Limestones
of carbonate sediments therefore occur in the tropi-cal–subtropical belt, some 30° north and south of theEquator, and most limestones of the Phanerozoicformed in low latitudes. Biogenic carbonate produc-tivity is highest in sea water of normal salinity in theshallow (less than 10m), agitated part of the photiczone (the depth down to which light penetrates, of theorder of 100–200m). Skeletal carbonate sands dooccur in higher latitudes, such as along the westerncoast of Ireland and Norway where calcareous redalgae (especially Lithothamnion) and Mollusca domi-nate the sediments, and also off southern Australia,where bryozoans are especially important. However,there are few ancient examples of these so-called cool-water carbonates (see papers in James & Clarke(1997) and Lukasik et al. (2000) for a Miocene exam-ple from South Australia).
Non-skeletal grains, such as ooids and lime mud, areprecipitated only in the warm shallow waters of thetropics. In the deeper-water pelagic environment, cal-careous oozes are developed extensively and they arecomposed principally of the skeletons of pelagic or-ganisms, Foraminifera and coccoliths, which live inthe photic zone. High rates of carbonate dissolution atdepths of several thousand metres result in little car-bonate deposition below this, the carbonate compen-sation depth. Limestones also form in lakes and soils.One overriding control on carbonate deposition is theinput of siliciclastic material. Many carbonate-producing organisms cannot tolerate the influx oflarge quantities of terrigenous mud.
For the petrographic study of limestones thin-sections are examined routinely, but acetate peels areuseful too and quick to make. The surfaces of a lime-stone are polished and then etched with dilute acid(5% HCl or acetic acid for 30–60s). The surface isthen covered in acetone and a piece of acetate sheet isrolled on, with care taken not to trap air bubbles. Afterat least 10min, the acetate is peeled off, and placed be-tween glass. It is now ready for the microscope. All textural details are faithfully replicated, but of course
110
4.1 Introduction
Biological and biochemical processes are dominant inthe formation of carbonate sediments, although in-organic precipitation of CaCO3 from seawater alsotakes place. Once deposited, the chemical and physicalprocesses of diagenesis can considerably modify thecarbonate sediment. Limestones occur throughout theworld in every geological period from the Cambrianonwards and reflect the changing fortunes, throughevolution and extinction, of invertebrates with car-bonate skeletons. In the Precambrian, carbonates arealso abundant, but they are commonly dolomite and many contain stromatolites, produced largely by microbes, especially the cyanobacteria (‘blue–greenalgae’).
The economic importance of limestones today lieschiefly in their reservoir properties, as about half of theworld’s major petroleum reserves are contained withincarbonate rocks. Limestones also are hosts to epige-netic lead and zinc sulphide deposits of the MississippiValley type and they have a wide variety of chemi-cal and industrial uses, including the manufacture of cement.
As a result of recent geological events, notably thePleistocene glaciation and a global sea-level lowstand,shallow-marine carbonate sediments are not widelydeveloped at the present time. In the past, shallowepeiric seas periodically covered vast continental areasso that limestones were deposited over many thou-sands of square kilometres. On a broad scale, ex-tensive carbonate deposition correlates with globalsea-level highstands. Organisms with carbonate skele-tons occur throughout the world’s seas and oceans so that carbonate sediments can develop anywhere.However, there are several factors, of which the mostimportant are temperature, salinity, water depth andsiliciclastic input, that control carbonate deposition.Many carbonate skeletal organisms, such as the reef-building corals and many calcareous green algae, re-quire warm waters in which to flourish. The majority
Compiled by : S.N.Mohapatra
Source : Sedimentary Petrology (Tucker)
Limestones 111
polarizers cannot be crossed. See Miller (1988a) forfurther details.
Much petrographic information is obtained fromthin-sections viewed in transmitted light but furtherdetail can be revealed through the technique ofcathodoluminescence (CL): bombarding a polishedthin-section (no cover slip) with electrons in a specialsmall vacuum chamber mounted on a microscopestage (see Marshall (1988) and Miller (1988b) for details). An example of CL is given in Plate 13c. Examining a thin-section under UV light and observ-ing the fluorescence can also reveal ‘hidden’ textures(see Dravis & Yurewicz, 1985). It is common practiceto stain a thin-section of a limestone, or the polishedsurface of a hand specimen before taking a peel, toshow the mineralogy (calcite or dolomite), and theiron content (ferroan or non-ferroan). Alizarin Red Sand potassium ferricyanide are used (for the recipe see Miller, 1988a), and calcite (non-ferroan) stains pink and ferroan calcite blue to mauve; dolomite (non-ferroan) does not take up the stain whereas ferroandolomite is turquoise–blue. Table 4.1 summarizes the features to look for in a carbonate thin-section and Table 4.2 gives a simple table for notes when describing a slide.
4.2 Mineralogy of carbonate sediments
In Recent sediments, two calcium carbonate mineralsdominate: aragonite (orthorhombic) and calcite (trigonal). Two types of calcite are recognized depend-ing on the magnesium content: low-magnesium calcitewith less than 4 mol.% MgCO3 and high-magnesiumcalcite with greater than 4 mol.%, but typically rang-ing between 11 and 19 mol.% MgCO3. By compari-son, aragonite normally has a very low Mg content(less than 5000p.p.m.) but it may contain up to 10000p.p.m. (1%) strontium, substituting for calcium. Themineralogy of a modern carbonate sediment dependslargely on the skeletal and non-skeletal grains present.Carbonate skeletons of organisms have a specific mineralogy or mixture of mineralogies (Table 4.3), al-though the actual magnesium content of the calcitesmay vary, being partly dependent on ambient watertemperature.
Aragonite is unstable at surface temperatures andpressures and in time high-Mg calcite loses its Mg.Thus all carbonate sediments with their original mixedmineralogy are converted to low-Mg calcite during
diagenesis. Grains and cement composed originally oflow-Mg calcite generally are perfectly preserved inlimestones; those originally of high-Mg calcite aremostly well preserved, like low-Mg calcite, but theymay show some microstructural alteration and minordissolution. Grains of aragonite are either replaced bycalcite with some retention of original structure (theprocess of calcitization), or dissolved out completelyto leave a mould, which later may be filled with calcite(a cement). A limestone also may be dolomitized,whereby dolomite, CaMg(CO3)2, replaces the CaCO3minerals and is precipitated as a cement (dolomitiza-tion is discussed in Section 4.8). Non-carbonate min-erals in limestones include terrigenous quartz and clay, and pyrite, hematite, chert and phosphate of diagenetic origin. Evaporite minerals, in particular gypsum–anhydrite, may be closely associated withlimestones (see Chapter 5).
4.3 Components of limestones
Limestones are very varied in composition but broadly the components can be divided into fourgroups: (i) non-skeletal grains, (ii) skeletal grains, (iii)micrite and (iv) cement. The common cement, sparite,and others are discussed in the section on diagenesis(Section 4.7).
4.3.1 Non-skeletal grains
Ooids and pisoids
Modern ooids are spherical–subspherical grains, consisting of one or more regular concentric lamellaearound a nucleus, usually a carbonate particle orquartz grain (Figs 4.1 & 4.2). Sediment composed ofooids is referred to as an oolite. The term ooid (for-merly oolith) has been restricted to grains less than 2mm in diameter and the term pisoid (formerly pisolith)is used for similar grains of a larger diameter. If onlyone lamella is developed around a nucleus, then theterm superficial ooid is applied (Fig. 4.1). Compositeooids consist of several small ooids enveloped by con-centric lamellae. Coated grain is a general term fre-quently used for ooids and pisoids, and includesoncoids, grains with a microbial coating (see Section4.3.3).
The majority of modern ooids range from 0.2 to 0.5mm in diameter. They typically form in agitated
112 Chapter 4
waters where they are frequently moved as sandwaves,dunes and ripples by tidal and storm currents, andwave action. On the Bahama platform, the ooids formshoals close to the edge of the platform. In the ArabianGulf the ooids form in tidal deltas at the mouth of tidalinlets between barrier islands along the Trucial Coast.Along the Yucatan shoreline (northeast Mexico),ooids are being precipitated in the shoreface and fore-shore zones. Depths of water where ooids precipitateusually are less than 5m, but they may reach 10–15m.
Practically all marine ooids forming today, such asin the Bahamas and Arabian Gulf, are composed ofaragonite and they have a high surface polish (Fig.
4.2). Bimineralic high-Mg calcite–aragonite ooidshave been recorded from Baffin Bay, Texas. Relictearly Holocene ooids of high-Mg calcite occur off theGreat Barrier Reef of Queensland and on the AmazonShelf.
The characteristic microstructure of modern arago-nitic marine ooids is a tangential orientation of acicu-lar crystals or needles, 2mm in length. Lamellae ofmicrocrystalline aragonite and of randomly orientedaragonite needles also occur. The sub-Recent high-Mgcalcite ooids have a radial fabric. Ooids contain organic matter, located chiefly between lamellae and inthe microcrystalline layers.
Table 4.1 Scheme for petrographic description for carbonate rocks
Hand specimenNote the colour; type(s) of grain(s); grain size; grain shape if significant; evidence of fossil breakage, disarticulation, abrasion, packing,
etc.; presence of micrite and/or sparry calcite if observable; physical and biogenic sedimentary structures; stylolites, etc.Any evidence of diagenetic alteration, e.g. dolomitization, silicification, dissolution, compaction
Thin-sectionCheck macroscopic features of thin-section by holding up to light and noting any lamination or large fossils or grainsGrains: type(s) —bioclasts, ooids, peloids, intraclasts/aggregates, etc.; size, sorting, shape, packing of grains; mineralogy/composition of
grains (in stained sections —ferroan/non-ferroan calcite/dolomite). Identify bioclasts from shape, internal structure and preservation.Ooids —determine original composition (aragonite or calcite). Peloids —micritized grains or faecal pellets?
Micrite (lime mud): grains mostly less than 4 mm and normally almost opaque and brownish in colour, but may be peloidal. Any aggradingneomorphism?
Cements: Usually coarser than 10 mm; identify types of cement: fibrous calcite, bladed calcite, syntaxial overgrowths, drusy calcite spar,poikilotopic calcite spar; note geometry: meniscus, isopachous, pore-filling; spar may be non-ferroan, ferroan, or zoned (use staining).Determine original mineralogy (aragonite, calcite, high-Mg calcite). Timing of cement precipitation, early versus late, pre- versus post-compaction. Where cemented —marine, meteoric, burial?
Replacement, recrystallization and neomorphism: minerals such as dolomite, silica (chert) or phosphate may replace calcite grains,micrite or cement. Neomorphism of grains, e.g. aragonitic shells, ooids or cements replaced by calcite (calcitization) and micritereplaced by microspar —aggrading neomorphism. Timing of replacement relative to compaction?
Dolomitization: scattered rhombs or pervasively dolomitized? Any fabric control: grains or micrite preferentially dolomitized? Determinedolomite texture and crystal shape/size/zonation. Is dolomite early or late, pre- or post-compaction; timing relative to calcitediagenesis. Any dedolomitization?
Compaction: look for evidence of mechanical compaction (broken bioclasts, broken micrite envelopes, spalled oolitic coatings or earlycements). Check for chemical compaction: sutured contacts between grains, pre-burial spar cementation, and stylolites: through-going pressure dissolution seams, post-spar cementation
Porosity: identify type: primary (intergranular, intragranular, cavity, growth, etc.) and secondary (fracture, dissolutional, intercrystalline —as through dolomitization)
Classify using the Dunham scheme (or Folk) after assessing the proportion of grains, cement and micrite, and dominant grain types.Where diagenetic alteration is significant, then qualify name, e.g. partly dolomitized bioclastic grainstone, silicified oosparite,recrystallized lime mudstone
InterpretationDepositional environment: use grain types and texture, e.g. most grainstones represent moderate to high energy shallow subtidal, many
lime mudstones–wackestones represent lower-energy, lagoonal or outer-shelf/ramp environments. Check bioclasts; these can indicateopen-marine, restricted, deep-water, shallow, non-marine, etc.
Diagenesis: identify early (near-surface) and late (burial) processes, determine timing of cementation. Assess degree of compaction; anydolomitization? Try to interpret in terms of pore-water chemistry (e.g. marine or freshwater from cement fabric, and Eh/pH fromferroan/non-ferroan nature of calcite/dolomite). Deduce succession of diagenetic events and porosity evolution
Limestones 113
Ooids and pisoids can also form in quieter-watermarine locations, such as in lagoons (e.g. Baffin Bay,Texas) and on tidal flats (e.g. the Trucial Coast).Ooids–pisoids also are formed in association withtepee structures on intertidal–supratidal flats (see Section 4.6.1), being precipitated in local pools and beneath cemented crusts. These low-energy coatedgrains commonly have a strong radial fabric, so thatthey do break relatively easily (e.g. Fig. 4.3). Thoseforming in a vadose situation, such as in associationwith tepees, are commonly asymmetric; they may alsoshow downward thickening laminae and a fitted fabricfrom in situ growth. Such vadose pisoids are a feature
of the back-reef facies in the Capitan Reef Complex ofTexas–New Mexico.
Ooids can form in high-energy locations in lakes, as in the Great Salt Lake, Utah and Pyramid Lake,Nevada. These lacustrine ooids are commonly dull,and they may have a cerebroid (bumpy) surface. TheGreat Salt Lake ooids are composed of aragonite andmany have a strong radial fabric. The Pyramid Lakeooids are bimineralic (low-Mg calcite–aragonite).
Structures resembling ooids–pisoids do form in cal-careous soils. They are usually composed of fine-grained calcite and have a poorly developed concentriclamination, which may be asymmetric. They form
Features Thin-section 1 Thin-section 2
Grains present and percentageBioclastsOoidsPellidsIntraclastsMicrite: uniform, peloidal, microsparitic
TextureRoundness, sorting, fabric, binding, framework
Cavity structuresUmbrella, geopetal, fenestrae —birdseye, laminar,
burrow, rootlet
CementsFibrous, bladed, drusy calcite, sparite, geometry
ReplacementsCalcitization, neomorphism, microspar,
micritization, dissolution
DolomiteCrystal shape, size, distribution, texture, fabric
preservation, timing
Evidence of compactionBroken grains and micrite envelopes, sutured
contacts, stylolites
PorosityIntergranular, intragranular, framework, mouldic,
intercrystalline, fracture, etc.
Name/microfacies
Depositional environment
Diagenetic events 1:2:3:
Table 4.2 Scheme for describingcarbonate rocks in thin-section
114 Chapter 4
Mineralogy
Low-Mg High-MgOrganism Aragonite calcite calcite Aragonite+ calcite
MolluscaBivalves X X XGastropods X XPteropods XCephalopods X (X)
Brachiopods X (X)
CoralsScleractinian XRugose + Tabulate X X
Sponges X X X
Bryozoans X X X
Echinoderms X
Ostracods X X
ForaminiferaBenthic (X) XPelagic X
AlgaeCoccolithophoridae XRhodophyta X XChlorophyta XCharophyta X
Table 4.3 The mineralogy of carbonateskeletons (X, dominant mineralogy; (X),less common). During diagenesis, thesemineralogies may be altered orreplaced; in particular, aragonite ismetastable and is invariably replaced bycalcite, and high-Mg calcite loses its Mg
Ooid
Diametertypically
0.2–0.5 mm
Concentric lamellae
Micritized lamella
Nucleus, skeletal fragmentor quartz grain
Tangential aragoniteneedles in mostmodern ooids
Radial fibrous calcitein most ancient ooids
Aggregate Peloid – composed of micrite
e.g.
A collection of grainscemented together
Superficial ooid—single lamella
Composite ooid
(a) (b)
A pellet, typically0.1–0.5 mmdiameter
Amorphous grain,many are micritizedskeletal grains
Fig. 4.1 The principal non-skeletal grainsin limestones: ooids, peloids andaggregates.
Limestones 115
largely by microbial processes and calcification of fungal and bacterial filaments. These calcrete pisoidsare commonly associated with laminated crusts,formed by the calcification of root mats (see Section4.6.1).
Ancient marine ooids
Ooids in the rock record are composed of calcite (lowMg), unless dolomitized or silicified. However, al-though there has been much discussion over the mat-ter, it is clear that some were originally calcite, whereasothers were originally aragonite. Former bimineralicooids also have been reported. Primary calcite ooids,whether in high-energy or low-energy facies, typicallyhave a radial texture of wedge-shaped, fibrous crystals(see Plate 6c). Under crossed polars, an extinctioncross is seen. The cortex of larger, originally calciticooids may have an inner radial part and an outer radial–concentric part. It is not easy to determinewhether calcitic ooids originally had a low or high Mgcontent. The presence of small dolomite crystals (microdolomites), evidence of minor dissolution of the cortex or a moderate to high iron content may indi-cate an original high-Mg calcite composition (see Section 4.7.1).
Ancient ooids originally of aragonite will have beenaltered during diagenesis to a greater or lesser extent(see Plate 6d). They may be replaced by calcite withsome retention of the original tangential structure (if high energy) or radial structure (if quiet water)through the presence of minute inclusions of organicmatter and/or relics of the aragonite. This replacementprocess is termed calcitization (also see Section 4.7.4).Alternatively, the aragonite of the ooids may be dis-solved out completely, to leave oomoulds. These holesmay be left empty to give the limestone an oomouldicporosity, or filled with calcite cement (see Plate 6d).Some ancient ooids have a fine-grained micritic tex-ture. As with modern ooids, this may be the result ofmicritization by endolithic microbial organisms (seeSection 4.3.3) or it may result from diagenetic alter-ation (neomorphism, see Section 4.7.3).
Origin of ooids
There has been much discussion on the origin of ooids;current ideas invoke biochemical or inorganic pro-
Fig. 4.2 Modern aragonitic ooids from the Bahamas withpolished surface.
Fig. 4.3 Pisoids, probably of vadose origin (vadoids) showingstrong radial structure and broken and recoated grains. Smallergrains are peloids. Rock is a grainstone, but it is 100% dolomite.The sediment was dolomitized with perfect preservation oforiginal texture. The clear spar is a dolomitic cement. DürrensteinFormation, Triassic. The Dolomites, Italy.
116 Chapter 4
cesses, a direct microbial origin being largely discard-ed now. Although a precise mechanism of inorganicprecipitation has not been demonstrated, seawater inshallow tropical areas is supersaturated with respectto CaCO3, so that this, together with water agitation,CO2 degassing and elevated temperature, might besufficient to bring about carbonate precipitation onnuclei. A biochemical origin hinges on the organic mu-cilage that coats and permeates the ooids. One view isthat bacterial activity within the organic matter cre-ates a microenvironment conducive to carbonate pre-cipitation. Some ooids have a proteinaceous matrix,also suggesting a biochemical process, because in or-ganisms it is amino acids that induce calcification. Abiological origin is supported by SEM examination ofmodern marine ooids, which shows that the aragoniterods and nanograins that form the cortex are identicalto those associated with endolithic and epilithic bacte-ria and mucilaginous films occurring on and within theooids. Laboratory synthesis of ooids has suggestedthat organic compounds in the water are instrumentalin the formation of quiet-water ooids with their radialfabric, but that ooids formed in agitated conditions areprecipitated inorganically.
The factors determining the primary mineralogy ofooids are water chemistry, especially Pco2, Mg/Caratio and carbonate saturation, and possibly the de-gree of water agitation. It is believed that aragoniteand high-Mg calcite ooids are precipitated when Pco2is low and Mg/Ca ratio high, and that low-Mg calciteooids form when Pco2 is high and Mg/Ca ratio low. Itis unclear what controls the precipitation of aragoniteooids as opposed to high-Mg calcite ooids but carbon-ate supply rate has been implicated. High carbonatesupply, as would occur in high-energy locations, isthought to favour aragonite precipitation.
Surveys of the original mineralogy of ooids throughthe Phanerozoic have shown that there is a secularvariation, with aragonitic ooids, which may be associ-ated with calcitic ooids (presumably high-Mg calciteoriginally) in the late Precambrian/early Cambrian,mid-Carboniferous through Triassic and Tertiary toRecent, and calcitic ooids (presumed to have had alow-to-moderate Mg content) dominant in the mid-Palaeozoic and Jurassic–Cretaceous (Fig. 4.4). Thispattern suggests that there have been subtle fluctua-tions in seawater chemistry through time, in Pco2and/or the Mg/Ca ratio. See Sandberg (1983) and thereview in Stanley & Hardie (1998). Some anomalies
do occur in the broad trend, notably in Upper Jurassicstrata where aragonitic ooids are recorded, as in theSmackover Formation of the US Gulf Coast subsur-face (see Plate 6d; Heydari & Moore, 1994).
The general trend shown in Fig. 4.4 appears to tie inwith the first-order, global sea-level curve, suggestingthat a geotectonic mechanism(s) is causing the subtlevariations in seawater chemistry, giving rise to the secular variation in ooid mineralogy. High sea-levelstands, correlating with calcite seas, are times of high rates of sea-floor spreading, when Pco2 can be expected to be relatively high from increased meta-morphism at subduction zones and the Mg/Ca ratiorelatively low from increased extraction of Mg2+ atmid-ocean ridges as sea water is pumped through (seeStanley & Hardie, 1998). Temporal exceptions to thebroad trend, as in the Late Jurassic, could be the resultof local conditions, the most likely being a raisedMg/Ca ratio from evaporite deposition, leading toaragonite precipitation.
Peloids
Peloids are spherical, ellipsoidal or angular grains,composed of microcrystalline carbonate, but with nointernal structure (Fig. 4.1 and Plates 6a, 7a,c & 8c).The size of peloids may reach several millimetres butthe majority are in the range of 0.1–0.5mm in diam-eter. Most peloids are of faecal origin and so can be re-ferred to as pellets. Organisms such as gastropods,
?
Aragonite threshold
?
PC C O S D C P T J k T
Rising
Falling
Sea-level fluctuations
Vail et alHallam
High-Mg calciteand aragonite
Calcite: generally lowMg content
Fig. 4.4 Mineralogy of marine, abiogenic, carbonate precipitatesthrough the Phanerozoic, compared with the first-order globalsea-level curve.
Limestones 117
crustaceans and polychaetes produce pellets in vastquantities. Faecal pellets have a regular shape and theyare rich in organic matter. They are most common in the sediments of protected environments such as lagoons and tidal flats. Pellets are very common inlimestones and many micritic limestones, seeminglywithout sand-sized grains, may actually be pelleted.The definition of pellets is commonly lost as a result ofdiagenetic processes, and the limestones may show aflocculent or clotted texture.
The term peloid includes micritized bioclastic grainsformed by alteration of skeletal fragments by micro-boring microbes and recrystallization (Section 4.3.3).They are more irregular in shape and are an impor-tant component of modern Bahamian carbonate sediments.
Aggregates and intraclasts
Aggregates consist of several carbonate particles cemented together by a microcrystalline cement orbound by organic matter. Such grains in the Bahamasare known as grapestones and form in relatively pro-tected shallow subtidal areas, usually beneath a thin,surficial microbial mat.
Intraclasts are fragments of lithified or partly lithi-fied sediment. A common type of intraclast in carbon-ate sediments is a micritic flake or chip, derived fromdesiccation of tidal-flat muds or disruption by stormsof partially lithified or cemented subtidal lime muds(Fig. 4.5). The latter are particularly common in thePrecambrian and Cambrian. An abundance of these
flakes produces flat-pebble or edgewise conglomer-ates, also called flakestones. They may show an imbri-cation of clasts (see Section 2.2.4).
One distinctive intraclast is a black pebble. Theseare carbon-impregnated pebbles, commonly with soil fabrics, which are associated with palaeosoils,laminated crusts and palaeokarsts, and may be re-worked into intraformational conglomerates. Theyform in soil horizons and may be the result of forestfires, or more likely organic-matter impregnation inwater-logged, reducing conditions (Shinn & Lidz,1988).
4.3.2 Skeletal components (excluding algae)
The skeletal components of a limestone are a reflectionof the distribution of carbonate-secreting inverte-brates through time and space (Fig. 4.6). Environ-mental factors, such as depth, temperature, salinity,substrate and turbulence, control the distribution anddevelopment of the organisms in the various carbon-ate environments. Throughout the Phanerozoic, vari-ous groups expanded and evolved to occupy the nichesleft by others that were declining or becoming extinct.The mineralogy of carbonate skeletons also variesthrough the Phanerozoic, like the inorganic pre-cipitates (Section 4.3.1), and this probably is also a reflection of tectonically forced shifts in seawaterchemistry (see Stanley & Hardie, 1998).
The main skeletal contributors to limestones are dis-cussed in the following sections, with comments ontheir recognition. Detailed accounts of skeletal struc-
Fig. 4.5 Intraclasts. Storm bed composedof rip-up clasts of micritic limestone.Cambrian. Qinhuandao, northeast China.
