like texture when viewed in thin se::tion (Fig. 5f). Closer inspection of the rhizolith shown in Fig. 7a reveals that micron-sized plates of calcite are present in the position of the middle lamella between adjacent cells (Fig. 7c and d). Similar fabrics have been found in other rhizoliths (Fig. 8). The rhizolith of Fig. 8a is composed of calcified parenchymatous cells and xylem vessels which have been preserved in plate-shaped and rhombic micrite (Fig. 8b-f). The protoplast or cell contents have not been calcified in this sample. The cells are now voids (Fig. 8d and e) or partially filled with needle fibres of calcite (Fig. 8c). Thus, it appears that the middle lamella, originally composed of or containing calcium pectate, is a preferential site for calcification. Thus, cell morphology of the roots is maintained. In thin section, calcite impregnated or replaced middle lamellae replicate the cellular pattern of root cells (Fig. 5d). In Fig. 5d root anatomy has not been preserved in detail but the concentric layering of cells can be discerned. This form of preservation is especially characteristic of rhizoliths which occur within sheet calcrete horizons.
GENESIS OF RHIZOLITHS
The role of plants in biological weathering is well known (Keller & Frederickson, I952). Plant roots accelerate weathering of rocks by exchanging H+ ions from the roots for Ca2+, Mg2+, K+, etc., ions in the rocks. Keller & Frederickson (1952) suggested that the surface chemistry of a plant root could be explained by the Debye-Hiickel double layer system. The root and adhering water film is surrounded by an ionic double layer of which the root has a strong negative charge which is balanced by a surrounding area of positive ions (typically H +) . Thus, the high concentration of H + ions in the diffuse ionic double layer around roots will accelerate weathering of surrounding minerals if the released metal cations are removed from the system. The above outlined mechanism of biological weathering may account for root borings and the breakdown of host materials, but the production of cemented cylinders of calcium carbonate around roots to form rhizocretions and the calcification of root tissues to form root petrifactions require further explanation. Ca2 + ions are not removed from the rhizosphere in environments conducive to the formation of rhizoliths within calcrete profiles. On the contrary, calcium carbonate is the stable mineral phase in such environments. The problem is to explain why this should be so. Gray & Williams (197I) have questioned the assumption that roots increase acidity in the sur rounding soil as a result of excretion of C02 and H+ ions. Some roots take up more anions than cations. Such roots maintain electrical neutrality by passing out HC0-3 ions rather than H + ions. In so doing, the pH of the surrounding soil is raised, rathe.r than lowered. This may trigger precipitation of calcium carbonate around roots, thus leading to the formation of rhizocretions. Johnson (I 967) listed further possible ways of forming rhizocretions. He stated (p. I 54): 'Root sheaths apparently form in one or more of five biochemical ways, dependent upon (I) the presence of organic acids exuded by living plant roots; (2) symbiotic relations between roots and certain soil bacteria; (3) symbiotic relations between roots and certain soil fungi; (4) the presence of some blue green soil algae which have calcium carbonate precipitating bacteria housed in their slime sheaths; (5) calcium exclusion properties of some plants which promote the precipitation of calcium carbon ate outside the root'. Although Johnson admitted 161
162
Rhizoliths in terrestrial carbonates
the possibility that the formation of rhizocretions may result from a combination of these processes, he favoured the first one. Carozzi (1967) described calcite-cemented sand stone around roots of Iroko trees from the Ivary Coast and attributed the cementation to calcareous secretions related to wounds kept unhealed by insect activity. Calvet et al. (1975) suggested that rhizocretions, occurring in Pleistocene aeolianites from Mallorca formed by: (1) progressive root penetration, pro ducing a closer packing of sand grains around the roots; (2) formation of a calcareous envelope (sheath), resulting mainly from the activity of micro organisms, the effects of organic acids and evapo transpiration; and (3) centripetal filling of chalky material following death and decay of the root. Kindle ( 1925, p. 744) suggested that the presence of certain bacteria, or of fungi, on. living roots may constitute the initial factor in the development of root concretions. That micro-organisms are present in and around roots has been noted in this study and elsewhere (Burges, 1958; Alexander, 1961; Gray & Williams, 1971; Russell, 1973; Klappa, 1979a, b). Whether they play an active role in, or are incidental to, the formation of rhizoliths is another question. The suggestions of Kindle (1925), Johnson (1967) and Calvet et al. (1975) are reasonable but not readily proved (Klappa, 1978a, p. 5 14). With respect to root petrifactions, it seems to be more than coincidental that sites for calcification in these structures correspond to naturally occurring calcium-rich layers within plant tissues, notably the middle lamella (organic cement of calcium pectate between cell walls). Thus, a substrate or template control appears to govern the form of preservation in petrified samples examined in this study (Figs 7 and 8). A similar control has been found for the formation of some calcified filamentous micro organisms (algae, fungi, actinomycetes; see Klappa, 1979a) which, together with rhizoliths, are common
and characteristic biogenetic carbonate structures of pedodiagenetic calcretes (Klappa, 1978c). The formation of tubules and rhizocretions, on the other hand, involves dissolution of mineral components within the rhizosphere and reprecipitation of some or all of the dissolved minerals around the root (Figs 3, 4, 5a-c) and/or introduction of CaC03-rich solutions from elsewhere. This process may take place during the life of the root (Fig. 3d) (Kindle, 1925) or during its decay. Hoffmeister & Multer (1965), in their description of an inferred sequence of events which led to the formation of a 'fossil mangrove reef rock' from Florida, suggested that the slow decomposition of buried root material released C02 which combined with available water, forming H2C03• This action dissolved calcite and produced carbonate-bearing solutions which percolated through pore spaces of the calcareous-quartzitic sand substrate. Reprecipi tation of the CaC03 in the sand immediately adjacent to the rotting root cemented quartz grains together, forming a hard cylindrical rim around the root (Hoffmeister & Multer, 1965, p. 851). They envisaged that the hard cylindrical rim (equivalent to root tubule of this study)' . . . slowly grew outward as the action continued and resulted in a coating con siderably thicker than the original periderm. At the same time continued decay of the organic material, surrounded by the hard but still porous rim, pro vided an environment for calcification within the woody structure and for replacement of the tissue itself by CaC03'. In some root tubules examined in this study the tubular wall is composed of cryptocrystalline calcite with virtually no porosity. Because the outer wall of the tubule forms an effective barrier between the decaying root within the tubule and the outer surface of the tubule itself, a change in the chemical micro environment as a result of decaying root organic matter would be unable to cause further buildup of CaC03 on the outer wall.
Fig. 9. (a) Vertical section through calcrete hardpan. Tubular voids are root moulds. Some moulds contain cylinders composed of microcrystalline calcite to give an alveolar texture. Polished slab. Cala Bassa, Ibiza; calcrete hardpan of Pleistocene age. (b) Alveolar texture consisting of ramifying micritic walls. White areas are root moulds. Black peloids (arrows), composed of cryptocrystalline calcite, are interpreted as calcified faecal pellets. Thin section, PPL. Same sample as a. (c) Detail of b showing transverse sections (T) and longitudinal sections (L) through micritic cylinders. Thin section, PPL. (d) Detail of micritic wall shown in c. Wall is composed of micron-sized calcite needles (cf tangential needle fibres; James, 1972). Arrow points to columnar calcite crystals which have their long axes perpendicular to needle calcite walls. Thin section, polarizers at 45°. (e) Cylinders composed of calcite needles oriented tangentially with respect to the surface of the cylinders but random with respect to the long axes of the cylinders. Tubular voids (arrows) are root moulds. SEM. Location: Tarragona, NE Spain. (f) Detail of e. Needle calcite wall has a banded fibrous fabric when viewed in transverse section and a hyphantic fibrous fabric when viewed in longitudinal section. Equant microcrystalline calcite precipitated on needle wall (arrow) leads to thickening of wall. SEM.
163
Colin F. Klappa
At many field outcrops in the study area, root tubules were observed around living roots (Fig. 3d). It was noted that root tubules started at some distance (mm) from the root-sediment interface on the outer periphery of the rhizosphere. Laboratory examination of these samples indicated that the outer diameter of the tubule is slightly greater than the maximum extent of root hair penetration into the sediment. In other words, the outer diameter of the tubule is approximately at, or just beyond, the rhizosphere. With decay of the root hairs the rhizo sphere decreases and precipitation of CaC03 can occur near the root surface since C02 evolution from root hairs is terminated. New root hairs will grow lower down the root as the growing root tip pene trates further into the sediment. Thus, formation of the tubules is viewed as a centripetal process, similar to that described by Calve! et al. (1975). However, the significant difference between the process envisaged here and that of Calve! et at. (1975) is that centripetal tubule formation can occur around living roots as well as around decaying roots. The root tubule, once formed, provides a conduit for downward percolating solutions. When the root within the tubule can no longer maintain viability (perhaps because of the tubule itself), C02 levels are reduced. The root, following its death, begins to decay and releases proteins and sugars which in crease alkalinity of the ambient rhizosphere. If precipitation of calcite takes place within or on the decaying root, some anatomical features of the root may be preserved; the end result being a petrified root (root petrifaction) surrounded by a root tubule.
ASSOCIATED FEATURES
Field and petrographic observations indicate that roots of higher plants are partly or totally respon sible for numerous and characteristic features of calcretes and calcretized aeolianites (Kiappa, 1978b). Apart from rhizoliths themselves, roots are pri marily responsible for the formation of vertically elongate glaebules or concretionary soil structures (Fig. 3b), sheet calcrete layers, brecciation textures and the formation of some tepee structures (Klappa, 1980), brittle fracture, channel and mouldic poro sity, and alveolar textures (cf. Esteban, 1974). Roots, together with symbiotic fungi, also are responsible for the enigmatic structure Micro codium as demonstrated by Klappa (1978a).
Calcretization involves modification or oblitera tion of precursor fabrics, textures and structures in a given host material and the production of new fabrics, textures and structures. Roots modify -and destroy rocks (e.g. 'rhizomicritization', results from dissolution of silt-sized or larger carbonate grains and/or cement and reprecipitation of released CaCOa as microcrystalline calcite). Roots also become calcified. Thus, roots are fundamental contributors to pedodia-genetic processes and resulting products of calcretization, the products being rhizoliths and related features as outlined above.
A CKNOWLEDGMENTS
This study evolved from part of a Ph. D. dissertation on calcretes from coastal regions of the western Mediterranean, completed at the University of Liverpool under the advisorship of Robin G. C. Bathurst. I wish to express my gratitude to him, and to Mateu Esteban, Francese Calvet and Lluis Pomar for their encouragement and stimulating dis cussion during the course of this work. I also thank s.- T. Williams for providing SEM facilities in the: Department of Botany, University of Liverpool andl C. J. Veltkamp for technical assistance. The manuscript was reviewed critically by J. A. D. Dickson, M. Esteban, J. D. Hudson and N. P. James; I am grateful for their comments and sug gestions which have improved this contribution considerably. Financial support for field and laboratory studies was provided by the Natural Environment Research Council (Research Studentship Grant No. GT4/75/ GS/131) which is gratefully acknowledged.
REFERENCES
M. (1961) Introduction to Soil Microbiology. Wiley & Sons, New York. 472 pp. AMERICAN GEOLOGICAL INSTITUTE t)972) Glossary of Geology. Washington, D.C. 805 pp. AMIEL, A.J. (1975) Progressive pedogenesis of eolianite sandstone. J. sedim. Petrol. 45, 513-519. B AL, L. (1975) Carbonate in soil: a theoretical con-· sideration on, and proposal for its fabric analysis. II. Crystal tubes, intercalary crystals, K fabric. Neth. J. Agric. Sci. 23, 163-176. BoYD, D.W. (1975) False or misleading traces. In: -The Study of Trace Fossils (Ed. by R. W. Frey), pp. 65-83. Springer-Verlag, New York. ALEXANDER,
164
Rhizoliths in terrestrial carbonates
C.J.R. (1975) Petrology of palaeosols and other terrestrial sediments on Aldabra, Western Indian Ocean. Phil. Trans. R. Soc. Land. 273, 1-32. BREWER, R. (1964) Fabric and Mineral Analysis of Soils. Wiley & Sons, New York. 470 pp. BROMLEY, R.G., CURRAN, H.A., FREY, R.W., GUTSCHICK, R.C. & SuTTER, L.J. (1975) Problems in interpreting unusually large burrows. In: The Study of Trace Fossils (Ed. by R. W. Frey), pp. 351-376. Springer Verlag, New York. BuRGES, A. (1958) Micro-Organisms in the Soil. Hutchin son, London. 188 pp. CALVET, F., PoMAR, L. & EsTEBAN, M. (1975) Las Rizocreciones del Pleistoceno de Mallorca. lnst. Invest. Ceo!. Univ. Barcelona, 30, 35-60. CAROZZI, A.V. (1967) Recent calcite-cemented sandstone generated by the Equatorial tree Iroko (Chiorophora cxcelsa), Daloa, Ivory Coast. J. sedim. Petrol. 37, 597-600. DURAND, J.H. (1949) Essai de nomenclature des croutes. Bull. Soc. Sci. Nautrelles Tunisie, 3 4, 141-142. EsTEBAN, M. (1974) Caliche textures and Microcodium. Boll. Soc. geol. !tal. 92, Suppl., 1973, 105-125. ESTEBAN, M. (1976) Vadose pisolite and caliche. Bull. Am. Ass. Petrol. Ceo!. 60,2048-2057. FAIRBRIDGE, R.W. & TEICHERT, C. (1953) Soil horizons and marine bands in the coastal limestones of Western Australia. J. Proc. R. Soc. New South Wales, 86, 68-87. GLENNIE, K.W. & EvAMY, B.B. (1968) Dikaka: plants and plant-root structures associated with aeolian sand. Palaeogeog. Palaeoclimat. Palaeoecol. 4, 78-87. GRAY, T.R.G. & WILLIAMS, S.T. (1971) Soil Micro organisms. Oliver & Boyd, Edinburgh. 240 pp. HARRISON, R.S. (1977) Caliche profiles: indicators of near-surface subaerial diagenesis, Barbados, West Indies. Bull. Can. Petrol. Ceo!. 25, 123-173. HOFFMEISTER, J.E. & MULTER, H.G. (1965) Fossil mangrove reef of Key Biscayne, Florida. Bull. geol. Soc. Am. 16,845-852. JAMES, N.P. (1972) Holocene and Pleistocene calcareous crust (caliche) profiles: criteria for subaerial exposure. J. sedim. Petrol. 42, 817-836. JOHNSON, D.L. (1967) Caliche on the Channel Islands. Miner. Jnf . Calif Div. Mines Ceo!. 20, 151-158. KELLER, W.D. & FREDERICKSON, A.F. (1952) Role of plants and colloidal acids in the mechanism of weathering. Am. J. Sci. 250, 594-608. KINDLE, E.M. (1923) Range and distribution of certain types of Canadian Pleistocene concretions. Bull. geol. Soc. Am. 34,609-648. KrNDLE, E.M. (1925) A note on Rhizocretions. J. Ceo!. 33,744-746. KLAPPA, C.F. (1978a) Biolithogenesis of Microcodium: elucidation. Sedimentology, 25, 489-522. KLAPPA, C.F. (1978b) Morphology, composition and genesis of Quaternary calcretes from the western BRAITHWAITE,
-
Mediterranean: a petrographic approach. Unpublished Ph.D. Thesis, University of Liverpool, 446 pp. KLAPPA, C.F. (1978c) Biogenetic carbonate structures in Quaternary calcretes, western Mediterranean. lOth Int. Sediment. Congr., Jerusalem, 365 (abstract). KLAPPA, C.F. (1979a) Calcified filaments in Quaternary calcretes: organa-mineral interactions in the subaerial vadose environment. J. sedim. Petrol. 49,955-968. KLAPPA, C.F. (1979b) Calcification and significance of soil filamentous micro-organisms in Quaternary calcretes, eastern Spain. Bull. Am. Ass. Petrol. Ceo!. 63,480. KLAPPA, C.F. (1980) Brecciation textures and tepee structures in Quaternary calcrete (caliche) profiles from eastern Spain: the plant factor in their forma tion. Ceo!. J. IS, 81-89. NoRTHROP, J.I. (1890) Notes on the geology of the Bahamas. Trans. N.Y. Acad. Sci. 10, 4-22. PERKINS, R.D. (1977) Depositional Framework of Pleisto cene Rocks in south Florida. Mem. geol. Soc. Am. 147, 131-198. PLAZIAT, J.C. (1971) Racines ou terriers? Criteres de distinction a partir de quelques exemples du Tertiaire continental et littoral du Bassin de Paris et du Midi de Ia France. Consequences paleographiques. Bull. Soc. geol. Fr. ser 7, 13, 195-203. READ, J.F. (1974) Calcrete deposits and Quaternary sediments, Edel Province, Western Australia. Mem. Am. Ass. Petrol. Ceo!. 22, 250-282. RussELL, E.W. (1973) Soil Conditions and Plant Growth, lOth ed. Longman, London. 849 pp. RuTH, N. ST J. (1927) Replacement vs impregnation in petrified wood. Econ. Ceo!. 22, 729-739. SARJEANT, W.A.S. (I 975) Plant trace fossils. In: The Study of Trace Fossils (Ed. by R. W. Frey), pp. 163179. Springer-Verlag, New York. SHERMAN, G.D. & IKAWA, H. (1958) Calcareous con cretions and sheets in soils near South Point, Hawaii. Pacific Sci. 12, 255-257. STEINEN, R.P. (1974) Phreatic and vadose diagenetic modification of Pleistocene limestone: Petrographic observations from sub-surface of Barbados, West Indies. Bull. Am. Ass. Petrol. Ceo!. 58, 1008-1024. STRAKHOV, N.H. (1970) Principles of Lithogenesis, vol. 3. Oliver & Boyd, Edinburgh. 577 pp. TEICHERT, C. (1950) Late Quaternary sea-level changes at Rottnest Island, Western Australia. Proc. R. Soc. Victoria, 59, 63-79. VALETON, I. (1971) Tubular fossils in the bauxites and the underlying sediments of Surinam and Guyana. Geologie Mijnb. 50, 733-741. WARD, W.C. (1975) Petrology and diagenesis of carbon ate Eolianites of northeastern Yucatan, Mexico. In: Studies in Geology, 2. Belize Shelf: Carbonate Sedi ments, Clastic Sediments and Ecology, pp. 500-571. Am. Ass. Petrol. Geol.
(Manuscript received 28 September 1979; revision received 18 February 1980)
165
Reprinted from
Sedimentology
(1980) 27 651-660
Calcrete profiles in the Eyam Limestone (Carboniferous) of Derbyshire: petrology and regional significance
A. E. A D A M S Department of Geology, University of Manchester, Manchester M13 9PL
ABSTRACT Calcrete profiles (caliche) have been recognized in the Eyam Limestone from the Lower Carboniferous in the area around Monyash, Derbyshire. They occur at the top of the flank facies surrounding carbonate-mud buildups ('knoll reefs'). Four units make up the complete profile. These are from base to top: (I) grain-supported sediment with rhizocretions, (2) matrix-supported sediment with alveolar texture, (3) pelleted calcrete, (4) laminar calcrete. Commonly one or more units are missing from the profile. Calcretes indicate subaerial exposure. The carbonate buildups of the Eyam Limestone were completely exposed soon after deposition, requiring a fall in sea-level probably in excess of 10m. This discovery demands a review of previous regional palaeoenvironmental studies.
INTRODUCTION
Calcretes (Caliche) form through the accumulation and re-distribution of carbonate in soil-profiles and indicate subaerial weathering and unconformity in otherwise marine limestone sequences (Reeves, 1970; Read, 1976). During the last few years there has been an increasing volume of literature on the subject of Pleistocene and Recent calcrete deposits. Authors such as Multer & Hoffmeister (1968), James (1972) and Read (1976) have emphasized the importance of recognizing calcretes, and in particu lar distinguishing them from superficially similar algal stromatolite deposits which indicate tidal or shallow subtidal environments. Read (1976) noted that there are few published descriptions of ancient calcrete profiles. In particu lar, there is a shortage of information on details of calcrete microtexture which may enable calcretes to be identified even where the more obvious large-scale features are absent. Palaeozoic examples described in the literature include those by Harrison & Steinen (1978) and Walls, Harris & Nunan (1975). The aim of this contribution, there fore, is to describe an example of an ancient calcrete profile with particular emphasis on microtextures, Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
and to discuss its implications for the regional geology. GEO L O G IC A L SETTIN G
The calcrete profiles discussed here occur in the Eyam Limestone of the area around Monyash, Derbyshire (Fig. 1). The Eyam Limestone occurs in the topmost part of the Carboniferous Limestone and has been assigned to the P2 zone on the basis of bivalve-goniatite faunas. Previous work on the Eyam Limestone includes stratigraphical studies (e.g. Shirley, 1959) and broad scale palaeoenvironmental studies (Brown, 1973). Biggins (1969) briefly des cribed the geology of the P2 zone in Lathkill Dale while making a comparison with coeval Carboni ferous Limestone at High Tor near Matlock 15 km to the southwest. In the Monyash area the Eyam Limestone conformably overlies pale-coloured, fossiliferous shelf limestones of the Monsal Dale Beds. The Eyam Limestone comprises up to 30 m of well-bedded, dark grey, fine-grained limestones. Included in the sequence are a number of lenticular bodies of pale-coloured, massive, fine-grained 167
A. E. Adams
KEY
lllllllJ
Black mudstone (Upper Carboniferous) Unconformity Basin fill facies Buildup facies
]
Eyam Limestone (0 to 30m), p2 zone
D
Monsal Dale Beds (base not seen in Monyash area), D2 zone
""'
Fault
�
Old quarry
Fig. 1.
Map showing distribution of principal rock-types and location of quarry exposures, Lathkill Dale. Grid line numbers refer to 1 km squares in National Grid Reference square SK.
limestone each surrounded by coarse crinoidal lime stones which exhibit depositional dips away from the massive core. These features are usually called 'knoll-reefs' although the author prefers the more general term carbonate-mud buildup since many of the attributes of reefs as reviewed by Braithwaite (1973) and Heckel (1974), such as wave resistance and ecological zonation, have not been demonstrated. The terminology of such structures and their origin has caused much controversy and the subject is reviewed by Wilson (1975). Following general usage, the massive fine-grained limestones of the buildup are here called the buildup-core facies, the coarse crinoidal limestones, the buildup-flank facies and the bedded strata above the buildups, the basin fill facies. The buildups generally occur at the base of the Eyam Limestone and rest directly on the Monsal Dale Beds, but a few occur higher in the sequence. In the area studied, higher formations have generally
been removed by erosion although in Monyash village there is a small patch of shale attributed to the: Upper Carboniferous (Fig. 1). PETRO LO GY OF THE CA LCRETES General description
The calcretes occur at the junction between the: coarse, crinoidal buildup-flank facies and the dark-· coloured fine-grained limestones of the basin-fill facies. This contact is well exposed in and around Ricklow Quarry near the head of Lathkill Dale (Fig. 1). The sediments on which the calcretes have developed are bioclastic grainstones and packstones (classification of Dunham, 1962) in which the bio .. clasts are crinoid plates and ossicles, together with fragments of brachiopods and bryozoans. The calcretes are characterized by a series of 168
Calcrete profiles in Eyam Limestone
structures which always occur in the same strati graphical order although not all profiles show the complete range of structures. The four units identi fied are, from base to top: (I) grain-supported sediment with rhizocretions, (2) matrix-supported sediment with alveolar texture, (3) pelleted calcrete, (4) laminar calcrete. These units are described in detail and the terminology explained below. The thickest calcrete profile observed is 50 em, but more usually they are 10-20 em thick. They are much thinner than many of the Quaternary profiles des cribed in the literature (see, for example, Read, 1974, 1976). A typical complete profile is illustrated in Fig. 2. ;;: '
c
� "
carbonate mudstone
]]
. :I
em
20 30
laminar calc rete
pelleted calcrete matrix·- supported sediment with alveolar texture
�
"' ·;:;
grain- supported sediment with rhizocretions
.E -" c
0 :;:: ' a.
0
�
3! "3 ..c
�
0"
'--0
"-0
0
o,--,
�
�
0 ,-..,
0
�
alveolar texture
·.•
calcrete pellets
# 0
host packstone or grainstone
By analogy with Quaternary examples, the tube structures are interpreted as coated plant roots. Perkins ( 1977) noted that the precipitation of fine grained calcite around small root tubules in Ple.isto cene calcretes from Florida may be so extensive as to produce a 'root-rock'. He called such structures root tube calcifications, while noting that similar structures described by other workers had been given different names. Harrison & Steinen ( 1978) figure 'root voids' from dense pelleted micrites in Recent and ancient calcretes which are similar structures to those described here. The nomenclature of plant root structures in sediments has been reviewed by Klappa ( 1978). He used the term 'rhizoliths' to include all organo sedimentary structures produced by roots. This would include simple root moulds and root casts as well as root petrifications. Klappa used the term 'rhizo cretion' for concretionary cemented coatings around living or decaying roots and described examples from Quaternary calcretes of the western Mediterranean. They comprise sheaths of micritized and micrite cemented carbonate grains around roots. Although the Carboniferous structures described here have no remaining traces of organic matter, they are very similar to the Quaternary structures described by Klappa and are thus interpreted as rhizocretions.
rhizocretions crinoid plates
Matrix-supported sediment with alveolar texture
'-f-.. shell fragmeuts
The second unit of the calcrete profile contains a higher proportion of fine-grained matrix than does the unit beneath. It is matrix-supported and has a transitional contact with the grain supported sedi ments below. This unit displays a characteristic texture comprising irregular-shaped but approxi mately equidimensional pores filled with sparry calcite separated by a network of interconnecting walls of micrite (Figs 4 and 5). The diameter of the pores varies from 0· 1 to 0·5 mm and the width of the walls from 0·01 to 0·1 mm. A similar structure was figured by Esteban ( 1974) in Pleistocene calcretes from Spain and named 'alveolar texture'. Esteban noted that the texture is apparently exclusive to calcrete and suggested that it might be formed by the disso lution of the interiors of pisoliths. In a later publica tion, however, Esteban stated that alveolar texture is 'probably related to rhizocretion fabrics' (Esteban, 1976). A comparable texture is described and illustrated by Braithwaite ( 1975) from Quaternary terrestrial sediments on the Indian Ocean Island of Aldabra. He
Fig. 2. Generalized complete calcrete profile· on top of buildup-flank facies� Eyam Limestone, Ricklow Quatry.
Grain-supported sediment with rhizocretions
The lowermost unit of the calcrete profile has a gradational contact with the host packstone or grainstone. The principal features of this unit, as seen in thin section, are thin micritic coatings on bioclasts and circular to irregular spar-filled cavities, having diameters of 0·25-o:s mm, surrounded by walls of fine-grained carbonate up to 0·2 mm thick (Fig. 3). The walls often.show an irregular concentric lamin ation and may include small aggregates of micrite (pellets). In this case they have a clotted appearance. These spar-filled cavities are sections through irregu lar tubes. In hand specimen these tubes impart a brown stain to an otherwise pale grey limestone. The micritic coatings are often continuous between individual sediment grains and have the appearance of binding the sediment together. Thus they formed after final deposition of the sediment grains. 169
A. E. Adams
Photomicrograph showing irregular spar-filled voids separated by a network of micrite walls. Interpreted as alveolar texture formed by the deposition of fine-grained carbonate around decaying rootlets. Scale bar represents 0· 5 mm.
Fig. 3.
Fig. 4. Photomicrograph showing irregular spar-filled voids surrounded by coatings of vaguely laminated micrite. Interpreted as rhizocretions (see text). Scale bar represents 0· 5 mm.
170
Calcrete profiles in Eyam Limestone
Fig. 5. Photomicrograph showing well-developed alveolar texture from immediately below laminar unit. Scale bar represents 0·5 mm.
called it 'vesicular structure' and suggested that it was caused by the formation of large numbers of gas bubbles during alternate wetting and drying of the sediments. Harrison & Steinen (1978) described arcuate sheaths of micrite surrounding a network of voids in Recent and Carboniferous calcretes. In the latter case voids are cement filled. These sheaths were interpreted as cement precipitates on the surfaces of rootlets and root hairs which originally occupied the voids. These were thus called 'arcuate branching root sheaths'. They are similar to the structures present in the Eyam Limestone, although the arcuate form of the micrite sheaths is not so evident. The most detailed description of alveolar texture is that of Klappa (1978) from Quaternary calcretes from the western Mediterranean. He states, 'alveolar texture consists basically of a number of cylindrical to irregular pores, which may or may not be filled with calcite cement, separated by a network of anastomosing micritic walls'. The pores have similar dimensions to those in the Carboniferous structures described here. In the Quaternary examples the walls consist of calcite needle fibres. Klappa ob served that whereas decaying roots in Quaternary calcretes might become coated with needle micrite, Jiving roots would always lack such a coating. He
therefore suggested that calcite precipitation was caused by a change in micro-environment produced by root decomposition. Partial collapse of decom posing roots followed by calcite precipitation would lead to the irregular structure seen in alveolar texture. Alveolar texture is thus a special type of rhizocretion fabric. Simple rhizocretions as described from the under lying unit, also occur occasionally in this unit. In some cases a trace of alveolar texture can be seen within such rhizocretions. Pelleted calcrete
The pelleted calcrete has a transitional contact with underlying units. In hand specimen it appears as pale grey or cream, structureless fine-grained sedi ment. In thin section a clotted texture is visible resulting from the aggregation of irregular micritic pellets up to 0·3 mm in diameter. Micritized shell fragments and unfilled or spar-filled voids are also present. Calcrete pellets are common components of many Quaternary profiles (Read, 1976). They may have formed by the alteration of skeletal grains during calcrete formation (James, 1972) or as cement 171
A. E. Adams
aggregates (Kiappa, 1978). The pelleted calcrete in the Eyam Limestone is the least common unit of the calcrete profile and where present forms a layer up to 1·5 em thick. Laminar calcrete
The laminar unit of calcrete profiles may not be volumetrically the most important part of the profile but it has generally drawn the most discussion, partly because it is readily visible in field studies. and because of its similarity to stromatolites. In Quater nary calcretes the laminar horizon commonly under lies loose soils (James, 1972; Multer & Hoffmeister, 1968; Read 1974, 1976). In the Eyam Limestone described here, the laminar unit rests sharply on the underlying units. This sharp contact is in contrast to the transitional contacts between other units of the profile. The laminar unit is up to 3 em thick and is a dense pale to dark grey micritic limestone. In quarry exposures the laminar unit can be seen. to be discontinuous, apparently filling slight depressions in the underlying sediment surface. The deposit is finely laminated and the laminae are undulating. As the unit thins laminae are successively cut out giving an impression of uncon formity with overlap (Fig. 6). The laminae average about 0·4 mm in thickness and have slightly un dulating boundaries. As with many Quaternary
calcretes, the laminae .in the calcretes described here are texturally similar and thus much less obvious in thin section than in hand specimen. They are differen tiated by variations in pigment content (Read, 1976). Few workers have considered the origin of laminar calcretes although Klappa (1979) has shown that some laminar calcretes in the Quaternary of the western Mediterranean may form through the activities of successive growths of lichens. In many places alveolar texture is well developed in the laminar unit. Flattened rhizocretions and other spar-filled voids are also present. Klappa (1978) records the presence of alveolar textures within laminar horizons in Quaternary calcretes from the western Mediterranean. Other vadose diagenetic features
Since calcrete profiles form above the water table, diagenetic features such as vadose internal sediments (Dunham, 1969) and dripstone and meniscus ce ments might be expected. Such features are recorded from Quaternary calcretes by Perkins (1977) and Klappa (1978) and from a Jurassic example by Bernoulli & Wagner (1971). It is difficult to demonstrate how much of the cement in the profiles from the Eyam Limestone was precipitated in the vadose zone. Firstly, much of the matrix is micritic sediment and sparry calcite is
Fig. 6. Polished sample of laminated unit showing irregular nature of laminae and microunconformities with overlap. Interpreted as laminar calcrete. Scale bar in em. 172
Calcrete profiles in Eyam Limestone
restricted to sheltered areas such as the undersides of shell fragments. This does not necessarily indicate vadose cementation. Secondly, in grainstones much of the cement is in the form of syntaxial overgrowths on crinoid fragments. Any preferential downwards growth of overgrowths during cementation in the vadose zone may have been obscured by earlier or later periods of cementation in the phreatic zone. In the Eyam Limestone downwards-thickening over growths, if present, are indistinguishable using a petrographic microscope even with stained sections. However, in a few cases there are examples of coarse prismatic calcite cements to be seen on the under surfaces of shell fragments, even where they are overlain by sparry calcite (Fig. 7). Such cements may have formed in the meteoric vadose zone. A possible example of vadose sediment is also illustrated by Fig. 7. Sediment in the cavity was formed after the sparry calcite described above (and so post-dates vadose cementation) and is plastered on the roof of the cavity as well as at the base. Perkins (1977) illustrates a similar feature in vadose-altered Quaternary limestones from Florida. He suggested that the centre space of the void, now occupied by blocky spar, was occupied by an air bubble formed
during desiccation and this caused sediment to be plastered on the roof of the cavity. Discussion
As previously mentioned, many workers have emphasized the importance of recognizing calcretes and distinguishing them from algal stromatolites. It is relevant therefore, to list briefly the criteria which aid identification of the calcrete in this particular example. (1) The presence of features exclusive to calcretes such as rhizocretions and alveolar texture. (2) The presence of distinctive sediment types occurring in a particular order, characteristic of Quaternary calcrete profiles. (3) The presence of 'clotted' textures in the pelle ted horizon, formed by the alteration of pre-existing sediment or by the aggregation of small cement crystals. (4) The nature of the laminar horizon-it possesses many of the characteristics of laminar calcretes as distinct from algal stromatolites as noted by Read (1976), for example, (i) laminations caused by differential pigmentation rather than by
Fig. 7. Photomicrograph showing possible examples of vadose cement and sediment (see text for discussion). Shell fragments are partly silicified. Scale bar represents 0·5 mm.
173
A. E. Adams
significant textural variations, (ii) laminar horizons discontinuous, tending to fill depressions in under lying surface rather than thickening over highs, (iii) presence of microunconformities within laminar zone and (iv) absence of bioturbation. It is also relevant to discuss differences between the profile in the Eyam Limestone and other calcrete profiles described in the literature. The smaller thickness of the profile has already been mentioned. Profile thickness is related, in part, to the length of time available for development, but the ,Processes occurring within calcretes are complex and as yet poorly understood (Kiappa, I978). Climate may also be an important factor. Robbin & Stipp (1979) used radiocarbon dating to determine the age of laminated crusts from Florida Keys. Calculated rates of accumulation varied from I cm/2000 years to I em/ 4000 years. The thickest-known development of the laminar zone in the Eyam Limestone is 3 em perhaps suggesting that subaerial exposure lasted at least 6000 years. Many Quaternary and ancient profiles show inter digitation of sediment types suggesting repeated profile development (Read, I976). The Carbonifer ous profile described here shows no such features and this is perhaps an indication that there was a single fairly short -lived episode of subaerial exposure. Although the distinction made between vadose pisoliths and calcrete pisoliths by Esteban (I976), has meant that pisoliths might not be as diagnostic of calcretes as was once thought, they are nevertheless present in many profiles (James, I972; Read, I974). No such features have been recognized here. In some Quaternary profiles (e.g. Shark Bay, Read, I974) pisoliths occur in loose soils overlying indurated laminar calcretes. Had these been present in the Carboniferous examples they might have been removed by erosion before or during the ensuing marine incursion. The exact sequence of sediment types in both Recent and ancient calcretes is highly varied (Read, I976). The sequence in the Eyam Limestone shows an increasing intensity of alteration towards the top of the profile as would be expected, but the reasons for the order in which the units occur is not clear. Further work on the processes occuring in Recent calcretes is needed.
elusive evidence that theEyamLimestone experienced at least one episode of subaerial exposure during the Lower Carboniferous. Because of the nature of the exposure around Ricklow Quarry it is not Clear whether the original crest of the buildup is now exposed. Thus it cannot be shown whether calcretes formed over the whole buildup or only where they can be seen today, at the base and on the lower flanks of the buildup. However, since the sides of the buildup dip fairly steeply (at up to 30°) it is likely that downslope movement of soils would at least cause thickening of the calcretes on the lower flanks. It is this interpretation which is included on the summary diagram (Fig. 8). At Shark Bay, Western Australia, Read (I974) has described Quaternary soils which are thickest on the flanks of dunes and thin towards dune crests and inter-dune depressions as a result of downslope movements. Since the calcretes in theEyamLimestone occur on the lower flanks of the buildups, and the buildups had a positive relief on the sea-floor during depo sition, the buildups must have stood well above sea A.
B.
Deposition of buildup core and flank facies
Relative drop in sea level, exposure of buildups and development of calcretes
_-
runoff
'""
,,,
C.
Gradual submergence of buildups, deposition of dark, lagoonal
D.
limestones around buildups
Further submergence, re -establishment of normal marine conditions, accumulation of coral and beari limestones
REGIONAL SI GNIFICANCE OF CA LCRETES
Summary diagram illustrating depositional and early diagenetic history of carbonate mud buildups, Eyam Limestone, Lathkill Dale.
Fig. 8.
The calcrete profiles described here present con174
Calcrete profiles in Eyam Limestone
level during formation of the calcretes. (Fig. 8). Today the buildups have a relief of 5-10 m above the base of the surrounding buildup-flank sediments. Even allowing for some differential compaction of buildup-flank and basin-fill facies over build-up-core facies, a fall of at least several metres in the level of the sea relative to the land would have been neces sary. Wilson (1975) in his summary on the origin of Carboniferous carbonate-mud buildups noted that many geologists believe such structures accumulated below wave-base, because of their fine-grained nature and Jack of any wave-washed or sorted talus. Furthermore since calcareous algae are generally absent from such structures, including the build-up core and flank facies of the Eyam Limestone, although abundant in many other Lower Carboni ferous limestones, it is likely that the buildups accumulated below the photic zone. Even making allowance for muddy waters reduc ing the depth of the photic zone, and a sheltered sea in which normal wave-base was high, it seems likely that a fall of sea-level much greater than the minimum of 5-10 m would have been required to bring the base of the buildups above sea-level. Laminated crusts have been described from the older D zone limestones of Derbyshire by Walkden (1974). These crusts, also interpreted as calcretes, are associated with palaeokarstic mammillated surfaces and are overlain by clay beds interpreted by Walkden as weathered volcanic ashes. Palaeokarstic surfaces associated with clay beds have not been recognized in the sections described here. Walkden estimated times of 30,000-100,000 years to be necessary for the development of the karstic surfaces and it is possible that the Eyam Limestone was not exposed sufficiently long for such features to develop. At Ricklow Quarry calcretes may be overlain directly by further lime stones without any noticeable parting, in other words they may occur completely within a unit which might be described as a single 'bed'. In sections where the laminar unit is absent, recognition of the calcrete is not possible in the field; rhizocretions and alveolar texture can only be distinguished by detailed sampling and the examination of acetate peels or thin sections under the microscope. Since such studies are lacking in the Carboniferous Limestone of many areas, it is probable that many similar examples exist elsewhere. Environmental syntheses of the Carboniferous Limestone in Derbyshire are few. An attempt has been made by Ford (1977) and he noted the probable transgression which resulted in the deposition of
progressively deeper-water sediment from the mass ive bioclastic Monsal Dale Beds, envisaged as shallow shelf deposits, through the dark well-bedded Eyam Limestone to the Longstone Mudstone which occurs in the area 5 km to the northeast of Monyash. This follows the typical Lower Carboniferous se quence of deepening water sediments envisaged by Ramsbottom (1973). Ford (1977) noted that the buildups may repre sent regressive phases which may be related to local tectonism contemporaneous with deposition rather than to the cyclic transgression concept for the Lower Carboniferous as a whole proposed by Ramsbottom (1973). Whether the period of subaerial exposure described here is only local or can be traced to other areas remains to be seen and will be difficult to show because of the paucity of exposure. Nevertheless a considerable change in sea-level apparently occurred around Monyash. If local changes of this magnitude are to superimposed on a regional cyclic pattern it will be very difficult to apply the new Lower Carbon iferous stratigraphy, demonstrated in other areas by Ramsbottom (1973), to the Peak District. The limestones immediately overlying the cal crete are thinly bedded dark-coloured limestones with a limited fauna. Such limestones have often been called 'basinal facies' in Derbyshire (see Ford, 1968). However, petrographic studies show that these limestones have a biota of gastropods, foraminifera and calcareous algae. It is suggested here that rather than being 'basinal', these limestones formed in restricted stagnant lagoons around the buildups as they became submerged. Fully marine fossils such as corals and brachiopods do not occur in any numbers until the level of the top of the buildups is reached (Fig. 8).
ACKN OWLE D GMENTS
The author would like to thank Mr G. S. Evans for assistance in the field and laboratory, Professor R. G. C. Bathurst for helpful discussions and Dr F. M. Broadhurst for his encouragement, and for critical comments on the manuscript.
REFERENCES
BERNOULLI, D. & WAGNER, C.W. (1971) Subaerial dia
genesis and fossil caliche deposits in the Calcare Massiccio Formation, Lower Jurassic, Central 175
A. E. Adams
Apennines, Italy. Neues Jb. Palaont. Abh. 138, 135149. BIGGINS, D. (1969) The structure, sedimentology and
KLAPPA, C. F. (1978) Morphology, composition and genesis of Quaternary calcretes from the western Mediter ranean: a petrographic approach. Unpublished Ph.D.
Thesis, University of Liverpool.
palaeoecology of a Carboniferous reef knoll at High Tor, Derbyshire. Unpublished Ph.D. thesis, Uni
KLAPPA, C.F. (1979) Lichen stromatolites: Criterion for
versity of London.
subaerial exposure and a mechanism for the formation of laminar calcretes (caliche). J. sedim. Petrol. 49,
BRAITHWAITE, C.J.R. (1973) Reefs: just a problem of semantics. Bull. Am. Ass. Petrol. Geol. 57,1100--1116. BRAITHWAITE, C.J.R. (1975) Petrology of palaeosols and
387-400. MULTER, H.G. & HOFFMEISTER, J.E. (1968) Subaerial
other terrestrial sediments on Aldabra, western Indian Ocean. Phil. Trans. R. Soc. Ser. B, 273, 1-32. BROWN, M.C. (1973) Limestones of the eastern side of the
laminated crusts of Florida Keys.
Bull. geol. Soc. Am.
79, 183-192.
PERKINS, R.D. (1977) Depositional framework of Pleisto
cene rocks in South Florida. In: Quaternary Sedi mentation in South Florida (Ed. by P. Enos and R. D. Perkins). Mem. geol. Soc. Am. 147, 131-198.
Derbyshire outcrop of the Carboniferous Limestone.
Unpublished Ph.D. Thesis, University of Reading. DuNHAM, R.J. (1962) Classification of carbonate rocks according to depositional texture. In: Classification of Carbonate Rocks (Ed. by W. E. Ham). Mem. Am.
RAMSBOTTOM, 'W.H.C. (1973) Transgressions and regres
sions in the Dinantian: A new synthesis of British Dinantian stratigraphy. Proc. Yorks. geol. Soc. 39,
Ass. Petrol. Geo/. 1, 108-121.
DuNHAM, R.J. (1969) Early vadose silt in Townsend Depositional En
567-607. READ, J.F. (1974) Calcrete deposits and Quaternary
vironments in Carbonate Rocks (Ed. by G. M. Friedman). Spec. Pub/. Soc. econ. Paleont. Miner.,
sediments. Edel Province, Shark Bay, Western Australia. In: Evolution and Diagenesis of Quaternary
Tulsa, 14,139-181.
Carbonate Sequences, Shark Bay, Western Australia. Mem. Am. Ass. Petrol. Geo/. 22, 250-282.
Mound (Reef), New Mexico. lri:
EsTEBAN, C.M. (1974) Caliche textures and Microcodium. Bull. Soc. Geol. It. (sup.) 92, 105-125. EsTEBAN, C.M. (1976) Vadose pisolite and caliche. Bull.
READ, J.F. (1976) Calcretes and their distinction from
Stromatolites. In: Stromatolites (Ed. by M. Walter), pp. 55-71. Elsevier Publishing Co., Amsterdam. REEVES, C.C., Jr (1970) Origin, classification and geologic history of caliche on the southern High Plains, Texas and eastern New Mexico. J. Geo/. 78,352-362. RoBBIN, D.M. & STIPP, J.J. U979) Depositional rate of laminated soilstone crusts, Florida Keys. J. sedim.
Am. Ass. Petrol. Geo/. 60, 2048-2057.
FORD, T.D. (1968) The Carboniferous Limestone. In: The Geology of the East Midlands (Ed. by P. C. Sylvester-Bradley and T. D. Ford), pp. 59-79. Leicester University Press. FoRD, T.D. (Ed.) (1977) Limestones and Caves of the Peak District. Geo. Abstracts, Norwich. HARRISON, R.S. & STEINEN, R.P. (1978) Subaerial crusts, caliche profiles, and breccia horizons. Comparison of some Holocene and Mississippian exposure surfaces, Barbados and Kentucky. Bull. geol. Soc. Am. 89,
385-395. HECKEL, P.H. (1974) Carbonate buildups in the geologic
record: a review. ln: L. F. Laporte). Spec.
Reefs in Time and Space (Ed. by Pub/. Soc. econ. Paleont. Miner.,
Tulsa, 18, 90-154. JAMES, N.P. (1972) Holocene and Pleistocene calcareous
crust (caliche) profiles: criteria for subaerial exposure. J. sedim. Petrol. 42, 817-836.
Petrol. 49, 175-180.
SHIRLEY, J. (1959) The Carboniferous Limestone of the Monyash-Wirksworth area, Derbyshire. Q. J. geol. Soc. Land. 114, 411-429.
WALKDEN, G.M. (1974) Paleokarstic surfaces in Upper Visean (Carboniferous) Limestones of the Derbyshire Block, England. J. sedim. Petrol. 44, 1232-1247. WALLS, R.A., HARRIS, W.B. & NUNAN, W.E. (1975) Calcareous crust (caliche) profiles and early subaerial exposure of Carboniferous carbonates, north eastern Kentucky. Sedimentology, 22, 417-440. WILSON, J.L. (1975) Carbonate Facies in Geologic History. Springer-Verlag, New York.
(Manuscript received 8 October 1979; revision received 20 February 1980)
176
Reprinted from Sedimentology (1983) 30159-179
A rendzina from the Lower Carboniferous of South Wales V. PA U L W R IG H T*
Department of Geology, University College, Cardiff
ABSTRACT
A thin calcrete-crust horizon from the Lower Carboniferous Llanelly Formation of South Wales consists of two parts. an upper laminated unit and a lower peloidal unit. The former is interpreted as a subaerial stromatolite and the latter as an A horizon of a palaeosol. Comparisons are made with the A horizons of rendzinas and it is concluded that the calcrete-crust represents a complete rendzina profile. This fossil rendzina contains abundant evidence of a soil fauna in the form of fecal pellets and small burrows.
GEOLOGICAL SETTING
INTRODUCTION
The calcrete-crust horizon has been named the Darrenfelen Pedoderm (Wright, 1981a). A pedoderm is a mappable palaeosol unit which has characteristics and stratigraphic relationships that permit its recog nition in the field (Brewer, Cook & Speight, 1970, p. I 06). It has been found at five localities in the outcrop area of the Llanelly Formation in South Wales. This formation comprises part of the atten uated Lower Carboniferous succession in the north east part of the South Wales coal field (Fig. lA), which consists in the main of shallow, subtidal and peritidal limestones deposited on the northern (land ward) edge of a carbonate shelf which covered much of South Wales (Wright, Raven & Burchette, 1981). The units comprising the sequence are shown in Fig. I (B). The Llanelly Formation is composed of four distinct members (Fig. l C); the Clydach Halt and Gilwern Clay members are floodplain deposits with sheet-flood, stream-flood and high-sinuosity channel sandstones and conglomerates, and claystones with calcrete profiles (Wright, 1982). The Penllwyn Oolite Member is a thin oolitic unit separated from the underlying Cheltenham Limestone Member by an oncolitic grainstone, the Uraloporella Bed, containing replaced aragonite cements and the problematical tubiform microfossil Uraloporella (Wright, 1981c). The Cheltenham Limestone Member consists of a . series of peloidal limestones of lagoonal to supratidal facies-type deposited as a facies mosaic (Wright,
There are now many descriptions from both recent and ancient carbonate sequences of so-called caliche or calcrete crusts. These horizons develop on subaeri ally exposed carbonate sediments and rocks, and studies of Recent and Pleistocene forms (e.g. Multer & Hoffmeister, 1968; James, 1972; Read, 1974; Harrison, 1977) have led to the recognition of many characteristic features which enable similar crusts to be recognized in ancient sequences (e.g. Walkden, 1974; Walls, Harris & Nunan, 1975; Harrison & Steinen, 1978; Somerville, 1979; Adams, 1980; Riding & Wright, 1981 ; Wright, 198l b). Subaerial crusts form in a variety of ways. Some are purely accretion ary, like a subaerial dripstone, others result from purely pedogenic processes, some from the activ ities of lichens (Kiappa, 1979) while others result from the calcification of algal mats (Krumbein & Giele, 1979). This paper aims to document a variety of features which occur in a 'calcrete crust'-like horizon in the Lower Carboniferous of South Wales, using infor mation from soil microscopy. This horizon contains abundant evidence that a soil fauna was active during its formation.
* Present address: Department of Earth Sciences, Open University, Walton Hall, Milton Keynes MK7 6AA, U.K.
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
177
V. P.
Wright
MILLSTONE GRIT (NAM URIAN)
DOW LAIS
� . .
[)
LIMESTONE
0
• CARBONIFEROUS LIMESTONE
LLANELLY FM.
A 0
/
OOLI TE 0
GROUP
0 LOWER LIMESTONE
GILWER N
SHALES CLAY MR. OLD RED SANDSTONE (DEVONIAN)
.
PENLLW Y N 0
B
OOLITE MR.
Sml
LIMESTONE MR. CLYDACH HALT MR.
c Fig. I. Geological setting of the Darrenfelen Pedoderm. (A) position of outcrop area. (B) major stratigraphic subdivisions in the outcrop area. (C) subdivisions of the Llanelly Formation.
1981a), although occasionally containing fining-up ward shoaling units (Wright, 198l b). The calcrete-crust described herein occurs 0.5-1 m below the Uraloporella Bed and has been found at five localities; at the Graig quarry on the eastern side of Gilwern Hill near Abergavenny (British Grid Refer ence S021, 2475, 1250); at the Clydach Halt Lime Works near Brynmawr (S021, 2342, 1261); at Llanelly Quarry near Brynmawr (S021, 2233, 1237); at Craig y Gaer near Brynmawr (S021, 2232, 1328), and Cwar yr Hendre near Tredegar (SOO I, 0995, 1492). Detailed descriptions of these localities have been given in Wright (198l a).
FIELD APPEARANCE
The calcrete crust appears as a thin (2-12 em) light grey to buff, fine-grained fenestral limestone. It is overlain at all localities by a thin green clay (Wright, 1981b, plate 1a) which at Craig y Gaer contains small ( l -3 em diameter) round and subspherical nodules of fine-grained spar identical to calcrete nodules de scribed from elsewhere in the Llanelly Formation (Wright, 1982). In the field two units can be recognized in the crust; first, there is a thin (under 4 em), finely laminated unit which may occur separately or overlie the second type, which consists of a thicker highly 178
Carboniferous rendzina (Fig. 2A). Some of the latter can be traced horizontally for distances of 30 em. Large spar-filled vertical cracks and subhorizontal clay-filled fractures also occur as well as irregular patches of the underlying grainstone (Fig. 2B). In thin section, the unit is seen to have a very variable fabric but is composed essentially of three components, grains, coatings of various types and fenestrae.
fenestral unit (Fig. 2) containing peloids and larger aggregates. The laminated unit has been figured and described in a earlier paper (Wright, 198lb). These horizons overly a peloidal grainstone, 10-20 em thick and the laminated unit has a sharp boundary where it overlies the peloidal grainstone but the peloidal fenestral unit has a gradational contact. These two horizons will be described separately.
Grains
PELOIDAL FENESTRAL UNIT
In vertically cut slabs, this unit is seen to contain farge numbers of irregular and horizontal tubular fenestrae
There are three distinct major grain types in this unit; large irregular peloids, small, well sorted peloids and peloid aggregates.
Fig. 2. Polished sections through the peloidal layer. (A) is from Llanelly Quarry, showing fenestrae; Note the trains of tubular fenestrae running from left to right across the section. (B) is from Craig y Gaer showing fenestrae and it incorporates darker patches of the associated peloidal grainstone. 179
Fig. 3. (A) Small pellets occurring between the larger peloids; scale bar is I mm long. (B) Concentration of pellets in the upper half of photograph overlying an area of poorly sorted peloids with irregular calcite coatings. Many of the pellets are compacted (welded) together. Scale bar is I mm long. (C) Area of compacted pellets traversed by tubular fenestrae; note the pellets in the fenestra at the top left, and the septa-like structures in some of the fenestrae; scale bar is 2 mm long. (D) Pellet coatings and bridges (arrowed) on larger sediment-grain peloids; scale bar is 0· 5 mm long. (E) Pellet coatings (arrowed) on peloids; scale bar is 0·5 mm long. (F) Tubular fenestra (burrow?) partially filled with pellets; note the irregular domed roof and chamber-like termination. The matrix consists of compacted pellets. Scale bar is I mm long.
Fig. 4. (A) Chamber-like structure within a complex tubular fenestra (see Fig. SF) showing pellets in the roof (arrowed). The
pellets are coated by a fine crust of fibrous calcite. Scale bar is 0·5 mm long. (B) Pellets in the roof (arrowed) of a large chamber like structure in a complex tubular fenestra (see Fig. SF). The pellets also have a fine fibrous crust. Note the sparry haloes around the peloids above the roof (see Fig. 7). Scale bar is 0·5 mm long. (C) Pellet septa (arrowed) in a tubular fenestra; note the sparry haloes around the overlying peloids. The small pellets are packed in between the large peloids. Scale bar is 0·5 mm long. (D) Irregular fenestra containing a pellet arch (arrowed); note geopetal crystal silt.overlying a pellet layer. The matrix consists of compacted pellets. Scale bar is l mm long. (E) Pellet bridge structure; is this a simple pellet bridge similar to those in Fig. 3D or a pellet tube-lining formed by some organism? Scale bar is l mm long. (F) A pellet tube structure. This is not simply a series of pellet bridges connecting peloids but a distinct pellet tube, but was it formed by a root simply pushing pellets aside or by some animal constructing a protective tube? Note the regular sparry halo around the peloid in the upper left quadrant. Scale bar is 0·5 mm long.
V. P.
Wright
Most of the peloidal fenestral unit is a poorly sorted , grainstone containing irregular, subrounded large peloids, 200-1200 11m in diameter, ostracode valves and quartz grains of medium sand size. This material is identical to the underlying peloidal grainstones interpreted as shallow, restricted subtidal deposits (Wright, 198la). The small peloids, which occur in huge numbers, are very well sorted, and consist of well rounded, spherical to ovoid peloids 20-50 11m in size but averaging 40 !J.m. They occur in a variety of distribu tions, e.g. they may occur as fillings between the larger peloids (Fig. 3A) or may make up the whole of the fabric (Fig. 3B) forming a finer, better sorted grain stone. These small peloids may also occur as com pacted masses (Fig. 3C), or may coat the larger peloids and form bridges between them (Fig. 3D, E). They may also occur inside fenestrae, as geopetal fills (Fig. 3F), or in the walls of fenestrae (Fig. 4A, B, C) and sometimes packed between the larger peloids in the walls (Fig. 4C). They also occur as septa-like structures within tubular fenestrae (Fig. 4C) or as arch-like structures within large fenestrae (Fig. 4D) or concentrically arranged in tubes in the intergranular spaces between very large peloids (Fig. 4E, F). These various distributions are shown schematically in Fig. 5. Small peloids are more abundant at the top of the peloidal unit, and sometimes form dense compacted masses (Fig. 3C) in the middle of the unit. These masses appear in hand specimen as grey bands, up to 1 em thick, transected by spar-filled tubular fenestrae. The pellets are less numerous towards the base of the unit where they often occur as coatings and bridges. The third grain type consists of compound grains, up to 5 mm in diameter made up of small peloids (Fig. 6A). These structures have smooth outlines and are not simply aggregates.
the peloidal limestones in the Cheltenham Limestone Member, no peloids have been encountered which show the same degree of sorting, or are the same shape and size as those in the peloidal fenestral unit. They are, therefore, considered to be of pedogenic or biological origin and not primary sedimentary grains. Similar peloids to these small forms have also been described as an important component of Recent calcrete crusts (James, 1972, p. 823 and Harrison, 1977, p. 133), and those described by Harrison are very similar indeed to those described here. Harrison noted the similarities between soil fecal pellets and such peloids but instead interpreted them as small nodules (glaebules of the soil terminology). It is surprising that few of the detailed descriptions of Recent and fossil calcrete crusts mentions fecal pellets, which are a very important component of many other soil types and especially those developed on carbonate parent materials such as rendzina soils (Bridges, 1978). The small peloids described here are identical in size, shape and degree of sorting to the fecal material of the smaller soil animals such as mites, collembolas and some enchytraeid worms, e.g. compare Fig. 3(A, B, F) with those in Babel (1975),
FABRIC TERM
skeletal grains
ooo 00
granular fabric
pellets
dropping fabric
-6 welded agglomeratic fabric
pellets between grains
pellet coats and bridges
coated and linked distribution
�
Interpretation
�0
The larger peloids, ostracode fragments and quartz grains are identical to those in the underlying peloidal grainstones and were presumably derived from them. Accepting that this unit, by virtue of its macro- and microscopic similarities to descriptions of calcrete crusts (see below), is a pedogenic deposit, then the sediment grains would be described as the skeleton grains of a soil (Brewer, 1964). The smaller peloids have no counterpart in the underlying peloidal grainstone, indeed during the examination of hundreds of peels and thin sections of
in tubular fenestrae or as pellet tubes
CD
tubulic distribution
QK;)
8:o Fig. 5. Fabrics and distributions of pellets. See text for
details.
182
Fig. 6. (A) Large compound pellets set in a matrix of smaller pellets. Many of these compound pellets resemble earthworm fecal material. Scale bar is 2 mm long. (B) Peloids (sediment grains) coated by a fine spar; note the meniscus-like thickenings (arrowed). Scale bar is 0·5 mm long. (C) Gravitational'cement' (arrowed) composed of very fine spar; scale bar is 0·5 mm long. (D) Irregular fenestra with a clay-lined bottom (white arrow) overlain by geopetal crystal silt (black arrow). Scale bar is 0·5 mm long. (E) The upper third of the photograph shows a silt-rich clay laminae thought to represent an argillan. This overlies peloids. Scale bar is l mm long.
V. P.
Wright
Bal (1970, 1973) and de Coninck eta/. (1974). Even though the peloids are now composed of micrite, they are interpreted as calcified fecal pellets because of their shape and size similarities to Recent soil fecal pellets, because of their high degree of sorting as compared with associated sedimentary grains and by their presence in the walls of burrow-like fenestrae (see below). These peloids are quite unlike calcrete glaebules which occur in other palaeosols in the Llanelly Formation (Wright, 1982). Fecal pellets are an important component in modern soils and Bal (1973) has provided a useful terminology for such pellets. Using this terminology these Carboniferous pellets would be described as spherical to ellipsoidal, medium fine excrements, and would be said to occur in a heaped distribution (in groups), from single (discrete pellets) to strongly welded (compacted) forms (see Fig. 5). The distribution of these fecal pellets provides additional information of processes which were at work in this soil. Those areas which lack fecal pellets, i.e. composed only of skeleton grains (large peloids), would be said to have a granular soil fabric (Brewer, 1964) (Fig. 5) and form grainstones in the petrographic sense similar to those beneath the peloidal unit. In descriptions of soil fabrics fecal pellets are not usually treated as skeleton grains since they form distinct fabrics, and a variety of specialized terms are available to describe them. The areas which consist only of fecal pellets (Fig. 3B) are said to form a dropping fabric (Babel, 1975) (Fig. 5) or a separated distribution in the sense of de Coninck eta!. (1974, p. 268). The areas where pellets fill the intergranular spaces between large peloids (skeleton grains) (Fig. 3A) are described as having an agglomeratic fabric (Kubiena, 1938, p. 146 and Brewer, 1974, p. 39) (Fig. 5). A coated distribution refers to the skeleton grains coated by pellets (Figs 3D, E and 5) and a linked distribution refers to skeleton grains linked by pellet bridges or braces (Figs 3D, E and 5) (de Coninck eta/., 1974, p. 268). The pellets associated with tubular fenestrae show a tubulic distribution in the sense of Bal (1973) and these are discussed at length below. The dropping fabric, the agglomeratic, coated and linked distributions and the areas of strongly welded pellets all probably owe their origin to concentrations caused by illuviation, the washing down of material in suspension, such that the pellets became mixed with the skeleton grains. Such concentrations in Recent soils have been called mecaconcentrations by Jongerius (1970, p. 320). Some mixing of pellets and
skeleton grains could also have been caused by faunal activity or by the churning of the horizon caused by shrink-swell cycles. By analogy with Recent soils the: pellets were probably produced in the upper, organic rich, part of the soil and were washed down. Such processes are well documented in Recent soils (e.g. Babel, 1975, p. 429; Bal, 1970, Jongerius & Schelling, 1960). The experiments of Wright & Foss (1968) have proved that silt-sized particles (and presumably pellets) up to 50 �m in diameter, are easily moved down through sand by flowing water. Evidence for the action of this process in the peloidal fenestral unit is clearly seen in the partial geopetal fills of many fenestrae (Fig. 3F) and by the overall decrease in the amount of pellets down through the unit. The high concentrations which occur locally in the middle of the horizon probably reflect areas where permeability was reduced resulting in a change in flow velocity and the deposition of the pellet load. Such a reduction in permeability in soils is usually caused by a change in grain size or packing or by the presence of grain coats. Bal (1970, p. 20) describes similar pellet concentrations in Recent soils. The aggregations of pellets to form dense welded (com pacted) masses is a common feature of fecal pellet rich soils today, and such compaction may result from pedoturbations (Jongerius, 1970) caused by faunal or root activity or by shrink-swell cycles. The most common cause, is however, the decay of the pellets, (Bal, 1970, p. 28; Bal, 1973; Bjorkhem & Jongerius, 1974; de Coninck eta/., 1974, p. 270; Jongerius & Schelling, 1960, p. 703). The pellets, as mentioned above, are similar in shape and size to those of the smaller soil arthropods such as the mites, collembolas and some enchytraeids (see also discussion on tubular fenestrae below). The: determination of the composition of soil faunas from their pellets is a difficult task and there are a number of pitfalls. First, pellets are not diagnostic of particular soil organisms for taxonomically different animal groups can produce very similar fecal pellets (Babel, 1975, p. 422); secondly, the size of a pellet is not always a reliable guide to the size of the animal which produced it because pellet size varies with growth stage of the animal (Bal, 1973, p. 64); thirdly the same: animal can produce markedly different pellets de-· pending on the type of vegetation it is feeding on (Bal, 1970; van der Drift, 1964, p. 79); lastly, pellets can shrink considerably on drying with a consequent change in shape (Bal, 1973, p. 65). Subsequent diagenetic changes such as calcification (see below) may also have affected both shape and size. Despite: 184
Carboniferous rendzina these problems, the pellets in the Darrenfelen Pedod erm are remarkable for the uniformity of shape and size which suggests a probable lack of diagenetic deformation, and that the fauna was probably of low diversity. The larger pellet aggregates which also occur in this unit are very similar to the larger fecal pellets in Recent soils (cf. Babel, 1975 and Bal, 1973). The smooth edges of these larger pellets argues against them being simple mechanical soil aggregates (crumbs). What inferences can be drawn about the productiv ity of the fauna from the pellet evidence? The huge numbers of pellets present might indicate that the fauna was very abundant or that it was active for a long period. In Recent soils, pellets can accumulate in considerable quantities where there is an absence of organisms ingesting them, e.g. earthworms normally consume the smaller fecal pellets to 'form larger aggregates rather like those described above (Babel, 1975, p. 428). Thus, pellets in a soil where secondary ingestors are absent have a higher preservation potential than those in other soils. Other factors can influence preservation potential, and the preservation potential of pellets is enhanced in dry conditions (Fitzpatrick, 1971, p. 223), the existence of which is proved by the occurrence of calcretes and evaporites in the Llanelly Formation (Wright, 1981a). The early calcification of these pellets would also have enhanced their chances of preservation (see below). Thin section examination of the pellets has only revealed a very fine micritic microstructure and no recognizable organic structures have been seen. The pellets would originally have been made largely or wholly of organic matter and their preservation suggests early calcification. This calcification is probably analogous to the calcification of fecal pellets in Recent carbonate environments, and although this calcification is common, it is not well understood (Bathurst, 1975, p. 364) but it may be related to bacterial decay. The early calcification of fecal pellets has been noted in Recent caliches in Spain by Klappa (1978, p. 189).
L. ( 1970) Morphological investigation and the role of the soil fauna in their genesis. Geoderma, 4, 5-36. BAL, L. (1973) Micromorphological analysis of soils. Lower levels in the organization of organic soil materials. Soil Surv. Pap. 6. Netherlands Soil Survey Institute, Wagen ingen, 174 pp. BATHURST, R.G.C. (1966) Boring algae, micrite envelopes and the Lithification of molluscan biosparites. Geol. J. 5, 15-32. BATHURST, R.G.C. (1975) Carbonate Sediments and their Diagenesis. Elsevier, Amsterdam. 658 pp. BJORKHEM, U. & JONGERJUS, A. (1974) Micromorphological observations in some podzolised soils from central Sweden. In: Soil Microscopy (Ed. by G. Rutherford), pp. 3 2Q-33 2. Limestone Press, Kingston, Ontario. BRAITHWAITE, C.J.R. (1975) Petrology of palaeosols and other terrestrial sediments on Aldabra, western Indian Ocean. Phil. Trans. R. Soc. B, 273, 1-3 2. BREWER, R. (1964) Fabric and Mineral Analysis of Soils. Wiley, New York. 470 pp. BREWER, R. (1974) Some considerations of micromorpho logical terminology. In: Soil Microscopy (Ed. by G. Rutherford), pp. 28-48. Limestone Press, Kingston, Ontario. BREWER, R.,COOK, K.A.W. &SPEIGHT,J.G. (1970) Proposal for soil stratigraphic units in the Australian Stratigraphic Code. J. geol. Soc. Aus. 17, 103-109. BREWER, R. & HALDANE, A.D. (1957) Preliminary experi ments in the development of clay orientation in soils. Soil Sci. 84, 301-308. BRIDGES, E.M. (1978) World Soils. Cambridge University Press. 128 pp. BULLOCK, P. & MACKNEY,.D. (1970) Micromorphology of strata in the Boyn Hill Terrace Deposits, Buckingham shire. In: Micromorphological Techniques and Applications (ED. by D. A. Osmond and P. Bullock). Soil Surv. Tech. Monogr. 2, 97-106. BUURMAN, P. (1980) Palaeosols in the Reading Beds (Paleocene) of Alum Bay, Isle of Wight, U.K. Sediment ology, 27, 593-606. DE CONINCK, F., RIGHI, D., MAUCORPS, J. & ROBIN, A.M. (1974) Origin and micromorphological nomenclature of organic matter in sandy spodosols. In: Soil Microscopy (Ed. by G. Rutherford), pp. 263-280. Limestone Press, Kingston, Ontario. DuNHAM, R.J. (1971) Meniscus cement. In: Carbonate Cements. John Hopkins Studies in Geology No. 19 (Ed. by 0. P. Bricker), pp. 297-300. Baltimore. ESTEBAN, M. (1974) Caliche textures and Microcodium. Suppl. Boll. Soc. geol. ita/. 92, 105-125. EsTEBAN, C.M. (1976) Vadose pisolite and caliche. Bull. Am. Ass. Petrol. Geol. 60, 2048-2057. FITZPATRICK, E. A. (1971) Pedology: a systematic approach to soil science. Oliver & Boyd, Edinburgh. 306 pp. FLETCHER, J.E. & MARTIN, P.W. (1948) Some effects of algae and moulds in the rain-crust of desert soils. Ecology, 29, 95-100. FRIEDMANN, l., LIPKIN, Y. &0CAMPO-PAUS, R. (1967) Desert algae of the Negev (Israel). Phyco/ogia, 6, 185-200. GOLUBIC, S. & CAMPBELL, S.E. (1979) Analogous microbial forms in Recent subaerial habitats and in Precambrian charts. Gloethece coerulea Geitler and Eosynechococcus moorei Hofmann. Precamb. Res. 8, 201-217. GROVER, G.M. & READ, J.F. (1978) Fenestral and associated
BAL,
REFERENCES
A.E. (1980) Calcrete profiles in the Eyam Limestone (Carboniferous) of Derbyshire. Sedimentology, 27, 651660. BABEL, U. (1975) Micromorphology of soil organic matter. In: Soil Components, Vol. 1; Organic Components (Ed. by J. E. Gieseking), pp. 369--473. Springer-Verlag, Berlin. ADAMS,
185
V. P.
Wright (I 974) Calcrete deposits and Quaternary sedi ments. Edel Province, Shark Bay, Western Australia. In:
vadose diagenetic fabrics of tidal flat carbonates, Middle Ordovician New Market Limestone, south western Vir ginia. J. sedim. Petrol. 48, 453-473. HARRISON R.S. (1977) Caliche profiles: indicators of near surface subaerial diagenesis, Barbados, West Indies. Bull. Can. Petrol. Geol. 25, 123-173. HARRISON, R.S. & STEINEN, R.P. (I 978) Subaerial crusts, caliche profiles, and breccia horizons. Comparison of some Holocene and Mississippian exposure surfaces, Barbados and Kentucky. Bull. geol. Soc. Am. 89, 385-395. JAMES, N.P. (1972) Holocene and Pleistocene calcareous crust (caliche) profiles: criteria for subaerial exposure. J. sedim. Petrol. 42, 817-836. JONGERIUS, A. (1970) Some morphological aspects of regrouping phenomena in Dutch soils. Geoderma, 4, 311331. JONGERIUS, A. & SCHELLING, J. (1960) Micromorphology of organic matter formed under the influence of soil organ isms, especially soil fauna. 7th int. Congress of Soil Science, vol. 2. Madison, Wisconsin, U.S.A. KLAPPA, C.F. (I 978) Morphology, composition and genesis of
READ, J. F.
Evolution and Diagenesis of Quaternary Carbonate Se quences, Shark Bay, Western Australia. Mem. Am. Ass. Petrol. Geo/. 22, 250-282.
R & WRIGHT, V.P. (1981) Palaeosols and tidal flat/lagoon Sequences on a Carboniferous carbonate shelf: sedimentary associations of triple disconformities. J. sedim. Petrol. 51, 1323-1339. RIGHI, D. & DE CONINCK, F. (1974) Micromorphological aspects of Humods and Haplaquods of the Landes du Medoc, France. In: Soil Microscopy (Ed. by G. Ruther ford), pp. 567-588. Limestone Press, Kingston, Ontario. ROLFE, W. D.l. (1980) Early invertebrate terrestrial faunas. In: The Terrestrial Environment and the Origin of Land Vertebrates (Ed. by A. L. Panchen) Systematics Association (Special Volume), IS, 117-157. Academic Press, London. SIMONSON, R.R. (1978) A multiple-process model of soil genesis. In: Quaternary Soils (Ed. by W. C. Mahoney), pp. 2-25. Geo. Abstracts, Norwich. SOIL SURVEY STAFF (1975) Soil Taxonomy, Agricultural Handbook 436. USDA, Washington. SOMERVILLE, l.D. (1979) A cyclicity in the early Brigantian (D2) Limestones east of the Clwydian Range, North Wales and its use in correlation. Geol. J. 14, 69-86. TERUGGI, M.E. & ANDREIS, R.R. (I 971) Micromorphological recognition of paleosolic features in sediments and sedimentary rocks. In: Paleopedology (Ed. by D. H. Yaalon), pp. 161-171. International Society of Soil Sciences and Israel University Press. VAN DER DRIFT, J. (I 964) Soil fauna and soil profile in some inland-dune habitats. In: Soil Micromorphology (Ed. by A. Jongerius), pp. 69-81. Elsevier, Amsterdam. WALKDEN, G.M. (I 974) Paleokarstic surfaces in the Upper Visean (Carboniferous) Limestone of the Derbyshire Block, England. J. sedim. Petrol. 44, 1232-1247. WALLS, R.A. HARRIS, W.B. & NUNAN, W.E. (1975) Calcareous crust (caliche) profiles and early subaerial exposure of Carboniferous carbonates, north eastern Kentucky. Sedimentology, 22, 417-440. WALLWORK, J.A. (I 970) The Ecology of Soil Animals. McGraw-Hill, New York. RIDING,
Quaternary calcretes from the Western Mediterranean: a petrographic approach. Unpublished Ph.D. Thesis. Uni
versity of Liverpool. C.F. (I 979) Lichen stromatolites: criterion for subaerial exposure and a mechanism for the formation of laminar calcretes (caliche). J. sedim. Petrol. 49, 387-400. KLAPPA, C.F. (1980) Rhizolith.s in terrestrial carbonates: classification, recognition, genesis and significance. Sedi mentology, 27, 613-629. KRUMBEIN, W.E. & GIELE, C. (1979) Calcification in a coccoid cyanobacterium associated with the formation of desert stromatolites. Sedimentology, 26, 593-604. KUBIENA, W.L. (1938) Micropedology. Collegiate Press, Ames, Iowa. KUBIENA, W.L. (1970) Micromorphological features of soil geography. Rutgers University Press, New Brunswick, New Jersey. 254 pp. McPHERSON, J. G. (1979) Calcrete (caliche) palaeosols in fluvial redbeds of the Aztec Siltstone (Upper Devonian), southern Victoria Land, Antarctica. Sedim. Geol. 22, 267285. MERMUT, A. & PAPE, TH. (1971) Micromorphology of two soils from Turkey, with special reference to the in situ formation of clay cutans. Geoderma, 5, 271-281. MEYER, R. (I 976) Continental sedimentation soil genesis and marine transgression in the basal beds of the Cretaceous in the east of the Paris Basin. Sedimentology, 23, 235-253. MEYERS, W.J. (1977) Chertification in the Mississippian Lake Valley Formation Sacramento Mountains, New Mexico. Sedimentology, 24, 75-106. MOLLER, G. (1971) Gravitational cement: an indicator for the vadose zone of the subaerial diagenetic environment. In: Carbonate Cements. John Hopkins Studies in Geology, No. 19 (Ed. by 0. P. Bricker), pp. 301-302. Baltimore. MULTER, H.G. & HOFFMEISTER, J.E. (1968) Subaerial laminated crusts of Florida Keys. Bull. geol. Soc. Am. 79, 183-192. PERKINS, R.D. (I 977) Depositional framework of Pleistocene rocks in South Florida. In: Quaternary Sedimentation in South Florida (Ed. by P. Enos and R. D. Perkins). Mem. geol. Soc. Am. 147, 131-198. KLAPPA,
W.C. (I 975) Petrology and diagenesis of eolianites of north eastern fuca tan, Mexico. In: Studies in Geology, 2,
WARD,
Belize Shelf: Carbonate Sediments, Clastic Sediments and Ecology, 500-571. American Association of Petrol
Geologists. N.L. (1976) Paleopedogenic palygorskite from the basal Permo-Triassic of Northwest Scotland. Am. Miner. 61, 299-302. WATTS, N.L. (1977) A comparative study of some Quaternary, Permo- Triassic and Siluro-Devonian calcretes. Unpublished Ph.D. Thesis. University of Reading.
WATTS,
N.L. (1980) Quaternary pedogenic calcretes from the Kalahari (Southern Africa): mineralogy, genesis and diagenesis. Sedimentology, 27, 661-686.
WATTS,
V.P. (1980) Climatic fluctuation in the Lower Carboniferous. Natunvissenschaften, 67, 252-253. WRIGHT, V.P. (198l a) Stratigraphy and Sedimentology of the
WRIGHT,
Llanelly Formation between Penderyn and Blorenge, South
Wales. WRIGHT,
186
Unpublished Ph.D. Thesis. University of Wales. V.P. (1981b) A subaerial stromatolite from the
Carboniferous rendzina Lower Carboniferous of South Wales. Geol. Mag. ll8, 97100. WRIGHT, V.P. (198lc) Algal aragonite-encrusted pisoids from a Lower Carboniferous schizohaline lagoon. J. sedim. Petrol. 51, 479-489. WRIGHT, V.P. (1982) Calcrete palaeosols from the Lower Carboniferous, Llanelly Formation, South Wales, Sedim. Geol. 33, 1-33. (Manuscript received
WRIGHT, V.P., RAVEN, M. & BURCHETTE, T.P. (1981) A Field Guide to the Lower Carboniferous Rocks near Aber gavenny. Department of Geology, University College
Cardiff. 76 pp. & Foss, J.E. (1968) Movement of silt-sized particles in sand columns. Proc. Soil Sci. Soc. Am. 32, 446448.
WRIGHT, W.R.
26 February 1982; revision received 28 June 1982)
187
Reprinted from Sedimentology (1986) 33 831-838
The role of fungal biomineralization in the formation of Early Carboniferous soil fabrics V. PAUL WRIGHT Department of Geology, Wills Memorial Building, University of Bristol, Queen's Road, Bristol BS8 IRJ
ABSTRACT
Paleosols in the Lower Carboniferous limestones of South Wales commonly contain needle-fibre calcite which is an unusual form of calcite recently shown to form by the calcification of fungal hyphae in present day soils. The needle-fibre calcite occurs in two associations in the paleosols: as coatings on sediment grains and as rhizocretions. The former can be compared with the microbial grain coatings of Quaternary calcretes. The latter represent the sites of fungal coats on roots and are interpreted as probable ectomycorrhizae, a symbiotic fungal sheath-root association. These findings suggest that biomineralization was important in the formation of soil fabrics during the Carboniferous as it is in present day soils.
INTRODUCTION
pseudomycelia in the soil science literature (e.g.
Calcretes have now been widely documented in
regarded, but never proved until recently, to be of
ancient sedimentary sequences and have proved useful
fungal origin. The origins of this form of calcite have
Kubiena,
1938; Fitzpatrick, 1984). They have been
for a variety of interpretive purposes. Studies of
been discussed by many workers and were reviewed
Quaternary soil carbonates have shown the impor
in Wright
tance of biomineralization in the formation of many
these elongate crystals are analogous to 'whisker'
( 1984). Some authors have speculated that
crystals formed by growth during extremely high
calcretes (Callot, Guyon & Mousain, 1985; Calvet, 1982; Ca1vet & Julia, 1983; Calvet, Pomar & Esteban, 1975; Esteban, 1974; Kahle, 1977; Klappa, 1978, 1979, 1980; Krumbein & Giele, 1979; Ward 1975, 1978). During a study of subaerial exposure surfaces
However, the illustration provided by these authors
within the Lower Carboniferous (Mississippian) of
more closely resembles lublinite, a bizarre form of
South Wales, a variety of biogenic calcrete fabrics
calcite consisting of stacked, en echelon flattened
degrees of supersaturation caused by rapid degassing of C02 and/or evaporation. This view has most recently been repeated by Given & Wilkinson
(1985).
have been found, including abundant calcified fungal
rhombs (Stoops,
hyphae. The aim of this paper is to describe the
has stressed the strong association between needle
1976; Ward, 1978). Wright ( 1984)
distribution of these fungal fabrics and to compare
fibre calcite and micro-organisms, especially fungi.
them with their Quaternary counterparts.
This view was also reached by Harrison
( 1977) and (1978) during detailed studies of Quaternary calcretes. Recently Chafetz, Wilkinson & Love (1985) Ward
have also suggested a possible biogenic influence.
NEEDLE FIBRE CALCITE
The most significant contribution to this problem Needle-fibre calcite consists of extremely elongate
has come from Callot, Guyon & Mousain
needles of low-magnesium calcite. The needles are
carefully documented the formation of needle-fibre
typically a few microns wide and up to several hundred
calcite by the calcification of soil fungal hyphae of the
( 1985), who
microns long. It is an unusual habit of calcite yet is
Basidiomycetes. A number of authors have noted that
very common in soils. Aggregates of needles resemble
fungi have the ability to concentrate various ions,
fungal mycelia and have long been referred to as
including calcium (Sihanonth & Todd,
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
189
1977), which
V. Paul Wright
can result in either the formation of CaC03 in the outer walls of hyphae (Callot, Mousain & Plassard, 1985) or in the formation of calcium oxalate which is easily broken down by bacteria allowing calcium to form bicarbonate and carbonates (Cromack et a!., 1977). The evidence presented by Callot, Guyon & Mousain ( 1985) shows that at least some forms of needle-fibre calcite are organically precipitated. While it cannot be ruled out that needle-fibre calcite has more than one origin most inorganically precipitated calcite in soils has a polyhedral micritic or more rarely a rhombic habit (Folk, 1974). Claims that needle-fibre calcite forms in abiotic settings by inorganic precipi tation have never been substantiated by experimen tation. The origin of lublinite, however, remains uncertain. Based on the above, the presence of needle-fibre calcite is taken as reliable evidence that fungi were once present in a soil or paleosol. However, such observations can be taken a stage further when the actual distribution of the fungi (needle-fibre calcite) is studied.
by exposure surfaces. The material described here comes from the Heatherslade Geosol, a paleosol which occurs at the top of the Chadian Gully (or Caswell Bay) Oolite in South Wales (Fig. I) (Wright, 1984). The top of the oolite is an irregular surface veneered by a calcrete crust in the Gower (Wright, 1982, 1984) but is overlain by a thick calcretised regolith and petrocalcic horizon as at Miskin near Cardiff (Riding & Wright, 1981) (Fig. 2). The needle-fibre calcite occurs in these paleosols in two associations: (i) within irregular micritic grain coatings on ooids and (ii) composing the walls of 'alveolar textures' (in the sense of Esteban, 1974), that is, elongate tubules with irregularly curved partitions (septae).
GRAIN COATS
The calcrete crust in the Gower and the regolith lithoclasts at Miskin both contain abundant micritic coats on grains. The individual grains (ooids) are coated by a single layer of dark micrite up to 500 llm in thickness. These coats are typically irregular in thickness and commonly display irregular protuber ances including elongate branched filaments, which average 27!lm wide (range 10-60 !lm) and up to several hundred microns long, which may form bridges between grains (Figs. 3, 4). Under the SEM, poorly preserved needles of calcite occur in parallel sets (Fig. 4F). The needles are similar in size to those in the tubular structures described
GEOLOGICAL SETTING
Subaerial exposure surfaces represented by palaeo karsts and paleosols are common in the Lower Carboniferous of South Wales. The sequence repre sents deposition on a carbonate ramp (Wright, 1986) and contains a number of oolitic sandbodies capped
,......
High ""'
Tor
Limestone Caswell Bay Mudstone HG
z <(
0
<( :I: ()
0 0
Gully
[
Oolite
sm
Fig. I. Locality map showing the positions of the two profiles described. Carboniferous Limestone in black. HG = Heatherslade Geosol.
190
Fungal biomineralization in Early Carboniferous soils
----+---- oolite
B
A
Fig. 2. Profiles in the Heatherslade Geosol. (A) Three Cliffs Bay, Gower. (B) Miskin near Cardiff where a prominent massive calcrete horizon (petrocalcic horizon) overlies the lithoclast and clay regolith horizon. below, being commonly 2-5 �m wide and up to 50 �m
occurrence of needle-fibre calcite with fungal remains
long. They pass laterally into micrite which appears
around the grain coatings. Knox
to be a product of their degradation and contains short
micro-organisms were also important in the construc
fragments of the needles set in micrite matrix. Clay
(1977) noted that
tion of coatings in calcareous soils in South Africa.
particles are widely distributed in these grain coats
Such microbially formed coatings typically show
and may range from trace amounts to comprising
irregularities around the grain surface and outgrowths
nearly half the volume of the coating. Microprobe
strikingly like those from the Heatherslade Geosol
analyses suggest that the clays are probably illite.
(see Calvet,
1982; Calvet & Julia, 1983). The presence
of needle-fibre calcite in the Carboniferous grain coats is evidence that fungal biomineralization was a
Interpretation Grain coatings similar to these have been widely documented in Quaternary calcrete soils (James, 1972;
contributing factor in their formation as it is in Quaternary calcretes. In the soil terminology such pedogenic carbonate coatings are called calcans or calcitans (Brewer,
Harrison,
1977; Knox, 1977) and work by Calvet (1982) and Calvet & Julia (1983) has shown that such
1964; Fitzpatrick, 1984).
coatings are constructed by various micro-organisms, especially fungi. These authors also noted the common
'ALVEOLAR TEXTURES' The tubular structures in the paleosol profiles consist
5) up to 100 mm long 1-3 mm in diameter. These are abundant in some lithoclasts and comprise up to 20% of the total volume
of branched tubular voids (Fig. and
of the clasts. They also form a very large component of the calcrete crust which veneers the top of the Gully Oolite at Three Cliffs Bay (Fig.
2). In thin section the 60 �m (range 20-180 �m) wide and up to 500 �m long. In the
tubes are cross-cut by arcuate septae averaging
Miskin material these septae are composed of a brown micrite whereas in the Three Cliffs Bay material they
Fig. 3. Drawing taken from photomicrograph of an oolitic lithoclast from Miskin, showing ooids (stippled) with irregular micritic coats (black). Calcite spar cement in white. Field of view is 6 mm wide.
consist of needle-fibre calcite (Fig.
6), which passes
into a similar brown micrite. This micrite shows a similar relationship to the needles as does the micrite
191
V. Paul Wright
Fig. 4. Photomicrographs of grain coats from the Heatherslade Geosol at Miskin. (A) Oolitic grainstone with micritised ooids coated by dark micrite showing irregular thicknesses around the ooids. Protuberances occur on some grains (e.g. large ooid in centre). Scale bar represents 750 [.Lm. (B-0) Micritic filaments (arrowed) extending from coatings and connecting grains. Identical fabrics have been described from Quaternary calcrete formed by micro-organisms (see text). Scale bar represents 0·5 mm for each. (E) Scanning electron photomicrograph showing part of a micritised ooid coated by micrite (calcan) with calcite spar in the lower right quadrant. The area shown in (F) is indicated by the arrow. Scale bar represents 50 [.Lm. (F) Higher magnification view of area indicated in (E) showing calcite needles. Scale bar represents 10 [.Lm. Both (E) and (F) .are polished specimens etched for 5 s in I% hydrochloric acid.
192
Fungal biomineralization in Early Carboniferous soils
Fig. 5. (A) Polished section through a lithoclast from the Heatherslade Geosol at Miskin. A bifurcating tubular structure runs from the upper left corner of the sample (arrow). The dark area in the lower right corner is a complex set of micrite-filled cracks. The lithoclast is an oolitic grainstone. (B) Photomicrograph of part of the tubular structure shown in (A). The tube is cross-cut by irregular, arcuate, bifurcating micritic septae forming an alveolar-septal structure. Scale bar represents 3 mm.
in the grain coats, so as to suggest that it has also formed by the breakdown of the needles. The Miskin material also has a thin, up to 200 J.lm wide, brown, isopachous fibrous cement on the septae, which occurs on other grains and surfaces in the profile.
(Wright, 1984) and were compared to 'alveolar texture' which has been described in detail from a variety of fossil and Recent calcretes (Esteban, 1974; Esteban & Klappa, 1983; Adams, 1980; Klappa, 1980) and represents encrustations of roots (Steinen, 1974; Harrison, 1977; Klappa, 1980; Calvet & Julia, 1983). The presence of needle-fibre calcite in these root encrustations (rhizocretions or rhizoliths in the sense of Klappa, 1980) is evidence of fungal activity around the root.
Interpretation
The tubular structures from Three Cliffs Bay have been described by the author in an earlier paper 193
V. Paul
Wright
Fig. 6. (A) Photomicrograph of alveolar-septal structure from the Heatherslade Geosol at Three Cliffs Bay. Scale bar represents 0·75 mm. (B) Scanning electron photomicro graph of part of a septum from Three Cliffs Bay showing well pre served needle-fibre calcite set in a · micntlc matrix. Polished and etched for 5 s in I% hydrochloric acid. Scale bar represents 25 Jlm.
onth & Todd, 1977). Thick, concentric fungal coats can form around the rootlet and if calcification occurred (calcification has been recorded in the Basidiomycetes by Callot, Guyon & Mousain, 1985) a series of calcareous coatings might form. On decay of a rootlet some of the coating may collapse or spall-off forming arcuate septae. An individual root tubule may be used by many rootlets during a soil's history making such structures more complex. However, the geologi cal range of the Basidiomycetes is not known with confidence but they may have appeared as early as the Silurian (Sherwood-Pike & Gray, 1985). The term 'alveolar texture' is widely used by geologists to describe similar arcuate septal structures to those described here (e.g. Esteban & Klappa, 1983).
DISCUSSION
The association of calcified fungi and rootlets requires further comment. The fungi may have been acting as decomposers around the decaying rootlet or may have been acting as parasites. However, associations of fungae and higher plants are numerous and complex in their nature and some have been invoked to explain unusual calcrete fabrics such as Microcodium (Klappa, 1978). In present-day soils the Basidiomycete fungi most commonly form ectomycorrhizal associations with roots. They form fungal coatings as a symbiotic relationship with the root. The fungi both accumulate and transport nutrients to rootlet cells and are particularly important in soils of low fertility (Sihan194
Fungal biomineralization in Early Carboniferous soils However the term 'alveolar structure' is also used by soil scientists to describe a soil fabric with abundant, circular-ovoid, irregular pores forming a honeycomb structure (Fitzpatrick, 1984, page 144-5). It results from the release of gases during wetting and drying or freezing and thawing and is therefore similar to the fenestral fabrics common in peritidal limestones (Shinn, 1983). It is suggested that the term used to describe the alveolar structures found widely in fossil and present day calcretes be modified to avoid confusion with the term used by soil scientists. Rather than abandon the term completely I suggest it be modified to alveolar-septal structure to distinguish it from the other use in which the septae are absent. The common occurrence of the needles merging into masses of brown micrite, often containing degraded relics, suggests that the needles were prone to disintegration or recrystallization. This seems to be a common feature of needle-fibre calcite even in present day soils. Knox (1977), describing modern calcrete soils in South Africa noted that needle-fibre calcites readily recrystallize. Calvet & Julia (1983) noted that such calcites break down into smaller 'rod shaped' crystals and eventually to micrite in the present day soils of north-east Spain. Needle-fibre calcite may have been more common in these Carboniferous soils and evidence for its presence in paleosols must be sought very carefully, as has recently been shown by Solomon & Walkden ( 1985) who discovered its presence in paleosols from North Wales only by using cathodoluminescence.
A. & MOUSAIN, D. (1985) Inter relations entre aiguilles de calcite et hyphes myceliens.
CALLOT, G., GUYON,
Agronomie, 5, 209-216.
D. & PLASSARD, C. (1985) Concentra tions de carbonate de calcium sur les parois des hyphes myce!iens. Agronomie, 5, 143-150. CAL VET, F. (1982) Constructive micrite envelopes developed in vadose continental environments in Pleistocene eolian ites of Mallorca (Spain). Acta Geologica Hispanica, 17, CALLOT, G., MOUSAIN,
169-178.
F. & JULIA, R. (1983) Pisoids in the caliche profiles of Tarragona, north east Spain. In: Coated Grains (Ed. by T.M. Peryt) Springer, Berlin, pp 45 6-473. CALVET, F., POMAR, L. & ESTEBAN, M. (1975) Las Rhizocre ciones del Pleistocene de Mallorca. Instituto de Investiga ciones Geo/6gicas Universidad de Barcelona, 30, 35-60. CHAFETZ, H.S., WILKINSON, B.H. & LOVE, K.M. (1985) Morphology and composition of non-marine carbonate cements in near-surface settings. In: Carbonate Cements (Ed. by N. Schneidermann and P.M. Harris). Spec. Pubis. Soc. econ. Paleont. Miner., Tulsa, 36, 337-347. CROMACK, K., SOLLINS, P., TODD, R.L., FOGEL, R., TODD, A.W., FENDER, W.M., CROSSLEY, M.E. & CROSSLEY, D.A. (1977) The role of oxalic acid and bicarbonate in calcium cycling by fungi and bacteria: some possible implications for soil animals. Ecology Bulletin 25, 246-252. ESTEBAN, M. (1974) Caliche textures and Microcodium. Bull. Soc. Geo/. Ita/. (supp.), 92 , 105-125. ESTEBAN, M. & KLAPPA, C.F. (1983) Subaerial exposure environment. In: Carbonate Depositional Environments. (Ed. by P.A. Scholle, D.G. Bebout and C.H. Moore), Am. Ass. Petrol. Geo/. Mem. 33, 1-54. FITZPATRICK, E.A. (1984) Micromorphology of Soils, Chap man & Hall Ltd., London. FOLK, R. L. (1974) The natural history of crystalline calcium carbonate: effect of magnesian content and salinity. J. sedim. Petrol., 44, 40-53. GIVEN, R.K. & WILKINSON, B. H. (1985) Kinetic control of morphology, composition, and mineralogy of abiotic sedimentary carbonates. J. sedim. Petrol., 55,109-119. HARRISON, R.S. (1977) Caliche profiles: indicators of near surface subaerial diagenesis, Barbados, West Indies.
CALVET,
ACKNOWLEDGMENTS
Canadian Petrol. Geol. Bull. 25, 123-173.
I thank Richard Carlton (Open University) for assistance with the electron microscoy, Simon Powell for preparing the photographs and Mavis Hardiman for typing the manuscript. I thank Dr. G. Callot for providing me with preprints of her work, and Henry Chafetz and Maurice Tucker for their constructive comments on an earlier version of this paper.
N.P. (1972) Holocene & Pleistocene calcareous crust (Caliche) profiles: criteria for subaerial exposure. J. sedim. Petrol., 42, 817-836. KAHLE, C. F. (1977) Origin of subaerial Holocene calcareous courts: role of algae, fungi and sparmicritization. Sedimen tology, 24, 413-435. KLAPP A, C.F. (1978) Biolithogenesis of Microcodium: elucidation. Sedimentology, 25, 489-522. KLAPPA, C.F. (1979) Lichen stromatolites: criterion for subaerial exposure and a mechanism for the formation of laminar calcretes (caliche). J. sedim. Petrol. 49, 387-400. KLAPPA, C.F. (1980) Rhizoliths in terrestrial carbonates: classification, recognition, genesis and significance. Sedi
JAMES,
REFERENCES
mentology, 27,613-629.
A.E. (1980) Calcrete profiles in the Eyan Limestone (Carboniferous) of Derbyshire. Sedimentology, 27, 651-
ADAMS,
Caliche profile formation, Saldanha Bay (South Africa) Sedimentology, 24, 657-674. KUBIENA, L.W. (1938) Micropedology, Collegiate Press, Ames, Iowa.
KNOX, G.J. (1977)
660
R. (1964) Fabric and Mineral Analysis of Soils. Wiley, New York.
BREWER,
195
V.
Paul Wright Carbonate Sediments, Clastic Sediments and Ecology, (Ed. by K.F. Wantland and W.C. Pusey) Amer. Ass., Petrol.
W.E. & GIELE, C. (1979) Calcification in a coccoid cyanobacterium associated with the formation of desert stromatolites. Sedimentology, 26, 593-604.
KRUMBEIN,
R. & WRIGHT, V.P. (1981) Paleosols and tidal flat lagoon sequences on a Carboniferous carbonate shelf. J. sedim. Petrol. 51, 1323-1339.
RIDING,
M.A. & GRAY, J. (1985) Silurian Fungal remains: probable records of the Class Ascomycetes. Letlzaia 18, 1-20.
SHERWOOD-PIKE,
E.A. (1983) Birdseyes, fenestral and shrinkage pores and loferites: a re-evaluation. J. sedim. Petrol. 53, 619-
SHINN, 628.
P. & TODD, R.L. (1977) Transfer of nutrients from ectomycorrhizal fungi to plant roots. Ecological
SIHANONTH,
Geol. Studies in Geol., 2 500-571.
WARD, W.C. (1978) Collectors guide to carbonate cement types, north eastern Yucutan peninsula. In: Geology. and Hydrogeology of North Eastern Yucatan. (Ed. by W.C. Ward and A.E. Weidie) New Orleans Geological Society, New Orleans, pp 147-169. WRIGHT, V.P. (1982) The recognition and interpretation of paleokarsts; two examples from the Lower Carboniferous of South Wales. J. sedim. Petrol, 52, 83-94. WRIGHT, V.P. (1984) The significance of needle-fibre calcite in a Lower Carboniferous palaeosol. Geol. J. 19, 23-32. WRIGHT, V.P. (1986) Facies sequences on a carbonate ramp: the Lower Carboniferous of South Wales. Sedimentology, 33,221-241.
Bulletin, 25, 392-397.
S.T. & WALKDEN, G.M. (1985) The application of cathodoluminescence to interpreting the diagenesis of an ancient calcrete profile. Sedimentology, 32, 877-896.
SOLOMON,
Similar thread-like outgrowths from grain coats have been recently described from Cambrian phoscrete profiles in Australia and are associated with phosphatized microbial tubules, see Southgate, R. N. (1986) Cambrian phoscrete profiles, coated grains and microbial processes in phosphogenesis: Georgina Basin, Australia. J. sedim. Petrol. 56, 429-441. Microbial micritic inter-granular networks have also recently been described from Bahamian beach rock by Beier, J. A. (1985) Diagenesis of Quaternary Bahamian beach rock: petrographic and isotopic evidence. J. sedim. Petrol. 55, Note added in proof:
R.P. (1974) Phreatic and vadose diagenetic modi fication of Pleistocene limestone; petrographic observa tions from subsurface of Barbados, West Indies. Bull. Am. Ass. Petrol. Geol. 58, 1008-1024.
STEINEN,
STOOPS, G .J. (1976) On the nature of'lublinite' from
Hollanta
(Turkey). American Mineralogist, 61, 172. WARD, W.C., (1975) Petrology and diagenesis of eolianites of north eastern Yucatan, Mexico. In: Belize Shelf: (Manuscript received 8 November
755-761.
1985; revision received 5 March 1986)
196
Reprinted from Sedimentology (1987) 34 991-998
Petrographic and geochemical analysis of caliche profiles in a Bahamian Pleistocene dune J. A. B E I E R Biogeochemical Laboratories, Department o fGeology, Indiana University, Bloomington, Indiana 47405, U.S.A.
A BSTRACT
Two caliche profiles from a Pleistocene carbonate dune on San Salvador Island, Bahamas, were examined by petrographic and geochemical analysis. Profile A is an immature buried caliche profile characterized by caliche pisolites, a friable crust and abundant Cerion. Profile B is a more well-developed caliche profile at the top of the dune which contains abundant pisolites, rhizomorphs, laminated calcrete, a breccia and abundant Cerion. Geochemical changes in caliche profiles relative to the host rock are an increase in Al203, Fe203 and 3 total organic carbon, a decrease.in Mg and Sr, and a decrease in o'3Cc.,b, o'80c.,b and o' C0,9. The magnitude of these changes is probably a function of the duration of subaerial exposure and resultant colonization by dune plants and associated microflora. Abundance of calcified filaments and needle-fibre crystals in profile A attests to the importance of microbial processes in the early development of caliche profiles. Biogenic structures are largely destroyed in profile B due to recrystallization, but indirect evidence of biological activity is retained in the form of carbon isotope values.
profiles at different stages of development from a Pleistocene dune on San Salvador Island, Bahamas. Diagenetic features of earliest caliche profile devel opment are preserved in an immature buried soil zone. Comparison with a more well-developed profile leads to an understanding of the sequential development and diagenetic alteration of caliche profiles. This information can be useful in reconstructing the depositional and diagenetic history of the entire dune sequence.
INTRO DUCTION
Holocene and Pleistocene caliche profiles capped by calcareous crusts (calcrete) form by progressive alter ation of the host rock during pedogenesis. Character istic textures and fabrics of caliche profiles include partial to total micritization of carbonate skeletal grains, and precipitation of calcrete around individual and composite grains and as laminated crusts. How ever, the processes whereby micritization occurs, especially the extent of biological involvement, are still not completely understood. Previous descriptions of microscopic diagenetic features in caliche profiles include studies of Holocene and Pleistocene sequences from Barbados (James, 1972), the Florida keys (Multer & Hoffmeister, 1968; Coniglio & Harrison, 1983), the Yucatan (Ward, 1975), Western Australia (Read, 1974), Southern Australia (Warren, 1983) and South Africa (Seisser, 1973; Knox, 1977). The purpose of this study is to combine petrographic examination and geochemical analysis of two caliche Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
GENERAL SETTIN G
San Salvador Island is located about 620 km ESE of Fort Lauderdale, Florida, and is situated along the eastern edge of the Bahamas platform (Fig. 1). The island is surrounded by a shallow shelf with abundant reefs and other carbonate-producing communities, beyond which water depths reach 5000 m on the steep continental slope. The interior of the island is dominated by arcuate 197
J. A. Beier
ao·
75'
300
35'
500
Miles Km
N
t Grand Bahama
' 25
��; d\ ��
� A dros 1 sland
' 25
\ ....._ � 0.
g
san Salvador Island
"'
75'
70'
Fig. I. Map showing location of study area.
buried by younger aeolian sediments. Profile A is characterized by the development of caliche pisolites and a thin, friable crust (Fig. 2b). �hells of the terrestrial gastropod Cerion are also abundant. The dune is capped by a more well-developed caliche profile (profile B), which exhibits more abundant caliche pisolites, especially on the flank of the dune (Fig. 2c), well-developed rhizomorphs, and laminated calcrete which occurs as crusts on top of grainstone and between pisolites. Profile B is capped with a breccia which consists of clasts of aeolianite cemented by calcrete. Although younger in age, profile B exhibits more well-developed vadose diagenetic features due to longer duration of subaerial exposure and pedoge nesis. According to Carew & Mylroie ( 1985) the Bluff was deposited sometime during the regression following the last interglacial high stand of sea-level (Substage 5e), but the exact age of dune emplacement has not been determined. Carew & Mylroie (in press) have also determined by 14C dating of pisolites and amino acid racemization dating of Cerion that soil zone· A formed about 60 000 yr BP and soil zone B about 15 000 yr BP.
dune ridges with shallow brackish to hypersaline lakes occupying the low inter-dune areas. Coastal landforms range from cliffs of aeolianite to crescentic sandy beaches between rocky headlands where Holocene beachrock is common (Beier, 1985). Carew & Mylroie ( 1985) have proposed a model for the Holocene and Pleistocene stratigraphic framework of San Salvador in which the deposition of subtidal, reef, beach and dune suites during high stands of sea level is interspersed with the development of palaeo sols and other exposure features associated with lower sea-levels.
The Bluff
The location of this study is a carbonate dune exposed on the SE coast of San Salvador known as the Bluff (Fig. 2a). The aeolianite at the Bluff is composed of poorly cemented skeletal grains, dominantly penero plid and miliolid forams, and red and green algae. Large-scale crossbeds attest to the aeolian origin of the dune. A very immature caliche profile (profile A) occurs about 5 m above sea-level (Fig. 2a), and is 198
Analysis of caliche profiles in a Bahamian Pleistocene dune
Fig. 2. (A) Photograph of carbonate dune at the Bluff, San Salvador Island, Bahamas, showing caliche profile A (a), a buried soil horizon, and caliche profile B (b) at the top of the dune. (B) Photograph of caliche profile A showing caliche pisolites and a poorly-developed crust. (C) Photograph of caliche profile B showing abundant caliche pisolites interspersed with laminated calcrete on the flank of the dune.
ments), and depletion occurs only in profile B, where the concentration is about 700 ppm. Isotopic compo sition of the host rock is about + 2·0%0 b13C, reflecting the marine origin of the aeolianite, and - 2·0%0 «5180, indicating some equilibration of the host rock with meteoric water. Caliche profiles show enrichment in 3 the light isotopes: b1 C is - 3·0%0 in profile A and -8·0%0 in profile B, but «5180 decreases only slightly to - 4·0%0 in both profiles. The abundance of total organic carbon is extremely low (0· 1%) in the host rock and in profile A, but increases to 0·5% in profile B. The isotopic composition of total organic carbon is - 19·0%0 in the host rock, - 2 1·0%0 in profile A and reaches - 25·0%0 in profile B.
GEOCHEMISTRY OF WHOLE-ROCK SAMPLES
Analysis of whole-rock samples from the Bluff reveal geochemical changes in aeolianite which has been altered by soil zone processes (Fig. 3). The magnitude of geochemical alteration of the host rock is apparently a function of caliche profile maturity, since enrichment or depletion of elements is generally more marked in profile B than in profile A. Aluminium and iron occur in extremely low concentrations in the host rock, but the concentration of both elements increases slightly in profile A and in profile B Fe203 reaches 0·5 wt.% and Al203 is 0· 15 wt.%. Magnesium in whole-rock samples is derived from high-Mg calcite skeletal grains (red algae, forams), and wt.% MgO is 2·5-3·0% in the host rock, 1·5% in profile A and 0·5% in profile B. The host rock also contains about 3000 ppm Sr derived from aragonitic skeletal grains (green algae, coral frag-
Discussion
The enrichment of iron and aluminium oxides in palaeosols, previously thought to represent the insol uble residue of carbonate rock dissolution, has been 199
J. A. Beier
%TOG
20
., ,,
-·
Fig. 3. Geochemical profiles of whole-rock samples for section shown in Fig. 2A. Samples for trace element analysis (AI203,
Fe203, MgO, Sr) were prepared by mixed-acid (HF, HN03, HCI04) dissolution of powdered samples and analysed by 3 induction coupled plasma optical emission spectroscopy. Samples for b1 C cacb and b180carb were powdered and vacuum roasted for 1 hr at 400°C to remove organics, then reacted in 103% H3P04 at 50°C and the evolved C02 was run on a Finnegan-MAT Delta E mass spectrometer. The abundance of total organic carbon was determined by reacting weighed, powdered samples with cold 1·0 M HCI for 24 hr to remove carbonate and the organic residue then combusted in a quartz tube with CuO added at 850°C. The amount of resultant C02 was measured to determine% TOC and the gas then analysed on the mass spectrometer 3 to determine b1 C0,g. All isotopic analyses were duplicated to ensure reproducibility.
the precipitation of diagenetic carbonates. A simple mass balance calculation shows that carbonate in profile B at 8·0%0 represents a mixture of about 50% soil-gas carbon and 50% carbon from host rock dissolution, assuming an average value of +2·0%0 for the host rock and - 18·0%0 for soil-gas carbon (unpublished data, Warren W. Wood, USGS, Reston, VA.). The contrast between high variability of 813C and the relative homogeneity of 8180 has been noted by Allan & Matthews ( 1982), and is attributed to rapid equilibrium of carbonates with the large, isotopically homogenous oxygen reservoir of meteoric water. The oxygen isotopic composition of meteoric water is controlled by the 8180 of local rainwater, which is about 2·0 to 3·0%0 in the Bahamas (Yurtesever, 1975). Aeolian sediment is derived from beach sand, and the oxygen isotopic composition of modern beach sand on San Salvador Island is about -0·5%0 (Beier, 1985), reflecting 8 180 of seawater. However, 8180 of aeolianite at the Bluff is about 2·0%0, which suggests equilibration with meteoric water even though the sediment is only poorly cemented and. it still retains its marine signature with respect to carbon. The slight decrease in 8180 in caliche profiles may be due to precipitation of diagenetic carbonates at 'slightly elevated temperatures in the surface environment, or it may represent more complete equilibration with meteoric water. The increase of organic carbon in caliche profiles
attributed in some cases to accretionary addition of material to soil zones (Ruhe, Cady & Gomez, 1961). This appears to be true for the enrichment of Al203 and Fe203 in caliche profiles at the Bluff since the concentration of these elements is extremely low in the host rock and there is no evidence of large-scale dissolution of aeolianite at this location. AI and Fe are most likely added to the soil zone via wind-borne continental detritus. According to Savoie & Prospero ( 1977), the source of mineral aerosols deposited in the northern tropical Atlantic is the arid and semiarid regions of West Africa. Increased deposition of AI and Fe oxides in very mature soil zones results in the formation of Terra rossa-like sediments, which char acterize older paleosols at other locations on San Salvador Island. Both caliche profiles at the Bluff show a decrease in Mg, but Sr decreases only in profile B, indicating that Mg-calcite components are more readily dissolved than aragonite components in this section. This suggests that vadose pore waters are probably near or above saturation for aragonite (Walter, 1985). The carbon in vadose diagenetic carbonates repre sents a mixture of carbon from two isotopically distinct sources: marine carbonate derived from dissolution of sediment grains, and organic carbon in the form of soil gas produced during pedogenesis (Gross, 1964). Therefore, carbon isotope values can reflect the extent to which soil zone processes influence
-
-
-
-
200
Analysis of caliche profiles in a Bahamian Pleistocene dune
which is isotopically lighter than that in the underlying aeolianite is due to the addition of organic matter derived from the terrestrial plants which colonize the dune surface.
(Fig. 4C-F). Such structures are also common in caliche profiles and have been observed by several authors. The close association of calcified filaments and needle-like crystals with caliche pisolites has been noted by Calvert & Julia ( 1983) and Coniglio & Harrison ( 1983). Calcified filaments such as those shown in Fig. 4C and D have been interpreted as algal filaments encrusted with secondary calcite on the basis of size and shape of the filaments (James, 1972; Klappa, 1979). Clusters of needle-like structures such as those in Fig. 4E and F have been interpreted as fungal hyphae by Kahle ( 1977). Rods of carbonate 12 �m long have also lj)een observed and interpreted as fungal by Kahle ( 1977, Fig. 4a, p. 422) and identical structures also occur infrequently inside the caliche pisolites examined in this study.
PETRO GRAPHY Host rock
Aeolianite below caliche profiles is commonly ce menteq by micrite or microspar. The aeolianite at the Bluff is in general poorly lithified. Throughout most of the sequence cement occurs only at grain boundaries and porosity is still high. However, micritic and sparry calcite cements have been isolated using a micro-drill from more well-cemented eolianite below paleosols at several other locations on the island and analysed for stable isotope values. Whole-rock values of about + 2·0%0 <513C are similar to <513C of isolated sedi ment grains, whereas cements range from -4·5%0 to -6·5%0 <513C. This indicates that pore waters contain significant quantities of soil gas-derived carbon even below the caliche profile.
Profile B
Caliche pisolites from profile B commonly contain sparry calcite cement, and many are hollow. SEM examination of these pisolite� shows that filaments are rare due to recrystallization of the pisolite centres to large ( 100-400 �m) calcite crystals. This calcite has <513C values of - 5·0 to -8·0%o· Examination of rhizomorphs in thin section shows a gradation from basically unaltered grainstone to partially micritized grainstone (Fig. 4B) in which randomly oriented needle-like crystals commonly fill voids, and finally to a completely micritized fabric in the centre of the rhizomorph. Randomly-oriented needle fibres are also very common in caliche profiles, and precipitation of these crystals has been attributed to rapid evaporation of super-saturated fluids (James, 1972). However, Knox ( 1977) pointed out that these crystals are similar to those found by Krumbein ( 1968) in cultures of microbes from the Nari limestone, and therefore microbial activity may be important in the precipitation of such crystals. Laminated calcrete crusts are common in profile B, and in thin section relict sediment grains are com monly seen within the calcrete. No filamentous structures were observed during SEM examination of calcrete crusts. Samples of laminated calcrete yield <51 3C values of- 8·0.to - 10·0%o·
Profile A
Coated grains and pisolites are common in caliche profiles and have been noted by numerous authors. In thin section, coated grains appear as individual sediment grains, usually micritized, with 20-150 �m thick coatings of concentrically laminated calcrete. Such grains have been referred to as diagenetic or caliche ooids (Siesser, 1973; Read, 1974). Examination of caliche pisolites in thin section reveals a 2-5 mm core of partially to totally micritized grains cemented by micrite. The pisolites are formed when these clusters of grains are encased by a 100500 �m thick rind of calcrete, which also may contain a few sediment grains. Caliche pisolite centres have b13C values of about - 3·0 to - 4·0%0, and calcrete which coasts the pisolites has <513C values of about -8·0%0• Filamentous structures can be seen in intergranular voids in the interior of the pisolites (Fig. 4A). These relatively large (50 �m) filaments with smaller (I 0 �m) filaments branching radially from them are probably rootlets of dune plants (Kiappa, 1979; Coniglio & Harrison, 1983). Examination of pisolites with the SEM reveals abundant calcified filaments and clusters of tangentially oriented needle-like crystals coating and cementing grains in the interior of the pisolites
Discussion
The abundance of filaments and needle-like crystals in profile A indicates that the formation of caliche pisolites during earliest caliche profile development is strongly influenced by microbial processes associated 201
J. A. Beier
400JLm
100JLffi
Fig. 4. (A) Thin section photomicrograph of the interior of a caliche pisolite from profile A showing rootlets of dune plants. (B) Thin section photomicrograph of the outer edge of a rhizomorph from profile B showing micritization of the host grainstone toward the centre of the rhizomorph. (C-F) SEM photomicrographs of the interior of a caliche pisolite from profile A. (C) Calcified tubules, probably of algal origin, which coat and bind together grains inside the pisolite, and appear as micrite cement in thin section. (D) Close-up of calcite (?) crystals encrusting tububles shown in (C). (E) Tangential needle-fibre crystals which also coat grains inside pisolites in profile A and may have a fungal origin. (F) Close-up of tangential needle-fibre crystals shown in (E). 202
Analysis of caliche profiles in a Bahamian Pleistocene dune
with colonization of the dune surface by higher plants. The proliferation of a rhizosphere flora is related to the loss of several types of plant metabolites from plant roots and subsequent rapid use of these com pounds by bacteria, fungi and algae (Webley, East wood & Gimingham, 1952; Whipps & Lynch, 1986). The continuous loss of large quantities of organic carbon from plant roots (as exudates, secretions, lysates and respired C02) to the surrounding soil (Whipps & Lynch, 1986) is probably the dominant source of isotopically light soil-gas C02 in vadose pore waters. The precipitation of micrite in caliche profiles has been attributed to various mechanisms involving the evaporation of carbonate-saturated pore waters in the near-surface vadose zone (Seisser, 1973; Warren, 1983). Most authors cite dissolution of sediment grains as an important source of carbonate, along with contributions by sea-spray (James,. 1972), rainwater (Warren, 1983), and possibly aerosol dust and terres trial flora (James, 1972). The carbon isotopic compo sitions of pisolites and laminated calcrete indicate that organic carbon, derived from plant roots and their associated micro-organisms, is indeed a major source of carbon in micrite precipitated in the caliche profiles examined in this study. Comparison of microscopic features from profiles A and B indicates that as caliche profile development proceeds, biogenic structures which give evidence of the origin of macroscopic features (e.g. pisolites) may be destroyed. However, indirect evidence of biological involvement may be preserved in the form of stable carbon isotopic compositions of diagenetic carbon ates. Evidence of biological processes may even be preserved in aeolianite below the caliche profile itself in the form of isotopically light cements whose textures are not influenced by pedogenic processes.
SUMMARY
Two caliche profiles at different stages of development were studied by petrographic examination and geo chemical analysis in order to determine the sequential development and diagenetic alteration of caliche profiles from San Salvador Island, Bahamas. Profile A is an immature caliche profile buried by younger aeolian sediments, which preserved diage netic features of earliest caliche profile development during previous subaerial exposure. It is characterized in the field by caliche pisolites, a thin, friable crust, and abundant Cerion. Geochemical analysis of whole-
rock samples shows profile A to be slightly enriched in Al203 and Fe203 over the host rock due to aerosol deposition. The concentration of Mg is slightly lower than the host rock due to dissolution of Mg-calcite grains; and light stable-isotope values indicate incor poration of soil-gas C02 into diagenetic carbonate. Abundant calcified filaments and needle-fibre crystals 3 in pisolites of profile A, as well as the b1 C of micrite, attest to the importance of microbial processes during early caliche profile development. Profile B, although younger in age, is more well developed than profile A, and contains abundant caliche pisolites, rhizomorphs, laminated calcrete and a breccia, as well as abundant Cerion. The enrichment of Al203 and Fe203 in this profile is greater than that in profile A because of the increased aerosol deposi tion. Mg and Sr are depleted due to dissolution of 3 aragonite as well as Mg-calcite components; and b1 C indicates an even greater amount of soil-gas C02 in diagenetic carbonate than in profile A. An increase in the amount of organic carbon in whole-rock samples 3 is coupled with a decrease in b1 C0,9 due to the addition of terrestrial organic matter. Although micritization and other features of soil zone diagenesis are more abundant in profile B, biogenic structures have largely been destroyed by recrystallization. However, carbon isotopic compositions still reflect the importance of biological involvement in caliche profile development. Differences in the extent of development of the caliche profiles could be the result of either longer duration of subaerial exposure or an increased rate of diagenesis during the development of soil zone B over the last 15 000 yr relative to about 60 000 yr BP when soil zone A formed. Dissolution/precipitation reac tions in the vadose environment are probably most influenced by the amount of rainfall. Several atmos pheric models conclude that climates were generally drier during glacial periods (Gates, 1976). However, there is little information on variations in the rates of rainfall in the Bahamas since the last interglacial interval, and no compelling evidence exists to suggest that climatic conditions would favour increased rates of diagenesis and aerosol deposition over the last 15 000 yr compared to 60 000 yr BP. Therefore, I conclude that soil zone B represents a much longer duration of subaerial exposure than soil zone A. This interpretation can be used in conjunction with the age estimates by Carew & Mylroie (in press) of the dune and the paleosols in reconstructing a tentative scenario for the depositional history of the Bluff. The dune was emplaced sometime between 125 000 yr BP 203
J. A. Beier
and 85 000 yr BP. A period of erosion following that initial deposition was initiated at some unknown time and must have given way to a period of non-deposition 60 000 yr BP, when soil zone A formed. Renewed deposition of aeolian sediment occurred shortly thereafter since the diagenetic features of soil zone A suggest only transient exposure. Subsequent erosion was followed by a period of non-deposition repre sented by soil zone B which began about 15 000 yr BP and continues today.
Pleistocene caliche from Big Pine Key, Florida. Bull. Can. Petrol. Geol., 31,3-13. GATES, L.W. (1976) Modeling the ice-age climate. Science, 191,1138-1144. GROSS, M.G. (1964) Variations in the 018/016 and C13fC12 ratios of diagenetically altered limestones in the Bermuda Islands. J. Geol. , 72,170-194. JAMES, N.P. (1972) Holocene and Pleistocene calcareous crust (caliche) profiles: criteria for subaerial exposure. J. sedim. Petrol., 42,817-836. C. F. (1977) Origin of subaerial Holocene calcareous
KAHLE,
crusts: role of algae, fungi and sparmicritisation. Sedimen tology, 24,413-435. C.F. (1979) Calcified filaments in Quaternary
KLAPPA,
calcretes: organa-mineral interactions in the subaerial vadose environment. J. sedim. Petrol., 49,955-968. KNOX, G .J. (1977) Caliche profile formation; Saldanha Bay (S. Africa). Sedimentology, 24,657-674. KRUMBEIN, W.E. (1968) Geomicrobiology and geochemistry of the 'Nari-lime-crust' (Israel). In: Recent Developments in Carbonate Sedimentology (Ed. by G. Mueller and G. M. Friedman), pp. 138-147. Springer-Verlag, Heidelberg. MULTER, H.C. & HOFFMEISTER, J.E. (1968) Subaerial laminated crusts of the Florida Keys. Bull. geol. Soc. Am.,
ACKNO WLE D GMENTS
This research was funded in part by grants from Sigma Xi and NASA (NRG 15-003- 1 18 to John M. Hayes). SEM work was supported by NSF (PCN 8212660). I thank Don and Kathy Gerace of the Bahamian Field Station on San Salvador Island for their assistance during field work, Mark Gilstrap for running the ICP, and Abhijit Basu for the use of his microscope. Reviews of the manuscript by Jim Carew, Eric Davaud, John Mylroie and Tony Randazzo were appreciated. Jim Carew and John Mylroie also kindly made their 14C and amino acid racemization dates available to me. Finally, I thank my advisor, John M. Hayes, for his support.
79,183-192.
J.F. (1974) Calcrete deposits and Quaternary sedi ments, Edel Province, Shark Bay, Western Australia. In:
READ,
Evolution and Diagenesis of Quaternary Carbonate Se quences, Shark Bay, Western Australia (Ed. by B. W. Logan et a!.). Mem. Am. Ass. Petrol. Geol., 22,250-282. R.V., CADY, J.G. & GOMEZ, R.S. (1961) Paleosols of
RUHE,
Bermuda. Bull. geol. Soc. Am., 72,1121-1142. D.L. & PROSPERO, J. M. (1977) Aerosol concentration statistics for the tropical North Atlantic. J. geophys. Res.,
SAVOIE,
82,5954-5963. (1973) Diagenetically formed ooids and intraclasts in South African calcretes. Sedimentology, 20,
SEISSER, W.G. 539-551.
REFERENCES
L.M. (1985) Relative reactivity of skeletal carbon ates during dissolution: implications for diagenesis. In: Carbonate Cements (Ed. by N. Schneiderman & P. M. Harris). Spec. Pubis Soc. econ. Paleont. Miner., Tulsa, 36,
WALTER,
J.R. & MATTHEWS, lt.K. (1982) Isotope signatures associated with early meteoric diagenesis. Sedimentology,
ALLAN,
3-16.
29,797-817.
W.C. (1975) Petrology and diagenesis of carbonate eolianites of Northeastern Yucatan Penninsula, Mexico. In: Belize Shelf Carbonate Sediments, Clastic Sediments, and Ecology (Ed. by E. G. Purdy et a!.). Am. Ass. Petrol.
Diagenesis of Quaternary Bahamian beachrock: petrographic and isotopic evidence. J. sedim.
WARD,
BEIER, J.A. (1985)
Petrol. , 55,755-761. CALVERT, F. & JULIA, R. (1983)
Pisoids in the caliche profiles of Tarragona (N.E. Spain). In: Coated Grains (Ed. by T. Peryt), pp. 456-473. Springer-Verlag, Berlin. CAREW, J.L. & MYLROIE, J.E. (1985) The Pleistocene and Holocene stratigraphy of San Salvador Island, Bahamas, with reference to marine and terrestrial lithofacies at French Bay. In: Pleistocene and Holocene Carbonate Environments on San Salvador Island, Bahamas (Ed. by H. A. Curran), pp. 11-61. Guidebook for Field Trip 2, GSA
Geol., Studies in Geology, 2,500-571. WARREN, J.K. (1983)
Pedogenic calcrete as it occurs in Quaternary calcareous dunes in coastal South Australia.
J. sedim. Petrol., 53,787-796.
D.M., EASTWOOD, D.J. & GIMINGHAM, C H. (1952) Development of a soil microflora in relation to plant succession of sand-dunes, including the 'rhizosphere' flora associated with colonizing species. J. Ecol., 40, 168-178. WHIPPS, J.M. & LYNCH, J.M. (1986) The influence of the rhizosphere on crop productivity. In: Advances in Microbial Ecology (Ed. by K. C. Marshall), 9,pp. 187-240.
WEBLEY,
1985 Annual Meeting, Orlando, Florida.
J.L. & MYLROIE, J.E. (in press) A refined geochron ology for San Salvador Island, Bahamas. In: Proc. 3rd Symp. Geology of San Salvador Island, Bahamas (Ed. by H. A. Curran). CCFL Bahamian Field Station. CONIGLIO, M. & HARRISON, R.S. (1983) Holocene and
CAREW,
(Manuscript received 10 October
.
YURTESEVER, Y. (1975) Worldwide Survey of Stable Isotopes
in Precipitation. Rept. Sect. Isotope Hydro!., IAEA, November, 1975, 40p.
1986; revision received 23 March 1987) 204
BIOLOGICAL ACTIVITY AND LAMINAR CALCRETES
Laminar calcretes are a distinctive feature of calcrete
Wright et al. describe Carboniferous and Jurassic
2),
Cretaceous laminar calcretes produced by the calci
coat impermeable bedrock substrates and occur in
fication of root mats. In some cases these have
profiles. They can cap petrocalcic horizons (Fig.
association with root mats and water-tables or with
formed as discrete horizons not associated with cal
calcretes forming at the capillary fringe. These dif
crete profiles or bed-rock surfaces and represent
ferent occurrences are discussed in the paper by
root mats formed above shallow water-tables in
Wright, Platt & Wimbledon.
aggrading flood-plain sediments. These papers stress the role of biological activity
The origin of such calcretes is complex. Kahle describes 'replacive' forms, representing an alter
in
ation zone caused, especially, by micro-organisms.
laminar calcretes consist of very finely laminated,
Krumbein & Giele describe another form of micro
dense micritic limestones lacking evidence of a bio
bial laminar calcrete formed at the surface by
logical origin (see introductory section, p. 14).
forming
laminar
calcretes.
However,
many
cyanobacteria. These are accretionary in origin.
Fig. 14. Microbial mat (cryptogamic earth) from the desert surface in Arches National Park, Utah. Similar crusts, composed of cyanobacteria, occur in many desert areas. Krumbein & Giele (this volume) describe calcified forms. These mats from Utah are only lightly calcified. For further details see Wright
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
(1989).
205
Reprinted from Sedimentology (1977) 24 413-435
Origin of subaerial Holocene calcareous crusts: role of algae, fungi and sparmicritisation
C H A R L E S F. K A H L E Department of Geology, Bowling Green State University, Bowling Green, Ohio
43403
ABSTRACT
The Pleistocene Miami Limestone that crops out on the lower Florida Keys is overlain by thin (16 em or less), discontinuous, Holocene calcareous crusts (caliche) that are usually laminated, composed dominantly of calcite micrite and may or may not incorporate part of the underlying limestone. Both allochems and sparry calcite cement in the former unit contain endolithic algae and fungi, borings and unicellular algae. Biogenic structures identical to those in the Miami Limestone also occur in the calcareous crusts but are somewhat less abundant in the latter unit versus the former unit. The calcareous crusts were formed in the vadose diagenetic environment. Some of the CaC03 necessary for the micrite that comprises the bulk of the crusts was probably derived from solution of carbonate from a soil cover and some from wind blown salt spray. Most of the micrite, however, was formed by replacement of the uppermost portions of the Miami Limestone. Replacement involved micri tisation of allochems and a previously unreported process, sparmicritisation, the degrading recrystallization of sparry calcite to micrite. Minor sp;umicritisation was caused by micrite calcification of endolithic fungi or algae within sparry calcite cement or by micrite precipitation in empty borings within such cement. Most sparmicritisation took place by dissolution of sparry calcite and concomitant precipitation of micrite in the space occupied previously by the dissolved spar. Such sparmicritisation is interpreted to be caused by chemical reactions involving the crystals, pore water which is moving slowly but steadily and organic com pounds released during bacterial decomposition of fungi, algae or both. It is recognized that sparmicritisation occurs in the marine diagenetic en vironment and is not, therefore, necessarily indicative of vadose diagenesis. Incomplete sparmicritisation is responsible for some of the clotted textures typically found within calcareous crusts and may explain such textures in many other carbonate rock types. A combination of sparmicritisation and micritisation has probably greatly influenced the porosity of many reefs and, in some cases, led to the formation of 'micritic reefs'.
INTRODUCTION
The original terminology used to describe what is herein termed calcareous crusts on the lower Florida Keys, after James (1972), was subaerial laminated crusts ( Multer Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
207
Charles
F.
Kahle
& Hoffmeister, 1968). Terms sometimes used synonymously with calcareous crusts include ' Nari' ( Krumbein, 1968), caliche ( Krumbein, 1968; James, 1972) and calcrete (James, 1972, p. 828). Holocene and Pleistocene calcareous crusts are important because they are thought to form only in the vadose diagenetic environment ( Multer & Hoffmeister, 1968; Krumbein, 1968; James, 1972; Steinen, 1974). Study of such crusts provides a basis for recognizing calcareous crusts that originated due to subaerial exposure of ancient rocks ( Blank & Tynes, 1965; Gill, 1973; Kahle, 1974a; Walls, Harris & Nunan, 1975). Calcareous crusts form under similar conditions controlled primarily by climate, chemistry and time, but the processes involved in their formation are not completely understood (Walls et a!., 1975). Although algae and fungi are known to occur in some calcareous crusts ( Krumbein, 1968; James, 1972), the nature of these microorganisms in such rocks remains to be established as well as to what extent they contribute to the formation of textures and calcite micrite in crusts. The geology and petrography of the Miami Limestone are described by Ginsburg (1957), Friedman (1964) and Hoffmeister, Stockman & Multer ( 1967). Detailed studies of certain textures in this unit have been made by Schneidermann, Sandberg & Wunder ( 1972) and by Sandberg, Schneidermann & Wunder ( 1973). Calcareous crusts on the lower Florida Keys have been studied previously by Multer & Hoff meister ( 1968) and by James ( 1972). Due to James (1972) the petrography of cal careous crusts is now well known. The present work provides the first detailed published account of the process herein termed sparmicritisation, * the conversion of sparry calcite cement to micrite. One purpose of this report is to document that calcareous crusts on the lower Florida Keys were derived primarily by way of sparmicritisation and micritisation of allo chems in portions of the Miami Limestone within the vadose diagenetic environment. Additional aims are to describe the nature and occurrence of endolithic algae, fungi and unicellular algae in the Miami Limestone and in associated calcareous crusts and to explore the role of these microorganisms in the processes of sparmicritisation and micritisation.
MATERIALS AND METHODS
Samples of the Miami Limestone and calcareous crusts were collected at, or near, the surface of Big Coppit, Sugar Loaf, No Name and Big Pine Keys in the lower Florida Keys. Maximum depth of sampling below ground surface was about 40 em. Covered and uncovered thin sections were studied petrographically, primarily using an oil immersion objective at a magnification of x 1250. Many thin sections were ground to about 10 fLm and etched in 0 ·5% H Cl to facilitate recognition of crystal boundaries and biogenic structures. Rock fragments and uncovered thin sections were decalcified; any remaining organic matter or biogenic structures were studied petro graphically. A I% solution of Malachite Green was used to selectively stain and facilitate recognition of biogenic structures and organic matter ( Kahle, Eutsler & Huh, 1973). About forty rock fragments or uncovered thin sections were coated with gold *The author first used the term sparmicritisation in a paper delivered at the 1974 A.A.P.G. S.E.P.M. meeting in San Antonio, Texas. The term was not given, however, in the abstract of this paper (Kahle, 1974b, p. 51). 208
Origin of calcareous crusts
and examined with a Hitachi H H S-2 R scanning electron microscope ( S E M). The use of fragments of uncovered thin sections permitted comparison between biogenic structures as seen using light microscopy and S E M. Mineral composition was de termined by standard X-ray diffraction techniques, light microscopy, artificial stain ing with Fiegels solution for aragonite, and on the basis of crystal habit as seen with S E M. Semi-quantitative analysis for some elements was carried out using a Kevex energy dispersive X-ray analyser attached to the S E M.