118 Chapter 4
ture and thin-section appearance are given in Bathurst(1975), Scholle (1978), Flügel (1982), Adams et al.(1984) and Adams & Mackenzie (1998). For the iden-tification of skeletal particles in thin-section the pointsto note are:1 shape (and size), bearing in mind that under the microscope a two-dimensional view only is given; look for other sections of the same fossil to determinethe three-dimensional shape;2 internal microstructure, which may be modified orobliterated by diagenesis;3 mineralogy —although in a limestone everythingwill be calcite, unless dolomitized or silicified, fabricevidence can be used to decide if a skeletal particle was originally aragonitic (staining a thin-section forferroan calcite and dolomite using Alizarin red S andpotassium ferricyanide (see Section 4.1) may provideadditional information —for example, skeletal com-ponents originally of high-Mg calcite may be replacedpreferentially by ferroan calcite);4 other features likely to be diagnostic, such as pres-ence of spines or pores.
Mollusca
Bivalves, gastropods and cephalopods occur in lime-stones from the Early Palaeozoic onwards. The bi-valves are a very large group with species occupyingmost marine, brackish and freshwater environments.Bivalves have been important contributors to marinecarbonate sediments, particularly since the Tertiaryfollowing the decline of the brachiopods. The modes
of life are very varied, too, including infaunal (living within the sediment), epifaunal (attached to ahard substrate), vagile (crawlers), nektonic (free-swimming) and planktonic (free-floating). Certain bi-valves, such as oysters, may form reef-like structures.During the Cretaceous, masses of aberrant, coral-likebivalves called rudists formed reefs in Mexico, south-ern USA, the Mediterranean region and the MiddleEast, for example. Fresh- and brackish-water lime-stones may be composed largely of bivalves; examplesoccur in the Upper Carboniferous, Upper Triassic(Rhaetic) and Upper Jurassic (Purbeck) of western Europe.
The majority of bivalve shells are composed of aragonite; some are of mixed mineralogy (the rudists,for example, Plate 7e); others, such as the oysters and scallops, are calcitic. Bivalve shells consist of severallayers of specific internal microstructure, composed of micron-sized crystallites (see Plate 7b,c). One com-mon shell structure is of an inner nacreous layer con-sisting of sheets of aragonite tablets, and an outerprismatic layer of aragonite (or calcite). If composedoriginally of aragonite, the internal structure of a fossilbivalve shell is likely to be poorly preserved or not preserved at all (see Fig. 4.7). The aragonite may bedissolved out completely to leave a mould, which subsequently may be filled by calcite (a cement). This isthe most common mode of preservation so that mostbivalve fragments in limestones are composed of clear,coarse drusy sparite (see Plate 7d; also Plates 6c & 9e).Alternatively, the aragonite of the shell may be re-placed by calcite (calcitized) so that faint relics of the
Cenozoic
Fora
min
ifer
a
Arc
hae
ocy
ath
ids
Spo
ng
es
Co
rals
Stro
mat
op
oro
ids
Bry
ozo
ans
Bra
chio
po
ds
Biv
alve
s
Gas
tro
po
ds
Cep
hal
op
od
s
Trilb
ites
Ost
raco
ds
Ech
ino
der
ms
Cal
care
ou
s al
gae
Stro
mat
olit
es
CretaceousJurassicTriassicPermianCarboniferousDevonianSilurianOrdovicianCambrian
Precambrian? ?
? ?
?
Fig. 4.6 Age range and generalizedtaxonomic diversity of principal carbonate-secreting organisms.
Plat
e 1
(a)L
ithic
gra
ins o
f sed
imen
tary
orig
in—
lam
inat
ed a
nd si
lty sh
ale,
als
o an
gula
rqu
artz
gra
ins.
Pla
ne-p
olar
ized
ligh
t.C
arbo
nife
rous
fluv
ial s
ands
tone
.C
anta
bria
ns, S
pain
. Fie
ld o
f vie
w
3¥
2m
m.
(b)L
ithic
gra
ins o
f vol
cani
c or
igin
. The
two
dark
gra
ins c
onsi
st o
f mafi
c ph
enoc
ryst
s in
a ve
ry a
ltere
d, o
rigin
ally
gla
ssy,
dar
kgr
ound
mas
s. T
he o
ther
gra
ins c
onsi
st
of m
inut
e fe
ldsp
ar la
ths i
n a
glas
sygr
ound
mas
s, a
nd h
ave
num
erou
s sm
all
dark
iron
-ric
h sp
ots.
The
cem
ent i
s cal
cite
.Pl
ane-
pola
rized
ligh
t. T
riass
ic sh
allo
w-
mar
ine
sand
ston
e. T
he D
olom
ites,
Ital
y.Fi
eld
of v
iew
3¥
2m
m.
(c)Q
uart
z gr
ains
: sev
eral
mon
ocry
stal
line
quar
tz g
rain
s sho
win
g bo
th u
nifo
rm a
ndun
it ex
tinct
ion;
pol
ycry
stal
line
quar
tzgr
ains
with
seve
ral a
nd m
any
subc
ryst
als,
som
e of
the
latt
er w
ith su
ture
d co
ntac
ts.
The
cem
ent i
s poi
kilo
topi
c ca
lcite
. Cro
ssed
pola
rs. P
erm
ian
aeol
ian
sand
ston
e.D
urha
m, E
ngla
nd. F
ield
of v
iew
3¥
2m
m.
(d)F
elds
par g
rain
s: m
icro
clin
e on
left
with
grid
-iron
twin
ning
and
ort
hocl
ase
on ri
ght
with
brig
ht sp
ecks
of a
ltera
tion
mat
eria
l(s
eric
ite).
Qua
rtz
grai
ns a
lso
pres
ent
(mos
tly m
onoc
ryst
allin
e w
ith u
nit
extin
ctio
n), t
he o
ne in
low
er c
entr
e w
ithov
ergr
owth
. Mus
covi
te fl
ake
extr
eme
right
show
ing
blue
col
our.
Gra
ins a
reco
ated
with
hem
atite
, giv
ing
red/
brow
nrim
. Cro
ssed
pol
ars.
Pre
cam
bria
n flu
vial
sand
ston
e. To
rrid
on, S
cotla
nd. F
ield
of
view
1.2
¥0.
8m
m.
(a)
(b)
(c)
(d)
facing p. 119
Plat
e 2
(a)M
usco
vite
and
kao
linite
in q
uart
zar
enite
. Mus
covi
te sh
ows o
rang
e co
lour
.K
aolin
ite fo
rms t
he sm
all d
ark
crys
tals
betw
een
the
quar
tz g
rain
s, w
hich
are
mos
tly m
onoc
ryst
allin
e w
ith u
nit
extin
ctio
n; th
e on
e on
the
left
has
min
ute
fluid
and
min
eral
incl
usio
ns. C
ross
edpo
lars
. Car
boni
fero
us fl
uvia
l san
dsto
ne.
Nor
thum
berla
nd, E
ngla
nd. F
ield
of v
iew
1.2
¥0.
8m
m.
(b)B
iotit
e m
ica
(bro
wn)
show
ing
effe
cts
of c
ompa
ctio
n, a
nd li
ght b
row
n m
atrix
cons
istin
g of
min
ute,
unr
esol
vabl
e cl
aym
iner
als,
iron
min
eral
s (da
rkbr
own/
blac
k), a
nd si
lt-gr
ade
quar
tz. A
lso
pres
ent a
re m
any
angu
lar q
uart
z gr
ains
(whi
te),
and
feld
spar
and
lith
ic g
rain
s.Pl
ane-
pola
rized
ligh
t. S
iluria
n tu
rbid
itegr
eyw
acke
. Sco
tland
. Fie
ld o
f vie
w 1
.2¥
0.8
mm
. (c
, d)Q
uart
z ar
enite
with
wel
l-dev
elop
edov
ergr
owth
s on
quar
tz g
rain
s, w
hich
are
mos
tly m
onoc
ryst
allin
e w
ith u
nit
extin
ctio
n. D
usty
look
ing
quar
tz g
rain
on
left
is o
f hyd
roth
erm
al o
rigin
and
is fu
ll of
fluid
incl
usio
ns. N
ote
clea
r ove
rgro
wth
.Re
d he
mat
ite c
oatin
g ar
ound
gra
ins.
Tw
ofe
ldsp
ar g
rain
s sho
w e
ffec
ts o
fdi
ssol
utio
n. R
ock
impr
egna
ted
with
blu
ere
sin
to sh
ow p
oros
ity (r
educ
edin
terg
ranu
lar a
nd d
isso
lutio
nal
intr
agra
nula
r). (
c)Pl
ane-
pola
rized
ligh
t;(d
)cro
ssed
pol
ars.
Per
mia
n ae
olia
nsa
ndst
one.
Cum
bria
, Eng
land
. Fie
ld o
fvi
ew 3
¥2
mm
.
(a)
(b)
(c)
(d)
Plat
e 3
(a)Q
uart
z ar
enite
with
ang
ular
look
ing
quar
tz g
rain
s fro
m o
verg
row
th c
emen
t.D
ust-
line
arou
nd g
rain
s vis
ible
in so
me
case
s. S
uper
mat
ure
sand
ston
e co
nsis
ting
of m
onoc
ryst
allin
e, u
nit-
extin
guis
hing
quar
tz g
rain
s. Z
ircon
, a h
eavy
min
eral
,up
per l
eft (
red)
. Cro
ssed
pol
ars.
Car
boni
fero
us m
arin
e sa
ndst
one.
Dur
ham
, Eng
land
. Fie
ld o
f vie
w
1.2
¥0.
8m
m.
(b)L
ithar
enite
with
sedi
men
tary
rock
frag
men
ts o
f fine
sand
ston
e an
d m
udro
ck(s
ome
show
ing
lam
inat
ion)
. Qua
rtz
grai
nsal
so p
rese
nt (c
lear
). N
ote
the
effe
cts o
fco
mpa
ctio
n: a
tigh
t-fit
ting
arra
ngem
ent
of g
rain
s, a
nd so
me
inte
rpen
etra
tion
and
squa
shin
g of
gra
ins t
oo. P
lane
-pol
ariz
edlig
ht. C
arbo
nife
rous
fluv
ial s
ands
tone
.C
anta
bria
ns, S
pain
. Fie
ld o
f vie
w
6¥
4m
m.
(c)A
rkos
e w
ith m
any
feld
spar
gra
ins
(dus
ty/d
irty
look
ing
com
pare
d w
ith c
lear
erqu
artz
gra
ins)
, and
hem
atite
coa
tings
arou
nd sa
nd g
rain
s; n
ote
that
hem
atite
isab
sent
whe
re g
rain
s are
in c
onta
ct. P
lane
-po
lariz
ed li
ght.
Pre
cam
bria
n flu
vial
sand
ston
e. To
rrid
on, S
cotla
nd. F
ield
of
view
1.2
¥0.
8m
m.
(d)A
rkos
e w
ith fe
ldsp
ar g
rain
s(o
rtho
clas
e, m
icro
clin
e) sh
owin
g in
cipi
ent
repl
acem
ent b
y se
ricite
(min
ute
brig
htcr
ysta
ls).
Cro
ssed
pol
ars.
Pre
cam
bria
nflu
vial
sand
ston
e. To
rrid
on, S
cotla
nd. F
ield
of v
iew
1.2
¥0.
8m
m.
(a)
(b)
(c)
(d)
Plat
e 4
(a, b
)Gre
ywac
ke. Q
uart
z, fe
ldsp
ar a
ndlit
hic
grai
ns a
re c
onta
ined
in a
fine
-gr
aine
d m
atrix
of c
hlor
ite a
nd si
lt-gr
ade
quar
tz: (
a)pl
ane-
pola
rized
ligh
t; (b
)cr
osse
d po
lars
. Silu
rian
turb
idite
grey
wac
ke. S
outh
ern
Upl
ands
, Sco
tland
.Fi
eld
of v
iew
3¥
2m
m.
(c, d
)Qua
rtz
over
grow
ths o
n qu
artz
grai
ns. T
he g
rain
surf
ace
is sh
own
by th
ere
d he
mat
ite c
oatin
g. B
ette
r ove
rgro
wth
son
mon
ocry
stal
line
quar
tz th
anpo
lycr
ysta
lline
qua
rtz
grai
ns. F
elds
par
grai
n up
per r
ight
has
no
over
grow
th a
ndsh
ows s
ome
effe
cts o
f dis
solu
tion.
Roc
kim
preg
nate
d w
ith b
lue
resi
n to
show
poro
sity
(red
uced
inte
rgra
nula
r): (
c)pl
ane-
pola
rized
ligh
t; (d
)cro
ssed
pol
ars.
Perm
ian
aeol
ian
sand
ston
e. C
umbr
ia,
Engl
and.
Fie
ld o
f vie
w 1
.2¥
0.8
mm
.
(a)
(b)
(c)
(d)
Plat
e 5
(a, b
)Cal
cite
cem
ent i
n qu
artz
are
nite
.La
rge
poik
iloto
pic
calc
ite c
ryst
als
encl
osin
g se
vera
l gra
ins.
Qua
rtz
grai
ns a
rew
ell-r
ound
ed m
onoc
ryst
allin
e, w
ith u
nit
and
undu
lose
ext
inct
ion,
and
poly
crys
talli
ne. F
elds
par g
rain
show
ing
split
ting
by c
alci
te c
ryst
al; d
etai
l in
(b).
Cro
ssed
pol
ars.
Per
mia
n ae
olia
nsa
ndst
one.
Dur
ham
, Eng
land
: (a)
field
of
view
3¥
2m
m; (
b)fi
eld
of v
iew
1.2
¥0.
8m
m.
(c)K
aolin
ite (s
mal
l bla
ck a
nd w
hite
crys
talli
tes)
bet
wee
n qu
artz
gra
ins (
mon
o-an
d po
lycr
ysta
lline
), pr
obab
ly re
plac
ing
afe
ldsp
ar g
rain
. Dis
tort
ed m
usco
vite
flak
e(s
how
ing
blue
col
our)
bet
wee
n qu
artz
grai
ns. C
ross
ed p
olar
s. C
arbo
nife
rous
fluvi
al sa
ndst
one.
Nor
thum
berla
nd,
Engl
and.
Fie
ld o
f vie
w 1
.2¥
0.8
mm
. (d
)Mud
rock
show
ing
effe
cts o
fco
mpa
ctio
n w
ith fr
actu
re o
f she
lls (t
hin
brac
hiop
ods)
, fol
ding
of l
amin
ae a
roun
dsh
ells
and
flat
teni
ng o
f bur
row
s (up
per
left
). M
inut
e si
lt-gr
ade
quar
tz a
nd sh
ell
debr
is d
isse
min
ated
thro
ugho
ut m
ud;
roun
d w
hite
gra
in (c
entr
e le
ft) i
s crin
oid
ossi
cle.
Car
boni
fero
us m
arin
e m
udro
ck.
Nor
thum
berla
nd, E
ngla
nd. F
ield
of v
iew
6
¥4
mm
.
(a)
(b)
(c)
(d)
Plat
e 6
(a, b
)Hol
ocen
e oo
ids c
ompo
sed
ofar
agon
ite sh
owin
g co
ncen
tric
stru
ctur
ean
d a
nucl
eus,
seve
ral o
f whi
ch a
repe
loid
s, a
nd o
ne in
the
cent
re o
f the
bigg
est o
oid
is a
bio
clas
t. S
truc
ture
less
oval
gra
in a
t low
er le
ft is
a p
eloi
d. A
lso
pres
ent i
s a m
enis
cus c
emen
t of c
alci
tepr
ecip
itate
d in
the
met
eoric
vad
ose
zone
.Th
e w
hite
are
as b
etw
een
grai
ns in
(a) a
ndth
e bl
ack
area
s in
(b) a
re p
ore
spac
e:
(a)p
lane
-pol
ariz
ed li
ght;
(b)c
ross
edpo
lars
. Jou
lters
Cay
, Bah
amas
. Fie
ld o
fvi
ew 1
.2¥
0.8
mm
. (c
)Prim
ary,
cal
citic
ooi
ds w
ith st
rong
radi
al-c
once
ntric
stru
ctur
e an
d nu
clei
mos
tly o
f pel
oids
. Biv
alve
frag
men
t, n
owco
mpo
sed
of c
lear
cal
cite
spar
cry
stal
s, is
also
coa
ted.
Not
ice
cont
act b
etw
een
grai
ns; a
litt
le in
terp
enet
ratio
n in
dica
ting
som
e bu
rial c
ompa
ctio
n be
fore
cem
enta
tion.
The
ooi
ds a
re c
onta
ined
in a
very
larg
e po
ikilo
topi
c ca
lcite
cem
ent
(her
e ap
pear
ing
whi
te).
Ool
itic
grai
nsto
ne, J
uras
sic.
Lin
coln
shire
,En
glan
d. F
ield
of v
iew
3¥
2m
m.
(d)F
orm
erly
ara
goni
tic o
oids
, now
com
pose
d of
cal
cite
with
poo
rpr
eser
vatio
n of
orig
inal
con
cent
ricst
ruct
ure
and
oom
ould
s (fil
led
with
blu
ere
sin)
. Som
e co
mpa
ctio
n of
oom
ould
sin
dica
ting
that
the
drus
y ca
lcite
spar
cem
ent i
s a b
uria
l pre
cipi
tate
. Ool
itic
grai
nsto
ne, S
mac
kove
r For
mat
ion,
Jura
ssic
. Sub
surf
ace
Ark
ansa
s, U
SA. F
ield
of v
iew
3¥
2m
m.
(a)
(b)
(c)
(d)
Plat
e 7
(a)P
eloi
ds; m
any
are
mic
ritiz
ed b
iocl
asts
and
ooid
s, so
me
are
faec
al p
elle
ts. M
icrit
een
velo
pe d
efine
s a b
ival
ve sh
ell t
hat h
asdi
ssol
ved
away
; biv
alve
shel
l with
in c
oate
dgr
ain
is re
plac
ed b
y co
arse
cal
cite
cry
stal
s.Th
e sp
arse
cem
ent c
onsi
sts o
f sm
all,
stub
by c
alci
te c
ryst
als o
f pro
babl
em
eteo
ric p
hrea
tic o
rigin
. Sm
all d
olom
iterh
ombs
als
o pr
esen
t. B
lue
resi
n sh
ows
poro
sity
. Jur
assi
c. D
orse
t, E
ngla
nd. F
ield
of
view
0.8
¥0.
8m
m.
(b,c
)Mod
ern
biva
lve
frag
men
t with
mic
rite
enve
lope
. She
ll, c
ompo
sed
ofar
agon
ite, c
onsi
sts o
f min
ute
crys
talli
tes
givi
ng a
swee
ping
ext
inct
ion
unde
rcr
osse
d po
lars
(c).
Abu
Dha
bi, U
AE.
Fie
ldof
vie
w 0
.8¥
1.0
mm
. (d
)Biv
alve
frag
men
t (th
e el
onga
te g
rain
)in
cen
tre
with
a p
rom
inen
t mic
rite
enve
lope
and
shel
l now
com
pose
d of
drus
y ca
lcite
spar
, a c
emen
t. T
he m
icrit
een
velo
pe h
as fr
actu
red
(in th
e ce
ntre
) as a
resu
lt of
com
pact
ion.
Als
o pr
esen
t are
num
erou
s pel
oids
, mos
t of
whi
ch a
re m
icrit
ized
bio
clas
ts, a
ndso
me
crin
oid
frag
men
ts, i
n a
spar
itece
men
t.U
rgon
ian,
Cre
tace
ous.
Ver
cors
,Fr
ance
. Fie
ld o
f vie
w 3
¥2
mm
. (e
)Hip
purit
id ru
dist
biv
alve
(the
con
ical
,at
tach
ed v
alve
) sho
win
g br
own,
wel
l-pr
eser
ved
fibro
us c
alci
te o
uter
wal
l, an
dth
inne
r inn
er w
all w
ith ta
bula
e, w
hich
orig
inal
ly w
ere
arag
onite
, but
are
now
com
pose
d of
cal
cite
spar
. Rou
nd a
reas
of
sedi
men
t (m
icrit
e) a
re th
e fil
ls o
f spo
nge
borin
gs. C
reta
ceou
s. P
rove
nce,
Fra
nce.
Fiel
d of
vie
w 6
¥4
mm
.(d
)(e
)
(a)
(b)
(c)
Plat
e 8
(a, b
)Cal
citiz
ed b
ival
ve sh
ells
. She
llsor
igin
ally
com
pose
d of
ara
goni
te b
utre
plac
ed b
y ca
lcite
with
som
e re
tent
ion
ofor
igin
al sh
ell s
truc
ture
. Cal
cite
cry
stal
scr
oss-
cut t
he sh
ell s
truc
ture
and
are
pseu
dopl
eoch
roic
(diff
eren
t sha
des o
fbr
own
on sl
ide
rota
tion)
: (a)
plan
e-po
lariz
ed li
ght a
nd (b
)cro
ssed
pol
ars.
Jura
ssic
. Dor
set,
Eng
land
. Fie
ld o
f vie
w 0
.8¥
0.8
mm
. (c
)Gas
trop
ods,
in lo
ng- a
nd c
ross
-sec
tion,
defin
ed b
y th
in m
icrit
e en
velo
pes (
blac
k).
The
shel
ls, o
rigin
ally
ara
goni
te, d
isso
lved
out a
nd v
oids
wer
e fil
led
by m
arin
e fib
rous
calc
ite c
emen
t (pa
le b
row
n). A
late
r cle
arca
lcite
cem
ent fi
lled
rem
aini
ng p
ores
(whi
te).
Jura
ssic
. Sic
ily, I
taly.
Fie
ld o
f vie
w 4
¥4
mm
. (d
)Bra
chio
pod
shel
l with
pun
cti,
som
efil
led
with
lim
e m
ud se
dim
ent,
show
ing
pres
erva
tion
of in
tern
al st
ruct
ure
cons
istin
g of
obl
ique
ly a
rran
ged
fibre
s.O
ther
gra
ins a
re m
icrit
ized
bio
clas
ts(p
eloi
ds).
Bioc
last
ic g
rain
ston
e.C
reta
ceou
s. V
erco
rs, F
ranc
e. F
ield
of v
iew
3¥
2m
m.
(e)B
rach
iopo
d sh
ells
and
spin
es w
ith w
ell-
pres
erve
d sh
ell s
truc
ture
, but
lim
esto
neha
s suf
fere
d co
mpa
ctio
n an
d m
any
shel
lsar
e br
oken
. Thi
n-se
ctio
n is
stai
ned
with
Aliz
arin
Red
S a
nd p
otas
sium
ferr
icya
nide
;bi
ocla
sts a
re p
ink
(cal
cite
) and
cem
ent i
sbl
ue (f
erro
an c
alci
te).
Cem
ent o
ccur
sw
ithin
cra
cks s
how
ing
that
it is
a b
uria
lpr
ecip
itate
. Crin
oids
and
cla
y (b
row
n) a
lso
pres
ent.
Bio
clas
tic p
acks
tone
.C
arbo
nife
rous
. Nor
thum
berla
nd,
Engl
and.
Fie
ld o
f vie
w 6
¥4
mm
.
(a)
(b)
(c)
(d)
(e)
Plat
e 9
(a)R
ugos
e co
ral (
Lith
ostr
otio
nsp
.)sh
owin
g in
tern
al p
late
s (se
pta,
tabu
lae
and
diss
epim
ents
). Th
e po
res w
ithin
the
cora
l are
par
tly fi
lled
with
an
initi
alis
opac
hous
fibr
ous m
arin
e ce
men
t and
then
by
inte
rnal
sedi
men
t. C
arbo
nife
rous
.D
urha
m, E
ngla
nd. F
ield
of v
iew
6¥
4m
m.
(b)S
cler
actin
ian
cora
l fro
m a
Jura
ssic
patc
h re
ef sh
owin
g va
riabl
e pr
eser
vatio
nof
stru
ctur
e as
a re
sult
of re
plac
emen
t of
the
orig
inal
ara
goni
te b
y ca
lcite
. Als
opr
esen
t, o
n th
e rig
ht, i
s a b
orin
g m
ade
bya
litho
phag
id b
ival
ve (s
hells
pre
sent
); th
ebo
ring
show
s a g
eope
tal s
truc
ture
, with
sedi
men
t bel
ow (d
ark)
and
dru
sy c
alci
tesp
ar a
bove
(whi
te).
York
shire
, Eng
land
.Fi
eld
of v
iew
6¥
4m
m.
(c, d
, e)F
oram
inife
ra. (
c)En
doth
yrac
idfo
ram
, crin
oid
to le
ft, c
ut b
y a
calc
ite v
ein.
Car
boni
fero
us. C
lwyd
, Wal
es. F
ield
of
view
4¥
2m
m. (
d)N
umm
ulite
s. B
iocl
astic
grai
nsto
ne, E
ocen
e. T
unis
ia. F
ield
of v
iew
6¥
4m
m. (
e)M
iliol
ids,
als
o bi
valv
efr
agm
ents
. Cre
tace
ous.
Ver
cors
, Fra
nce.
Fiel
d of
vie
w 4
¥2
mm
.
(a)
(b)
(d)
(c)
(e)
Plat
e 10
(a)D
asyc
lad
alga
e. C
apita
n, P
erm
ian.
Texa
s, U
SA. F
ield
of v
iew
4¥
2m
m.
(b)C
alca
reou
s red
alg
ae, L
ithot
ham
nion
in lo
ngitu
dina
l sec
tion
(sho
win
g se
ason
algr
owth
zon
es) a
nd c
ross
-sec
tion.
The
grai
ns h
ere
are
cem
ente
d by
isop
acho
ushi
gh-M
g ca
lcite
mar
ine
cem
ent.
Rec
ent.
Beliz
e. F
ield
of v
iew
1.0
¥0.
8m
m.
(c)C
alci
fied
mic
robe
s, R
enal
cis.
Dev
onia
n.G
uilin
, Chi
na. F
ield
of v
iew
4¥
2m
m.
(d)M
oder
n m
icro
bial
mat
com
pose
d of
dolo
mite
. The
fila
men
ts o
f the
cyan
obac
teria
are
cle
arly
vis
ible
, but
the
dolo
mite
cry
stal
s are
subm
icro
scop
ic.
Rece
nt. B
aham
as. F
ield
of v
iew
4¥
2m
m.
(e)S
trom
atol
ite (m
icro
bial
mat
)co
mpo
sed
of m
icrit
e la
min
ae a
ndla
min
oid
fene
stra
e, w
ith so
me
intr
acla
sts
from
des
icca
tion.
Car
boni
fero
us.
Gla
mor
gan,
Wal
es. F
ield
of v
iew
5¥
4m
m.
(f)S
trom
atol
ite c
ompo
sed
of m
icrit
ic a
ndgr
ainy
lam
inae
, and
smal
l spa
r-fil
led
fene
stra
e. P
reca
mbr
ian.
Flin
ders
,A
ustr
alia
. Fie
ld o
f vie
w 4
¥2
mm
.
(a)
(b)
(e)
(d)
(c)
(f)
Plat
e 11
(a)C
oral
with
ara
goni
te c
emen
t —ne
edle
san
d a
botr
yoid
with
in th
e co
ralli
tes,
and
lime
mud
inte
rnal
sedi
men
t (bl
ack)
and
pelo
ids i
n ot
her p
ores
. Cro
ssed
pol
ars.
Rece
nt. B
eliz
e. F
ield
of v
iew
6¥
4m
m.
(b)I
sopa
chou
s hig
h-M
g ca
lcite
cem
ent
arou
nd sk
elet
al g
rain
s (in
clud
ing
calc
areo
us re
d al
gae)
. Rec
ent f
ore-
reef
debr
is. B
eliz
e. F
ield
of v
iew
1.2
¥0.
8m
m.
(c)O
oliti
c gr
ains
tone
with
bra
chio
pod
and
crin
oid
frag
men
ts (w
ith th
in m
icrit
een
velo
pe—
blac
k) c
emen
ted
by e
arly
isop
acho
us fi
brou
s mar
ine
cem
ent
(cal
cite
), th
en so
me
inte
rnal
sedi
men
t of
pelo
ids,
follo
wed
by
drus
y ca
lcite
spar
cem
ent.
Car
boni
fero
us. G
lam
orga
n,W
ales
. Fie
ld o
f vie
w 6
¥4
mm
. (d
)Har
dgro
und
with
ooi
ds su
rrou
nded
by
thin
isop
acho
us m
arin
e ce
men
t frin
ge a
ndth
en p
ore
spac
e fil
led
by li
me
mud
, now
mic
rite.
Cem
ente
d ro
ck th
en c
ut b
yan
nelid
bor
ings
, whi
ch la
ter fi
lled
with
quar
tz g
rain
s. Ju
rass
ic. G
louc
este
rshi
re,
Engl
and.
Fie
ld o
f vie
w 6
¥4
mm
.
(a)
(b)
(c)
(d)
Plat
e 12
(a,b
)Fus
ulin
id fo
ram
inife
r cem
ente
d by
radi
axia
l fibr
ous c
alci
te.T
he c
ryst
als a
reco
lum
nar a
nd c
loud
y w
ith in
clus
ions
, and
have
und
ulos
e ex
tinct
ion
unde
r cro
ssed
pola
rs. S
mal
l are
a of
cle
ar c
alci
te sp
ar:
(a)p
lane
-pol
ariz
ed li
ght,
fiel
d of
vie
w
6¥
4m
m; (
b)c
ross
ed p
olar
s, fi
eld
of v
iew
3¥
2m
m. C
apita
n, P
erm
ian.
Texa
s, U
SA.
(c,d
)Syn
taxi
al c
alci
te o
verg
row
th c
emen
ton
crin
oid
grai
n. E
arly
par
t of o
verg
row
thca
lcite
clo
udy
with
incl
usio
ns is
pro
babl
y a
mar
ine
prec
ipita
te; c
lear
late
r ove
rgro
wth
a bu
rial p
reci
pita
te. G
rain
s mai
nly
pelo
ids,
mic
ritiz
ed b
iocl
asts
and
faec
al p
elle
ts,
show
ing
conc
avo-
conv
ex/in
terp
enet
rativ
eco
ntac
ts, i
ndic
atin
g so
me
com
pact
ion.
Very
thin
isop
acho
us c
alci
te c
emen
tfr
inge
s aro
und
grai
ns se
en in
(d)a
repr
obab
ly m
arin
e pr
ecip
itate
s. R
ock
also
cut b
y th
in c
alci
te v
eins
. (c)
Plan
e-po
lariz
ed li
ght;
(d)C
ross
ed p
olar
s.C
reta
ceou
s. A
lps,
Fra
nce.
Fie
ld o
f vie
w
3¥
2m
m.
(a)
(b)
(c)
(d)
Plat
e 13
(a)C
alci
te c
emen
t at g
rain
con
tact
s and
irreg
ular
ly a
roun
d gr
ains
, ind
icat
ing
near
-su
rfac
e, m
eteo
ric v
ados
e en
viro
nmen
t.La
ter p
reci
pita
tion
of la
rge
poik
iloto
pic
calc
ite (b
lack
, in
extin
ctio
n) to
ok p
lace
durin
g bu
rial a
fter
som
e co
mpa
ctio
n.C
ross
ed p
olar
s. C
arbo
nife
rous
.G
lam
orga
n, W
ales
. Fie
ld o
f vie
w
0.8
¥0.
8m
m.
(b, c
)Cal
cite
spar
. (b
)Cal
cite
spar
und
er p
lane
-pol
ariz
edlig
ht sh
owin
g dr
usy
fabr
ic (c
ryst
al-s
ize
incr
ease
aw
ay fr
om th
e su
bstr
ate)
and
prom
inen
t tw
in p
lane
s. (c
)Sam
e fie
ld o
fvi
ew u
nder
cat
hodo
lum
ines
cenc
esh
owin
g de
licat
e gr
owth
zon
es re
sulti
ngfr
om su
btle
var
iatio
n in
man
gane
se a
ndiro
n co
nten
ts. T
riass
ic. G
lam
orga
n, W
ales
.Fi
eld
of v
iew
2¥
2m
m.
(d)S
utur
ed (m
icro
styl
oliti
c) c
onta
cts a
ndco
ncav
o-co
nvex
con
tact
s bet
wee
n gr
ains
(mic
ritiz
ed o
oids
and
bio
clas
ts) a
ndm
echa
nica
l fra
ctur
e of
low
er g
rain
.C
alci
te sp
ar b
etw
een
grai
ns is
a p
ost-
com
pact
ion
buria
l cem
ent.
Jura
ssic
.Bu
rgun
dy, F
ranc
e. F
ield
of v
iew
2¥
2m
m.
(e)M
icro
spar
–pse
udos
par f
orm
edth
roug
h ag
grad
ing
neom
orph
ism
with
foss
il re
lics (
brac
hiop
od sp
ine
on ri
ght)
.C
arbo
nife
rous
. Yor
kshi
re, E
ngla
nd. F
ield
of v
iew
2¥
2m
m.
(f)O
oliti
c gr
ains
tone
with
scat
tere
ddo
lom
ite rh
ombs
pre
cipi
tate
d af
ter e
arly
com
pact
ion
(see
gra
in c
onta
cts)
, bef
ore
calc
ite sp
ar c
emen
t. C
arbo
nife
rous
.G
lam
orga
n, W
ales
. Fie
ld o
f vie
w 2
¥2
mm
.
(a)
(b)
(c)
(d)
(e)
(f)
Plat
e 14
(a)D
olom
itize
d oo
lite
(no
relic
s of o
rigin
algr
ains
) with
styl
olite
, hig
hlig
hted
by
iron-
rich
clay
. Xen
otop
ic d
olom
ite (a
nhed
ral
crys
tals
) bel
ow st
ylol
ite a
nd id
ioto
pic
dolo
mite
(euh
edra
l cry
stal
s) a
bove
.In
terc
ryst
allin
e po
rosi
ty sh
own
thro
ugh
impr
egna
tion
with
blu
e re
sin.
Ara
bFo
rmat
ion.
Off
shor
e U
AE.
Fie
ld o
f vie
w 3
¥2
mm
. (b
)Dol
omiti
zed
grai
nsto
ne w
ithm
oder
ate
pres
erva
tion
of o
rigin
al o
oids
.In
terc
ryst
allin
e po
rosi
ty is
pre
sent
, sho
wn
up b
y im
preg
natio
n w
ith b
lue
resi
n.C
reta
ceou
s. O
ffsh
ore
Ang
ola.
Fie
ld o
fvi
ew 6
¥4
mm
. (c
)Bar
oque
dol
omite
: coa
rse
crys
tals
with
undu
lose
ext
inct
ion.
Roc
k is
an
oolit
e bu
tth
ere
are
no re
lics.
Cro
ssed
pol
ars.
Car
boni
fero
us. G
lam
orga
n, W
ales
. Fie
ldof
vie
w 2
¥2
mm
. (d
)Ded
olom
ite: c
rinoi
dal g
rain
ston
e w
ithov
ergr
owth
s con
tain
ing
scat
tere
ddo
lom
ite rh
ombs
that
hav
e be
en re
plac
edby
cal
cite
. The
dar
k m
ater
ial i
s iro
nox
ide/
hydr
oxid
e, su
gges
ting
the
dolo
mite
was
orig
inal
ly fe
rroa
n. C
arbo
nife
rous
.N
orth
umbe
rland
, Eng
land
. Fie
ld o
f vie
w 2
¥2
mm
. (e
)Dol
omite
mou
lds:
gra
inst
one
with
scat
tere
d do
lom
ite rh
ombs
that
hav
ebe
en d
isso
lved
out
to g
ive
a go
od p
oros
ity,
as sh
own
by b
lue
resi
n. S
tylo
litic
con
tact
sbe
twee
n gr
ains
. Jur
assi
c. B
urgu
ndy,
Fran
ce. F
ield
of v
iew
0.8
¥0.
8m
m.
(a)
(b)
(c)
(d)
(e)
Plat
e 15
(a)A
nhyd
rite:
min
ute
crys
tals
of a
nhyd
rite
form
ing
nodu
les w
ith so
me
clay
sedi
men
tbe
twee
n (b
row
n) a
nd la
rge
repl
acem
ent
anhy
drite
cry
stal
s. C
ross
ed p
olar
s.Pe
rmia
n. C
umbr
ia, E
ngla
nd. F
ield
of v
iew
6¥
4m
m.
(b)A
nhyd
rite
and
seco
ndar
y gy
psum
:sm
all c
ryst
als a
nd la
ths o
f anh
ydrit
e(b
right
col
ours
) bei
ng re
plac
ed b
y la
rge,
low
bire
frin
genc
e, p
orph
yrot
opic
gyp
sum
crys
tals
. Cro
ssed
pol
ars.
Per
mia
n(Z
echs
tein
). Te
essi
de, E
ngla
nd. F
ield
of
view
6¥
4m
m.
(c, d
)Hem
atiti
c iro
nsto
ne w
ith h
emat
ite-
impr
egna
ted
crin
oid
and
bryo
zoan
frag
men
ts in
a c
alci
te c
emen
t: (c
)pla
ne-
pola
rized
ligh
t; (d
)cro
ssed
pol
ars.
Bioc
last
ic g
rain
ston
e, C
arbo
nife
rous
Rhiw
bina
Iron
ston
e. G
lam
orga
n, W
ales
.Fi
eld
of v
iew
2¥
2m
m.
(e)B
erth
ierin
e–ch
amos
ite o
oids
show
ing
dist
orte
d sh
apes
and
ele
phan
tine
feat
ures
in c
alci
te c
emen
t. P
lane
-pol
ariz
ed li
ght.
Jura
ssic
. Raa
say,
Sco
tland
. Fie
ld o
f vie
w
2¥
2m
m.
(a)
(b)
(c)
(d)
(e)
Plat
e 16
(a)B
erth
ierin
e–ch
amos
ite o
oids
, with
som
e sh
ape
dist
ortio
n, in
a si
derit
ece
men
t par
tly a
ltere
d to
goe
thite
(bro
wn)
.Pl
ane-
pola
rized
ligh
t. Ju
rass
ic. Y
orks
hire
,En
glan
d. F
ield
of v
iew
3¥
2m
m.
(b)G
lauc
onite
gra
ins (
gree
n) in
dar
k ch
alk
with
pla
nkto
nic
fora
min
ifers
and
in a
pebb
le (o
n rig
ht) o
f pho
spha
te (p
ale
brow
n), w
ith a
ngul
ar q
uart
z gr
ains
inbo
th. P
lane
-pol
ariz
ed li
ght.
Cre
tace
ous.
Born
holm
, Den
mar
k. F
ield
of v
iew
3
¥2
mm
. (c
, d, e
)Pho
spho
rites
: all
plan
e-po
lariz
edlig
ht; a
ll fie
lds o
f vie
w 2
¥2
mm
. (c)
Phos
phor
ite p
elle
ts. P
erm
ian.
Idah
o, U
SA.
(d)F
ish
scal
es a
nd b
ones
in p
hosp
horit
em
udst
one.
Per
mia
n. Id
aho,
USA
. (e)
Bone
frag
men
ts. T
riass
ic. G
louc
este
rshi
re,
Engl
and.
(a)
(b)
(c)
(d)
(e)
Limestones 119
internal structure (growth lines) are preserved (seePlate 8a), and there are minute inclusions of aragoniteleft in the calcite (see Section 4.7). Calcitic bivalvesnormally will retain their original structure and themost common types are foliaceous (thin parallelsheets) and prismatic. Bivalve fragments in thin-section will be seen as elongate, rectangular to curvedgrains, typically disarticulated.
Gastropods are ubiquitous throughout shallow-marine environments. They also occur in vast num-bers, but low species diversity, in hypersaline andbrackish waters, such as on tidal flats and in estuaries,because certain species are able to tolerate fluctuationsand extremes of salinity. Most gastropods are benthic,vagile creatures. The encrusting vermetiform gastro-pods, often confused with serpulids, form reef-likestructures in the tropics and Carboniferous. The small,conical pteropods are important in Cenozoic pelagicsediments.
The majority of gastropods have shells of aragonitewith similar internal microstructures to bivalves. Theinternal microstructure of fossil gastropods also israrely seen because the original aragonite is mostly dis-solved out and the void filled by calcite cement. Gas-
tropod fragments can be recognized easily under themicroscope by their shape, although this is very muchdependent on the plane of section (Fig. 4.7 and Plate8b). Gastropods may resemble certain foraminifersbut the latter are usually much smaller and composedof dark micritic calcite. (See Plate 9c,d,e.)
Of the cephalopods, nautiloids and ammonoids arerelatively common in limestones of the Palaeozoic andMesozoic and belemnites occur in Mesozoic lime-stones. They were wholly marine animals with a domi-nantly nektonic or nekto-planktonic mode of life, aswith the modern Nautilus, octopus and cuttlefish. Thecephalopods are more common in pelagic, relativelydeep-water deposits. Examples include the Ordovi-cian–Silurian Orthoceras limestones of Sweden, theDevonian Cephalopodenkalk and Griotte of westernEurope and the Jurassic Ammonitico Rosso of the European Alpine region. Nautiloid and ammonoidshells were originally aragonitic and so in limestonesthey are typically composed of calcite spar with littleinternal structure. The shape, normally large size andpresence of septa are the features to note. Belemniteguards were made of calcite and have a strong radial-fibrous fabric in cross-section.
Bivalve
Brachiopod
Gastropod
Echinoderm Foraminifera
If originallycalcitic
If originallyaragonitic
or
Drusy sparite, nointernal structurepreserved
Neomorphic calcite withrelics of internalstructure
Structurepreserved
Variable shape
Typically drusysparite with nointernal structure
Originalstructure preserved endopunctae, pseudo-punctae+–
Variable shape
Single calcitecrystal
Cloudyappearance
Syntaxial over-growth of sparite
Echinoid spines
Variable shapee.g.
Micritic or fibrouswall textureFig. 4.7 Typical thin-section appearance
of bivalve, gastropod, brachiopod,echinoderm and foraminiferal skeletalgrains in limestone.
120 Chapter 4
Brachiopods
Brachiopods are particularly common in Palaeozoicand Mesozoic limestones of shallow-marine origin.These were largely benthic, sessile organisms; a fewspecies were infaunal. Only in rare cases, such as inthe Permian of west Texas, did the brachiopods contribute to reef development. At the present timebrachiopods are an insignificant group of marine invertebrates.
Although in section brachiopod shells are similar to those of bivalves in shape and size, most articulatebrachiopods were composed of low-Mg calcite, sothat the internal structure is invariably well preserved.The common structure is a very thin, outer layer of calcite fibres orientated normal to the shell surface,and a much thicker, inner layer of oblique fibres (Fig.4.7). It can be difficult to distinguish brachiopod shellsfrom bivalves with foliaceous calcite. However, cer-tain brachiopods have modifications to the shell, withpunctae and pseudopunctae. In the punctate bra-chiopods, such as the terebratulids, fine tubes (endo-punctae) perpendicular to the shell surface perforate the inner layer and are filled with sparite or micrite.Pseudopunctae, as occur in the strophomenid group,are prominent rod-like prisms within the shell. Inar-ticulate brachiopods, composed mostly of chitin orchitinophosphate, are rare in limestones. (See Plates8c,d & 11c.)
Cnidaria (especially corals)
The Cnidaria include the Anthozoa (corals), of whichtwo ecological groups exist today: hermatypic coralsthat contain symbiotic dinoflagellate algae (zooxan-thellae) in their polyps and ahermatypic corals with-out such algae. Because of the algae, hermatypic coralsrequire shallow, warm and clear seawater. They are the reef-forming corals at the present time, being mainly responsible for the reef framework, which is re-inforced by red algae. Ahermatypic corals can occur atmuch greater depths and tolerate colder waters. Theylocally form build-ups. The rugose and tabulate coralswere important in Silurian and Devonian reefs, andmany Triassic reefs contain scleractinian corals. Someof the latter may well have been ahermatypic. Corals,both solitary and colonial, and coral debris occur inmany non-reefal limestones.
The Palaeozoic rugose and tabulate corals were
composed of calcite, most probably high-Mg calcite,so that preservation generally is very good (see Plate9a). Scleractinian corals (Triassic to Recent), on theother hand, have aragonitic skeletons and so normallyare poorly preserved in limestones (see Plate 9b). Identification of coral is based on such internal fea-tures as septa and, where present, other internal platesin the Rugosa and Scleractinia, and tabulae in the Tabulata. Corallite form and colonial organization arealso important. The microstructure of Palaeozoic cal-citic and later aragonitic corals is very similar, chieflyconsisting of fibres in spherulitic or parallel arrange-ments, which form linear structures called trabeculae,or sheets.
Echinodermata
Echinoderms are wholly marine organisms that in-clude the echinoids (sea-urchins) and crinoids (sea-lilies). In modern seas, echinoids inhabit reef andassociated environments, locally in great numbers, butcrinoids are restricted to deeper waters and are in-significant as producers of carbonate sediment. In thePalaeozoic and Mesozoic, fragments of echinoderms,especially the crinoids, are a major constituent of bioclastic limestones. Many deep-water limestone turbidites are composed of crinoidal debris, derivedfrom shallow platforms.
Echinoid and crinoid skeletons are calcitic; modernforms generally have a high Mg content. Echinodermfragments are easily identified because they are com-posed of large, single calcite crystals, individual grainsthus showing unit extinction. In many cases, a sparitecement crystal has grown syntaxially around theechinoderm fragment (Fig. 4.7). Echinoderm grainshave a dusty appearance, especially relative to a sparitecement overgrowth, and they may show a porousstructure filled with micrite or sparite. (See Plates 9c,12a,b & 15c,d.)
Bryozoa
Although these small, colonial marine organisms are significant suppliers of carbonate sediment only locally at the present time (notably to cool-water carbonates, as off southern Australia), they have in the past contributed to the formation of reef andother limestones, particularly in the Palaeozoic. Examples include the Mississippian mud-mounds of
Limestones 121
southwestern USA and Europe, the Permian reefs ofTexas and western Europe, and the Danian Chalk of Denmark.
Modern bryozoan skeletons are composed of eitheraragonite or calcite (commonly high-Mg calcite) or amixture of both. There are many types of bryozoansbut the fenestrate variety, including the fenestellids,are seen most frequently in sections of Palaeozoic lime-stones. The skeleton consists of foliaceous calcite withround holes (the zooecia, where the individuals of thecolony used to live) filled with sparite or sediment (seePlate 15c).
Foraminifera
Foraminifera are dominantly marine Protozoa, mostlyof microscopic size. Planktonic foraminifers dominatesome pelagic deposits, such as the Globigerina oozesof ocean floors and some Cretaceous and Tertiarychalks and marls. Benthic foraminifers are common inwarm, shallow seas, living within and on the sediment,and encrusting hard substrates.
Foraminifera are composed of low- or high-Mg cal-cite, rarely aragonite. Foraminifera are very diverse inshape but in section many common forms are circularto subcircular with chambers. The test wall is dark andmicrogranular in many thin-walled foraminifers suchas the endothyracids and miliolids, but light-colouredand fibrous in larger, thicker species, such as the rotali-ids, nummulitids and orbitolinids. (See Plate 9c,d,e.)
Other carbonate-forming organisms
There are many other organisms that have calcareousskeletons but contributed in only a minor way to lime-stone formation, or were important for only short periods of geological time.
Sponges (Porifera). Spicules of sponges, which maybe composed of silica or calcite, occur sporadically insediments from the Cambrian onwards. The impor-tance of spicules is as a source of silica for the forma-tion of chert nodules and silicification of limestones(Section 4.8.4). At times, sponges provided the frame-work for reefs and mounds; examples include lithistidsponges in the Ordovician, calcisponges in the Permi-an of Texas and Triassic of the Alps, and silicisponges(now calcitized) in the Jurassic of southern Germany.Sclerosponges are the dominant reef-forming organ-
ism in some modern Caribbean reefs. Stromato-poroids, once considered hydrozoans, are now classi-fied as a subphylum of the Porifera. Stromatoporoidswere marine colonial organisms that had a wide vari-ety of growth forms, ranging from spherical to lami-nar, depending on species and environmental factors.Stromatoporoids were a major reef organism in the Sil-urian and Devonian, commonly in association withrugose and tabulate corals, and grew up to a metre ormore in size. Archaeocyathids, also probably sponges,formed reefs in the lower Cambrian of North America,Morocco, Siberia and South Australia.
Arthropods. Of this group, the ostracods (Cambrianto Recent) are locally significant in Tertiary limestonesand some others. They live at shallow depths in ma-rine, brackish or freshwater environments. Ostracodshave small (around 1mm in length), thin bivalvedshells, smooth or ornamented, composed of calcitewith a radial-fibrous structure. Trilobites (Cambrianto Permian), with a similar skeletal structure, occur lo-cally in Palaeozoic shelf limestones, but never in rock-forming quantities.
Calcispheres. These are simple spherical objects, upto 0.5mm in diameter, composed of calcite (usuallysparite), in some cases with a micritic wall. They areprobably some form of alga, although an affinity with Foraminifera has been suggested. They occur inmany Palaeozoic limestones, particularly fine-grainedmicrites of back-reef or lagoonal origin.