TERMS
The-term micrite (Folk, 1965) is used for calcite crystals less than about 4 fLm in size which typically appear brown and poorly resolved under the petrographic microscope. The term sparmicritisation is proposed to describe the transformation of sparry calcite into micrite by any known or inferred process; such micrite can be termed sparmicrite. In keeping with prior usage ( Bathurst, 1966; Winland, 1968; Friedman & Sanders 197 1; Lloyd, 1971; Margolis & Rex, 1971), the term micritisation is used only to describe the transformation of allochems to micrite by a process that may or may not be specified. In addition, micritisation is considered equivalent to what Purdy ( 1968) terms 'recrystallization to cryptocrystalline carbonate' which he equates to micritisation. In this report the term filament applies to tubular structures lacking calcification, often branching in form, known or inferred to be living, dead or dormant algae or fungi regardless of the mode of occurrence of the structure (see below). A calcified filament is one wherein Ca C03 has replaced the filament or has been precipitated on or within the filament; the latter results in partial or complete encrustation of the filament by crystals of CaC03 ( Schraer, 1970; Schroeder, 1972). Herein, a boring refers to an excavation in a crystal or rock known or interpreted to have been caused by filamentous algae or fungi but which contains no trace of the organism responsible for the boring. An endolith is a compound structure composed of a filament found in, and pro ducing its own boring ( Lukas, 1973). A chasmolith is a fungal or algal filament within a pore space not created by the organism responsible for the filament ( Lukas, 1973). It is possible for two or more filaments of the same microorganism to be both endo lithic and chasmolithic. It is also possible for a single filament to be both endolithic and chasmolithic ( Schroeder, 1972). The term treppelith (from treppe-to trap) is proposed for unicellular algae that are trapped within crystals of sparry calcite cemerit, along crystal boundaries of such cement, or in ooids or micrite and which can not be shown to occur in a pore space.
GENERAL GEOLOGY
The lower Florida Keys comprise a tectonically stable series of islands extending from the area of Big Pine Key southwestward to Key West ( Multer & Hoffmeister, 1968, Fig. 1). Maximum land elevation is about 4 m and the climate is sub-tropical. 209
Charles F. Kahle
Hoffmeister et al. ( 1967) applied the name Miami Limestone to rocks termed pre viously the Miami oolite in south Florida and the Florida Keys. These authors divided the Miami Limestone into an upper oolitic facies and a lower bryozoan facies. Nearly all of the rock that crops out on the lower Keys, except for calcareous crusts, is the oolite facies of the Miami Limestone. This facies is equivalent in age to the Key Largo Limestone reef facies and accumulated as a series of oolite shoals adjacent to, and over the top of, the Key Largo Limestone ( Hoffmeister et al., 1967). The reported thickness of the Miami Limestone in the lower Keys is from about 1·5 to about 7 ·6 m ( Hoffmeister et al., 1967, p. 188). The Miami Limestone in the lower Keys is mainly an oosparite; subordinate rock types include pelsparite, oomicrite, biosparite and bio micrite. Allochems in the Miami Limestone are usually a mixture of aragonite and calcite but may be composed entirely of either mineral. The matrix between allochems varies from sparry to micritic calcite as also noted by Friedman ( 1964, p. 80 1). Calcareous crusts are composed of calcite micrite except for those portions which contain voids or fractures partly or completely filled with sparry calcite cement. Some crusts reach a thickness of about 16 em (Fig. 1); Multer & Hoffmeister ( 1968, p. 183) reported a maximum thickness of crusts of about 6 em. Crusts are usually laminated as seen megascopically (Fig. I) but as seen using light microscopy they tend to have a clotted texture. The crusts are distributed randomly on the surface of the Keys and typically follow the topography. Breccia fragments of the Miami Limestone and of black, micritic carbonate occur in many crusts (Fig. 1; Multer & Hoffmeister, I 968, p. I87). The calcareous crusts, based on 14C age-dating, were formed within the past 4395±90 years according to Multer & Hoffmeister ( 1968). The age-dating of the crusts is suspect, however, since the underlying Miami Limestone has been exposed for at least 80,000 years (Noel James, personal communication).
Fig. I. Polished slab. Oolitic Miami Limestone (Pleistocene) overlain by calcareous crusts (Holocene) containing breccia fragments of the former unit. Big Pine Key. Scale numbered in em. 210
Origin of calcareous crusts NATURE AND OCCURRENCE OF BIOGENIC STRUCTURES AND STRUCTURES RESEMBLING BIOGENIC STRUCTURES
Biogenic structures, the occurrence of which is shown in Table 1 , are: (a) endo lithic or chasmolithic algae in which the filaments may or may not be calcified and borings interpreted to be caused by filamentous algae (Fig. 2a, b, c, d, e); (b) uni cellular algae (Fig. 2f, g, h); (c) endolithic or chasmolithic fungi which may or may not be calcified and borings interpreted to be due to fungi (Figs 3 and 4a, b, c); worms (Fig. 4d); and plant fibres (Fig. 4e).
Table 1. Summary of occurrence of biogenic structures. See text for details. X=present; O=not observed. Numbers in parenthesis give the maximum depth in em that the structures were observed below ground level. Maximum depth of sampling of the Miami Limestone below ground level was 40 em; maximum thickness of the calcareous crusts is 16 em
Miami limestone Type of biogenic structure Endolithic algae calcified by calcite Chasmolithic algae calcified by calcite Chasmolithic algae calcified by aragonite Endolithic algae of unknown composition Algal? borings Treppelithic algae Endolithic and chasmolithic fungi Endolithic and chasmolithic fungi calcified by calcite Chasmolithic fungi calcified by aragonite Fungal? borings Worms Plant fibres
Calcareous crusts
Sparry calcite cement
Allochems
X(l6)
X(40)
X(40)
X(l6)
0
0
0
X(40)
0
X(l6)
0
0
X(l6) X(16) X(l6)
X(40) X(40) X(40)
X(40) X(40) X(40)
X(16)
X(40)
0
X(16)
X(30)
X(30)
X(16) X(16) X(16)
X(40) 0 X(30)
X(40) 0 X(30)
Calcareous crusts and the Miami Limestone contain a variety of spine-like and tube-like structures that may or may not have bulbous nodes (Fig. 4f); they are composed of a mixture of organic matter and calcite. These are considered to be non biogenic structures because their characteristics correspond to non-biogenic structures produced by organic decay in laboratory experiments ( Mc Cunn, 1972). Fungal structures of various types (Table 1) are the dominant biogenic structures in both the Miami Limestone and in calcareous crusts. None of the fungal or algal structures can be further identified as to division, class, genus or species ( Bold, 1973) 211
Fig. 2. Examples of biogenic structures; see text for explanation of terms used to describe structures. (a) Thin section, plane light; longitudinal and cross-sectional views of endolithic algae calcified by calcite micrite in sparry calcite cement. Miami Limestone. (b) SEM micrograph; endolithic algae calcified by calcite. Miami Limestone. (c) Thin section, plane light; same type of structure as shown in Fig. 2b. Miami Limestone. (d) SEM micrograph; chasmolithic algae (upper right to lower left) calcified by aragonite battons (see text) A possible chasmolithic fungi, also calcified by battons, extends from the lower right toward the upper left. Miami Limestone. (e) SEM micrograph; algal endolith of unknown composition in a matrix of micrite calcite. Calcareous crusts. Boring in lower left was presumably made by the same type of algae. Smaller 2-3 Jlm borings in upper right are inter preted to be of fungal origin. (f) Thin section; plane light; treppelithic unicellular algae concentrated along the boundaries (dashed lines) of crystals of sparry calcite cement. Miami Limestone. (g) Thin section, plane light; degraded treppelithic algae associated with the degrading recrystallization of crystals of sparry calcite cement to micrite (sparmicritisation). Miami Limestone. Boundaries of cement crystals where known are shown by solid lines and by dashed lines where approximate. (h) Thin section, plane light; traces of treppelithic algae within calcareous crusts.
Origin of calcareous crusts
because the structures lack sufficient morphological detail ( Rex L. Lowe, William Spencer, Alexander H. Smith, personal communication). Even when algae and fungi preserved in rocks display abundant morphological details, further taxonomic classification is difficult or impossible without morphological and physiological study in cultures ( Bold, 1973; Moore- Landecker, 1972).
Algae
Endolithic algae calcified by calcite. These may contain a central hollow core (Fig. 2a), range in diameter from 5 to 30 J.lm but are mostly between 15 and 20 J.lm and may be straight, curved or branching. Mucilaginous sheaths with identical dimensions and branching charateristics were found by decalcification of rock fragments and un covered thin sections. These algae are similar to algal filaments shown by Perkins & Halsey ( 197 1; Figs 6 and 10) from Carolina shelf sediments. Chasmolithic algae calcified by calcite (Fig. 2b). The average diameter of these structur�s is 20 fLm and they consist of either a single straight or Y-shaped filament. A close correspondence occurs in the appearance of these structures as seen with the S E M (Fig. 2b) and using light microscopy (Fig. 2c). Chasmolithic algae calcified by aragonite (Fig. 2d). Diameters range from about 7 fLm to 10 fLm. Endolithic algae of unknown composition (Fig. 2e). Such structures have smooth surfaces and diameters from 5 fLm to 10 fLm. Algal? borings. These range in diameter from 10 to 20 fLm and average 15 fLm (Fig. 2e). Treppelithic unicellular algae. They are colourless to green to brown and occur in the Miami Limestone in ooids, in cement (Fig. 2f, g) and in calcareous crusts (Fig. 2h). They are typically spherical, ranging in diameter from 1 to 4 fLm with an average diameter of 3 fLm. Most are single spheres but some are in the form of two or three spherical segments separated by one or three medial sections (Fig. 2f). These algae are tentatively identified as the green algae Palmella ( Palmer, 1962) or Chlorococcum ( Bold, 1970), or the blue-green algae once known as Anacystis (Drouet & Daily, 1956) now termed Chlorococcus ( Palmer, 1962). All of these algae have been reported from fresh water and Chlorococcum is a common soil algae ( Bold, 1970).
Fungi
Endolithic and chasmolithic fungi. Such structures are usually in the form of endoliths but may occur as chasmoliths; they range in diameter from 1 to 4 fLm but . average 2 fLm. These are the most common biogenic structures in the Miami Limestone and in the calcareous crusts. They occur ( Table I) in the Miami Limestone as endo liths in crystals of sparry calcite cement (Fig. 3a, b, c), in allochems (Fig. 3d), within cement in calcareous crusts (Fig. 3e) and in the micrite that makes up the bulk of the crusts (Fig. 3b, e). They also occur as chasmoliths in calcareous crusts (Fig. 3f) and in the Miami Limestone. Whether as endoliths or chasmoliths, the structures are typically single, straight, isodiametric bodies except for those that contain apparent conidia or sporangia (Fig. 3a, d, e). Clamp connections, suggestive of Basidiomycetes ( Moore Landecker, 1972, p. 9 1) may occur in some of the structures (Fig. 3d). Evenly to un evenly spaced septa and cells (Fig. 3a, b, c, d) and hyaline walls are visible in some of 213
Fig. 3. Examples of biogenic structures; see text for explanation of terms used to describe structures. (a) Thin section, plane light; fungal endoliths in crystals of sparry calcite cement. Miami Limestone. (b) Thin section, plane light; same type of structures as in Fig. 3a in sparry calcite cement in the Miami Limestone (bottom) and (faintly visible) in calcareous crusts (top). Compare this figure with Golubic, 1969, p. 749, Fig. 3. (c) Thin section, plane light; fungal endoliths (appear as dark lines) and fungal borings (appear as light lines) in sparry calcite cement. Miami Limestone. The somewhat mottled appearing areas in portions of the photograph, as along the top, are areas of brown dis coloration often associated with fungal structures in crystals of sparry calcite cement. (d) Thin section, plane light; endolithic fungal filaments preserved in situ in the interior of an allochem. Miami Limestone. In situ preservation of the structures was accomplished by impregnating uncovered thin sections with Lakeside-70 followed by decalcification. (e) Thin section, plane. light; endolithic fungi with possible conidia or sporangia in sparry calcite cement. Calcareous crusts. Note sparmi critisation (see text) adjacent to some of the endoliths and the faint traces of endoliths in the micrite surrounding the former pore space. (f) SEM micrograph; chasmolithic fungi of unknown composition in calcareous crusts. Note corroded relics of sparry calcite cement from the Miami Limestone. (g) SEM micrograph; combined chasmolithic and endolithic fungi calcified by calcite micrite. Calcareous crusts. Note corroded relics of sparry calcite cement from the Miami Limestone.
Origin of calcareous crusts
the structures using light microscopy (Dr Alexander H. Smith, personal communica tion). Many of the structures contain a dark brown refractory material (Fig. 3b) presumed to be of organic origin. These fungal structures (Fig. 3a-f ) closely resemble several genera of the class Hyphomycetes as described by Kendrick & Carmichael (I973), especially the genus Humicoloa which is common in soils (Gilman, I957, p. 327). Combined endolithic and chasmolithic fungi calcified by calcite (Fig. 3g). These average 2 I.Lill in diameter and may represent the calcified equivalents of the structures described directly above. Fungal chasmoliths calcified by aragonite (Fig. 4a). Such structures are calcified by crystals of aragonite morphologically identical to aragonite crystals that Loreau & Purser ( I973) refer to as battons in ooids in the Persian Gulf. The individual aragonite crystals have lengths from 2·0 to 0·2 [LID and widths from 0· 14 to 0·4 [LID but an average length/width ratio of I ·0-0·3 [LID. In the Miami Limestone, battons calcify algal (Fig. 2d) and fungal (Fig. 4a) chasmoliths, occur infrequently as cement(Figs 2d and 4a), and comprise portions of laminae in ooids. Battons also make up less than I% of the matrix of calcareous crusts. Fungal? borings (Fig. 4b, c). The diameter and configuration of these structures is similar or identical to the characteristics of fungal structures described above and such structures are considered, therefore, to be of fungal origin. Many of the fungal? borings are empty (Fig. 3b) whereas others are partially (Fig. 4c) or completely filled with crystals of calcite micrite. Worms
Features interpreted as worms (Fig. 4d) are characterized by one smooth side that may be longitudinally striated and by an opposed segmented side. The structures range in diameter from about 20 [Lm to 30 [LID. Plant fibres
The characteristics of these structures (Fig. 4e) are identical to known plant fibres (Fahn, 1967, Fig. 39).
ALTERATION OF MIAMI LIMESTONE INTO CALCAREOUS CRUSTS Micri tisation
Nearly all allochems in the Miami Limestone show some evidence of micritisation. The amount of micritisation is usually greatest within the uppermost 3-5 em of the Miami Limestone in contact with calcareous crusts (Fig. 5a). Within this interval, termed the zone of alteration, micritisation of allochems is greater than in allochems below this zone. Allochems within this zone tend to show more micritisation the closer they are to calcareous crusts (Fig. 5a, b). All stages of transition involved in micritisation occur proceeding upward through the zone of alteration into the margin of calcareous crusts (Fig. 5a, b) and then into the main body of crusts (Fig. 5c, d). In some cases, half of a heavily micritised ooid occurs in the Miami Limestone while the other half, showing even more evidence of micritisation, occurs within calcareous crusts (Fig. 5a). Micritisation of ooids within the zone of alteration or within calcareous 215
Charles F. Kahle
Fig. 4. Examples of biogenic structures (a- e) and nonbiogenic structures (f). (a) SEM micro
graph; chasmolithic fungi calcified by aragonite battons (see text). The walls of the pore (upper left and lower right) are lined by cement made up of battons identical to those which calcify the en dolithic fungi. Miami Limestone. (b) SEM micrograph; fungal borings in calcareous crusts. (c) SEM micrograph; fungal boring partially lined with crystals of calcite micrite in a crystal of sparry calcite cement. Miami Limestone. (d) SEM micrograph; structures interpreted as worms; calcareous crusts. (e) SEM micrograph; plant fibres in calcareous crusts. Insert shows the same type of structure as seen in thin section with a petrographic microscope. Maximum width of the structure shown in the insert is 20 J.lm. (f) Decalcified thin section; plane light; non-biogenic structures. Miami Limestone. Com pare this figure with McCunn, 1972, Fig. 7.
216
Fig. 5. Nature of textural alteration associated with the transformation of the Miami Limestone into calcareous crusts. (a) Thin section, plane light. Zone of alteration (see text) in the topmost portion of the Miami Limestone in contact with calcareous crusts showing micritisation of allochems and sparmicritisation of sparry calcite cement (faintly visible between allochems). (b) Thin section, plane light. Contact (dashed) between Miami Limestone and overlying calcareous crusts. Note relics of allochems and of sparry calcite cement from the Miami Limestone in the calcareous crusts. (c) Thin section, plane light. Relic of allochem from the Miami Limestone in calcareous crusts. Note merging of micrite created by micritisation of allochem with micrite making up the matrix of surrounding calcareous crusts. Ooid contains numerous endolithic fungi. (d) Thin section, plane light; 'ghosts' of ooids from the Miami Limestone in calcareous crusts. (e) Thin section, plane light. Portion of ooid rim that has undergone aggrading recrystallization. Miami Limestone. Boundaries of crystals of sparry calcite indicated by dashed lines. The intrusion of patches of micrite into crystals of sparry calcite (I), patches of micrite that cross-cut such crystals and their boundaries (D), and the presence of rhomboid-shaped patches of micrite (R) indicate that sparmicritisation occurred after aggrading recrystallization. (f) Thin section, plane light. Sparmicritisation concentrated primarily along the boundaries of crystals of sparry calcite cement; Miami Limestone. Portions of some crystals (arrows) show a brown discoloration. Fungal endoliths appear as dark lines some of which branch. P is a pore.
Charles F. Kahle
crusts may be initiated anywhere within ooids. Alternatively, micritisation may begin around the periphery of ooids (Fig. Sc) and gradually move inward until traces of the ooids are nearly eliminated (Fig. Sd). Most allochems, but especially ooids within the zone of alteration and within calcareous crusts, contain endolithic fungi (Fig. Sc).
Aggrading recrystalli zati on
Such recrystallization is rare in both the Miami Limestone and in calcareous crusts. In the Miami Limestone aggrading recrystallization is confined almost entirely to allochems, especially ooids (Fig. Se).
Sparmicritisation
Sparmicritisation occurs within both the Miami Limestone (Figs Se, f; 6a, b, c, d) and in calcareous crusts (Fig. 6e, f ). In the Miami Limestone, sparmicritisation is un common outside of the zone of alteration (Fig. Sa, b, f ) noted above and, within this zone, is initiated at different sites. Sparmicritisation may begin in ooids that have previously undergon_e aggrading recrystallization (Fig. Se). Within sparry calcite cement in the Miami Limestone, sparmicritisation is initiated typically along crystal boundaries and spreads inward until portions or all of such crystals are transformed into micrite (Figs Sf; 6a, b). Sparmicritisation begins frequently along the contact between calcareous crusts and the Miami Limestone by way of finger-like projections of micrite that extend away from the crusts along crystal boundaries of sparry calcite within the Miami Limestone (Fig. 6a). Continuation of the process leads to transforma tion of the crystal margins to micrite and, eventually, to complete sparmicritisation (Fig. 6a, b). Complete sparmicritisation may result in triangular areas of micrite along the margins of crusts (Fig. 6b) that retain the outline of crystal faces of the original sparry calcite such as the crystal shown in the middle of Fig. S. Like micritisation (Fig. Sc, d), sparmicritisation does not occur entirely within the Miami Limestone; it also occurs in calcareous crusts within pore cements (Fig. 6e) and in cement in fractures (Fig. 6f ). That sparmicritisation does not always go to completion within the alteration zone (Fig. Sa) is shown by the occurrence in calcareous crusts of relics of sparry calcite cement from the Miami Limestone (Figs 3f, g and 6b). Endolithic algae and fungi, algal and fungal borings and treppelithic algae within sparry calcite cement are often accompanied by discolouration of all or parts of the white or cream coloured spar crystals to shades of light brown (Fig. 3c and Sf ). This phenomenon may or may not be associated with sparmicritisation. Where sparmicrite is adjacent to areas of discolouration the former often intrudes into or cross-cuts the areas of discolouration; the reverse relationship is less common. This suggests that the sparmicritisation and discolouration can occur either concurrently or sequentially. Most, but not all cases of sparmicritisation, in both the Miami Limestone and in calcareous crusts are associated with endolithic fungi (Figs 3, Sf and 6b, e, f ) and, based upon petrographic examination of uncovered thin sections to which malachite green stain had been applied, with organic matter. Areas of sparry calcite that display different degrees of sparmicritisation are richer in organic matter than areas of such cement which have not undergone any sparmicritisation. An estimated S% of all sparmicrite has replaced organic matter. 218
Fig. 6. Sparmicritisation in the Miami Limestone and in calcareous crusts. (a) Incipient sparmicritisa tion represented by 'spikes' of micrite extending away from calcareous crusts along the boundaries of sparry calcite cement in the Miami Limestone (near centre). The 'spikes' of micrite grade into a more advanced stage of sparmicritisation represented by crystals of sparry calcite enveloped by micrite (lower left) that grade, in turn, into a patch of sparmicrite in the Miami Limestone (bottom centre). (b) Thin section, plane light; calcareous crusts on left and Miami Limestone on right. Nearly com plete sparmicrrtisation of a former crystal of sparry calcite cement in the Miami Limestone (tri angular area of micrite left of centre). Note areas of sparmicritisation along the boundaries of crystals of sparry calcite and projecting away from the former into the centre of the latter. Fungal endoliths appear as dashes or dashed lines. (c) SEM micrograph. Crystals of sparry calcite cement that have undergone partial sparmicritisation in the alteration zone (see text) in the Miami Limestone. (d) SEM micrograph; same as (c) except a different part of the alteration zone. Hole (H) is a possible algal boring. (e) Thin section, plane light. Pore within calcareous crusts that was filled by sparry calcite cement that was subsequently bored by fungi. Each of the finger-like areas of sparmicrite contains one or more fungal endoliths aligned parallel or sub-parallel to the length of the 'finger'. The sparmicrite occurs within or directly adjacent to the endoliths. (f) Thin section, plane light. Fracture within crusts that contains a mixture of sparry calcite cement and sparmicrite. All areas of sparmicrite contain abundant fungal endoliths; such endoliths are absent or few in number in the sparry cement.
Charles
F.
Kahle
INTERPRETATIONS
Origin of calcareous crusts
Reasons why calcareous crusts on the lower Florida Keys, and those elsewhere, form in the vadose environment are covered by Blank & Tynes ( 1965), Multer & Hoffmeister ( 1967) and James ( 1972) and are not repeated here. Formation of cal careous crusts is restricted to the air-rock interface and close by and never involves the main part of the vadose zone (Noel James, personal communication). Key issues in considering the origin of calcareous crusts on the lower Florida Keys involve the origin of micrite that makes up most of the crusts and the origin of clotted textures and laminations that are common attributes of the crusts. The most important evidence bearing on the origin of the micrite that makes up the bulk of the calcareous crusts is considered to be the following. (I) Portions of calcareous crusts commonly intrude into the Miami Limestone and the crusts contain fragments of the. limestone (Fig. 1). (2) The uppermost portion of the Miami Limestone contains irregularly shaped areas of micrite from a fraction of I mm to about 4 em in maximum length (Fig. 1). The micrite in these areas can often be observed using a hand lens or a microscope to intrude into or cross-cut allochems in the surrounding limestone. (3) Allochems and sparry calcite cement in the outermost portion of most breccia fragments of the Miami Limestone in calcareous crusts (Fig. I) are partially altered to micrite whereas allochems and sparry calcite cement within the interior of such fragments show less or no such alteration. (4) The amount of micritisation and sparmicritisation increases upward in the uppermost 3-5 em of the Miami Lime stone (Fig. 5a, b). (5) Calcareous crusts contain abundant ghosts of ooids (Fig. 5c, d) and other allochems. The ooids and other allochems did not originate in the crusts. They often retain sufficient detail in terms of size and structure (Fig. 5c) to demonstrate that they are relics of portions of the Miami Limestone (Fig. 2a) transformed into calcareous crusts. (6) Calcareous crusts contain many crystals of sparry calcite that have undergone different degrees of sparmicritisation and are surrounded, or nearly so, by micrite (Fig. 3f, g). Many of these crystals retain morphological characteristics (Fig. 3g) which can be matched in every detail to crystals of sparry calcite cement in the underlying Miami Limestone. (7) It can be demonstrated that all stages of grada tion occur between micritisation of allochems and sparmicritisation using the Miami Limestone as one end member and calcareous crusts as the other end member (Figs 2f, g, h; 5; 6). At least some of the CaC03 necessary for the micrite making up the calcareous crusts was probably derived from an outside source such as leaching of soils or from salt spray ( Multer & Hoffmeister, 1968; James, 1972). But based upon evidence pre sented previously and summarized in points 1-7 above, the author believes that the great majority of micrite comprising the calcareous crusts resulted from replacement of the Miami Limestone by micrite as a result of micritisation and sparmicritisation. Thus, as opposed to an outside source of CaC03 ( Blank & Tynes, 1965, p. 1390), formation of crusts mainly involved a local source of CaC03 derived from dissolution of allochems and sparry calcite cement comprising the Miami Limestone. A combination of light microscopy and S E M shows that clotted textures in calcareous crusts as seen using light microscopy are due to: (a) incomplete sparmi critisation (Figs 5a and 6c), (b) differences in the size of micrite resulting from spar220
Origin of calcareous crusts
micritisation (Fig. 6c, d), (c) the occurrence of algal and especially fungal endoliths in sparry calcite cement in calcareous crusts that originally were endoliths in sparry calcite cement in antecedent Miami Limestone (Fig. 3b), (d) borings· in crusts (Fig. 4b), and (e) the irregular distribution within crusts of 'clumps' of organic matter, some of which are similar or identical in appearance (Fig. 4f ) to structures produced by organic decay in laboratory experiments ( McCunn, 1972, Figs 4 and 7). Laminations in calcareous crusts are due to organic staining rather than to differ ences in the size of crystals within different laminae, as also noted by Multer & Hoffmeister ( 1968). If the organic staining was somehow present initially in the Miami Limestone, the laminations represented by the staining would have been destroyed during the transformation of the rock into calcareous crusts by way of micritisation and sparmicritisation. The staining must have occurred either during or after the time that the calcareous crusts were being formed. A possible model for the staining is provided by laboratory experiments in which decay of organic matter in a high calcium carbonate brine produced laminations in sediment ( Mc Cunn, 1972, Fig. I), apparently due to staining by organic compounds, that closely resemble the laminations in calcareous crusts (Fig. I). Sparmicritisation occurred within the crusts as well as participating in the altera tion of parts of the Miami Limestone into crusts as noted previously. Within some pores and fractures filled with sparry calcite cement within crusts, sparmicritiSation proceeded sufficiently to cause merging of micrite produced by this process with micrite formed previously in adjacent portions of the crusts (Figs 3e and 6e, f ). Sparmicritisation probably obliterated many sparry calcite filled pores and fractures within crusts.
General nature of sparmicritisation
An estimated 5% of all sparmicritisation is caused by boring of sparry calcite cement by fungi or algae followed by calcification of the algae (Fig. 2a) or fungi (Fig. 3g) or by filling of empty borings by micrite (Figs 3c and 4c). The mechanism of both algal ( Golubic, 1969) and fungal ( Silverman & Munoz, 1970, p. 985) boring of sparry calcite is one in which the organisms produce acid that dissolves the calcite. What caused subsequent calcification of algae and fungi in the samples is unknown. It could have been caused by the matabolic processes of the organisms or by conditions external to the organisms which may or may not have been influenced by the meta bolism of the borer ( Schraer, 1970; Schroeder, 1972; Alexandersson, 1974). Likewise, it is unknown what caused micrite filling of empty algal and fungal borings. In either case, the steps represented by the dissolution of the calcite and the formation of micrite probably did not occur concurrently. Kobluk ( 1975; and personal communica tion) has shown that after endolithic algae initially infest crystals of sparry calcite, the algal filaments are not calcifei d by micrite until 8 weeks later, or longer. Assumedly, the same observation would apply to endolithic fungi. It is inferred from Bathurst ( 1966) that the boring-filling mode of sparmicritisation would require days or years between the time of initial boring of sparry calcite and filling of the empty bores by micrite. About 95% of all sparmicritisation involves partial to complete dissolution of crystals of sparry calcite and precipitation of micrite in the space occupied previously by the dissolved portion(s) of the crystals. In examples of this type of sparmicritisation, 221
Charles
F.
Kahle
use of the X-ray analyser attached to the SE M shows that portions of crystals of sparry cement that have not undergone sparmicritisation have the same elemental 'signature' for elements besides calcium as the micrite which replaced parts of the crystals (Fig. 6c). In calcite crystals in different stages of this type of sparmicritisation, it is impossible on an ultrascale to distinguish between small areas of calcite leaching and of calcite precipitation (Figs 3f and 6c, d; 7). The mechanisms of dissolution and precipitation must occur concomitantly. For the sake of brevity this type of sparmicritisation is referred to below as CD P sparmicritisation. Concomitant dissolution-precipitation (CDP) sparmicritisation
This type of sparmicritisation must involve ground water because calcareous crusts are formed in the vadose diagenetic environment ( James, 1972). It may also in volve the metabolism of algae, decomposition of organic matter and kinetic factors, as discussed below. Metabolism of algae
Based upon experiments by Schmalz & Swanson ( 1969) and the reasoning of Alexandersson ( 1974) algal metabolism could, in theory, lead to CDP sparmicritisa tion. In considering this idea, it should be noted that the combined volume of fila mentous and unicellular algae is estimated to be about 0·5% of the solid volume of the Miami Limestone and even less for calcareous crusts. This volume of algae could con ceivably have been responsible for a far greater volume of metabolic C02 ( Alexander sson, 1974). Also, the original volume of algae in the Miami Limestone and in calcareous crusts may have been far greater than is now apparent. But, many of the green unicellular algae in the Miami Limestone are presumably alive (Fig. 2f ), even if dormant, and such algae are almost never associated with examples of CDP sparmi critisation. Also, this type of sparmicritisation is not associated with filamentous algae. Because of these relationships, it appears unlikely that algal metabolism played any role in CD P sparmicritisation. Decomposition of organic matter
An intimate relationship occurs between the concentration of all of the following or some combination thereof with CDP sparmicritisation: (a) algal and especially fungal endoliths (Figs 2a, 3 and 5f ), (b) non-biogenic structures (Fig. 4f) similar or identical in appearance to non-biogenic structures known to result from decomposi tion of organic matter in laboratory experiments ( Mc Cunn, 1972, Fig. 7), and (c) organic matter. These relationships do not prove a cause and effect relationship but they do provide evidence for the idea that decomposition of organic matter can some how bring about chemical changes in ground water that can cause CDP sparmicritisa tion. The validity of this idea is enhanced by the observed replacement of some organic matter by sparmicrite generated by CDP sparmicritisation. Direct evidence that decomposition of organic matter is involved in CDP sparmicritisation is provided by unicellular algae. If such algae have not undergone decomposition they are green in colour, display good morphological detail and are not associated with CDP sparmi critisation (Fig. 2f ). A direct correlation occurs between decomposition of such algae, manifested by a change in colour from green to progressively darker shades of brown and progressive loss of morphological detail versus increasing amounts of CD P 222
Origin of calcareous crusts
Fig. 7. SEM micrographs showing sparmicritisation. (a) Relic crystal of sparry calcite cement from the
Miami Limestone in calcareous crusts. A portion of the crystal has undergone sparmicritisation (area outline by dashed line). Crystal is located about 50 J..lm above the contact between the Miami Lime stone and overlying calcareous crusts. Sparmicritisation of this crystal is interpreted to have involved a semi-continuous process of concomitant dissolution-precipitation whereby a dissolution 'front' followed by a zone in which backfilling occurred in the form of micrite precipitation, moved part way through the crystal. The planar surfaces (P) above the crystal surface appear to be the walls of a mold. This mold may have formed by dissolution of a portion of the crystal shown or by way of dissolution of another crystal that occurred above the crystal shown. (b) Close-up of the central portion of Fig. lOa.
sparmicritisation (Fig. 2g, h). The initial stages of such sparmicritisation (Fig. 2g) are very similar to experimentally produced corrosion along cleavage traces of sparry calcite crystals resulting from bacterial decomposition of algae in an aqueous medium (Wagner & Schwartz 1965, Fig. 10). Based upon the above and work by Purdy (1963b, 1968) there now seems suffi cient evidence, both direct and indirect, to conclude that in aqueous media bacterial decomposition of organic matter can cause concomitant dissolution-precipitation 223
Charles
F.
Kahle
type sparmicritisation. Probably a series of chemical reactions are involved, triggered initially by decomposition of organic matter, as suggested by Purdy (1963b, p. 486) but the chemistry of the reactions in nature is unknown. Ammonia and ammonium are involved in the reactions noted by Purdy (1963b, p. 486). The results of experiments by Paul & Curtin (1975, Figs I and 2) involving brownish discolouration of organic crystals by ammonia gas are strikingly similar to examples of brownish discolouration associated with fungal endoliths or borings, or both, in crystals of sparry calcite cement (Figs 3c and Sf ). Although boring of crystals of sparry calcite by fungal or algal endoliths does not, in itself, cause COP sparmicritisation, it is suggested that endoliths play a role in this type of sparmicritisation in the following ways. First, endoliths increase the surface area of spar crystals making it statistically more likely that ground water can come into contact with the crystals. It could be argued that ground water would be unable to pass in and out of living endoliths but this might be accomplished by diffusion ( Bold, 1973; Schraer, 1970). Alternatively, living endoliths can create porosity located be tween themselves and the walls of their borings for ground water to migrate into spar crystals. S E M shows that the diameter of borings made by endoliths is often wider than the diameter of the endoliths themselves (Fig. 2e). Probably, however, ground water movement through areas occupied by endoliths is greater after the endoliths have died and undergone partial or complete decomposition versus living endoliths. Second, dead endoliths provide the major source of organic matter involved in CO P sparmicritisation. Third, bacterial decomposition of organic matter is enhanced by endoliths. Bacteria, carried by ground water, could gain access to organic matter by way of methods outlined under point one above. An alternative hypothesis is that living endoliths are hosts for bacteria and, after the endoliths infest spar crystals and die, the bacteria cause decomposition of the dead endoliths. Kinetics
Kinetics are thought to play an important role in vadose diagenesis, especially the dissolution-precipitation mechanisms involved in the recrystallization of aragonite to calcite ( Schmalz, 1967; Matthews, 1974). This type of recrystallization will not occur if ground water movement is nonexistent, too slow or too fast but will occur if water movement is slow but steady ( Matthews, 1974). Comparison suggests that slow but steady movement of ground water is also required for COP sparmicritisation, in addition to decomposition of organic matter noted previously. It is also possible that organic compounds derived from decomposition of organic matter could affect the kinetics of CO P sparmicritisation by inhibiting or accelerating certain reactions ( Jackson & Bischoff, 1971). Micritisation
Using modern Bahamian marine allochems as an analogue ( Purdy, 1963a, b, 1968; Bathurst, 197 1), allochems in the Miami Limestone underwent various degrees of micritisation in the marine environment in which the allochems were formed ( Hoff meister et a!., 1967). Marine micritisation can probably be caused by a number of pro cesses which are still not understood ( Bathurst, 1966; Purdy, 1963a, 1968). Blank & Tynes ( 1965) and James ( 1972) have shown previously that micritisation of allochems occurs in association with the formation of calcareous crusts in the 224
Origin of calcareous crusts
vadose diagenetic environment. Vadose micritisation is most apparent in those portions of the Miami Limestone referred to previously as the alteration zone com prising the uppermost 3- S em of this unit immediately adjacent to calcareous crusts (Fig. Sa, b). Within this zone the amount of micritisation and sparmicritisation typically increase sympathetically toward calcareous crusts. Many of the same fungal endoliths, calcified fungal filaments and fungal borings within this zone penetrate both sparry calcite cement and allochems. Thus, micritisation and sparmicritisation within the zone of alteration appear to have occurred concurrently. Although many allo chems were micritised within the alteration zone (Fig. Sa), micritisation of some allochems occurred after the allochems were incorporated into calcareous crusts (Fig. Sc, d). Evidence from both light microscopy and S E M shows that vadose micritisation of allochems within the zone of a lteration and within calcareous crusts was only rarely caused by the boring-filling mode of micritisation that occurs in some allochems in the marine environment ( Bathurst, 19 66). Instead, the great majority of micritisation of this type appears to have be�n caused by the same processes involved in concomitant dissolution-precipitation sparmicritisation, discussed previously.
EXTRAPOLATION OF FINDINGS
The fact that both algae and fungi can bore into crystals of sparry calcite cement in the vadose diagenetic environme nt as shown in this report is not surprising, except possibly in terms of the environment in which the borings occurred. It has been shown previously that algae and fungi are capable of boring into calcite crystals of sparry calcite cement, Iceland Spar, and spar-sized calcite crystals within fossils in environments rangin g from peritidal to deep marine ( Kohlmeyer, 1969; Fisher & Garrison, 19 67; Goubic, 19 69; Gatrall & Golubic, 1970; Rooney & Perkins, 1972; Perkins & Tsentas, 1973). In none of these cases, however, did the author(s) allude to what Is herein termed sparmicritisation. David Kobluk (personal communication) has shown to the author examples of sparmicritisation associated with algal borings of crystals of Iceland Spar in which the crystals were immersed in artificial sea water or in natural sea water in the Caribbean. Additional examples of sparmicritisation may be recorded in Pleistocene limestone flakes from the Bahamas (Purdy, 19 63a, Plate 4c), in Jurassic algal limestones from Scotland ( Hu dson, 1970, Fig. 7b) and in certain Upper Carboniferous rocks from Spain ( De Meijer, 197 1, Plate 1, Fig. 4). Sparmicritisation is probably more common than realized previously in carbonate sediments and rocks regardless of their depositional and diagenetic history. Algal and fungal endoliths in sparry calcite cement in carbonate rocks in the geologic record may be more common than realized previously for two main reasons . First, they are easily overlooked because of their small size. Second, algal and fungal endoliths have been portrayed as being associated entirely with allochems ( Bathurst, 19 66, 197 1; Friedman et a!., 197 1; and many others), and not with cement. Sparmicritisation is not confined to calcareous crusts on the lower Florida Keys. Blank & Tynes ( 19 6S) showed that degrading recrystallization of sparry calcite cement, herein termed sparmicritisation, was a key process involved in the formation of calcareous crusts from the Edwards Limestone ( Cretaceous) in west-central Texas. It is predicted that sparmicritisation will be discovered to have played a major role in the 225
Charles F. Kahle
origin of calcareous crusts besides those described herein and by Blank & Tynes ( 1965). Considering that algae, and possibly fungi , are known to be abundant in many reefs (Friedman et al. , 197 1 ; 1974, p. 822) and that reefs can be affected by both marine and non-marine environments ( Schroeder , 1972; Gill 1973), reefs may con situte ideal places for sparmicritisation. Many 'micritic reefs ' may be due to a com bination of micritisation and sparmicritisation. Furthermore , sparmicritisation may have been involved in the evolution of many pelsparites , pelmicrites and carbonate rocks displaying a clotted texture ( Bathurst , I97 I).
CONCLUSIONS
From outcrops on the lower Florida Keys, rock samples were collected from (a) the uppermost 40 em of the Miami Limestone where this unit is not overlain by c:l careous crusts, (b) calcareous crusts which range in thickness from about 1 em to I6 em and which form a discontinuous cover over the Miami Limestone and (c) the uppermost 1-39 em of the Miami Limestone below calcareous crusts. Detailed study of these samples suggests : ( 1) Calcareous crusts on the lower Florida Keys have been bored intensively by endolithic fungi and to a lesser extent by endolithic algae. A criterion useful in dis tinguishing calcareous crusts from algal stromatolites may be that the former typically contain more fungal structures than do the latter. (2) Brownish discoloration of sparry calcite cement occurs in the vadose dia genetic environment , probably due to reactions between the crystals and organic gases. Additional work may show that the colour of sparry calcite cement is a useful guide to reconstructing diagenetic environments. (3) Sparmicritisation , the transformation of sparry calcite cement to micrite, in volves different mechanisms. Minor sparmicritisation is due to boring of sparry calcite cement by fungal or algal filaments followed by micrite filling of the empty borings created after the filaments decompose, or by way of micrite calcification of the organisms responsible for the borings. Most sparmicritisation is caused by con comitant dissolution-precipitation. These mechanisms result , in turn , from the co m bined effects of (a) slow but steady movement of ground water , downward due to gravity and upward due to capillary movement and (b) chemical changes in the ground w�ter caused by chemical compounds liberated from the bacterial decom position of organic matter. The present work supports ideas alluded to by Purdy ( 1963a , b; 1968 , p. 192) that decomposition of organic matter can provide the key ingredients necessary for degrading recrystallization of CaC03 and that many car bonate textures ascribed to aggrading recrystallization may , in fact , represent de grading recrystallization instead. (4) Calcareous crusts on the lowe r Florida Keys were formed in the shallow vadose environment , primarily by the in situ micritisation of allochems and by sparmicritisation of cement in the Miami Limestone. A much smaller volume of the micrite in calcareous crusts may have resulted from precipitation of CaC03 derived from salt spray or leaching of soil , or both. (5) Sparmicritisation can occur in marine diagenetic environments besides the vadose diagenetic environment and is probably responsible for textures found in some 226
Origin of calcareous crusts
pelsparites and pelmicrites as well as clotted textures found in many carbonate rocks. Sparmicritisation, along with micritisation, may account for many 'micritic reefs'. (6) Attention has been focused previously upon the congruent ideas that the primary effect of diagenesis in the vadose diagenetic environment is the achievement of mineralogic stability (Friedman, 1964; Mathews, 1974) and that recrystallization due to dissolution-precipitation mechanisms is only possible in this environment if more than one carbonate phase is present ( Schmalz, 1967; Matthews, 1974). Within this environment, however, significant diagenetic changes can occur after mineralogic stability has been achieved, and involve the presence of only a single carbonate phase, calcite. Dissolution-precipitation mechanisms in the vadose diagenetic environment can cause not only the recryst allization of aragonite to calcite, but also the degrading recrystallization of calcite to calcite.
ACKNO WLEDGMENTS
The author thanks Robert Eutsler, Tim Jackson, Ron Martin, Walter Reas, Ron Shaw and Fred Siegel for providing study material; the following members of the Department of Biology, B owling Green State University: Rex L. Lowe and Robert Romans for help in identifying biogenic structures and Richard C. Crang and Pat Ansbaugh for, respectively, making a scanning electron microscope available and for operating the instrument; William Spencer, who was a graduate student in the afore mentioned department at the time work was being carried out for this study, for helping me to learn more about algae and fungi; Alexander H. Smith, Department of Botany and University Herbarium, University of Michigan, for listening to some of my ideas contained in this paper and for clarifying for me the algal or fungal affinities of some of the biogenic structures and David R. Kobluk of Mc Master University for sharing with me the results of his research on algae. I greatly appreciate critical reviews of manuscript drafts by Louis I. Briggs, Noel P. James, D avid R. Kobluk and four unnamed reviewers. I am grateful to the Graduate Research Committee of Bowling Green State University for providing funds for some of the research.
REFERENCES ALEXANDERSSON, T. (1974) Carbonate cementation in coralline algal nodules in the Sakagerrak,
North Sea. J. sedim. Petrol. 44, 7-26. BATHURST, R.G.C. (1966) Boring algae, micrite envelopes and lithification of molluscan biosparites. Geol. J. 5, 1 5-32.
BATHURST, R.G.C. (1971) Carbonate Sediments and their Diagenesis, pp. 620. Developments in Sedimentology, 12. Elsevier, Amsterdam, London, New York.
BLANK, H.R. & TYNES, E.W. (1965) Formation of caliche in situ. Bull. geol. Soc. Am. 76, 138 7-1392. BoLD, H.C. (1970) Some aspects of the taxonomy of soil algae. Ann. N. Y. Acad. Sci. 175, 60 1-616. BoLD, H. C. (1973) Morphology of Plants. Harper & Row, New York. DE MEIJER, J.J. (1971) Carbonate petrology of algal limestones (Lois-Ciguera Formation), Upper
Carboniferous, Leon, Spain. Leidse geol. Med., 47, 1-97. DROUET, F. & DAILY, W.A. (1956) Revision of the Coccoid Myxophyceae. Butler Univ. Botanical Studies 12, 218 pp.
FAHN, A. (1967) Plant Anatomy. Pergamon Press, Oxford, England. 227
Charles F. Kahle A . G. & GARRISON, R.E. (196 7) Carbonate lithification on the sea floor. J. Geol. 75, 488-496. R.L. (1965) Some aspects of recrystallization in ancient limestones. In : Dolomitization and Limestone Diagenesis (Ed. by L. C. Pray and R. C. Murray). Spec. Pubis Soc. econ. Paleont. Miner., Tulsa, 13, 14-48. FRIEDMAN, G.M. (1964) Early diagenesis and lithification in carbonate sediments. J. sedim. Petrol. 34, 777-813. FRIEDMAN, G .M . & SANDERS, J.E. ( 1971) Micrite envelopes of carbonate grains are not exclusively of photosynthetic algal origin. Sedimentology, 16, 89-96. FRIEDMAN, G .M., AMIEL, A.J. & SCHNEIDERMANN, N. ( 1974) Submarine cementation in reefs : Ex amples from the Red Sea. J. sedim. Petrol. 44, 8 16-825. GATRALL, M. & GoLUBIC, S. ( 1970) Comparative study on some Jurassic and Recent endolithic fungi using scanning electron microscope. Geol. J. Spec. Issue, 3, 167-178. GILL, D. ( 1973) Stratigraphy, facies, evolution and diagenesis of productive Niagaran Guelph reefs and Cayugan sabkha deposits, the Belle River Mills gas field, Michigan basin. Final Research Report, Mich. Gas Assoc., 282 pp. GILMAN, J . C . ( 1957) A Manual of Soil Fungi. Iowa State Univ. Press. Ames, Iowa. GINSBURG, R.N. (1957) Early diagenesis and lithification of shallow-water carbonate sediments in south Florida. In : Regional Aspects of Carbonate Deposition (Ed. by R. J. Le Blanc and J. G. Breeding). Spec. Pubis. Soc. econ. Paleont Miner., Tulsa, 13, 80-100. GoLUBIC, S. (1969) Distribution, taxonomy and boring patterns of marine endolithic algae. Am. Zoologists, 9, 747-751. HOFFMEISTER, J.E. , STOCKMAN, K.W. & MULTER, H. G. ( 1967) Miami Limestone of Florida and its Recent Bahamian counterparts. Bull. geol. Soc. Am. 78, 175- 190. HuDSON, J.D. (1970) Algal limestone with pseudomorphs after gypsum from the Middle Jurassic of Scotland. Lethaia, 3, 11-40. JACKSON, T.A. & BISCHOFF, J.L. ( 1971) The influence of amino acids on the kinetics of the recrystal lization of aragonite to calcite. J. Geol. 79, 493-497. JAMES, N.P. (1972) Holocene and Pleistocene calcareous crust (caliche) profiles : criteria for subaerial exposure. J. sedim. Petrol. 42, 8 1 7-836. KAHLE, C.F., EuTSLER, R. & HuH, J.M.S. ( 1973) Nature and significance of borings and other structures on and within oolite allochems. Bull. Am. Ass. Petrol. Geol. 57, 787. KAHLE, C. F. ( 1974a) Nature and significance of Silurian rocks at Maumee quarry, Ohio. In : Silurian Reef-Evaporite Relationships (Ed. by R. V. Kesling). Mich. Basin Geol. Soc., pp. 31-54. KAHLE, C. F. ( 1974b) Algae, fungi and transformation of sparry calcite to micrite. Abs. Prog. Am. Ass. Petrol Geol.-Soc. econ. Paleont. Miner. , Tulsa , pp. 5 1. KENDRICK, W.B. & CARMICHAEL, J.V. ( 1973) Hyphomycetes. In : The Fungi: An Advanced Treatise (Ed. by G . C. Ainsworth, F. K. Sparrow and A. S. Sussman), Vol. IVA, pp. 323-509. Academic Press, New York.
FISHER, FoLK,
KoBLUK,
D.R. ( 1975) Destruction and alteration of carbonates by boring (endolithic) algae. Abs.
Prog. geol. Soc. Am. 7, no. 6, 798. KoHLMEYER, J.
( 1969) The role of marine fungi in the penetration of calcareous substances.
Am. Zoologists, 9, 741-746. KRUMBEIN,
W. E. (1968) Geomicrobiology and geochemistry of the 'Nari-Lime-Crust' (Israel). In :
Recent Developments in Carbonate Sedimentology in Central Europe (Ed. by G. Muller and
G. M. Friedman). Springer-Verlag, Heidelberg, New York, pp. 138-147. R.M. ( 1971) Some observations on Recent sediment alteration ('micritisation') and the poss ible role of algae in submarine cementation. In : Carbonate Cements. Johns Hopkins Univ. Stud. Geol. 19, 72-79.
LLOYD,
LOREAU, J.P. & PURSER,
B.H. (1973) Distribution and ultrastructure of Holocene oois in the Persian Gulf. In : The Persian Gulf (Ed. by B. H. Purser), pp. 279-328. Springer-Verlag, New York.
LUKAS,
K.J. ( 1973) Taxonomy and ecology of the endolithic microflo ra ofreef corals with a review of the
literature on endolithic microphytes. Ph.D. thesis, University of Rhode Island.
McCuNN, H.J. ( 1972) Calcite and aragonite phenomena precipitated by organic decay in high lime concentrate brines. J. sedim. Petrol. 42, 150-154. S. & R Ex, R.W. ( 1971) Endolithic algae and micrite envelope formation in Bahamian oolites as revealed by scanning electron microscopy. Bull. geol. Soc. Am. 82, 843-852.
MARGOLIS,
228
Origin of calcareous crusts MATTHEWS,
R.K. (1974) A process approach to diagenesis of reefs and reef associated limestones. In :
Reefs in Time and Space (Ed. by L. F. Laporte). Spec. Pubis Soc. econ. Paleont. Miner. , Tulsa, 18, 234-256.
E. ( 1972) Fundamentals of the Fungi. Prentice-Hall, Englewood Cliff, N. J. H.G. & HoFFMEISTER, J.E. (1968) Subaerial laminated crusts of the Florida Keys. Bull. geol. Soc. Am. 79, 183-192. PALMER, C.M. (1962) Algae in Water Supplies : An Illustrated Manual on the Identification, Signifi cance and Control of Algae in Water Supplies. U.S. Dept. HEW Public Health Service Pub. 657, Washington, D.C. PAUL, I. C. & CuRTIN, D.Y. ( 1975) Reactions of organic crystals with gases. Science, 187, 19-26. PERKINS, R.D. & HALSEY, S.D. ( 1971) Geologic significance of microboring fungi and algae m Carolina shelf sediments. J. sedim. Petrol. 41, 843-853. PERKINS, R.D. & TsENTAS, C. I. ( 1973) Microbial destruction of carbonate substances 'planted' on the northeastern St Croix shelf. Abs. Prog. geol. Soc. Am. PuRDY, E.G. ( 1963a) Recent calcium carbonate facies of the Great Bahama bank, 1. Petrography and reaction groups. J. Geol. 71, 334-355. PURDY, E.G. ( 1963b) Recent calcium carbonate facies of the Great Bahama bank, 2. Sedimentary facies. J. Geol. 71, 472-497. PURDY, E.G. (1968) Carbonate diagenesis : an environmental survey. Geol. Rom. VII, 183-228. RoONEY, W.S., JR. & PERKINS, R.D. ( 1972) Distribution and geologic significance of microboring organisms within sediments of the Arlington reef complex, Australia. Bull. geol. Soc. Am. 83, 1139-1150. SANDBERG, P.A., SCHNEIDERMANN, N. & WUNDER, S.J. ( 1973) Aragonitic ultrastructural relics in calcite-replaced Pleistocene skeletons. Nature, Land. 245, 133-134. SCHMALZ, R.F. (1967) Kinetics and diagenesis of carbonate sediments. J. sedim. Petrol. 37, 60-67. ScHMALZ, R.F. & SwANSON, F.J. ( 1969) Diurnal variations in the carbonate saturation of seawater. J. sedim. Petrol. 39, 255-267. SCHNEIDERMANN, N., SANDBERG, P.A . & WUNDER, S.J. ( 1972) Recognition of early cementation of aragonitic skeletal carbonates. Nature, Land. 240, 88-89. ScHRAER, H. ( 1970) Biological Calcification: Cellular and Molecular Aspects. Appeleton-Century Crofts, New York. ScHROEDER, J.H. (1972) Calcified filaments of an endolithic algal in Recent Bermuda reefs. N. lb. Geol. Pa!eont. Mh. , Jg 1972, 16-33. SILVERMAN, M.P. & MuNoz, E. F. ( 1970) Fungal attack on rock : solubilization and altered infrared spectra. Science, N. Y. 169, 985-987. STEINEN, R.P. (1974) Phreatic and vadose diagenetic modification of Pleistocene limestones : petro graphic observations from subsurface of Barbados, West Indies. Bull. Am. Ass. Petrol. Geol. 58, 1008-1024. WAGNER, E. & ScHWARTZ, W. ( 1965) Geomikrobiologische Untersuchungen. IV. Untersuchungen tiber die mikrobielle Verwitterung von Kalkstein im Karst. Z. al!g. Mikrobiol. 5, 52-76. WALLS, R.A., HARRIS, W.B. & NUNAN, W.E. ( 1975) Calcareous crust (caliche) profiles and early subaerial exposure of Carboniferous carbonates, northeastern Kentucky. Sedimentology, 22, 417-440. WINLAND, H.D. (1968) The role of high-Mg calcite in the preservation of micrite envelopes and textural features of aragonite sediments. J. sedim. Petrol. 38, 1320-1325. MoORE-LANDECKER,
MULTER,
(Manuscript received 23 March 1 976; revision received 4 October 1 976)
229
Reprinted from Sedimentology (1979) 26 593-604
Calcification in a coccoid cyanobacterium associated with the formation of desert stromatolites
W. E. K RUMBEIN
and
C. G IELE
Environmental Laboratory, University of Oldenburg, P.O. Box
D-2900
25 03,
Oldenburg, 0/db., Germany
ABSTRACT In the Borrego Desert (Californi�) and in the Sinai Desert (Israel) laminated, microbially mediate:! carbonate crusts have been found and analysed biologically and mineralogically, and further studied with scanning electron microscope methods combined with energy dispersive X-ray analyses. All morphological and biological features of the extant crusts justify the term 'desert stromatolite', a term applied to stromatolites from desert regions which form under permanent exposure to the atmosphere. These stromatolites are never covered by standing water, and running water (heavy rainfall) covers them for only a few hours.during the ye�r. Carbonate deposition is achieved principally by the cyanobacterium Pleurocapsa sp. which exhibits characteristic yet different stages of calcification. Calcification occurs in the sheaths of single cells (including baeocytes) as well as in mature colonies. The specificity for calcification in Pleurocapsa sp. is discussed.
INTRODUCTION
Stromatolites have at present several different definitions. In our contribution we use the definition given in Walter (1976 p. l .): 'Stromatolites are organosedimentary structures produced by sediment trapping, binding and/or precipitation as a result of growth and metabolic activity of microorganisms, principally cyanophytes.' Walter further states that water cover is necessary. Awramik, Gebelein & Cloud (1978), however, state that formation in the terrestrial environment is possible, but as yet not recorded with recent or ancient examples. Monty (1976) has justified arguments that the term cryptalgal structures may be applied when one wishes to talk exclusively of structures of non-skeletal algal origin. In the study of stromatolitic structures, it has been frequently stated that one of the crucial questions asked is whether the micro-organisms merely act as sediment traps and binding agents, or whether they are actively participating in lithification by the precipitation of car bonates. Krumbein (1978) has extended this view by showing that precipitation of Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
231
W. E. Krumbein and C. Giele
carbonates within stromatolitic structures may occur by different processes: (a) inorganic precipitation not controlled in any way by biological processes, (b) inorganic precipitation initiated by biogenic changes in the chemical environment, (c) precipita tion of minerals on or in the vicinity of cells, related to their morphology and distri bution, but initiated by changes in the immediate c;hemical environment of the cell, and (d) by species-dependent specific calcification within or on cells or parts of cells. Desert calcareous crusts (calcretes) have been studied in relation to these terms and questions. MATERIAL AND METHODS
Samples were of stromatolitic crusts from the desert region around Borrego Springs, California (33° 15·1' N, 116° 22·5' W), and from Ein Netafim, 10 km from the town of Elat (Israel) in the Sinai desert. In both cases, small wadis or depressions on mountain slopes experience traces of running water during the time of heavy desert rains. Annual rainfall in the Borrego region averages 200-400 mm over a mean of 30 days. In the Sinai, rainfall averages between 15 and 50 mm over a mean of 10 days. The Borrego calcrete crusts were in a granite valley, while the calcretes from Sinai were from a small wadi depression close to Ein Netafim, in a small Cretaceous graben parallel to the main rift valley of the Arava (Figs I and 2). Calcrete crusts of possibly different origins occur at both sites. The crusts at both sites are between 1 and 3 em thick. Only laminated calcretes were collected and analysed. Samples were taken preferentially where active growth of photosynthetic organisms was recognized, usually as green or black spots in or on the calcretes. Thin sections were prepared and analysed under laboratory conditions. Carbonate chips with associated photosynthetic organisms were placed on culture media, and parts of the chips at the same time analysed using light microscopy, scanning electron (SEM) microscopy and electron dispersive X-ray analysis (EDX). For SEM and EDX studies, specimens of the crust were fractured, re-wetted, fixed in glutaraldehyde, critical-point dried, coated with gold, and analysed using a Cambridge Instruments S 180 connected to an Ortec 6230 EDX system, a number of samples was also examined in a Philips PW 1410 X-ray spectrometer. RESULTS
Although samples have been collected from different continents and on different bedrocks, the laminated calcrete crusts are of striking similarity (Figs 3 and 4). They are distinctly laminated in waves of alternating dark and bright laminae. Thin sections (Figs 5 and 6) reveal the same laminated pattern. Extremely fine-grained carbonate is interbedded with quartz and feldspar particles although clay minerals also occur. The study of thin sections and X-ray analyses implies that the calcretes also have a capacity for sediment trapping. In the case of the Borrego sample, debris of plants has also been found. In both cases the biological structures were not detectable in the thin sections. This may be due to the fact that no precaution was taken to preserve the biological material during the process of cutting and polishing. In both sets of samples light microscopy and SEM analyses indicate living cyano bacteria and chemo-organotrophic bacteria. At the crust surface they were still 232
Calcification in desert stromatolites
Fig. 1. Small wadi depression in Cretaceous limestone of Ein Netafim. The bedrock is covered by
calcrete crusts. Fig. 2. Calcrete crust dissected from bedrock. Left part shows crusts turned upside down. Sample with camera cover was taken home.
233
W. E. Krumbein and C. Giele
Fig. 3. Part of specimen in Fig. 2 after formatizing and cutting. Scale in mm. Fig. 4. Cross-section of calcrete crust from the Borrego desert.
234
Calcification in desert stromatolites
Fig. 5. Thin section of calcrete from Ein Netafim. Scale bar, I 00 J.llll. Fig. 6. Thin section of calcrete from Borrego desert. Scale bar, 150 J.lm.
235
W. E. Krumbein and C. Giele
photosynthesizing, and heterotrophic bacteria were active and motile. A visually conspicuous community consisting in both cases of the dominating organisms Pleuro capsa sp. and Plectonema sp. (LPP group) grows over the surface of the crusts at various places. In the case of samples of Ein Netafim Aphanothece sp. (Synechococcus) and Gloeocapsa sp. were also found to occur in smaller numbers. Only Pleurocapsa sp. and Plectonema sp were found to undergo calcification. The calcification of Plectonema sp. is described elsewhere (Krumbein & Potts, 1979). Different stages were recognized in the calcification of Pleurocapsa cells, all giving rise to various preserved structures within the laminated crusts (Fig. 7). Non-calcified cells were infrequently observed, because the collection of material was done during dryness. Their distribution appeared to be random (Fig. 7a and b). Enlargement of cells caused by the crystallization was occasionally discernible (Fig. 8). It should be stated, however, that the crystallization pattern demonstrated in Fig. 8 was the only case we have found in several samples, while hundreds of examples were found for the crystal growth pattern described in Fig. 7. Calcite deposition seems to originate within the sheath, which is deformed in its outer parts by crystal growth (Fig. 7c and d). Absolute proof of whether the deposition of carbonate starts within the capsule or on the capsule surface can only be given by TEM sections, which we were unable to analyse. The crystals are small, thin platelets (100-200 nm) orientated perpendicular to the surface, thus giving a misleading appearance of spherulitic structure (Fig. 8). Very often crystal platelets are situated normal or at smaller angles to each other thus forming a card-house structure (Fig. 7d). The calcification pattern and the shape of crystals are very different from that in Plectonema sp. (LPP group) (Krumbein, Rippka & Waterbury 1979). Weathering and ageing produce a smoother surface appearance. This may be due to additional micritic carbonate filling after the decay of the slime material of the capsules (Fig. 9), but also to differential redissolution during ageing (Fig. 7f and g). By these two processes, indistinguishable in SEM-micrographs (Figs 7f, 7g, 9), the originally very well defined cell structures (Fig. 7c;d, e, h and i) seem to become 'fused' together in a single layer (Fig. 7f). Cracks, which are definitely not artefacts of preparation, are present in both samples in these stages. The holes (Fig. 7g) are dissolution phenomena always occurring in older parts of the calcifying community. They may be caused by boring fungi or filamentous bacteria. Despite the well defined .fixation methods, these, however, were not seen. Another feature of dissolution of carbonate previously deposited is visible in Fig. 7d, where mushroom-like structures are formed. In desert environments, as well as in beachrocks, one always finds the phenomenon of competing microbially induced carbonate dissolution and precipita tion (Monty, 1976). This competition may lead to eventual accretion, destruction or to steady state equilibrium (Krumbein, 1969; 1979). New cycles of calcification are indicated (I) by a younger layer of cells on top of older smoothed carbonate crusts as in Figs 7f and 8 and (2) by baeocytes calcifying immediately after release or after attaining the size of new large colonies. Treatment with diluted acetic acid under the transmittant light microscope revealed also that fully mature colonies and young cells have the capacity to calcify. Plectonema (Krumbein & Potts,. 1979) calcifies by forming small elongated needles while Gloeocapsa sp and Aphanothece sp. do not calcify at all. Comparison of light microscopy data and SEM-data show clearly that the small cells shown in Fig. 7h are identical with the baeocytes in Fig. 7b. The baeocytes (Fig. 7b and h) are in a size order of 0·5 f..lill which corresponds to small bacterial cells. Bacteria in light 236
Q
r;'5; '"' l:l
6· N w -..J
�i}l!;� �t··�.� 'ti· ' ��� ��.-� <,�.
\
.•·
...
" ,.
;: s· l:l..
"' "' "' :::t "'
� ;::!
l:l
Ci ;::.,
�
Fig. 7. Different stages of calcification in
P!eurocapsa sp.: (a) Cells from surface thallus; commonly 3·6-4·8 �m diameter,
forming a pseudoparenchymatous syste:n and later pseudofilaments in single rows of cells, olive green; sheath usually clear, sometimes yellow; scale bar,
5 �m. (b) Parental cell (after treatment with dilute acid), showing numerous baee>cytes and 2 �m. (c) Early calcification of cell shea.ths in cells which settle
outer sheath; baeocytes=O·S �m diameter; scale bar,
on silicate minerals; scale bar, 3 �m. (d) Partially eroded calcified sheath; the edgy, irregular surface pattern is attributed to the deposition of calcite crystals within the sheath; scale bar,
1 �m.
[continued on next page]
� � ;:><: ..., �
� N (.;J 00
��
-.-
- •. t.' I ..._:,-
, • ·,
t\f
'If!�.&�� �
c::r "'
s-
!:> ;::: .,_
0 CJ «;· "'
[Fig. 7 cont] (e) Pseudofilament showing the single row of cells, assymetrically displaced; parenchymatous cell groups are adjacent to the 'filament'; the crystals are corroding, hereby producing a smoother surface; scale bar, 3 Jlm.(f) Overview of a final stromatolitic layer; cell structures are less distinct; corrosion and desiccation cracks visible; scale bar, 3 Jlm. (g) Partial breakdown of a layer (see f) by corrosion and boring (?); empty and broken sheaths are visible in the upper region; scale bar, 11 Jlm. (h) New calcifying structures (baeocytes) settling on an older layer; scale bar, 1 Jlm. (i) Ca (KO() line scan of calcified sheaths on a feldspar background; sheaths have a significantly higher Ca (KO() signal and significantly lower K (KO() signal than the background; scale bar, 111m.
Calcification in desert stromatolites
Fig. 8. Atypical calcification of P!eurocapsa sp. This pattern was found only once while hundreds of cells were calcifying as shown in Fig. 7. Scale bar, 211m.
microscopy preparations and SEM micrographs did not calcify at all in the case of the desert stromatolite samples, though calcification is known for bacteria in stromatolitic environments (Krumbein, Cohen & Shilo, 1977; Krumbein, 1978). X-ray analyses of the carbonates from the stromatolites yielded exclusively low Mg calcite (5 mol/;; MgC03). Fig. 7i shows an EDX-profile over calcified Pleurocapsa sp. cell aggregates on a background of quartz minerals. Ca is highly enriched on the calcified cells. This method does not allow an estimate of Mg ratios within the carbonate since it is only semi-quantitative.
DISCUSSION
It is clear from Figs 3-6 that true stromatolitic structures occur within the calcrete crusts collected in California and Israel. Analyses of the micro-organism associations clearly demonstrate that the laminated stromatolites are produced by their activities, namely calcification, and the trapping of windblown or sheet-flood transported particles. According to the literature (Walter, 1976; Awramik et a!., 1978) such biogenic laminated structures have not yet been described for desert environments. In some respects, they are comparable to subaerial stromatolites of the supratidal zone described by Monty & Hardie (1976). Awramik eta!. (1978) state that periods of submergence or wetting are necessary for stromatolitic development. The formation of desert stromatolites with extremely short wetting periods is in itself astonishingly similar to supratidal stromatolitic crusts discussed by Monty (1973). The interesting point in our study, in addition to the similarities to supratidal 239
W. E. Krumbein and C. Giele
Fig. 9.