4.3.3 The contribution of algae and microbes tolimestones
Algae and microbes make a major contribution tolimestones by providing skeletal carbonate particles,trapping grains to form laminated sediments and at-tacking particles and substrates through their boringactivities. Many of the Precambrian limestones were atleast in part produced by microbes and algal–microbial limestones are widely distributed through-out the Phanerozoic. Four groups of algae are important: red algae (Rhodophyta), green algae(Chlorophyta), yellow–green algae (Chrysophyta)and cyanobacteria (formerly blue–green algae). Relevant texts are Walter (1976), Flügel (1977), Wray(1977), Monty (1981), Toomey & Nitechi (1985),Riding (1990) and Riding & Awramik (2000).
122 Chapter 4
Rhodophyta (red algae)
Calcareous algae of the Rhodophyta, such as theCorallinaceae (Carboniferous to Recent) and Soleno-poraceae (Cambrian to Miocene) have skeletons composed of cryptocrystalline calcite, which is pre-cipitated within and between cell walls. In section, a regular cellular structure is present (see Plates 10b &11b). Modern coralline algae have a high Mg contentin the calcite, which is related to water temperature(higher values for a given species in warmer waters).Many of the coralline algae encrust substrates and ifthis is a pebble or shell then nodules, referred to asrhodoliths, develop. Encrustations may be massiveand rounded, or delicately branched, depending onecological factors. One of the most important roles of these red algae is in coating, binding and cementingthe substrate, particularly in modern reefs. In tem-perate and arctic carbonate sands, the red alga
Lithothamnion is a major contributor. In the Palaeo-zoic and Mesozoic, red algae of the Solenoporaceaeare locally abundant and in some cases participate inreef formation.
Chlorophyta (green algae)
Three algal groups are important: the Codiaceae,Dasycladaceae (both Cambrian to Recent) andCharaceae (Silurian to Recent). The Characeae are incompletely calcified (low-Mg calcite), so that onlystalks and reproductive capsules are found in lime-stones. Modern (and fossil) forms are restricted tofresh or brackish water.
With the dasyclad algae, calcification is also incom-plete and involves the precipitation of an aragoniticcrust around the stem and branches of the plant. Dasy-clads are marine algae that tend to occur in shallow,protected lagoonal areas of the tropics. Under the microscope, circular, ovoid or elongate shapes areseen, representing sections through the stems orbranches (see Plate 10a).
The codiacean algae include Halimeda and Penicil-lus (Fig. 4.8), two common genera of Caribbean and
Fig. 4.8 Calcareous green algae. (a) Penicillus, which on deathand disintegration produce micron-sized aragonite needles. (b)Halimeda, which gives rise to sand-sized grains. Recent. FloridaKeys, USA.
(a) (b)
Limestones 123
Pacific reefs and lagoons. Halimeda is a segmentedplant that on death and disintegration generatescoarse, sand-sized particles. In thin-section, they looklike Swiss cheese. Penicillus, the shaving-brush alga, isless rigid, consisting of a bundle of filaments coated inneedles of aragonite. Death of this and other algae pro-vides fine-grained carbonate sediment (lime mud ormicrite, Section 4.3.4).
Phylloid algae are a group of late Palaeozoic algae that have a leaf or potato-crisp shape. Some be-long to the codiaceans, whereas others have moreaffinities with red algae. Phylloid algae are one of themain components of shelf-margin mud-mounds in theUpper Carboniferous to Lower Permian of southwestUSA.
Chrysophyta (coccoliths)
Coccolithophorids (Jurassic to Recent) are planktonicalgae that have a low-Mg calcite skeleton consisting ofa spherical coccosphere (10–100mm diameter) com-posed of numerous calcareous plates, called coccol-iths. In view of their size, these algae are studied withthe scanning electron microscope. The coccoliths arechiefly disc-shaped, commonly with a radial arrange-ment of crystals (Fig. 4.9). Coccoliths are a significant
component of modern deep-water carbonate oozes,particularly those of lower latitudes. They form thebulk of Tertiary and Cretaceous chalks and occur inred pelagic limestones of the Alpine Jurassic (see Fig.4.17).
Micrite envelopes
A large proportion of skeletal fragments in modernand ancient carbonate sediments possess a dark mi-crite envelope around the grains (see Plates 7b,c,d, 8b& 11c). The envelope is produced mostly by endolithicbacteria that bore into the skeletal debris. Followingvacation of the microbores (5–15mm in diameter),they are filled with micrite. Repeated boring and fillingresult in a dense micrite envelope, which is the alteredouter part of the skeletal grain. This process of graindegradation may eventually produce a totally mi-critized grain, i.e. a peloid (Section 4.3.1), devoid ofthe original skeletal structure (see Plates 7a, 8c, 12c &13d). Some bioclasts, especially the corals, molluscsand foraminifers, are more susceptible to microboringthan others, and this has implications for the preserva-tion of grain assemblages (Perry, 1998). Inorganic recrystallization of grains on the sea floor also takesplace, destroying the original skeletal structure (Reid& Macintyre, 1998).
Many other organisms bore into skeletal grains and carbonate substrates; examples include clionidsponges (in Plate 7e), bivalves (such as Lithophaga, seePlate 9b and Fig. 2.42d), and polychaetes, all of whichproduce larger borings and cavities, and fungi, whichproduce bores of 1–2mm in diameter. The micrite filling the borings may be precipitated physico-chemically or biochemically through decompositionof the microbes. Micrite envelopes due to endolithiccyanobacteria can be used as a depth criterion, indi-cating deposition within the photic zone (less than100–200m), but the grains can be transported togreater depths.
Stromatolites and microbialites
An important role of cyanobacteria (with other organ-isms such a diatoms, fungi and nematodes) is the for-mation of microbial mats, formerly called algal mats.These organic mats occur on sediment surfaces inmany low- to mid-latitude, marine and non-marineenvironments, from moderate depth subtidal through
Fig. 4.9 Scanning electron micrograph of coccoliths from pelagicooze, Shatsky Rise, northwest Pacific. Also present (lower right) isa larger discoaster, another type of nanoplanktonic alga.
124 Chapter 4
to supratidal marine areas, and fresh to hypersalinelakes and marshes. They form planar sheets, columnsand domes (Figs 4.10–4.13). The cyanobacteria aremainly filamentous varieties, common mat-forminggenera being Lyngbya, Microcoleus, Schizothrix, Scy-tonema and Oscillatorea, although unicellular coc-coid forms such as Endophysalis also occur. A matusually has a specific community that together with environmental factors produces a mat of particularmorphology and structure. Areas where microbialmats of various types are developing today include the Bahamas, the Arabian Gulf and Shark Bay, WesternAustralia.
The cyanobacteria are mucilaginous and this, to-gether with their filamentous nature, results in thetrapping and binding of sedimentary particles to produce a laminated sediment, a microbialite or stromatolite (Figs 4.12 & 4.14). Stromatolites occurthroughout the geological record but are particularlyimportant in the Precambrian, where they have beenused for stratigraphic correlation. The lamination inmany modern intertidal mats consists of couplets ofdark organic-rich layers alternating with light, sediment-rich laminae. Microbial filaments may be
Fig. 4.10 Tidal-flat microbial mats fromAbu Dhabi, Arabian Gulf. The microbialmats (dark areas) are desiccated into smallpolygons and have a thin covering ofrecently deposited carbonate sediment.Large polygonal desiccation cracks occurin the slightly higher, partly lithifiedgypsiferous carbonate sediment (lightarea). Knife (circled) for scale.
Fig. 4.11 Columnar stromatolites from agitated intertidal zone,Shark Bay, Western Australia. Columns increase in size upwardsand join with adjacent columns to form large domal structures.
Fig. 4.12 Microbially laminated carbonate sediments from anintertidal flat. Shark Bay, Western Australia. Desiccation cracksand cavities (fenestrae) are present.
Limestones 125
preserved (see Plate 10d). Laminae are usually lessthan several millimetres thick, but some sediment lam-inae may reach a centimetre or more. The alternatinglaminae reflect growth of the microbial mat (organiclayer) followed by sedimentation, and then trappingand binding of the sediment particles into the mat, asthe microbial filaments grow through to form a neworganic layer at the surface once more. In ancient stro-matolites, the laminae are usually alterations of densemicrite, perhaps dolomitized, and grains, such aspeloids and fine skeletal debris (see Plate 10f). Micro-bially laminated sediments commonly show small cor-rugations and irregularities in thickness, which serve
to distinguish them from laminae deposited purely byphysical processes. There may be evidence of desicca-tion —broken laminae, intraclasts and laminoid fenes-trae (see Plate 10e). The laminae may also constitutelarger-scale domes and columns. A diurnal growthpattern has been demonstrated for the laminae in somesubtidal mats, but in other cases, such as on tidal flatsand in ephemeral lakes, mat growth is probably sea-sonal or related to periodic wettings, and sedimenta-tion is erratic, being controlled largely by stormfloodings.
Microbial mats give rise to a range of laminatedstructures (microbialites). The simplest are planarstromatolites (or microbial laminites). They typicallydevelop on protected tidal flats and so may show desic-cation polygons (Fig. 4.10) and contain laminoid fen-estrae (elongate cavities: Fig. 4.12, Plate 10e andSection 4.6.3) and evaporite minerals or their pseudo-morphs. Domal stromatolites, where the laminae are continuous from one dome to the next, occur onthe scale of centimetres to metres. Columnar stroma-tolites are individual structures, which may be severalmetres high (Figs 4.13 & 4.14). Complex stromato-lites, such as occur in Precambrian strata, may be combinations of domes and columns, such as a largecolumnar structure with linked domes internally, orthey may show branching of columns. They can formbuild-ups many tens of metres high and laterally bevery extensive.
One further type of microbial-sediment structure isa nodule (ball), or oncoid (Fig. 4.15). Some have an internal concentric lamination, which may be asym-
Fig. 4.13 Stromatolite domes andcolumns in a few metres of water in ahigh-energy tidal channel. Exuma Cay,Bahamas.
Fig. 4.14 Stromatolite consisting of a large dome with internalcrinkly laminae. Late Precambrian. Anti-Atlas, Morocco.
126 Chapter 4
metric. Many oncoids are composed of dense micrite,whereas others are more clotted. The latter fabric alsooccurs in stromatolites where ones with a poor lami-nation and a clotted texture are called thrombolites.They probably have been formed by coccoidcyanobacteria.
In a notation for describing stromatolites (Fig. 4.16),domal stromatolites are referred to as laterally linkedhemispheroids (LLH), columnar stromatolites are vertically stacked hemispheroids (VSH) and oncoidsare spherical structures (SS). There is also a binomialclassification with form genera and species, particu-
larly used for Precambrian examples. Common generainclude Conophyton, Colleniaand Cryptozoon.
The morphological variation of microbialites de-pends largely on environmental factors such as waterdepth, tidal and wave energy, frequency of exposureand sedimentation rate. For example, large columnarstructures in Shark Bay (Fig. 4.11) are restricted to intertidal and subtidal areas in the vicinity of head-lands; small columns and domes occur in less agitatedbay waters, and low domes and planar mats dominateprotected tidal flats (Fig. 4.10). In the Bahamas, largestromatolite columns (almost ‘reefs’) are growing in
Fig. 4.15 Oncoids: spherical to irregularmicrobial balls. Some showing markedasymmetry, through periods of stationarygrowth. Carboniferous. Fife, Scotland.
Laterally linkedhemispheroids (LLH)
Vertically stackedhemispheroids (VSH)
Close lateral linkage, type LLH-C
Spaced lateral linkage, type LLH-S
Constant basal radius, type SH-C
Variable basal radius, type SH-VPlanar laminites or planar
stromatolites Oncoids(microbial balls)Typically slightly irregular
or crinkly lamination, may bedesiccated. Laminoid fenestraecommon Fig. 4.16 Common types of stromatolite
with terminology.
Limestones 127
shallow, high-energy tidal channels (Fig. 4.13; Macin-tyre et al., 1996). The microstructure of microbialmats and stromatolites is also variable; it is thoughtlargely to be a reflection of the microbial community,as noted above.
At the present time, microbial mats are quite re-stricted in their occurrence, being developed exten-sively only in hypersaline tidal-flat and freshwaterenvironments. Planar stromatolites are a feature ofperitidal facies throughout the geological record. Thedearth of domal and columnar stromatolites in thePhanerozoic, notably in shallow, normal-marine fa-cies, is attributed to the grazing activities of organisms,especially the gastropods. The absence of such meta-zoans in the Precambrian and earliest Phanerozoic wasone of the main factors in the widespread and diversestromatolite development at that time, including manyin subtidal and deeper-water settings.
Modern marine microbial mats are largely unlithi-fied, whereas those of fresh and hypersaline watersmay be cemented through biochemical or physico-chemical precipitation of carbonate. There are par-ticular microbes, common in the Palaeozoic, that dohave calcified filaments. The group name Porostroma-ta has been applied to these, which include Girvanella,Ortonella, Garwoodia and Cayeuxia. The structuresconsist of tubes or filaments with micritic walls, con-sidered to be calcified sheaths, arranged in an irregularspaghetti-type or more ordered, radial fashion. Theytypically form nodules (skeletal oncoids) but also stro-matolites. In some cases, the algae are associated inti-mately with encrusting foraminifers, as in Osagia andSphaerocodium nodules. More bush-like, branchingstructures (dendrolites) of calcified microbes are com-mon in the early Palaeozoic, commonly forming reefalstructures, and include Renalcis and Epiphyton (seePlate 10c).
For a discussion of the geological history of micro-bial carbonates see Riding (2000), and papers in Rid-ing & Awramik (2000).
4.3.4 Lime mud and micrite
Many grainy limestones have a fine, usually dark matrix and many others are composed entirely of fine-grained carbonate. This material is micrite (microcrys-talline calcite), with a grain size generally less than 4mm. Electron microscope studies have shown thatthe micrite is not homogeneous but has areas of finer or
coarser crystals, and intercrystalline boundaries thatmay be planar, curved, irregular or sutured (Fig. 4.17).Micrites are susceptible to diagenetic alteration andmay be replaced by coarser mosaics of microspar(5–15mm) through aggrading neomorphism (Section4.7.4).
Carbonate muds are accumulating in many modernenvironments, from tidal flats and shallow lagoons tothe deep-sea floor. There are many possible sources oflime mud. Carbonate muds of south Florida and theBahama Platform have been the subject of much re-search. They occur in the shallow subtidal, less-agitated central parts of the platform to the west of An-dros Island, and in lagoons, such as the Bight of Abaco.Fine carbonate sediments also occur on tidal flats andon the slopes and in deep basins around the platform(‘periplatform ooze’). The mud in many subtidal areasconsists predominantly of aragonite needles and lathsa few microns in length; only some 20% of the sedi-ment is recognizably biogenic. Inorganic precipitationas a result of evaporation has been postulated for theGreat Bahama Bank. The occasional ‘whiting’, a sud-den milkiness of the sea resulting from suspendedaragonite needles, may be the actual inorganic precip-itation taking place, although stirring of bottom mudsby shoals of fish can produce the same effect (see Shinn et al., 1989). The disintegration of calcareous
Fig. 4.17 Transmission electron micrograph of a Jurassic micriticpelagic limestone showing variation in size and shape of micritecrystals, and two coccoliths consisting of elongate, radiallyarranged crystals. Ammonitico Rosso, Jurassic. Austrian Alps.
128 Chapter 4
green algae has been widely regarded as a majorprocess of lime-mud production, from studies in SouthFlorida and the Bight of Abaco. When algae such asPenicillus (Fig. 4.8a) break down, a vast quantity ofaragonite is released. Measurements of growth ratesand calculations of the mass of aragonite producedhave shown that sufficient mud is produced to accountfor all the fine-grained sediments. Indeed, it appearsthat there is an overproduction so that algal disinte-gration in lagoons could be the source of mud forneighbouring tidal flats and peri-platform areas (Fig.4.18). Hence the shallow subtidal zone is referred to asthe ‘carbonate factory’. However, detailed studieswith the SEM and trace elements (e.g. Milliman et al., 1993) have concluded that much of the lime mud on the Great Bahama Bank is a direct precipitate,whereas that of Florida Bay is largely green algal in ori-gin. The Sr/Mg ratio of algae is less than two, whereasit is more than four in inorganic aragonite.
Three other processes that produce lime mud, but invariable or limited quantities are:1 bioerosion, where organisms such as boringsponges and microbes attack carbonate grains andsubstrates;2 mechanical breakdown of skeletal grains throughwaves and currents;3 biochemical precipitation through microbial photosynthesis and decomposition.
Carbonate mud, largely of skeletal origin, forms
subtidal banks in Florida and Shark Bay, where sea-grasses and algae trap and bind the sediment (e.g.Bosence, 1995). Carbonate muds of the deep-oceanfloors are oozes composed chiefly of coccoliths, withlarger foraminiferal and pteropod grains.
In the lagoons along the Trucial Coast of the Arabian Gulf, inorganic precipitation is probably tak-ing place. The aragonite needle muds contain highstrontium values (9400p.p.m.), close to the theoreti-cal for direct precipitation from that lagoonal water.There is also a paucity of calcareous algae in the regionand the possibility of other aragonitic skeletons con-tributing is precluded by their low strontium values.
There usually is little evidence in a limestone for theorigin of the micrite. Nanofossils, in particular coccol-iths, can be recognized with the electron microscope insome pelagic limestones (Fig. 4.17), but on the wholethere is little to indicate a biogenic origin for the majority of ancient micrites. From the studies of modern lime muds, it is tempting to suggest that manyancient shallow-marine micrites were the product ofcalcareous green-algal disintegration. However, thepossibility of inorganic precipitation in the past can-not be ruled out.
The original mineralogy of lime muds will have beenimportant in their diagenesis; those with a high con-tent of aragonite will have been more susceptible to neomorphism and microspar formation. Originalaragonite-dominated precursor muds can be recog-nized through SEM study of lightly etched micrites, re-vealing minute crystal relics of aragonite (Lasemi &Sandberg, 1984). In grainstones and other coarse lime-stones, micrite could well be a cement, rather than a matrix (Section 4.7.1). In addition, fine carbonatesediment up to silt grade may filter into a porous lime-stone soon after deposition or during early diagenesis.Such geopetal or internal sediment can be recognizedby its cavity-filling nature (see Plates 9a,b & 11a,c,d).
4.4 Classification of limestones
Three classification systems are currently used, eachwith a different emphasis, but the third, that of Dunham, based on texture, is now used more widely.1 A very simple but often useful scheme divides lime-stones on the basis of grain size into calcirudite (mostgrains >2mm), calcarenite (most grains between 2mmand 62mm) and calcilutite (most grains less than 62mm).
Mechanical disaggregation ofcarbonate grains
Bioerosion
Chemical/biochemicalprecipitation
Formation
Pellets
Disaggregation
Into solution
Mudtransportedoffshore insuspension
Erosion ofDeposition on
Tidal flats
Disaggregationof calcareousgreen algae
Lime mud<62μm
on lagoon floor
Fig. 4.18 Lime-mud budget for a lagoon in the Bahamas.
Limestones 129
2 The classification scheme of R.L. Folk (Fig. 4.19),based mainly on composition, distinguishes threecomponents: (a) the grains (allochems), (b) matrix,chiefly micrite and (c) cement, usually drusy sparite.An abbreviation for the grains (bio —skeletal grains,oo —ooids, pel —peloids, intra —intraclasts) is used asa prefix to micrite or sparite, whichever is dominant.Terms can be combined if two types of grain dominate,as in biopelsparite or bio-oosparite. Terms can bemodified to give an indication of coarse grain size, as in biosparrudite or intramicrudite. Other categories ofFolk are biolithite, referring to a limestone formed insitu, such as a stromatolite or reef-rock; and dismi-crite, referring to a micrite with cavities (usually spar-filled), such as a birdseye limestone (Section 4.6.3).3 The classification of R.J. Dunham (Fig. 4.20) di-vides limestones on the basis of texture into: grain-stone, grains without matrix (such as a bio- or
oosparite); packstone, grains in contact, with matrix(this could be a biomicrite); wackestone, coarse grainsfloating in a matrix (could also be a biomicrite); and amudstone, micrite with few grains. Additional termsof A.F. Embry & J.E. Klovan give an indication ofcoarse grain size (floatstone and rudstone), and of thetype of organic binding in boundstone during depo-sition (bafflestone, bindstone and framestone). Theterms can be qualified to give information on composi-tion, e.g. oolitic grainstone, peloidal mudstone orcrinoidal rudstone.
As a result of diagenetic modifications to lime-stones, care must be exercised in naming the rock. Forexample, a homogeneous-looking micrite may be apeloidal mudstone, and micrite in a bioclastic, grain-supported rock could be cement, compacted pellets(i.e. grains), primary sediment (i.e. matrix) or internalsediment (infiltrated geopetal sediment). The second
Skeletal grains(bioclasts)
Principal grainsin limestone
Ooids
Peloids
Intraclasts
Limestone formedin situ
Biosparite
Oosparite
Pelsparite
Intrasparite
Biolithite
Cemented by spariteLimestone types
Biomicrite
Oomicrite
Pelmicrite
Intramicrite
Fenestrallimestone–dismicrite
With a micrite matrix
Fig. 4.19 Classification of limestonesbased on composition.
Original components not boundtogether during deposition
Contains lime mud
Mud-supported
Less than10%
grains
More than10%
grains
Grain-supported
Lacks mudand isgrain
supported
Originalcompon-
entsbound
together
Deposit-ional
texturenot
recogniz-able
Crystallinecarbonate
Original componentsnot organically bound
during deposition
Original componentsorganically boundduring deposition
>10% grains >2mm
Matrixsupported
Supportedby > 2mmcompon-
ents
Organismsact as
baffles
Organismsencrust
and bind
Organismsbuild arigid
framework
Mudstone Wackestone Packstone Grainstone Boundstone Crystalline Floatstone Rudstone Baffle stone Bindstone Framestone
Fig. 4.20 Classification of limestones based on depositional texture.
130 Chapter 4
example is basically a question of separating a grain-stone from a packstone/wackestone. Many pack-stones are actually compacted wackestones.
The depositional environments and facies of lime-stones are considered in Section 4.10, where the typi-cal limestone rock types of each environment arediscussed briefly. The composition of limestones canbe taken further by point-counting the different graintypes present and calculating their percentages. A vis-ual estimation of the grain percentages can be madeusing Fig. 2.49 in Chapter 2. Triangular diagrams canillustrate the composition for three principal compo-nents (e.g. ooids, bioclasts and peloids, or crinoids,brachiopods and bivalves). In this way different microfacies can be recognized.
4.5 Limestone grain size and texture
For the most part carbonate sediments are formed insitu. Although some may be transported from shelf tobasin by turbidity currents or slumps, and from aninner to outer shelf by storms, the majority of lime-stones accumulated where the component grains wereformed or have been subjected to only limited trans-port by wave and tidal currents. The skeletal grains of carbonate sediments vary greatly in size and shape.Interpretations of limestone deposition are thus to alarge extent based on the types of grain present be-cause these will often provide concise information onthe depth, salinity, degree of agitation, etc. This is notto say that the grain size and degree of sorting androunding are unimportant. Although the grain sizewill be a reflection largely of the size of the carbonateskeletons of the organisms living in the area and of themany biological factors involved in their breakdown,the physical factors of waves and currents will alsocontribute, and in cases dominate. A measure of thegrain size, then, will often give useful additional infor-mation, reflecting the energy level of the environment,or energy gradient of the area. Where one is dealingwith grainstones, the grain-size parameters discussedin Section 2.2.1 can be applied. It must be borne inmind, however, that carbonate particles are hydrody-namically different from quartz grains (see Kench &McClean, 1996). Apart from complications arisingout of shape, carbonate grains commonly have a lowerdensity because of pores and contained organic matter.The degree of sorting and rounding of skeletal grains(use Figs 2.4 and 2.6) can be useful in certain bioclastic
rocks, such as those of shelves and ramps wherechanges in these features could indicate proximity to ashoreline or zone of higher wave and tidal current ac-tivity. Some limestones of course, such as oolitic andpeloidal grainstones, are very well sorted and roundedanyway. In interpreting the energy level of a deposi-tional environment from a rock’s grain-size param-eters and texture, one is assuming that the sedimentsurface was in equilibrium with the hydrodynamicregime. With carbonate sediments this may not havebeen the case. In the modern shallow-marine environ-ment, the sediment surface is commonly covered in asurficial microbial mat that stabilizes the sediment, en-abling it to withstand current velocities up to five timesthose eroding nearby sediments lacking a microbialcover. During diagenesis, evidence of the mat could be destroyed. The probable wackestone would be in-terpreted as a low-energy deposit, whereas in fact peri-odically it was subjected to high current velocities.
Measurements of carbonate grain size in ancientlimestones are made by point-counting slides or ac-etate peels under the microscope. For loose carbonatesands, the use of a settling chamber/sedimentation balance is recommended because this gives a better indication of the hydraulic behaviour than the grain-size distribution obtained by sieving.
In a general way the amount of micrite or lime mudin a limestone reflects the degree of agitation; limemuds tend to be deposited in quiet lagoons or on outerramps, as well as on tidal flats and in the deep sea, inbasins and periplatform areas. Increasing agitationleads to a decrease in the micrite content and increasein grain-support fabric and sparite content; sortingand rounding of grains then improves in the grain-stone/biosparite. Interpretations must be made withcare though because lime muds can accumulate inhigher-energy environments, trapped and stabilized by sea-grass or a surficial microbial mat, which leavesno record in the sediment, and micrite can be precipi-tated as a cement during early diagenesis (Section4.7.1).
4.6 Sedimentary structures of limestones
Limestones contain many of the sedimentary struc-tures occurring in sandstones described in Section 2.3,but some structures are found only in carbonate sedi-ments. Demicco & Hardie (1995) presented a descrip-tion of sedimentary structures in limestones.
Limestones 131
4.6.1 Bedding planes, hardgrounds, tepees andpalaeokarstic surfaces
As in siliciclastic sediments, bedding planes generallyrepresent a change in the conditions of sedimentation.The changes may have been subtle or short-lived. Thebedding planes are mostly the result of changes in sediment grain size or composition. Thin clay seamsalso commonly define the bedding in limestone succes-sions. However, with limestones it is not uncommon tofind that bedding planes have been affected by dissolu-tion as a result of overburden pressure (Section 4.7.5).Through this, originally gradational bed boundaries,such as a limestone passing up into a mudrock, or agrainstone into a lime mudstone, may become sharp.In many platform limestones, the ‘bedding’ planes arenot primary depositional surfaces, but they have beenproduced by pressure dissolution during burial. This ismost obvious where, for example, a ‘bedding’ surfaceoccurs within a graded bed, or where it cross-cuts, at alow angle, a clear primary bedding surface. Thesepseudobedding planes actually account for much ofthe stratification in shallow and deep-water limestones(Simpson, 1985).
One particular type of bedding plane is a hard-ground surface. Hardgrounds are horizons of synsedi-mentary cementation, taking place at or just below thesediment surface. Where a hardground surface formedthe sea floor, it was commonly encrusted by sessile benthic organisms, such as corals, serpulids, oysters,foraminifers and crinoids, and bored by polychaeteannelids, certain bivalves and sponges. Hardgroundsurfaces may cut across fossils and sedimentary struc-tures. Two types of hardground surface can be recog-nized: a smooth, planar surface, formed by abrasion(Fig. 4.21), and an irregular, angular surface formedby dissolution (a corrosional hardground surface).The first type is more common in shallow subtidal sediments where waves and currents are able to move oolitic and skeletal sands across lithified sediment to produce a planar erosional surface. Corrosional hardground surfaces are more com-mon in pelagic limestones where periods of non-sedimentation allow sea-floor cementation and disso-lution. The identification of a hardground is importantbecause it demonstrates synsedimentary submarinecementation. Hardground surfaces may become mineralized and impregnated with iron hydroxides,Fe–Mn oxides, phosphate and glauconite. Hard-
grounds develop from loose sediments through firm-grounds to lithified layers, and associated with thisthere may be a change in the fauna, particularly of theburrowing organisms, as the sedimentation rateslowed down. Hardgrounds occur throughout thePhanerozoic and modern ones are forming at the pre-sent time off Qatar in the Arabian Gulf and onEleuthera Bank, Bahamas (also Section 4.7.1).
One distinctive feature of peritidal limestones is thetepee structure (Fig. 4.22). Tepees are disruptions ofthe bedding into ‘pseudoanticlines’ and in plan viewthe tepee crests form a polygonal pattern. Tepees occuron the scale of tens of centimetres to several metresacross. They mostly form on intertidal–supratidal flatsas a result of the cementation and expansion of the surface-sediment layer. Upward movement (resur-gence) of ground water, marine or meteoric, is a con-tributory factor in some cases. Elongate cavities (sheetcracks) commonly form beneath the uplifted slabs andin these, pisoids may form, as well as vadose cementssuch as dripstone and flowstone. These tepees usuallyare associated with planar stromatolites, desiccationcracks and intraclast conglomerates. Modern exam-ples of these tepees, some with the spelean–pisoid as-sociation, occur on supratidal flats and around salinelakes in South and Western Australia. Tepees also formin the submarine environment, for example, wherehardground surfaces have expanded through the ce-ment precipitation. These are well developed off the
Fig. 4.21 Planar hardground surface encrusted with Ostrea(oysters) and penetrated by annelid and bivalve borings (seen oncut face). These features demonstrate synsedimentary sea-floorcementation of the sediment (see also Plate 11d). Jurassic.Gloucestershire, England.
132 Chapter 4
Qatar Peninsula. Tepees are prominent in many an-cient peritidal sequences; classic examples occur in theback-reef facies of the Permian Capitan reef complexof Texas–New Mexico and in the Triassic of theDolomites of northern Italy. See the review of Kendall& Warren (1987) for more information and M. Mutti(1994).
Another particular type of bedding discontinuitypeculiar to limestones is a palaeokarstic surface (Fig.4.23). When carbonate sediments become emergent,then dissolution through contact with meteoric water
produces an irregular, pot-holed surface. This dissolu-tion commonly takes place beneath a thin soil cover,and the soil itself may be preserved as a discontinuousclay seam or bed immediately above the dissolutionsurface. The term karst is applied to these dissolutionfeatures, which are typical of more humid climaticareas. In a well-developed karst system, pot holes and caverns may form many tens or even hundreds of metres below the surface. Breccias form by cave collapse and deposition from subterranean streams;speleothems and flowstones are precipitated too. Im-portant karst developed in the Lower Ordovician ofTexas and Oklahoma, and in central China (OrdosBasin) and are major hydrocarbon reservoirs. Exam-ples are given in Esteban & Klappa (1983) and James& Choquette (1988); also see Vanstone (1998), Molina et al. (1999) and Purdy & Waltham (1999).
Laminated crusts form upon and within upliftedcarbonate sediments, as a type of calcrete or caliche(Section 4.10.1). In most cases, they are calcified rootmats (Wright et al., 1988), and usually they are closelyassociated with vadose pisoids and black pebbles. In ancient limestones, laminated crusts can be mis-taken for stromatolites, but their association withpalaeokarstic surfaces and palaeosoils indicates a sub-aerial, pedogenic origin.
Bedding surfaces and their significance in carbon-ates are discussed in Hillgärtner (1998).
4.6.2 Current and wave structures
All the current structures of siliciclastic rocks occur in limestones: wave and current ripples, cross-lamination, cross-bedding on all scales, planar or flatbedding, small scours to large channels, HCS in stormbeds, bundled cross-beds and reactivation surfaces intidal sands and sole structures on the bases of stormbeds and turbidites. Post-depositional structures re-sulting from dewatering and loading are also com-mon. For details of these sedimentary structures seeSection 2.3. The lack of clay in many limestones, to-gether with the effects of surface weathering, maymake internal structures difficult to discern in the field.Careful observations, perhaps aided by polishing,etching or staining cut blocks, will often bring them tolight.
The same importance is attached to current struc-tures in limestones as in sandstones. They are essentialto environmental interpretation and facies analysis,
Fig. 4.22 Tepee structure in Triassic lacustrine littoral dolomite.Glamorgan, Wales.
Fig. 4.23 Palaeokarst. Cretaceous limestone (white) ispenetrated by Eocene sands (grey), which also contain blocks ofthe host rock. Sub-Alpine Chains, France.
Limestones 133
giving valuable information on depositional process,palaeocurrents, depth and water turbulence. Intra-clasts of lime mud (flakes) and large fragments ofskeletal debris are generally more common in lime-stones. They may be concentrated through currentwinnowing to form lag deposits, or transported bystorms to give rudstones and floatstones (flakestones,Fig. 4.5). These beds may show imbrication of clasts,reverse or normal grading. Elongate fossils are com-monly aligned parallel to, or normal to, the current direction, so that they too can give a palaeocurrent in-dication.
4.6.3 Cavity structures
Depositional and early diagenetic cavity structures are common in limestones, and there are many types.Some are partly filled with sediment that has beenwashed in to occupy the lower part of the cavity, withthe space above occupied by sparite cement (see Plates9a,b & 11a,c,d). Such cavity fills are known asgeopetal structures, and they are a most useful way-upindicator (sparite at the top of course). Geopetal struc-tures also record the horizontal at the time of sedimen-tation (acting as a spirit level) and in some cases showthat there was an original depositional dip (as in fore-reef limestones, for example). Umbrella structuresare simple cavities beneath convex-up bivalve and brachiopod shells and other skeletal fragments. In-traskeletal cavities occur in enclosed or chamberedfossils, such as gastropods, foraminifers and am-monoids. Growth cavities are formed beneath theskeletons of frame-building organisms, corals andstromatoporoids, for example, where they build outabove the sediment or enclose space within the skeletons.
Fenestral cavities or ‘birdseyes’ are small cavitiesthat occur particularly in peloidal mudstones of inter-tidal–supratidal environments. The majority are spar-filled only, but some may be sediment-filled. Threemain types can be distinguished:1 irregular fenestrae, the typical ‘birdseyes’ (Fig.4.24), several millimetres across, equidimensional toirregular in shape;2 laminoid fenestrae, several millimetres high andseveral centimetres long, parallel to bedding (Figs 4.12& 4.25);3 tubular fenestrae, cylindrical, vertical to subverticalin arrangement, several millimetres in diameter.
Irregular fenestrae in abundance form the so-calledbirdseye limestones (my favourite). They are ascribedto gas entrapment in the sediment and desiccation andso are a characteristic intertidal-facies indicator. Simi-lar structures also occur in subtidal grainstones associ-ated with early cementation and hardgrounds. Beachsands also contain irregular fenestrae, called keystonevugs, from the movement of water and air through thesediment.
Laminoid fenestrae develop in laminated sedi-
Fig. 4.24 Birdseyes (fenestrae) in lime mudstone. Fenestrae arefilled with sparry calcite although some contain internalsediment. Carboniferous. Clwyd, Wales.
Fig. 4.25 Laminoid fenestrae and birdseyes filled with calcite inmicrobial and peloidal dolomite. The dolomite is probablypenecontemporaneous in origin. Triassic. Austrian Alps.
134 Chapter 4
ments, particularly planar stromatolites, from thedecay of organic matter, and desiccation and parting oflaminae. Tubular fenestrae are formed mainly by bur-rowing organisms, but plant rootlets produce similartubes. Although common in tidal-flat sediments, tubu-lar fenestrae will also occur in shallow, subtidal sedi-ments. Fenestral limestones occur throughout thegeological record; one notable example is the loferiteof the Alpine Triassic (Fig. 4.25).
Small vugs to extensive cavern systems on the scaleof tens of metres form as a result of surface and near-surface karstic dissolution of limestones (see Section4.6.1). Dissolution also may take place during deepburial in the formation of hydrothermal karst.
A further particular type of cavity is stromatactis(Fig. 4.26). This is common in carbonate mud-moundsof the Palaeozoic such as the Waulsortian ‘reefs’ andother mounds of the European Carboniferous and Devonian, the Devonian Tully Limestone of NewYork, Silurian sponge ‘reefs’ of Quebec and Ordovi-cian mud-mounds of Nevada. Stromatactis cavitieshave an irregular, unsupported roof and a flat floor,formed by internal sediment. The cement is invariablya first generation of fibrous calcite, followed by drusysparite. Most stromatactis cavities are a few centime-tres long, but they may reach tens of centimetres. Theorigin of the cavities has led to much discussion andspeculation. Stromatactis in Upper Devonian mud-mounds has been interpreted as the product of recrys-tallization of algal, cyanobacterial and bacterialcolonies. Fibrous calcite, which is a feature of moststromatactis cavities and normally is thought of as a
marine cement (see Section 4.7.1), also has been inter-preted as microbial in origin. Various inorganic originsare available: sediment collapse and dewatering; win-nowing of uncemented sediment beneath lithifiedcrusts or beneath gelatinous microbial mats; and dis-solution during deep burial. A popular mechanismnow is that stromatactis formed within a spongeframework, from the decay of uncemented sponge tis-sue (Bourque & Boulvain, 1993; and papers in Montyet al., 1995). It is likely that stromatactis structures canform through several of these processes.
Two types of cavity that form in partly lithified or cemented limestone are sheet cracks and neptuniandykes. Sheet cracks are cavities generally running par-allel to the bedding, which have planar walls, althoughsome may have irregular roofs. Neptunian dykes cutacross the bedding and may penetrate down many metres from a particular bedding plane. Both sheetcracks and neptunian dykes are filled with internal sediment, in some cases with fossils a little youngerthan the adjacent limestone, if the cavities opened onto the sea floor. Spectacular examples of both types, inpelagic limestones and penetrating down into underly-ing platform carbonates, occur in the Triassic andJurassic of the Alps and Spain (e.g. Molina et al.,1995). Sheet cracks and neptunian dykes formthrough small tectonic movements during sedimenta-tion and/or some slight downslope movement of sedi-ment, causing fracturing of the lithified or partlylithified limestone mass.
Fractures and veins filled with calcite, of varioustypes, are common in limestone. Although they can beearly structures, like neptunian dykes, the majorityform through later fracturing, and this is usually in response to tectonic stresses rather than purely diagenetic processes.
4.7 Carbonate diagenesis
The diagenesis of carbonates involves many differentprocesses and takes place in near-surface marine andmeteoric environments, down into the deep-burial en-vironment. It is most important in occluding and gen-erating porosity in the sediment. Six major processescan be distinguished: cementation, microbial micriti-zation, neomorphism, dissolution, compaction anddolomitization. Carbonate diagenesis mostly involvesthe carbonate minerals, aragonite, calcite and dolo-mite, but other minerals such as quartz, feldspar, clays,
Fig. 4.26 Stromatactis cavities filled with fibrous calcite anddrusy calcite. Mud-mound, Upper Devonian. Ardennes, Belgium.
Limestones 135
phosphates, iron oxides and sulphides, and evaporitesalso may be involved. The diagenesis of carbonate sed-iments begins on the sea floor; in fact depositional anddiagenetic processes may be going on at the same time.As a reef is growing or carbonate sand is being movedby the waves, cements may be precipitated within in-traskeletal cavities and grains altered by micritization.The latter process has been described in Section 4.3.3,and results in the formation of micrite envelopesaround bioclasts and completely micritized grains.Micrite envelopes play an important role during diage-nesis by maintaining the shape of an aragonite bioclas-tic grain after its dissolution.
Cementation is the major diagenetic process pro-ducing a solid limestone from a loose sediment andtaking place principally where there is a significantthroughput of pore-fluid saturated with respect to thecement phase. The mineralogy of the cements dependson water chemistry, particularly Pco2 and the Mg/Caratio, and carbonate supply rate (see Section 4.7.1).Neomorphism is used to describe replacement and re-crystallization processes where there may have been achange of mineralogy. Examples include the coarsen-ing of crystal sizes in a lime mud/micrite (aggradingneomorphism) and the replacement of aragonite shellsand cements by calcite (calcitization). Many lime-stones have suffered dissolution as a result of the pas-sage of pore-fluids undersaturated with respect to thecarbonate phase present. This is a major process innear-surface, meteoric diagenetic environments, andmay lead to the formation of karst (see Section 4.6.1),but it can also take place on the sea-floor and duringdeep burial. The secondary porosity created by car-bonate dissolution is important in some hydrocarbonreservoirs. Compaction takes place during burial, re-sulting in a closer packing of grains, their fracture andeventual dissolution where in contact. Chemical com-paction leads to stylolites and dissolution seams, whenburial exceeds many hundreds of metres of over-burden. Dolomitization is a major alteration processfor many limestones and the dolomite, CaMg(CO3)2,may be precipitated in near-surface and burial envi-ronments. There are a number of models for dolo-mitization, but the matter is still one of great debate.
Three major diagenetic environments are distin-guished: marine, near-surface meteoric and burial (seeFig. 4.27). In the marine environment, diagenesis takesplace on and just below the sea-floor in both shallowand deep water, and in the intertidal–supratidal
zone. Meteoric diagenesis can affect a sediment soonafter it is deposited if there is shoreline progradation ora slight sea-level fall, or it may operate much laterwhen a limestone is uplifted after burial. The burial environment, the least well known, begins at a depthbelow the sediment surface of tens to hundreds of metres, that is, below the zone affected by surface pro-cesses, down to several kilometres where meta-morphic dehydration reactions and wholesalerecrystallization take over.
4.7.1 Marine diagenesis
Marine diagenesis in Recent carbonate sediments
Intertidal–supratidal diagenesis. Cementation in theintertidal zone produces cemented beach sands knownas beachrock. Beachrocks are most common in thetropics and subtropics but they do occur along temper-ate shorelines. They are composed of the same sedi-ment that forms the surrounding loose beach sand; thisusually is calcareous but it may have a substantial oreven dominant siliciclastic component. Beachrockscan form quickly, as is evidenced by the inclusion ofanthropogenic objects such as beer cans. Beachrockformation probably takes place a few tens of centime-tres below the surface of the beach, but in many placesit is exposed through storm action, and then it can beeroded (to produce intraclasts), encrusted, and boredand grazed by intertidal organisms.
Meteoricvadose
Marinevadose
Protectedinner shelf
Shelf-margin Basin
Shallow to deepburial environment
Marine phreaticMeteoricphreatic
Mud+muddy sandSandshoals Reef Fore-reef
Active circulation
Little circulation
Lake
Mixingzone
sl
Fig. 4.27 Carbonate diagenetic environments, schematicallydrawn for a rimmed shelf with unconfined aquifers. Where thereare confined aquifers as a result of impermeable strata, then it ispossible, for example, for meteoric waters to penetrate deepbeneath the marine shelf and even emerge on the sea floor assubmarine springs.
136 Chapter 4
The cements in modern beachrocks are aragoniteand/or high-Mg calcite. Aragonite typically occurs asfringes from 10 to 200mm thick of acicular crystals,orientated normal to the grain surfaces (Figs 4.28 &4.29). In many cases the cement fringes are iso-pachous, i.e. of equal thickness (Fig. 4.30), indicating marine phreatic (below the water table) precipitation
where pores were constantly water-filled. Asymmetriccement fringes, thicker on the underside of grains, andmeniscus cements, concentrated at grain contacts (Fig.4.30), are recorded from some beachrocks and indi-cate precipitation in the marine vadose zone. High-Mgcalcite is usually a dark micritic cement coating grainsor filling pores. Micritization of grains is common inbeachrocks and calcified microbial filaments are alsopresent. If the meteoric ground-water table is high inthe backshore area, then low-Mg calcite cements maybe precipitated in beachrocks there in the upper inter-tidal zone.
Two processes are important in beachrock formation:1 purely physico-chemical precipitation throughevaporation of seawater when the tide is out and CO2-degassing of seawater as it is pumped through the sandby waves and the rising and falling tide;2 biochemical–microbial precipitation, involving microbial photosynthesis, bacterial calcification andthe decomposition of organic matter.
Cemented surface crusts occur in the high intertidalto low supratidal zone of tidal flats in carbonate areas. Along the Trucial Coast of the Arabian Gulf,aragonite-cemented crusts are commonly brecciatedor polygonally cracked, and expansion of the crusts asa result of the cementation leads to pseudoanticlines or tepee structures (Section 4.6.1; Kendall & Warren,1987). The resurgence of continental ground wateronto supratidal flats may contribute to crust and tepeeformation. Beneath the surface crust, dripstone cements, vadose pisoids and aragonite botryoids
Fig. 4.28 Beachrock with cements of acicular aragonite andmicritic high-Mg calcite: the latter is the dark coating aroundgrains. The aragonite cement has grown syntaxially on the bivalvegrain, which shows a two-layer shell structure. Plane polarizedlight. Great Barrier Reef, Australia.
Fig. 4.29 Scanning electron micrograph of beachrock cementshown in Fig. 4.28. Where the coating of acicular aragonitecrystals has come off the grain, the high-Mg calcite micritecement is visible.
Isopachouscement
Gravity (stalactitic) andmeniscus cements
Fig. 4.30 The geometry of first-generation cements: isopachouscement, indicative of precipitation in phreatic zones where allpores are filled with water (typical feature of low-intertidal andsubtidal cements); and gravity (stalactitic) and meniscus cements,indicative of vadose-zone precipitation, as occurs in high-intertidal, supratidal and shallow-subsurface meteoric situations.
Limestones 137
may develop. Supratidal crusts in the Bahamas are cemented by dolomite (Section 4.8.1, see Plate 10d).
Shallow-subtidal cementation. Sea-floor diagenesisin low-latitude, shallow-marine areas mostly involvescementation and microbial micritization. The formeris more common in high-energy areas, where seawateris pumped through the sediment, whereas the latter ismore common in quieter-water areas, such as back-reef lagoons. Thus an active marine phreatic diage-netic environment can be distinguished from astagnant one. In higher-latitude, shallow-water areas,seawater is undersaturated with respect to CaCO3 andso inorganic precipitation of cement does not takeplace. Dissolution of grains may occur, however.
In the stagnant, marine phreatic diagenetic environ-ment, cementation is restricted to intraskeletal cav-ities, such as occur within gastropod and foraminiferalbioclasts. The formation of grapestones and aggre-gates is common in this environment, with microbialbinding and filament calcification usually involved. In these areas too, micritization of grains by recry-stallization is taking place on the sea floor (Reid &Macintyre, 1998).
Shallow-subtidal cementation of loose carbonatesand to produce surface crusts and lithified layers israre but it is taking place in a few metres of water offthe Qatar Peninsula, Arabian Gulf and on EleutheraBank, Bahamas. Off Qatar, these modern hard-grounds are being bored and encrusted by organisms,and polygonal cracks and tepee structures haveformed through expansion of the cemented layer. Thecements are mainly acicular aragonite, with some micritic high-Mg calcite. The acicular cements formisopachous fringes (see Plate 11b), but if well devel-oped and completely filling the pore space, then apolygonal pattern is observed in thin-section, from the meeting of the cement fringes on adjacent grains.The precipitation of cements is probably the result ofturbulent bottom conditions and the pumping ofCaCO3-supersaturated seawater through the sedi-ments, particularly in areas of slow sedimentation.Microbial filaments, commonly calcified, are usuallypresent, and they may have provided some initial sta-bilization of the sediment, permitting the cementationto begin.
Marine cements can be developed extensively inreefs (see papers in Schroeder & Purser, 1986). There isa wide variety of cement morphologies but the miner-
alogy is either aragonite or high-Mg calcite. Aragoniteoccurs mostly as acicular fringes (like those in Fig.4.28), and needle meshworks, but one prominent typeis the botryoid: isolated or coalesced mamelons, whichreach 100mm in diameter (see Plate 11a). They consistof fanning fibrous crystals, commonly twinned to givea pseudohexagonal cross-section. High-Mg calcite occurs as bladed cements, 20–100mm long and lessthan 10mm wide, forming isopachous fringes (seePlate 11b). They are more common than aragonite insome reefs, but rare in others. High-Mg calcite also oc-curs as a micritic cement. This coats grains and linescavities, but more commonly forms peloidal struc-tures, abundant in skeletal cavities and forming surficial crusts on corals in some instances. The peloidsare 20–60mm in diameter and are arranged in grain-stone to packstone textures. The origin of the peloidshas given rise to much discussion, principally overwhether they are inorganic precipitates, microbial pre-cipitates, faecal pellets, or detrital sediment. The con-sensus now is for a microbial origin with precipitationtaking place within and around clumps of bacteria (seeChafetz, 1986).
The precipitation of cements in reefs contributes to-wards the generation of a solid framework, but reefsare also subject to extensive bioerosion. Microbial or-ganisms, clionid sponges and lithophagid bivalves allbore into carbonate skeletons in reefs; sponges alsogenerate much fine debris. Fish and other organismsgraze on corals and generate sediment too. Internalsedimentation of this material into primary and sec-ondary cavities is widespread, and these sedimentsmay then become cemented.
The distribution of cements in many reefs appears torelate to the circulation of water; cementation is gener-ally more intense along the windward margins, whereseawater is constantly pumped through the reef. How-ever, on a smaller scale, the occurrence of cements maybe very patchy and varied, both in terms of extent,mineralogy and morphology. One cavity may containacicular aragonite, an adjacent cavity high-Mg calcitepeloids, whereas another close by may be empty (as inPlate 11a). One important control on the amount ofcementation is local permeability, which determinesthe fluid-flow rates through the reef-rock (Goldsmith& King, 1987). This also may control the mineralogyitself.