Calcified cells of
Pleurocapsa sp. showing smooth parts. It was difficult to define, even in direct
observation of the oscillator screen and altering the information by grey levelling and structuring electronically, whether this is corrosion or addition of micritic calcite. Ca-line scans of such samples did not show higher or lower amounts of Ca at the respective patches. Scale bar, 2 JJm.
subaerial crusts, is the identity of the stromatolites from different deserts as well as the identity of the microbial communities and the specific calcification pattern in Pleuro capsa sp. Horodyski & Yonder Haar (1975) have described premineralization of a coccoid cyanobacterium which they called Entophysalis sp. They attributed the calcification to photosynthetic C02 uptake. The figures and micrographs in their paper, however, bear a strong resemblance to the form discussed here (Krumbein et. a!., 1979). Krumbein ( 1978) also commented on the presence of Pleurocapsa and Entophysalis in the Bahia California mats. The marine species Pleurocapsafuliginosa is known to form microfossils (Knoll, Barghoorn & Golubic, 1975) and Potts (1977) has described marine calcareous stromatolites from a hypersaline tidal flat in the lagoon of Aldabra Atoll, Indian Ocean, where the dominant forms were P/eurocapsa jit!iginosa and Entophysa/is granulosa. In all these previously described cases, calci fication was mainly explained by C02 uptake or bacterial degradation of organic matter. Monty (1967) has described calcification of coccoid and filamentous forms . within a generally calcifying environment. In the case of the two desert stromatolites described here, several arguments strongly suggest specificity of calcification in P/ectonema (Krumbein & Potts, 1979) and P/eurocapsa. Golubic (1973) has summar ized the controversial discussion about specificity of calcification in a very clear way: 'Since the formation of CaC03 crystals within the thalli of cyanophytes does not conform to the rules of inorganic mineralogy, a specific algal influence has been 240
Calcification in desert stromatolites
postulated.' and 'The evidence that such crystal formations or modifications are species specific in Cyanophyta (Wallner, 1934), however, is not conclusive.' Several authors have previously mentioned appearance and shape of carbonate crystals forming along the filaments or capsules of cyanobacteria (e.g. Monty, 1967; Riding, 1977; Wallner, 1935). Several arguments for specificity of calcification in cyanobacteria can be added from our findings in desert stromatolites described here. (1) Two species of cyanobacteria occurring in spatially separate environments calcify and preserve their morphologies during this process. (2) The crystal shape in one case is elongated and consists of spicules perpendicular to the filament (Krumbein & Potts, 1979) while, in the case of Pleurocapsa, platelets are formed of much smaller size and different shapes. The two carbonate forms may occur in one single aggregate. (3) Bacteria which sometimes calcify in stromatolitic environments were not found to calcify in this aerobic environment (Krumbein, 1978). (4) Two other species of cyanobacteria (Aphanothece sp. and Gloeocapsa sp.) do not calcify at all in these environments. To some extent this rules out the attribution to nonspecific C02 uptake and subsequent carbonate precipitation. In addition to this, a freshwater tufa found directly in the perennial spring of Ein Netafim exhibited typical calcification patterns caused by water supersaturated with C02 which, coming into equilibrium with the atmosphere outside the rock, immediately causes calcification. This was enhanced by mosses and other plants growing in the spring and along the walls of the waterfall. Astonishingly also in this environment we found non-calcifying filamentous and coccoid cyanobacteria. Concerning specificity of calcification in cyanobacteria, we conclude that more evidence for this postulation has been added. We feel strongly, however, that careful laboratory experiments with axenic cultures of the organisms in question are necess ary, and that the localization of the crystals should be defined by TEM methods. Unfortunately, Pleurocapsalean cyanobacteria are relatively slowly growing organisms and our isolates in the Oldenburg cyanobacteria collection so far have not yielded laboratory made calcite. Concerning the occurrence of laminated calcretes in desert regions, we conclude that the great similarities in texture and fabrics, as well as the two identical major species contributing to calcification in the Borrego Desert and in the Sinai mountains, give further evidence of the importance of cyanobacteria in the formation of subaerial stromatolites.
ACKNOWLEDGMENTS
The work was supported by DFG grant Kr 333/12 and Kr 333/13. We thank G. Koch, I. Raether, P. Rongen and M. Rosenow for technical aid. M. Potts has con tributed largely to the determination of forms according to the botanical code and to field and laboratory analyses. The major findings of the work were presented at the lAS conference, 1978, in Jerusalem (Potts & Krumbein, 1978). The comments of several participants during the discussion of the paper have improved the presentation of data. In addition we are indebted more than usual to the reviewers of the first version, whose written and in one case (C. L. V. Monty) oral comments have made resubmission possible and improved the interpretation. It is lamentable that because of his untimely death we are unable to express our gratitude and appreciation of his 241
W. E. Krumbein and C. Giele
comments to Conrad Gebelein himself. They must have been among the last things he has communicated to the field, in which he was guiding us for so long. REFERENCES S.M., GEBELEIN, C.D. & CLOUD, P. (1978) Biogeologic relationships of ancient stromato lites and modern analogs. In: Environmental Biogeochemistry and Geomicrobio/ogy, Vol. I The Aquatic Environment. (Ed. by W. E. Krumbein), pp. 165-178. Ann Arbor Science, Ann Arbor. GOLUBIC, S. (1973) The relationship between blue-green algae and carbonate deposits. In: The Biology of Blue-Green Algae (Ed. by N. G. Carr and B. A. Whitton), pp. 434-472. Blackwell Scientific Publications, Oxford. HoRODYSKI, R.J. & YONDER HAAR, S. (1975) Recent calcareous stromatolites from Laguna Mormona (Baja California). J. sedim. Petrol. 45, 894-906.. KNOLL, A. H., BARGHOORN, E.S. & GOLUBIC, S. (1975) Pa/aeop/eurocapsa wopfneri Gen. et sp. nov.: a late Precambrian alga and its modern counterpart. S. Proc. natn. Acad. Sci. U.S.A. 72, 24882492. KRUMBEIN, W.E. (1969) Deer den Einflul3 der Mikroflora auf die exogene Dynamik (Verwitterung und Krustenbildung). Geo/. Rdsch. 58, 333-363. KRUMBEIN, W.E. (1978) Algal mats and their lithification. In: Environmental Biogeochemistry and Geomicrobio/ogy, Vol. I The Aquatic Environment (Ed. by W. E. Krumbein), pp. 209-225. Ann Arbor Science, Ann Arbor. KRU:YIBEIN, W.E. (1979) Photolithotrophic and chemoorganotrophic activity of bacteria and algae as related to beachrock formation and degradation (Gulf of Aqaba, Sinai). Geomicrobiol. J. 1, 139-203. KRUMBEIN, W.E., CoHEN, Y. & SHILO, M. (1977) Solar Lake (Sinai). 4. Stromatolitic cyanobacterial mats. Limno/. Oceanogr. 22, 635-656. KRUMBEIN, W.E. & Porrs, M. (1979) Recent Girvanella from a freshwater environment in the Borrego Desert, south California. Geomicrobiol. J. 1, 211-218 KRUMBEIN, W.E., RIPPKA, R. & WATERBURY, J.B. (1979) Schematische bakteriologische Gliederung der Cyanophyten im Vergeleich zur phykologischen. In: Cyanobakterie-Bakterien oder Algen? (Ed. by W.E. Krumbein), pp. 107-130. Univ. Oldenburg, Oldenburg. MoNTY, C.L.V. (1967) Distribution and structure of Recent stromatolitic algal mats, eastern Andros Island, Bahamas. Ann. Soc. geol. Be/g. Bull. 90, 55-100. MoNTY, C.L.V. (1973) Precambrian background and Phanerozoic history of stromatolite com munities, an overview. Ann. Soc. geo/. Be/g. Bull. 96, 585-624. MoNTY, C.L.V. (1976) The origin and development of cryptalgal fabrics. In: Stromatolites (Ed. by M. R. Walter) Developments in Sedimentology, 20, pp. 193-250. Elsevier Publishing Co., Amsterdam. MoNTY, C.L.V. & HARDIE, L.A. (1976) The geological significance of the freshwater blue-green algal calcareous marsh. In: Stromatolites (Ed. by M. R. Walter), Developments in Sedimentology, 20, pp. 447-478. Elsevier Publishing Co., Amsterdam. Porrs, M. (1977) Studies on blue-green algae and photosynthetic bacteria in the lagoon of Aidabra Atoll. Unpublished Ph.D. Thesis, University of Durham, England. RIDING, R. (1977) Problems of affinity in Paleozoic calcareous algae. In: Fossil Algae, Recent Results and Developments (Ed. by E. Fli.igel), pp. 57-60. Springer-Verlag, Berlin. WALLNER, J. (1934) Zur Kenntnis des unter pflanzlichem Einflul3 gebildeten Kalkspates. Planta, 51-55. WALLNER, J. (1935) Zur Kenntnis der Kalkbindung in der Gattung Rivularia. B:!ih. Bot. Zentralbl. B 34, 151-155. WALTER, M.R. (Ed.) (1976) Stromatolites. Developments in Sedimentology, 20, Elsevier Publishing Co., Amsterdam. AwRAMIK,
(Manuscript received 27 January 1978; revision received 17 November 1978)
242
Reprinted from Sedimentology (1988) 35 6 03-6 20
Biogenic laminar calcretes: evidence of calcified root-mat horizons in paleosols V. P. W R I G H T Department of Geology, University o f Bristol, Wills Memorial Building, Queen's Road, Bristol BS8 1 RJ
N. H . P L A T T* Department of Earth Sciences, University of Oxford, Parks Road, Oxford OXJ 3PR
W. A. W I M B L ED ON Geological Conservation Unit, Nature Conservancy Council, Northminster House, Peterborough
ABSTRACT
Laminar calcretes are described from the Lower Carboniferous of South Wales, the Upper Jurassic of southern England and the Upper J urassic-Lower Cretaceous of northern Spain. They are interpreted as calcified root-mats (horizontal root systems) and are compared with other examples in the geological record and with possible modern analogues. All three occurrences consist of virtually identical, centimetre to decimetre-thick, locally organic carbon-rich, laminar micrites containing up to 5 0% by volume of millimetre-sized typically calcite-filled, tubular fenestrae set in an irregular but very finely laminated matrix. It is suggested that root-mat calcretes are probably very common in the geological record in peritidal, lacustrine margin and floodplain deposits, but owing to their crudely biogenic microstructure, they more closely resemble cryptalgal laminites than do other laminar calcretes. The recognition of such root-mat calcretes in sedimentary sequences not only provides evidence of subaerial exposure and vegetation cover but can also indicate positions of palaeo-water-tables in certain circumstances.
horizons resemble microbial stromatolites, and this has led to a debate on their differentiation (Read, 1976). The aim of this paper is to describe a type of laminar calcrete which has been found in three distinctly different geological settings, and are inter preted as calcified root mats. These calcretes more closely resemble microbial stromatolites than other laminar calcretes and they could be easily mistaken for such stromatolites.
INTRODUCTION
Paleosols and other types of subaerial exposure features are now being widely recognized within ancient shallow water limestone sequences and conti nental facies. They can provide important information on base-level and sea-level changes, sedimentation rates and early diagenetic processes and porosity formation. Descriptions of subaerial surfaces and calcretes on Recent and Pleistocene limestones pro vide a suite of criteria useful for recognizing such surfaces in the rock record (Harrison & Steinen, 1978; Esteban & Klappa, 1983; Braithwaite, 1983). Some of the most commonly recognized features of this type are calcrete crusts which coat exposed limestone surfaces, or are associated with calcrete profiles. Such
LAMINAR CRUSTS
Laminar calcretes ('croute zonaire' of the French usage, e.g. Freyet & Plaziat, 1982) are a characteristic feature associated with subaerially exposed limestones and with calcareous and calcrete-bearing soils. These laminar horizons are generally only a few centimetres or tens of centimetres thick and display a gently
*Present address : Geologisches Institut, Universitat Bern, Baltzerstrasse I, CH-3 01 2 Bern, Switzerland. Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
243
V. P. Wright, N. H. Platt and W. A. Wimbledon
undulating form, especially when coating irregular topographies where they commonly thicken into hollows. They can occur even on very steep or near vertical surfaces and can also completely coat objects to form coated-grains (ooids and pisoids). The laminar horizons generally have sharply-defined lower and upper boundaries and are commonly dark in colour, reflecting the presence of disseminated organic matter. The lamination is typically on a millimetre scale and is due to variations in microstructure or in the degree of organic staining. The microstructure is variable but commonly consists of dense micritic or microsparitic matrix with various forms of fenestrae and rootlet moulds. Sediment grains and soil fragments can be incorporated into such horizons. Some exhibit convo luted laminations and resemble microbial stromato lites, but differ from them in that they show an absence of the upwardly thickening, domed laminations so typical of stromatolites (Read, 1976). They may show evidence of phases of dissolution and typically contain micro-unconformities in the laminae. Some are asso ciated with abundant evidence of micro-borings, related to the activities of lichens (Klappa, 1979). Detailed illustrations and descriptions of these laminar calcrete characteristics are to be found in numerous papers (for example, Multer & Hoffmeister, 1968; Read, 1974, 1976; Braithwaite, 1975; Harrison, 1977; Netterberg, 1980; Blume!, 1982; Esteben & Klappa, 1983). A variety of different laminar calcrete types has been described but in broad terms three categories can be defined (Fig. I); surficial, pedogenic and capillary-rise zone.
SURFICIAL
Lcm Go
crust
�
T
I
-
limestone
PEDOGENIC
�m CAPILLARY RISE ZONE
[m :_:: :·:- -:· .. . . ..
.
·.
.
.
.
sediment
.
.
: ���g �=L�ca lcrete �.,...,--�,..,..---....+.,,___ water table · -
·. ·. ·.
Fig. l. Types of laminar calcrete (see text).
and calcium oxalate to create 'lichen stromatolites'. These studies suggest that the differentiation of calcrete forms and stromatolites, which had been a focus of interest previously (Read, 1976) may be even more difficult because some subaerial (pedogenic) laminar horizons are biogenic in origin. Klappa ( 1979) did provide a number of criteria for recognizing lichen stromatolites and. Coniglio & Harrison ( 1983) have used them to identify similar horizons in the Quater nary of Florida.
Surficial laminar calcretes
These form at the rock-atmosphere interface on lithified substrates. Many such laminar calcretes found exposed at the present time originally formed within soil profiles and have been exposed by erosion and differentiating them from true surficial forms can be difficult. Surficial laminar calcretes form in a similar manner to travertine in that calcium carbonate precipitates out from water flowing over the substrate or when it is ponded in depressions (Multer & Hoffmeister, 1968; Arakel, 1982; Lauman, 1973). However, other surficial forms are biogenic in origin and represent calcified microbial crusts (desert stro matolites of Krumbein & Giele, 1979). Klappa ( 1979) suggested that some laminar carbonate horizons are formed by lichens, which not only alter carbonate substrates but can also produce calcium carbonate
Pedogenic laminar calcretes
Laminar horizons can also form in association with soil profiles and usually cap either underlying im permeable horizons such as calcrete hardpans (Read, 1974; Netterberg, 1980; Arakel, 1982; Goudie, 1983) or underlying bed-rock, or cemented surficial sedi ments (Multer & Hoffmeister, 1968). The thickness of the overlying soil can vary from only a few centimetres, as in the case of rendzina-type soils, to a metre or 244
Biogenic laminar calcretes
more with some Aridsols and Xerosols. The generally held explanation for the formation of these soil covered laminar horizons is that carbonate-rich water percolating through the soil profile is ponded above the impermeable layer and the calcium carbonate is precipitated to form an upwardly accreting horizon (Gile, Peterson & Grossman, 1966; Read, 1974; Arakel, 1982). As soils aggrade, new separate horizons can form, thus leading to the development of multiple laminar horizons (Arakel, 1982). An alternative explanation for the origin of laminar forms capping hardpan calcretes has been offered by Blume! ( 1982). He suggested that they are surficial forms formed by precipitation from run-off flowing over exhumed hardpans which could be later covered by more soil and so preserved within a profile. Some of these soil-covered laminar calcretes contain a high percentage of tubiform pores, notably those from Florida described by Multer & Hoffmeister ( 1968) and Perkins ( 1977). We believe that these represent possible analogues for the examples de scribed in this paper.
not described in detail from three distinctly different geological settings. They come from the Lower Carboniferous (Dinantian) of Miskin, South Wales, the Upper Jurassic (Portlandian)of Swindon, southern England and the Upper Jurassic-Lower Cretaceous of the Cameros Basin, northern Spain. We use the term calcrete in the sense of Goudie ( 1983) for all the carbonate accumulations described occur within pa leosol profiles. Lower Carboniferous laminar calcrete from Miskin
A distinctive laminar calcrete occurs in the Heather slade Geosol, a complex polygenetic calcrete profile capping the Chadian Gully Oolite (syn. Caswell Bay Oolite) at Miskin, near Cardiff, South Wales (Fig. 2). Details of the locality and stratigraphy have been given by Riding & Wright ( 198 1) and profile details were provided by Wright ( 1986, 1987). The laminar calcrete coats cobble to boulder-sized fragments of calcretized oolitic grainstones and reach a thickness of 35 mm; however, lateral extent was impossible to determine because the soil was modified by a later phase of Carboniferous pedogenesis (Wright, 1986) and because quarry blast-damage made lateral corre lation difficult. It occurs within a calcrete conglomer ate layer interpreted by Riding & Wright ( 1981) as a regolith overprinted by a massive calcrete (petrocalcic horizon). The regolith material contains a large suite of calcrete and vadose fabrics, including micritic grain coats, inter-granular fungal networks, needle-fibre calcite, tubiform rootlet moulds associated with alveolar-septal structure, probable calcified faecal pellets, meniscus cements, geopetal textures, and circumgranular cracks (Riding & Wright, 1981; Wright, 1986, 1987). The laminar calcrete, which is almost black in colour, contains centimetre-sized angular fragments of the calcretized oolitic substrate set within a crudely laminar fabric, containing a few distinctive darker, denser laminae. The base of the horizon is sharply defined. Thin sections show that it has a highly irregular, contorted micro-fabric with smooth-walled calcite cement-filled cylinders, set in a matrix with a contorted micro-laminar fabric. (The term micro laminar is here used to describe the very fine fibrous microstructure seen in thin sections, while we reserve the term 'laminated' for the coarser-scale layering seen in hand specimen) . The cylindrical fenestrae which comprise 10-50% of the volume of the horizon are smooth-walled, and are oriented subparallel to sub-perpendicular to the underlying substrate (Fig.
Capillary rise-zone laminar calcretes
A third type of laminar calcrete has been recognized forming just above the water-table in present day dunes in southwestern Australia (Semeniuk & Meagher, 1981; Semeniuk & Searle, 1985). Evapora tion of water above the water-table, in the zone of capillary rise or updraw by deep-rooted plants leads to supersaturation of calcium carbonate and causes precipitation. The result can be the formation of an impermeable massive calcrete which acts as a sub strate for the precipitation of a laminar horizon from ponded waters. The term non-pedogenic has been used by Semeniuk & Meagher ( 1981) to describe such horizons. However, we prefer the term capillary-rise zone laminar calcrete because surficial laminar cal cretes could also be described as 'non-pedogenic'. Laminar calcretes forming at very shallow water tables arguably can be pedogenic and some ancient examples are described in this paper. The distinction between pedogenic and capillary-rise zone types becomes meaningless in such cases.
DESCRIPTION OF
LAMINAR
CALCRETE
During our separate studies of paleosols we have found identical laminar calcretes of a type hitherto 245
V. P. Wright, N. H. Platt and W. A. Wimbledon
z <(
High Tor
-'""' -'""'
Limestone Caswell Bay Mudstone HG
0
0
25km
• Carboniferous
Gully Oolite
[
sm
Limestone
Fig. 2. Locality map of Miskin showing stratigraphic position. H G , Heatherslade Geosol. For further locality details see
Riding and Wright ( 1 9 8 1 ).
locally passes into poorly preserved alveolar-septal fabric. Well-preserved examples of this latter fabric do occur (Fig. 3c) and the curved septae are mainly 20-50 11m wide and less than 100 11m long. The horizon also contains scattered ooids, presum ably derived from the substrate, and small aggrega tions of rounded, sub-spherical, well-sorted peloids 40-50 11m in diameter (Fig. 3b) resembling the faecal pellets found in associated horizons (Wright, 1987). Amorphous organic-carbon is also visible in thin sections.
3a). the cylinders are filled with a drusy or equant medium to coarsely crystalline calcite cement, are up to 3 mm in diameter and up to 10 mm in length. They probably represent meandering, curved tubes which are cut by the plane of section. However, many show evidence of compression and have non-paralh!l mar gins (Fig. 3a), indicating that many, at least, wer"e empty calcite-walled tubes before cementation by calcite cement. The sparry calcite-filled cylinders have walls of dark micrite which are micro-laminar (Fig. 3b), that is, they are made up of several sheets of micrite up to 100 11m thick with individual layers traceable for up to a millimetre, separated by microspar layers which are up to 200 11m thick but are mostly in the range 50100 Jlm. Rarely, some of the micro-laminae have poorly preserved partitions connecting the micrite layers giving rise to a crude cellular structure. A few tubes have crudely pelleted walls, the margins of which are defined by irregular micrite peloids up to I 00 11m in diameter. When the cylindrical structures are viewed in cross section the micro-laminar structure is clearly seen to be concentric around them. Where the cylinders are densely packed, the inter-pore space only contains the micrite and microspar lamellae; where they are further apart, the micro-laminar zone passes into areas filled with a cloudy microspar exhibiting a much more crudely micro-laminated texture (Fig. 3c) which
Upper Jurassic laminar calcrete from Swindon
The second example, occurs in the Great Quarry in Swindon, Wiltshire (British Grid Reference SU 153836) within the 2-m thick Town Gardens Member of the Portland Stone Formation (Portlandian) (Fig. 4) (Sylvester-Bradley, 1940; Wimbledon, 1976, 1981). The Town Gardens Member comprises a laterally highly variable sequence of fenestral lime mudstones, peloidal and quartz-rich bioclastic marine limestones and sandstones, some non-marine limestones and two prominent laminated calcretes. The two horizons each directly overlie a medium-bedded, marine limestone with large bio-moulds and 0·3 m below the lower unit is a distinctive sandy micritic limestone with mm sized rhizocretions with alveolar-septal structures. Pebbles and boulders, including clasts of these marine 246
Fig. 3. Photomicrographs of the laminar calcrete from Miskin. (a) highly fenestral fabric. The long axes of the cylindrical fenestrae are subparallel to the base of the horizon. N ote the larger irregular calcite-filled vugs. Scale bar represents 2·5 mm; (b) detail of microfabric, showing the dark micritic coating around the cylinders (transverse sections of tubes) and the crudely micro-laminar structure of the matrix. A small pellet concentration is arrowed. Scale bar rep(esents 0·5 mm; (c) note dark micritic coatings around the spar-filled cylindrical fenestrae, passing into the crudely micro-laminar zone. Alveolar septal fabric (arrowed) occurs within a micrite-rimmed tubule, and the diffuse fabric resembling alveolar-septal structure above the two micrite-coated cylinders. Scale bar represents 0·5 mm.
V. P. Wright, N. H. Platt and W. A. Wimbledon
? ,...
·.·.·.· .. ;: ; •
0
•
•
. . .
.
WEALDEN? . ...
. . .
T 1 I I .I .I .I
.. .
PURBECK LI MESTONE
'I Ill ;::,
FORMAT ION
' ' ' ' ' ' '
0
' '
B Berriasian
•
• •
'
•
·
•
•
CIS
® Town
-
Swlndon
l5
Sand
Q)
Gardens
Member
&
Stone Member
PORTLAND STONE FORMATION
Cockly Bed
Kimmeridgian
Oxfmd;an
\
:.·· ..·. ·: ·.·: .< .·. ' . . .
,...,...,.,..."',..�
Q).�
0
I
0
""' ""
I
0
0
""
I
0
I
I
0
I
I
0
0
""' ""' ...... 0
0
CLAY
O
�����;::;
Callovian
FORMATION
:>;;/// ':.'-
�::; ,r-----
Bathonian
.
.
I
-en 'Den 'Dec
KIMMERIDGE
·
.
'-
"\
.· ..�.... '\.. "-. "'-"" . .
.
.
.
Fig. 4. Stratigraphy of the Swindon area showing the position (asterisk) of the laminar calcrete within the Town Gardens
Member. (See text).
limestones, occur both between them and also beneath the lower one. The upper, thicker, porous marine limestone bed, with its laminated calcret e, is overlain by a fragmented white micritic limestone with freshwater gastropods. Both porous beds and their laminar calcretes are laterally discontinuous and the upper bed, in particular, is reduced to a mass of large cobble to boulder-sized, subrounded to rounded, spherical to oblate clasts for tens of metres along strike. These relationships suggest phases of in situ reworking during deposition of the unit, reminiscent of regolith horizons. A similar series of deposits has been described by Perkins (1977) from the Pleistocene of Florida. These cla�ts, and the parent rock from which they were derived, are calcite cemented, bioclastic, peloi-
dal, medium to coarsely grained sandstones and
limestones. The matrix around the clasts is a calcareous
clay-rich sand. This study concentrates on the lower much thicker
laminar calcrete. Thin sections of the sandstone clasts show a variety of calcrete textures such as floating sediment grains set in a dense micritic matrix passing
down into a grain-supported fabric in which the grains
exhibit irregular, non-laminated micritic coatings up to a few hundred microns thick. Some small (average I mm diameter), subspherical nodules with sharply defined margins occur, containing several quartz grains surrounded by micrite. Shell moulds are abundant and are only partially occluded by finely crystalline bladed calcite cement. The micrite matrix also exhibits sub-millimetre wide irregular cracks, 248
Biogenic laminar calcretes
downward into the substrate as sub-horizontal to sub vertical, irregular (wavy) dark, micritic stringers. These are up to several centimetres in length and averaging 1 mm in width (Fig. 5). Internally these stringers have the same microstructure as the laminar calcrete. Thin sections of the calcrete are strikingly like those at Miskin in exhibiting small tubiform cylinders (here as open pores) with micritic and microspar micro laminar walls (Fig. 6a). The only differences with this horizon and that at Miskin are that alveolar-septal structures are more common, and that dense, micritic laminae occur. These latter occur both within the · main part of the horizon, separating irregular cylin drical-micro-laminar layers, and near its top where they are more common and several millimetres thick (Fig. 6b). These thicker, dense zones have very few pores, exhibit a very fine wavy lamination and include layers of finely crystalline inclusion-rich, brown bladed calcite (Fig. 6b).
including circumgranular cracks. Bifurcating tubular pores are also common, up to 500 f.!m in diameter, and locally showing alveolar-septal structure. This suite of fabrics shows that the unit underlying the laminar calcrete was modified by pedogenic processes. The laminar calcrete itself is up to 50 mm thick (Fig. 5) but it thins over projections on the substrate to 15 mm. It is light buff-beige in colour and contains fragments of the underlying lithology up to 10 mm in diameter. The laminae, which have relief of one to two millimetres, are of two types: dark laminae, 1 mm or less in thickness which are traceable for several centimetres, separated by lighter coloured, highly contorted laminae. The waviness and contorted appearance is due to an abundam:e of small tubiform pores (under 1 mm diameter) occurring within the lighter coloured layers. The pores themselves are traceable laterally for only one or two millimetres. The base of the horizon is sharp and shows no evidence of previous endolithic activity at the contact. However, from the base there are extensions passing
Laminar calcretes from the Upper Jurassic-Lower Cretaceous of the Cameros Basin, northern Spain
Similar calcretes (Fig. 9b) occur in the Upper Jurassic Lower Cretaceous Tierra de Lara Group (Group I of Salomon, 1982) of the western Cameros Basin of northern Spain (Platt, 1986) (Fig. 7). They occur in both siliciclastic (Sefiora de Brezales Formation) and carbonate-dominated sequences (Rupelo Formation) (Fig. 7), and four different facies associations have been recognized (Fig. 8). 1 Conglomerate/sandstone association (Fig. 8.1) Laminar calcretes occur in the alluvial Sefiora de Brezales Formation both within sandstones and capping channelized conglomerates. These clastic sediments have been interpreted as sandflat and streamflood deposits, respectively (Platt, 1986). The laminar calcretes which are 0·05-0·20 m thick, form laterally discontinuous, undulating horizons traceable for up to 5 m. They either cap the conglomerates, suggesting that their formation took place after the abandonment of the channels, or occur within the sandstone units where they are associated with cross cutting, 1-2 mm diameter, carbonate stringers inter preted as fine-scale rhizocretions (Platt, 1986) (Fig. 8).
Fig. S. Polished sample of the laminar calcrete from the Town Gardens Member. Note the biomoulds in the calcretized sandstone beneath the horizon. A small micrite stringer is arrowed. The mottled appearance of the sandstone is due to differential cementation by pedogenic micrite which occurs as grain coats.
2 Marl/paleosol association (Fig. 8.2) Laminar calcretes occur in both the upper part of the Sefiora de Brezales Formation and in the basal Rupelo 249
V. P. Wright, N. H. Platt and W. A. Wimbledon
Fig. 6. Photomicrogr.aphs of the Town Gardens Member laminar calcrete. (a) porous, micro-laminar microfabric showing fine
tubular and larger irregular pores (some with a crude septal structure resembling alveolar-septal structure) set in a micro laminar matrix). Scale bar represents 0·5 mm. (b) dense micritic laminae forming layers several millimetres thick. Scale bar represents 0·5 mm.
250
Biogenic laminar calcretes ' "'
c
'iii c 0 ·E·E (/)C.. O :J �·� a:o ,-"! ClCI: "'"' C.> Q_ :::l
__]_
HORTIGUELA
�(!J
FORMATION
b__i�L
oncoidal limestones
shallow, littoral
and marls
lacustrine
pedogenically -
palustrine
modified limestones
(/)
z
250
:J
�
Rio
0
---
Cabrera Member
L!.l
0 -
c.. :J 0
u <
(!J
!tl
<
·-
<
f-
L!.l a:
(/)
u
io
200
1< ::!! a: 0
a:
c
L!.l
marls and limestones
Member
a:·
::OJ.Sl
Member
CD
a: L!.l i=
100
Las Vinas Member
red marls pedogenicallymodified limestones
_l l L
a:
Ql
open lacustrine
150
Ladera
<
�
evaporitic lakes
charophyte-ostracod
de Lara
Cl
�
chert
limestones
0 -' w 0. :::l
....J
floodplain
yellow vuggy
Mambrillas
u_
a:
·-
l l
z
red marls
r.1
red marls
floodplain
palustrine
floodplain
-
.
50
1-
_ _
SENORA DE
,
BREZALES
u -
FORMATION
(/) (/)
< a: :J ...,
-
� .
.
. 0 Up.Oxl. -Lr.Kimm.
TA LVEILA FORMATION
Lower Callovian
A
•
·
·
.
ti:::!
.
.
.
•
•
� •
sandstones with
l/
channelised
sandflat-wadi
conglomerates
deposits
and nodular carbonate paleosols limestones and cross -bedded arenites
C:r::i\1'1v I
limestones
littoral marine
I
LITHOLOGIES
INTERPRETATION
C-C astrovido E-Espejon, H-Hortezuelos M-Mamolar
ostu<jy area
ML-Mambrillas d e Lara SB-Senora de Brezales
S PAIN
B
T-Talveila
�Okm
c
�km
Fig. 7. Stratigraphy and locality of the Cameros Basin sequence. The study area is shown in B and localities (and outcrop area) shown in C (see Platt, 1 986).
251
V. P. Wright, N. H. Platt and W. A. Wimbledon 1. CONGLOMERATE AND SANDSTONE ASSO CIATION
���§:..
root mat larqe rhizolith sandstone
r,.,-,+'F,:·."·C"'4- polyqonal
mottling
channel
2. MARL/PALEOSOL ASSO CIATION
rhizoliths many fine carbonate stringers '-----=--=---'-0
�
3. UN CONFORMITY ASSO CIATION directly on unconformity filling
joints
and fissures
nd
- -- -----
]gf 2
2m
�
0t mat
rubbly paleosol
Middle Jurassic · limestones 0
0
4. "PALUST RINE" CARBONATE ASSOCIATION palustrine limestones with
--:::J����
glaebules and microkarst root crusts up to 20cm occasional green marl partings (discontinuous) thin charophyte marls
red marls 5-20m red paleosols
palustrine limestones
� �
Fig. 8. Lithofacies associations of the laminar calcretes in the Cameros Basin sequence (see text).
upon relatively slow and discontinuous sedimentation. The limestones are micritic but are highly brecciated with abundant fine-scale anastamosing and circum granular spar-filled cracks averaging 0·5 mm wide. Further evidence of subaerial exposure and pedogenic modification includes often intense yellow-orange red-purple mottling, carbonate nodule (glaebule) de velopment, occasional 1-5 mm 'black pebbles', and isolated I mm diameter concentrically-coated spar filled tubular voids (interpreted as fine root tubules). Larger-scale 10-20 mm wide cracks also occur; these are filled either by grey, impure carbonate containing occasional floating clastic grains, or, more rarely, by densely-laminated micrite.
Formation (lower part of the Ladera Member) associated with interbedded decimetre- to metre-scale red marls or variably brecciated grey and mottled limestones (Fig. 8). The marls are generally structure less and homogeneous, although where laminar cal cretes are present they commonly show yellow-red purple mottling. The marls contain a very sparse microfauna of ostracod and charophyte fragments, abundant disemminated clay, and scattered fine detrital grains up to 250 Jlfil in size. The marls have been interpreted as the distal suspension deposits of a lake marginal alluvial plain (Platt, 1986). The absence of primary depositional features (such as lamination) is attributed to thorough bioturbation consequent 252
Biogenic laminar calcretes
The marls commonly contain 20-30 mm diameter vertical tubular structures up to 0· 3 m in length which have been interpreted as rhizocretions (Platt, 1986). The marls alternate with the laminar horizons, which are generally 10-50 mm in thickness. The marls are commonly of similar thickness and the alternation may be regular and repeated many times in vertical sequence. Total thicknesses of regularly-bedded marl laminar calcrete alternations are commonly 0· 5-1·0 m (as at Espejon, La Gallega and Talveila), but may locally reach 20 m (as at Senora de Brezales). However, the thicknesses of the laminar carbonate horizons and of the interbedded marls are locally variable. The calcretes may be only 10-20 mm thick, laterally impersistent over 0·2-0·5 m, and interbedded with marl units 0·5m or more in thickness; at other . localities (e.g. Talveila), they are more laterally continuous (individual horizons may be traceable for up to 20 m or more) and separated by only thin ( < 10 mm) red marl partings. Despite this variability, all of the scales of marl/calcrete alternation result in a gross centimetre-to-decimetre-scale laminar fabric which is strikingly apparent at outcrop. Locally, the combination of both vertical elements (rhizocretions, larger-scale desiccation cracks) and horizontal ele ments (laminar calcretes, desiccation sheet cracks) may create a crudely reticulate fabric (Fig. 8.2). At one locality (Senora de Brezales), sequences of stacked 20-50 mm bedded laminar carbonate units occur as gentle antiformaljdomal structures 5 m across, up to 2-5 m across and with a surface relief of 0·5 m (Fig. 9a). Field evidence suggests these structures are original and not tectonic in origin.
(beneath a soil), or along joints extending downwards from, the unconformity surface. These laminar horizons associated with unconform ity surfaces show a fabric similar to the other laminar forms, although micro-unconformities in the laminae seem to be more common. At Castrovido, the laminae are locally strongly blackened, a feature reminiscent of the crusts from the Carboniferous of Miskin, South Wales.
4 Palustrine carbonate association (Fig. 8.4) Laminar horizons are also developed associated with thick (up to 80 m) sequences of marginal lacustrine palustrine carbonate facies in the Ladera and Rio Cabrera Members of the Rupelo Formation. The limestones are pale grey in colour and are interbedded with rare, 0·5 m-thick, dark grey charophyte-rich marls. The limestones themselves are micritic mud stones and wackestones, and contain only rare charophyte stems and gyrogonites, ostracods and molluscan fragments. They show abundant evidence of pedogenic modification (Salomon, 1984; Platt, 1985) and are commonly brecciated on a millimetre scale owing to the presence of many fine spar-filled circumgranular and anastomosing cracks. Rounded carbonate nodules 1 mm in diameter are also abun dant. The limestones are locally strongly pink-purple mottled and contain centimetre-scale cavities with drusy sparry calcite and geopetal vadose cements with included peloids. These cavities are similar to the 'microkarst cavities' described from Cretaceous and early Tertiary palustrine carbonates of Languedoc by Freytet & Plaziat (1982) and Santander by Garcia Mondejar et a!. (1985). These Cameros Basin limestones have been inter preted as lake margin deposits. Periodic prolonged exposure of shallow lake carbonate muds led to pedogenic modification and solution (Platt, 1986). Laminar calcretes occur at many intervals within the Rupelo Formation but are usually very thin (520 mm) and laterally discontinuous on a decimetre to metre scale. At Mambrillas de Lara they occur interbedded with thin limestones (0·1-0·3 m thick) separated by 5-20 mm green marl partings. The thin limestones contain 1 mm diameter, concentrically micrite-coated tubules identical to rhizocretions de scribed from Quaternary calcretes by Klappa (1980) (Platt, 1986). Other laminar calcretes up to 1 m in thickness also occur at Mamolar and Rupelo (Platt, 1986). They are
3 Unconformity association (Fig. 8.3) 10-20 mm thick, gently undulating laminar horizons are also developed on Lower to Middle Callovian limestones at a prominent unconformity in the north of the study area (e.g. at Mambrillas de Lara and at Castrovido) (Fig. 8.3). The laminar horizons may locally occur beneath the main unconformity surface (at Mambrillas they are found down to 1 m below it), or may occur immediately beneath rubbly breccias which are developed on the unconformity surface. At Castrovido, the laminar calcretes and breccias occur only where the unconformity surface is not scoured by the conglomerate channels of the basal Senora de Brezales Formation. It is inferred, then, that the laminar horizons were developed in interchannel or upstanding areas, often at the exposure surface itself 253
V. P. Wright, N. H. Platt and W. A. Wimbledon
Fig. 9. (a) Thick multiple domal laminar calcrete from the Senora de Brezales Fm; east side of road from Espejon to La
Gallega, 3 km north of Espejon and 5 0 0 m north of the chapel of Nuestra Senora de Brezales. From the marl/paleosol association. (b) polished sample of laminar calcrete from the Rio Cabrera Member, 3 0 0 m west of Rupelo. Specimen taken from the top of a one metre thick horizon capping a palustrine sequence. From a palustrine carbonate association.
254
Biogenic laminar calcretes
rhizoliths), which are calcareous accumulations around living or dead roots (Calvet, Pomar & Esteban, 1975; Klappa, 1980, p. 621; Semeniuk & Meagher, 1981, p. 57; Arakel, 1982, p. 1 16; Wieder & Yaalon, 1982, p. 2 12; Mount & Cohen, 1984, p. 268). (4) The tubular structures, whether calcite-cemented or open locally contain, or pass laterally into, areas displaying alveolar-septal structure. This is a fabric probably formed by calcified fungal hyphae associated with roots (Phillips & Self, 1987; Wright, 1986). Such structures are not diagnostic of root activity, but they are very commonly associated with rootlets and so provide circum stantial evidence for the laminar calcretes being rootlet related.
present near the top of the carbonate members (such as the Ladera and Rio Cabrera Members) or at major bedding planes which have been interpreted as depositional breaks (Platt, 1986). The latter are associated with micro-brecciated horizons indicating prolonged exposure and karstification. Thus, these thicker horizons seem to coincide with major deposi tional hiatuses in the sequence.
Petrography
Thin sections (Fig. 10) show that all these laminar calcretes in the Cameros Basin sequences are virtually identical to those from the Miskin and Swindon sections. The only differences are that small pellet concentrations (Fig. lOa) are more common and alveolar-septal structures (Fig. lOb) are more abun dant but the characteristic, highly contorted cylindri cal structures and micro-laminar micrite structure is essentially identical to that in the other two occur rences (Fig. JOe).
The calcretes are considered to represent densely interwoven rootlet horizons which were calcified, possibly while the rootlets were alive, by micritic and microsparitic calcite. Significantly Klappa (1980, p. 625) in his studies of Quaternary calcretes, noted that this type of concentric encrustation was especially characteristic of rhizocretions within sheet calcrete horizons, such as those described here. Where the original rootlets were densely intertwined the whole rock has a micro-laminar microstructure, but where the rootlets were less closely spaced the pore-space was filled by alveolar-septal structure representing fungal activity around root hairs, rootlets or root voids. The denser micritic layers noted in the Swindon material, especially at the top of the horizon, probably represent simple precipitation without root calcification. Calcified root mats have been described, although not in detail, by a number of authors. Cohen (1982) and Mount & Cohen ( 1984) have described horizontal, planar root mats up to 50 mm thick, covering hundreds of square metres, with tubes averaging 5 mm diameter, from the Koobi Fora Formation (Plio-Pleistocene) of Kenya. Perkins (1977), in a study of the Pleistocene of Florida, has briefly described dense masses of calicified rootlets which he referred to as 'root-rock'. Calvet & Julia (1983) have referred to 'root stromato lites' which overlie lithified substrates in recent calcrete profiles from Tarragona (Spain) (see also Julia & Calvet, 1983). Klappa ( 1980, p. 6 15) has offered the term rhizolite for a rock showing structural, textural and fabric details determined largely by the activity or former activity of plant roots. We favour this term and classify the laminar calcrete described in this paper as rhizolite laminar calcretes.
INTERPRETATION
The laminar calcretes are considered to be pedogenic in origin both because they are intimately associated with pedogenic profiles and because they internally contain pedogenic fabrics. The striking similarities between the three types described indicate a common origin which we believe to be related to rootlet activity for the following reasons: ( I ) The spar-filled cylindrical fenestrae and tubiform pores either represent burrows or rootlet struc tures. The former origin seems unlikely because of the absence of burrow features such as menis cate fills or pellet concentrations. Furthermore, the walls are not simply soil matrix pushed aside but have a distinctive concentric structure for which no obvious burrow analogue can be found. (2) The laminar calcretes are underlain by cylindrical structures (stringers) which bifurcate downwards, a fabric suggestive of a rootlet origin. The stringers have the same internal structure as the calcrete. (3) The microstructure consists of concentric milli metre-diameter spar-filled cylindrical fenestrae or tubular pores with micritic and microsparitic coatings. Similar features have been described and illustrated from Quaternary rhizocretions (or 255
256
Biogenic laminar calcretes
--
-
HARDPAN CALCRETE OR BEDROCK
----
a
b alveolar-septal structJre
l==c-=:==::=-= ==�,_
micrite lamellae root
c Fig. 11. Schematic diagrams to show the occurrences of dense root mats. (a) the mat has formed above on impenetrable hardpan calcrete or bedrock. (b) the mat forms at or just below the water-table. (c) typical microfabric of calcified root mats as described in this paper (see text and Figs 3, 6 and I 0). The absence of large rhizocretions in the crusts described in this paper suggests that the vegetation cover consisted of relatively small plants.
DISCUSSION
The occurrence of rhizolite laminar calcretes in these three sequences requires comment. Root mats appear to form in two settings (Fig. I I): firstly, they occur above lithified zones, either bedrock or calcrete; secondly, they form in areas with very shallow phreatic waters where the roots extend to, or just below the water-table surface (Cohen, 1982; Mount & Cohen, 1984). The latter examples occur especially in shallow lacustrine or palustrine settings. The Miskin and Swindon examples occur on substrates which show evidence of extensive calcretization for example the development of grain coatings and calcrete matrix with floating sediment grains. The rhizolite horizons probably formed above a relatively impervious 'hard pan' layer. However, as noted earlier, some of the Cameros Basin horizons occur within a complex of floodplain-lake margin, palustrine settings and some may be found as discrete units within marl or lake margin carbonate sequences and not only in associa-
tion with calcrete profiles of pedogenic or capillary rise zone type. The marl and palustrine-associated forms probably represent the shallow phreatic type of root mat and may reflect the positions of former water tables. In this case their classification as pedogenic or capillary-rise type is ambiguous. Mount & Cohen (1984), in work based on their studies of the Plio-Pleistocene of Kenya, offered criteria for distinguishing rhizocretions formed in well-drained settings from those formed in shallow phreatic settings. Those formed in horizontal root mats at or below the local water-table lacked obvious vadose features such as meniscus and pendant cements, but contained abundant clay and plant debris, as well as Mn concentrations as high as 4·5 cation percent. These features indicate formation in water-saturated conditions with reduced Eh condi tions. Few horizons described in this study contain obvious vadose cements (although such cements occur in associated horizons) and the clay and organic
Fig. 10. Photomicrographs of laminar calcrete showing typical microfabrics. Sections are from specimen shown in Fig. 9b. (a) irregular and cylindrical fenestrae surrounded by micro-laminar micrite. Note cluster of pellets (arrowed). Scale bar represents I mm. (b) irregular spar-filled fenestrae and micro-laminar micrite. Note micritic cylinder within one of the pores. This probably represents the light calcification of the outer margin of a rootlet or root hair. Scale bar represents 0·5 mm. (c) detail of microfabric. Scale bar represents 0·5 mm.
257
V. P. Wright, N. H. Platt and W. A. Wimbledon
matter content of the Cameros material is no different to the other two types. Likewise the Mn content (as assessed from cathodoluminescence characteristics) is no different to the others. The pedogenic features and field associations of the Cameros calcretes seem to be more reliable palaeoenvironmental indicators than these other criteria. The Cameros rhizolite calcrete examples are partic ularly significant. The wide range of facies associations and palaeoenvironmental settings inferred for their occurrences demonstrates conclusively that they are not unusual features developed only under particular and specialized conditions. They occur in a wide variety of settings including inter-channel, floodplain, marginal and ephemeral lake environments. Calcified root-mat horizons are probably common in many ancient calcrete profiles and we have noted a number of descriptions and illustrations of very similar features in Carboniferous calcretes from the United States (Harrison & Steinen, 1978, fig. 6E; Goldham mer & Elmore, 1984, fig. 4C & D; Prather, 1985, p. 215). Strasser & Davaud (1982) have illustrated crudely similar horizons from the Purbeck facies of France. Similarly, possible Quaternary examples have been described briefly by Braithwaite (1975, p. 10 & 20) from the soils of Aldabra. One problem in recognizing these root-mat horizons is their highly 'fenestral' and biogenic fabric. This means that they could be mistaken for true stromato lites or tufas even more easily than other laminar calcretes. However, in the three occurrences described here the presence of alveolar-septal structure has been ubiquitous and may provide a useful criteria for recognizing fossil laminar root-mat calcretes.
mats developed on lithified calcretized horizons, while some from the latter area represent mats developed at shallow water-tables. Such laminar calcretes are referred to as rhizolite laminar calcretes and are probably common in ancient paleosols from peritidal, palustrine and floodplain deposits. They are also probably important compo nents of present day calcrete soils but have not been described in detail. As a result of their plant origin they have a strong biogenic fabric and can be easily mistaken for microbial stromatolites or tufas. An awareness of rhizolite laminar calcretes may prove useful not only for the recognition of subaerial exposure zones but also in palaeohydrological studies enabling fossil' water-tables to be indentified. They also provide evidence of rooted vegetation and surprisingly identi cal horizons have been found in Carboniferous to Cretaceous deposits despite probable differences in the plant originators.
ACKNOW LEDGMENTS
N. H. P. would like to thank: Dr H. G. Reading and many colleagues at Oxford for encouragement during the course of his work in Spain; colleagues in Madrid, Bilbao and Geneva for assistance in the field and NERC for financial support. W. A. W. gratefully acknowledges the help of and discussions with Dr D. Bridgland, D. Cripps and A. Adams, and also the latter for his assistance in having slides and peels produced. V. P. W. acknowledges the financial assistance of the Nuffield Foundation for Studies of Late Palaeozoic and Mesozoic paleosols. Fransesc Calvet (University of Barcelona) is thanked for helpful discussions. Sabrina Sadri and Jane Hawker typed the manu script, Simon Powell and Andy Wetherman prepared the photographs, and Alma Gregory and Pam Baldaro prepared the diagrams. Our thanks go to them all for their help. We especially wish to thank Judith McKenzie, Pierre Freytet and an anonymous referee for their constructive reviews of an earlier version.
CONCLUSIONS
Distinctive laminar calcretes containing abundant millimetre-sized, calcite-cemented, cylindrical fenes trae or tubular pores, separated by concentrically or crudely laminated matrix with alveolar-septal struc ture, have been found in three different settings: (i) within a thick calcrete and regolith profile from the Lower Carboniferous of South Wales; (ii) as a laminar horizon in calcretized marginal marine sandstones and limestones from the Upper Jurassic of Swindon, southern England; and (iii) within fluvial channel, floodplain and palustrine deposits of Upper Jurassic� Lower Cretaceous age from northern Spain. The laminar calcretes are interpreted as calcified root mats. The former two occurrences represent root
REFERENCES ARAKEL, A. V.
( 1 982) Genesis of calcrete in Quaternary soil " profiles, Hutt and Leeman lagoons, Western Australia. J. sedim. Petrol, 52, 1 09-125. BLUMEL, W.O. ( 1 982) Calcretes in Namibia and south-east
258
Biogenic laminar calcretes
coccoid cyanobacterium associated with the formation of desert stromatolites. Sedimentology, 26, 593-604. LATTMAN, L . H . ( 1 973) Calcium carbonate cementation of alluvial fans in southern Nevada. Bull. geol. Soc. Am., 84, 301 3-3028. MOUNT, J . F . & COHEN, A.S. ( 1 984) Petrology and geochem istry of rhizoliths from Plio-Pleistocene fluvial and marginal lacustrine deposits, east Lake Turkana, Kenya. J. sedim. Petrol., 54, 263-275. MULTER, H . G . & HOFFMEISTER, J . E . ( 1 968) Subaerial laminated crusts of the Florida Keys. Bull. geol. Soc. Am., 79, 1 83- 192. NETTERBERG, F. (1 980) Geology of southern African cal cretes : I . Terminology, description, macrofeatures, and classification. Trans. geol. Soc. S. Afr., 83, 255-283. PERKINS, R.D. ( 1 977) Depositional framework of Pleistocene rocks in south Florida. Mem. geol. Soc. Am., 147, 1 3 1-198. PHILLIPS, S . E . & SELF, P . G . ( 1 987) Morphology, crystallog raphy and origin of needle-fibre calcite in Quaternary pedogenic calcretes of South Australia. Aust. J. Soil Res. , 25, 249-444. PLATT, N . H . ( 1 985). Fresh-water limestones from the Cameros Basin (North Spain). 5th Reg. Mtg. int. Ass. Sedim. , Lerida (Spain, 1985, Abstr.), 363-366. PLATT, N . H . ( 1 986). Sedimentology and tectonics of the
Spain : relations to substratum, soil formation and geo morphic factors. In : Aridic Soils and Geomorphic Processes (Ed. by D. H. Yaalon). Catena (Suppl.), 1, 67-82. BRAITHWAITE, C . J . R . ( 1 975) Petrology of palaeosols and other terrestrial sediments on Aldabra, western Indian Ocean. Phil. Trans. R. Soc. B., 273, 1-32. BRAITHWAITE, C.J . R. ( 1 983) Calcrete and other soils in Quaternary limestones : structures, processes and applica tions. J. geol. Soc. London, 140, 3 5 1 -363. CALVET, F. & JULIA, R. ( 1 983) Pisoids in the caliche profiles ofTarragona (north east Spain). I n : Coated Grains (Ed. by T . M . Peryt), pp. 456-473. Springer-Ver)ag, Berlin. CALVET, F., POMAR, L. & EsTEBAN, M . ( 1 975) Las rizocre ciones del Pleistocene de Mallorca. Inst. Invest. Geol. Univ. de Barcelona, 30, 35-60. COHEN, A.S. ( 1 982) Paleoenvironments of root casts from the Koobi Fora Formation, Kenya. J. sedim. Petrol., 52, 401 -4 1 4. CONIGLIO, M. & HARRISON, R.S. ( 1 983) Holocene and Pleistocene caliche from Big Pine Key, Florida. Bull. Can. Petrol. Geol. , 31, 3-1 3. ESTEBAN, M . &. KLAPPA, C.F. ( 1 983) Subaerial exposure environment. I n : Carbonate Depositional Environments (Ed. by P. A. Scholle, D. G. Bebout & C. H . Moore). Mem. Amr. Ass. Petrol. Geol., 33, 1-54. FREYTET, P. & PLAZIAT, J-C. ( 1 982) Continental carbonate sedimentation and pedogenesis-late Cretaceous and early Tertiary of southern France. Contr. Sediment., 12, 2 1 3 pp. E. Schweizerbart'sche Verlagsbuchhundlung, Stuttgart. GARCIA-MONDEJAR, J . , HII'IE�, F . M . , PUJALTE, V. & READ ING, H . G . ( 1 985) Sedimentation and tectonics in the western Basque-Cantabrian area (northern Spain) during Cretaceous and Tertiary times. Excurs. Guide, 6th Eur. Reg. Mtg, int. Assc. Sedim. , 307-392. G ILE, L.H., PETERSEN, F.F. & GROSSMAN, R.S. ( 1 966) Morphological and genetic sequences of carbonate accu mulation in desert soils. Soil Sci., 101, 347-360. GOLDHAMMER, R.K. & ELMORE, R.D. ( 1 984) Paleosols capping regressive carbonate cycles in the Pennsylvanian Black Prince Limestone, Arizona. J. sedim. Petrol., 54, 1 1 24- 1 1 37. GouDIE, A.S. ( 1 983) Calcrete. In : Chemical Sediments and
Western Cameros Basin, Province ofBurgos, Northern Spain.
Unpubl. DPhil. Thesis, University of Oxford, 250 pp." PRATHER, B.E. (1 985) An Upper Pennsylvanian desert paleosol in the O-Zone of the Lansing-Kansas City Groups, Hitchcock County, Nebraska. J. sedim. Petrol., 55, 2 1 3-22 1 . READ, J.F. ( 1 974) Calcrete deposits and Quaternary sedi ments, Edel Province, Shark Bay Western Australia. Mem. Am. Ass. Petrol. Geol. , 22, 250-282. READ, J . F . ( 1 976) Calcretes and their distinction from stromatolites. In : Stromatolites (Ed. by M. R. Walter). Developments in Sedimentology, 20, 255-27 1 . Elsevier, Amsterdam. RIDING, R. & WRIGHT, V.P. ( 1 9 8 1 ) Paleosols and tidal-flat/ lagoon sequences on a Carboniferous carbonate shelf. J. sedim. Petrol., 51, 1 323- 1 3 39. SALOMON, J. ( 1 982) Les formations continentales du Juras sique superieur-Cretace inferieur (Espagne du Nord, Chaines Cantabrique et NW Iberique. Mem. Geol. Univ. Dijon, 6, 228 pp. SALOMON, J. ( 1 984) Paleopedologie et redistributions dans les formations continentales du Jurassique superieur du Bassin de Soria (Espagne). Abstr. 5th Eur. Reg. Mtg. int. Ass. Sedim. , 393-394. SEMENIUK, V. & MEAGHER, T.D. ( 1 9 8 1 ) Calcrete in Quater nary coastal dunes in southwestern Australia : a capillary rise phenomenon associated with plants. J. sedim. Petrol., 51, 47-68. SEMENIUK, V. & SEARLE, D.J. ( 1 985) Distribution of calcrete in Holocene coastal sands in relationship to climate, southwestern Australia. J. sedim. Petrol., 56, 86-95. STRASSER, A. & DAVAUD, E . ( 1 982) Les croutes calcaires (calcretes) du f'urbeckien du Mont-Salt!ve (Haute Savoie), France. Eclog. geol. Helv., 75, 287-30 1 . SYLVESTER-BRADLEY, P.C. ( 1 940) The Purbeck Beds of Swindon. Proc. geol. Ass., 5 1 , 349-399. WIEDER, M. & YAALON, D.H. ( 1 982) Micromorphological
Geomorphology : Precipitates and Residua in the Near Surface Environment (Ed. by A. S. Goudie & K. Pye), pp. 93- 1 3 1 .
Academic Press, London. R.S. ( 1 977) Caliche profiles : indicators of near surface subaerial diagenesis, Barbados, West Indies. Bull. Can. Petrol. Geol., 25, 1 23-173. HARRISON, R.S. & STEINEN, R.P. ( 1 978) Subaerial crusts, caliche profiles, and breccia horizons : Comparison of some Holocene and Mississippian exposure surfaces, Barbados and Kentucky. Bull. geol. Soc. Am., 89, 38 5-396. JULIA, R. & CALVET, F. ( 1 983) Description e interpretacion de las texturas y microtexturas de caliches recientes del camp de Tarragona y Penedes (Catalunya). Libra Jubilar J. M. Rios, 3, 6 1 -96. I . G . M . E . KLAPPA, C . F . ( 1 979) Lichen stromatolites : criterion for subaerial exposure and a mechanism for the formation of laminar calcretes (caliche). J. sedim. Petrol., 49, 387-400. KLAPPA, C.F. ( 1 980) Rhizoliths in terrestrial carbonates : classification, recognition, genesis and significance. Sedi mentology, 27, 6 1 3-629. KRUMBIEN, W . E . & G!ELE, C. ( 1 979) Calcification in a HARRISON,
259
V. P. Wright, N. H. Platt and W. A. Wimbledon
fabrics and developmental stages of carbonate nodular forms related to soil characteristics. Geoderma, 28, 203-
(Ed. by J. C. W. Cope). Spec. Rep. geol. Soc. London, 15. V.P. ( 1 986) The role of fungal biomineralization in the formation of early Carboniferous soil fabrics. Sedimen
WRIGHT,
220. W . A. ( 1 976) The Portland Beds (Upper Jurassic) of Wiltshire. Wiltshire archaeol. nat. Hist. Mag. ,
WIMBLEDON,
tology, 33, 8 3 1-838. (I 987) The ecology of two early Carboniferous paleosols. I n : European Dinantian Environments (Ed. by J . Miller, A . E . Adams & V. P. Wright), p p . 345-358. Geol. J. Spec. Issue, 1 2. Wiley, London.
WRIGHT, V .P.
71, 3-1 1 . W IMBLEDON, W.A. ( 1 98 1 )
In : A correlation of Jurassic Rocks
in the British Isles. Part Two : Middle and Upper Jurassic
(Manuscript received 15 June 1987; revision received 26 January 1988)
260
ASPECTS OF CALCRETE PETROGRAPHY
This set of papers illustrates a range of fabrics and
floating, fractured and partially replaced silicate
processes associated with calcretes.
grains interpreted as vadose in origin. The style of
Tandon & Narayan describe authigenic carbonates
fracturing of grains described in this paper is also
from the late Cainozoic of India. Coated-grain car
widely seen in other 'clast' types in calcretes; for
bonates formed in channels while calcretes developed
example in the 'Old Red Sandstone' calcretes de
on floodplains. Spar-cemented conglomerates occur
scribed by Tandon & Friend in this volume (see
also and probably represent groundwater calcrete.
their fig. 6a). This fracturing is a product of the
The pedogenic calcretes are more finely crystalline
tendency of calcite (and gypsum) to grow displacively
than the groundwater forms which are pore-filling
as mosaics, probably a function of the inability of
cements. Similar meteoric cementation occurs in
calcite to form adhesive bonds with non-carbonate
many Quaternary continental alluvial deposits yet
grains; such grains are more likely to be displaced
records of similar cements in the older geological
rather than be simply cemented by calcite growth. It is displacive growth which characterizes alpha
record are relatively rare. The role of displacive growth is discussed by
fabrics. Tandon & Friend describe alpha calcrete fabrics
Buczynski & Chafetz. The calcite cements contain
Fig. 15. Calcite crystals from Quaternary calcrete, Blackdom surface, near Carlsbad, New Mexico. The morphology reflects phases of crystal growth and dissolution. Many crystals clearly show evidence of considerable etching. Such phases of growth and dissolution have been detected in ancient calcretes using cathodoluminescence (see Wright & Peeters, 1989).
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
261
Aspects of Calcretes Petrography formed in late Devonian to early Carboniferous
Wales. This micritic calcrete developed on shallow
palaeo-vertisols.
of
marine limestones and, while micritization has oc
replacive and displacive growth of pedogenic car
curred, much of the micrite is revealed under catho
They
document
the
effects
bonate, as well as possible neomorphism and dis
doluminescence (CL) to be needle-fibre calcite. By
solution. Direct evidence of biogenically-induced
using CL these authors are able to show the complex
precipitation is absent. Their descriptions of neo
phases of formation including several dissolution
morphic processes are particularly interesting and
events. These were generally confined to the walls of
they note the presence of relic micritic cathodolu
larger pores and brecciation features which acted as
minescence patterns within the neomorphic rhombic
fluid
calcites. What is unclear is how this relic pattern is
around
pathways. such
Extensive
conduits
dissolution
whereas
little
can
occur
dissolution
subsequently removed from these rhombs (those
occurs around the interparticle areas; this reflects
farther away from the neomorphic front apparently
the increased dissolution rates in conduit systems
lack these relics; their fig.
7)
where flow is rapid and turbulent as against the
without the loss of the
slower rates of dissolution achieved by intergranular
delicate growth zones in the rhombs.
laminar flow.
Solomon & Walkden provide a detailed study of a late Mississippian (Asbian) calcrete from North
262
Reprinted from
Sedimentology (1981) 28 353-367
Calcrete conglomerate, case-hardened conglomerate and cornstone - a comparative account of pedogenic and non-pedogenic carbonates from the continental Siwalik Group, Punjab, India S. K. T AN D ON and D EV EN D R A N A R AY AN
Department of Geology, University of Delhi, Delhi 11000 7, India
ABSTRACT
The occurrence of authigenic carbonates formed in three different environmental situations, within the continental Siwalik Group, has been used to compare the lithological and petrographic characters of the contrasted lithofacies. The three lithofacies are: (I) calcrete conglomerate, (2) case-hardened conglomerate, (3) cornstone (pedogenic, nodular calcrete). The calcrete conglomerate facies laterally intertongues with the channel conglomerates. It consists of pisolites which are interpreted to have formed from carbonate-rich spring waters emerging on to the gravelly substrate of dry, abandoned channels. The laminae characteristics of these pisolites are distinctly different from those of marine origin and also from comparable biogenic materials. Case-hardened conglomerate occurs in the youngest part of the Siwalik stratigraphic column, in boulder conglomerates having limestone as the principal clast component. This lithofacies has resulted from cementation of the conglomerate through continued dissolution and re-precipitation of calcite, by meteoric water, downwards from the surface. lt displays a coarsely crystalline, sparry calcite cement with no evidence for displacive growth or replacement by calcite. Cornstones (nodular calcrete) occur in several sedimentary cycles of the Middle Siwalik Sub-Group. These are immature and commonly associated with thinly-bedded sandstones (levee) and red shales (overbank). This lithofacies is a result of concentration of carbonate through capillary action associ ated with pedogenic activity. Ooids developed in cornstone are essentially micritic in nature and usually composed of less than five indistinct laminae.
INTRODUCTION
This study aims to compare the non-pedogenic and pedogenic continental carbonate rocks from the Siwalik Group of a part of the Sub-Himalayan zone in Punjab. The Siwalik Group consists of an exten sive suite of Middle Miocene to Late Pleistocene fluviatile sediments flanking the southern fringe of the Lesser Himalaya and forming the outermost range of the Himalayan Chain. Lithologically, the Siwalik Group consists of sandstones, inte rbedded fine clastics and conglomerate. Tandon & Rangaraj (1979) recorded the presence of calcrete conglomer ate, cornstone (pedogenic, nodular calcrete) and Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
carbonate-cemented sandstones at various strati graphic levels in the Siwalik Group in Punjab. The study of calcrete has gained considerable im portance because of its widespread occurrence in semi-arid climates-an association which has been used as an important factor in palaeoenvironmental and palaeoclimatic studies. Quaternary calcrete is being increasingly documented from various regions of the world (Goudie, 1971, 1973; Reeves, 1976). Secondary carbonate accumulations in the conti nental rocks generate overlapping textural and
263
S. K. Tandon and D. Narayan
N 6
... .
Baklah
5
2
6
Dhar Kalan 2 Dhar Khurd
5
.
0
1 km
...__.
CD Upper S iwo I ik format 10n 8 C1:J Upper S1wOI1k formot1on A [TI Middle Siwalik formot1on C [TI M1ddle Siwol1k formatiOn 8 CD M1ddle S1wOIIk formation A [TI Lower Siwalik sub- group IZJ Thrust plane Fig. 1.
� Calcrete c onglomerate � C ornst one � Case-hardened boulder
c onglomerate
�Recen t calcareous tufa copping on � r i chly carbonate c emented sandstones
Location map showing the important carbonate lithic types of the Siwalik sequence.
structural frameworks, leading to confusion in the interpretation of ancient sequences. The purpose of this study is to document and compare the field and petrographic characters of various continental·, carbonate rocks from an essentially fluviatile, molassic sequence. Significantly, pisolites compar able to the Holocene pisoliths, described by Risacher & Eugster (1979), are being documented from the ancient rock record for the first time. The locale of this study is delineated by latitudes 32° :12' -32° 30'N and longitudes 75° 45' -7 0° 55' E ( Fig. 1 ).
TERMINOLOGY
The terminology of calcretes and related rocks is a difficult problem because several terms are in use for terrestrial materials cemented by calcium carbonate. In contrast to the ambiguously used term 'caliche', calcrete has always been used specifc i ally for materials cemented by calcium carbonate (Goudie, 1972). Goudie has excluded the usage of 'calcrete' for speleothems, spring deposits, beach-rock (marine deposits ) and lacustrine algal stromatoliths. The 264
Carbonates from the Siwalik Group, India
source for carbonate supply of the pisolites, in the calcrete conglomerate under description, has been related to spring-fed water, but the formation of the pisolites has taken place primarily within the vadose zone. A second generation, inter-pisolitic cement of phreatic origin is also present. However, this cement is a relatively minor component, and since the major processes of accumulation of calcium carbonate took place in the vadose zone we prefer to use the term 'calcrete conglomerate'. The specific connotation of the terms used for the description of the various types of carbonate rocks is given below: Calcrete conglomerate (after Goudie, 1972): a term for terrestrial materials consisting dominantly of calcium carbonate, ranging in states from nodular to highly indurated and involving cementation of, accumulation in and/or replacement of greater or lesser quantities of soil, rock or weathered material, primarily within the vadose zone. Cornstone (after Buckland, 1821 and Steel, 1 974): consists of concretionary carbonate in host rocks of Table l.
a marly character, occurs preferentially in associa tion with floodplain deposits, particularly within the upper parts of fluvial, fining-upward cyclothems and is analogous to pedogenic caliche (calcrete ). Case-hardened conglomerate (after Lattman, 1 97 3): preferential cementation of the surface of alluvial and colluvial deposits, where exposed in vertical or steep slopes. It is formed by surface water flowing over steep exposures causing solution of calcareous fines and redeposition as cement.