Aragonite cements of intertidal to shallow-subtidalsediments usually have a high strontium content, up to
138 Chapter 4
10000p.p.m., and Mg content of about 1000p.p.m.or less. The high-Mg calcite cements are typically be-tween 14 and 19 mol.% MgCO3, but Sr is low ataround 1000p.p.m.
Deeper-water cementation. Cemented carbonatesediments have been recovered from the ocean floorsat depths down to 3500m, mostly from areas of negli-gible sedimentation, such as seamounts, banks andplateaux. The limestones consist chiefly of planktonicforaminifera, molluscs and coccoliths cemented by amicritic calcite. The limestones are commonly boredand may be impregnated with phosphate and ferro-manganese oxides. On the deeper-water slopes aroundthe Bahama Platform (700–2000m deep), periplat-form ooze is being cemented by micritic low-Mg calcite. The sediment itself consists of planktonicforaminifera, coccoliths and shallow-water materialderived from the platform. Cemented crusts and hard-grounds pass downslope into patchily cemented sedi-ment and nodules, and then there is little cementationon the deeper-water slopes. Originally aragonitic bio-clasts have been leached and high-Mg calcite grainshave lost their Mg (Dix & Mullins, 1988). In theMediterranean and Red Sea, where there are and/orhave been warmer bottom waters, thin surface crustsand nodules of lithified pelagic sediment are cementedby micritic high-Mg calcite. Acicular and micriticaragonite cements pteropod layers on the floor of theRed Sea.
Carbonate cement precipitation in these deeper-water environments is mainly a reflection of the veryslow sedimentation rates, which allow interaction be-tween sediment and seawater. The CaCO3 is derivedfrom seawater and the dissolution of less stable grainsin the sediment; the type of cement precipitated and itsMg content are determined by the precipitation rateand the water temperature; as the latter goes down, sodoes the mol.% MgCO3.
Marine dissolution. Dissolution of carbonate grainsdoes take place on the shallow sea floor in higher lati-tudes where seawater is undersaturated with respect to the carbonate phase. The saturation state of sea-water with regard to CaCO3 decreases with lower temperature (and so broadly with increasing depth),and seawater becomes undersaturated in respect ofaragonite and high-Mg calcite before low-Mg calcite.Dissolution of bioclasts is taking place in slope sedi-
ments off the Bahama Platform at depths of a few hun-dred metres, and in the geological record relativelyshallow sea-floor dissolution of aragonite is recordedfrom hardgrounds in the Jurassic and Ordovician(Palmer et al., 1988). These were times when shallow,low-latitude seawater was supersaturated with regardto calcite, but not aragonite, too, as is the case today.Sea-floor dissolution of aragonite is common in Mesozoic deeper-water pelagic limestones; this isclearly shown where, through sea-floor loss of theiraragonitic shell, ammonites are preserved as casts encrusted by sessile organisms and Fe–Mn oxides.Carbonate dissolution increases with increasing depth until the CCD (carbonate compensation depth),below which limestones are not deposited (see Section4.10.7).
Marine diagenesis in ancient limestones
In many limestones, there is abundant evidence for diagenesis taking place on the sea-floor or just below.Marine cements are common in ancient reef-rocks,and hardgrounds are well known too in the geologicalrecord (see Section 4.6.1). However, the cementsthemselves are variable; some were similar to those ofthe Recent and were composed of aragonite and high-Mg calcite, although they are now calcite (low Mg).On the other hand, other marine cements in ancientlimestones do not appear to have exact modern equivalents and were precipitated as calcite with fabrics somewhat different from modern cements.
General features for the recognition of marine cements in limestones are:1 they are the first-generation cement;2 they usually form isopachous fringes around grainsor cavity walls;3 they usually are/were fibrous in nature;4 they may be cut by borings or include skeletal debris;5 they may be associated closely with internal sediments;6 the crystals are usually non-ferroan and non-luminescent;7 they are succeeded by clear calcite spar.
Ancient marine cements are shown in Plates 11c,d& 12a,b.
Ancient marine aragonite cements. Aragonite ce-ments are very rarely preserved as aragonite in lime-
Limestones 139
stones because the mineral is metastable. They are al-tered to calcite in a similar way to aragonite bioclasts,either through wholesale dissolution and then laterfilling of the void by calcite spar, or through calcitiza-tion, whereby calcite replaces the aragonite across athin-fluid film, with dissolution of aragonite on oneside and precipitation of calcite on the other. In thisprocess, some traces of the original cement texturemay be preserved within the replacement calcite crys-tals by the pattern of minute relics of the aragonite orinclusions of organic matter (Fig. 4.31). The replace-ment calcite crystals, a type of neomorphic spar (seeSection 4.7.4), are usually irregular to equant in shape,cross-cutting the original acicular pattern of the arag-onite cement. This calcite may have a relatively high(several thousandp.p.m.) strontium content inheritedfrom the aragonite. Some ancient calcitized aragonitecements have distinctive ‘square-ended terminations’along the cement fringes. The gross morphology of an-cient aragonite cements is similar to that of Recentcases, isopachous fringes and botryoids. Althoughsome marine isopachous cement fringes in limestones
were calcite originally, botryoids appear to have been aragonitic only; somewhat similar structures, in calcite, do form in meteoric, spelean–vadose diage-netic environments. Descriptions of ancient aragoniteare included in Tucker & Hollingworth (1986), fromthe Permian of northeast England, and Roylance(1990), from the Carboniferous of the Paradox Basin,USA.
Ancient marine calcite cements. The most commonmarine calcite cement in ancient limestones is fibrouscalcite: elongate crystals normal to substrate, and usu-ally cloudy or dusty in appearance relative to later cal-cite spar. A columnar growth form, with length-to-width ratio of more than 6 :1 and width greater than10mm, can be distinguished from a more acicular vari-ety, with a much higher length/width ratio and widthof around 10mm. Columnar fibrous calcite is abun-dant in reef cavities and in stromatactis structures ofPalaeozoic mud-mounds (see Fig. 4.26). The more aci-cular type is common in grainstones but occurs in reefstoo (see Plate 11c,d).
Fibrous calcite has a range of fabrics from unit-extinguishing crystals to undulose-extinguishing radi-axial fibrous calcite (RFC), with divergent optic axes,and fascicular-optic fibrous calcite (FOFC), with con-vergent optic axes (Fig. 4.32 and Plate 12a,b). In RFC(the most common type) the extinction swings acrossthe crystal in the same direction as rotation of the
Fig. 4.31 Calcitized aragonite botryoid. (a) Photomicrograph ofirregularly shaped neomorphic calcite cross-cutting the pattern ofthe original aragonite crystals. Crossed polars. (b) Scanningelectron micrograph of minute aragonite crystal relics in calcitereplacing an aragonite botryoid. Permian reef (Zechstein).Durham, England.
(a) (b)
140 Chapter 4
microscope stage. Twin planes are common in fibrouscalcite and in RFC they are concave away from thesubstrate. There has been much discussion over theorigin of fibrous calcite, with a replacement origin longpopular, but it is now regarded generally as a primaryprecipitate, with the peculiar fabrics attributed tosplit-crystal growth (Kendall, 1985).
One particular point of interest with fibrous calciteis whether it was originally low- or high-Mg calcite. Itis not easy to decide, as high-Mg calcite loses its Mgduring diagenesis, but several lines of evidence can beused. Fibrous calcite usually is cloudy with inclusions,and some of these may be minute crystals of dolo-mite. Others will be fluid-filled inclusions or empty.Microdolomites are taken to indicate an original high-Mg calcite mineralogy; this is supported by theirpresence in formerly high-Mg calcite bioclasts such asechinoderms. In addition, fibrous calcite after high-Mg calcite may still contain several mol.% MgCO3after the diagenetic loss of most of the Mg. This mag-nesium memory can be detected only by geochemicalmeans, such as with use of the electron probe. Fibrouscalcites originally of high-Mg calcite may pick up someFe2+, if this is in the pore waters when the high- to low-Mg calcite transformation is taking place. Althoughthe fabrics of fibrous calcite are mostly unchanged dur-ing diagenesis, some neomorphism may occur, espe-cially if the crystals were originally high-Mg calcite.Dissolution may occur along twin boundaries, for example.
In addition to fibrous calcite, there is evidence insome limestones for equant sparry calcite being a ma-rine precipitate. Although mostly of meteoric or burialorigin, it does occur as the first-generation cement insome Jurassic and Ordovician hardgrounds. Syntaxialovergrowths on echinoderm debris, also generally re-
garded as near-surface meteoric or burial precipitates,formed early in some hardgrounds too. Early, inclusion-rich zones of crinoidal overgrowths areprobably marine (see Plate 12c).
Similar textures to the micritic and peloidal high-Mg calcite precipitates in modern reefs do occur in an-cient reefs, and there are the same arguments over theirorigin. The peloidal structures are particularly com-mon in Triassic reefs and some Jurassic ones too (e.g.Sun & Wright, 1989).
Discussion of marine cements
The mineralogy of modern tropical shallow-marinecements is mostly either aragonite or high-Mg calcite,and in the geological record, marine cements origi-nally of aragonite, high-Mg calcite and low-Mg calcitecan be identified. In a similar manner to ooids (see Sec-tion 4.3.1), there appears to be a secular variation inmarine cement mineralogy through the Phanerozoic(see Fig. 4.4). Aragonite botryoids, for example, one ofthe distinctive marine cement morphologies, are com-mon in the Cenozoic and Permo-Triassic, but appar-ently absent from the mid-Palaeozoic and Jurassic–Cretaceous, where fibrous calcite is the dominant ce-ment type. As with ooids, the controls on cement min-eralogy are likely to be the seawater Mg/Ca ratio, Pco2and carbonate-supply rate (Fig. 4.33). Aragonite has asimilar stability to magnesium calcite, with around 12 mol.% MgCO3. It is unclear what controls whetheraragonite or high-Mg calcite is precipitated. The pres-ence of Mg2+ and SO4
2– in seawater does result in a ki-netic hindrance to calcite precipitation, so favouringaragonite. High carbonate supply appears to favouraragonite precipitation too, so that where fluid-flowrates are higher, as in very permeable reef-rocks andlime sands, aragonite will be precipitated in preferenceto high-Mg calcite, which will tend to occur in less per-meable sediments. In some situations, substrate con-trol is another factor, with cement crystals being thesame mineralogy as the substrate, and in optical conti-nuity (syntaxial) too. See the review of Given &Wilkinson (1985) for more information.
4.7.2 Meteoric diagenesis
Near-surface meteoric diagenesis mostly involvesfresh water and the major processes are carbonate dis-solution, cementation and the formation of soils. The
Undulose extinctionCurved twin planes
Subcrystals
Unit extinctionStraight twin planes
Fascicular-optic Radiaxial Radial–fibrous
Fig. 4.32 Common types of fibrous calcite.
Limestones 141
position of the ground-water table is important andthe vadose zone above is distinguished from thephreatic zone below. In the vadose zone, pores periodi-cally contain water, air or both, and an upper zone ofinfiltration is distinguished from a lower zone of per-colation. Rainwater is undersaturated with respect toCaCO3 so that dissolution is one of the main processesoperating in the infiltration zone. As the water movesdown through the vadose zone, it may become super-saturated with respect to CaCO3 so that precipitationoccurs. This is normally of calcite (low Mg), as mete-oric water has a very low Mg/Ca ratio. In the phreaticzone, pores are fluid-filled all the time, and the nor-mally fresh water gives way downwards to more salinewater at depths of several hundred metres or more. Incoastal regions, the phreatic meteoric ground-waterrealm passes into a mixing zone with seawater. It hasbeen suggested frequently that the mixing zone is animportant location for dolomitization (see Section4.8.2), but there is now much doubt about this. Cli-mate is a major influence on meteoric diagenesis be-cause it controls the availability of meteoric water and
also affects the temperature and degree and nature ofplant cover and soil development. Climate also deter-mines the type and extent of karstic dissolution.Palaeokarstic surfaces (see Section 4.6.1) are now wellknown in the geological record and can be recognizedfrom their morphology (an irregular pot-holed sur-face) and association with soil crusts and other fea-tures of subaerial exposure. Subsurface karst is thesystem of caves and fissures that may extend great dis-tances down from the surface.
There have been several studies of modern to Pleis-tocene limestones to document the progressive alter-ation of marine sediments with increasing exposure tometeoric water. In the early stages, low-Mg calcite isprecipitated on the surfaces of grains as an (a) iso-pachous (uniform thickness) fringe if precipitated inthe phreatic zone, below the water table where all poresare completely filled with water, or (b) asymmetricfringe, thicker on the underside of grains (a dripstoneeffect) or located at grain contacts (a meniscus effect), ifprecipitated in the vadose zone (Fig. 4.30, and Plates6a,b & 13a). Syntaxial overgrowths on echinoderm
Equant calcite cement in cold, deeper-water, low-latitude sediments andshallow-water temperate sediments
Equant high-Mg calcite cements in reefs (rare)
Acicular high-Mg calcite cements in reefs and lime sands
Acicular aragonite cements in reefs and lime sands
Equant low-Mg calcite spar cements in meteoric environments
Acicular low-Mg calcite in speleothems and travertines
Acicular aragonite in speleothems (rare)
1
2
3
4
5
6
7
1 2 3 4
5 6 7
High-Mg Calcite
Marine water
Meteoric water
Low-Mg Calcite AragoniteEq
uant
Acicular
Increasing rate of CO2– supply3
Incr
easi
ng
am
bie
nt
Mg
/Ca
Fig. 4.33 Schematic illustration ofrelationship between fluid Mg/Ca ratio, rateof carbonate-ion supply, crystal morphology(equant or acicular) and mineralogy ofinorganic precipitates (low-Mg calcite, high-Mg calcite or aragonite). The Mg/Ca ratio ofthe meteoric line is 0.3 and of the marineline 5.2. Low-Mg calcite is up to 9 mol.%MgCO3 and high-Mg calcite is above. Theposition of modern, naturally occurringprecipitates.
142 Chapter 4
grains also begin to develop in this early stage. Nextgenerally follows the loss of Mg from the high-Mg cal-cite, leaving a sediment of low-Mg calcite and arago-nite (Fig. 4.34). The next stage is aragonite dissolutionand reprecipitation of the CaCO3 as drusy calcite spar.Aragonite skeletal grains are dissolved and the voidsleft are then filled with calcite, the shape of the grainsbeing maintained by the earlier cement fringe or a mi-crite envelope (Section 4.3.3). The main feature of thisdissolution–reprecipitation process is a loss of internalstructure in the aragonitic skeletal grains. The repre-cipitation of calcite may follow soon after aragonitedissolution, or, there may be a much longer time gap, al-lowing some compaction of the skeletal voids beforecalcite precipitation, producing broken and fracturedcement fringes and micrite envelopes (seen in Plates 7c& 13d). Some aragonite grains are replaced in situ, i.e.calcitized (Section 4.7.4), so that the internal structureis preserved to some extent (see Plate 8a). The last stageof meteoric diagenesis would involve further precipita-tion of low-Mg calcite to fill all remaining voids. Inmost Pleistocene limestones, this final stage has rarelyproduced a fully cemented limestone; porosity may stillreach 20%. A further phase of cementation is thus re-quired to reduce the porosity to less than 5%, the typi-cal value of ancient limestones.
Near-surface, meteoric vadose drusy calcite spar isgenerally non-ferroan, as pore waters are usually oxi-dizing. However, if there is decomposing organic mat-ter, the water-flow rate is low and Fe2+ is available, thenphreatic calcite spar may be ferroan.
Near-surface, meteoric calcite cementation takesplace in many continental sediments, such as wadi
gravels, scree sediments and aeolian sands, where porewaters enriched in CaCO3 are evaporated. Also in themeteoric diagenetic environment, soils are formed, especially calcretes (caliches). Near-surface sedimentsare cemented and altered through pedogenesis, andsome distinctive fabrics are produced, such as needle-fibre calcite, alveolar texture, vadose pisoids and lami-nated crusts from calcification of root mats (seeSection 4.10.1).
Ancient meteoric calcite cements
In the geological record, the most obvious products ofmeteoric diagenesis are the calcareous soils (calcretes)and palaeokarsts, but the vadose cement morpholo-gies are not that uncommon (see Plate 13a) and it ispossible that some drusy calcite spar, the ubiquitouscement of ancient limestones, was precipitated in anear-surface, meteoric phreatic environment. Theoriginal mineralogy of a carbonate sediment is also afactor in the extent of meteoric diagenesis. Wherethere is a high percentage of metastable aragonite andhigh-Mg calcite, then the sediment has a much higherpotential for dissolution and cementation. Sedimentsdominated by the more stable low-Mg calcite will havea much lower diagenetic potential. This aspect of me-teoric diagenesis is explored further in Hird & Tucker(1988) and James & Bone (1989).
4.7.3 Calcite spar
The cement that occupies the majority of the originalpore space in many limestones is a clear, equant calcite,
Low-Mg calciteMineralogy of ancient limestones(unless dolomitized)
Aragonite High-Mg calcite
C
B
A
Initial cementation by low-Mg calcite
Loss of Mg from high-Mgcalcite grains
Dissolution of aragonite and precipitation of calcite (low Mg)
A
B
CFig. 4.34 Triangular diagram indicatingthe mineralogical range of modernmarine carbonate sediments (stippled)and the path of meteoric diagenesis.
Limestones 143
referred to as sparite or calcite spar (see Fig. 4.35).Sparite possesses a number of features, which, takentogether, allow its cement interpretation. These are:1 its location between grains and skeletons, and within original cavities;2 its generally clear nature, with few inclusions;3 the presence of planar intercrystalline boundaries;4 a drusy fabric, i.e. an increase in crystal size awayfrom the substrate or cavity wall;5 crystals with a preferred orientation of optic axesnormal to the substrate.
The fabric characteristics of sparite 4 and 5 are re-flections of the preferred growth direction of calcite,parallel to the c-axis. Calcite spar is invariably precipi-tated after the fibrous calcite described earlier, which ismostly a marine cement. In some cases, there is a layerof internal sediment between the two cement genera-tions. Calcite cement may also take the form of largepoikilotopic crystals, several millimetres to centime-tres across (as in Plate 6c). The large crystals are the re-sult of a low nucleation rate and slow growth, perhapsbecause pore fluids were only just saturated with re-
spect to CaCO3. Where echinoderm grains, and otherscomposed of a single calcite crystal, are present in thelimestone, then the sparite cement may precipitatesyntaxially (in optical continuity) upon the grain toproduce an overgrowth (see Plate 12c,d). Preferentialcement growth upon such single crystal grains may en-velop adjacent, small, polycrystalline grains.
Calcite-spar crystals are commonly delicately zonedas a result of subtle variations in Fe and Mn contents.The zonation can be revealed by staining with AlizarinRed S plus potassium ferricyanide, or by observing theluminescence (see Plate 13b,c and Section 4.1). Study-ing the zonation pattern in calcite spar in a limestoneformation on a regional scale allows a cement stratig-raphy to be erected from which the larger-scale hydrol-ogy of the basin can be reconstructed (e.g. papers in Sedimentary Geology, 65(3/4), 1989; Horbury &Robinson, 1993; Budd et al., 1995).
Carbon and oxygen isotopes are being used increas-ingly in the study of carbonate cements. Fibrous calcitecommonly has a marine signature in both d13C andd18O, whereas calcite spar tends to show more nega-tive d18O, reflecting precipitation at a higher tempera-ture during burial and/or precipitation from meteoricwater (see papers in Schneidermann & Harris (1985)and Horbury & Robinson (1993); and Tucker &Wright, 1990). Fluid inclusions in calcite also provideinformation on the chemistry of pore fluids and thetemperature of precipitation (see Emery & Robinson,1993; Goldstein & Reynolds, 1994).
Origin of calcite spar. There has been much discus-sion of the environment of precipitation of sparite, themain cement in all medium- to coarse-grained lime-stones and the fill of most cavities in fine-grained lime-stones. Although drusy sparite is the typical cement ofnear-surface meteoric diagenesis, in many limestonesthere is evidence for sparite precipitation after com-paction. The occurrence of broken and fractured fi-brous cement fringes and micrite envelopes aroundformerly aragonite skeletal grains indicates mechani-cal compaction resulting from overburden pressureafter aragonite dissolution and before sparite precipi-tation (see Plate 7c). In other limestones, sutured con-tacts between grains show that chemical compactionoccurred before sparite precipitation (see Plate 13a).The pore waters for deep-burial cementation will be either connate (modified seawater buried with thesediments) or meteoric (or a mixture of the two). In
(a) (b)
(c) (d)
Equant spardrusy mosaic
Compromise boundariesin drusy mosaics
Polikilotopic sparSyntaxialovergrowth
Fig. 4.35 Calcite spar. (a) Drusy calcite spar, the most commontype, characterized by an increasing crystal size away from thesubstrate towards the cavity centre. (b) Growth zones in calcitespar showing how straight crystal boundaries (compromiseboundaries) develop between adjacent crystals. (c) Syntaxialovergrowths where calcite spar cement is in optical continuitywith a host grain, schematically shown here by twin planes goingfrom cement into grain. (d) Poikilotopic calcite spar, large crystalsenveloping several grains.
144 Chapter 4
fact, in many sedimentary basins it does appear thatpore waters are largely of meteoric origin, althoughsalinity is much higher than that of near-surface freshwater. There are three possible sources for the CaCO3:(a) the pore water itself, (b) pressure dissolution with-in the limestones or at deeper levels and (c) dissolutionof CaCO3, mainly skeletal aragonite, in calcareousshales interbedded with the limestones. For precipita-tion of calcite from trapped seawater the Mg/Ca ratiowould have to be lowered; the adsorption of Mg ontoclay minerals could have this effect. Pressure dissolu-tion has been considered of major importance in ce-mentation in view of the great quantities of CaCO3that clearly have been dissolved from pressure dissolu-tion planes, and because of the vast quantities ofCaCO3 that are required from somewhere to producethe fully cemented limestone formations that are seenat the Earth’s surface. As noted in Section 4.7.1, insome limestones there is evidence for calcite spar hav-ing been precipitated on the sea floor.
4.7.4 Neomorphism
Some diagenetic processes involve changes in the mineralogy and/or fabric of the sediment. For theseprocesses of replacement, once loosely referred to asrecrystallization, the term neomorphism is now usedto include all transformations between one mineraland itself or a polymorph. There are two aspects toneomorphism: the wet polymorphic transformationof aragonite to calcite and the wet recrystallization ofcalcite to calcite. Both processes are wet because theytake place in the presence of water, through dissolu-tion–reprecipitation; dry, solid-state processes, such asthe inversion of aragonite to calcite or recrystallizationsensu stricto of calcite to calcite, are unlikely to occurin limestones, where diagenetic environments are al-ways wet. Most neomorphism in limestones is of theaggrading type, leading to a coarser mosaic of crystals.Two common types are (a) microspar–pseudospar formation from micrite and (b) the calcitization oforiginally aragonitic skeletons, ooids and cements.Degrading neomorphism results in a finer mosaic ofcrystals.
A scheme for describing textures and fabrics of neomorphic limestones and dolomites, and other sedi-ments such as evaporites that have been precipitated,crystallized or recrystallized, is given in Table 4.4.
Microspar–pseudospar; aggrading neomorphism
It is not uncommon to find that in fine-grained lime-stones, the micritic matrix (less than 4mm) has been locally or even totally replaced by microspar (crystalsizes between 4 and 10mm) and pseudospar (10–50mm) (see Plate 13e). This neomorphic spar can berecognized by:1 irregular or curved intercrystalline boundaries,commonly with embayments (contrasting with theplane intercrystalline boundaries of sparite cement,Section 4.7.3);2 very irregular crystal-size distribution and patchydevelopment of coarse mosaic;3 gradational boundaries to areas of neomorphicspar;4 presence of skeletal grains floating in coarse spar.