GEOLOGICAL BACKGROUND
The Siwalik Group of this area has been subdivided into six lithostratigraphic units by Tandon & Rangaraj (197 9). A summarized account is given in Table I. The fluvial nature of these clastic sediments is clear from the sedimentary organization of channel facies (sandstones ) and interbedded floodplain facies (mudstone/siltstone/shale). The presence of typical fining-upward cycles along with large channel struc-
Correlation between the lithostratigraphic succession in the area of study and the type succession from Potwar, Pakistan
Siwalik Group Upper Siwalik Sub-Group
Middle Siwalik Sub-Group
Lower Siwalik Sub-Group
Age
Potwar
Area of study
Lithological content
Lower Pleistocene
Boulder conglomerate
Upper Siwalik Formation B
Approx. 600m thick-<:onglomerate with clasts of varying lithologies and coarse sand intercala tions (the upper 150-200m show the pheno menon of case-hardening)
1·3-3 Myr (Keller et a!. 1977)*
Pinjor-Tatror
Upper Siwalik Formation A
Approx. 500m thick-light brown to buff coloured, medium-grained, calcareous sand stones; pebble to cobble conglomerate and coarse siltstones (about 2 m thick horizon of calcrete conglomerate)
8·6 Myr (Barndt eta!., 1978)
Dhok-Pathan
Middle Siwalik Formation C
Approx. 200m thick-grey coloured, medium to coarse-grained, pebbly sandstone; sandy con glomerate, mudstone and shale (irregular patches of calcrete conglomerate)
Nagri
Middle Siwalik Formation B
Approx. 900m thick-sandstone; siltstone; mudstone; shale and marl (cornstone horizons associated with the overbank facies of fining upward c:rcles)
10·2 Myr (Barndt eta!., 1978)*
Chinji
Middle Siwalik Formation A
Approx. 900m thick-grey, richly micaceous and calcareous, thick sandstones and shale-chip conglomerates. Interbedded siltstones and shales (calcareous sandstones with carbonate concre tions)
Middle Miocene
Lower Siwalik
Lower Siwalik Sub-Group
Maroon to reddish-brown calcareous shales with interbedded sandstones (carbonate clots in shale)
* These ages refer to the chronology described for sections which lie 150-200 km to the west of the sequence under consideration, and their extrapolation is approximate. Parentheses indicate the dominant carbonate type at various stratigraphic levels.
265
S. K. Tandon and D. Narayan
tures also affords adequate evidence of fluvial origin. Johnson et a/. (I 979) have indicated that the Siwalik sediments represent complex stream systems of the Neogene and Quaternary which are akin to the sediments developed today in the Indo-Gangetic plain. Changes in river patterns from fluvio-deltaic to entrenched, meander-belt type to braided stream and alluvial-fan types are demonstrable for the Siwalik Group. Tandon & Rangaraj (1979) have shown that the Siwalik fluvial sequence represents a basinward shift of the depositional system in time.
formations of the Siwalik Group. Carbonate cemented and profusely concretionary sandstones occur particularly in Middle Siwalik Formation A and Upper Siwalik Formation A. The best exposure of calcrete conglomerate is in the southern belt of the Upper Siwalik Formation A ( Fig. 1). Case-hardened conglomerates occur beneath the fault scarp face near Bakloh and cornstones are well developed in the Putoli Khud section ( Fig. 1). Non-pedogenic carbonate rocks
Calcrete conglomerate
DESCRIPTION OF THE CARBONATE PROFILES
Secondary carbonate accumulation is observed in all
The calcrete conglomerate facies occurs as discon tinuous, pod-shaped bodies within the channel con glomerate of the Upper Siwalik Formation A. The host rock is a sandy conglomerate consisting of
--1 - -- l -
-_
_
l - - -:1
-------
--------1
A
B Index
�� @:� j·�luo/j E= J ����
Calcrete
Cornstone /Carbonate clots Red-shale Muds tone
� F1ne
L.:.:..:.J
conglomerate
gramed muddy
sandstone
� Med1um to sandstone
L.:..:.:J
� �
coarse
grained
Red indurat e d mudstone Large-scale
cross beddmg
� P lane lamination/Horizontal � bedd1 ng
� �
E=:J j;,; 0 qj [ID
Scoured surface Boulder conglomerate Limestone clast
�Coarse grained calc1te cement �and colc1te ve1ns
Small··scale r1pple crosslam1nat10n
Fig. 2. Vertical stratigraphic sections showing the nature of occurrence of principal continental carbonates. Locations
of the sections: (A) calcrete conglomerate near Patuan; (b) calcrete conglomerate in Chakki river; (C) cornstone in Putoli river and (D) case-hardened conglomerate near Bakloh (refer to Fig. I for locality names).
266
Carbonates from the Siwalik Group, India
clasts of quartzite, sandstone, shale and volcanic rock fragments. It is underlain, with a sharp contact, by fine-grained mudstone (Fig. 2A). Laterally; the calcrete conglomerate horizon varies within a few tens of metres from completely cemented/replaced sandy conglomerate to partially replaced conglomer ate. This unit grades into the channel conglomerate facies when traced along the strike. In the Chakki section (Fig. I), the rocks under- and overlying the calcrete conglomerate consist of a poorly cemented, buff-coloured, medium-grained sandstone (Fig. 2B). The calcrete conglomerate consists mainly of pisolites of varying form and size (0·1 to over 8 em) in a partially altered matrix. The pisolites occur in clusters of hundreds and are oval/elliptical/circular in outl ine. The number and thickness of concentric laminae around the nucleus vary widely. The con. centric structure consists of alternating light- and dark-coloured laminae. The nuclei of the pisolites consist of varied rock types ranging from mudstone, siltstone, shale, metamorphic and volcanic rock fragments to even earlier formed aggregates of pisolites. In some of the pisolites, the size of the nucleus is very large, with only a thin rind covering it. Petrology of the calcrete conglomerate
The calcrete conglomerate consists mainly of piso lites of varying size, detrital grains and sparry calcite cement (Fig. 3). The pisolites exhibit a concentric habit with many individual� having 20-30 distinct laminations around a detrital nucleus (Fig. 4). The
shape and size of the nucleus displays a strong control on that of the rind. The inner laminae tend to follow the outline of the nucleus. This control exerted by the shape of nucleus, however, diminishes outwards from the inner laminae. X-ray diffraction analysis and Alizarin Red- S stained thin-sections indicate that the rind is composed entirely of calcite of alternating light and dark colour. The darker laminae consist of extremely fine-grained (1 -3 1-1m) calcite, whereas the light-coloured laminae consist of sparry calcite which shows a transverse growth from the wall of the light-coloured laminae. Micritic laminae are probably the result of rapid precipitation from supersaturated solutions and surface accretion of very fine-grained sediment (Harrison, 1977; Braithwaite, 1 979). The light-coloured laminae are composed of bladed sparry calcite (20 1-1m in length). This bladed calcite, with its coarser grain size, is probably the result of extreme supersaturation, slower precipitation and the presence· of impurities in the solution. Due to the presence of micro- to submicro-inclusions, most of the sparry crystals are cloudy. The presence of meniscus and 'dripstone' cement (Fig. 5) clearly indicates that the formation of these pisolites took place in the vadose environ ment. In several pisolites, the laminae have a tendency to be wavy and bunched, re flecting the nature of microtopography of the original surface on which the calcite has precipitated. Presumably their forma tion was either due to differential precipitation on the pre-existing surface, due to quiescent conditions
Fig. 3. An overview of the pisolitic calcrete conglomerate. Note the stylolitic seam in lower left corner (thin-section, crossed nicols; scale bar
=
200 1-1m).
267
S. K. Tandon and D. Narayan
Fig. 4. A large pisolite from the calcrete conglomerate showing more than 20 distinct laminations (negative 1 em). prJnt from an acetate peel; scale bar =
( Donahue, 19 78) or due to re-solution of calcite from that precipitated earlier on this surface (Braithwaite, 1 979). Bun-shaped and cabbage-like aggregations of laminae, simulating the morphology of biogenic structures like stromatolites, are commonly present.
as well as the nucleus, especially in pisolites which have a mudstone nucleus. These fractures have been healed by the second-generation calcite which also cemented the pisolites. Obviously, these fractures were produced after the formation of the rind. Inter penetrating pisolites (Fig. 6) and formation of stylolitic seams are also .observed. In view of the evidence in favour of pressure solution, passive vadose compaction of the type proposed by Clark (19 79) did not occur. Mechanical indentation (note the plastic deformation of laminae in Fig. 6) also indicates the presence of directed, external pressure. The pressure needed to produce these features was, probably, related to overburden, to which these pisolites were subjected after the rind formation and prior to complete precipitation of the second generation calcite cement. The inter-pisolitic area is filled up with a second generation sparry calcite cement exhibiting an equigranular, crystalline, mosaic fabric. Individual crystals are roughly e quidimensional, light yellow in colour, and range in size from 100 to 250 1-1m. The texture of this cement suggests it to be of phreatic origin. This cement, however, is a minor component and does not exceed 10-20% of the total carbonate. Replacement of detrital grains by calcite is limited to corrosion of margins of quartz grains and, in places, partial replacement of mudstone r0ck fragments. Brecciated detrital grains, the displaced parts of which can be matched (Fig. 7) indicate that dis placive calcite growth also took place.
For purposes of comparison, and also to preclude an organic origin for these pisolites, thin-sections of e quatorial planes of concentric algal forms from various formations in India were studied. An imme diately revealing contrast is that the detrital content of algal materials is not only considerably less than that of the calcrete conglomerate, but also the range of variation in size and lithology of the detrital com ponent is less in algal structures. Other points of difference include slightly diffuse laminae; transverse growth of calcite crystals from laminae interfaces is rare in biogenic materials; laminae thickness is highly variable; colour differentiation between laminae is more subtle; and thickening and thinning of laminae is far more variable and common. In situ fracturing in the laminae of the pisolites as well as the nucleus has also been observed. Most of these fractures have been healed by sparry calcite. Some of the fractures in the nuclei were healed prior to rind formation. Other fractures cut across the rind
Case-hardened conglomerate
The conglomerate, belonging to the Upper Siwalik formation B, shows the p henomenon of case hardening. In general, the boulder conglomerates show little cementation, which is evidenced by the rapid rate of erosion of the clasts. However, the uppermost 200m of this unit shows an increase in limestone clast component up to 8 0%. Such out crops are observed to be cemented strongly ( Fig. 2D ). Continued dissolution and reprecipitation of calcite, by meteoric water percolating down, has led to selective cementation of vertical scarp faces. This process is still continuing, and road cuttings which are about 5 0 years old also show case-hardening. The depth to which calcite has precipitated seldom exceeds 5 em. Calcite veins, consisting of a coarsely crystalline, white variety are common. Ill-sorted, muddy-matrix-rich conglomerates are relatively weakly cemented. 268
Carbonates from the Siwalik Group, India
Fig. 5.
Pisolite showing typical 'dripstone' cement indicative of vadose cementation. Calcrete conglomerate (thin 200 �m). section, crossed nicols; scale bar =
6. Smaller pisolite penetrating into the rind of a larger pisolite. Note the bending and breaking of laminae of the larger pisolite due to the pressure exerted by the smaller pisolite (thin-section, crossed nicols; scale bar= 50 �m).
Fig.
placive found.
Petrology of the case-hardened conglomerate
The framework of the case-hardened conglomerate consists of rock fragments of various metamorphic
growth
or
recrystallization
has
been
Pedogenic carbonate rocks
rocks and limestones. The inter-clast area is filled up with well-developed, equant and bladed calcite (Fig. 8) ranging in size from 150 �m to over I· 5 mm. The sparry calcite occurs as a pass ive, pore-filling cement. Individual crystals show well-developed cleavages and straight, intercrystalline boundaries. Micrite is absent. No evidence of replacement by calcite, dis-
Cornstone
The cornstones are well developed in the Middle S iwalik Formation B. These are associated with th inly-bedded levee sandstones and overbank red shales and occur in the upper part of the fluvial fining-upward cycles (F ig. 2C). Light grey-coloured, 269
S. K. Tandon and D. Narayan
Fig. 7. Brecciated detrital quartz grains with matching outlines. Fracturing attributed to displacive growth of calcite
(thin-section, crossed nicols; scale bar
=
50 IJm).
Fig. 8. Sparry calcite cement from the case-hardened conglomerate showing well-developed cleavage and straight, intercrystalline boundaries (thin-section, crossed nicols; scale bar 200 1-1m). =
medium-grained, channel sandstone showing well developed cross-stratification, forms the base of these cycles. These are succeeded by overbank deposits of very soft and friable red shales. The average thick ness of a cycle is 30 m. Cornstone occurs as bands of hard, indurated, fine-grained carbonate within the red shales. The host rocks are highly calcareous silt stone and mudstone. Abundant carbonate clots or segregations, ranging in size from less than I to over 5 em, are distributed randomly in these bands. There are, on average, two to three thin cornstone profiles
in the overbank facies of each cycle. The bands are less than I m thick. Most of the cornstone consists of small (1 --6 em ) irregular carbonate clots corre sponding to Type I of Steel (197 4 ). The carbonate segregations are also observed to extend downwards in the form of pipes. There is a marked increase in the carbonate content vertically upwards through the profile. The association of cornstone with levee sand stone indicates frequent breaching of the channel banks and consequent inundation, arresting the development of the cornstone profile. 27 0
Carbonates from the Siwalik Group, India
Fig. 9. Cornstone showing a secondar� mud-support fabric. The dark patches are authigenic ferruginous glaebules
Fe-oxide stained micritic masses (thin-section, crossed nicols; scale bar
=
50 1-1m).
Fig. 10. Micritic ooids developed in cornstone. Note the fuzzy laminae (thin-section, crossed nicols; scale bar
=
50 1-lm).
However, the presence of detrital quartz grains which show in situ brecciation, due to introduction of calcite, indicates that displacive growth of calcite was also important. No significant biogenic. activity that could have caused mechanical dislocation (Klappa, 1979), has been observed. The micrite has been partially recrystallized, giving rise to neomor phic microspar (10-15 1-1m). Voids are filled with microspar and, in places, radial arrangement of. crystals, similar to 'flower spar' described by James ( 1972), is also observed. Microspar calcite crystals
Petrology of the cornstones
The cornstone is composed of corroded, detrital, grains floating in a micritic matrix (Fig. 9) which, in places, shows a peloidal habit. Microspar channels are common and, where the crystal size is larger, may display circumgranular cracking. The floating tex tL!re, in these cornstones, is explained by partial to total replacement of the fine-grained overbank sedi ments, while the larger but sparsely distributed detrital quartz grains show only corroded margins. 271
. '
..
.,
S. K. Tandon and D. Narayan
.····
'.... •.
•
.
·' .·
. . :.
f
. .
..
.
... .
.
.
·
..
.
. '
.
r.
.
.
.
"'" �.
1.'1
l·
. .· .
.
. , •
.
·'.If'
'.
.
;
.
.
.,
:.
..
.
.
I
. '
.
� '
.
.
. ..
'
.
.
.....
..
..
r1,.·-" �� . ... . •�
·.. .
. . __
.
.. ' ... <
.
• 1�, '· • •• .
,.
! "'·-.....
-� : ·.
•
• �-
. . .. " : .·- . " ;.. •
'
'•
' ' ·-:�..' r .> ·•-:1'-:.·. :. .
. .
·" ' .. ,. ·'
.
.
. . �.
"'
.
. .
:� ;c (
.·
. ,, .•
. . ..
"
.. . I
"· ·· "' .
'
' .��}t •. ',,.."-"!: "':"At _ .. . ' . ;· \' � · " � · . ,, . -;...\"'•.. ; ·f!�-�-'• -.- ,... _:d!! · - · .. . ·rrJ _, ::"'·. · .�·'1. : � �� -;.t . � ';(. : • . .. · A.... -.. � ·� ;. . t ., . �- � �1:.·� �.;, ,.-··.-· � -,;� :; . ii!J. ...:. � f' � '� : · 1? ..,;,� -·..-;,.� t . . .,.. ,;., 'l!':i��. ,"',�: � . .
.
•..;./ ·'� � {.
· c . ' �,· .;;-: ::..,1.,·'1· . ·,,.; ..�/t-' .� ... . -� . ..,., , .. . : >. .• ' , ; '. .'. ,---,· ·..,j • . ._ """"' ·-· ,.. ' . . .,...._ , ., '
·. :
r-· ....
. ...
"'
.. . .
•
•
. >,
'•
•
. . . . ' ·. .• . • ., : I .. \. � _,. � ·· ' .. '>.. • . ,. .• . ·_, . ... . . ·' . '. _.,. ·» •.. ' .. . :�. •• . ' _ �.... ·. '. . .. '' .. .... . . ,,_ . . � . . _, '· "' .• . ' . ·. , � _. ., .. .. • ' . . .. . .. •, . .. .. - . •. • .'. . ,_ "' ·•. ....'. ' <•·''·; . ' . ... ... . . . · . : . 'fi!" �' ., .. . flt!l •, ' ... . • " ' •. . , .{:· .. . ,. ' . ',. "' � . _, , . "- -.. · • , , .,,_,. ' t. . ,·. . •. � .. "' . ·· ·i .. .�:.:.� ? •' '"'' .. ........·.? _ ·':tf • • ' • ·� · .... . ... :'1.,; , •. ·: '. 1\• •...: · � � .. ' ._. ·'' ' . -.. • • • J:.' . . ', . ·� . ,. • • •.;.# ...,.,. . . '-J ti.r< &;:4 • • .. .,. .J..
!
.
.
.,f . . .
,- ·
. ·!· '
Fig. 11 Recent kankar (nodular calcrete) displaying typical mud-support fabric (corroded detrital grains floating on . 50 �m). a micritic matrix (thin-section, crossed nicols; scale bar =
...
...
,
"'
...
Oetntal gra1ns
Legend· r:l Calcrete L:J conglomerate
.
•
M Case- hardened conglomerate
�
� �
Cornstone Recent
konkar
Q Biogenic laminated L2.J structures ff=il Oato from published work
l-=:J
(I) Knox
(1977)
(2) Sehgal and Stoops (1972)
( 3) (4)
Watts
(1978)
McGannon ( 1975)
(5)0ohanoyke
Mic ritiC
COIC1te
Sparry
(1977 )
COIC1te
Fig. 12. Triangular diagram based on the relative contents of detrital grains/micritic calcite/sparry calcite of the various continental carbonate rocks. Note the distinctly separate fields of the case-hardened conglomerates and equatorial sections of concentric stromatolites.
commonly show irregular, craggy boundaries. In some of the cornstones, ooids (concentric glaebules) are developed. In sharp contrast to the pisolites of the calcrete conglomerate facies, these ooids are composed almost entirely of micrite and are indis tinctly laminated (Fig. 1 0). The laminae are fuzzy, less than five, and are defined by subtle variations in
colour of the dense micnttc carbonate. These features, as stated by Esteban (I 9 76), are character istic of caliche pisolites. For purposes of comparison, a few samples of Recent 'Kankar' (Nodular Calcrete), developed in various localities in Haryana, India, were studied. Thin-sections revealed a micrite-supported fabric 272
Carbonates from the Siwalik Group, India
Table 2. Comparative petrographic features of the main continental carbonate rocks of the Siwalik sequence
Carbonate type
Laminae characteristics
Calcrete conglomerate
Pisolites with Mainly sparry alternate light- and calcite, replacement limited to dark-coloured laminations; corrosion of boundaries distinct, margins laminae more than 10, lighter laminae consist of radially grown sparry calcite. Darker laminae consist of micrite; laminae tend to be spherical in outline away from the nucleus
Case-hardened conglomerate Cornstone
Pisolites rare, laminae indistinct, micritic in nature, less than five in number
Type of calcite cement
Detrital grains
Other features
Mainly quartz and rock fragments. Detrital grains show microfaulting and in situ brecciation
Holocene pisoliths (associated with spring-fed surface pools) of Risacher & Eugster (1979) also show alternate clear and dark bands. Clear bands consist of radially grown sparry calcite. Size up to 20 em, laminations distinct and more than 10 in number. Primary fluvial oolites of McGannon (1975) have sharp boundaries, as many as 20 laminations and consist of alternate light and dark laminae
Exclusively sparry calcite
Mainly limestone rock fragments and quartz
No corrosion of detrital grains. Calcite occurs as an inert porefilling cement
Mainly micritic calcite
Corroded quartz grains floating on a micritic matrix
Overall mud-support fabric. Authigenic ferruginous clots probably formed by the alteration of detrital biotite. The spar is neomorphic in origin
geological formations in India. Data on modal com position have also been generated by analysing pub lished photographs (given in the works of Sehgal & Stoops, 197 2; Steel, 1974; McGannon, 1975; Dahanayake, 1977; Knox, 1977; Watts, 197 8). A total of thirty-nine analyses have been plotted in a triangular diagram with detrital grains, micritic carbonate and sparry calcite as the three compo nents. The detrital grains in the calcrete conglomerate include all grains occurring as nuclei as well as in inter-pisolitic areas. In thin-sections showing neo morphic aggradation, areas of microspar have been included under the category of sparry carbonate. The triangular plot (Fig. 1 2) shows fairly distinct fields for the cornstones and calcrete conglomerate, biological materials (concentric stromatolites and algal nodules) and case-hardened conglomerate. The diagram reveals the low content of detrital grains in biological structures. In contrast, samples from all other groups show a higher content of detritus. The distinction between case-hardened conglomerate and cornstone/calcrete conglomerate is based on the almost complete absence of micritic carbonate. However, the distinction between cornstone and calcrete conglomerate is not very sharp on this diagram. The distinction between the non-pedogenic calcrete conglomerate with the pedogenic nodular
(Fig. 11). The detrital grains show encrustation by calcite. The micrite is granular and less dense than in the Siwalik cornstones. Voids are lined with micro spar growing radially inwards towards the centre of the void. Medium-sized (10-15 �m) neomorphic microspar is seen to cut across the micrite areas and most commonly occurs in voids or as recrystalliza tion rims around framework grains. Fe oxide-stained micritic masses are more dominant in the Siwalik cornstones and may be related to alteration of biotite through time. This is inferred from the fact that the amount of detrital mica is considerably less in corn stones than in the contemporaneous, non-calcretized red shales.
PETROLOGICAL CLASSIFICATION
On the basis of the differences discussed so far, it seems to be worthwhile to examine the possibility of discriminating between these three continental car bonate rock types, using the relative content of detrital grains, micntic carbonate and sparry car bonate. Modal counts have been made on a wide variety of samples including. the materials under description, together with recent soil carbonate aggregations, and biological materials from various 273
·
S. K. Tandon and D. Narayan
calcrete (cornstone ) can be made on the basis of the number and nature of the laminae (Table 2 ). It would be interesting to point out that the samples of pedogenic calcrete and fluvial oolites, on which data have been generated from published photographs, show a higher content of spar compared to our samples. This may be on account of error introduced due to recognition problems while working with pub lished photographs.
I
D ISCUSSION
The calcrete conglomerate occurs in the field as irregular, lenticular units which interfinger with the channel conglomerate of the fluvial Upper Siwalik formation A. In general appearance, the calcrete conglomerate is indistinguishable from the associated fluvial deposits. This field setting, together with the fact that the size of the pisolites is much larger than most marine ooids, excludes a marine origin. Piso lites formed in continental environments are not very common and freshwater pisolites are rare. Most previous data pertain to lacustrine and spring deposits. The calcrete conglomerate has originated in the channel conglomerate facies of a fluvial system. Under such conditions, the source for the carbonate is rather problematic. Most outcrops of the calcrete conglomerate have carbonate contents in the region of 70- 80%. Obviously, the few limestone clasts in the channel conglomerate could not have supplied the large volume of carbonate required for its forma tion. Cementation is bound to a particular deposi tional cycle and thus has to be viewed in time as a process operative during that cycle. The model for origin of the calcrete conglomerate should also ex plain the carbonate supply within the system. A cue has been taken from the development of calcareous tufa in the vicinity of most present-day spring orifices in the area. The spring water is originating from thick, multistoried, carbonate-cemented, chan nel sandstones. Water discharged from these springs has a very high concentration of calcium and bicar bonate ions. The tufa that is being developed is micritic, and shows development of laminar struc tures comparable to the pisolitic laminae. It may be logical to assume the existence of a comparable pro cess during the formation of the calcrete con glomerate facies. It is, therefore, proposed that CaC03-rich water, emanating from springs, dis charged on to dry, abandoned channels with
n
][
THICK MULTISTORIED CAR BONATE CEMENTED SANDSTONES Fig. 13.
Schematic diagram to illustrate the development of the calcrete conglomerate. Stage I shows normal fluvial deposition in a gravel stream. Stage II shows the avulsion of the channel. Stage III shows the development of the calcrete conglomerate through carbonate supplied from spring waters.
gravelly beds. The quantity of carbonate in the calcrete conglomerate would require intermittent periods of enhanced flow from the springs. Pisolites grow slowly, totally immersed in the natant fluid and are formed at times of low flow when water volume is decreased by evaporation (Braithwaite, 1 979).' Percolation of CaC03-rich water would take place on to the gravelly substratum, resulting in the develop27 4
Carbonates from the Siwalik Group, India
Upper Siwalik Formation-S (alluvial fan type and proximal braided stream type deposits)
Case
Hardened
Conglomerate
Upper Siwalik Formation- A (braided
stream type deposits)
Calcrete
Middle Siwalik
Formation
Conglomerate
-B
(entrenched meander belt type deposits)
Cornstone
Fig. 14. Schematic diagram to illustrate the regional sedimentological framework and distribution of the three main
carbonate types in time and space.
water and the turbulence would immediately result in the physicochemical precipitation of calcite from water super-saturated in it ( McGannon, 1975). The patchy distribution, thickness attributes of the pro file and the lack of any variation of the carbonate content in the vertical profile are all explained by this scheme of events. The cornstone facies occurs in the upper parts of the overbank deposits. It is, in places, interbedded with the thinly-bedded levee sandstones. In the sand stones of the Upper Siwalik Formation A, the corn stone occurs as a thin, surficial layer on the buff coloured sandstones. These carbonate units are closely comparable in their outcrop and petrographic characters to the Quaternary pedogenic carbonate accumulations. Subaerial exposure of the sediments
ment of pisolitic calcrete conglomerate (Fig. 13). McGannon ( 19 75) has proposed a spring-fed river environment for the formation of primary fluvial oolites. Risacher & Eugster ( 1979) have postulated an evaporite setting for the development of Holocene pisoliths in spring-fed surface pools. The character istics of the pisoliths described by them match closely with those from the Siwalik Group. The idea that phases of evaporation did exist finds support from the work of Krynine ( 1937) who advocated a lack of deposition due to aridity at the base of the Upper Siwalik Sub-group in the Potwar region of Pakistan. On the basis of various features, including dis rupted laminae, the environment of formation was, in part, turbulent. Loss of carbon dioxide from the water on emergence from the springs, warming of 275
S. K. Tandon and D. Narayan
can be explained because of the emergence of some of the distal floodplain areas consequent upon fre quent avulsion of channels. Johnson (1977) has advanced evidence to show that some of the argillaceous horizons of comparable age in the Hari Talyangar area (lying to the east of the present area of study) are palaeosols. Wind-blown sediments have not so far been documented from the associated rock record. Therefore, carbonate supply from wind blown dust has not been considered as a major con tributing source. The absence of relict biogenic structures like rhizocretions, etc. is taken as an argument against the possibility of biogenic contri bution as a major source of carbonate supply. Most of the underlying Siwalik sandstones and other con temporaneous sediments are rich in carbonate cement; the carbonate content, in places, reaches up to 50 % of the total volume. Thus it may be reason able to assume that they supplied the carbonate re quired for the formation of the cornstones. Fig. 14 is a schematic diagram to illustrate the regional sedimentological framework and the dis tribution of the three main carbonate types in time and space.
REFERENCES BARNDT,
CONCLUSIONS
Three carbonate lithic types from the continental Siwalik Group inducting both pedogenic and non pedogenic varieties have been documented for their field and petrographic characteristics. Amongst the several field an:d petrographic characters which may be used for discriminating these contrasted facies, it is proposed that the following criteria are most useful: (a) Association of the non-pedogenic carbonate accumulations with the channel facies and associa: tion of the pedogenic carbonate accumulations with the overbank facies. (b) The higher ratio of micrite/spar in the pedo genic nodular calcrete (cornstone). (c) The poor development and smaller size of concentric bodies in the pedogenic carbonate in con trast to the well-developed, large pisolites in the non pedogenic carbonate accumulations. (d) The sharply distinct nature of the laminae in the case of the non-pedogenic carbonate accumula tions. (e) Dominantly neomorphic sparry calcite in the pedogenic carbonates.
276
J., JOHNSON, N.M., JOHNSON, G.D., OPDYKE, N.D., LINDSAY, E.H., PILBEAM, D. & TAHIRKHELI, R.A.K. (1978) The magnetic polarity stratigraphy and age of the Siwalik Group near Dhok Pathan village, Potwar Plateau, Pakistan. Earth planet. Sci. Lett. 41, 355-364. BRAITHWAITE, C.J.R. (1979) Crystal textures of recent fluvial pisolites and laminated crystalline crusts in Dyfed, South Wales, J. sedim. Petrol. 49, 181-193. BucKLAND, W. (1821) Description of the quartz rock of the Lickey Hills. Trans. Geol. Soc. London, 5, 506544. CLARK, D.N. (1979) Patterns of porosity and cement in ooid reservoirs in Dogger (Middle Jurassic) of France: discussion. Bull. Am. Ass. Petrol. Geol. 63, 676-679. DAHANAYAKE, K. (1977) Classification ofOncoids from the Upper Jurassic carbonates of the French Jura. Sedim. Geol. 18, 337-353. Do NAHU E, J. (1978) Pisolite. In: Encyclopaedia of Sedimentology (Ed. by R.W. Fairbridge and Joanne Bourgeois), pp. 582-583. Dowden, Hutchinson & Ross, Stroudsburg, Pennsylvania. Esl'EBAN, C.M. (1976) Vadose pisolites and caliche. Bull. Am. Ass. Petrol. Geol. 60, 2048-2057. GouDIE, A. (1971) A regional bibliography of calcrete. In: Food, Fibre and the Arid Lands (Ed. by W.G. McGinnies, B.J. Goldman and P. Paylore), pp. 421437. University of Arizona Press. GouDIE, A. (1972) On the definition of calcrete deposits. Z. Geomorph. 16, 464-468. GoUDIE, A. (1973) Duricrusis in Tropical and Subtropical Landscapes. Clarendon Press, Oxford. 174 pp. HARRISON, R.S. (1977) Caliche profiles, indicators of near-surface sub-aerial diagenesis. Barbados, West Indies. Bull. Can. Petrol. Geol. 25, 123-173. JAMES, N.P. (1972) Holocene and Pleistocene calcareous crust (caliche) profiles, criteria for sub-aerial exposure. J. sedim. Petrol. 42, 817-836. JOHNSON, G.D. (1977) Paleopedology of the Ramapi thecus-bearing sediments, North India. Geol. Rdsch. 66, 192-216. JOHNSON, G.D., JOHNSON, N.M., OPDYKE, N.D. & TAHIRKHELI, R.A.K. (1979) Magnetic reversal strati graphy and sedimentary tectonic history of the Upper Siwalik Group, Eastern Salt Range and Southwestern Kashmir. In: Geodynamics of Pakistan (Ed. by A. Farah and K.A. DeJong), pp. 149-165. Geological Survey of Pakistan, Quetta. KELLER, N.M., TAHIRKHELI, R.A.K., MIRZA, A.M., JOHNSON, G.D., JOHNSON, N.M. & OPDYKE, N.D. (1977) Magnetic polarity stratigraphy .of the Siwalik deposits, Pabbi Hills, Pakistan. Earth planet. Sci. Lett. 36, 187-201. KLAPPA, C.F. (1979) Comment on 'Displacive calcite: evidence from recent and ancient calcretes'. Geology, 7, 420--421. KNOX, G.J. (1977) Caliche profile formation, Saldanha Bay (South Africa). Sedimentology, 24, 657-674. KRYNINE, P.D. (1937) Petrography and genesis of the Siwalik series. Am. J. Sci. 34, 422-446.
Carbonates from the Siwalik Group, India
L.H. (1973) Calcium carbonate cementation of Alluvial Fans in Southern Nevada. Bull. geol. Soc. Am. 84, 3013-3028. McGANNON, D.E. (1975) Primary fluvial oolites. J. sedim. Petrol. 45, 719-727. REEVES, C.C. (1976) Caliche. Estacado Books, Lubbock, Texas. 253 pp. RISACHER, F. & EUGSTER, H.P. (1979) Holocene piso liths and encrustations associated with spring-fed surface pools, Pastos Grandos, Bolivia. Sedimentology, 26, 253-270. SEHGAL, J.L. & STOOPS, G. (1972) Pedogenic calcite accumulation in arid to semi-arid regions of the Indo-
Gangetic alluvial plain of erstwhile Punjab (India). Their morphology and origin. Geoderma, 8, 59--72. STEEL, R.J. (1974) Cornstone (fossil caliche)-its origin, stratigraphic and sedimentological importance in the New Red Sandstone, Western Scotland. J. Geol. 82, 351-369. TANDON, S.K. & RANGARAJ, S. (1979) Sedimentary tectonics of the Siwalik Sequence, southeast of the Ravi Structural Re-entrant. In: Structural Geology of the Himalaya (Ed. by P. S. Saklani), pp. 273-283. Today and Tomorrow, Delhi. WATTS, N.L. (1978) Displacive calcite: evidence from recent and ancient calcrete. Geology, 6, 699-703.
LATTMAN,
(Manuscript received
6
November 1 979; revision received 14 August 1 980)
277
Reprinted from Sedimentology (1987) 34 837-843
Siliciclastic grain breakage and displacement due to carbonate crystal growth: an example from the Lueders Formation (Permian) of north-central Texas, U.S.A. C HR I S B U CZYN SK I and H ENRY S. CHA F E TZ Department ofGeosciences, University ofHouston, Houston, Texas 77004, U.S.A.
ABSTRACT
The Lueders Formation (mid-Permian) in Baylor County, Texas, is an intercalated suite of fluvial siliciclastic, shallow marine siliciclastic, and shallow marine carbonate strata. There are at least two generations of carbonate cements (probably originally composed of calcite) in the fluvial sandstones where fractured grains are observed. These cements represent the initial stages of caliche formation. Cementation is envisioned as a two step process. In the first step, calcite cements form from supersaturated fluids in a freshwater, vadose environment as a meniscus cement at grain contacts. Areas of cement formation are restricted. to these sites because fluid distribution is restricted to these sites. Stresses generated by the growth of cements at grain contacts are transmitted through and concentrated at quartz/ quartz grain point-contacts until the stress is sufficient to fracture quartz grains, even though the ultimate strength of calcite is less than that of quartz, per unit area. This process occurs too rapidly to be accommodated by pressure solution. In the second phase of cementation, cement nucleation is no longer restricted by vadose conditions. In this phase, calcite growth can no longer result in quartz grain breakage; rather, the quartz grains are dispersed in poikilotopic calcite cement.
INTRODUCTION
described similar occurrences from samples of calcrete profiles of the Devonian Old Red Sandstone from Scotland. The Lueders Formation crops out at several locali ties to the east of lake Kemp in Baylor County, Texas (Fig. 1). In this area, the formation is a maximum of 36 m thick with a dip of approximately 0.5° to the west (Berman, 1969). The Lueders Formation was deposited in an environment which varied between fluvial and shallow marine during a major regression and which was punctuated by several small transgressions. As water level fell during the regression, fluvial and/or deltaic distributary channels were cut through previously deposited sediments and sedimentary rocks. Isolated caliche nodules formed in overbank material and in old channel-fill deposits. Quartz sand, intraforma tional shale clasts, and caliche clasts were deposited in the channels while clay minerals and quartz sand
The ability of a growing crystal to exert a linear force on its surroundings has been amply demonstrated by laboratory experiments (Becker & Day, 1905, 1916; Taber, 1916; Rothrock, 1925; Correns, 1949). Re cently, the same phenomenon has been observed in the geological record during the course of combined field and laboratory studies (Weyl, 1959; Dapples, 1971; Wieder&Yaalon, 1974; Reeves, 1976; Assereto & Kendall, 1977; Watts, 1977, 1978; Parnell, 1983; and others). Although most investigators have been concerned with silicates, Rothrock (1925) conducted laboratory experiments which indicate that, under certain circumstances, the force of crystallization of carbonate cements can result in the fracturing of quartz grains. An example of this phenomenon, that is, quartz grain breakage and separation by the growth of carbonate cements, occurs in the sandstones from the Lueders Formation (mid-Permian). Braithwaite (private communication) and Parnell (1983) have Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
279
C. Buczynski and H. S. Chafetz
(a)
(b)
N
Km. ,.... ,.... I 10 0 5
Kilometers ""' ""' 0
Dam
Dolomite (mapped) Dolomite (inferred)
0.5
+
Outcrop Locations
Fig. l. (A) Location of the field area in Baylor County, Texas. (B) An expanded view of the field area showing the distribution of the dolomite beds in the Lueders Formation (after Garrett, Lloyd & Laskey, 1937).
were deposited as overbank material during floods. During minor transgressions, energy conditions de creased as water depth increased and clay minerals were deposited in a shallow marine environment on top of channel-fill and the associated overbank deposits. Ultimately, carbonate production and accu mulation took place when the influx of clay minerals decreased. The cycle began again with the continua tion of the regression. This process resulted in complex cycles of interfingering terrestrial siliciclastic, marine siliciclastic, and marine carbonate strata and, in addition, the formation of caliche profiles during periods of exposure and non-deposition. Since depo sition, the carbonate strata have been pervasively dolomitized. The Lueders Formation is tectonically undeformed
in the field area. Based on stratigraphic relationships and the thickness of younger strata, it is estimated that the Lueders Formation was never buried more deeply than 440-600 m. There is no interpenetration of quartz or carbonate grains in the sandstones or in the carbonates. Soft clay clasts in the sandstones show the effects of only mild compaction. The sandstones are cemented primarily by clay minerals, and there fore, deep burial would have forced quartz grains together and squeezed the soft grains into the pore spaces of the rock. This was not observed; the clasts are still quite distinct and discrete. There is no evidence of stylolitization in the carbonates which, if present, should be readily apparent due to the high percentage of insoluble clay minerals. In addition, there is excellent preservation of moulds in the 280
Siliciclastic grain breakage due to carbonate crystal growth
carbonates. All of this suggests that the strata were never deeply buried, nor subjected to significant lithostatic stresses.
THE SANDSTONES
Fluvial and overbank sandstones in the Lueders Formation consist primarily of very fine sand-sized, subangular to subrounded, monocrystalline, inclusion free quartz grains. Other constituents include detrital clay minerals, feldspars, polycrystalline quartz, calcite and dolomite clasts, and heavy minerals. Quartz grains account for 25-75% of the rock. Detrital and authigenic clay minerals comprise from a few per cent to over 60% of a given sandstone. Almost all of the authigenic clay comes from weathered feldspars. Calcite and poikilotopic barite have cemented small areas of the sandstones. Calcite cementation is not pervasive anywhere in the sandstones; cements usually occur in small patches of impure, poikilotopic crystals ranging from I mm to 5 mm in diameter. In some areas, these patches coalesced into larger patches. In some instances, cementation has followed the struc ture of burrows and/or roots. These cements are associated with a reducing environment and cemented patches, as well as the areas immediately surrounding them, are often grey whereas the remainder of the rock is oxidized. At least two periods of calcite cementation occurred in the sandstones and each consisted of two separate steps. Some of the cementation took place either prior to, and/or contemporaneously with, barite formation and prior to dolomitization. The next generation of calcite cementation began after barite formation and after the previously formed cements had been dolom itized. These second-generation calcite cements are not dolomitized. The siliciclastic grains in the carbonate-cemented areas display a 'floating' texture (Fig. 2). Point count data indicate that siliciclastic grains comprise 68-74% of the carbonate-free areas (the rest being composed of a clay matrix) whereas siliciclastic grains comprise only 26-43% of the carbonate-cemented areas with carbonate cement comprising the remainder of these areas. The decrease in abundance of siliciclastic grains is due to both displacement and replacement of these grains, with displacement being the dominant process. Significantly, fractured quartz grains have been observed only in carbonate-cemented areas. Many fractured grains have undergone separation (Fig. 3). The 'jigsaw-puzzle fit' of some of these pieces
Fig. 2. The left half of this pair of photomicrographs (A= plane light, B =crossed polars) is calcite-cemented and the right half is 'normal', non-calcite-cemented sandstone. The contact between the two areas runs essentially vertically through the centre of the field of view (arrows). The expanded, open fabric of the framework grains is readily apparent in the calcite-cemented area when compared with the non-calcite-cemented area. The light area in (B) is poikilotopic calcite cement.
leaves no doubt that they are now separated pieces of what were previously single, monocrystalline grains (Fig. 4A). These pieces must have undergone separa tion after deposition. One possible interpretation is that some of these 'separated' pieces are separated only by a vein of replacement calcite running through a quartz grain and that no fracturing or displacement have taken place. However, these pieces sometimes have slightly different crystallographic orientations of their axes (Fig. 4B). This difference in orientation means that they must have undergone some rotation 281
C. Buczynski and H. S. Chafetz
Fig. 3. Thin section photomicrograph taken under crossed polars showing several fractured quartz grains (arrows) 'floating' in a calcite matrix.
and the calcite separating the individual pieces can not be a simple replacement vein. In addition, broken edges of the grains do not display a corroded appearance as do many of the outer edges of partially replaced siliciclastic grains (Fig. 5). The logical and reasonable conclusion, therefore, is that these pieces have been broken and displaced after corrosion, and that corrosion followed deposition. Fig. 4. Photomicrograph of a broken quartz grain in calcite cement displaying an essentially perfect 'jigsaw-puzzle fit' of the two pieces. The pieces have undergone some rotation relative to one another and lateral movement away from each other. The rotation is indicated by the difference in crystallographic orientation displayed by the two pieces when viewed through crossed polars (B). (A) and (B) show the same field of view through plane light and crossed polars, respectively.
PROPOSED PROCESS FOR GRAIN BREAKAGE
Becker & Day (1916), Taber (1916), Rothrock (1925), and Watts (1978), among others, have all noted that growing crystals are able to exert a force on their surroundings. Before the stresses created by the displacive growth of a calcite cement can fracture quartz grains, two conditions must be met. Firstly, crystals must be in contact with supersaturated solutions, otherwise they would not grow (Taber, 1916). Secondly, the stress created by the growing crystals must be transmitted through, and concen trated at, quartz grain boundaries, just as an axe splits wood by concentrating the force along the sharp edge of the blade. This second condition is necessary because the ultimate strength of calcite (140 bars at 24°C with no confining pressure) is about two orders of magnitude less than that of quartz [25 kbars at 24°C with no confining pressure (Handin, 1966)]. Riecke (1895, as quoted by Correns, 1949, and Robin, 1978) observed that if crystals were to grow
displacively while submersed in a supersaturated solution, an additional condition had to be met: there must not be any unstressed crystals present. Un stressed crystals would grow at the expense of stressed crystals and no displacive growth would take place. We maintain that this condition may not apply if displacive growth occurs in the vadose zone because calcite crystals are precipitating from a supersaturated solution that is not constantly being replenished, rather they are forming rapidly as the last bit of fluid at grain -contacts completely evaporates; a non equilibrium condition. There are two lines of evidence which clearly suggest that the growth of carbonate cements in the 282
Siliciclastic grain breakage due to carbonate crystal growth
In the first step of the fracturing/dispersion process, carbonate-enriched waters percolate down through the sediment in the vadose zone after every rainfall. As evaporation takes place and the sediments begin to dry, the remaining water becomes more nearly saturated with respect to calcite. Eventually there will be a small amount of supersaturated fluid present at grain boundaries and in grain fractures (Fig. 6A). As the last of the water evaporates, calcite crystals form a meniscus cement. This meniscus cement can do one of two things depending on the stresses holding the grains together. If the stress between the two grains is greater than that of the force of crystallization of calcite, calcite crystals cannot force the grains apart; rather they must form adjacent to the point of grain contact, and no displacement of the grains will occur. If the stress between the grains is less than the force of crystallization of calcite, then the grains will be forced apart by the growing crystals (Fig. 6B). The transmission of stresses from the greater surface area of the growing carbonate to progressively smaller surface areas at grain-to-grain point contacts will eventually concentrate enough stress to fracture quartz grains (Fig. 6C). Assuming, for a given mineral, that the force of crystallization and the force required to fracture a crystal of that mineral are of the same order of magnitude (Becker & Day, 1905), calcite would have to crystallize over an area of about 20 mm2 to cause pure, monocrystalline, inclusion-free quartz grains in point contact over 0.1 mm2 to fracture. As most quartz grains have some flaws or weaknesses, (e.g. partially annealed occult fractures) the actual forces required to fracture them may be much smaller. Crystallization from supersaturated fluids occurs too rapidly to be accommodated by pressure solution and the interpenetration of quartz grains; consequently, this 'rapid' build-up in stress is released by grain fracturing. Once grains have fractured, subsequent influxes of water will lead to the growth of carbonate cement in the newly created fractures. This enlarges the fractures and fills them with calcium carbonate. After each pulse of water dries, the individual fragments of each broken grain will be held together as a cohesive unit by the cement. The second step of the fracturing/dispersion process takes place when the sediments are in the phreatic zone. When sediments are below the water table, the pore space is filled with fluid and cements could form anywhere within the void spaces. Their growth would not be restricted to a meniscus cement at grain contacts and within grain fractures. At this stage, new
Fig. 5. Many grains show evidence of having been partially replaced. Thin section photomicrograph, showing crenu lated, irregular boundaries are characteristic of replaced grains (arrows). Under crossed-polars.
Lueders Formation sandstones has exerted sufficient force to: (I) fracture quartz grains; and, (2) disperse quartz grains in carbonate cement. Firstly, as stated above, fractured grains have been observed exclu sively in carbonate-cemented areas. The fracture patterns closely resemble experimental fracture pat terns created at grain contacts (Gallagher et a!., 1974, figs 18 and 28) and the borders of the fractures do not display corrosive textures characteristic of the par tially replaced grains. These findings also support our interpretation that the carbonate separating pieces of what were formerly single grains is not a replacement of a portion of the siliciclastic grains, but a fracture fill deposit. Some of the fractures may have been present as occult fractures prior to cementation and the cements may have broken the grains along these pre-existing planes of weakness. However, unbroken, fractured grains have never been observed outside of cemented areas whereas occasional to abundant broken grains are present in cemented areas (Fig. 3). This means that grains must have been fractured and broken after deposition, by some force that did not act uniformly on the sediments and was in some way related to carbonate cementation. Secondly, there are fewer grains per unit area in the cemented patches as substantiated by the point count data. It is clear that growth of cement has dispersed quartz grains in carbonate-cemented areas. The fracturing and ultimate dispersion of quartz grains in a carbonate cement is envisioned as a two step process similar to that described by Braithwaite (private communication). In the first step, the growth of carbonate cement at grain contacts in the vadose environment fractures the grains. In the second step, grains are displaced by further cement growth. 283
C. Buczynski and H. S. Chafetz
nuclei form and the cements begin growing outward from both the old and new nuclei. This displaces grains at the periphery of the growing carbonate away from the centre of nucleation. This outward growth also disperses grains that were trapped in the growing cement.
SUMMARY AND CONCLUSIONS
Sandstones in the Lueders Formation were cemented by poikilotopic barite, and carbonates (probably calcite). There were two generations of calcite cemen tation and each took place in two steps. In the first step, sediments in the freshwater vadose zone were periodically wetted. Calcite cements formed at grain contacts and in intragranular fractures as the sedi ments dried and the pore fluids became supersaturated with respect to calcite. The stresses created by calcite growth were transmitted through quartz grain contacts and the stresses increased too rapidly to have been accommodated by pressure solution. Ultimately, this rapid increase in stress at quartz/quartz point contacts resulted in the fracturing of quartz grains. Cements also precipitated in these newly formed fractures and expanded them. The second phase of cementation began when the sediments were in the phreatic zone. In a phreatic environment, cement nucleation and growth would not be limited to grain contacts and intragranular fractures and, therefore, quartz grains could not be fractured. The second step of the process is responsible for the dispersal of grains in calcite cement, and eventually, the complete cementation of the sandstone by calcite. In the Lueders Formation the second phase of cementation is only in the inchoate stages of development. Fig. 6. A two-dimensional schematic representation of a hypothetical process in which the force of crystallization due to calcite precipitation over a relatively large surface area is concentrated at a point, resulting in quartz grain breakage. Grains not involved in the transmission of stresses due to cement growth have been omitted from (B) and (C) for simplicity. (A) Due to evaporation in the vadose zone, waters supersaturated with respect to calcite concentrate at grain contacts (stippled areas). (B) Calcite crystallizes (black areas) and begins to push grains away from each other; however; the stress at some grain contacts is too high to permit calcite growth to push grains apart. (C) The stress from the growing calcite between many grains is transmitted through grain contacts and is additive, until a grain fractures and reduces the stress.
ACKNOWLEDGMENTS
This research was conducted as part of an MS thesis project at the University of Houston. We would like to acknowledge the generous financial support of the University of Houston Geology Foundation, the American Association of Petroleum Geologists, and Conoco. In addition, we would like to thank Dr William Carlson and Dr James Boles for their careful reviews of earlier versions of this manuscript.
284
Siliciclastic grain breakage due to carbonate crystal growth
REFERENCES AssERETO,R.L.A.M. &KENDALL, C. G. Sr. C. (1977)Nature,
origin and classification of peritidal tepee structures and related breccias. Sedimentology, 24, 153-210. BECKER, G.F. & D AY, A.L. (1905) The linear force of growing crystals. Proc. Washington Acad. Sci., 7, 283-288. BECKER, G.F. & D AY, A.L. (1916) Note on the linear force of growing crystals. J. Geo/., 24, 313-333. BERMAN, D.S. (1969) Vertebrate fossils from the Lueders Formation, Lower Permian of North-Central Texas. PhD Thesis, University of California, Los Angeles, California, 160 pp. CORRENS, C.W. (1949) Growth and dissolution of crystals under linear pressure. Discuss. Faraday Soc., 5, 267-271. DAPPLES, E.C. (1971) Physical classification of carbonate cement in quartzose sandstones. J. sedim. Petrol., 41, 196-204. GALLAGHER, J.J. JR., FRIEDMAN, M., HANDIN, J. & SOWERS,
G.M. (1974) Experimental studies relating to microfrac tures in sandstone. Tectonophys., 21, 203-247. GARRETT, M.M., LLOYD, A.M. & LASKEY, G.E. (1937) Geologic Map of Baylor County, Texas, (Revised). Univer sity of Texas at Austin, Bureau of Economic Geology. HANDIN, J. ( l 966) Strength and ductility. In: Handbook of Physical Constants (Ed. by S. P. Clark), pp. 223-289. Mem. geo/. Soc. Am., 97. PARNELL, J. (1983) Ancient duricrusts and related rocks in .
perspective: a contribution from the Old Red Sandstone. In: Residual Deposits: Surface Related Weathering Process and Materials (Ed. by R. C. L. Wilson), pp. 197-209. Blackwell Scientific Publications, Oxford, England. REEVES, C.C. JR. (1976) Caliche - Origin, Classification, Morphology and Uses. Estacada Books, Lubbock, Texas, 233 pp. RIECKE, E. (1895) Ueber das Gleichgewicht zwischen einem festen, homogen deformirten Korper und einer fliissigen Phase, insbesondere iiber die Depression des Schmelzpunktes durch einseitige Spannung. Ann. Physik., 54, 731-738. RoBIN, P.-Y. F. (1978) Pressure-solution at grain-to-grain contacts. Geochim. Cosmochim. Acta, 42, 1383-1389. ROTHROCK, E.P. (1925) On the force of crystallizing calcite. J. Geol., 33, 80-83. TABER, S. (1916) The growth of crystals under external pressure. Am. J. Sci., 41, 532-556. WATTS, N.L. (1977) Pseudoanticlines and other structures in some calcretes of Botswana and South Africa. Earth Surf Processes, 2, 63-74. WArrs, N. L. (1978) Displacive calcite: evidence from recent and ancient calcretes. Geology, 6, 699-703. WEYL, P.K. (1959) Pressure solution and the force of crystallization - a phenomenological theory. J. geophys. Res., 64, 2001-2025. WIEDER, M. & YAALON, D.H. (1974) Effect of matrix composition on carbonate nodule crystallization. Ceo derma, 11, 95-121.
(Manuscript received 10 June 1986; revision received 5 February 1987)
285
Reprinted from Sedimentology ( 1989) 36 1113-1126
Near-surface shrinkage and carbonate replacement processes, Arran Cornstone Formation, Scotland
S . K. T ANDON* and PETER F . FRIENDt *Department of Geology, University ofDelhi, Delhi 110007, India; fDepartment of Earth Sciences, University of Cambridge; Downing Street, Cambridge CB2 3EQ, UK
ABSTRACT
A 47 m thick succession of conglomerates, sandstones and mudstones, of Late Devonian or Early Carboniferous age, outcrops at Fallen Rocks in northeast Arran (western Scotland). It is defined here as the type section of the Arran Cornstone Formation. At numerous levels in the succession, varieties of fissures and carbonate concretions formed during the accumulation of the Formation. The fissures opened as a result of drying-shrinkage, and were closed again either by filling with different sediment, or by wetting and expansion of the fissure wall sediment. Carbonate concretions form complete beds, discontinuous, bedding-concordant sheets, or bedding-discordant nodules or rods (the rod cornstones). These concretions formed close enough to the surface to be incorporated, after erosion and redeposition, as clasts into overlying beds. The concretions were formed by growth of micrite, mainly by replacement, but shrinkage displacement played an important role in subsequently fracturing and reworking the micrite. The micrite was also locally replaced by microspar and spar, and this involved dissolution and precipitation. No independent evidence of biological influence in any of these processes has been found.
INTRODUCTION
The Late Devonian orEarly Carboniferous Cornstone Formation of Arran overlies the non-marine Upper Old Red Sandstone, and contains a remarkable variety of carbonate concretions. There are many other Scottish occurrences of concretionary carbonates ('cornstone') at a similar stratigraphic level (Burgess, 1961; Chisholm & Dean, 1974; Leeder, 1976; Read & Johnson, 1967; Parnell, 1983a,b) and most authors have implied a non-marine and pedogenic origin for the concretions. In the southern British Isles, concre tionary carbonates, probably pedogenic, are also abundant in both the Lower and Upper Old Red Sandstones (Allen, 1965, 1974; Gardner & Horne, 1981; MacCarthy, Gardner & Horne, 1978; Pick, 1964; Tunbridge, 1981; Watts, 1978). Much of the work on these carbonates has focussed on understand ing the relationships of the carbonates with the fluvial facies in which they occur. In contrast, we have analysed the morphology and petrography of the Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
Arran Cornstone carbonates, with the prime object of discovering the processes that have acted locally within the sediment. We have been particularly concerned to question whether the carbonate was formed during (1) pedogenesis, (2) a phase of near-, but sub-surface groundwater activity or (3) later burial or re-emergence diagenesis. Work on the textures and fabrics of rather similar concretionary carbonate rocks from elsewhere has provided important new understanding of these local processes (Freytet & Plaziat, 1982; Klappa, 1978, 1980a,b; Nagtegaal, 1969; Peryt, 1983; Steel, 1974; Watts, 1978, 1980). All of these studies show that there has been growth, in the host rock, of carbonate of greater volume than that required for the filling of the original pores. Most such rocks also contain evidence of the repeated brecciation, dissolution and reprecipitation of the carbonate. The Arran carbonate concretions occur in sand287
(a)
(
utton's Unc
'
o�f�r
S. K. Tandon and P. F. Friend
(b)
mlty
.
t
0
·.::·.::.
Grld
km
North
Dalradian metasediments (early Palaeozoic)
co
Northern Granile
Fallen Rocks
(early Tertiary) through wood
Carboniferous and Permian Arran Cornstone Formation
North Sannox Old Red Sandstone with lavas (v)
55' 38' N
5'16'W
Fig. l. (a) outline geological map of north-eastern Arran, based on I: 50 000 Geological Survey map of Arran, 1972. Fallen Rocks locality and three other Arran Cornstone outcrop areas are shown. (b) access map for Fallen Rocks locality, NE Arran.
stones and mudstones in which fissures of various sizes and forms are common. The involvement of these fissures in the carbonate growth is an important process that we shall discuss below. The methods we have used range from study of the form and occurrence of the concretions and fissures in the outcrop, to microscopic work. Thin-section petrog raphy has been based on both stained and unstained material, particularly using a Wild M400 Leitz microscope with a zoom facility. Cathodoluminesc ence work has been particularly profitable, and we used a chamber with a cold cathode operating at 30 kV and 800/J.!A. Some clay-mineral identification has been carried out using normal X-ray diffraction techniques on acid-insoluble residues from the mud stones, and preliminary chemical work was carried out on calcite types using an electron microprobe. All analytical work was carried out in the Department of Earth Science, University of Cambridge.
STRATIGRAPHY
Definition and age of the Arran Cornstone Formation
As a type exposure for the Formation, we specify a coastal locality in northeast Arran, just north of the 'Fallen Rocks' (Fig. 1). The 47 m section measured by us along the coast at this locality is presented in Fig. 2. The lower stratigraphical contact with coarse polymict conglomerates typical of the Upper Old Red Sandstone (Friend, Harland & Hudson, 1963) occurs at Grid Ref. NS 003483. The upper, faulted, contact with sediments that are grey-green in contrast to the reds, greys and browns of the Cornstone Formation occurs at Grid Ref. NS 002485. The succession dips northwards at about 30°. Two other coastal localities (Fig. 1) at which the Formation is exposed, with rather different features, are at Hutton's Unconformity (NR 935520), and 288
Near-surface shrinkage and carbonate replacement
wide, 50-I 00 mm high) between the fills of lighter coloured-sand (Fig. 3a). Some sediment contains zones of diffuse 'vertical fabric' (Figs 2 & 3) where large numbers of anastomosing fissures have formed and closed but not been filled by later sediment of distinctive types. This intrastratal reworking by the opening and closing of fissures appears to have been widespread, due not only to the opening and closing of one set of fissures, but to the superimposition of later fissure systems.
northern Corrie (NS 023442). However, we have not made detailed studies at these localities. No fossils have been found in the Formation, so its age is uncertain. About 100 km to the east, the Upper Old Red Sandstone contains vertebrate fossil evidence of Late Devonian (Fammenian) deposition, but it is quite possible that the Arran Upper Old Red was a 'Lower Carboniferous (Old Red Facies)' deposit (Westoll, 1977). Above the Cornstone Formation, the next securely correlated horizon is the Corrie Lime stone, well within the succession mapped as Carboni ferous in Arran, and dated as latest Early Carboniferous (basal P2' latest Dinantian) (George et al., 1976). All we can conclude, therefore, is that the Arran Cornstone Formation is either Late Devonian or Early Carboniferous in age.
Fissure-fill 'diamictites'
Some of the sandy mudstones high in the section have a diamictite-like texture, where small pebbles and granules of vein quartz 'float' in a finer matrix (Fig. 4). Zones with this texture pass, by transition, vertically and laterally, into zones where fine pebbles and granules form the fill of fissures with distinct margins. In some areas these clearly-defined fissure fills pass upwards into an irregular-based, but sharply defined bed of coarse sediment similar to the fissure fill. In other cases, it is clear that smaller amounts of coarse sediment have been introduced during fissure formation and followed by a succession of episodes of wetting and drying or fissure wall collapse, resulting in the thorough mixing of fine matrix and coarser fill to produce a diamictite-like texture (Fig. 4).
General description of measured section
The Arran Cornstone Formation consists of a variety of distinct types and colours of sedimentary rock: creamy coarse and medium-grained sandstones, red muddy fine to very fine-grained sandstones, red mudstones, creamy conglomerates, pebbly sand stones, grey nodular and bedded limestones. We consider the Formation to be non-marine because of the total absence of marine fossils, and the similar interpretation for the stratigraphic units below and above. Carbonate concretions occur in all the above rock types and their different form and situations will be described below.
MINERALOGY OF HOST SEDIMENTS
The sandstones consist largely of quartz grains, many of them polycrystalline. Minor amounts of plagioclase and rock fragments are generally present. Haematite pigment is abundant in the red sandstones. Grains of microcrystalline calcite, apparently locally derived detritus, are present. The grain-to-grain contacts in some sandstone beds indicate a degree of framework compaction. The only clay-minerals identified by routine X-ray diffraction of the fine-grained fractions of the mud stones are illite and chlorite.
Fissures
Fissures and polygonal cracks of varying sizes occur throughout the measured section (Fig. 2). In many beds, fissure systems of different sizes occur together, and suggest many episodes of wetting and shrinkage. The largest polygonal cracks (Fig. 3) occur towards the top of the section. They are up to 30 mm wide, and form irregular polygons up to 0· 3 m in width, representing a linear shrinkage of about I 0%. Within a bed the fissures have sometimes opened up and coalesced to produce linking systems running nearly sub-parallel to the general bedding. Examples of both incomplete and complete polygons are common. The material filling the fissures varies in grade from granule-size to very fine sand. Carbonate concretions occur between the fissures, sometimes as small nodules at the margins of granule-grade fills, sometimes as bedding perpendicular, elongate nodules (30-50 mm
DISTINCTIVE CARBONATE PRECIPITATION SITES IN THE SEQUENCE Sandstone cements
Some sandstone beds contain distinct and irregular layers with carbonate cement. The cement is sparry, 289
S. K. Tandon and P. F. Friend
32
�
,,J,
15
.
� .
fault
47
27
"--
�
39
\:;
(f) w 0:: 1w 2
�!
\:: _ ,,, ·�
�
• • ••
• • .. . •
25 � "'�
4 .
��
••
'
�
�20 ��Q
28
• -. . ..
.
D
0
'
. 0
37
55
9
;,.. ,
,
. .
'
�, �.
44
:��
HY .. v .. vv v·••
�
•
YyY
I
f ,, 0
•'•
ORS
.!.
�
17
24
Fig. 2. Sedimentary log measured through Arran Cornstone Formation at Fallen Rocks locality. Thicknesses in metres above the basal contact with Old Red Sandstone conglomerates are indicated to the left of the log, and the numbers assigned to the beds are indicated (odd numbers only) to the right. Carbonate concretions are shown in black, with some indication of their morphology. Fissures (shown by open fissure symbol), 'vertical fabric' (VF), cross-bedding, pebbles, diamictites (D) are indicated. Cementation of sandstones is shown by stipple.
290
Near-surface shrinkage and carbonate replacement
Fig. 3. Fissuring in Arran Cornstone Formation. (a) Sandstone-filled fissures extend down from a sandstone sheet, carbonate nodules occur below and between some of these filled fissures (bed 10). Pen is 0· 15 m long. (b) 'Vertical fabric' interpreted as formed by repeated cycles of fissuring and fissure closing (bed 6). Elongate carbonate concretions tend to be subparallel to the 'vertical fabric'. Pen is 0·15 m long. (c) Polygonal fissure system (bed 54) filled with granule-grade conglomerate. Hammer is 0·3 m long. (d) Bed 54 (below) showing well-defined, granule conglomerate filled fissures, is overlain by bed 55 which has a well-bedded lower part (apparent dip 30° to right) but mainly consists of homogeneous diamictite of granules and small pebbles 'floating' in a mudstone matrix. The diamictite was formed by repeated fissuring, and fissure closure, with introduction of granules. Hammer is 0·3 m long. �
and appears to have grown in the interparticle pores. This carbonate could have formed at any stage after deposition of the host sand.
fragments a few em across and composed of a patchy carbonate. Several generations of cracks with distinct margins are present, filled by coarse calcite. In addition to the patchy nature of the carbonate, other textures visible are floating quartz grains, crystallaria, gravi tational cements, and occasional pisoliths of micrq crystalline calcite. Large fenestrae are also common, containing bladed crystals of calcite, apparently formed as phreatic cement filling voids. These features make this bed strikingly comparable to modern 'hardpan' calcretes forming in the inter fluves of semi-arid fluvial systems (Reeves, 1976; Watts, 1980).
Bedding-concordant carbonates Brecciated whitish-grey limestone
One bed of compact, indurated white to whitish grey, intensely brecciated limestone stands out prominently in the measured section (Fig. 2, No. 16). The bed varies in thickness up to about 0· 3 m, and pinches out altogether over a distance of tens of metres. Its upper and lower contacts, with medium-grained sandstone, are sharp though irregular. Patchily distributed, vertically elongate areas of reddish sandstone occur as relicts within the limestone. Polished slabs of this limestone reveal individual
Discontinous grey massive carbonate
This carbonate type consists of grey massive and indurated carbonate in irregular, discontinuous sheets, 291
S. K. Tandon and P. F. Friend
fragments of laminar carbonate. They prove that concretionary carbonates similar to those visible in the present outcrop were available for transport and deposition throughout the sedimentation of the Corn stone Formation.
(a)
fissure opens by drying
Bedding-discordant carbonates Rod cornstone
(b)
complete
closed by
fill !exotic)
wetting
(c)
repeated opening and closing:
"vertical fabric··
Fig. 4. Diagram summarizing (a) fissure opening; (b) different processes resulting in closure of fissure, i.e. (1) complete filling by exotic sediment. (2) filling by a mixture of exotic sediment and collapse of fissure walls, (3) filling by expansion of walls during wetting of sediment; (c) repeated opening and closing may generate a diamictite, where exotic clasts become thoroughly mixed with fissure-wall mud, or 'vertical fabric' where old fissure traces pervade the sediment thoroughly.
that extend for no more than 3 m parallel to bedding in the medium to coarse sandstones and fine conglom erates that are the normal host rock. In some beds, carbonates of this form have grown parallel to, and within, cross-stratification foresets, forming sheets 0· 3-0· 5 m thick. Patchy carbonate groundmass tex tures, and the presence of floating quartz grains with irregular corroded margins, are evidence of in situ modification and growth of the carbonate.
This very distinctive concretionary carbonate type consists of vertical to sub-vertical rods in either red fine-to-medium grained sandstone or red mudstone (Fig. 5). The sizes of the rods vary up to maxima of 0·07 m in width and 0·5-0·6 m in length. The percentage of any bed occupied by concretionary carbonate varies from 20-30% in the lower third to 70-80% towards the top of the bed. Vertically elongate carbonate nodules also occur between fully developed rods. Where rods are large and occupy a high percentage of the bed, neighbouring rods may be in contact with each other, producing coalescence of the individual rods. Vein-like carbonate stringers locally transverse the detrital sediment between the rods. In some cases the rods are clustered into groups that amalgamate upwards, and splay out, like fingers, downwards, Many rods tend to taper at their lower terminations. Their upper ends often terminate below a zone, about 50 mm thick, of local erosion and deposition of clasts of concretionary carbonate and detrital matrix. Thin sections show a diffusely patchy, micrite texture with interspersed patches of microspar. Float ing grains of quartz are common in the rod, and occasionally show fracturing by displacive growth of calcite. The calcite of the rod margins tends to be coarser (microspar) than that of the interior, although spar-filled fenestrae are common in the micrites of the rod centres. Although vertically elongate structures are com monly reported from modern and ancient calcretes (Reeves, 1976; Goudie, 1983), such well-developed rods as these have rarely been described, to the best of our knowledge. Below we shall discuss the possible influence of fissure growth on their formation.
Carbonate clast conglomerate
Nodular concretions
Thin (up to 0·1 m), discontinuous, conglomerate units contain carbonate clasts in addition to clasts of vein quartz, feldspar and quartzite, cemented by later calcite. The carbonate clasts consist of microcrystal line calcite, pisoliths with concentric layering, and
More-or-less equidimensional carbonate nodules are an abundant feature of the red mudstones, particularly towards the top of the measured section (Fig. 2). Individual nodules are several em across and may occur eithe r as isolated or coalescing forms. 292
Near-surface shrinkage and carbonate replacement
Fig. 5. (a) Rod cornstone of bed I (Fig. 2) with apparent dip �!5° to right. Length of visible scale is 40 em. (b) Bedding surface in bed 13. A few carbonate concretions have grown to form polygonal sheets, each one closely fitting the fissure polygons in which they have replaced the original sandstone. Length of pen is 0·15 m.
Polished surfaces cut through the nodules show them to have been intensely fractured, with fragments from a few mm to a few em across. Local reduction (discolouration) of the red host sediments, and red pigmentation of parts of the nodules are common. Sand-filled fissures are also often associated with the nodular beds. Various patchy, floating-grain and brecciation textures are well developed in the nodules.
Folk, 1962). This micrite is characterized in thin section by a diffuse patchiness due partly to variations in crystal size, but mainly to differences in darkness as seen with transmitted light (Fig. 6a & b). This patchy texture has elsewhere led to the term 'clotted micrite' (James, 1972), and probably reflects varia tions in the amount of fine-grained clay and iron oxide present. Most of the micrites luminesce a dull orange, but they vary slightly in intensity from patch to patch. Rounded micrite glaebules or pisoliths, up to 0. 1 mm in diameter, and usually darker in transmitted light, are locally present. They are frequently traversed by calcite spar-filled cracks, and occasionally show concentric structure (Fig. 6a). 'Floating' quartz grains are also normally present (Fig. 6a & b), usually well separated from each other by the general micrite groundmass. The detailed features of their margins will be discussed below. Coarse-grained carbonate occurs both in bodies with sharply-defined margins (fenestrae or crack fills), or with more diffuse margins (patches). The material may be microspar or spar (crystals up to about 500 �J.m). Compared with the micrite crystals, spar sized crystals are large enough to commonly display patterns of luminescence, demonstrating their local growth patterns.
Carbonate concretions within polygons betweenfissures
In some beds, vertically elongate and irregular carbonate concretions, a few to several em across, occur in the polygonal prisms of sandstone defined by the fissures already described (Fig. 5b). They are usually located in the centres of the host rock prisms, at places extend as far as the margins of the fissures, but have never been seen to extend into the fissures. The concretions consist of micro-crystalline calcite, with abundant floating grains of quartz, and fenestral spar fills. MICROMORPHOLOGICAL FEATURES AND THEIR CATHODOLUMINESCENCE General features
Spar-micrite growth patterns
The general groundmass of the material making up the concretions is micrite (crystals less than 4 11m;
Some spar bodies with sharp margins appear to be the result of the passive fill of cracks (Fig. 6b) that were 293
S. K. Tandon and P. F. Friend
Fig. 6. Both thin sections are from bed 16. (a) Typical clotted micrite with glaebules(one with concentric structure) and floating quartz grains. Calcite veins have cut the glaebules and surrounding sediment at different stages. (b) Typical clotted micrite showing vein fills, fenestrae and patches of spar and microspar. Length of both scale bars is 0·1 mm.
Fig. 7. Cathodoluminescence photograph of sample from bed 12. The central part and most of the upper part of the area are formed of spar, showing peripheral growth of the crystals. Micrite below and above shows replacement of the micrite fabric by spar-sized crystals, presumably by solution and precipitation about certain centres in the micrite. Scale bar is 0·2 mm long.
empty (of solids) at the time of spar growth. The evidence for this is the abrupt cross-cutting relation ships of the crack surfaces, and the initial lining of these surfaces with microspar before the main spar fill occurred.
Figure 7 shows part of a spar-filled crack traversing micrite. Within the micrite individual small crystals luminesce faintly, but are surrounded by a network of bright luminescing material that gives an overall effect of brightness to the micrite. The network of bright 294
Near-swface shrinkage and carbonate replacement
a late stage precipitation of bright luminescing material on crystal surfaces that often show evidence of solution before precipitation. Two areas with micrite are present. In the largest, several sharp-edged cracks filled with microspar appear to be empty crack fill features, and these are cross-cut by the surrounding spar a·n.d appear therefore to be earlier than it. The other tp.icrite area shows two separate wisps of micrite that appear to be relicts of a larger area now replaced by spar. The form of the luminescence in this area shows. the peripheral growth of the spar, and also evidence of the solution of its surface before the latest, bright luminescing material formed. Crack and fill sequences are also present and embayments suggest solution of the micrite surface along with precipitation of the late (bright luminescing) spar. In summary, the micrite appears to have formed by calcite growth about numerous, closely spaced nuclei, followed by further precipitation along the partly open, and possibly moving crystal interfaces. Discrete open spaces and cracks in this material became the sites for open space spar-fill. In other cases, spar-sized crystals grew by solution and replacement of the micrite. This may have been helped by cycles of intercrystal movement and cracking.
material in places passes into what is clearly a network of cracks. Within the main spar area, numerous crystals show evidence of concentric growth in all directions within the plane of the section. The important observation is that there is a transitional zone, of variable width, where spar-sized calcite crystals with concentric growth stratigraphy also show a division into fine-grained (micrite) component crystals. We interpret this effect to indicate progressive replacement by the spar crystals of an earlier micrite texture (Fig. 8). This apparently involved (I) prefer ential and local solution around the margins of the component crystals of the micrite mosaic and (2) growth, stage by stage, of the new spar crystal stratigraphy. Intercrystal cracking may have been an important process in opening up the micrite mosaic for solution and replacement. Figure 9(a & b) shows a predominance of spar, with evidence of peripheral growth of the spar crystals, and
Replacement of floating quartz grains and clay
(a)
(b)
aggregates by carbonate
Fig. 8. Diagram illustrating first stages in replacement of a micrite by a spar texture. (a) Initial mosaic of micrite crystals with intercrystal boundary areas open to pore-water solution and reprecipitation of calcite. (b) Selective replacement of patch in relicts. (c) Patch now replaced by spar, and extension outwards of replacement zone with potentially distinctive cathodoluminescence.
Isolated quartz grains are a feature of both micrite and spar areas in the concretions (Fig. I 0 a-d). In many cases in a micrite area, the quartz grain has a zone of spar surrounding it (Figs lOa, b & I I a, b). In some cases this zone has so distinct an outer margin, that we considered the possibility that the zone had
Fig. 9. Plane-polarized light (a), and cathodoluminescence (b) views of a thin section from bed 2. The views show largely spar, although with luminescing margins, and some inter-spar micrite. Relict fragments of micrite are visible at the bottom right, and bottom left of centre. The latter shows evidence of replacive growth of the spar at the expense of the micrite. Scale bar 0·2 mm.
295
S. K. Tandon and P. F. Friend
Fig. 10. (a) Floating quartz grains in micrite are surrounded by rims of spar. Scale bar is 1·0 mm, bed 23. (b) Floating quartz grains in micrite surrounded by spar zones,which appear to have formed by solution and replacement of the quartz. Scale bar is 0·2 mm. (c) Transmitted light, (d) cathodoluminescence. Growth patterns in (d) show that the spar has grown by replacement of the floating grains, particularly on some faces. Scale bars are 0·2 mm.
b) shows a progressive change towards the surface of the quartz grain from a micrite with an overall luminescence intensity, by transition towards a mi crospar, and then a zone of spar, next to the quartz, that again shows evidence of growth inwards towards the quartz grain. One surface of the quartz grain is very irregular, and we suggest that this irregularity indicates that the replacement of the quartz proceeded to a lesser extent than was the case with the other, smoother surface. Figure 11 (c & d) shows wisps and pockets of material that appear opaque with transmitted light, and we interpret this as iron-oxide-rich clay material. This material is surrounded by calcite microspar, with spar in closest proximity to it. The brightest luminesc ing material is the spar which we can again interpret as a result of a late replacement event of the clay by the calcite. Another possibility is inward displacement of the clay by growth of the calcite resulting in reduction of the clay volume. In view of the--l'trength
formed as a grain-coating or calcitan (Brewer, 1964; Reeves, 1976, p. 72) on the surface of the quartz grain. However, we prefer the idea that these very distinct spar zones may have formed by filling of a crack around the grain caused by shrinkage or solution of the micrite. This effect of circumgranular cracking has been reported by other workers (Reeves, 1976, p. 67; Goudie, 1983, p. I 04). Luminescence work on spar near some of the grains provides important information. Figure 10 (c & d) shows a number of quartz grains surrounded by spar with a few pockets of microspar. The spar shows very clearly luminescence stratigraphy demonstrating pe ripheral growth of the crystals, and the form of the stratigraphic surfaces indicates that the growth has been towards the quartz grains, and, we think, replacing the outer parts of the quartz grains. The form of the curved surface of the large quartz grain strongly suggests calcite replacement of the quartz, rather than mechanical displacement. Figure II (a & 296
Near-surface shrinkage and carbonate replacement
Fig. ll. (a) Transmitted light,(b) cathodoluminescence; (b) shows that the quartz grain has been replaced by successive growth on two faces and is highly irregular and indented on the other. (c) Transmitted light, (d) cathodoluminescence. A number of relicts of (opaque) clay aggregates have been strongly indented by growth of late-stage (bright luminescing) spar. This is interpreted as evidence of replacement of the clay aggregate by spar. Scale bar is 0·2 mm in all cases.
of displacement by repeated drying and wetting of the ground surface. Although shrinkage of soils due to drying is a common present-day occurrence where smectite clays are present (Bowles, 1984, p. 206), it is a general feature of most soils. We stressed the important observation that the bedding discordant carbonate concretions tend to occur between fissures filled by coarse sediment, and if the concretions are elongate (rod cornstones), they are parallel to the dominant fissure direction.
of the evidence for replacement of the quartz grains, we favour replacement and not displacement of the clay.
DISCUSSION OF PROCESSES FORMING THE ARRAN CORNSTONES General conclusions
We consider that the distinctive concretions and fissures of the Arran Cornstone formed under near surface conditions, either by pedogenic or ground water processes.
Special controls in the Amin setting
The widespread occurrence of concretionary carbon ate across Scotland at a similar stratigraphic level appears to reflect a distinctive climatic episode. But in Arran, fissuring and the rod cornstones are unusually highly developed. One reason for this may have been· the extreme local variability of water and sediment input, such as one might expect near a tectonically
Fissures as surface shrinkage cracks
The abundant fissures-some filled by introduced sediment, and others filled by expansion of the previously shrunk host sediment-provide evidence 297
S. K. Tandon and P. F. Friend
active escarpment along the line of the Highland
Our outcrop scale survey of carbonate rocks in the
Boundary (Friend eta/., 1963). The area is also unusual
sequence shows that bed-by-bed variation of carbon
compared with the other Scottish concretionary
ate type is normal. With the exception of the
outcrops in containing lavas intercalated within the
sandstones with carbonate cement, which could have
sediment sequence. These lavas may have weathered
formed at various stages during burial, all the other
to provide smectite-rich muds that would, on drying,
concretionary forms appear to have grown at an early
have yielded strongly developed fissure patterns and
enough
therefore encouraged rod cornstone development.
overlying beds. The form of the concretionary growth
This speculation however is not supported by any
varied from simple rounded nodules in homogenous
stage to be reworked into immediately
independent evidence of the original presence of
mudrocks, to various levels of rod development, and
smectites.
fissure polygon replacement, to growth of a continuous concretionary bed. Our micromorphological interpretation indicates
Absence of direct evidence of biological activity
early growth of micrite. We suggest that this involved
It is clear that organisms have a strong, if not
replacement by carbonate growing about closely
controlling, influence on the formation of many
spaced nuclei, but it also involved fracturing and
present-day soil profiles. In special circumstances (e.g.
possibly other forms of displacement, that produced
preceding air-fall tuffs) evidence of the major influence
the 'clotted' or patchy texture, the pisoliths, floating
of organisms has been found even on early Devonian surfaces (Allen
quartz grains and mudrock relicts. Our luminescence
& Williams, 1981, 1982; Allen, 1986).
work has produced clear evidence of continuing
However, no direct evidence of the role of organisms
replacement, particularly by carbonate spar, of ma
has been preserved in Arran. Our rod cornstones show
terial around the floating exotic particles, and there is
a resemblance in form and size to some rhizoliths (e.g. Klappa,
also evidence of open space filling of voids and cracks
1980b), particularly those that formed by
by spar.
progressive destruction and replacement of plant roots
Our general conclusion is
in the Cretaceous and Cenozoic carbonate-rich soils of southern France (Freytet
that the carbonate
concretions grew mainly by replacement, but that this
& Plaziat, 1982). The
continued through complex, repeated cycles of shrink
work in France has found intermediate stages in their
age displacement, even solution, and void filling.
development that convincingly support this explana
Some
tion. In the rod cornstone case, roots may also have
80 km to the ESE of Fallen Rocks on Arran, (1961) carried out a
in south Ayrshire, Burgess
been important although we have found no independ
pioneering study on material of broadly similar age.
ent evidence of their presence.
The concretionary carbonates are much thicker at his Ayrshire locality, and show some development of
Processes of carbonate growth
profiles consisting of distinctive lithological types. Thin sections made by us from Burgess's locality at
It is important to clarify our use of words in considering
Craigdullyet, Ayrshire, show a number of features
the local processes active within sediment near the
that we have not seen in our Arran material, e.g.
surface. Displacement of material involves its physical
laminated (?coated) grains, and particles with a lacey
movement, relative to some surrounding frame of
fabric similar to that described by Wrigi".t
reference. This may be due to local patterns of ( l ) change) or
(1982), or
with a needle-fibre calcite texture as described by
shrinkage (due, for example, to drying o r chemical
Solomon
(2) expansion [due, for example, to crystal
& Walkden (1985).
We consider that the distinctive features of the
growth or organism displacement (growing, moving)]. Replacement of material involves no physical move
Arran Cornstone Formation are consistent with a very
ment, but volume for volume solution and precipita
early diagenetic (pedogenic and ground water con
tion of other material. Open space filling involves the
trolled) environment, in which an unusually varied
passive fill by material (either mechanically, particle,
accumulation of sediments was repeatedly reworked
by particle, or chemically by precipitation) of spaces
by drying (shrinkage) and wetting (swelling), and
created ( l ) during original sedimentation as pores
repeated variations of water table and water percola tion. During these complex sequences, calcite grew,.
between particles,
(2) by shrinkage displacement due to drying, (3) by solution or (4) by organic decay (e.g.
mainly by replacement, but also locally by displace
root marks).
ment and void fill.
298
Near-surface shrinkage and carbonate replacement
Old Red Sandstone and the Highland Boundary in Arran, Scotland. Trans. Edin. geol. Soc., 19, 363-425. GARDNER, P.R.R. & HORNE, R.R. (1981) The stratigraphy of the Upper Devonian and Lower Carboniferous clastic sequence in southwest County Wexford. Bull. geol. Surv. Irl., 3,51-77. GEORGE,T.N.,JOHNSON, G.A.L.,MITCHELL,M.,PRENTICE, J.E., RAMSBOTTOM, W.H.C., SEVASTOPULO, G.M. & WILSON, R.B. (1976) A correlation of Dinantian rocks in the British Isles. Spec. Rep. geol. Soc. London, 7, 87 pp. GouDIE, A.S. (1983) Calcrete. In: Chemical Sediments and Geomorphology (Ed. by A.S. Goudie and K. Pye), pp. 93131. Academic Press,London. JAMES, H.P. (1972) Holocene and Pleistocene Calcareous crust (calcite) profiles, Criteria for subaerial exposure. J. sedim. Petrol., 42, 817-836. KLAPPA, C.F. (1978) Morphology, composition and genesis of
ACKNOWLEDGMENTS S. K. Tandon wishes to acknowledge receipt of a (British) Royal Society Bursary, and a D. S. T. Government of India, travel grant, which made his sabbatical visit to Cambridge possible. In the Cam bridge Department, Dr. J. A. D. Dickson introduced us to the use of cathodoluminescence and has provided us with may helpful ideas; Ben Harris and Ron Lee provided valuable technical support. We benefitted greatly from a critical discussion of an early version of this paper with Professor P. Freytet (Universite de Paris Sud, Orsay, France). Cambridge Earth Sciences Contribution No.
1057.
Quarternary calcretes from the Western Mediterranean: a petrographic approach. Unpublished PhD Thesis, Univer
sity of Liverpool. KLAPPA, C.F. (1980a) Brecciation textures and structures in Quarternary calcretes(caliche) profiles from eastern Spain: the plant factor in their formation. Geol. J., 15,81-89. KLAPPA, C.F. (1980b)Rhizoliths in terrestrial carbonates: classification, recognition, genesis and significance. Sedi mentology, 27,613-629. LEEDER, M.R. (1976) Palaeographic significance of pedo genic carbonates in the topmost Upper Old Red Sandstone of the Scottish Border Basin. Geol. J., II, 21-28. MACCARTHY,I.A.J,GARDNER, P.R.R. &HORNE, R.R(1978) The lithostratigraphy of the Devonian--early Carbonifer ous succession in parts of Counties Cork and Waterford, Ireland. Bull. geol. Surv. Irl, 2, 265-305.
RE FERENCES
ALLEN, J.R.L. (1965) Sedimentation and palaeography of the Old Red Sandstone of Anglesey, North Wales. Proc. Yorks. geol. Soc., 35,139-185. ALLEN, J.R.L. (1974) Studies in fluviatile sedimentation: implications of pedogenic carbonate units,Lower Old Red Sandstone,Anglo-Welsh outcrop. Geol. J., 9, 181-208. ALLEN, J.R.L. (1986) Pedogenic carbonates in the Old Red Sandstone facies (late Silurian - early Carboniferous) of the Anglo-Welsh area, southern Britain. In: Paleosols: Their recognition and Interpretation (Ed. by V. P. Wright). Blackwell Scientific Publications,Oxford,315 pp. ALLEN, J.R.L. & WILLIAMS, B.P.J. (1981) Sedimentology and stratigraphy of the Townsend Tuff Bed (Lower Old Red Sandstone) in South Wales and the Welsh Borders. J. geol. Soc. London, 138,15-29. ALLEN, J.R.L. & WILLIAMS, B.P.J. (1982) The architecture of an alluvial suite: rocks between the Townsend Tuff and Pickard Bay Tuff beds (early Devonian), southern Wales. Phil. Trans. R. Soc. B., 297, 51-89. BowLES, J.E. (1984) Physical and Geotechnical Properties of Soils, 2nd edn,McGraw-Hill,New York,478 pp. BREWER, R. (1964) Fabric and Mineral Analysis of Soils. Wiley,New York,470 pp. BURGESS, J.C. (1961) Fossil soils in the Upper Old Red Sandstones of south Ayrshire. Trans. geol. Soc. Glasgow, 24, 138-153.' CHISHOLM, J.l. & DEAN, J.M. (1974) The Upper Old Red Sandstone of Fife and Kinross: a fluviatile sequence with evidence of marine incursion. Scott. J. Geo/., 10,1-30. FOLK, R.L. (1962) Spectral subdivisions of limestone types. In: Classification of carbonate rocks (Ed. by W.E. Ham). pp. 62-84. Mem. Am. Ass. Petrol. Geol., Tulsa, OK., 279 pp. FREYTET, P. & Plaziat, J.C. (1982) Continental carbonate
NAGTEGAAL,P.J.C. (1969) Microtextures in recent and fossil caliche. Leidse Geol. Meded., 42, 131-142. PARNELL, J. (1983a) The Cothall Limestone. Scott. J. Geol., 19,215-218. PARNELL,J. (1983b) Ancient Duricrusts and related rocks in perspective: a contribution from the Old Red Sandstone. In: Residual Deposits: Surface Related Weathering Processes and Materials (Ed. by R.C.L. Wilson), pp. 197-209. Blackwell Scientific Publications,Oxford. PERYT, T.M. (ed.) (1983) Coated Grains. Springer-Verlag, Berlin,655 pp. PICK,M.C. (1964) The stratigraphy and sedimentary features of the Old Red Sandstone,Portishead Section, north-east Somerset. Proc. geol. Ass., 75,199-221. READ, W.A. & JOHNSON, S.R.H. (1967) The sedimentology of sandstone formations within the Upper Old Red Sandstone Measures west of Stirling. Scott. J. Geol., 3, 247-267. REEVES, C.C. (1976) Caliche. Estecado books, Lubbock, (TX),233 pp. SOLOMON, S.T. & WALKDEN, G.M. (1985) The application of cathodoluminescence in interpreting the diagenesis of an ancient calcrete profile. Sedimentology, 32, 877-896.
sedimentation and pedogenesis-Late Cretaceous and Early Tertiary of southern France. Contrib. in sediment 12E.
STEEL, R.J. (1974) Cornstone (fossil caliche}-its origin, stratigraphic and sedimentologic importance in the New Red Sandstone,Western Scotland. Geol. J., 83,351-369.
Schweizerbartische Verlagsbuchhandlung (Nagele u. Ob ermiller),Stuttgart,213 pp. FRIEND, P.F., HARLAND, W.B. & HUDSON, J.D. (1963) The
TUNBRIDGE,I.P. (1981) Old Red Sandstone sedimentation299
S. K. Tandon and P. F. Friend
WESTOLL, T.S. (1977) Northern Britain. In A correlation of Devonian rocks of the British Isles (Ed. by M. R. House, J. B. Richardson, W. G. Chaloner, J. R. L. Allen, C. H. Holland & T. S. Westall). Spec. Rep. Geol. Soc. London, 8, llOpp. WRIGHT, V.P. (1982) Calcrete Paleosols from the Lower Carboniferous Llanelly Formation, South Wales. Sedi ment. Geol., 33, 1-33.
an example from the Brownstones (highest Lower Old Red Sandstone) of south central Wales. Geol. J., 16, 111124. WAITS,N.L. (1978) Displacive calcite: evidence from recent and ancient calcretes. Geology, 6, 699-703. WATIS, N .L. (1980) Quarternary pedogenic calcretes from the Kalahari (Southern Africa): mineralogy, genesis and diagenesis. Sedimentology, 27,661-686.
(Manuscript received 18 November 1987; revision received 25 January 1989)
300
Reprinted from Sedimentology
(1985) 32 877-896
The application of cathodoluminescence to interpreting the diagenesis of an ancient calcrete profile S T E P H E N T . S OL O M O N* and G O R D O N M . W A LK D E N Department o fGeology, Marischal College, University ofAberdeen, Aberdeen AB9 1AS, U.K.
ABSTRACT
A laterally extensive calcrete profile has been identified in the Late Asbian (Lower Carboniferous) shallow marine shelf limestones of the Llangollen area, North Wales. The upper surface of the profile is defined by a laterally discontinuous palaeokarstic surface and by laminated calcareous crusts which developed within the underlying limestone. The profile contains a unique series of early pore-filling vadose cements which only occur down to I m below the palaeokarstic surface. Cathodoluminescence reveals that these cements pre-date the late pore filling meteoric phreatic cements which occur throughout local Asbian lithologies. A spar cement stratigraphy has been established for the calcrete profile. Subaerial vadose cements comprise two generations of non-luminescent cement, followed by a brightly luminescent generation which occasionally shows an acicular habit. This needle-fibre calcite represents the final stage of vadose cementation. Precipitation of vadose cements was contemporary with subaerial alteration and micritization of the limestone. Textures, visible only with cathodoluminescence, provide evidence of recurrent periods of fabric dissolution. The most extensive phase of dissolution occurred immediately after the precipitation of the non-luminescent subaerial vadose cements. Several different textures have been recorded, each reflecting the morphology of a partially dissolved substrate. Dissolution textures are generally confined to the walls of the larger pores and to early brecciation fractures. These probably acted as fluid pathways in the calcrete during early subaerial diagenesis. Much of the non-marine micrite in the calcrete profile appears as needle-fibre calcite under cathodoluminescence. This acicular calcite was probably formed in response to localized supersaturation of meteoric pore fluids caused by periods of near-surface evaporation. Since needle-fibre luminescence is strongly variable, these ambient conditions are not believed to have directly controlled the activator ion concentrations of cementing pore waters. N eedle-fibre calcite is considered to be a cement precipitate which has almost completely recrystallized to micrite, probably during the late stages of subaerial diagenesis. Two generations of subaerial micrite which define a 'micrite stratigraphy', have been distinguished under cathodoluminescence. Reconstructing the diagenetic history of this ancient calcrete profile has revealed that subaerial alteration was multistaged, with many diagenetic processes acting simultaneously during a single phase of emergence.
INTRO DUCTION
The use of cathodoluminescence to examine an ancient calcrete profile has provided much additional infor mation about the diagenetic processes operative during the subaerial alteration of a marine carbonate sediment. A cement stratigraphy has been established
for the calcrete and we demonstrate that pore-filling cementation and subaerial alteration occurred con comitantly, within 1 m of an emergent surface. Textures, visible only with cathodoluminescence, provide evidence of recurring periods of extensive dissolution during subaerial e xposure. Micritic fabrics characteristic of calcretes (particu larly alveolar-like pore bridging structures and grain
*Present address: Conoco (U.K.) Ltd, Park House, 11 6 Park Street, London W 1 Y 4N N, U.K. Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
301
S. T. Solomon and G. M. Walkden coatings) appear as needle-fibre calcite under catho doluminescence. The origin of this needle-fibre calcite and its relationship to subaerial vadose micrite is examined together with the control of needle-fibre calcite on the formation of calcrete fabrics. The simultaneous operation of many subaerial diagenetic processes during a single phase of emerg ence is revealed from a detailed diagenetic history reconstruction using combined transmitted light and cathodoluminescence petrography. Local geology and previous studies
An almost complete succession of Late Asbian (Lower Carboniferous) shallow marine shelf limestones is exposed on the western scarp face of the Eglwyseg Rocks, 5 km north of Llangollen, North Wales (Fig. 1). Recent studies by Somerville (1977, 1979) and Gray (1981) have divided the succession into cyclic units. These reflect sedimentation during minor fluctuations in Late Asbian sea-levels, each cycle beginning with a period of transgression, followed by gradual regression and finally subaerial emergence. Ancient emergent surfaces have been described by Somerville (1977, 1979) and Gray (1981) and are similar both in outcrop and thin section to other Carboniferous calcretes (e.g. Walkden, 1974; Walls, Harris & Nunan, 1975; Harrison & Steinen, 1978; Wright, 1983) and also Quaternary calcretes (notably, Multer & Hoffmeister, 1968; James, 1972; Braith waite, 1975; Harrison, 1977). This report describes a I m deep calcrete profile which formed during one of these episodes of subaerial emergence. The calcrete profile has been traced for approximately 3 km along the NW-SE trending outcrop and corresponds to the top of Somerville's (1977) 'Cycle 7' and Gray's (1981) 'Eg. 8 surface'. Below, the petrology of the calcrete profile is examined briefly, before considering the petrographic evidence for the major diagenetic processes which led to its development.
PETROLOG Y OF T HE CALCRETE PROFILE
The calcrete profile is laterally extensi ve and has been studied at four locations along the western scarp face of the Eglwyseg Rocks (Fig. 2). In outcrop, this ancient emergent surface is defined by a laterally discontinuous mammillated palaeokarstic surface with low relief (up to 0·15 m deep). This is overlain by
a discontinuous clay palaeosol up to 30 mm thick. The development of calcrete fabrics in outcrop is laterally variable. Discontinuous brown-stained laminated cal careous crusts (up to 0· 2 m thick) underlie the emergent surface. Texturally these laminated crusts can be divided into two types (after Walkden, 1974) which are commonly interstratified and rapidly intergrading: (I) Densely laminated dark brown stained crusts (0·02-0·1 m thick). Laminae are approximately I mm thick, sharply defined, subhorizontal and locally continuous (described in thin section below). (2) Porous light brown-stained crust (up to 0·1 m thick) with discontinuous irregular laminae and a finely brecciated fabric. Round and planar spar filled vugs of millimetre size are common, forming irregular branching networks. Lithologies with a 'rubbly'/'stylonodular' weathering expression often partially mask calcrete fabrics in outcrop (Fig. 2, Joe. D). Therefore, polished slabs and thin sections are required to identify the depth of the ancient calcrete profile. In thin section, the calcrete profile is characterized by abrupt transitions from patches of relatively unaltered marine sediment into regions which are intensely micritized. The micritized regions contain remnants of marine allochems together with a variety of micrite fabrics. Some primary porosity is preserved together with secondary pore types including circum granular brecciation fractures, calcified root tubes and solution vugs. Pore space is filled by a sequence of low-magnesium calcite cements with both subaerial vadose and meteoric phreatic characteristics.
Micrite fabrics of the calcrete profile
Micrite fabrics present in the calcrete profile are interpreted as the product of both replacement alteration of marine allochems and carbonate precip itation. Micrite formed in both marine and non marine diagenetic environments: (I) Marine micrite is present in most lithologies of the Asbian succession as a centripetal alteration of marine allochems. It can be distinguished from the products of non-marine alteration where it underlies the first generation of non-luminescent subaerial vadose cements (Table 1). (2) Non-marine micrite is present only i mmediately below ancient emergent surfaces. In this calcrete
302
Cathodoluminescence and calcrete diagenesis
I
I
N. WALES
-
BRIGANTIAN CARBONATES
D
ASBIAN CARBONATES
D
BASEMENT BEDS
EZJ
ORDOVICIAN & SILURIAN SLATES
Fig. 1. Outcrop of Dinantian limestones in North Wales (shown in black) and geology of the region immediately north of Llangollen showing the location of four outcrops of the calcrete profile (A-D).
B
c
D
GR 2220 4524
GR 2239 4409
GR 2290 4334
A GR 2203 4515
metres
q_<>_
OoOoO
M
0·8 km
-< >-
I I I _l _!____!_I_ _!_
B
1_1_ 1_11
,-_1-_li
II-II
CALCRETE
-<
PROFILE
M
r:_o.:!. :?_ o_o::_(
MASSIVE WHITE GRAINSTONES
GREY PACKSTONES AND GRAINSTONES
GRAINSTONES WITH 'RUBBLY/STYLONODULAR' OUTCROP EXPRESSION
0
T
··_.! -I l LLL..l-
§
LAMINATED CALCAREOUS CRUST. NO PALAEOKARSTIC SURFACE
B
MAMMILLATED PALAEOKARSTIC SURFACE & LAMINATED CALCAREOUS CRUST
EJ
PALAEOKARSTIC SURFACE WITH STEEP SIDED DEEP MAMMILLAE
I-;< �1
GREY WACKESTONES
D
CALCIFIED PLANT ROOT TUBULES
BLACKENED CALCRETE PEBBLES IN CLAY PALAEOSOL
Fig. 2. Outcrop diagrams of the calcrete profile developed in different lithologies at four locations along the western scarp face of the Eglwyseg Rocks.
303
S. T. Solomon and G. M. Walkden cement occasionally occur enclosed in a micritic matrix (Fig. 3). Since the non-luminescent spar shows no evidence of displacive growth (unlike W atts, 1978) this texture is considered to have formed by the aggradational/constructive growth of micrite (perhaps by nucleation on pellets). This could cause reduction of the original interparticle porosity and fabric expansion, finally producing a dense micrite. This may have occurred in the manner described by Harrison (I 977) or Cal vet (I 982).
profile non-marine micrite often has a patchy distribution and displays fabrics which are closely similar to both the Recent and ancient calcretes referred to above. The principal subaerial micrite fabrics are summarized below: Dense micritic laminar fabric
Laminae are subhorizontal, undulating and are de fined by colour variations from dark grey to grey brown. They vary in thickness from 50 to 350 mm and have irregular, slightly diffuse boundaries. Laminae are laterally continuous for several tens of millimetres and are commonly interspersed with regions of grey brown uniform micrite.
Pore lining and bridgingfabric
Arcuate bridges of 'micrite' subdivide and line pores producing a fabric similar to alveolar texture (Esteban, 1974). This fabric occurs in primary pores, subaerial brecciation fractures and calcified plant root tubules. Cathodoluminescence reveals that the arcuate bridges of'micrite' consist of luminescing needle-fibre calcite. This suggests that pore lining and bridging fabric may have a different origin from other fabrics listed above. This is discussed in detail later.
Pelleted fabric
Pellets comprise subrounded, roughly spherical micri tic grains, usually with slightly diffuse irregular outlines. Diameters range from 30 to 100 11m, averag ing 40 11m. Pellets are structureless but may be amalgamated into irregular shaped masses. Well sorted pelleted fabrics may pervade the entire matrix or occur as restricted irregular patches, occasionally associated with calcified plant root tubules (see below).
Secondary porosity
Partially micritized allochems
Two major pore types evolved during subaerial diagenesis :
Identifiable marine allochems occur within the fabrics described above, their structures being invaded by the micrite of the surrounding matrix. This suggests that micritization took place partly by in situ alteration of manne grams.
Brecciation fracture porosity
These millimetre scale, irregular or curved fractures cross cut both the micritized calcrete matrix and a generation of non-luminescent spar cement. Allo chems are frequently separated from the enclosing
'Floating' grain fabrics
Allochems coated by clear prismatic non-luminescent
Fig. 3. Thin section, plane polarized light; 'floating' grain fabric. A brachiopod fragment (BR) coated by clear prismatic spar (CS) is enclosed by the surrounding micritic matrix. Field of view: 2·05 mm.
304
Cathodoluminescence and calcrete diagenesis matrix producing textures similar to James' (1972) 'exploded jigsaw puzzle'.
Calcifiedplant root tubules The presence of tubular pores in the calcrete profile is indicated by the occurrence of both circular (internal diameter 30-750 11m) and elongate pore sections lined by micrite and/or cement. Generally, elongate pores have a roughly constant width (up to 650 11m). They are straight or slightly sinuous and, at location B (Fig. 2), show single dichotomous branching. The signifi cantly decreased diameter of the bifurcations (criteria of Klappa, 1980), together with the absence of micritic pellets from the central void of the tubular pores (unlike Wright, 1983), suggests that these tubular pores are formed as the result of plant root penetration and not by burrowing animals. The walls of these tubular pores are preserved in one of two ways:
(I ) Micrite: alternating dark orange-brown and grey stained irregular micritic laminae form roughly concentric layers around the central hollow of the tubular pore (Fig. 4A). This type of preservation is characteristic of the strongly micritized and densely laminated regions of calcareous crusts. In structure, it is probably similar to that of rhizoliths recorded by Klappa (1980) and may have formed by calcite impregnation of the plant root tubule either during decay or perhaps while still living. (2) Fibrous calcite: laminated, radial fibrous brown calcite lines pore walls (approximately 50 11m thick) in some places. Characteristically, it shows rapidly fading blue-purple luminescence which may be caused by an argillaceous or organic residue being destroyed in the beam heat. To conclude, this Lower Carboniferous calcrete profile shows many petrographic fabrics which are common in other examples of both modern and ancient calcretes.
(A) Thin section, plane polarized light; a calcified plant root tubule with a wall structure of irregular micritic laminae which form roughly concentric layers around the spar cement filled central hollow. (B) Under cathodoluminescence; narrow, brightly luminescent zones of micrite/needle-fibre calcite commonly occur in the immediate vicinity of calcified plant root tubules. Field of view: 1-4 mm.
Fig. 4.
305
S. T. Solomon and G. M. Walkden PROCESSES OF SUBAERIAL
(I) Clear blocky-prismatic spar (CS) (Table I, column
DIAGENESIS CONTRIBUTING TO
2): this inclusion-free spar has a variable morphology and lines both inter- and intra-particle pores. Crystals with prismatic morphology have length: width ratios up to 4: I . Blocky morphology is characterized by roughly equant crystals up to I 00 �m width. Although this cement shows a slight tendency to thicken between allochems (Fig. 5), true meniscus and gravitational morphologies have not been observed. This generation of clear spar is the first spar cement to coat substrates and fills aragonite moulds thus oost-dating aragonite dissolution. Occasionally, thin, medium-bright orange luminescent subzones with pitted boundaries are visible in this essentially non-luminescing cement (Fig. 58). Clear non-luminescent sparcement accounts for an estimated 5% of initial porosity fill in larger pores of the calcrete profile. (2) Light brown-coloured spar (B) (Table I , column 2): light brown-coloured spar shows a variable morphol ogy but occupies a constant position in relation to the cement sequence, either coating early clear spar or occurring directly upon marine sediment substrates. Brown-coloured cements may be subdivided into two principal morphologies:
CALCRETE FORMATION
The petrographic interpretation of ancient calcretes is often hindered by the wide variety and irregular development of calcrete fabrics and cements. The petrographic complexity of calcretes reflects the brief and intermittent operation of numerous genetic processes in the subaerial diagenetic environment, together with the uneven movement of fluids through the parent sediment. This report examines new evidence from cathodoluminescence petrography for three major diagenetic processes which altered marine sediments during subaerial exposure: cementation, recurring dissolution and subaerial micritization.
CEMENTATION
Early cementation is one of the most important diagenetic alterations during subaerial exposure, being responsible for partial lithification of the otherwise relatively unconsolidated marine sediment. Up to 30% of porosity was occluded by early subaerial vadose cements near the base of this I m deep calcrete profile.
(a) Isopachous ( < 20 �m thick) brown fibrous calcite with numerous fine opaque (organic?) inclusions (Fig. SA; within (I )). lndividial fibres are only a few micrometres wide and are orientated perpen dicular to substrate. This morphology commonly occurs as overgrowths on early clear spar (above) in optical continuity with the underlying cement and is volumetrically unimportant. (b) Pale brown-coloured, stubby prismatic or blocky calcite; stubby prismatic cements are approxi mately 80 �m x 30 �m, while blocky crystals are inequant with widths ranging from 5 to 30 �m. This second morphology increases in volumetric importance downwards through the calcrete pro file at locations C to D (Fig. 2). At I m they form up to 30% of pore filling spar cement. Below this they are absent.
Spar cement stratigraphy
The cements of the calcrete profile may be divided into a three-part stratigraphy based on their distribu tion and their appearance under cathodoluminescence (Fig. 5 and Table I, column I). The first non luminescent cements have a restricted vertical strati graphic range occurring only to a depth of I m below the subaerial exposure surface. In contrast, the later pore-filling cements occur in lithologies throughout the Asbian succession. Cements are described below in order of decreasing age :
Brown colouration may be caused by organic matter and records a geochemical change in the composition of cement-precipitating pore fluids which is not marked by changes in luminescence; brown-coloured spar is non-luminescent. Both the clear, and brown-coloured generations of non-luminiscent cements have an erratic distribution within the calcrete profile and variable morphology, suggesting a subaerial vadose origin. Despite this, the possibility that the clear non-luminescent cement
Non-luminescent cements (I ) (Table I , column I ) These cements are characterized by irregular patchy distribution and variable volumetric importance in different parts of the profile. Non-luminescent cements are typically non-substrate selective and have non ferroan calcite mineralogies (artificial stain of Dick son, 1965). On the basis of plane light petrography (Table I, column 2) they have been subdivided into two generations: 306
Cathodoluminescence and calcrete diagenesis
Fig. 5. (A) Thin section, plane polarized light; an interparticle pore in the calcrete profile containing the full sequence of spar cements. BR: brachiopod, SM: submarine micritised rim of the brachiopod, CR: crinoid ossicle, CS: clear spar (with blocky/ prismatic form elsewhere), B: light brown-stained spar, IS: clear, inequant, blocky spar. (B) Under cathodoluminescence; I : non-luminescent cements, 2: first brightly luminescent cement, 3: non-luminescent cements with brightly luminescent pitted subzones, 4: final pore-filling medium luminescent cements with smooth subzones. Field of view: 1 ·9 mm.
307
Table I. Interpreted cementation and fabric formation history of a Late Asbian calcrete profile; Llangollen, north W ales. Horizontal lines represent synchronous
surfaces 'SPAR CEMENT STRATIGRAPHY' SEQUENCE OF DIAGENETIC EVENTS
Cll 00--Cf) 11:1- w O
CATHODOLUMINESCENCE CHARACTERISTICS
PLANE POLARISED
CALCRETE LIGHT
FABRIC
CHARACTERISTICS
FORMATION
(COL.2)
(COL.3)
(COL.1)
medium lum. final pore filling cements with smooth subzones (4)
clear inequant
non-luminescent cements with brightly lum. pitted subzones (3)
blocky spar (IS)
� :-l
�
-Cll w 0 00
I
Cll w z w Cl
�
needle-fibre calcite with variable luminescence
clear needle-fibre calcite
0 ;:: second generation micrite (acicular
first brightly luminescent cement
(2)
thin, clear cement coating
C.L. signature)
...J <( a: w < m :::> Cll
----;;; W
Dissolution Hiatus
non-luminescent
cements ( 1)
light brown stained spar with a variable form (B)
clear spar with blocky/prismatic form (CS)
r . ...
generation micrite
c - 0
.. -;: .. ..
;: .. ., ;: .£>"
�
0 0 �
0 w Cll 0 0 <( >
..
l
.. " " .. m
.,
;:s >:>..
� !
0
';;
�
Q
c 0
0 ;:s
c 0
.. " " • 0
� ;;;::: >:>..
�
Cathodoluminescence and calcrete diagenesis ((CS); Table I, column 2) may have been precipitated in the meteoric phreatic zone before subaerial expo sure, cannot be completely discounted. In contrast, the light brown-coloured spar (B; Table I, column 2) is demonstrably a subaerial vadose cement, having been precipitated concurrently with subaerial breccia tion (demonstrated below) and being absent below the I m deep calcrete profile. First (major) bright orange luminescent cement (2) (Table I, column I) This distinctive 'marker' cement within the cathodol uminescence cement stratigraphy coats earlier non luminescent cements as a thin layer (Fig. 5B). Its boundaries are sharply defined and pitted. This brightly luminescent cement is believed to have formed under the influence of subaerial diagenetic conditions because of its association with needle-fibre calcite (discussed later). Non-luminescent cements with brightly luminescent pitted subzones andfinal pore-filling medium luminescent cements with smooth subzones (3) and (4) (Table I, column I) I n transmitted light this sequence of pore-filling cements is characterized by clear inequant blocky crystals which increase in size dramatically towards pore centres (Fig. 5A and Table I, column 2). On the basis of transmitted light petrography and cathodoluminescence subzone morphology (after cri teria of Meyers, 1974), cements (3) and (4) (Fig. 5 and Table I, column I) are interpreted to have been precipitated during meteoric p hreatic diagenesis. This finding is in accordance with earlier petrographic and geochemical work of Walkden & Davies ( 1983) and Walkden & Berry ( 1984) who described a closely similar sequence of meteoric phreatic cements (their zones 1-3) from Asbian successions in Anglesey and Derbyshire ·respectively. Meteoric phreatic cements form an estimated 70-85% of pore-fill in lithologies of the calcrete profile. Relationship of cementation to calcrete fabric formation
By combining transmitted light and cathodolumines cence petrography, the timing of cement precipitation in the calcrete profile can be related to the evolution of fabrics which formed in response to other subaerial diagenetic processes.
Subaerial brecciation Cross-cutting relationships (Fig. 6A, B) show that clear, prismatic, non-luminescent spar cement pre dates brecciation. Brown-stained inclusion rich spar cements are both cross-cut by brecciation fractures and form the first cement to coat brecciation fracture walls. This shows that subaerial brecciation occurred during the precipitation of brown stained non luminescent cements (Table 1, column 3). Younger generations of cement are u naffected by subaerial brecciation. Calcification ofplant roots Both subaerial cements and micrites define calcified plant root tubules indicating that plant root growth/ decay occurred concurrently during early cementation and subaerial micritization (Table 1, column 3). Subaerial micritization The exact relationship between precipitation of non luminescent cements and the onset of subaerial micritization is unclear. I f ' floating' grain fabrics (Fig. 3) are formed by aggradational micritization (consid ered earlier), this may indicate that the non-lumines cent cements (I) (Table 1, column 1) were precipitated before or concurrently with the onset of subaerial micritization. It will be shown that subaerial micriti zation occurred both before and after a major dissolution event (Dissolution Hiatus; Table I) and possibly continued after the precipitation of needle fibre calcite. These observations and interpretations show that the non-luminescent cements (I) (Table 1, column 1) were precipitated concurrently with subaerial breccia tion, micritization and the formation of calcified plant root tubules. This implies that the non-luminescent cements formed during subaerial vadose diagenesis.
RECURRING DISSOLUTION
Textures visible only with cathodoluminescence pro vide evidence that dissolution occurred repeatedly during subaerial vadose diagenesis. Dissolution sur faces cut both cements and allochems and are coated by a bright orange luminescent phase which marks their presence. In Fig. 6, dissolution surfaces are visible in the coral wall structure (Fig. 6B, b) and along the inter-crystalline bou ndaries of non-lumines cent subaerial vadose cements (Fig. 6B, a). 309
S. T. Solomon and G. M. Walkden
Fig. 6. (A) Thin section, plane polarized light: a coral fragment in the calcrete profile cut by a subaerial brecciation fracture
(f) filled by spar cements. A: unbrecciated intra-particle pore, B: brecciated intra-particle pore, C: coral wall. (B) Under cathodoluminescence; dissolution textures formed during subaerial diagenesis. l , 3, 4: pore-filling cements (as Fig. 5B) a, b, c: locations referred to in the text. Field of view: l· 35 mm.
Only the non-luminescent subaerial vadose cements display dissolution textures of this type, demonstrating that dissolution occurred before the onset of meteoric phreatic cementation.
bright luminescent phase defining them. In brecciated pores (Fig. 6B, pore B) the bright orange luminescent phase not only 'invades' intercrystalline boundaries (Fig. 6B, a) but also forms a thin, even coating on the terminations of the non-luminescent subaerial vadose cements (Fig. 6B, c). In adjacent unbrecciated pores showing the same cement stratigraphy (Fig. 6B, pore A and Fig. 7); again, this first bright luminescent
Origin of the textures
The process responsible for forming the textures can be determined by examining the characteristics of the 310
Cathodoluminescence and calcrete diagenesis 1). This petrographic observation is of major impor tance since it establishes that the bright luminescent phase is a cement precipitate. This implies that the textures were produced by erosion/degradation of substrates which occurred before precipitation of the first bright orange luminescent cement. The brightly luminescent cement merely acts as a passive marker highlighting the morphology of the solution corroded substrate. We consider dissolution by calcium carbon ate undersaturated pore fluids to be the most likely cause of substrate erosion. The texture seen with cathodoluminescence varies with substrate type. Three major textures are visible within the calcrete profile defined by the first bright orange luminescent cement filling the following substrates:
(I) Dissolution ofnon-luminescent subaerial vadose cements
Fig. 7. Thin section, cathodoluminescence; an unbrecciated
intraparticle pore in the same coral as Fig. 6(B). The same sequence of spar cements fill the pore but dissolution textures are absent. Field of view: 0· 7 mm.
Comparison of transmitted light and cathodolumi nescence views of Fig. 6 shows that the bright orange luminescent cement has filled space opened along intercrystalline boundaries; the prismatic crystalline morphology of the non-luminescent cement control ling the appearance of the texture seen with cathodo luminescence. Laminar pores are believed to have been created
phase coats the non-luminescent subaerial vadose cements, but does not 'invade' intercrystalline bound aries. This first bright orange luminescent phase occurring in both pores is an identifiable part of the cathodoluminescence cement stratigraphy; being the first brightly luminescent cement (2) (Table I, column
Fig. 8. Thin section, cathodoluminescence, a dissolution texture developed in the wall structure of a cora]· to a limited depth from a circumgranular brecciation fracture (f). A: unbrecciated pore showing no associated dissolution textures. Field of view: 2· 3mm.
311
S. T. Solomon and G. M. Walkden along intercrystalline boundaries which represent more favourable sites for dissolution (higher free energy) than the non-luminescent calcite crystals. (2) Dissolution offibrous calcite walls ofmarine allochems The bright luminescent cement also fills rugose coral wall ultrastructure (Figs 6B and 8). This texture, seen with cathodoluminescence, is closely comparable with the rugose coral wall ultras�ructure figured by Sand berg (1975) (Fig. 9). It seems likely that this texture
Fig. 9. Sandberg (1975, plate 4-S.E.M; rugose coral wall ultrastructure). We thank Dr P. Sandberg for kindly permitting us to use this photomicrograph. Fig. 10. Thin section, cathodoluminescence; dissolution
was formed by dissolution along the inter-crystalline boundaries of calcite fibres which constitute the coral wall ultrastructure. The resulting laminar pores would have been filled by bright orange luminescent cement producing the texture shown in Fig. 8. It is expected that brachiopods and other fibrous walled allochems, similarly affected by dissolution and filled by brightly luminescent cement, will show a texture reflecting the arrangement of calcite fibres in their wall ultrastru c ture.
textures developed in matrix micrite and fibrous calcite walled allochems immediately adjacent to a (now) spar cement filled subaerial brecciation fracture. Field of view: l-4mm.
probably because of the microcrystalline size of the substrates. The examples cited above show that the morphology/structure of the partially dissolved sub strate controls the appearance of the texture seen with cathodoluminescence. The brightly luminescent ce ment defining the textures only penetrates a limited distance into the substrate from the palaeo-pore surface (e.g. Fig. 10). This is interpreted to indicate that dissolution acted at pore-substrate interfaces and selectively dissolved underlying substrates to a limited depth. Dissolution textures only occur adjacent to subaerial
(3) Dissolution of micrite substrates (including the products of both marine and subaerial micritization) I nfilling brightly luminescent cement defines a fine random anastomosing texture (Fig. 10). Even at high magnification the resolution of this texture is poor,
312
Cathodoluminescence and calcrete diagenesis brecciation fractures and in the larger inter-, intrapar ticle and mouldic pores within the calcrete profile. The genetic implications of these observations are considered below. A model for dissolution during subaerial diagenesis
It is proposed that extensive fabric dissolution and weakening occurred during subaerial vadose dia genesis. This was probably caused by strongly under saturated (with respect to CaC03) fluids travelling through the developing calcrete profile. These fluids (percolating rainwaters?) moved through larger pores and along subaerial brecciation fractures which may have acted as fluid pathways within the relatively impermeable calcrete (which was already partially cemented and micritized). Progressive dissolution of substrates occurred to a limited depth in the walls of pores through which the dissolving fluids travelled. The solution-etched surfaces of substrates were sub sequently filled by a bright orange luminescent cement producing the characteristic textures described. Pores containing non-luminescent cements or other sub strates which show no evidence of dissolution are believed to have been closed or simply by-passed during this stage of d iagenesis (e.g. Fig. 6B, pore A and Fig. 7). Fig. 11. Thin section, cathodoluminescence; cemented in Recurring dissolution
terparticle pore in calcrete profile. Brightly luminescent cement subzones (a and b) define two dissolution surfaces which pre-date the major dissolution event defined by textures associated with first major bright luminescent subzones. Field llf view: 0· 57 mm.
So far, only one major d issolution hiatus has been discussed. However, some lithologies within the calcrete profile show evidence of two earlier dissolu tion hiatuses which are recognizable by the same cathodoluminescence textures, but defined by rela tively dull orange luminescent cements (Fig. 11; subzones a and b). This proves that fabric dissolution was a recurring process during the precipitation of non-luminescent subaerial vadose cements.
ination of these micrite fabrics with cathodolumines cence has revealed that many have an acicular cathodoluminescence signature. In particular, the micritic pore-lining and pore-bridging fabrics appear as a dense mesh of luminescent calcite needles up to 50 11m long and 5J.!m wide (Fig. 12). Small calcite needle-fibres are occasionally visible in transmitted light, but their petrographic characteristics are best studied with cathodoluminescence.
SUBAERIAL MICRITIZATION AN D TH E FORMATION OF N E E DL E-FIBR E CALCIT E
Much of the original marine matrix of lithologies in the calcrete profile has been subaerially altered to micrite which displays a variety of characteristic fabrics (already described). Normally, the term micrite (Folk, 1959) is used to describe calcite crystals 1-4 Jlm in diameter which typically appear brown and are poorly resolved in transmitted light. However, exam-
Cathodoluminescence petrography of needle-fibre calcite
Needle-fibres are straight, unbranched and usually form steep sided rhombs or, more rarely, rods with blunt prismatic terminations. They have variable packing density and are commonly arranged in one of the following textures:
313
S. T. Solomon and G. M. Walkden
Fig. 12. (A) Thin section, plane polarized light; subaerial 'micrite' from the calcrete profile showing arcuate pore bridging structure. (B) Under cathodoluminescence; subaerial 'micrite' appears as dense mesh of brightly luminescent needle-fibre calcite. Note that needle-fibre meshes act as substrates for later pore-filling cements (non-luminescent and medium luminescent). Field of view: 1·65 mm.
(1) Random, or subparallel orientated acicular meshes. (2) Tangential surface coatings on pore walls. (3) Bifurcating arcuate pore bridging structures con sisting of subparallel orientated needles (this texture closely resembles alveolar texture of Esteban, 1974).
All three textures can occur within the same pore (Fig. 138). The timing of needle-fibre calcite growth relative to other stages of cementation can be determined by cement stratigraphy ; the needle-fibre calcite has acted as a substrate for later meteoric phreatic cements (Fig. 128), but overlies non-luminescent subaerial vadose 314
Cathodoluminescence and calcrete diagenesis
Fig. 13. (A) Thin section, plane polarized light; subaerially brecciated coral (C) in the calcrete profile. f: circumgranular brecciation fracture. Early non-luminescent prismatic cements (CS, B as Fig. 5A) line the intraparticle pore which is partly filled by pore bridging 'micrite' (PM). (B) Under cathodoluminescence; pore bridging 'micrite' appears as needle-fibre calcite lining the intraparticle pore and coating the dissolution-eroded surface of early non-lum'inescent cements. Dissolution textures also occur in association with the circumgranular brecciation fracture. Field of view: l·l mm.
The luminescence of needle-fibre calcite varies from bright orange to dull brown. Non-luminescent needle fibre calcite has been recorded by the authors in an Asbian calcrete from Cockermouth, Cumbria. This suggests that luminescence colour is not directly determined by the geochemical or physical conditions
cements (Fig. 13B). Needle-fibres always coat disso lution eroded surfaces and may show the same luminescence as the filling cement (Fig. 13B). Conse quently, we believe that needle-fibre calcite formed during the later stages of subaerial diagenesis (Table 1). 315
S. T. Solomon and G. M. Walkden that permit growth of acicular calcite. Needle-fibre luminescence also varies on a microscopic scale forming two common fabrics:
also may have formed by in situ overprinting of the original marine fabric.
( l ) Zoned needle-fibres : brightly luminescing needle fibres which occasionally show micrometre-sized non-luminescent cores. This fabric probably rep resents continued growth of needle-fibre calcite from pore water with changing geochemistry (e.g. Eh, pH, activator or quencher ion concentration). (2) Irregular patchy, or layered fabrics : such fabrics are caused by variations in luminescence which occur as irregular patches and layers on scales of ten to hundreds of micrometres (Fig. 14B). Contacts are typically irregular and gradational over a few tens of micrometres. These variations in luminescence could reflect activator ion com position changes of the inflowing cementing pore water. Alternatively, they may be caused by the presence of organic compounds in the calcrete profile which have locally influenced pore water geochemistry. Decaying plant roots could have created localized geochemical effects (Klappa, 1980) which may account for the brightly lumines cent zones often associated with calcified plant root tubules (Fig. 4B).
Origin of needle-fibre calcite morphology
Needle-fibre calcite has been recognized in transmit ted light by numerous authors and occurs in both Recent and ancient calcretes and also in soils. Explanations for the origin of needle-fibre habit fall into two categories; either organically or inorganically controlled growth. Workers preferring organic controls (Ward, 1970; Klappa, 1979, 1980 ; Calvet & Julia, 1983; Wright, 1984) have cited the relationship of needle-fibres to calcified plant root structures and have suggested that needle-fibre calcite formed as an indirect result of the activity of micro-organisms. In particular, some authors consider the biochemical reactions operating in the microenvironments around plant roots and fungal hyphae to be important. In this study, catho doluminescence has shown that needle-fibre calcite is abundant and widespread throughout the calcrete profile. This may imply a pervasive genetic process rather than one restricted to biological microenviron ments. Consequently, although biogenic and pedo genic processes may have contributed to the formation of needle-fibre calcite an inorganic process is favoured as the dominant crystal growth control. Buckley (1951) indicated that needle crystals may be produced by extreme supersaturation of the precipitating solution. James (1972) discussed inor ganic controls and suggested that 'needle-fibres' in a 'laminated crust profile' of Pliestocene reef limestones in North Barbados crystallized from rapidly evapora ting pore solutions, which may have quickly reached high degrees of supersaturation. Knox (1977) and Braithwaite (1983) also came to this conclusion. Harrison (1977) described similar needle-fibres which were mostly confined to root voids. Harrison (1977) suggested that supersaturation required for needle production might only have been satisfied in the plant root microenvironment. Needle-fibre calcite from the Quaternary succession of Barbados is particularly well illustrated by Esteban & Klappa (1983) and closely resembles the needle-fibre calcite seen in the Carbon iferous calcrete profile of this study. Tightly packed acicular meshes of calcite crystals (pseudomycelium) are a common nearsurface feature of recent soils (e.g. Fitzpatrick, 1971, fig. 83; Bal, 1975) usually occurring in carbonate-rich horizons approximately I m thick. Generally, pseudomycelium structures are believed to form rapidly (often within
Partially micritized marine allochems have been identified occurring within meshes of needle-fibre calcite (Fig. 14). This suggests that needle-fibre calcite
Fig. 14. Thin section, cathodoluminescence; needle-fibre calcite displaying irregular patchy luminescence. Partially subaerially micritized foraminifera occur within a mesh of needle-fibre calcite. Field of view: 1-4 mm.
316
Cathodoluminescence and calcrete diagenesis months) due to precipitation of leached calcium carbonate in the soil profile in response to evaporation. The needle-fibres of this Lower Carboniferous calcrete profile are closely similar to those cited above and, by comparison, are also believed to have been precipitated from pore solutions which became CaC03 supersaturated in response to near-surface evaporation. The relationship between needle-fibre calcite and subaerial micrite
Establishing the genetic relationship between needle fibre calcite and subaerial micrite has been aided by combining transmitted light and cathodolumines cence petrography. Transmitted light petrography has shown that needle-fibre calcite does not occur in optical continuity with the enclosing cement. This indicates that needle-fibres are not a relic or pseudo morph phase but were formed as we now see them. Needle-fibres always occur in close association with micrite, but because they form self-supporting pore bridging structures without the presence of micrite (Fig. 15), it seems likely that needle-fibre calcite cement and micrite formed independently. However, cathodoluminescence reveals that the micrite phase is not simply a fill of the acicular calcite meshes (Fig. 12B). Consequently, a more complex diagenetic relationship must be considered. Similar needle-fibre calcite/micrite associations have been described in transmitted light in studies of Quaternary calcretes by many authors (Harrison, 1977; H arrison & Steinen, 1978; Klappa, 1978, 1980). Both Knox (1977) and Calvet & Julia (1983) have
noted that in Recent calcretes needle-fibre calcite rapidly became unstable and recrystallized to micrite. Wright (1984) also considered this mode of degrada tion when describing needle-fibre calcite from a Dinantian calcrete in South Wales. Calvet & Julia (1983) noted that needle-fibre calcite 'reorganizes' (by degrading neomorphism) into crypto-microcrystalline anhedral crystals of low magnesium calcite. The needle-fibre calcite of this study may have undergone the same transformation. This is supported by slight crystal form differences from well-defined steep-sided needle-fibres (Fig. 12B) to relatively di ffuse rod-like crystals with blunt terminations (Fig. 13B). While crystal orientation may partly account for the apparent difference in form, it cannot explain the relatively diffuse appearance of many needle-fibres, particularly those associated with dense micritic fabrics. To conclude, the needle-fibre calcite cement may have partially neomorphosed to form micrite, the pore-bridging micritic fabrics reflecting the original arrangement of calcite needle-fibres.
Needle-fibre calcite, a common feature of ancient calcretes?
Additional examples of this needle-fibre calcite/ micrite relationship have been recorded by the authors in Asbian calcretes from both Derbyshire and Cum bria. It is predicted that the use of cathodolumines cence to examine calcretes will reveal that needle fibre calcite is a more common feature of ancient calcrete profiles than the geological literature currently indicates.
15. Thin section, plane polarized light; pore lining and bridging needle-fibre calcite occurring in close association with similar fabrics in the 'micrite' matrix of the calcrete profile. Field of view: 0· 72 mm.
Fig.
3 17
S. T. Solomon and G. M. Walkden distinguished primarily by its acicular cathodolumi nescence signature. Pore lining and bridging fabrics are characteristic, although smaller and more densely packed needle-fibres occur in association with micritic laminae and other subaerial micrite fabrics. This may indicate that second generation micrite contributed to the continued formation of earlier micrite fabrics ; possibly by the alteration of needle-fibre calcite (already discussed). Cathodoluminescence cement stratigraphy shows that second generation micrite overlies both non-luminescent subaerial vadose ce ments and the major dissolution hiatus. It also fills brecciation fractures demonstrating that it is the younger of the two micrite generations present in this calcrete (Table 1). Where acicular cathodolumines cence signature is poorly defined and dissolution textures and other identifiable cements are absent, the distinction of these two micrite generations is not possible. To conclude, during the formation of this ancient calcrete profile subaerial micritization appears to have been a continual process, initially altering the original marine sediment, but later also affecting needle-fibre calcite cements which partially occluded porosity.
Micrite stratigraphy
Two generations of non-marine micrite can be distinguished primarily on the basis of cathodolumi nescence signature. However, their different relative ages can be demonstrated only when dissolution textures and a full spar cement stratigraphy can be identified. Under cathodoluminescence first generation mi crite appears diffuse with no distinguishable crystal form and a dull to intermediate orange luminescence. In transmitted light it has a grey-brown colour and displays all the micrite fabrics described, except pore lining and bridging textures. Dissolution textures and brecciation fractures cross-cut this generation of micrite (Fig. 16 and Table 1). Second generation micrite (Table I , column 3) is
DIAG EN E TIC HISTOR Y
The combined use of transmitted light and cathodo luminescence microscopy has revealed the interrela tionships of the major diagenetic processes operative during subaerial emergence. This has enabled the diagenetic history of the calcrete profile to be reconstructed (Table 1 ). Subaerial alteration of the marine sediment was multistaged, involving the simultaneous action of many diagenetic processes during a single phase of emergence in the Late Asbian. Subaerial diagenesis was probably halted by the onset of the next phase of cyclic shelf sedimentation and the influx of marine pore water into the calcrete profile.
CONCLUSIONS
( ! ) A laterally extensive ancient calcrete profile has been identified in the Late Asbian shallow marine shelf limestones of the Llangollen area, North Wales. (2) In thin section, lithologies of the calcrete profile have a strongly altered and variable fabric, possessing micritic textures and secondary pore types of subaerial diagenetic origin. (3) Using cathodoluminescence to determine rela-
Fig. l6. Thin section, cathodoluminescence; two generations
of subaerial micrite distinguished in cathodoluminescence. First generation micrite (m l ) is offset by a subaerial brecciation fracture together with a brachiopod fragment (BR) and non-luminescent cements ( l -as Fig. 5). Second generation 'micrite'/needle-fibre calcite (m2) and meteoric phreatic cements infill the brecciation fracture. Field of view: 2·05 mm.