Aggrading neomorphic textures are prominent inmicritic limestones and they can give a mottled, almostbrecciated appearance, of ‘clasts’ of neomorphicspar/microspar in a micritic matrix. Such pseudo-breccias are common in the British Carboniferous(Solomon, 1989). Neomorphism may have takenplace within an original inhomogeneous sediment,such as that resulting from bioturbation. Aggradingneomorphism involves the growth of certain crystals
Table 4.4 Terms for describing textures and fabrics of crystalmosaics in sedimentary rocks
For crystal shape: anhedral —poor crystal shapesubhedral — intermediate crystal shapeeuhedral —good crystal shape
For equigranular xenotopic —majority of crystals anhedralmosaics: hypidiotopic —majority of crystals
subhedralidiotopic —majority of crystals euhedral
For inequigranular porphyrotopic —where larger crystals mosaics: (porphyrotopes) are enclosed in a
finer-grained matrixpoikilotopic —where larger crystals
(poikilotopes) enclose smaller crystals
Size-scale: micrometre-sized 0–10 mmdecimicrometre-sized 10–100 mmcentimicrometre-sized 100–1000mmmillimetre-sized 1–10mmcentimetre-sized 10–100mm
Limestones 145
at the expense of others. It is likely that growth takesplace in solution films and cavities between crystals, by syntaxial precipitation on pre-existing crystals. The CaCO3 will be derived from dissolution of submicrometre-sized crystals and inflowing pore wa-ters. Studies of fine-grained limestones by SEM haveshown that much of the microspar is actually a cementrather than the result of aggrading neomorphism.Lasemi & Sandberg (1984) have found relics of arago-nite in some ancient micrites and so were able to distin-guish aragonite-dominated precursor muds (ADP)from calcite-dominated ones (CDP).
Calcitization of aragonite grains and cements
Bioclasts and ooids composed originally of aragoniteare now mostly drusy sparite in limestones, throughdissolution of the aragonite and later precipitation of calcite into the void (Section 4.7.1). Where calcitespar has not been precipitated, then biomoulds andoomoulds are present. In some cases, however, thegrains have been replaced by calcite with no interven-ing void phase, a process referred to as calcitization.Where this has occurred, features to note are:1 relics of the internal structure of the shell, preservedthrough inclusions of organic matter and minute crys-tals of aragonite;2 an irregular mosaic of small and large calcite crys-tals, with wavy, curved or straight intercrystallineboundaries;3 a brownish colour to the neomorphic spar, owing toresidual organic matter, which imparts a pseudo-pleochroism to the crystals (see Plate 8a).
Aragonite cements also endure either dissolution orcalcitization, but more tend to be calcitized (see Fig.4.31).
Degrading neomorphism
Degrading neomorphism, whereby large crystals ofCaCO3 are replaced by smaller calcite crystals, aprocess of ‘crystal diminution’, is rare in limestones,and mostly has occurred through tectonic stress orvery low-grade metamorphism. The process is mosteasily observed to have taken place in echinodermgrains. Micritization of skeletal grains by endolithicmicrobes is not a neomorphic process of course, butdoes result in a fine-grained mosaic.
4.7.5 Compaction
As discussed in Section 2.9.1, increasing overburdenpressure leads to compaction in sediments and twocategories are recognized: mechanical and chemical.Mechanical compaction may begin soon after deposi-tion, whereas chemical compaction normally requiresmore than several hundred metres of burial.
Mechanical compaction in grainy sediments leads toa closer packing of the grains and a rotation of elon-gate bioclasts towards the plane of the bedding. As theoverburden pressure increases, fracture of bioclastsmay take place and micritic grains may becomesquashed and deformed. If there are early cementsaround grains, these can be spalled off, as can ooliticcoatings of grains. Mechanical compaction may leadto collapse of micrite envelopes, if there has been anearlier phase of dissolution of the aragonitic bioclasts(see Plate 7c).
Lime muds generally suffer more compaction dur-ing very shallow burial as the sediments dewater. Thismay lead to burrows being compressed, especially ifthere is a high clay content in the sediment. Alternat-ing clay-poor and clay-rich, fine-grained limestones, acommon deep-water facies, usually show the differen-tial effects of compaction well, with shell fracture andburrow flattening in the latter, but not the former.Compaction can lead to the formation of skeletalpackstones from skeletal wackestones, as a result ofthe closer packing of grains.
Chemical compaction is the result of increased solu-bility at grain contacts and along sediment interfacesunder an applied stress. Mostly this is the result ofoverburden but tectonic stresses also give rise to pressure-dissolution effects. Three common texturesresult from chemical compaction: fitted fabrics, stylo-lites and pressure-dissolution seams. In a grainstonewith little or no early cement, sutured and concavo-convex contacts develop between grains (see Plate13d). If the pressure dissolution between grains is in-tense, then a fitted fabric may be produced. This can beon a microscopic scale, as in oolitic and echinodermgrainstones, or on a macroscale, between intraclasts,fossils and early lithified diagenetic nodules and bur-row fills in a muddy, compactible sediment. A stylo-breccia texture may be produced.
Stylolites are through-going sutured surfaces thatcut grains, cement and matrix indiscriminately (see
146 Chapter 4
Plate 14a; see also Fig. 9.10). Clay, iron minerals andorganic matter, the insoluble residue from the lime-stone’s dissolution, are usually concentrated along thestylolites. In more argillaceous limestones, dissolutionseams are smooth, undulose and anastomosing hori-zons of insoluble residue. They generally pass aroundgrains and early diagenetic nodules, and where abun-dant, the term flaser limestone is used.
Pressure dissolution is an important process accen-tuating bedding planes, particularly between moremuddy and more grainy sediments, and it may alsolead to the development of pseudobedding planes (seeSection 4.6.1 and Simpson, 1985). Calculations haveshown that considerable amounts of CaCO3 can beliberated by pressure dissolution, so that this process isoften cited as one of the main sources of CaCO3 forlimestone cementation, especially of late-diagenetic,burial calcite spar. Burial diagenesis is reviewed inChoquette & James (1987).
4.8 Dolomitization, dedolomitization andsilicification
It is not uncommon to find that ancient limestoneshave been partially or even completely dolomitized.The dolomite so formed can be replaced by calcite inthe process of dedolomitization (Section 4.8.3). Lime-stones also can be silicified to various degrees (Section4.8.4). These diagenetic processes commonly result in an obliteration of sedimentary and petrographic details.
4.8.1 Dolomites
The mineral dolomite is a rhombohedral carbonate be-longing to the 3̄ trigonal/hexagonal crystal system. Ideally, it consists of an equal number of Ca2+ and Mg2+ ions arranged into separate sheets with planes ofCO3
2– anions between. The well-ordered nature of thedolomite lattice results in a series of superstructure re-flections on X-ray diffraction (XRD), which are notpresent in the structurally similar calcite. The height ofthe ordering d015 peak relative to the d110 peak gives ameasure of the degree of ordering. Most moderndolomites have a low degree of ordering comparedwith older dolomites. The term protodolomite was introduced for Ca–Mg carbonates made in the labo-ratory with no, or very weak, ordering reflections.Many natural dolomites are also not stoichiometric
(Ca :Mg is 50 :50), but have an excess of Ca2+ ions, upto Ca :Mg of 58 :42. The Ca2+ substitution for Mg2+ in-creases the lattice spacing, and this also can be meas-ured with XRD by the shift in the position of the d104peak (see Hardy & Tucker (1988) for details of thetechnique). Iron substitution is common in dolomites,giving ferroan dolomite with a few mol.% FeCO3;ankerite CaMg0.5Fe0.5(CO3)2 may be associated.
The replacement of CaCO3 minerals by dolomiteand the precipitation of dolomite cement may takeplace soon after the sediments have been deposited, i.e.penecontemporaneously and during early diagenesis,or a long time after deposition, usually after cementa-tion, during burial. The term primary has often beenapplied to dolomite, implying a direct precipitate fromsea or lake water. In fact the majority of dolomites haveformed by replacement of pre-existing carbonate min-erals, although dolomite cements are common. Theword dolomite is used for both the mineral and therock type; the term dolostone has been used for the latter. Carbonate rocks are divided on the basis ofdolomite content into:
limestone 0–10% dolomitedolomitic limestone 10–50% dolomitecalcitic dolomite 50–90% dolomitedolomite (dolostone) 90–100% dolomite
To convey an indication of grain or crystal size in a dolomite, the terms dolorudite, dolarenite,dolosparite and dolomicrite can be used. In manycases, if the original structure has not been destroyedcompletely, the dolomites can be described in terms ofDunham’s (or Folk’s) classification (see Section 4.4),preceded by the word dolomitic, or prefixed by dolo-.For the description of textures and fabrics the schemeof Table 4.4 can be used.
Two common types of dolomite mosaic are xeno-topic and idiotopic, the former of anhedral crystalswith curved to serrated and irregular crystal bound-aries, and the latter of euhedral rhombic crystals (seeFig. 4.36 and Plate 14a). The features to note when describing dolomite textures are shown in Table 4.5.
The preservation of the original limestone texture ina dolomite varies from completely fabric destructivewith no obvious relics of the original sediment (seePlate 14a,c), to fabric retentive, with good to perfectpreservation of original structure (see Fig. 4.3 andPlate 14b for examples). Dolomitization also may befabric selective, only destructively replacing the matrix
Limestones 147
and not the grains, or only replacing certain bioclasts.Dolomite rhombs may be distributed randomlythrough the limestone (as in Plate 13f). Original crys-tal/grain size and mineralogy are important controls.An original micritic sediment generally is dolomitizedto a fine-grained mosaic, so that primary sedimentarystructures are still preserved. Grains of high-Mg cal-cite, such as red algae, some foraminifers and echino-derms, can be dolomitized with little fabric alteration(mimic replacement). By way of contrast, aragoniticgrains (e.g. molluscs) are either dolomitized withmuch fabric alteration, or the aragonite is dissolvedout and the mould filled by dolomite cement (in a simi-lar manner to aragonite altering to calcite). Low-Mgcalcite grains may resist dolomitization or be dolomi-tized destructively. The timing of dolomitization isalso a factor, because if it is late, during burial, then it isvery likely that the original, mixed-mineralogy sedi-ment has already stabilized to low-Mg calcite, so thatthe dolomite is fabric destructive.
Dolomite crystals are commonly zoned; in manycases the inner part is more cloudy (from fluid inclu-sions or calcite relics) and the outer part is clear.Dolomite cements, as opposed to replacements, occurin primary and secondary cavities in many limestonesand dolomites. They vary from cavity linings of clear,‘limpid’ rhombs to drusy mosaics similar to calcitespar (as in Fig. 4.5).
One type of dolomite, which may be a replacementor a cement, is baroque or saddle dolomite (see Plate14c). It is also known as pearlspar. The crystals gener-ally are large (many millimetres) and have conspicu-ous, curved crystal faces. In thin-section, they havecurved cleavage and undulose extinction. They com-monly contain inclusions (fluids or mineral relics) andmany are ferroan. Baroque dolomite commonly is as-sociated with sulphide mineralization, hydrothermalactivity and also hydrocarbons. It often is consideredtypical of burial dolomitization and the characteristiclattice distortion is attributed to variations in the con-centration of Ca2+ ions adsorbed onto the growingcrystal surface (see review of Spötl & Pitman, 1998).
The distribution of dolomites in the stratigraphicrecord is not even, and it has been said frequently thatdolomites increase in abundance back in time.Dolomites do appear to be more common than lime-stones in the Precambrian, and this has led to the sug-gestion that seawater had a different compositionthen, so that dolomite could be precipitated directly, or could replace CaCO3 more easily. Alternative views are that dolomitization environments weremore prevalent as a result of palaeogeographical andpalaeoclimatic differences, or that simply throughbeing older, the limestones have had more time inwhich to become dolomitized. A survey by Given &Wilkinson (1987) suggested two maxima of dolomiteoccurrence in the Phanerozoic (Jurassic–Cretaceousand mid-Palaeozoic), which broadly correspond to thehighstands in the first-order global sea-level curve, indicating a role of geotectonics and hydrosphere–atmosphere chemistry in dolomitization. Sun (1994)suggested that the abundance of Phanerozoicdolomites coincided with periods of extensive periti-dal deposition and large-scale evaporite basins. Burnset al. (2000) thought that periods of more extensivedolomitization correlated with decreased oxygen lev-els in the atmosphere and oceans, and that this fosteredmore active communities of anaerobic microbes,which promoted dolomite formation.
(a) (b)
Xenotopic Idiotopic
Fig. 4.36 Dolomite textures. (a) Xenotopic. (b) Idiotopic.
Table 4.5 Terms for describing dolomite textures
Dolomite crystal size unimodal or polymodalDolomite crystal shape anhedral, subhedral, euhedralDolomite mosaic xenotopic, hypidiotopic, idiotopicCrystal type limpid, rhombic, baroqueDolomite cement isopachous, drusyCaCO3 grains unreplaced or replaced or moulds; if
replaced: partial or complete, mimic or non-mimic
CaCO3 matrix Unreplaced or replacedVoid-filling dolomite limpid, rhombic, drusy or baroque
148 Chapter 4
4.8.2 The origin of dolomites and dolomitizationmodels
There is still much debate and argument over the ori-gin of dolomite, particularly concerning the pervasivedolomitization of extensive limestone platforms. Oneproblem with dolomite is that it is difficult to manufac-ture in the laboratory under sedimentary–diageneticconditions using natural waters. It is thus difficult to unravel the chemical controls on dolomite pre-cipitation. Seawater is supersaturated with respect to dolomite, but with its highly ordered structure,dolomite appears to be inhibited from direct precipita-tion by various kinetic factors (the high ionic strengthof sea-water, Mg2+ hydration and low CO3
2– activity)so that aragonite and high-Mg calcite, with their sim-pler structure, are precipitated instead. The majorconsiderations in the dolomitization of a limestone arethe source of the Mg2+ and the mechanism by whichthe dolomitization fluids are moved through the rocks (see reviews of Tucker & Wright, 1990; Purser et al.,1994). Models for dolomitization are shown in Figs4.37 & 4.38.
At the present time, dolomites are being precipitatedin intertidal–supratidal sediments of the Bahamas,Florida and Trucial Coast (Arabian Gulf). Thedolomite itself, poorly ordered and calcium-rich, con-sists mostly of 1–5mm rhombs, occurring within thesediments or forming hard surface crusts. There ismuch evidence in the form of dolomitized gastropodsand pellets for replacement, but direct precipitationalso takes place (Lasemi et al., 1989). In some of these modern peritidal occurrences, it appears that the dolomite is being precipitated from seawater,
(a) Evaporation
Stormrecharge
Vadose
Phreatic
Vadose
Phreatic
Sabkha/evaporation
Evaporation
(b)
Lagoon
Seepage–reflux Reflux
(d)
Meteoric
MarineMixing-zoneunconfined coastal aquifer
(c)
sl 1sl 2sl 3
Falling sea-level/evaporative drawdown
Freshwaterrecharge
(e)
Impermeablestrata
Mixing-zone deep confined aquifer
(f)
Compaction
Burial compaction
Fig. 4.37 Models of dolomitization, illustrating the variety ofmechanisms for moving dolomitizing fluids through thesediments. For seawater dolomitizing models see Fig. 4.38.
Seawater dolomitization
Tidalpumping
Reflux
Oceancurrentpumping
Deep salineflow
Mix
Kohoutconvection
Dolomite
Geothermal heat
Fig. 4.38 Models for seawater dolomitization of limestones, allbasically different ways of pumping seawater through acarbonate platform.
Limestones 149
evaporated to a greater or lesser extent. The Mg/Caratio of the pore fluids is raised by the precipitation ofaragonite and gypsum–anhydrite, and this promotesdolomite precipitation. Reduced SO4
2– also helps. TheCoorong in South Australia is a series of coastal lakeswhere dolomite is being precipitated, but not in asso-ciation with evaporites. The dolomite forms whereMg2+-rich continental ground water discharges intoshallow lakes and is evaporated. Dolomite does pre-cipitate directly from saline lake waters; this occurs inVictoria, Australia (Deckker & Last, 1989), whereMg/Ca ratios are high from weathering of basaltsnearby. In the shallow-marine environment, dolomiteis forming in Baffin Bay, a shallow, periodically hyper-saline lagoon with restricted connection to the Gulf ofMexico and in a lagoon near Kuwait.
In the geological record, many fine-graineddolomites with evidence of peritidal deposition (planar stromatolites, birdseyes, polygonal cracks, tepees and intraclasts) have probably formed through evaporitic dolomitization and direct dolomite pre-cipitation. Such dolomites generally show good preservation of the sedimentary structures, as in Plate 10e.
A model for dolomitizing shallow subtidal and reeflimestones is seepage–reflux, whereby high Mg/Caratio fluids are generated in lagoons and beneath tidalflats and sabkhas (Section 5.2.1) by evaporation andthese descend into the subsurface through density con-trasts with marine pore waters. Unfortunately, thereare no large-scale modern examples, but the model isapplied frequently to dolomite formations closely as-sociated with evaporites, such as those in the UpperPermian (Zechstein) basin of northwest Europe.Dolomitization of platforms and reefs around evapor-ite basins could well occur during drawdown (Fig.4.37c, see Kendall, 1989).
Another model that has been very popular, especial-ly for relatively early, near-surface dolomitization of limestone where there are no associated evaporites,is the meteoric–marine, mixing-zone model. As notedabove, seawater is supersaturated with respect todolomite but because of kinetic factors, dolomite nor-mally does not precipitate unless the Mg/Ca ratio israised (as in the evaporative model), and dolomite isforced out of solution. Therefore, it has been argued,dolomite precipitation is more likely to take placefrom dilute solutions (where there would be fewer interfering ions present) and at slow crystallization
rates. Calculations have suggested that the mixing ofmeteoric ground waters with up to 30% sea waterwould cause undersaturation with respect to calcitebut increasing saturation for dolomite. Dolomitiza-tion has thus been predicted for seawater–meteoric-water mixing zones, where salinities are reduced butMg/Ca ratios are maintained. In the mixing model, theMg2+ is derived from seawater, the Mg/Ca ratio ismaintained above unity and ground-water movementpumps the dolomitizing fluid through the limestones.Although a most attractive hypothesis and applied tomany pervasively dolomitized limestones without associated evaporites in the 1970s and 1980s, the mixing-zone model has come in for much criticism andis now abandoned. In essence, the very slow rate ofdolomite precipitation (because of the ordered natureof the crystals) relative to calcite dissolution wouldmean that dolomite could not precipitate in significantquantities in a mixing zone. See Machel & Mountjoy(1986) and Hardie (1987) and reviews in Tucker &Wright (1990) and Purser et al. (1994).
Much dolomite is precipitated during the burial of alimestone formation, but there has been much debateas to whether this could result in the pervasive dolomi-tization of whole carbonate platforms. The drivingforce in this model is considered to be compaction of basinal mudrocks and the expulsion of Mg2+-richfluids into adjacent platform-margin limestones. TheMg2+ is thought to be derived from clay minerals, marine pore waters and high-Mg calcite, but mass-balance calculations have suggested that there is insufficient Mg2+ for large-scale dolomitization. Thehigher temperature of the burial environment shouldmean that some of the kinetic obstacles to dolomiteprecipitation are overcome. However, another prob-lem appears to be that there is not a continuous sourceof fluid in the burial environment. Hydrothermal fluids may circulate, especially through fractures andfaults and lead to local veins of baroque dolomite.Scattered rhombs, dolomite crystals along stylolitesand pressure-dissolution seams, and late cavity-fillingdolomite cements are common forms of burialdolomite in many limestones. Burial dolomites gener-ally are coarse and fabric destructive; baroquedolomite is an especially common form. Burialdolomites have been well described from the Cam-brian Bonneterre of Missouri, the Upper Devonian of Alberta (e.g. Dix, 1993; Drivet & Mountjoy, 1997),the Carboniferous of England and the Triassic of the
150 Chapter 4
Dolomites (see papers in Purser et al., 1994; Kupecz et al., 1997; and Yoo et al., 2000).
In recent years, attention has focused on the pos-sibility that sea water alone, perhaps with a little modi-fication, is the dolomitizing fluid, and a number ofmechanisms have been put forward to drive seawaterthrough the sediment (see Fig. 4.38). Discoveries ofmodern dolomite in the marine environment are sig-nificant here. Beneath Enewetak atoll in the Pacific,Eocene strata at a depth of 1250–1400m are beingdolomitized. As a result of the cooler water at thisdepth, seawater is just undersaturated with respect to calcite, but still supersaturated with regard todolomite. Seawater is pumped through the atoll byoceanic tides and upwards by thermal convection as aresult of the high heat flow out of underlying volcanicbasement. Within the Bahama Platform, there also is astrong circulation of sea water as a result of ocean cur-rents such as the Gulf Stream impinging on the Ba-hama escarpment. This ocean current/tidal pumpingcould play a major role in dolomitization because itdrives huge volumes of seawater through carbonateplatforms. The high geothermal gradient beneath car-bonate platforms is thought to generate a large-scaleconvection system, which draws cold seawater intothe platform. Diagenetic changes, including dolomiteprecipitation, in periplatform ooze on the Bahamaslopes have been attributed to this Kohout convection.Lagoonal or platform-interior seawater is frequently alittle more saline (40–45‰) than open-ocean water(35‰) and this is sufficient to cause reflux into the underlying sediments. Restudy of tidal-flat dolomitesin the Florida Keys has suggested that they are formingthrough tidal pumping of Florida Bay water throughthe sediments, rather than through simple evaporationof seawater. Experiments have suggested that SO4
2– inseawater is an important kinetic inhibitor of dolomiti-zation. The SO4
2– is reduced by microbial activity andthis takes place in sediments containing organic mat-ter. Dolomite being precipitated from marine porewaters within anoxic pelagic sediments in the Gulf ofCalifornia and off Peru, and in lagoonal muds ofBrazil, could well be the result of microbial mediationand sulphate reduction. Indeed, there is a growingbody of evidence that microbes play a major role indolomite precipitation generally; see review of Burnset al. (2000), and Teal et al. (2000). One other mecha-nism that will move seawater through sediments is relative sea-level change. There also is much fluid
movement within the meteoric–marine mixing zone,which will generate pore-water circulation within the adjacent marine phreatic zone. Thus, now, sea-water dolomitization is very popular; it provides thenecessary Mg ions and there are several powerfulprocesses for circulating seawater through a carbonate platform.
The origin of dolomites remains an enigma but care-ful field, petrographic and geochemical study does per-mit an elimination of some dolomitization models andimplication of others. An analysis of the stable isotopeand trace-element geochemistry of dolomites, as wellas fluid inclusion data, can help to distinguish differentgenerations and to make inferences on the nature ofthe pore waters involved: hypersaline/marine/mixed/connate. 87/86Sr isotope data can be useful too, to iden-tify the origin of the fluids, and even date the time ofdolomitization.
An important consequence of many styles ofdolomitization is that porosity is increased (see Plate14a,b). Dolomite has a more compact crystal structurethan calcite so that theoretically the complete dolomi-tization of a limestone results in a porosity increase of13%, as long as there is no subsequent compaction orcementation. In addition, dissolution of relict calcite indolomitized limestone or increasing calcite dissolutionrelative to dolomite precipitation during the dolomiti-zation process creates extra porosity. Thus, dolo-mitization is important for hydrocarbon reservoirpotential. Many oilfields of western Canada, for ex-ample, are in dolomitized Devonian reef limestones(see papers in Roehl & Choquette (1985) for case his-tories). Many papers on dolomite are contained inShukla & Baker (1988) and Purser et al. (1994).
4.8.3 Dedolomitization
Dolomite may be replaced by calcite to produce lime-stone again. This calcitization process is referred to asdedolomitization and predominantly takes placethrough contact with meteoric waters. Calcite re-placement of dolomite commonly is associated withthe dissolution of gypsum–anhydrite, a near-surfacephenomenon as well (Section 5.5). Burial dedolomiti-zation also may occur. Recognition of ‘dedolomites’ issimilar to that of replaced evaporites, a question ofnoting dolomite crystal shapes (rhombohedra) occu-pied by calcite (pseudomorphs), or calcite crystalswith replacement fabrics (see neomorphic spar,
Limestones 151
Section 4.7.4) containing small relict inclusions ofdolomite (see Plate 14d). In some cases the originallimestone texture is partially regenerated on de-dolomitization; in other instances layers and concre-tions of fibrous calcite randomly and completelyreplace the dolomite. The dissolution of dolomiterhombs may lead to a mouldic porosity (see Plate 14e).
4.8.4 Silicification
Silicification, like dolomitization, can take place during early or late diagenesis. It takes the form of selective replacement of fossils or the development of chert nodules and layers (Section 9.4). Silica also occurs as a cement in some limestones. The main typesof diagenetic silica in limestones are: (a) euhedralquartz crystals, (b) microquartz, (c) megaquartz and(d) chalcedonic quartz. They are described in Section9.2 and shown in Figs 9.1, 9.2 & 9.11. Both length-fastand length-slow chalcedonic quartz occur and the lat-ter may indicate the former presence of evaporites(Section 5.5). Sponge spicules are the main source ofsilica, together with diatoms and radiolarians (Section9.4).