318
Cathodoluminescence and calcrete diagenesis precipitation of meteoric phreatic cements. (d) Needle-fibre calcite is considered to be a cement precipitate which may have almost completely recrystallized to micrite during the late stages of subaerial diagenesis. (e) In many parts of the calcrete profile two genera tions of non-marine micrite can be distinguished using cathodoluminescence. This two-part 'mi crite stratigraphy' requires the presence of either the full spar cement stratigraphy or dissolution textures and brecciation fractures to prove the different ages of subaerial micrite.
tive ages, the calcite cements of the calcrete profile have been divided into a three-part cement strati graphy : (I ) non-luminescent cements (subaerial vadose); (2) first (major) bright orange luminescent cement (subaerial vadose); (3) non-luminescent cements with brightly lumi nescent pitted subzones and final pore filling medium luminescent cements with smooth subzones (meteoric phreatic). (a) Precipitation of the non-luminescent subaerial vadose cements was concomitant with subaerial brecciation, micritization and the calcification of plant root tubules. (b) Subaerial vadose cements have a restricted verti cal stratigraphic range ; they occur both in the laminated calcareous crust of the calcrete profile and to depths of l m in the immediately underlying lithologies, but are absent from the rest of the Late Asbian succession.
(6) Combined transmitted light and cathodolumi nescence petrography show that the diagenetic history of the calcrete profile was multistaged, with many subaerial diagenetic processes acting simultaneously during a single phase of emergence.
AC KNO WLEDGMENTS
We would like to thank Drs V. P. Wright, C. J. R. Braithwaite and W. C. Ward for critically reading this manuscript and providing helpful suggestions.
(4) Textures, visible only with cathodolumines cence, are characteristic of recurring fabric dissolu tion. Dissolution took place during, and immediately after, precipitation of non-luminescent subaerial va dose cements.
REFERENCES
(a) The texture seen with cathodoluminescence is controlled by the microstructure/crystal morphol ogy of the partially dissolved substrate. (b) Dissolution textures are generally confined to the walls of larger pores and brecciation fractures which probably acted as fluid pathways in the calcrete during early subaerial diagenesis.
BAL, L. ( 1 975) Carbonate in soil : a theoretical consideration on, and proposal for its fabric analysis. Crystic, calcic and fibrous plasmic fabric. Neth. J. agric. Sci. 23, 1 8-35. BRAITHWAITE, C.J.R. ( 1 975) Petrology of palaeosols and other terrestrial sediments on Aldabra, western Indian Ocean. Phil. Trans. R. Soc. B, 273, 1-32. BRAITHWAITE, C . J . R . ( 1 983) Calcrete and other soils in Quaternary limestones : structures, processes and applica tions. J. geol. Soc. London, 140, 351-363. BuCKLEY, H.E. ( 1 9 5 1 ) Crystal Growth. Wiley, New York. CALVET, F. ( 1 982) Constructive micrite envelope developed in vadose continental environment in Pleistocene eoliantes of Mallorca (Spain). A cta geol. Hisp. 17, 1 69-1 78. CALVET, F. & JULIA, R. ( 1 983) Pisoids in the caliche profiles of Tarragona (N.E. Spain). In: Coated Grains (Ed. by T. M. Peryt), pp. 456-473. Springer-Verlag, Berlin. DICKSON, J . A. D . ( 1 965) A modified staining technique for carbonates in thin section. Nature, 205, 587. EsTEBAN, M . ( 1 974) Caliche textures and microcodium. Suppl. Bull. Soc. geol. ita/. 92, 1 05-1 25. EsTEBAN, M. & KLAPPA, C . F. ( 1 983) Subaerial exposure. In: Carbonate Depositional Environments (Ed. by P. A. Scholle, D. G. Bebout and C. H. Moore). Mem. Am. Ass. Petrol.
(5) Much of the original marine matrix of litholo gies in the calcrete profile has been subaerially altered to form micrite. Under cathodoluminescence this micrite often has an acicular appearance consisting of meshes of calcite needle-fibres. (a) The luminescence of needle-fibres is highly vari able suggesting that activator ion concentrations are not controlled by the geochemical or physical conditions required for the growth of acicular calcite. (b) Needle-fibre calcite was probably formed in response to localized supersaturation of meteoric pore fluids caused by periods of near-surface evaporation. (c) Cement stratigraphy demonstrates that needle fibre calcite formed after early subaerial vadose cementation and dissolution, but before the
Geo/. 33.
FITZPATRICK, E.A. ( 1 97 1 ) Pedology : a Systematic Approach to Soil Science. Oliver & Boyd, Edinburgh. FoLK, R.L. ( 1959) Practical petrographic classification of limestones. Bull. Am. Ass. Petrol. Geo/. 43, 1-8. GRAY, D.I. ( 1 98 1 ) Lower Carboniferous shelf pa/aeoenviron-
319
·
S. T. Solomon and G. M. Walkden ments in North Wales. Unpublished Ph.D. Thesis. Univer sity of Newcastle on Tyne. HARRISON, R.S. ( 1 977) Caliche profiles : indicators of near surface subaerial diagenesis, Barbados, West Indies. Bull. Can. Petrol. Geol. 25, 1 23-1 73. HARRISON, R.S. & STEINEN, R.P. ( 1 978) Subaerial crusts, caliche profiles, and breccia horizons : comparison of some Holocene and Mississippian exposure surfaces, Barbados and Kentucky. Bull. geol. Soc. Am. 89, 385-396. JAMES, N . P. ( 1 972) Holocene and Pleistocene calcareous crust (caliche) profiles : criteria for subaerial exposure. J. sedim. Petrol. 42, 8 1 7-836. KLAPPA, C. F. ( 1978) Morphology, composition and genesis of Quaternary calcretes from the western Mediterranean : a petrographic approach. Unpublished Ph.D. Thesis, Univer
sity of Liverpool. KLAPPA, C.F. ( 1 979) Calcified filaments in Quaternary calcretes : organo-mineral interactions in the subaerial vadose environment. J. sedim. Petrol. 49, 955-968. KLAPPA, C.F. ( 1 980) Rhizoliths in terrestrial carbonates : classification, recognition, genesis and significance. Sedi mentology, 27, 6 1 3-629. KNOX, G.J. ( 1 977) Caliche profile formation, Saldanha Bay (South Africa). Sedimentology, 24, 657-674. MEYERS, W.J. (1974) Carbonate cement stratigraphy of the Lake Valley Formation (Mississippian) Sacramento Mountains, New Mexico. J. sedim. Petrol. 44, 837-86 1 . MULTER, H . G . & HOFFMEISTER, J .E. ( 1968) Subaerial laminated crusts of the Florida Keys. Bull. geol. Soc. Am. 79, 1 83-192. SANDBERG, P.A. ( 1 975) Bryozoan diagenesis : bearing on the
(Manuscript received 2 January
nature of the original skeleton of rugose corals. J. 587-606. SOMERVILLE, I.D. ( 1977) The sedimentology and stratigraphy Palaeontol. 49,
of the Dinantian limestones in the Llangollen area and East of the Clwydian range, North Wales. Unpublished Ph.D.
Thesis, Queen's University of Belfast. SOMERVILLE, I. D. ( 1 979) Minor sedimentary cyclicity in Late Asbian (Upper Dl) limestones in the Llangollen district of North Wales. Proc. Yorks. geol. Soc. 42, 3 1 7-34 1 . WALKDEN, G . M . ( 1974) Palaeokarstic surfaces i n Upper Visean (Carboniferous) limestones of The Derbyshire Block, England. J. sedim. Petrol. 44, 1 232-1 247. WALKDEN, G . M . & BERRY, J . R. ( 1 984) Syntaxial over growths in muddy crinoidal limestones : cathodolumines cence sheds new light on an old problem. Sedimentology, 31, 251 -268. WALKDEN, G . M . & DAVIES, J. ( 1 983) Polyphase erosion of subaerial omission surfaces in the Late Dinantian of Anglesey, North Wales. Sedimentology, 30, 861 -878. WALLS, R.A., BURLEIGH HARRIS, E. & N UNAN, W.E. ( 1 975) Calcareous crust (caliche) profiles and early subaerial exposure of Carboniferous carbonates, northeastern Ken tucky. Sedimentology, 22, 4 1 7-440. WARD, W.C. ( 1 970) Diagenesis of Quaternary eo/iantes of NE Quintana Roo, Mexico. Ph.D. Thesis, Rice University, Houston, Texas. WATTS, N .L. ( 1 978) Displacive calcite : evidence from recent and ancient calcretes. Geology, 6, 699-703. WRIGHT, V.P. (1 983) A rendzina from the Lower Carboni ferous of South Wales. Sedimentology, 30, 1 59-179. WRIGHT, V.P. (1984) The significance of needle-fibre calcite in a Lower Carboniferous paleosol. Geol. J. 19, 23-32.
1985 ; revision received 2 May 1985)
320
CALCRETES AND PALUSTRINE CARBONATES
Palustrine carbonates (marsh carbonates) are wide
nates.
spread in continental deposits, particularly the later
limestones have more negative values, reflecting
Stable
isotope
data from
the
palustrine
Mesozoic and Tertiary of Europe. Platt, in describing
extensive pedogenic modification and the input of
material from the Cameros Basin in Spain, discusses
light 'meteoric' oxygen and light organic carbon
the problems of distinguishing calcretes from pedo
isotopes.
genically modified lacustrine and palustrine carbo-
pebbles'.
These
limestones
also
contain
'black
Fig. 16. Palustrine limestone, Oligocene, Bembridge Limestone, Isle of Wight, England. The original sediment, a marsh/pond bioclastic wackestone, has been affected extensively by pedogenic processes. In (A) calcrete nodules occur within the sediment, which also contains peloids and coated grains. A large dissolution void formed and was filled with darkened intraclasts. In (B) a laminar calcrete is succeeded by sediment containing peloids, coated grains (some with gastropod nuclei, others lime mud), and intraclasts of laminated crust, perhaps derived from rhizobrecciation. An elongate cavity near the top is filled by internal sediment and then calcite spar. Original sediment, with moulds of gastropods and bivalves, occurs at the very bottom and top of the specimen.
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
321
Reprinted from Sedimentology (1989) 36 665-684
Lacustrine carbonates and pedogenesis: sedimentology and origin of palustrine deposits from the Early Cretaceous Rupelo Formation, W Cameros Basin, N Spain NI G E L H. P L A T T Geologisches lnstitut, Universitiit Bern, Baltzerstrasse 1, CH-3012 Bern, Switzerland
ABSTRACT
The Berriasian Rupelo Formation of the W Cameros Basin consists of a 2-200 m thickness of marginal and open lacustrine carbonate and associated deposits. Open lacustrine facies contain a non-marine biota with abundant charophytes (both stems and gyrogonites), ostracods, gastropods and rare vertebrates. Carbonate production was mainly biogenic. The associated marginal lacustrine ('palustrine') facies show strong indications of subaerial exposure and exhibit a wide variety of pedogenic fabrics. Silicified evaporites found near to the top of the sequence reflect a short hypersaline phase in the lake history. The succession was laid down in a low gradient, shallow lake complex characterized by wide fluctuations of the shoreline. Carbon and o xygen stable isotope analyses from the carbonates show non-marine values with ranges of b13C from -7 to -l l%0 and <5180 from -3 to -7·5%0. Differences in the isotopic composition of open lacustrine carbonates are consistent with sedimentary evidence of variation in organic productivity within the lake. Analyses from the entire sample suite plot on a linear trend; isotopic compositions become lighter with increasing evidence of pedogenic modification. This suggests progressive vadose zone diagenesis and influence of meteoric waters rich in soil-derived C02• The stable isotope data thus support evidence from petrography and facies relations that 'palustrine' limestones form through pedogenic modification of lake carbonates.
INTRODUCTION
Studies of modern and ancient lacustrine carbonates have emphasized sedimentation in relatively deep, stratified lakes, commonly those with high-gradient, temporally stable, 'marl-bench' type margins (e.g. Murphy & Wilkinson, 1980; Brown & Wilkinson, 1981), or those with low-gradient, wave-influenced margins, as represented in parts of the Green River Formation of the W USA (Williams & Picard, 1974). In contrast, carbonate sedimentation in low-gradient, low-energy, unstratified lakes has received relatively little attention outside Southern Europe, where spec tacular sequences of shallow lake carbonates devel oped under warm, arid to semi-arid climatic conditions during the Mesozoic and Tertiary. Freytet & Plaziat ( 1982) and other authors (e.g. Cabrera, Colombo & Robles, 1985) have described ephemeral carbonate lake margin facies as 'palustrine'. This term is broadly Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
equivalent to 'paludal' ( = swampy, marshy; e.g. Treese & Wilkinson, 1982) and is the approximate lacustrine analogue for 'peritidal' or 'paralic'. This paper summarizes the results of a detailed investigation of an extensive lacustrine/palustrine carbonate sequence from the Berriasian (earliest Cretaceous) Rupelo Formation of the W Cameros Basin in Northern Spain (Platt, 1 986). Many docu mented 'palustrine' sequences (e.g. Glass &. Wilkin son, 1980; Freytet & Plaziat, 1982; Wells, 1983; Cabrera et a/., 1985) are from foreland basin-type tectonic environments. In contrast, the Rupelo For mation is an example from an extensional rift setting. Deposition post-dated Late Jurassic fault reactivation but preceded a sharp acceleration of subsidence at around the Valanginian. The Rupelo Formation may thus provide a useful model for shallow, ephemeral 323
Nigel H. Platt
the NNE of Madrid (Fig. 1). The Late Jurassic-Early Cretaceous was a time of rapid sedimentation over much of the Iberian Peninsula. Rifting associated with the opening of the Bay of Biscay and the North Atlantic caused subsidence rates to increase dramati cally in several major Mesozoic rift basins, for example in the Cameros (Salomon, 1982, 1983) and S Vasco Cantabrian Basins (Sbeta, 1985) of N Spain and in the Lusitanian Basin of W Portugal (Leinfelder, 1987). Upper Jurassic-Lower Cretaceous continental se quences up to 5 km thick were deposited above Late Jurassic unconformities in each of these basins. In the W Cameros Basin this succession is over 2 km in thickness (Platt, 1986). The lower part of the sequence, the Tierra de Lara Group, consists of two formations (Fig. 2): the predominantly siliciclastic Senora de Brezales Formation, which is up to 75 m thick, of ?Kimmeridgian to Berriasian age, and lies unconformably upon a 500 m thick sequence of marine Jurassic carbonates; and the overlying Rupelo For mation, which comprises a series of lacustrine facies carbonates traceable over more than 2500 km 2, and up to 200 m in thickness. Presence of the charophyte Globator incrassatus indicates a Berriasian age (P. 0. Mojon, pers. commun., 1985). The remaining 2 km of Lower Cretaceous sedimentary rocks consist of fluvial conglomerates, sandstones and mudstones, with only thin oncoidal carbonates near the base.
lake carbonate deposition in the isolated sectors of developing rift systems. Modern analogues with comparable tectonic settings may include lakes devel oped in isolated intermontane basins within the modern extensional province of the semi-arid Basin and Range in the United States (Fouch & Dean, 1982; Dean & Fouch, 1983). The ephemeral nature of the palustrine environment is evident from the abundant indications of subaerial exposure present in palustrine limestones. Esteban & Klappa ( 1983) noted that palustrine limestones com monly show similar fabrics to those of calcisols. On the basis of petrographic evidence and facies relation ships, Freytet (1973) suggested that palustrine lime stones formed through pedogenic modification of lacustrine carbonate muds, although geochemical evidence to support this hypothesis has not been presented to date. Sedimentological analysis of the Rupelo Formation presented here is complemented by a stable isotope study designed to investigate the origin of palustrine limestones.
GEOLOGICAL SETTING
The Cameros Basin lies in the Northern Spanish provinces of Burgos, Soria and La Rioja, 200 km to
oo
10° w
I VASCO-CANTABRIAN
CAMEROS
�
BASIN 40° N Madrid
' : ·' · �· '
40° N
•
.
(J
...
N
D 0
100km
oo
10° w
Fig. l. Map of Iberia showing location of the Cameros Basin and of other basins mentioned in text.
324
Sedimentology and origin of palustrine deposits - � - --; - ------ ---L-- ---- - - - ---0
0
0
0
marls and
littoral
la c u strin e
oncoidal limestones
ZURRAMUJERES MEMBER
z 0
:
_ _ _-__ ______ _ ___ _ _
i= distal
marls
alluvial
< :::;: a: 0 LJ..
SAN MARTIN MEMBER
0
Cf)
0 a. a: ::> 0 0 a: w a. (!'
r o o t mats
I I I I I I I I I
WI I I I �� .:�3:¥411 P1 II I
pedogonically modified
palustrine
limestones
RIO CABRERA MEMBER
I I I I I I I I I
a. ::> 0 a: (!'
red marls
z 0
evaporitic;:
chert
i= <
:::;:
yellow/vuggy limestones
charophyte·ostrac<:'d
marls
mudstones & wackestones rarer grainstones
II II ---� l
�I I
I I I I 111 II
0
� .
.
MAMBRILLAS DE LARA
0 ...J w a.. ::l a:
red marls
palustrine
<( a: <( ...J
MEMBER
lacustrine
root mats
LADERA MEMBER
pedogenically modified
c:
.. 'iii ..
·;:::
Q;
m
w
0
limestones
Il l� I I�--
.. :.·..
open
a: 0 LJ..
Cf) ::> 0 w
u <( 1-
w
a: u a: w
!=
0 ...J
d i st a l alluvial
red marls
LAS VIN AS MEMBER
.. '+. ---------------+--------r----------------+------� wen z 0 w 0
red sands
< ...J i= < a: 0 w :::;: lZ CX: a: wmo u.. en
;1i
sand flat
/wadi
-?--u
en
ll:Cf) W<( a. a: 0.::> ::>...,
Fig. 2. Stratigraphy of part of the Upper 1urassic-Lower Cretaceous continental sequence.
THE BERRIASIAN RUPELO
lacustrine complex. The. Rupelo Formation is domi
FORMATION
nated by shallow water limestones rich in charophytes (including calcified stems) and displaying abundant
The Rupelo Formation may be lithostratigraphically
evidence of subaerial exposure and pedogenesis. Distal
and biostratigraphically correlated with other latest
alluvial sediments and minor evaporites are also
Jurassic-earliest Cretaceous ('Purbeckian') marginal
intercalated. Depositional facies represented in the
marine carbonate sequences of NW Europe, for example in S England (West,
Rupelo Formation can be subdivided into five main
1 975; Francis, 1986)
associations:
and in the Jura' Mountains of France (Strasser & Davaud,
1 982) and Switzerland (Strasser, 1 987, 1 988).
( 1 ) open lacustrine carbonates (2) reworked lacustrine carbonates
However, the Rupelo Formation is apparently unique among Purbeckian successions because it shows no
(3)
indications of marine influence. Sedimentological and
(4) distal alluvial
faunal evidence indicates deposition in a very shallow
(5) evaporitic
325
marginal lacustrine-'palustrine' carbonates
Nigel H. Platt
Facies association 1: open lacustrine carbonates
This association comprises well-bedded 30-1 00 em mid- to dark-grey mudstones and wackestones in terbedded with grey marls. The limestones (Fig. 3A) contain gastropods 1-3 em in size, some with partial geopeta1 fillings of internal sediment, 0·5-2 mm Cyprideae ostracods, and many charophyte stems and gyrogonites. The marls are especially rich in charo phytes. Vertebrate bones and bone fragments up to 30 em in size occur in the marls and mudstones. Vertebrate footprints 30-50 em across were also observed. Polygonal desiccation cracks occur on the top surface of some beds, and fenestral cavities 1 2 em across showing geopetal fills o f crystal silt are also locally present. Lamination is virtually absent; only in the darkest of the limestone beds does a diffuse and disturbed lamination occur (Fig. 3B). Rarely, however, some of the interbedded grey marls display mm-scale fine graded silt laminae rich in 1 0-20 J.Lm subangular quartz grains. These occur only in the NW of the basin, where rare 1 0-50 em beds of limestone con glomerate and cross-bedded sandstone with erosive bases were also found. Elsewhere, siliciclastic material is virtually absent. Interpretation
The biota of these facies consists entirely of 'freshwa ter' elements typical of lacustrine environments with
salinities of less than 5%0 (Sbeta, 1976); Cyprideae ostracods, charophytes, and gastropods. The predom inance of mudstone textures suggests low-energy sedimentation. Carbonate production was mostly biogenic; carbonate mud was probably derived chiefly from degraded calcified charophyte debris. Modern charophytes commonly live in warm, shallow, alkaline waters, in depths of less than 10 m, or down to a maximum of 1 5-20 m (Cohen & Thouin, 1987). Charophyte stems are delicate structures, and the preservation of some intact stems provides further evidence for periods of low-energy deposition without significant trjinsport. Spherical reproductive gyrogon ites are more easily transported, even by relatively gentle currents. This may explain their paucity in beds where stems are abundant and their concentration in a few darker carbonate mudstone beds where stems are absent. The footprints and polygonal cracks indicate shal low depths and occasional emergence, but their preservation rules out pedogenic reworking. Indeed, the limestones show little evidence of pedogenic modification, although the fenestral cavities may be solution vugs, also suggesting occasional emergence. The general absence of lamination in these carbon ates implies bioturbation of the sediments and low salinity, oxygenated bottom waters, and indicates that stratification of lake waters was not permanent. The scarce diffuse lamination noted was commonly dis turbed. The fine clastic laminae are similar to laminae
Fig. 3. Open lacustrine carbonates. (A) Wackestone with abundant charophyte stems in transverse section; gastropods assoctated. Scale bar= I mm. Interpretation : open lacustrine biogenic carbonates. (B) Fine, mm-scale graded silt laminae defined by terrigenous material. Ostracod test (arrowed) probably reworked. Scale bar= I mm. Interpretation : deposition from gentle, perhaps r iver-generated, density currents.
326
Sedimentology and origin of palustrine deposits
reported from modern West African lakes by Talbot (I 984) and interpreted as the deposits of small-scale lacustrine density currents. Density currents may have been river-generated ; the occurrence of clastic lami nae associated with the thin clastic units exclusively in the NW of the basin points to terrigenous input into the lake from that direction. The preservation of lamination indicates periods of reduced bioturbation, either as a result of increased salinity or 02 deficiency at the lake floor during times of increased organic productivity. Facies association 2: reworked lacustrine carbonates
Associated with, or laterally equivalent to, the 'open lacustrine' carbonates, are an association of wacke stones, grainstones, and intraformational conglomer ates. Wackestones contain angular intraclasts. Grading is locally developed above sharp, erosive bases. The intraclasts are from 0·5-2 mm in size and most are of pale beige micrite. Many intraclasts contain mm sized peloidal structures, which are commonly associ ated with fine, curving sparry cracks. Dark 0·5-3 mm intraclasts are also locally abundant. These are angular and their colour ranges from black to mid-grey, commonly within a single pebble. Grainstones are present in 1 m beds and consist of angular mm-sized charophyte mudstone clasts ce mented with coarse spar (Fig. 4A). The clasts show no evidence of pedogenic modification. Cathodolumi nescence reveals the presence of two principal non ferroan calcite cement generations: a first, non luminescent phase, and, rarely, a later brightly luminescent and locally zoned cement generation (Fig. 4B). Intraformational limestone conglomerates are com monly developed at prominent bedding planes. They are difficult to recognize in the field because there is little lithological contrast between the irregular, weakly-rounded, grey micrite clasts 0·5-2 em across and the mid-grey marly matrix. Rare dark clasts also occur in these conglomerates, but here they are generally larger, up to 3 em across. The conglomerates drape sharp, erosive bases which are locally scoured to form shallow channels 2-3 m wide and up to 30 em deep. Interpretation
Each of these carbonate types contains reworked clasts.
The intraclasts present in the wackestones show evidence of pedogenic modification such as the mm sized peloids (these spheroidal structures resemble pedogenic nodules or 'glaebules'; Brewer, 1964), and the associated curving sparry cracks are similar to 'circumgranular cracks' described from pedogenic carbonates (e.g. by Swineford, Leonard & Frye, 1958, and Ward, 1975). The intraclasts were presumably derived either from brecciated marginal lake deposits or from immediately underlying beds. Their concen tration above sharp erosive surfaces suggests that they were washed in during storms. The dark intraclasts can be compared to 'black pebbles' ('cailloux noirs' of the French usage; Cayeux, 1935), common in carbonates from aeolian (Ward, Folk & Wilson, 1970), lacustrine (Freytet, 1973), and pedogenic (Esteban & Klappa, 1983) settings. They probably document input of stabilized organic matter from a vegetated hinterland (Strasser & Davaud, 1983; Strasser, 1984). Shinn & Lidz (I 987) related formation of black pebbles at modern subaerial exposure surfaces to burning of organic matter in forest fires. Shinn & Lidz warned against possible confusion with subtidally-blackened clasts, but the gradation of blackening and angular nature of the larger pebbles observed here are features cited by Shinn & Lidz ( 1987) as characteristic of subaerial fire blackening. The presence of smaller (0· 5-1 mm) dark intraclasts in the wackestones suggests abrasion during transport. The intraclasts present in the grainstones are of open lacustrine facies. They contain charophyte stems and show no evidence of pedogenic modification. However, the erosion of lithoclasts suggests early lithification. Winnowing may have been performed subaqueously within the lake, for example by traction currents or by small storm-generated waves on exposed shorelines or offshore bars. The dominance of early non-ferroan, non-lumines cent blocky calcite cement (probable low Mn, low Fe compositions) points to predominantly oxidising con ditions during cementation (Scoffin, 1987). The rare later development of zoned, luminescent cement suggests increased incorporation of Mn, possibly associated with the onset of reducing conditions. This evolution is consistent with cementation in the meteoric phreatic environment during lake regression and subsequent transgression (cf. Scoffin, 1987). The intraformational conglomerates also indicate internal reworking of open lacustrine facies. The only externally-derived clasts are the black pebbles. The channelling probably records periods of lower lake 327
Nigel H. Platt
Fig. 4. Reworked lacustrine carbonate. (A) Grainstone consisting of mm-sized poorly-sorted (rounded 1-3 mm; irregular 530 mm) clasts of charophyte mudstone set in coarse sparry cement. Scale bar= I mm . Interpretation : high-energy, possibly storm reworking. (B) Grainstone under luminescence. Clasts of charophyte mudstone show dull luminescence. Most spar cement is non-luminescent, but with wea k zoning; rare later generation of brightly luminescent cement filling remaining void space. Scale bar= I mm. Interpretation: meteoric diagenesis.
level. However, the absence of pedogenic features indicates that emergence was not prolonged. Rapid, transient shoreline retreat would have exposed open lacustrine sediments, which were then also locally eroded and reworked by streams of internal drainage plying the wide exposed lake flats.
with 0·5-1 mm subangular or subrounded red intra clasts. Laminar horizons contain numerous predominantly horizontal, cylindrical I mm diameter, spar-filled cavities, which are surrounded by a lining of weakly concentric laminated beige carbonate (Fig. 5B). Fine septae 0· 1 mm across locally protrude from the cavity walls and interlink to produce intricate fabrics. The porous structure occurs rarely in discrete patches or as isolated tubules, but is more commonly developed in irregularly laminated, bedded units several em in thickness which are laterally continuous for 0· 1 -5 m or more. Limestones with geopetally-filled cavities contain fenestrae which are 2-20 mm in length and 1-5 mm in height, are elongate parallel to bedding, and show planar cavity floors but smoothly-rounded roofs (Fig. SC). The cavity-fillings consist of single or multiple generations of fine calcite microspar (grain size approximately I 0 11m), and an upper, later coarse blocky void-filling non-ferroan calcite cement (grain size 100-200 11m). Rounded peloids 0·5-1 mm across occur 'floating' in the cavities and locally show a coarsening-up trend in grain-size within a single cavity. Cavities commonly interlink to form complex networks with perched geopetal fills of internal sediment and crystal silt (Fig. SD) or may be entirely
Facies association 3: marginal lacustrinefpalustrine carbonates
Microbrecciatedjpeloidallimestones do not generally contain faunal remains, although rare, poorly pre served charophyte stems occur within the matrix. Mottling is common, mostly from purple-red to grey, although yellow mottling also occurs. Many of the limestones are brecciated, and in thin section cavities and cracks are abundant. These are now spar-filled and display a variety of morphologies. Fine sparry cracks may be angular, or circumgranular structures developed around 0·5-2 mm rounded peloids (Fig. SA), which rarely show a weak and irregularly concentric structure. Smaller, 0·05-0· 1 mm pellets are also common. These are also partially surrounded by fine spar-filled cracks. Angular, black to mid-grey intraclasts from 2 to 5 mm (and rarely up to 30 mm) in size also occur, although they are not abundant. These locally show grey-brown rims and are associated
328
Sedimentology and origin of palustrine deposits
filled with fine microspar. This hinders their recogni tion from the host micrite. Other sparry cavities include unlined 1 mm diameter tubular voids and rarer angular 1-3 mm cavities filled with brown spar. Centimetre-sized carbonate nodules are abundant in some limestone beds. Larger, vertical nodular structures 5-15 em in diameter and up to 50 em long may be associated with rough vertical fracturing, giving rise to 'prismatic fabric'. Reticulate fabrics also occur. These show a combination of vertical nodular structures with horizontal laminar bedded units containing rom-scale cylindrical structures (see above). Interbedded with the limestones are several 0·51 m rubbly red horizons, consisting of mottled, impure carbonates containing many subangular red clasts 210 mm in diameter. Strings of pale grey carbonate, 1 mm in diameter and 5-30 mm in length, are also common, forming individual, concentrically micrite coated, cylindrical tubules. The red material drapes irregular basal contacts and locally passes down vertically along joints. Underlying beds have micro brecciated peloidal fabrics; overlying beds have rare red clasts 1-5 mm in size. Green marl units 1-10 em in thickness occur at a few localities. Samples taken from Rupelo Formation green marls were analysed by Deconinck & Strasser (1987) and found to be composed chiefly of smectite, with minor quantities of kaolinite. Interpretation
Evidence of pedogenesis was noted by Mensink & Schudack ( 1982) and Salomon ( 1984). Mottling, desiccation brecciation, glaebule development, micro karst cavities and root structures are all common features in the Rupelo Formation. Microbrecciated/peloidal limestones show mot tling, which is characteristic of sediment subjected to repeated wetting and drying and may develop during pedogenesis as a result of fluctuating Eh-pH condi tions or through redistribution of iron oxide/hydroxide particles (Buurman, 1980). The rom-scale peloidal structures (Fig. 5A) are 'glaebules' (Brewer, 1964), nodules formed during pedogenesis through concen tration of soil 'plasma' (colloidal or soluble soil material ; Brewer, 1964). Desiccation shrinkage and rotation during emergent periods then results in circumgranular cracking and formation of rounded soil grains or 'peds' (individual soil aggregates ; Brewer, 1964). The smaller 0·05 mm pellets are similar to micro-
pellets recorded from paleosols by Esteban & Klappa (1983) and Wright (1983) and interpreted as calcified faecal pellets. Their presence suggests the existence of a burrowing infauna. The dark intraclasts are similar to the 'black pebbles' reported above from the wackestones and grainstones. However, the larger size of the black pebbles in the marginal lacustrine carbonates proba bly reflects closer proximity to the vegetated hinter land. The common alteration, especially of the smaller pebbles, to a brown-red colour is characteristic of this facies and apparently reflects the effects of pedogenesis after deposition. This may explain the scarcity of small black pebbles in this facies association, as possibly only the largest ones survived alteration. The fabric of laminar horizons (Fig. 5B) is similar to that previously described as 'laminar calcrete' (James, 1972), 'laminar crusts' (Multer & Hoffmeister, 1968) or 'croute zonaire' in the French usage (Freytet & Plaziat, 1982). However, in this case, the abundant cylindrical spar-filled cavities are associated with complex networks of fine septal structures. This alveolar texture (Esteban, 1974) is a root-associated fabric recently attributed to calcification of fungal hyphae (Wright, 1986; Phillips & Self, 1987). The cavities are interpreted as root voids. They may have been created by the horizontal root networks of subaerial or shaliow-water marsh plants, or alterna tively by charophyte holdfasts. When they are abun dant, the bedding-parallel orientation of the cavities defines a crudely laminated fabric. Somewhat similar 'root mats' were described from the Plio-Pleistocene fluvial and marginal lacustrine deposits of Kenya by Mount & Cohen ( 1984). Laminar biogenic horizons with this morphology (rhizolite laminar calcretes) have been recognized from carbonate paleosols of a variety of stratigraphic ages (Wright, Platt & Wimble don, 1988) and provide evidence for periodic coloni zation by plants. The limestones with geopetal cavities (Figs 5C & 5D) are identical to carbonates with microkarst cavities described by Freytet & Pla:i:iat ( 1982). The cavities may have been originally produced by root systems, but subsequent solutional enlargement is suggested by the �ounded, embayed cavity margins. The cavities were then partially filled with internal sediment and crystal silt, probably during successive periods of desiccation and emergence prior to spar cementation. Internal sediment may have been de rived from the cavity walls, but the coarsening upwards size distribution of the peloids suggests that they were washed in during progressive widening. 329
Nigel H. Platt
Fig. 5. Marginal lacustrine carbonates. (A) Rounded glaebules (example marked 'g') 0· 5-2 mm in diameter. Partial and complete development of spar-filled circumgranular cracks (example arrowed). Scale bar= I mm. Interpretation : grain shrinkage and rotation on desiccation. (B) Laminar fabric showing tubular, spar-filled voids with concentric-laminated micrite lining. Scale bar= I mm. Interpretation : root mat (laminar rhizolite horizon) consisting of many small rhizoliths (Wright et a/., 1989) . (C) Elongate cavity with partial geopetal fill of fine crystal silt, later spar cement. C ircumgranular cracks in host sediment. Scale bar= I mm . Interpretation: 'pseudo-microkarst' (Frey<et & Plaziat, 1982). Cavity possibly formed by root penetration; rounded margins suggest later solutional expansion . (D) Complex network of interlinked cavities containing perched geopetal fills of crystal silt. Note transverse section of charophyte stem (arrowed) . Later sparry cement. Scale bar= I mm. Interpretation : brecciation as a result of strong development of 'pseudo-microkarst' (Freytet & Plaziat, 1982).
cavities may be 'bird's-eyes', small vugs formed through shrinkage and expansion or gas bubble formation and preserved as a result of early lithifica tion (Shinn, 1968, 1983), or may represent now dissolved authigenic minerals. The nodules present in the nodular limestones are
The crystal silt commonly postdates the peloids, favouring a later origin through precipitation, proba bly from vadose diagenetic fluids ('vadose silt'; Dunham, 1969). The unlined cavities are probably small-scale invertebrate burrows. The rarer, brown-spar-filled 330
Sedimentology and origin of palustrine deposits
Purbeck to the weathering of soil material and to the erosion of ancient massifs (Sladen, 1983), with the less easily transported kaolinite deposited mostly in areas proximal to sources of detrital input. In the French and Swiss Jura, Deconinck, Strasser & Debrabant (1988) noted that illite occurred in those areas nearer to marine influences, and suggested that illite was formed by conversion from detrital smectite as a result of repeated wetting by K-rich marine waters and subsequent drying in a hypersaline environment. In the case of the Rupelo Formation, the minor quantities of kaolinite present in the green marls are consistent with evidence from the carbonates of generally low detrital input. The dominance of smectite in the Rupelo Formation green marls reflects little conver sion to illite, and might suggest low salinity environ ments with no significant influx of K-rich waters. This, in turn, would be consistent with the absence of marine indicators in the Rupelo Formation.
interpreted as pedogenic structures. Cm-sized caliche nodules are common features of calcisols (Esteban & Klappa, 1983). Fabrics with columnar aspect were reported from marginal lacustrine facies by Freytet ( 1973). Vertical nodular structures are also typical features of paleosols (e.g. Cohen, 1982; Parnell, 1983; Brookfield & Sahni, 1986), and may be produced by root systems. Combinations of vertical structures (pneumatophores and geotrophic roots) with horizon tal features (lateral roots, desiccation sheet cracks and bedded root mats) can give rise to prismatic and reticulate fabrics (cf. Hoffmeister & Multer, 1965). An alternative explanation comes from the work of Dubiel, Blodgett & Bown ( 1987), who described somewhat similar vertical structures from the conti nental Upper Triassic of Colorado, which they interpret as fossil lungfish burrows. However, no lungfish remains were found here. The intercalated red horizons record supply of fine grained terrigenous sediment onto the exposed carbon ate surface. Fine sedi.ment may have come from incursions of the alluvial plain into the central lake areas during low lake stands or as wind-blown dust. The grey carbonate stringers present within the red horizons are interpreted as isolated root traces. Their concentrically micro-laminar wall structure is similar to that present in the laminated units and their cylindrical, locally branching, morphology is identical to that of rhizocretions described from Quaternary paleosols by Klappa ( 1980). The red horizons are thus thought to represent paleosols developed over, and locally penetrating, karstified limestone surfaces formed during prolonged periods of shoreline retreat. Thin green marls with smectite-illite-kaolinite mineralogies are common in Late Jurassic-Early Cretaceous marginal carbonate 'Purbeck' sequences (Deconinck & Strasser, 1987). Similar mineralogies were also reported from paleosols in the Lower Carboniferous (Mississippian) of South Wales by Robinson & Wright ( 1987). Illitization of smectites is commonly attributed to increasing temperature as burial diagenesis proceeds (Hower et a/., 1976; Nadeau et a/., 1985), but Eberl, Srodon & Northrop ( 1986) suggested that K-fixation necessary for illitization of smectites co1,1ld be achieved at surface temperatures by repeated wetting and drying. This led Robinson & Wright ( 1987) to suggest that some mixed layer illite smectite could be produced from smectite during pedogenesis (although this process has yet to be documented from modern soils). Deconinck & Strasser ( 1987) attributed input of detrital smectite and kaolinite in the European
Facies association 4: distal alluvial
Red marls locally make up as much as 50% of the Rupelo Formation sequence. These red, or red-green mottled, calcareous and silty mudstones occur as monotonous units up to 20-40 m thick. The marls are virtually structureless both at outcrop and in thin section, and contain only rare faunal or floral remnants in the form of a few fragmented ostracod shells and broken charophyte stems. White micritic limestones occur interbedded with the marls in lenses up to 5 m in thickness and 20 m in lateral extent. Interpretation
The red marls are interpreted as distal alluvial sediments that were deposited out of suspension in a floodplain environment. Their structureless nature suggests thorough bioturbation. Lithologically similar sandy mudstones were reported from the Late Creta ceous-Early Tertiary of SW France by Freytet & Plaziat ( 1982). These deposits were also associated with shallow water marginal lacustrine carbonates and clastics, and were likewise interpreted as distal floodplain deposits. Alternatively, the red marls might represent poorly-developed paleosols, a hypothesis supported by the local presence of mottling. However, no other pedogenic fabrics were discernible. The white micritic limestone lenses are interpreted as the deposits of isolated ponds developed in depressions on the floodplain. These ponds may have
331
Nigel H. Platt
been the sites of intermittent biogenic production and/or inorganic precipitation of carbonate where desiccation led to periods of subaerial exposure, resulting in some pedogenic modification. Facies association 5: evaporitic
Chert occurs in a 2-80 em thick chert bed towards the top of the Rupelo Formation sequence. This bed is traceable over much of the basin. The chert may show nodular fabrics (Fig. 6A). It is generally white or translucent grey-blue. Polygonal angular voids (25 mm) are common and l-5 mm angular patches of coarser silica are evident in thin section. Minor amounts of length-slow chalcedony and a few clustered 0·1 mm dolomite rhombs also occur. Yellow limestones are associated with the cherts. These carbonates may be nodular, consisting of microspar with abundant fine sparry cracks, or, more commonly, vuggy (Fig. 6B). The vuggy limestones consist of elongate 0· 5 x 2 mm calcite strings arranged in an irregular mesh network and separated by abundant angular interstitial voids 2-10 mm across, giving the rock a highly porous, sponge-like texture. Interpretation
The chert contains minor amounts of length-slow chalcedony, which is indicative of sulphate replace ment (Folk & Pittman, 1971). The cherts are thought to represent silicified evaporites. Nodules, crystal
shapes and voids similar to those present here have been described by a number of authors (e.g. West, 1964; Chowns & Elkins, 1974; Tucker, 1976; Milliken, 1979; Arbey, 1980) and interpreted as evaporite silicification fabrics. The dominantly trapezoidal pseudomorph shapes are probably gypsum forms, although inclusions of sulphates are not preserved. Features associated with subaqueous evaporites, such as depositional lamination or evidence of wave or current structures, are absent. However, a change in the net water budget, brought about by increased aridity or by tectonically-induced change in drainage may have caused a drop in water level eventually leading to the evaporite deposition in a continental sabkha environment. There is no evidence for a biogenic silica source. However, given sufficient input of dissolved silica to lake waters, then silicification may be achieved inorganically under high pH conditions. In alkaline lakes such as the Coorong (Peterson & von der Borch, 1965) and Lake Magadi (Eugster, 1967, 1969), direct precipitation of cristobalite silica or silicification of evaporite phases is possible. However, the high limestone content of the Rupelo Formation and evidence for the presence of primary gypsum argues against highly alkaline waters (see Drever, 1982). The common vuggy texture of the yellow limestones is consistent with evaporite solution. Dissolution of evaporite minerals (e.g. sulphates) could create an open porous fabric of this type. An alternative origin could involve dedolomitization of a cellular dolomite.
Fig. 6. Evaporitic facies. (A) White chert displaying undulating nodular fabric (cf. 'chicken-wire' texture). Scale bar= 2 mm. Interpretation : replacement of evaporites, probably sulphates, by silica. (B) Yellow vuggy carbonate showing extremely porous, open mesh fabric ofO·5 x 2 mm diameter calcite strings and interstitial angular voids. Scale bar= 2 mm. Interpretation: evaporite replacement .
332
Sedimentology and origin of palustrine deposits
subsidence within the basin (Platt, 1989). Thin Rupelo Formation sequences (10 m or less) are composed entirely of marginal lacustrinej'palustrine' and distal alluvial facies. These were deposited where subsidence was gentle. The evaporitic, open lacustrine and reworked carbonate facies occur only in the thickest sequences (25 m or more). These were deposited m areas of strongest subsidence.
FACIES ASSOCIATIONS AND SEQUENCES
Characteristic facies associations in the Rupelo Formation are outlined and interpreted in Table I . large-scale evolution o f facies in vertical sequence within the Rupelo Formation is indicated in Fig. 2. Climate
Regressive sequences
Platt ( 1989) suggested that long-term climatic changes were responsible for the evolution of environments observed through time (Fig. 2; Table 2). Semi-arid conditions prevailed during deposition of the distal alluvial Las Vinas Member, and persisted through sedimentation of the palustrine Ladera Member. Deposition of open lacustrine facies (Mambrillas de Lara Member) suggests a more humid climate, but subsequent greater aridity led to the deposition of evaporites before a return to ephemeral lake carbonate deposition (Rio Cabrera Member).
Marginal lacustrine facies are characteristically ar ranged in 'regressive' sequences (Fig. 7A). These sequences show upward increases in the degree of pedogenic modification. However, soil profiles are commonly incomplete and/or superimposed. The frequency of desiccation points to short-term climatic variation (seasonal or between individual years). The influx of red silts and marls reflects increase in terrigenous input, as a result of lake regression, climatic change or increased tectonic activity.
Tectonics
Graded sequences
Rapid lateral facies and thickness changes between individual fault blocks reflect the pattern of differential
Open lacustrine and reworked carbonate facies (Fig. 7B) show small-scale vertical variation in the
Table I. Characteristic facies associations within the Rupelo Formation, with environmental interpretations.
Interpretation
Lithofacies association
Periodically-inundated floodplain, ponded carbonates Marginal lacustrine, frequently exposed, terrigenous progradation Marginal lacustrine, less frequent exposure Clastic input at lake margin, river-generated density currents Open lacustrine, varying energy Open lacustrine, shallow, unstratified perennial lake
Red marls-white limestone lenses Peloidal limestones-rubbly horizons-red marls Green marls-root mats-white carbonates with microkarst Limestone conglomerates-arenites-clastic laminated mudstones/marls Grainstones-wackestones Marls /mudstones with charophytes, ostracods, gastropods , bones, prints Dark grey marls /mudstones, diffuse lamination Grey marls-wackestones-intraformational egis
Open lacustrine, deeper water, possible oxygen-poor bottom Open lacustrine, reworking in drainage channels on emergence Des iccating lake flat, evaporitic Periodic des iccation in marginal lacustrine carbonate flat
Cherts-yellow limestone-yellow vuggy limestones Wackestones-mudstones with desiccation cracks-red marls
Table 2. Interpretation of environments through Rupelo Formation time. Facies reflect variation in climate, carbonate production and clastic supply .
Member Las Vinas Ladera Mambrillas de Lara Rio Cabrera
Carbonate production
Climate Semi-arid Semi-arid More humid Arid, semi-arid
Minor, in ponds Important, freshwater, interrupted Important, freshwater, more cont inuous Important later in freshwater
333
Clastic supply High, from NW Low Low to zero Low
SEDIMENTARY STRUCTURES AND FOSSILS
�
LITHOLOGY chert vuggy yellow limestone grey marl intraformational limestone conglomerate green marl I clay
ertebrate bones
••
vertebrate tracks
<;{
mottling
charophytes (stems, gyrogonites)
breccia I paleosol
0
micritic intraclasts
limestone
)�-( polygonal cracks � microkarst >;:;
laminar rhizolite horizon red mudstone I marl
• dark intraclasts ("black pebbles")
limestone with microkarst
C §
circumgranular cracks, glaebules gastropods r-- ostracods root traces (rhizoliths)
/..
t I
grading regressive sequences
8
A
)...._/
-)-1.....
@ 0 c
1..0
•
1..0
@"
•
)��
t
• §,<' •of'2"
+
t t
@
"
•o •e
• •o@& I 0 • @&" I •oe
t t 0 c 0
0 DO 00
:e����
t
•o&" e •o"'
0 w
I
I.
p gr r
p gr r
texture
texture
Fig. 7. Typical sedimentary logs. (A) Facies associations 3 & 4: marginal lacustrine and associated distal alluvial faci es associations. Facies are arranged in shallowing upward sequences. (B) Facies associations I, 2 & 5: open lacustrine, reworked carbonates and evaporitic. Grading represents change in energy consequent upon transgression and storm events.
334
Sedimentology and origin of palustrine deposits " "PALUSTRINE marginal lacustrine pedogenesis on emergence
periodically inundated alluvial plain - red marls
&
silts
vegetation -root crusts and rhizocretions
grainstones on wave-dominated shoreline
fluctuating shorelines
rare dark organically-richer laminated mudstones
karst surface - on emergence terrigenous input
OPEN LACUSTRINE mudstones and wackestones charophyte
�red soils
gyrogonites
ostracods gastropods sauropods
Fig. 8. Schematic environmental model for the Rupelo Formation, adapted from palustrine model of Cabrera, Colombo & Robles ( 1985).
size and abundance of intraclasts. This is particularly evident in the wackestones. Grading above erosional surfaces probably reflects decreasing energy during transgression or storms.
dominantly from biogenic sources, chiefly from calci fied charophyte stems, which were easily broken up. Some organic matter was land-derived, notably as 'black pebbles', which may represent residue from burning of terrestrial organic debris. Lake transgressions reworked intraclasts from marginal into open lacustrine environments, and low stage lake-flat drainage channels reworked exposed open lacustrine facies during periods of shoreline retreat. Grainstones made up of open lacustrine facies intraclasts were deposited as gravels on exposed shorelines and shoals, perhaps during storms. Marginal lacustrine areas. The abundance of pedo genic features and the rare presence of a poorly preserved freshwater flora in the carbonates of this facies association implies that they are lacustrine carbonates that have been extensively modified during pedogenesis. A marginal setting in a low gradient, shallow ephemeral lacustrine setting is inferred. The shorelines of low-gradient lakes (e.g. Lake Chad;
ENVIRONMENTAL SYNTHESIS
Faunal and floral evidence indicate that the Rupelo Formation was deposited in a shallow, carbonate producing (hardwater) lake. Figure 8 shows a sche matic model for deposition during Rupelo Formation time. The main sub-environments represented in the Rupelo Formation depositional model are as follows. Open lacustrine areas. The dominance of unlami nated deposits reflects intense bioturbation and im plies shallow water depths without permanent stratification, although rare preservation of laminated fabrics may reflect periods of higher organic produc tivity within the lake. Carbonate derivation was
335
Nigel H. Platt
Reading, 1982) fluctuate greatly as a result of seasonal and longer-term climatic variations, so that periods of submergence and exposure alternate. Frequent oscillations of the shoreline led to the exposure and modification of marginal lacustrine facies. Pedogenic modification resulted in loss of original fabrics and development of mottling, sparry cracks, microkarst cavities, and brecciation, as well as development of macro- and micro-scale root structures. Colonization by abundant shallow-water marsh and land plants formed a fringing vegetated zone around the lake margins, which acted as an effective barrier to terrigenous clastic input. A possible modern analogue is provided by Ruby Lake, Nevada, in the Basin and Range of the United States (Fouch & Dean, 1982), where open lake areas are surrounded by fringing marshlands with emergent aquatic plants, grasses and shrubs. Floodplain environment. In low-gradient, shallow lakes, small-scale fluctuations in water level result in exposure of large areas, and occasional influx of fine grained alluvial material is common. The red marls and silts of the Rupelo Formation were deposited on alluvial plains around the lake. These occasionally encroached over large areas of the basin. Deposition in the distal alluvial setting was mainly from suspen sion. Carbonate sedimentation was confined to ponds developed on the floodplain. Only on prolonged exposure did the alluvial systems prograde sufficiently into the central lake areas to introduce significant quantities of clastic material. However, the occur rences noted above (see facies association I) of thin conglomerate and sandstone beds and mm-scale graded silt laminae in the NW of the basin suggests some clastic input from that direction. Evaporitic environment. During the later part of Rupelo Formation time, periods of extreme desiccation led to the widespread establishment of hypersaline conditions and the formation of evapor ites. Evaporite precipitation was associated with diagenetic silicification and dolomitization (and pos sibly subsequent dedolomitization).
lithofacies and components rather than investigate vertical trends in whole rock compositions, which were likely to be highly complex as a result of the heterogeneous fabrics and common superposition of pedogenic profiles. Samples were taken from only a few localities, and virtually all from a small geograph · ical area in the NW of the basin. Sample powder (1·5-5 mg) was drilled out under the microscope from Lakeside-mounted thin sections 150 f.lm in thickness using a fine engraving tool. Care was taken to sample from the micritic components; fine sparry cracks and bioclasts were avoided as far as possible. As a check on compositional heterogeneity, some analyses were controls carried out on samples drilled from different areas of the same· thin section or from different thin sections cut out of the same hand specimen. Isotopic data were obtained by standard techniques (McCrea, 1950). Results were calibrated against MCS-8, a known laboratory standard (reprod ucibility shown in Table 3), and the values expressed in parts per thousand with reference to the PDB standard (Craig, 1957). Results
Carbon and oxygen stable isotopic values for the Table 3. Carbon and oxygen stable isotope data from the Rupelo Formation.
Sample 3l 6B/l 339A/l 339B/l 34l A/l 341B/l B34l 343A/I 344/l 344/2 V35A/l V35 /2 5 60A/l 5 60B/l V30/l V39/I V39/2 V39/3 Vl l 4 NP77 V91/l V82B/l
STABLE ISOTOPES Methods
A total of 2 1 carbon and oxygen stable isotope analyses were performed on selected carbonates from the Rupelo Formation. The sampling philosophy for this initial small sample suite was designed to characterize the isotopic composition of different carbonate
Description palustrine palustrine palustrine palustrine palustrine black pebble palustrine-root mat palustrine palustrine palustrine palustrine open lacustrine open lacustrine open lacustrine-laminated open lacustrine open lacustrine open lacustrine open lacustrine+ fenestrae open lacustrine-black pebble crystal silt crystal silt
.513C
C%o)
.5'80 C%o)
-8· 5 -8·7 -8·2 -8·5 -8·1 -9·2 -9·2 -9-4 -8·1 -8·0 -8· 2 -7·1 -7-4 -7· 5 -7·8 -7·9 -7·7 -8·5 -9·0 -10·9 -10·2
-5·5 -6-4 -5·0 -5·8 -5·1 -6·7 -7·2 -5· 7 -4·7 -4· 3 -5·5 -4· 4 -4·6 -H -4·5 -5·3 -4·6 -4·6 -7·1 -7-4 -8· 0
Laboratory standard MCS-8 .513C = -0·700; .5180= -9·177 l 0 measurements: Standard deviation: .513C +I- 0·07; .5180 +/- 0·17
336
Sedimentology and origin of palustrine dep osits
freshwater limestones' (Keith & Weber, 1964) and of pedogenic carbonates (Talma & Netterberg, 1983), although they are strongly negative even within these fields. Three possible processes which might have affected the stable isotope values are :
Rupelo Formation (Table 3; Fig. 9) all lie on an approximately linear trend in the negative b 1 3 C, negative () 1 80 quadrant, with spreads of b 1 3 C from - 7 to - 1 1 %0 and of b 1 80 from - 4 to - 8%0. The 'palustrine' samples showed considerable iso topic inhomogeneity (b 1 3 C + /-0·6%0; b 1 8 0 + / - l · l %0). Open lacustrine carbonates were more homogeneous (b 1 3 C + / - 0·2%0; () 1 80 + / - 0·3%0). Samples from the open lacustrine facies association showed heavier values than palustrine samples. The heaviest isotopic compositions were recorded from a laminated open lacustrine biomicrite. Of the palustrine samples analysed, the lightest compositions recorded were from a 'root mat' (laminar rhizolite horizon) and from blackpebbles. Themostnegative valuesofall were obtained from samples of the geopetal crystal silt in fenestral cavities.
(I) Progressive evaporitic fractionation in an ephem eral lake settling towards 'heavier' isotopic com positions (compare Lloyd, 1966; Salomons, Goudie & Mook, 1978; Oberhansli & Allen, 1987). (2) Progressive modification due to increasing influ ence of fractionated meteoric groundwaters and isotopically-light, soil-derived C02 in near-sur face, vadose diagenetic (Gross, 1964; Allan & Matthews, 1977, 1982) or pedogenic (Beier, 1987; Cerling, Bowman & O'Neill, 1988) settings. Soil derived C02 has b 1 3 C from - 22 to - 8%o (Cerling, 1984). (3) Variation in the organic productivity of lake waters. Photosynthesis results in 1 3 C enrichment of dissolved C02 and hence favours more positive () 1 3 C values for lake carbonates deposited during times of higher productivity and more negative values when productivity is lower (McKenzie, 1985; Oberhansli & Allen, 1987).
Interpretation
The compositional heterogeneity of the palustrine samples reflects their complex fabrics. These speci mens contain a variety of pedogenic components including glaebules, micro-pellets, rhizolith carbonate etc., as well as fine micritic matrix. The open lacustrine samples have simpler fabrics with fewer components and show greater isotopic homogeneity. The values obtained lie close to those of 'average -10
-9
-7
-8
-6
Evaporitic fractionation would tend to produce isotopically lighter carbonates at times of high lake stand and isotopically heavier carbonates during low lake stands when the residual lake waters were more
• •
'
1!1 D 1!1
•
-3
-4
-5
-2
-1
0
. .·.. .
• •
@
••
• palustrine
e open
o palustrine
@ open lacustrine - fenestrae 0 open lacustrine - laminated
1!1 palustrine - black pebbles
- root
.o1. crystal silts
0
mat
lacustrine
Fig. 9. Plot of carbon and oxygen stable isotope data from the Rupelo Formation, showing correl ation of facies and isotopic compos ition.
337
Nigel H. Platt
evolved. However, the observed correlation between isotopic composition and facies shows the opposite tendency : open lacustrine limestones are isotopically heavier than palustrine limestones collected from the same locality. This trend is more consistent with processes 2 and 3. The trend is thus interpreted to represent the effects of increasing pedogenic/vadose diagenetic alteration of open lacustrine limestones, whose initial isotopic compositions reflected varia tions in organic productivity. The final isotopic composition after diagenesis depends upon the isotopic composition of the initial carbonate, the composition of the diagenetic fluids and the degree of exchange. The initial isotopic composition of unmodified open lacustrine carbonate was probably close to that of the laminated lacustrine carbonate (<5 1 3 C = - 7%0, <51 8 0 = - 3 %0) , which is likely to have been deposited during a period of relatively high organic prod uctivity and which shows no petrographic evidence of bioturbational mixing or exposure. Since there were probably no C4 plants during the Cretaceous, the soil carbonate <5 1 3 C value would probably have been about - 12%0 (Cerling, pers. commun., 1988). Thus the <5 1 3 C and <5 1 8 0 values from the crystal silts, at around - I 0 5%0 and 7 5%0 respectively, are likely to represent the closest approx imations to the isotopic compositions of the soil-gas charged vadose diagenetic fluids. The isotopic compositions of the other samples represent a spectrum reflecting both lower organic productivity and variable degree of isotopic exchange with vadose diagenetic fluids. The unlaminated open lacustrine limestones show typical isotopic composi tions of <5 1 3 C = - 7·8%0; <5 1 8 0 = - 4 ·5%0. The lack of lamination reflects bioturbation in oxygenated bottom waters, and this suggests less intense organic produc tivity. There is little evidence of exposure. Otherwise similar open lacustrine limestones containing fenes trae have slightly more negative isotopic values, particularly in b 1 3 C (<51 3 C = - 8%0 , <5 1 8 0 = - 4·6%0) , suggesting some contact with isotopically light vadose diagenetic fluids. Palustrine limestones show typical compositions of <5 1 3 C = - 8 ·5%0 , <5 1 8 0 = - 5·5%0. These more negative values reflect more extensive pedogenetic modifica tion. The palustrine limestone samples showing the lightest isotopic compositions are from root mats and black pebbles (<51 3 C = - 9%0, <5 1 8 0 = - 7%0). These soil components are likely to record prolonged contact with isotopically light soil-derived C02. Microscopic observation also shows slightly greater contents of organic carbon. ·
-
D I S C U SS I O N
Palustrine limestones commonly show similar fabrics to those of calcisols; indeed Esteban & Klappa ( 1983, p. 40) stressed the problem of distinguishing between palustrine and pedogenic carbonates. In his compre hensive study of Quaternary calcretes from Spain, Klappa ( 1978) described pedogenic features including crusts, vertical roots, horizontal roots, root mats, root rock ('rhizolite'), solutional cavities, glaebular con glomerates, black pebbles (especially in coastal areas), sheet calcrete, soils, and brecciated textures. All of these occur in the Rupelo Formation and in many other 'lacustrine' and 'palustrine' carbonate sequences (see, for example, Stanley & Collinson, 1979 ; Freytet & Plaziat, 1982 ; Cabrera et a!., 1985). An explanation for the similarity between palustrine and pedogenic carbonates comes from Freytet's (1973) suggestion that palustrine limestones formed through pedogenic modification of lacustrine carbonate mud stones. Stable isotope data from the Rupelo Formation supports this hypothesis. Esteban & Klappa ( 1983) recognized a 'continuous spectrum between lacustrine carbonates, palustrine carbonates and caliche'. The stable isotope study presented here suggests that this spectrum of fabrics is mirrored by a parallel spectrum of isotopic compositions. The fabrics reflect both the depositional conditions (laminated or unlaminated lacustrine carbonate) and the later effects of pedoge nesis. The isotopic compositions are likewise inter preted as the result of variations in initial composition caused by changing lake productivity, and of second ary modification due to diagenesis in the vadose zone. Isotopic compositions of the lacustrine-palustrine carbonates reveal that the near-surface diagenetic processes acting on freshwater carbonates after lacus trine desiccation (i.e. those associated with the formation of palustrine li mestones) are geochemically analogous to the processes of meteoric vadose diage nesis known to act on marine carbonates on regression and subaerial exposure. Distinctive characteristics of the lacustrine-palustrine environment appear to be isotopically lighter initial carbonate compositions and the potential for more rapid and more frequent regression than in marginal marine carbonate settings.
·
CON CLUSIONS
The Berriasian Rupelo Formation of the Camero.s Basin is a lacustrine-palustrine sequence deposited in a developing continental rift (Platt, 1989). Brief 338
Sedimentology and origin of palustrine dep osits
periods of stable lacustrine sedimentation led to uninterrupted deposition of charophyte-gastropod carbonate mudstones and wackestones, with rare laminites. However, the lake was generally ephemeral with widely fluctuating shorelines reflecting the response of a low-gradient lake margin to oscillating lake stands. Extensive areas were subject to repeated periods of exposure and strong pedogenic modification during successive intervals of shoreline retreat. Intercalated red marls are interpreted as fine grained distal alluvial sediments deposited on an extensive muddy floodplain fringing the lake complex. A brief phase of more intense desiccation led to the short-lived establishment of hypersaline conditions, as documented by cherts and vuggy limestones both showing evidence of evaporite replacement. The Rupelo Formation shares many features in common with palustrine deposits described in the literature (e.g. Glass & Wilkinson, 1 980; Freytet & Plaziat, 1 982; Wells, 1 983; Cabrera et a!., 1 985). However, it is notable firstly for its tectonic environ ment (in an extensional rift rather than in a foreland basin setting), for the virtual .1bsence of interbedded coarse clastic deposits, for the paucity of facies free from indications of pedogenic modification, and also perhaps for its age (it may represent a unique freshwater example of the European 'Purbeckian' facies). Carbon and oxygen stable isotope compositions of the open lacustrine carbonates show variation consist ent with sedimentary evidence for changes in organic productivity within the lake. Stable isotqpe data from the marginal lacustrine carbonates support the petro graphic evidence for the action of pedogenic and vadose diagenetic processes during formation of palustrine limestones, and are consistent with an origin through progressive modification of open lacustrine carbonates during subaerial exposure.
minescence. Stable isotope analyses were carried out at the British Geological Survey in London with assistance from B. Spiro and Lynda Thrift. Thanks also go to Andre Strasser & Eric Davaud at Geneva, to Steve Wright and Ghassan Al-Murani at Oxford, and to Professor Albert Matter, K. Ramseyer and H. Dronkert at Berne for their interest. T. D. Fouch and T. E. Cerling are warmly thanked for their helpful reviews of the final manuscript. R E F E R E N CES
ALLAN, 1 . R . & MATTHEWS, R . K . (1977) Carbon and oxygen isotopes as diagenetic and stratigraphic tools : surface and subsurface data, Barbados, West Indies . Geology, 5, 1620. ALLAN, 1 . R & MATTHEWS, R . K . (1982) Isotope s ignatures associated with early meteoric diagenesis. Sedimentology, 29, 797- 817. ARBEY, F . (1980) Les formes de silice et !'identification de evaporites dans les formations silicifiees. Centr. Rech. Expl. Prod. ElfAquitaine Bull., 4, 309- 365. BEIER, 1 .A. ( 1987) Petrographic and geochemical analysis of caliche profiles in a Bahamian Pleistocene dune . Sedimen tology, 34, 991-998. BREWER, R. ( 1964) Fabric and Mineral Analysis of Soils. Wiley, New York, 470 pp . BROOKFIELD, M . E . & SAHNI, A. (1986) Palaeoenvironments of the Lameta. Beds (Late Cretaceous) at 1 abapur, Madhya Pradesh, India : Soils and biotas of a semi-arid alluvial plain. Cretaceous Res., 8, 1- 14. BROWN, R . E . & WILKINSON, B. H . (1981) The D raney Limestone : Early Cretaceous lacustrine carbonate depo sition in western Wyoming and southeastern Idaho. Contribs Geology, Univ. Wyoming, 20, 23- 31. BUURMAN, P. (1980) Palaeosols in the Reading Beds (Palaeocene) of Alum Bay, Isle of Wight, U K . Sedimentol ogy, 27 , 593- 606. CABRERA, L., COLOMBO, F. & ROBLES, S. ( 1 985) Sedimenta tion and tectonics interrelationships in the Palaeogene marginal alluvial systems of the S E E bro Basin ; transition from alluvial to shallow lacustrine environments . Excursion .
Guide, 6th Eur. Regional Mtg., Int. Ass. Sedimentologists,
pp. 393- 492. CAYEUX, L. (1935) Les Roches si!dimentaires de Ia France. Roches Carbona tees. Masson, Paris, 447 pp. CERLING, T.E. (1984) The stable isotopic composition of modern soil carbonate and its relationship to climate. Earth planet. Sci. Lett. , 71, 229-240. CERLING, T. E . , BOWMAN, 1 . R. & O'NEIL, 1 . R . (1988) An isotopic study of a fluvial-lacustrine sequence : the Plio Pleistocene Koobi Fora sequence, East Africa. Palaeo geogr. Palaeoc/im . Palaeoeco/., 63, 335-356. CHOWNS, T . M . & ELKINS, 1.E. (1974) The origin of quartz geodes and cauliflower cherts through silicification of anhydrite nodules . J. sedim. Petrol., 44, 885-903. COHEN, A.S. (1982) Paleoenvironments from the Koobi Fora Formation, Kenya . J. sedim . Petrol., 52, 401- 414. COHEN, A.S. & THOUIN, C. (1987) Nearshore carbonate deposits in Lake Tangany ika. Geology, 15, 414-418. CRAIG, H. (1957) Isotopic standard for carbon and oxygen
A C K NOWLE D G M E N TS
This paper preseuts data submitted as part of a NERC-funded D.Phil. thesis at the Department of Earth Sciences, University of Oxford. I thank Harold Reading for his supervision, and Paul Wright at Bristol for constant encouragement and for sugges tions which helped to improve an earlier version of the manuscript. Thin sections were produced by Phil Jackson & Jon Well.s ; help with photography came from Richard McAvoy in Oxford and Andy Werthe mann in Berne. Steve Baker assisted with cathodolu339
Nigel H. Platt
and correction factors for mass-spectrometric analysis of carbon dioxide. Geochim. Cosmochim. Acta, 12, 133- 149.
GLASS, S. W. & WILKINSON, B. H . (1980) The Peterson Limestone - Early Cretaceous lacustrine carbonate depo sition in western Wyoming and southeastern Idaho. Sediment. Geo/., 27, 143-160. GRoss, M . G. (1964) Variations in the 1 80/160 and 1 3Cf 1 2C ratios of diagenetically altered limestones in the Bermuda Islands . J. Geol. Chicago, 72, 17o-- 194. HOFFMEISTER, J.E. & MULTER, H . G . (1965) Fossil mangrove reef of Key Biscayne, Florida. Bull. geo/. Soc. Am., 76, 845-852. HOWER, J., ESLINGER, E . V. , HOWER, M . E . & PERRY, E . A. ( 1976) Mechanisms of burial metamorphism of argilla ceous sediment : 1) Mineralogical and chemical evidence. Bull. geol. Soc. Am., 87, 725-737. JAMES, N.P. ( 1 972) Holocene and Pleistocene calcareous crust (Caliche) profile: Criteria for subaerial exposure. J. sedim. Petrol., 42, 817- 836. KEITH, M.L. & WEBER, J . N. ( 1964) Carbon and oxygen isotopic composition of selected limestones and fossils . Geochim. Cosmochim. A cta, 28, 1 787- 1816. KLAPPA, C . F . (1978) Morphology, Composition and Genesis of
DEAN, W . E . & FoucH , T.D. (1983) Lacustrine Environment. In: Carbonate Depositional Environments (Ed. by P. A. Scholle, D. G. Bebout & C. H. Moore). Mem . Am. Ass. Petrol. Geo/., 33, 96- 130. 0ECONINCK, J.F. & STRASSER, A. ( 1 987) Sedimentology, clay mineralogy and depositional environment of Purbeckian green marls (Swiss and French Jura). Ec/og. geol. Helv. , 79, 753-772. 0ECONINCK, J.F., STRASSER, A. & 0EBRABANT, P. (1988) Formation of illitic minerals at surface temperatures in Purbeckian sediments (Lower Berriasian, Swiss and French Jura). Clay Miner., 23, 91-103. DREYER, J . I . ( 1 982) The Geochemistry of Natural Waters. Prentice Hall, Englewood Cliffs , 388 pp. DUBIEL, R . F . , BLODGETT, R.F. & BOWN, T.M. (1987) Lungfish burrows in the Upper Triassic Chinle and Dolores Formations, Colorado Plateau. J. sedim. Petrol., 57, 5 1 2-5 21. DUNHAM, R.J. (1969) Early vadose silt in Townsend mound (reef), New Mexico and Texas . In : Depositional Environ ments in Carbonate Rocks. (Ed. by G. M. Friedman). Spec. Pubis Soc. econ . Paleont. Miner., Tulsa, 14, 182- 191.
Quaternary Calcretes from the western Mediterranean : a Petrographic Approach. Unpublished PhD Thesis, Univer
s ity of Liverpool, 446 pp. KLAPPA, C. F. ( 1980) Rhizoliths in terrestrial carbonates : classification, recognition, genesis and s ignificance. Sedi mentology, 27, 613-629. LEINFELDER, R. R . (1987) Multifactoral control of sedimen tation patterns in an ocean marginal basin - the Lusitanian Basin (Portugal) during the Jurassic and Cretaceous. Geo/. Rdsch., 76, 5 99-631. LLOYD, R. M. ( 1966) Oxygen isotope enrichment of sea-water by evaporation. Geochim. Cosmochim. A cta, 30, 801-814. McCREA, J . M . (1950) 0n the isotopic chemistry ofcarbonates and a paleotemperature scale. J. chem. Phys., 18, 849- 857. McKENZIE, J . A. (1985) Carbon isotopes and productivity in the lacustrine and marine environment. In : Chemical Processes in Lakes. (Ed. by W. Stumm), Wiley, New York, pp. 99- 118. MENSINK, H . & SCHUDACK, M . (1982) Caliche, Bodenbildung und die paliiogeographische Entwicklung an der Wende mariner Jura /Wealden in der westlichen S ierra de los Cameros (Spanien). N. Jb. Geo/. Paliiont. Abh., 163, 4980. MILLIKEN, K . L . (1979) The silicified evaporite syndrome two aspects of silicification h istory of former evaporite nodules from southern Kentucky and northern Tennessee. J. sedim. Petrol., 49, 245- 256. MOUNT, C . F . & COHEN, A.S. (1984) Petrology and geochem istry of rhizoliths from Plio-Pleistocene fluvial and marginal lacustrine deposits, east Lake Turkana, Kenya . J. sedim. Petrol., 54, 263-275. MULTER, H . G·. & HOFFMEISTER, J.E. (1968) Subaerial laminated crusts of Florida Keys. Bull. geol. Soc. Am., 79, 183-192. MURPHY, D . H . & WILKINSON, B. H . (1980) Carbonate deposition and facies distribution in a central Michigan marl lake. Sedimentology, 27, 123'- 1 35 . NADEAU, P. H . , WILSON, M.J., MCHARDY, W . J . & TAIT, J . M . (1985) The conversion of smectite to illite during
EBERL, D.O., SRODON, J. & NORTHROP, H . R . (1986) Potassium fixation in smectite by wetting and drying. In: Geochemical Processes at Mineral Surfaces (Ed. by J . A. Davies & K . F. Hayes). Am. chem. Soc. Symp. Ser., 323/ 14, 296- 326.
•
EsTEBAN, C. M . (1974) Caliche textures and 'Microcodium'. Boll. Soc. geo/. Ita/., 92 (Suppl.), I 05- 125. ESTEBAN, M. & KLAPPA, C. F . (1983) Subaerial exposure environment. In: Carbonate Depositional Environments. (Ed. by P. A. Scholle, D. G. Bebout & C. H. Moore). Mem. Am. Ass. Petrol. Geo/., 33, 1-63. EuGSTER, H.P. (1967) Hydrous sodium silicates from Lake Magadi, Kenya ; precursors of bedded chert. Science, 157, 11 77- 1 180. EUGSTER, H.P. (1969) Inorganic bedded cherts from the Magadi area, Kenya . Cont. Miner. Petrol., 22, 1-33. FOLK, R.L. & PITTMAN, J.S. (1971) Length-slow chalcedony: a new testimony for vanished evaporites. J. sedim . Petrol., 41, 1045- 1058. FoucH, T.D. & DEAN, W . E . (1982) Lacustrine and associated clastic depositional environments. In: Sandstone Deposi tional Environments. (Ed. by P. A. Scholle & D . Spearing). Mem. Am. Ass. Petrol. Geol., 32, 87- 1 14.
FRANCIS, J. E . (1986) The calcareous paleosols of the basal Purbeck Formation (Upper Jurassic), Southern England . In: Paleosols : their Recognition and Interpretation. (Ed. by V. P. Wright). Oxford, Blackwell, pp. 112- 1 38. FREYTET, P. ( 1 973) Petrography and paleoenvironment of carbonated continental deposits with a particular reference to Upper Cretaceous and Lower Eocene of Languedoc (southern France). Sediment. Geol., 10, 25-60. FREYTET, P. & PLAZIAT, J .C . (1982) Continental carbonate sedimentation and pedogenesis-Late Cretaceous and Early Tertiary of Southern France. Contr. Sedimentol., 12, 213 pp.
·
340
Sedimentology and origin of palustrine deposits
diagenes is : evidence f rom some illitic clays from benton ites and sandstones . Mineral. Mag., 49, 393-400. 0BERHANSLI, H. & ALLEN, P . A. (1987) Stable isotopic s ignatures of Tertiary lake carbonates, Eastern Ebro Basin, Spain . Palaeogeogr. Palaevclim. Palaeoecol., 60, 5 975. PARNELL, J. (1983) Ancient duricrusts and related rocks in perspective : a contribution f rom the Old Red Sandstone. In : Residual Deposits : Surface Related Weathering Proc esses and Materials. (Ed. by R. C. L. Wilson). Spec. Pub!. geol. Soc. London, 1 1 , 197- 209. PETERSON, M . N . A. & VON DER BORCH, C.C. (1965) Chert : modern inorganic deposition in a carbonate-precipitating locality. Science, 149, 1501-1503. PHILLIPS, S . E . & S E LF, P.G. (1987) Morphology, crystallog raphy and origin of needle-fibre calcite in Quaternary pedogen ic calcretes of South Australia. Aust. J. Soil Res., 25, 249- 444. PLATT, N . H . (1986) Sedimentology and Tectonics of the
SHINN, E . A. & LIDZ, B. H . (1 987) Blackened l imestone pebbles : fire at subaerial unconformities. In : Paleokarst (Ed. by N. P. James & P. W. Choquette). Sp ringer-Verlag, New York, 11 7- 131. SLADE'N, C . P . (1983) Trends in Early Cretaceous clay mineralogy in NW Europe. Zittelania, 1 0, 349- 357. STANLEY, K . O. & COLLINSON, J. W . (1979) Depositional h istory of Paleocene-Lower Eocene Flagstaff Limestone and coeval rocks, Central Utah . Bull. Am. Ass. Petrol. Geol., 63, 31 1- 323. STRASSER, A. (1984) Black pebble occurrence and genesis in Holocene carbonate sediments (Florida Keys, Bahamas and Tun isia). J. sedim. Petrol., 54, 1097-1 109. STRASSER, A. (1987) Detaill ierte Sequenzstratigraphie und ihre Anmeldung : Beispiel aus dem Purbeck des schweiz erischen und franzosischen Jura. Facies, 17, 237-244. STRASSER, A. (1988) Shallowing-upward sequences in Pur beckian peritidal carbonates (lowermost Cretaceous), Swiss and French Jura Mountains. Sedimentology, 35 , 369383. STRASSER, A. & DAVAUD, E. (1982) Les croutes calcaires (calcretes) du Purbech�n du Mont-Saleve (Haute-Savoie), France. Eclog. geol. Helv. , 75, 287-301. STRASSER, A. & DAVAUD, E. (1983) Black pebbles of the Purbeckian (Swiss and French Jura) : lithology, geochem istry and orig in. Ec!og. geol. Helv. , 76, 551-5 80. SWINEFORD. A. , LEONARD, A. B. & FRYE, J.C. (1958) Petrology of the Pliocene p isolitic limestone in the Great Plains. Bull. Kansas geol. Surv., 130, 97-116. TALBOT, M.R. (1984) Prel iminary results from sediment cores from Lake Bosumtwi, Ghana. Palaeoecol. Afr. Surr. Isis. , 16, 1 73- 192. TALMA, A.S. & NETTERBERG, F. (1 983) Stable isotopic abundances in calcretes. In : Residual Deposits : Surface Related Weathering Processes and Materials. (Ed . by R. C . L. Wilson). Spec. Pub!. geol. Soc. London, I I, 221- 233. TREESE, K.L. & WILKINSON, B. H . (1982) Peat-marl deposi tion in a Holocene paludal-lacustrine basin - Sucker Lake, Michigan . Sedimentology, 29, 375-390. TuCKER, M.E. (1976) Replaced evaporites from the Late Precambrian of F innmark, Arctic Norway . Sediment. Geol., 16, 193- 204. WARD, W .C. (1975) Petrology and diagenesis of carbonate eolian ites of northeastern Yucatan Peninsula, Mexico. In :
Western Cameros Basin, Province ofBurgos, Northern Spain.
Unpublished D . Phil Thesis, University of Oxford, 250 pp. PLATT, N . H . (1989) Climatic and tectonic controls on sedimentation of a Mesozoic lacustrine sequence : The Purbeck of the W Cameros Basin, N Spain. Palaeogeogr. Palaeoclim. Palaeoecol., 70. READING, H . G. (1982) Sedimentary basins and global tectonics. Proc. geol. Ass., 93, 321- 35 0. ROBINSON, D. & WRIGHT, V.P. (1987) Ordered illite-smectite and kaolinite-smectite : Pedogenic minerals in a Lower Carboniferous paleosol sequence, South Wales. Clay Miner., 22, 1 09-118. SALOMON, J. (1982) Les formations continentales du Juras s ique superieur-Cretace inferieur (Espagne du Nord, Chaines Cantabrique et NW Iberique). Mem. Geol. Univ. Dijon, 6, 228 pp. SALOMON, J. (1983) Les phases 'fosse' dans l'histoire du bassin de Soria (Sierra de los Cameros) au Jurassique superieur-Cretace inferieu r : relation entre tecton ique et sedimentation. Centr. Rech. Expl. Prod. ElfAquitaine Bull., 7, 399- 407. SALOMON, J. (1984) Paleopedologie et redistributions dans les formations continentales du Jurassique superieur du Bassin de Soria (Espagne). Abstracts, 5th Eur. Regional Mtg., Int. Ass. Sedimentologists, pp. 393-394. SALOMONS, W., GOUDIE, A. & M OOK, W . G . (1978) Isotopic composition of carbonates in recent sediments and soils from western Europe, Africa and India. Earth Surf Proc., 3, 48- 5 7. SBETA, A.M . (1976) Sedimentology of a Jurassic Carbonate
Belize Shelf-Carbonate Sediments, Clastic Sediments, and Ecology (Ed. by K. F. Wantland & W. C. Pusey). Am. Assoc. Petrol. Geol. Studies in Geology, 2, 5 00-571.
WARD, W.C., FOLK, R.L. & WILSON, J . L. (1970) Blackening of eolianite and caliche adjacent to saline lakes, Isla Mujeres, Quintana Roo, Mexico . J. sedim. Petrol., 40, 5 48555. WELLS, N . A. (1983) Carbonate depos ition , physical limnol ogy and environmentally-controlled chert formation in Paleocene-Eocene Lake Flagstaff, Central Utah . Sedi ment. Geol., 35 , 263- 296. WEST, I . M . (1964) Evaporite diagenesis in the Lower Purbeck beds of Dorset. Proc. Yorks. geol. Soc., 34, 315-330. WEST, I . M . (1975) Evaporites and associated sediments o f the Purbeck Formation (Upper Jurassic) of Dorset. Proc. geol. Ass., 86, 205- 225. WILLIAMS, C.R. & PICARD, M . D . (1974) Petrology o f
Sequence in the Eastern Cantabrian Mountains of Spain.
Unpublished D . Phil Thesis, University of Oxford, 200 pp . SBETA, A. M . (1985) Sedimentology of the marine and lacustrine Jurassic-basal Cretaceous carbonates in the SW Cantabrian mountains of north Spain. Abstracts, 6th Eur. Regional Mtg, Int. Ass. Sedimentologists, pp. 420- 423. ScoFFIN, T.P. (1987) An Introduction to Carbonate Sediments and Rocks. Blackie, Glasgow, 274 pp . SHINN, E . A. (1968) Practical s ignificance of birdseye struc tures in carbonate roGks. J. sedim. Petrol., 38, 215- 223. SHINN, E . A. (1983) Birdseyes, fenestrae, shrinkage pores and loferites-a re-evaluation. J. sedim. Petrol., 53, 619-628.
341
Nigel H. Platt
carbonate rocks of the Green River Formation (Eocene). J. sedim . Petrol., 44, 738- 759. WRIGHT, V.P. (I 983) A rendzina from the Lower Carboni ferous of South Wales. Sedimentology, 30, 159-179. WRIGHT, V.P. (I 986) The role of fungal biomineralization in
the formation of early Carboniferous soil fabrics . Sedimen tology, 33, 8 31-838. WRIGHT, V . P . , PLATT, N.H. & WIMBLEDON, W . A. (1988) Biogenic laminar calcretes: evidence for calcified root mat horizons in paleosols. Sedimentology, 35, 6 03-620.
( Manuscript received 16 June 1988 ; revision received 12 October 1988)
342
REFERENCES
ADAMS, A . E . (1980) Calcrete profiles in the Eyam Lime
modes du development. Docum. Lab. Geol. Fac. Sci. Lyons,
stone (Carboniferous) of Derbyshire: petrology and regional
62, 137-235.
significance. Sedimentology, 27, 651-660.
BOQUET, E ., BORONAT, A. & RAMOS-C ORMENZANA,
ADAMS, A . E . & CossEY , P.J. (1981) Calcrete develop
A. (1973) Production of calcite (calcium carbonate) crystals
ment at the junction between the Martin Limestone and the
by soil bacteria is a general phenomenon. Nature, 246, 527-528.
Red Hill Oolite (Lower Carboniferous), south Cumbria. Proc. Yorks. Geol. Soc., 43, 411-431.
BRAITHWAITE, C . J.R. (1975) Petrology of palaeosols and
ALLEN, B. L. (1985) Micromorphology of Aridisols. In: Soil Micromorphology and Soil
other terrestrial sediments on Aldabra, western Indian
Classification (Ed. by L.A.
Ocean. Phil. Trans. R. Soc. Lond. B, 273, 1-32.
Douglas & M.L. Thompson). Soil Sic. Soc. Am. Spec.
BRAITHWAITE, C.J.R. (1983) Calcrete and other soils in
Pub!. No. 15,197-216.
Quaternary limestones:
ALLEN, J.R. L. (1974a) Studies in fluviatile sedimentation:
structures, processes and appli
cations. J. geol. Soc. Lond., 140,351-363.
implications of pedogenic carbonate units, Lower Old Red
BRAITHWAITE,
Sandstone, Anglo-Welsh outcrop. Geol. J., 9, 181-208.
grain
ALLEN, J.R.L. (1974b) Sedimentology of the Old Red
breakage
C .J.R. in
(1989)
Displacive
sandstones. J.
sedim.
calcite
and
Petrol.,
59,
258-266.
Sandstone (Siluro-Devonian) in the Clee Hills area, Shrop
BREWER, R. (1964) Fabric and Mineral Analysis of Soils.
shire, England. Sedim. Geo/., 12, 73-167.
Wiley,New York, 470pp.
AMUNDSON, R.G., CHADWICK, O.A., SOWERS, J.M.
BROOK, G . A ., FOLKOFF, M . A . & B o x , E .O . (1983)
& DoNER, H. E. (1988) Relationship between climate and
A world model of soil carbon dioxide. Earth Surf. Proc.
vegetation and the stable carbon isotope chemistry of soils in the Eastern Mojave Desert, Nevada. Quat. Res.,
Landfm., 8, 79-88.
BuczYNSKI,
29,
245-254.
C. & CHAFETZ, H.S. (1987) Siliciclastic
grain breakage and displacement due to carbonate crystal
AMUNDSON, R.G., C HADWICK, O . A ., SowERS, J.M.
growth: an example from the Lueders Formation (Permian)
& D o NER,H. E . (1989) The stable isotope geochemistry of
of north-central Texas, USA. Sedimentology, 34, 837-843.
pedogenic carbonates at the Kyle Canyon,Nevada. Soil Sci.
CALLOT, G., GuYON, A. & MousAIN, D. (1985) Inter
Soc. Amer.
J., 53, 201-210.
relations entre aiguilles de calcite et hyphes myceliens.
A RAKE L, A. V. (1982) Genesis of calcrete in Quaternary soil
Agronomie, 5, 209-216.
profiles, Hutt and Leeman lagoons, Western Australia. J.
CALLOT, G. & 1AILLARD, B. (1987) Apports de Ia Ioupe
sedim. Petrol., 52, 109-125.
binoculaire a !'etude des interfaces sol/racine et sol/cham
ARAKEL, A. V. (1986) Evolution of calcrete in palaeodrain
pignon. In Micromorphologie des Sols (Ed. by N. Fedoroff & M.A. Courty), pp. 73-79. Assoc. Fran�aise pour !'Etude
ages of the Lake Napperby area, central Australia. Palaeo
du Sol.
geogr. Paleoclim. Palaeoecol., 54, 283-303.
CALVET, F. (1982) Constructive micrite envelopes developed
ARAKEL, A.V., JACOBSON, G., SALEHI, M. & HILL, C. M. (1989) Silicification of calcrete in palaeodrainage
in vadose continental environments in Pleistocene eolianites
basins of the Australian arid zone. Austr. J. Earth Sci., 36,
of Mallorca (Spain). Acta. Geol. Hispan., 17, 169-178.
73-89.
CALVET,
F. & JuLIA,
R. (1983) Pisoids in the calcite
profiles of Tarragona, north-east Spain. In: Coated Grains
ARAKEL, A.V. & McC oNCHIE, D. (1982) Classification and genesis of calcrete and gypsite lithofacies in palaeo
(Ed. by T.M. Peryt),pp. 456-473. Springer-Verlag,Berlin,
drainage systems of inland Australia and their relationship
Heidelberg, New York.
to carnotite mineralisation. J. sedim. Petrol., 52, 1149-1170.
CALVET, F. PoMAR, L. & EsTEBAN, M. (1975) Las riz
ATKINSON, T.C. (1977) Carbon dioxide in the atmosphere
cretiones del Pleistoceno de Mallorca. Ins/. Invest. Geol.
of the unsaturated zone: an important control on ground
Univ. Barcelona., 30, 35-60.
water hardness in limestones. J. Hydro/., 35, 111-123.
CALVO, J.P., ALONSO, A.M., & GARCIA DEL CURA, M . A . (1986) Depositional sedimentary controls on sepiolite
B AL, L. (1975) Carbonate in soils: a theoretical consideration on, and proposal for its fabric analysis. I and II. Neth. J.
occurrence
Agric. Sci., 23, 18-35, 163-176.
Geogaceta., 1, 25-28.
BEIER, J. A. (1987) Petrographic and geochemical analysis of
CARLISLE,
caliche profiles in a Bahamian Pleistocene Dune. Sedimen tology,
in
Paracuellos
de
Jarama,
Madrid
Basin.
D. (1980) Possible variation in the calcrete
gypcrete uranium model. Open-File Rep. US. Dept. Energy,
34, 991-998.
GJBX-53(80), 38pp. CARLISLE, D. (1983) Concentration of uranium and vana
BERTRAND-SARFATI, J. & MOUSSINE-POUCHKINE, A. (1983) Pedogenic and diagenetic fabrics in the Upper Pro
dium in calcretes and gypcretes. In: Residual Deposits (Ed.
terozoic, Sanyere Formation (Gourma,Mali). Precambrian
by R.C.L. Wilson). Geol. Soc. Lond. Spec. Pub!. No. 11,
Res., 20, 225-242.
BoDERGAT,
185-195.
A.M. (1974) Les microcodiums, milieux et
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
C ER u N G, T. E. (1984) The stable isotopic composition of
343
References modern soil carbonate and its relationship to climate. Earth
calcite. In: Carbonate Cements (Ed. by O.P. Bricker). Johns
Plan. Sci. Lett., 71, 229-240.
Hopkins
University
Studies
in
Geology,
Vol.
19,
pp. 167-168.
CERLING, T . E . & HAY, H.L. (1986) An isotopic study of
FRIEDMAN, E. & O NEIL, J. (1977) Data of Geochemistry,
palaeosol carbonates from Olduvai Gorge. Quat. Res., 25,
'
6th edn. Compilation of stable isotope fractionation factors
63-78.
of geochemical interest. US Geol. Survey Prof. Pap.
GERLING, T .E., QUADE, J., WANG, Y. & BOWMAN,
440-KK.
J. R. (1989) Carbon isotopes in soils and palaeosols as
FREYTET, P. & PLAZIAT, J.C. (1982) Continental carbon
ecology and palaeoecology indicators. Nature, 34, 138-139. CHADWICK, O . A ., HENDRICKS, D .M . & NETTLETON.,
ate sedimentation and pedogenesis- Late Cretaceous and
W. D. (1987) Silica in duric soils. I. A depositional model.
Early Tertiary of southern France. Contr. Sedimentol. 12, 213pp.
Soil. Sci. Soc. Am. J., 51 , 975-982.
GARDNER, L.G. (1984) Carbon and oxygen isotope com
CHADWICK, O . A . & NETTLETON, W.D . (1990) Micro morphological evidence of adhesive and cohesive forces in
position of pedogenic CaC03 from soil profiles in Nevada
soil cementation. In: Soil Micromorphology: A Basic and
and New Mexico, USA. Isotop. Geosc., 2, 55-73. GILE, L.H. & GROSSMAN, R. B. (1979) The Desert Project
Applied Science (Ed. by L.A. Douglas), pp. 207-212.
Elsevier,
Monograph. US Dept. Agric., Soil Conservation Serv.
AMUNDSON,
GILE, L .H., HAWLEY, J.W. & GROSSMAN, R.B. (1981)
R. G. (1989) Morphology of calcic crystals in clast coatings
Soils and Geomorphology in the Basin and Range Area of
Developments
in
Soil
Science
No.
19.
984pp.
Amsterdam. CHADWICK,
O . A .,
SOWERS,
J.M.
&
from four soils in the Mojave Desert region. Soil. Sci. Soc.
Southern New Mexico - Guidebook to the Desert Project.
Am. J., 53, 211-219.
Memoir 39, New Mexico Bureau of Mines and Mineral
CHAFETZ, H.S. &
BuTLER,
J.C. (1980) Petrology of
Resources, 222pp. GILE,
recent caliche pisolites, spherulites and speleothem deposits of central Texas. Sedimentology, 27, 497-518. Mistassini
Regolith,
Quebec,
Canada. Pre
GILE,
cambrian Res., 19, 285-299.
F.F. &
GROSSMAN,
R.B.
L.H.,
PETERSON,
F.F. &
GROSSMAN,
R.B.
(1966) Morphological and genetic sequences of carbonate accumulations in desert soils. Soil Sci., 100, 347-360.
COHEN, A. S. (1982) Palaeoenvironments of root casts from the Koobi
PETERSON,
accumulations. Soil Sci., 99, 74-82.
CHOWN, E.H. & CATY, J.L. (1983) Diagenesis of the Aphebian
L.H.,
(1965) The K horizon: a master soil horizon of carbonate
Formation, Kenya. J.
52,
GoLDSTEIN, R.H. (1988) Paleosols of Late Pennsylvanian
COURTY, M . A ., DHIR, P. & RAGHAVAN, H. (1987) Microfabrics of calcium carbonate accumulations in arid
GoODFRIEND, G . A . & MAGARITZ, M. (1988) Palaeosols
sedim.
Petrol.,
401-414.
strata, New Mexico. Sedimentology, 35, 777-803. and late Pleistocene rainfall fluctuations in the Negev
soils of western India. In: Micromorphologie des Sols
Desert. Nature, 322, 144-146.
(Ed. by N. Fedoroff & M.A. Courty), pp. 227-234. Assoc. Fran�aise pour !'Etude du Sol.
GouDIE, A. S. (1973) Duricrusts in Tropical and Subtropical
CROMACK, K., SOLLINS, P., TODD, R.L., FOGEL, R.,
GouDIE, A. S. (1983) Calcrete. In: Chemical Sediments and
Landscapes. Clarendon Press, Oxford.
TODD, A .W., FENDER, W.M., CROSSLEY, M.A. &
Geomorphology (Ed. by A.S. Goudie & K. Pye), pp. 93-
CROSSLEY, D . A . (1977) The role of oxalic acid in bicar
131. Academic Press, London, New York.
bonate in calcium cycling by fungi and bacteria: some poss
G RAUSTEIN, W. C., CROMACK, K. & SOLLINS, P. (1977)
ible implications for soil animals. Ecol. Bull., 25, 246-252.
Calcium oxalate: occurrence in soils and effect on nutrient
DREES,
L.R. & WILDING,
and geochemical cycles. Science, 198, 1252-1254.
L.P. (1987) Micromorphic
record and interpretations of carbonate forms in the Rolling
GREGG,
sedim. Petrol., 54, 908-931.
DREYER, L., FONTES, J.CH. & RICHE , G. (1987) Iso
HARRISON, R. S. (1977) Caliche profiles: indicators of near
topic approach to calcite dissolution and precipitation in soils under semi-arid conditions. Chern. Ceo/. Geosci. Sec.),
J.M. & SIBLEY, D .F . (1984) Epigenetic dolo
mitization and the origin of xenotopic dolomite textures. J.
Plains of Texas. Geoderma, 40, 157-175.
surface subaerial diagenesis, Barbados, West Indies. Bull.
(Isotop.
66, 307-314.
Can. Petrol. Ceo/., 25, 123-173.
HARRISON,
DucLoux, J., BuTEL, P. & DuPuis, T . (1984) Micro sequence mineralogique des carbonates de calcium dans une
R.S. &
STEINEN,
R.P. (1978) Subaerial
crusts, caliche profiles and breccia horizons. Comparison of
accumulation carbonatee sous galets calcaires, dans !'ouest
some Holocene and Mississippian exposure surface, Bar
de Ia France. Pedologie, 34, 161-177.
bados and Kentucky. Bull. Ceo/. Soc. Am., 89, 385-395.
DucLoux, J. & DuPUIS, T. (1987) Influence de Ia matiere
HAY, R.L. & REEDER, R.J. (1978) Calcretes of Olduvai
organique des sols sur Ia cristallogenese des carbonates de
Gorge and the Ndolanya Beds of Northern Tanzania. Sedi
calcium. In: Micromorphologie des Sols (Ed. by N. Fedoroff
mentology, 25, 649-673.
HAY, R.L. & WIGGINS, B. (1980) Pellets, ooids, sepiolite
& M.A. Courty), pp. 315-321. Assoc. Fran�aise pour
!'Etude du Sol.
and silica in three calcretes of the south-western United
DuRAND, J.H. (1963) Les croutes calcaires et gypseuses en
States. Sedimentology, 27, 559-576. HuBERT, J.F. (1978) Paleosol caliche in the New Haven
Algerie: formation et age. Bull. Ceo/. Soc. Fr., 5, 959-968. EsTEBAN, M. (1974) Caliche textures and Microcodium.
Arkose, Newark Group, Connecticut. Palaeogeogr. Pal
Boll. Soc. Ceo/. Ita/., 92, Suppl. 1973, 105-125.
aeoclim. Palaeocol., 24, 151-168.
HUTTON, J.T. & DIXON, J.C. (1981) The chemistry and·
EsTEBAN, M. & KLAPPA, C.F. (1983) Subaerial exposure environment. Mem. Am. Assoc. Petrol. Ceo/., 33, 1-54.
mineralogy of some South Australian calcretes and associated
F oLK, R.L. (1971) Caliche nodule composed of rhombic
soft carbonates and their dolomitisation. J. geol. Soc. Aust.,
344
References 28, 71-79.
KRUMBEIN, W.E. & GIELE, C. (1979) Calcification in a
JACOBSON, G., ARAKEL, A.V. & CHEN YlliAN (1988)
coccoid - cyanobacterium associated with the formation of desert stromatolites. Sedimefllology, 26, 593-604.
The central Australian groundwater discharge zone: evol
LAT TMAN,
ution of associated calcrete and gypcrete deposits. Austr. J.
L.H.
&
SIMONBERG,
E .M. (1971)
Case
hardening of carbonate alluvium and colluvium, Spring
Earth Sci., 35, 549-565.
Mountains, Nevada. J. sedim. Petrol., 41, 274-270.
JAILLARD, B . (1987) Les structures rhizomorphes calcaires: modele de reorganisation des mineraux du sol par les racines.
LEEDER, M.R. (1975) Pedogenic carbonates and floodplain
Inst. Nat de Ia Rech. Agron. Lab. de Science du Sol,
sediment accretion rates: a quantitative model for alluvial arid-zone lithofacies. Geol. Mag., 112, 257-270.
Montpellier (thesis), 228pp. JAILLARD, B. & CAL LOT, G. (1987) Action des racines sur
LoNGMAN, M.W. (1977) Factors controlling the formation
Ia segregation mineralogique des constituants du sol. In:
of microspar in the Bromide Formation. J. sedim. Petrol., 47, 347-350.
Micromorphologie des Sols (Ed. by N. Fedoroff & M.A.
MACHETTE, M.N. (1985) Calcic soils of the south-western
Courty), pp. 371-375. Assoc. Frans:aise pour !'Etude du Sol.
United
MAIZELS,
crust (caliche) profiles: criteria for subaerial exposure. J.
Geol. Soc. Am. Spec. Pap. No.
203,
J.K.
(1987)
Plio-Pleistocene
raised
channel
systems of the western Sharqiya (Wahiba), Oman. In: Desert
sedim. Petrol., 42, 817-836.
Sediments: Ancient and Modem (Ed. by L.E. Frostick & I.
JONES, B. (1988) The influence of plants and micro-organ
Reid). Geol. Soc. Lond. Spec. Pub!. No. 35, pp. 31-50.
isms on diagenesis in caliche: example from the Pleistocene
MALIVA,
Ironshore Formation on Cayman Brae, British West Indies.
R.G. (1989) Displacive calcite syntaxial over
growths in open marine limestones. J. sedim. Petrol., 59,
Bull. Can. Petrol. Geol., 3 6, 191-201.
397-403.
JoNES, B. & N G, K .- C. (1988) The structure and diagenesis
MANN, A.W. & DEUTSCHER, R.L. (1978) Genesis prin
of rhizoliths from Cayman Brae, British West Indies. J.
ciples for the precipitation of carnotite in calcrete drainages
sedim. Petrol., 58, 1002-1007.
in Western Australia. Econ. Geol., 73, 1724-1737.
JONES, B. & SQUAIR, C . A . (1989) Formation of peloids in
MANN, A.W. & HORWITZ,
plant rootlets, Grand Cayman, British West Indies. J. sedim.
R.C. (1979) Groundwater
calcrete deposits in Australia. Some observations from
Petrol., 59, 457-467.
Western Australia. J. Geol. Soc. Aust., 26, 293-303.
KAHLE, C. F . (1977) Origin of subaerial Holocene calcareous
MANZE, U. & BRUNNACKER, K. (1977) Ober das Verhalten
crusts: role of algae, fungi and sparmicritisation. Sedimen
der Sauerstoff- und Kohlenstoff-Isotope in Klakkrusten und
tology, 24, 413-435.
Kalkstufen des mediterranen Raumes und der Sahara. Z.
KENDALL, e.G. ST C. & WARREN, J. (1987) A review of
Geomorph., 21, 343-353.
the origin and setting of tepees and their associated fabrics. Sedimentology,
States.
pp. 1-21.
JAMES, N .P. (1972) Holocene and Pleistocene calcareous
34, 1007-1027.
MAGARITZ, M. (1986) Environmental changes recorded in the upper Pleistocene along the desert boundary, southern
KLAPPA, C.F. (1978) Biolithogenesis of Microcodium: elu
Israel. Palaeogeogr. Palaeoclim. Palaeoecol., 53, 213-403.
cidation. Sedimentology, 25, 489-522. KLAPPA, C.F. (1979a) Lichen stromatolites: criterion for
MARION,
G.M.,
SCHLESINGER,
W.H. &
FONTEYN,
subaerial exposure and a mechanism for the formation of
P. J. (1985) Caldep: a regional model for soil CaC03
laminar calcrete (caliche). J. sedim. Petrol., 49, 387-400.
(caliche) deposition in south western deserts. Soil Sci., 139, 468-481.
KLAPPA, C . F . (1979b) Calcified filaments in Quaternary cal
MAYER, L., McFADDEN, L . D . & HARDEN, J.W. (1988)
cretes: organo-mineral interactions in the subaerial vadose
Distribution of calcium carbonate in desert soils: a model.
environment. J. sedim. Petrol., 49, 955-968.
Geology, 16, 303-306.
KLAPPA, C . F . (1980a) Rhizoliths in terrestrial carbonates:
McFADDEN, L . D . & TINSLEY, J. (1985) Rate and depth of
classification, recognition, genesis and significance. Sedi mentology, 27, 613-629.
pedogenic-carbonate accumulation in soils: formulation and testing of a compartmefll model. Geol. Soc. Am. Spec. Pap.
KLAPPA, C.F. (1980b) Brecciation textures and tepee struc
No. 203, pp. 23-42.
tures in Quaternary calcrete (caliche) profiles from eastern
McGRATH, D .B . & HAWLEY, J.W. (1987) Geomorphic
Spain: the plant factor in their formation. Geol. J., 15,
evolution and soil-geomorphic relationships in the Socorro
81-89. K LAPPA, C. F. (1983) A process-response model for the for
area, central New Mexico. In: Guidebook to the Socorro
mation of pedogenic calcretes. In: Residual Deposits (Ed. by
Area (Ed. by V.T. McLemore & M.R. Bowie), pp. 58-67.
New Mexico Bureau of Mines and Mineral Resources.
R. C. L. Wilson). Geol. Soc. Lond. Spec. Pub!. No. 11, pp.
McPHERSON, J. G. (1979) Calcrete (caliche) palaeosols in
211-220.
fluvial red beds of the Aztec Siltstone (Upper Devonian),
KNo x , G. F. (1977) Caliche profile formation, Saldanha Bay (South Africa). Sedimentology, 24, 657-674.
South
Victoria
Land,
Antarctica.
Sedim.
Geol.,
22,
267-285.
KRUMBEIN, W.E. (1968) Geomicrobiology and geochem
MILNES, A. R. (1991) Calcrete. In: Rocks and Soils (Ed. by
istry of the 'Nari-Lime-Crust' (Israel). In: Recent Develop ments in Carbonate Sedimentology in Central Europe (Ed. by
I.P. Martini & W. Chesworth). Elsevier, Amsterdam, New
6. M. Friedman), pp. 138-147. Springer-Verlag, Berlin,
York. (In press.) MILNES,
Heidelberg, New York.
A .R. & HUTTON,
J.T. (1983) Calcretes in
Australia. In: Soils: an Australian Viewpoint, pp. 119-162.
KRUMBEIN, W. E. (1979) Calcification by bacteria and algae.
CSIRO. Academic Press, London, New York.
In: Biogeochemical Cycling of Mineral-Forming Elements
M ouNT, J.F. & C oHEN, A . S . (1984) Petrology and geo
(Ed. by P.A. Swaine), pp. 47-68. Elsevier, Amsterdam,
chemistry of rhizoliths from Plio-Pleistocene fluvial and
New York.
345
References marginal lacustrine deposits, East Lake Turkana, Kenya. J.
RANSON, M . D . & BIDWELL, O.W. (1990) Clay move
sedim. Petrol., 54, 263-275.
ment and carbonate accumulation in Ustolls of central
MucHEZ, PH. & VIANE, W. (1987) Dolocretes from the Lower
Carboniferous
of
the
Campine-Brabant
Kansas, USA. In: Soil Micromorphology: a Basic and
Basin,
Applied Science (Ed. by L.A. Douglas). Developments in
NAYLOR, H., T URNER, P., VAUGHAN, D .J. & FALLICK,
READ, J. F. (1974) Calcrete deposits and Quaternary sedi
Belgium. Pedologie, 37, 187-202.
Soil Science Vol. 19, pp. 417-423.
A. E. (1989) The Cherty Rock, Elgin: a petrographic and
ments, Edel Province, Shark Bay, Western Australia. Mem.
isotopic study of a Permo-Triassic calcrete. Geol. J., 24,
Am. Assoc. Petrol. Geol., 22, 250-282.
205-221.
READ, J. F. (1976) Calcretes and their distinction from stro
N ETTERBERG, F. (1967) Some road making properties of
matolites. In: Stromatolites (Ed. by M.R. Walter), pp. 55-
South African calcretes. Proc. 4th Reg. Conf. African Soil
71. Elsevier, Amsterdam, New York.
Mech. Fndn. Engng., Cape Town, 1, 77-81.
REEVES, C.C. (1976) Caliche: Origin, Classification, Mor
N ETTERBERG, F. (1980) Geology of southern African cal cretes. I. Terminology,
description,
phology and Uses. Estacado Books, Texas.
macrofeatures and
REEVES, C . C . (1983) Pliocene channel calcrete and suspen
classification. Trans. Geol. Soc. S. Afr., 83, 255-283.
parallel drainage in West Texas and New Mexico. In: Re
NETTLETON, W .O. & PETERSON, F .F . (1983) Aridisols.
sidual Deposits (Ed. by R.C.L. Wilson). Geol. Soc. Lond.
In: Pedogenesis and Soil Taxonomy. II. The Soil Orders
Spec. Pub!. No. 11, pp. 179-183. REHEIS, M .C. (1987) Gypsic soils on the Kane Alluvial Fans,
(Ed. by L.P. Wilding, N.E. Smeck & G.F. Hall), pp. 165215. Elsevier, Amsterdam, New York.
Big
Honi
County,
Wyoming.
US
Geol. Surv. Bull.
1590-C, 39pp.
PARADA, C. B., LONG, A. & DAVIS, S .N . (1983) Stable isotopic composition of soil carbon dioxide in the Tucson
RIDING, R. & WRIGHT, V.P. (1981) Paleosols and tidal
Basin, Arizona, USA. Isotop. Geosc., 1, 219-236.
fiat/lagoonal sequences on a Carboniferous carbonate shelf.
PENTECOST, A . & TERRY, C. (1989) Inability to demon
J. sedim. Petrol., 51, 1323-1339.
strate calcite precipitation by bacterial isolates from traver
RoBBIN, D .M. & STIPP, J.J. (1979) Depositional rate of
tine. Geomicrobiology J., 6, 129-135.
laminated soilstone crusts, Florida Keys. J. sedim. Petrol.,
PHILLIPS, S .E . & MILNES, A .R. (1988) The Pleistocene
49, 175-180.
ROBINSON, D. & WRIGHT, V.P. (1987) Ordered illite/
terrestrial carbonate mantle on the south-eastern margin of the St Vincent Basin, South Australia. Austr. J. Earth Sci.,
smectite and kaolinite/smectite as possible primary minerals
35, 463-481.
in a Lower Carboniferous paleosol sequence, South Wales? Clay Minerals, 22, 109-118.
PHILLIPS, S.E., MILNES, A.R. & FOSTER, R.C. (1987) Calcified filaments: an example of biological influences in
SAIGAL, G. C. & WALTON, E. K. (1988) On the occurrence
the formation of calcrete in South Australia. Aust. J. Soil
of displacive calcite in Lower Old Red Sandstone of Car
Res., 25, 405-428.
noustie, eastern Scotland. J. sedim. Petrol., 58, 131-135.
PHILLIPS, S .E . & SELF, P.G. (1987) Morphology, cry·
SALOMONS, W., GOUDIE, A. & MOOK, W.G. (1978)
stallograpliy and origin of needle-fibre calcite in Quaternary
Isotopic composition of calcrete deposits from Europe,
pedogenic carbonates of South Australia. Aust. J. Soil Res.,
Africa and India. Earth Surf. Proc., 3, 43-57. SALOMONS,
25, 429-444.
PLATT, N.H. (1989) Lacustrine carbonates and pedogenesis:
W. &
MooK,
W. G. (1986) Isotope geo
chemistry of carbonates in the weathering zone. In: Hand
2 (Ed.
sedimentology and origin of palustrine deposits from the
book of Environmental Isotope Geochemistry, Vol.
Early Cretaceous Rupelo Formation, West Cameros Basin,
by P. Fritz & J. Ch. Fontes), pp. 239-269. Elsevier, Am
North Spain. Sedimentology, 36, 665-684.
sterdam, New York.
QuADE, J., CERLING, T.E. & BOWMAN, J.R. (1989a)
SARKAR, S. (1988) Petrology of caliche-derived peloidal cal
Development of Asian monsoon revealed by marked ecolo
cirudite/calcarenite in the late Triassic Maleri Formation of
gical shift during the latest Miocene in northern Pakistan.
the Pranhita-Godavari Valley, south India. Sedim. Geol.,
Nature, 342, 163-166.
55, 263-282.
QuADE, J., CERLING, T .E . & BOWMAN, J.R. (1989b)
ScHLESINGER,
W .H. (1985) The formation of caliche
Systematic variations in the carbon and oxygen isotopic
in
composition of pedogenic carbonate along elevation tran
Cosmochim. Acta., 49, 57-66.
sects in the southern Great Basin, United States. Geol. Soc.
soils
SEARL,
of
the
Mojave
Desert,
California. Geochim.
A. (1988) Pedogenic dolomites from the Oolite
Group (Lower Carboniferous), South Wales. Geol. J., 23,
Am. Bull., 101, 464-475.
147-156.
RABENHORST, M.C. & WILDING, L.P. (1986) Pedogene
SEHGAL,
sis on the Edwards Plateau, Texas. III. New model for the
J.L. &
STOOPS,
G. (1972) Pedogenic calcite
formation of petrocalcic horizons. Soil Sci. Soc. Am. J., 50,
accumulations in arid and semi-arid regions of the Indo
693-699.
Gangetic
RABENHORST, M.C., WILDING, L.P. & WEST, L.T. (1984) Identification of pedogenic carbonates using isotope and microfabric
analyses. Soil
Sci.
Soc.
Am.
J.,
alluvial
plain
of
Erstwhile
Punjab
(India).
Geoderma, 8, 59-72.
SEMENIUK, V. (1986) Calcrete breccia fioatstone in Holo cene sand developed by storm-uprooted trees. Sedim Geol.,
48,
48, 183-192.
125-132. RAGHAVAN, H. & CouRTY, M . A . (1987) Holocene and
SEMENIUK,
V. & MEAGHER, T . D . (1981) Calcrete in
Pleistocene environments in the Thar Desert (Didwana,
Quaternary coastal dunes in south western Australia: a
India). In: Micromorphologie des Sols (Ed. by N. Fedoroff
capillary-rise phenomenon associated with plants. J. sedim.
& M.A. Courty), pp. 371-375. Assoc. Fran�aise pour !'Etude du Sol.
SHINN, E . A . & LIDz, B.H. (1988) Blackened limestone
Petrol., 51, 47-68.
346
References pebbles: fire at subaerial unconformities. In: Paleokarst (Ed.
posure of Carboniferous carbonates, north eastern Kentucky.
by N.P. James and P.W. Choquette), pp. 117-131. Springer
Sedimentology, 22, 417-440.
Verlag, Berlin, Heidelberg, New York. S HLEMON,
WATTS, N. L. (1977) Pseudo-anticlines and other structures
R.J. (1978) Quaternary soil-geomorphic re
in some calcretes of Botswana and South Africa. Earth Surf. Proc. Landfm., 2, 63-74.
lationships, south eastern Mojave Desert, California and Arizona. In Quaternary Soils (Ed. by W.C. Mahaney), pp.
WATTS, N. L. (1978) Displacive calcite: evidence from recent
187-207. Geo Abstracts, Norwich. SINGER,
and ancient calcretes. Geology, 6, 699-703.
A. (1988) Illite in aridic soils, desert dusts and
WATTS, N. L . (1980) Quaternary pedogenic calcretes from
desert loess. Sedim. Geol., 59, 251-259.
the Kalahari (southern Africa): mineralogy, genesis and dia
SOBECKI, T .M. & WILDING, L.P. (1982) Calcic horizon
genesis. Sedimentology, 27, 661-686.
distribution and soil classification in selected soils of the
WIEDER, M. & YAALON, D .H. (1974) Effect of matrix
Texas Coast Prairie. Soil Sci. Am. J., 46, 1222-1227.
composition on carbonate nodule crystallization. Geoderma,
SOBECKI, T.M. & WILDING, L.P. (1983) Formation of
11, 95-121.
calcic and argillic horizons in selected soils of the Texas
WIEDER, M. & YAALON, D .H. (1982) Micromorpholog
Coast Prairie. Soil Sci. Am. J., 47, 707-715.
ical fabrics and developmental stages of carbonate nodular
S oiL SuRVEY STAFF (1975) Soil taxonomy: a basic system of
forms
USDA -SCS Agric. Handbook No. 436. US Government
to
soil
characteristics.
Geoderma,
28,
WRIGHT, V.P. (1982) Calcrete palaeosols from the Lower
Printing Office, Washington, DC. 754pp.
Carboniferous Llanelly Formation, South Wales. Sedim.
SOLOMON, S .T . & WALKDEN, G.M. (1985) The appli
Geol., 33, 1-33.
cation of cathodoluminescence to interpreting the diagenesis
WRIGHT, V.P. (1983) A rendzina from the early Carboni ferous of South Wales. Sedimentology, 30, 159-179.
of an ancient calcrete profile. Sedimentology, 32 , 877-896. SouTHGATE,
related
203-220.
soil classification for making and interpreting soil surveys.
R.N. (1986) Cambrian phoscrete profiles,
WRIGHT, V. P. (1984) The significance of needle-fibre calcite
coated grains and microbial processes in phosphogenesis,
in a Lower Carboniferous palaeosol. Geol. J., 19, 23-32.
Georgina Basin, Australia. J. sedim. Petrol., 56, 429-441.
WRIGHT, V. P. (1986a) The role of fungal biomineralization
STEEL, R.J. (1974) Cornstone (fossil caliche): its origin,
in the formation of early Carboniferous soil fabrics. Sedi
stratigraphic and sedimentological importance in the New
mentology, 33, 831-838.
Red Sandstone, west Scotland. J. Geol., 82, 351-369.
WRIGHT, V. P. (1986b) Pyrite formation and the drowning of
STRASsER, A. (1984) Black pebble occurrence and genesis in
a paleosol. Geol. J., 21, 139-149.
Holocene carbonate sediments (Florida Keys, Bahamas and
WRIGHT, V.P. (1989) Terrestrial stromatolites and laminar
Tunisia). J. sedim. Petrol., 54, 1097-1109.
calcretes: a review. Sedim. Geol., 65, 1-13.
STRASSER, A. & DAVAUD, E. (1983) Black pebbles of the
WRIGHT, V. P. (1990a) Syngenetic formation of grainstones
Purbeckian (Swiss and French Jura): lithology, geochemistry
and pisolites from fenestral carbonates in peritidal settings: discussion. J. sedim. Petrol., 60, 309-310.
and origin. Ecol. geol. Helv., 76, 551-580. SUCHECKI, R.K . , HUBERT, J.F. & BIRNEY DE WET,
WRIGHT,
C. C. (1988) Isotopic imprint of climate and hydrogeo
V. P. (1990b) Estimating rates of calcrete for
mation and sediment accretion in ancient alluvial deposits.
chemistry on terrestrial strata of the Triassic-Jurassic Hart
Geol. Mag., 127, 273-276.
WRIGHT, V.P. (1990c) A micromorphological classification
ford and Fundy Rift Basins. J. sedim. Petrol., 58, 801-811. TALMA, A.S. & NETTERBERG, F. (1983) Stable isotope
of fossil and recent calcic and petrocalcic microstructures In:
abundances in calcretes. In: Residual Deposits (Ed. by R.C.
Soil Micromorphology: A Basic and Applied Science (Ed. by
L. Wilson). Geol. Soc. Lond. Spec. Pub. No. 13, pp.
L.A. Douglas). Developments in Soil Science, Vol. 19, pp.
221-233. TANDON,
401-407. Elsevier, Amsterdam. S .K . &
FRIEND,
P.F . (1989) Near-surface
shrinkage and carbonate. replacement processes,
WRIGHT, V.P. & PEETERS, C. (1989) Origins of some early
Arran
Carboniferous calcrete fabrics revealed by cathodolumine
36,
scence: implications for interpreting the sites of calcrete
S .K . & NARAYAN, D . (1981) Calcrete con
WRIGHT, V.P., PLATT, N .H . & WIMBLEDON, W.A.
Cornstone
Formation,
Scotland.
Sedimentology,
1113-1126. TANDON,
formation. Sed. Geol., 65, 345-353.
glomerate, a case-hardened conglomerate and cornstone- a
(1988) Biogenic laminar calcretes: evidence of calcified root
comparative
mat horizons in paleosols. Sedimentology, 35 , 603-620.
account
of
pedogenic
and
non-pedogenic
carbonates from the continental Siwalik Group, Punjab,
WRIGHT, V.P. & ROBINSON, D. (1988) Early Carboni
India. Sedimentology, 28, 353-367.
ferous floodplain deposits from South Wales: a case study of
TuCKER, M.E. & WRIGHT, V.P. (1990) Carbonate Sedi mentology. Blackwell Scientific· Publications, Oxford,
the controls on paleosol development. J. geol. Soc. Lond., 145, 847-857.
424pp.
WRIGHT, V.P. & WILSON, R.C.L. (1987) A terra-rosa like complex from the Upper Jurassic of Portugal. Sedi
WALKDEN, G.M. (1974) Palaeokarstic surfaces in Upper Visean (Carboniferous) limestone of the Derbyshire Dome.
mentology,
34, 259-273.
Y AALON, D.H. (1988) Calcic horizon and calcrete in Aridic
J. sedim. Petrol., 44, 1232-1247.
soils and paleosols: progress in the last twenty two years.
WALLS, R.A., HARRIS, W.B. & NUNAN, W.E. (1975) Calcareous crust (caliche) profiles and early subaerial ex-
Soil Sci. Soc. Am. Agronomy Abstr.
347
Index
Entries in
107,109
acicular networks
bold type refer to tables; italic type indicates figures
Arran Cornstone Formation
aeolianite Bahamas
198,199,201
bridges
153,154,155 Saldanha Bay 99,100, 106 Tanzania 35,36, 37 algae 211, 212,213 chasmolithic 211, 212, 213 endolithic 211, 212,213
petrology of calcrete profile 302-5
rhizoliths in
metabolism of
calcarenite
petrography
Holocene origin
calcification
subaerial diagenesis
209
calcite
16-17, 95 171
169, 170,
and fungal biomineralization
191,
193, 194-5
bridge and point-contact
138-9
3
calcite
26
chasmolithic algae·
213
non-luminescent
chasmolithic fungi
215
sandstone
endolithic algae
213
spar
endolithic fungi
215
Cerion
79,80,82,83, 86-7
306, 307,308, 309
289,291
306 198,203
analcime
20
needle-fibre see needle-fibre calcite
chasmoliths
precipitation
chemistry of calcretes
213
chasmolithic fungi
215
Kalahari Aridisols augite
79, 88
calcium oxalate
42
57,58
236,237,238
beta calcretes
Bahamas 197-204
6
petrography
22
South Africa
15, 17-18 95-204
197-204
Eyam Limestone
189-96
laminar calcretes see laminar calcretes, biological activity 115-48
rhizoliths
149-65
Saldanha Bay South Wales bird's eye vugs black pebbles
301-20 177-87,245,246,247
297 293,294 293,
294,295
190,
191,192 264,265,266-7,273 265,268,273, 275 petrology 269,270
conglomerate
definition
265
274 267, 268,269,270 sandstone association 249,252 development of
301-20
306-9
cementation and calcrete fabric
307,308, 309
formation micrite fabrics
349
petrology cornstone
302, 304
265
Arran Formation rod
318
needle-fibre calcite formation
Calcretes Edited by V. Paul Wright and Maurice E.Tucker © 1991 The International Association of Sedimentologists ISBN: 978-0-632-03187-0
107, 201
case-hardened
general features
history
14,36, 37
caliche profiles
293-7
295,296,
carbonate replacement
spar-micrite growth patterns
107,109,110-11 boulder/cobble calcrete 3 brecciation 12,305,309, 310
79,81-2,85
52
and fungal biomineralization
Arran Cornstone Formation
137-8
clotted texture coated grains
cathodoluminescence
cementation
Saldanha Bay
189-96
South Wales
and calcrete diagenesis
Microcodium
77,78,81-2
clinoptilolite,Kalahari
177-87
215,216
4
18,19
Kalahari
18
167-76
330
borings
clays
6
3
morphological
North Wales
10,12,14-15,321,327,
-
pedogenic calcretes
fungal biomineralization
97-114
329 fungal
200
323-42
Carboniferous calcretes
5 6
4,
micromorphological
97-114
Eyam Limestone
167-76
fungal biomineralization
Microcodium
hydrological
167-76
Cameros Basin, Spain
biogenic features of calcretes Bahamas
197-204
Eyam Limestone
6
6
dolomite content
organic carbon content
isotopic composition
3-5
calcretes and dolocretes
133,137
calcrete profiles Bahamas,caliche profiles
213
152
Chondrites classification
20
44, 45, 46,47
79
213
Chlorococcus
in Microcodium
baeocytes
209
Chlorococcum
138-9
154
southwestern United States
3,12
43
chlorite, Kalahari
139-40
and plant growth rhizoliths
chabazite
138-40
and plant decay
chasmolithic algae
308, 309
42-3
Kalahari
aragonite
105, 106
bright orange luminescent
213 236,241
313-18
cements
139-40
Anacystis
Aphanothece spp.
306
subaerial micritization
138-40
and plant growth
17-18
306,307,
spar cement stratigraphy
308, 309
and plant decay calcified soil
304,305
secondary porosity
220-1
6
Eyam Limestone
3
207-9
terminology
222
alveolar septal structure
311, 312, 313
106
cal<>areous crusts
211, 213 224,304, 312
allochems
309,310,
recurring dissolution
treppelithic
alpha calcretes
313-18
291
270 105, 106,304
Siwalik Group
287-300
292,293
Siwalik Group petrology
269-70,273,
271,272-3
275
Index
crumb aggregates crystallaria
illite
52
17
cyanobacteria
7, 15,205
calcification in
77-8,82
Tanzania
36
isotope ratios
231-42
Bahamas
214-15 fungi
20-2 199,200
336, 337-8
306
diamictites
249-56 249,252
152 222-4
allochems
petrography
312-13
non-luminescent subaerial vadose cements recurring
classification
Tanzania
69-94
43
open
209
evaporitic facies
biogenic
332
formation of calcretes
crusts
alveolar textures
191, 193,194
geological setting grain coats
pedogenic
189-90
212, 213,214-15
213,214-15
and Microcodium
rhizolite surficial types of
293,
294 palustrine carbonate fabrics 78-9,82
G/obater incrassatus
324
Gloeocapsa spp.
330
255
mineralogy 245,246,247
petrography Recent
12
hardpan calcrete
246,248,249,250
honeycomb calcrete 215
3
127,129, 136 127,129,136 120-1,122,
123,124,126 terminology
Kalahari
12,13-15
120
algae
biogenic aspects
preliminary observations
58, 59
217,218
211, 212,213
alteration into calcareous crusts
350
Kalahari
128, 132 115
79
Saldanha Bay 215,
108,109
Tanzania
42-3
mineralogy
20-2
133-4
136-8
microspar
207-29
217, 218,219
ultrastructure
Microcodium elegans
38,39,40,41-3
133-8
134-6
synthesis of model
88-9
aggrading recrystallization 207-29
sae
and terrestrial vegetation
Miami Limestone
3
shape
141-2
125-6
Southeastern Spain
1
massive calcretes
Holocene calcareous crusts
126-32
120-1, 122, 123, 124, 126
sample preparation
southwestern United States 154
125,126,127,
comparisons with Ancient
37,38 244
magnesium
8, 9-10
125-33
132-3
128, 129,131
243-60
macrofeatures of calcretes
236,241
groundwater calcretes
120, 121, 125
optical properties
244-5
143-4
119, 120
laboratory analysis
249,251, 252-3, laterites
117,118,
Ibiza-Pleistocene
172
254-5,256
glauconite, Kalahari
120-5
geological importance
Upper Jurassic-Lower Cretaceous
Arran Cornstone Formation
Humicoloa
setting
244,245
245-55
Upper Jurassic 14,17,23
1 1 6, 141-2
116, 141-2
geographic location and geological
244
Tanzania
134-5,137-8
138-40
epis de mas
231-42
Lower Carboniferous
190
190, 191, 192
needle-fibre calcite chasmolithic
rhizoliths
207-29
141-2
128,130, 136
colonies en laminae
history
interpretation
140-1
calcification
Eyam Limestone
189-96
6, 13, 16-17, 1 1 5-48 126
arrangement
243-5
description of
7, 15,212,21 3,214-15
318 209, 215, 217-18,224-5
comparisons with Recent
205-60
304
106,107,109
308, 309,313-18
Ancient
capillary rise-zone
8
304
304 302,304
abundance
243-60
stromatolites
152
biomineralization
gypsum
Microcodium
calcification in desert
209 289,290, 291, 297
glaebules
323-42
3, 14
calcareous crusts
11
endolithic
micritization
327,328
biological activity
fungi
331
326-7,335
laminar calcretes·
240
Entophysalis granulosa
Fucosopsis
79
pore lining and bridging fabric
328-9, 330-1,335-6
reworked
240
191,
partially micritized allochems
subaerial
Entophysalis spp.
fissures
304
stratigraphy
lacustrine carbonates marginal
filaments
304
floating grain fabrics
petrography
272-3
79
Rupelo Formation
105,106
ferricretes
201, 203
dense laminar fabric
Saldanha Bay
Kalahari
293,
294,295,296,297
Kalahari 252, 253
kaolinite
6
79,80,83,88
endoliths
255,256
unconformity association Kalahari calcretes
Kalahari druses
209
193, 194, 195
kankar, Siwalik Group
6
dolomite
micrites
pelleted fabric
1,8
217, 218, 219
215,216
and fungal biomineralization
311-12
309-13
dolocrete
252-3,254 252, 253,255
312
micrite substrates
249,
palustrine carbonate association
fibrous calcite walls of marine
worms
Bahamas
marllpalaeosol association
dissolution and concurrent precipitation
215,216
Arran Cornstone Formation
Conglomerate/sandstone association
289,290,292
Diplocraterion
2, 246,248,249,250
Jurassic- Lower Cretaceous calcretes
90
215, 217-18
plant fibres
sparmicritization Jurassic calcretes
301-20 subaerial
209,210
micritization
cathodoluminescence studies Kalahari
212, 213,214-15
geology
Rupelo Formation
diagenesis
211, 212-13,
biogenic structures
Kalahari
Index
Arran Cornstone Formation Kalahari calcretes
182, 184-5 179-85
per descensum model petrocalcic horizons
44
montmorillonite
Kyle Canyon
77,82
morphology
petrography
19-20
caliche profiles
102
needle-fibre calcite
316,317
16,95
56,57
Bahamas
317
313, 314,315,316 313-18
fungal origin
189-90
nephelinite
316-17
201, 202-3 168,169,170,171, 255,256
replacement 287-300
rhizoliths
pisolites
154,156-61
264,266,267,268,269,
263-77
Pleurocapsa spp. 56, 58,
precipitation
39,40, 42
opal,southwestern United States
236,237,239,240 240
silica
3
26,40,42
isotopic values
1,2,12, 23-94,
Tanzania
328-9,330-1 8,9,20
southwestern United States
59
1,3
10,11-12
pedotubule calcrete
3 171-2
southwestern United States 59-60,61,62-3
169,170
293,
294,295 caliche profiles cements
108,109
306,307, 308-9
sparmicritization, 209,217, 218, 219,
255
rhizoliths
13-14, 149-65
branching
genesis of
concomitant dissolution precipitation
rhizolites
164
151-4
161, 163-4
351
matter
222,223-4 224
metabolism of algae
150-1
internal structures
222-4
decomposition of organic kinetics
151, 152
field characteristics 58,
20
spar
226-7
classification
179-85
263-77
152
see also microspar; sparmicritization
150-1,156,157, 161,
associated features
26,42,52,304
Eyam Limestone South Wales
179,180,
181-2,183-5
Eyam Limestone
26-7
pelleted calcrete
177,178
163
69-94
soil profiles
178,179
peloidal fenestral unit rhizocretions
4
classification
279-85
282,283,284
Arran Cornstone Formation
field appearance geological setting
76-7, 78,82-3,84,85
pedogenic calcretes
Tanzania
smectite 177-87
323-42
Siwalik Group, India Skolithos
rendzina,Lower Carboniferous
sedimentology and origin
Kalahari
51-68
25-49
253,255,
321-42
76-7, 78-9, 82-3,
84,85 proposed process
southwestern United States
21-2
palustrine carbonates
palygorskite
69-94
58,59
43-4
siliciclastic grain breakage
Kalahari
213
Kalahari
79,80-1,82,85-6
silica-sesquioxide gel
261
marginal
1, 11
silicates, Kalahari
1, 10
57, 58,
12
southwestern United States
Quaternary calcretes
152
palaeo-calcretes Palmel/a
77,82-3,84,85
Kalahari
58-9 Ophiomorpha
281,282
8,9,20,23-4
59 silcretes
6
pseudo-ooids
sepiolite
southwestern United States
105,106
289,
291
Kalahari
236
point-contact cement powder calcrete
59-60, 63,64, 65,66-7
134-7
Arran Cornstone Formation Lueders Formation
Pleurocapsa fuliginosa
271
southwestern United States
151,153,154,156,162,
164,305,309
59-60,
215,216
ooids
26
156,158,159,160,
sandstones
267,268,269
31,32,33-4,37,38
plant fibres
4
151, 154,162
root petrifactions
roots,and Microcodium
southwestern United States
Plectonema spp.
Tanzania
151, 153,156 255,257, 258
root tubules
199, 201, 202
Siwalik Group
Olduvai Gorge see Tanzania calcretes Siwalik Group
root casts
26
Tanzania Ogallala cap rock
150-1
152
root mats
279-85
63,64, 65,66-7
270
152, 154
terminology
43
Bahamas
non-pedogenic carbonates, Siwalik
151, 153,155
substrate
root moulds
313-14,315,
152,153, 154, 156,162
161
phillipsite
3
156,157,158,159,
151,152,153
types of
near-surface shrinkage and carbonate
Siwalik Group
317
28, 42
Group
size
105,106,107,108
siliciclastic grain breakage
18
nodular calcrete
root tubules
316
110-11
and subaerial micrite neomorphism
301-20
needle-fibre calcites
origin of morphology
152, 154,162
root petrifactions
shape
laminar calcretes
formation of
156,157
152,156
root moulds
172,173-4
202, 203
154, 156-61
160, 161
Eyam Limestone
cathodoluminescence petrography
Saldanha Bay
root casts
52
15, 16-18, 19-20,
South Africa
Bahamas
151-2,153,155
rhizocretions
caliche profiles
ancient calcretes
154
orientation petrography
8
calcrete diagenesis
needle-fibre calcites
154,155
mineralogy
261-320
3
classification
location
179,180,181,182,183
South Wales 35-44
3
Kalahari
10, 17, 23, 95
interpretation
154
Tanzania calcretes Mollisols
peloids grains
132-3
Microcodiuin rhizoliths
289
76-81
154,157
nature of
222
221-2,223
stromatolites,calcification in methods of study
232
231-42
Index
41, 42-3
structure grumeleuse Synechococcus spp.
236
United States, Quaternary calcretes
vadose cements Tanzania calcretes tepees
textural inversion
15
Miami Limestone Formation subaerial, dissolution of
209
105, 106,
Eyam Limestone Formation
172,
173
14-15
treppeliths tubules
25-49
107, 109
Vertisols vugs
51-68
310,
3
108, 109, 330, 332
Wahiba Sands worms
215,
224-5 zeolites
38
Zoophycos
311-12
352
8
216
152