4.9 Porosity in carbonate sediments
The porosity of carbonate sediments shortly after deposition is very high: sand-sized sediments around50%, lime mud around 80%. Porosity is lost or re-duced through cementation, compaction and pressuredissolution, and gained through dissolution, dolomiti-zation and tectonic fracturing.
Porosity in limestones can be divided into two main types: primary (depositional) and secondary (diagenetic–tectonic). Three common types of pri-mary porosity are:1 framework porosity, formed by rigid carbonateskeletons such as corals, stromatoporoids and algae,especially in reef environments;2 interparticle porosity in carbonate sands, depend-ent on grain-size distribution and shape;3 porosity in carbonate muds provided by fenestrae(birdseyes) and stromatactis (Section 4.6.3).
Secondary porosity includes:1 moulds, vugs and caverns formed by dissolution ofgrains and rock, commonly through leaching by mete-oric ground waters, but also by basinal (connate) wa-ters;
2 intercrystalline porosity produced through dolomitization;3 fracture porosity, formed through tectonic pres-sures, and through collapse and brecciation of lime-stone as a result of dissolution, such as of interbeddedevaporites, or the limestone itself in karstification.
Primary porosity, and also secondary, is commonlyfacies controlled. Certain facies, such as reefs, fore-reefs and oolites have high primary porosities, where-as others have low porosities, lagoonal micrites andouter-ramp carbonates, for example, unless affectedby the diagenetic–tectonic processes leading to poro-sity creation. Studies of carbonate facies distributions,cementation patterns and diagenesis, in particulardolomitization, coupled with porosity–permeabilitymeasurements are thus all required to detect any reser-voir potential. Examples of carbonate hydrocarbonreservoirs are: the Upper Jurassic Arab Formation of Saudi Arabia, with a primary intergranular anddolomite porosity; Middle and Upper Devonian reefand fore-reef limestones of western Canada and theOrdovician Trenton Limestone of northeastern USA,both with a porosity, at least in part, as a result ofdolomitization; the Ellenburger of Oklahoma andTexas, with karstic porosity, and the Upper Creta-ceous Chalk of the North Sea and the Tertiary AsmariLimestone of Iran, both with fracture porosity. Car-bonate reservoirs are described in Roehl & Choquette(1985), Kupecz et al. (1997) and Harris et al. (1999),and porosity is reviewed by Moore (1989) and Lucia(1995, 1999).
4.10 Carbonate depositional environmentsand facies
4.10.1 Non-marine carbonate sediments
Lacustrine limestones
Lacustrine carbonates are of three principal types: (a) inorganic precipitates, (b) algal/microbial sedi-ments and (c) skeletal sands. Inorganic precipitation,producing lime muds, mostly takes place throughevaporation, but CO2 loss, as a result of plant photo-synthesis or pressure–temperature changes, and mixing of fresh stream or spring water with saline lake water, also causes carbonate precipitation. Tufamounds are spectacularly developed in some lakes(e.g. Mono Lake, California) as a result of sublacus-
152 Chapter 4
trine springs. Precipitation in shallow, agitated zonesmay produce ooids, as in the Great Salt Lake, Utah andPyramid Lake, Nevada. The mineralogy of the carbon-ate mud and ooids precipitated depends largely on theMg/Ca ratio of the water. Aragonite, calcite (high andlow Mg) and dolomite (as mud not ooids) may all beprecipitated.
Lime muds also may be produced through the activ-ities of algae, cyanobacteria and microbes, and fromphytoplankton blooms. The main role of cyanobacte-ria, however, is in the formation of stromatolites, com-mon in modern lakes (e.g. Great Salt Lake) and inmany ancient lake formations (e.g. the Green RiverFormation, Wyoming and Utah and the PlioceneRidge Basin of California). Oncoids also occur, thosefrom Lake Constance, Switzerland, being especiallywell known. Skeletal sands contain fragments of cal-careous algae, such as Chara, as well as bivalves andgastropods.
Lacustrine carbonates are arranged in a similar fa-cies pattern to their marine counterparts. Stromatolite‘reefs’ and ooid shoals occur in more agitated, shallowwaters, with lime muds occurring shoreward on lit-toral flats and in protected bays, and in the centraldeeper parts of lakes. One characteristic feature oflake-basin deposits is a rhythmic lamination, consist-ing of carbonate–organic matter couplets, often inter-preted as seasonal in origin.
Two broad types of lake can be distinguished:1 hydrologically open —these have an outlet and soare relatively stable;2 hydrologically closed —these have no outflow andso are subject to rapid changes in lake level throughfluctuations in rainfall and outflow.
Evaporation may exceed inflow so that a saline lakedevelops, where evaporites may be precipitated (seeSection 5.4.2). Perennial and ephemeral closed lakesare recognized. In the geological record, both open-and closed-lake carbonates occur in the Green RiverFormation, western USA, and perennial–ephemeralsaline-lake carbonates, mostly dolomites, occur in theCambrian Officer Basin, South Australia (Southgate etal., 1989). Stromatolites, laminites, fish beds anddolomites were formed in the thermally stratified Devonian Orcadian lake of northeast Scotland(Trewin & Davidson, 1999).
Shallow-water, muddy lacustrine facies are com-monly modified by pedogenesis and the term palus-trine is used for these deposits, which usually are
nodular and mottled. They are widely developed in theTertiary of the Mediterranean area, Cretaceous ofSpain (Platt, 1989), and the Devonian Catskills ofNew York (Dunagan & Driese, 1999). The Evergladesof Florida are a good modern analogue (Platt &Wright, 1992).
Reviews and descriptions of lake carbonates can befound in Matter & Tucker (1978), Dean & Fouch(1983), Anadon et al. (1991) and Talbot & Allen(1996).
Calcrete or caliche
In many parts of the world where rainfall is between200 and 600mmyear–1, and evaporation exceeds thisprecipitation, calcareous soils are formed. They aretypically seen in river floodplain sediments but theyalso develop in other continental sediments (aeolian,lacustrine and colluvial deposits), and in marine sedi-ments too, should they become subaerially exposed.Many terms are applied to these pedogenic carbonatesbut calcrete and caliche, the latter chiefly in the USA,are widely used. Calcrete occurs in several forms, fromnodules to continuous layers, with massive, laminatedand pisolitic textures. Many calcretes form in theupper vadose zone through a per descensum process ofdissolution of carbonate particles in the upper A horizon of the soil profile and reprecipitation in thelower B horizon (see Fig. 3.7). They also may formwithin or just below the capillary-rise zone, where theyare referred to as phreatic or ground-water calcretes,well known from southwest Australia. Some of theseare composed of dolomite (dolocrete; e.g. Colson &Cojan, 1996). Calcretes develop in time from scatteredto packed nodules (Fig. 4.39), to a massive limestonelayer. The amount of time involved varies but is of theorder of several to tens of thousands of years.
The characteristic fabric of calcretes is a fine-grained equigranular calcite mosaic with floatingquartz grains through displacive growth (Fig. 4.40).Grains and pebbles may be split in this displaciveprocess. Replacement of some grains also takes place.Circumgranular cracks and spar-filled veins also arecommon. Many calcretes possess spar-filled tubules,formerly occupied by rootlets, and lacey, micritic‘septa’ within these, formed by calcification of fungalfilaments, give the so-called alveolar texture. Rhi-zocretions (rootlet encrustations) and microbial graincoatings are common in some calcretes. Pisoids
Limestones 153
(‘vadoids’) form in some calcareous soils from calcifi-cation of fungal–bacterial clusters. Some of these showevidence for in situ growth: dominant downward-directed laminae, reverse-grading and a fitted, almostpolygonal arrangement of pisoids. Others accrete dur-ing imperceptible downslope movement of the soil.Laminated crusts are associated with calcretes andpisoids and some form from calcification of root mats.Black pebbles may be present (Section 4.3.1). Reviewson calcretes have been presented by Wright & Tucker(1991) and Retallack (1997). Specific descriptions in-clude Williams & Krause (1998) for Devonian cal-
cretes in floodplain mudrocks, and Beier (1987) formodern Bahamian calcretes.
4.10.2 Marine carbonates and carbonate platforms
Thick successions of mostly shallow-marine limestoneare a feature of the geological record. Carbonate plat-form is a general term widely used for these and theytypically develop along passive continental margins, inintracratonic basins, failed rifts and foreland basins.Five major categories of carbonate platform are recog-nized: the rimmed shelf, ramp, epeiric, isolated anddrowned platforms (Fig. 4.41). Each platform type hasa particular pattern of facies and facies succession. Therimmed shelf (Fig. 4.42) is a shallow-water platformwith a distinct break-of-slope into deeper water. Reefsand carbonate sand bodies occur along the high-energy shelf margin, restricting the circulation in theshelf lagoon behind to a greater or lesser extent. Alongthe shoreline, depending on the energy level and tidalrange, tidal flats or a beach-barrier complex will bepresent. Debris from the rimmed-shelf margin is shedonto the adjacent slope and into the basin. Modernrimmed shelves occur off South Florida, Belize andQueensland (Great Barrier Reef). The carbonate ramp(Fig. 4.43) is a gently sloping surface with a generallyhigh-energy, inner-ramp shoreline passing offshore toa quiet, deeper-water outer ramp, affected periodicallyby storms. Along the shoreline there may be a beach-barrier–tidal-delta complex, with lagoons and tidal
Fig. 4.39 Calcrete, consisting of closely packed elongate nodulesthat have grown in a river floodplain sediment. Old RedSandstone, Devonian. Gloucestershire, England.
Fig. 4.40 Photomicrograph of modern calcrete showing densemicritic calcite, which has displaced the quartz grains so that thelatter are now not in contact. There is also an envelope of coarsercalcite crystals around each quartz grain; this is a typical feature ofcalcretes. Plane-polarized light. Recent calcrete. Almeria, Spain.
Rimmed shelfwidth 10–100km Ramp
Width 10–100km
Epeiric platform Width 102–104 km
Isolated platform Width 1–100km
Drowned platform
Fig. 4.41 Different types of carbonate platform.
154 Chapter 4
flats behind, or a beach-ridge/strandplain system.Large reefs generally are not present on ramps, butpatch reefs may occur on the inner ramp, and pinnaclereefs and mud-mounds in deeper water on the outerramp. In effect, the carbonate ramp is equivalent to asiliciclastic open shelf. Modern carbonate ramps arelocated off the Yucatan Coast of Mexico, the TrucialCoast of the Arabian Gulf and Shark Bay of WesternAustralia. The epeiric platform is a very extensive(100–10000km across), relatively flat cratonic areacovered by shallow sea. Along the margin, there maybe a gentle (ramp-like) or steep (shelf-like) slope intothe adjoining basin, but the margin is not a major fea-ture of the platform. Within the platform itself, theremay be deeper-water basins, surrounded by ramps andrimmed shelves. Epeiric platforms are dominated bylow-energy, shallow subtidal–intertidal sediments.
Storms and tidal currents may be important. There areno good modern examples of epeiric platforms of thesize that existed in the past, but the interior of theGreat Bahama Bank and Florida Bay may be close analogues.
Isolated or detached platforms are surrounded bydeep water and so are very much affected by prevailingwind and storm directions. They vary in size from afew kilometres to a few hundred kilometres across.The Bahamas is a large, modern example; smaller onesoccur off the Belize shelf and in the Red Sea. Detachedplatforms consist of pure carbonate sediment becauseterrigenous material, apart from wind-blown dust, isexcluded. The margins of these platforms may be shelfor ramp in character. Drowned platforms are ones thathave suffered a relatively rapid sea-level rise so thatdeeper-water facies are deposited over shallow-water
Basin Slope Carbonate rimmed shelf
Below fair weatherwave base
Maximumwave action Protected Subaerial
Shales/pelagic
limestones
Re-sedimentedcarbonates
Reefs andcarbonate
sand bodies
Lagoonal and tidal flatcarbonates
Supratidalcarbonates
Mudstones Grain/rud/floatwackestones
Boundstones/grainstones
Wackestones–mudstones Mudstones
Fig. 4.42 General facies model andlimestone types for a rimmed shelf.
BasinCarbonate ramp
Deep ramp Shallow ramp Back ramp
Below fair weather wave base Wave-dominated Protected/subaerial
Shale/pelagic
limestone
Thin beddedlimestones
Storm deposits mud mounds
Beach–barrier/strandplain/sand shoalspatch reefs
Lagoonal–tidal flat–supratidal carbonates,evaporites palaeosols,
palaeokarsts
Mudstones Grain/wacke/mudstones
Grainstones Wackestones–mudstones
sl
FWWB
SWB
Fig. 4.43 General facies model andlimestone types for a carbonate ramp.
Limestones 155
facies. Many pelagic limestones were deposited inthese situations.
The type of carbonate platform developed is deter-mined largely by tectonics and relative sea-levelchange, although one type commonly evolves into another. For discussions, reviews and case histories ofcarbonate platforms see Wilson (1975), Crevello et al.(1989), Tucker & Wright (1990), Tucker et al. (1990),Loucks & Sarg (1993), Wright & Burchette (1996),Harris et al. (1999) and Insalaco et al. (2000). The car-bonate sediments of the major depositional environ-ments are now briefly described.
4.10.3 Intertidal–supratidal carbonates
Tidal flats are areas regularly to rarely covered bywater, dominated by weak currents and wave action.They are developed extensively upon epeiric platformsand they occur along the shorelines of low-energyshelves and ramps, typically behind beach-barriersand around lagoons. Tidal-flat carbonates are domi-nantly lime mudstones, commonly peloidal, althoughlocal lenses of coarser sediment (grainstone) may rep-resent tidal-channel fills. Fenestrae (Section 4.6.3) arethe characteristic structure, giving rise to the distinc-tive birdseye limestone (Fig. 4.24). The fauna may berestricted in diversity; gastropods in particular mayabound, together with ostracods, foraminifers and bivalves. Thin, coarse layers of subtidal skeletal grainsmay occur, transported onto the tidal flat by storms.Microbial mats and stromatolites (Section 4.3.3) aretypical of tidal-flat deposits. Many are simple planarvarieties, showing desiccation cracks and laminoidfenestrae (e.g. Figs 4.12 & 4.25). Small domes may de-velop and in higher-energy areas, columnar stromato-lites. Bioturbation is common, and rootlets may occur.Synsedimentary cementation of tidal-flat sedimentscan produce surface crusts, which may expand to formtepee structures (see Fig. 4.22) and megapolygons(Section 4.6.1). Crusts may break up to give intra-clasts, reworked into edgewise conglomerates andflakestones (Fig. 4.5). Penecontemporaneous dolomi-tization may take place, giving fine-grained dolomitemosaics (Section 4.8.1). In arid climatic areas, theevaporite minerals gypsum–anhydrite, and possiblyhalite, will develop in the sediment. They may be pre-served as pseudomorphs (Section 5.5). Slight upliftand contact with meteoric waters may result inpalaeokarstic surfaces, laminated crusts, calcretes,
vadose pisoids and black pebbles. Tidal-flat sedimentscommonly occur at the top of shallowing-upward cycles (see Section 4.10.9 and Fig. 4.54). Modern carbonate tidal flats and their ancient equivalents havebeen described by Ginsburg (1975), Hardie (1977,1986), Shinn (1983), Tucker & Wright (1990) andDemicco & Hardie (1995).
4.10.4 Lagoonal limestones
Lagoons are subtidal areas located behind barriers,which may be reefs or carbonate sand shoals. Theyoccur on rimmed shelves (shelf lagoon) and along theinner part of carbonate ramps. Protected, lagoonal-type environments are widely developed on epeiricplatforms too. Organisms living in lagoons, and there-fore the sediments accumulating in these predomi-nantly quiet-water areas, depend largely on the degreeof restriction and permanence of the barrier. Lagoonsmay be normal in terms of salinity, as in the lagoons ofatolls; brackish where there is much freshwater runoff,as in the inner part of Florida Bay, for example; or hypersaline to a greater or lesser extent, such as theinner part of the Great Bahama Bank, Shark Bay(Western Australia), and the lagoons of the TrucialCoast (Arabian Gulf). The sediments are variable ingrain size, although many are carbonate muds, rich inpeloids. Aggregates are common. Towards the barrierthe muds pass into coarser sediments, and coarseskeletal debris may be derived from small, coral patchreefs, which commonly grow in lagoons of normalsalinity. The lagoon floor is dominated by molluscs,green algae and foraminifers. The green algae in par-ticular are a major sediment contributor (Section4.3.3), and microbes play a significant role in skeletalbreakdown and production of micritized grains. Surfi-cial microbial mats and sea-grasses may cover the la-goon floor, as in the Bahamas and Shark Bay (Section4.3.3). Bioturbation is intensive, largely through theactivities of crustaceans and bivalves. Sedimentarystructures may be poorly developed, although thin,vaguely graded beds of coarser grains and shell lagsmay be formed through periodic storm reworking.
Lagoonal limestones are common in the geologicalrecord, particularly in back-reef and back-barrier situations passing shorewards into tidal-flat facies.Many are wackestones and mudstones, in some caseswith fossils preserved in growth position. Examplesare the Amphipora wackestones of Devonian back-
156 Chapter 4
reefs in western Canada, Europe and Australia, thecalcisphere-rich porcellanites of the Carboniferous,and the thick-shelled megalodont bivalve facies of Triassic back-reefs.
4.10.5 Intertidal–subtidal carbonate sand bodies
These sand bodies occur in areas of high tidal currentand wave activity and include barriers, beaches, theshoreface and tidal deltas along ramp shorelines, andshoals and banks along exposed, rimmed-shelf mar-gins. Depths of deposition are mostly less than 5–10m.The sediments are carbonate sands and grainstonescomposed of ooids and rounded and sorted skeletalgrains. The latter are fragments of normal-marine organisms: corals, bivalves, foraminifers and algae,and echinoderms and brachiopods in the Mesozoic–Palaeozoic. Sedimentary structures are ubiquitous,chiefly cross-bedding on all scales, perhaps with herringbone cross-bedding through tidal-current reversals (Fig. 2.23), reactivation surfaces, HCS+SCS, also planar bedding, scours and channels (seeSection 2.3).
Carbonate shoreline sands are similar to their silici-clastic equivalents (see Section 2.11.5), in terms oftheir facies succession and sedimentary structures. The northeast Yucatan coast of Mexico is a modernexample of a carbonate strandplain and the TrucialCoast of the Arabian Gulf is a barrier–lagoon–tidal-delta system. The Jurassic Lincolnshire Limestone of eastern England was deposited in such a barrier-island–lagoon system and the Jurassic Great Oolite of southern England was also deposited along aninner-ramp shoreline. Oolites in the South Wales Carboniferous were deposited in strandplains andbarriers of a carbonate ramp (Burchette et al., 1990). The Jurassic Smackover Formation of the Gulf Coast Rim was deposited in similar inner-rampenvironments, as were Mississippian grainstones inthe Illinois Basin (see papers in Wright & Burchette(1998)).
Shelf-margin sand bodies are best developed alongwindward shelf-margins, where skeletal debris alsomay be derived from nearby reefs. The shoals are elon-gate parallel to the shelf-break and traversed by tidalchannels that have lobes of sediment (spillovers) located at their ends. The shoals are covered in sand-waves and dunes that mostly are orientated lagoon-wards. Lily Bank on the northeast margin of the Little
Bahama Bank is a good example of this sand-bodytype. Through time, windward sand shoals may pro-grade into the lagoon to generate a muddy to grainy,coarsening-upward unit. Continued growth of a sandshoal may lead to the development of a sand flat, withislands. This has happened at Joulters Cay in the Ba-hamas over the past 5000years. In areas where tidalcurrents are strong, sand bodies are ridges orientatedmore normal to the shelf break, with grass-coveredmuddy sands in the depressions between. Such ooliteridges are prominent on the platform margins at theheads of the deep channels cutting into the BahamaBanks (e.g. Schooner and Exuma Cays, see Gonzalez& Eberli, 1997). Lime sand also accumulates alongmore leeward shelf-margins, and here off-shelf trans-port of sediment to the adjoining foreslope is a majorprocess in platform progradation. This has occurredalong the western side of the Great Bahama Bank, gen-erating clinoforms (see Section 4.10.8 and Eberli &Ginsburg, 1989).
Occurring in high-energy locations, where water is continuously pumped through the sands, thesegrainstones commonly have marine cements, andhardgrounds may be formed. Ancient examples ofshelf-margin, carbonate sand bodies include Zech-stein oolites in northeast England (Kaldi, 1986). Seepapers in Harris (1984), the review of Handford(1988) and Mutti et al. (1996).
Oolitic and skeletal grainstones are important hydrocarbon reservoirs. The Jurassic Smackover For-mation of the Gulf Coast Rim has substantial oil re-serves. The type of porosity varies considerably overthe region, from primary intergranular to dissolution-al to intercrystalline where dolomitized. The originalmineralogy of the ooids, aragonite to the north (land-wards) and calcite to the south (basinwards) is a majorfactor in porosity evolution (Heydari & Moore,1994).
Carbonate sands are also deposited in deeper waterson carbonate ramps by the action of storms. The sedi-ments are grainstones and packstones, especiallywhere depths are close to wave-base, passing intowackestones and lime mudstones in deeper water.Storm waves and currents generate HCS in abovestorm-wave-base lime sands, and shell lags and gradedbeds below storm wave-base (see Section 2.3.2). Bio-turbation is widespread. Common skeletal compo-nents are molluscs, chiefly bivalves, foraminifers andcoralline algae, and brachiopods and echinoderm
Limestones 157
debris in the Mesozoic and Palaeozoic. Terrigenous siltand clay are an important constituent of many outer-ramp and deep-shelf limestones. Storm-depositedlimestones (‘tempestites’) are described by Aigner(1984) from the Triassic in Germany, Handford(1986) from the Mississippian of Arkansas and Jennette & Pryor (1993) from the Ordovician of Ohio.
4.10.6 Reefs and carbonate build ups
Although coral reefs are one of the most familiar andmost studied of modern carbonate environments,there are many other types of reef developing at thepresent time and preserved in the geological record.Such carbonate build ups, a term widely used for locally formed limestone bodies that had originaltopographic relief, are a common feature of many car-bonate formations going back to the Precambrian.The literature on modern and ancient reefs is vast; reviews and compilations include Toomey (1981),James (1983), Fagerstrom (1987), Geldsetzer (1989),Loucks & Sarg (1993), Monty et al. (1995), Harris et al. (1999) and Wood (1999).
The term reef itself is best restricted to a carbonatebuild up that possesses(ed) a wave-resistant frame-work constructed by organisms, but to be clear the
term ecological reef or organic framework reef can beused for this. Specific types of such ecological reef,shown in Fig. 4.44, are patch reef, small and circular inshape; pinnacle reef, conical; barrier reef, separatedfrom the coast by a lagoon; fringing reef, attached tothe coast; and atoll, enclosing a lagoon. Other termsfrequently used are bioherm for local in situ organicgrowth with or without framework; biostrome for lat-erally extensive in situ growth with or without frame-work; organic bank or loose skeletal build up for anaccumulation of mostly skeletal sediment, chieflythrough trapping or baffling; mud-mound or mudbank (formerly reef knoll) for an accumulation ofmostly lime mud (micrite), probably by trapping andbaffling.
Many different organisms can be and have been in-volved in the construction of reefs. At the present time,the main reef builders are corals and coralline algae;others of limited importance are sponges, serpulids,oysters and vermetid gastropods. In the past, practi-cally all invertebrate groups have at one time or another contributed to reef growth. Special mentioncan be made of cyanobacteria and microbes (formerlyblue–green algae) constructing stromatolite bioherms/biostromes in the Precambrian and Cambrian (locallyin younger rocks too), stromatoporoids in the Ordovi-cian to Devonian, rugose corals in the Silurian to
+ ++ + ++
++
++
++
++
+
++
++
+
+
++
+
+
++
++
++ +
+
+
+
+
+
++
Patch reef Shelf-marginreef complex/barrier reef
Back-reef Reef
Fore-reef
Shelf
Pinnaclereefs
Mudmounds
Fringingreef Atoll
RampForeslope
Fig. 4.44 The main types of reef and theirlocation.
158 Chapter 4
Carboniferous, scleractinian corals from the Triassicto Recent, phylloid algae in the Carboniferous andPermian, sponges in the Triassic–Jurassic and rudistidbivalves in the Cretaceous (see Kiessling et al. (1999)for a review).
Organisms in reefs take three roles: (i) the frame-builders, those providing a skeletal framework (coralsat the present time); (ii) the frame-binders and en-crusters, organisms which consolidate the framework,such as calcareous algae and bryozoans; and (iii) thereef-users, such as boring bivalves and sponges, preda-tory fish and echinoderms. With many ancient reefs itis clear that there was no true solid framework, butmuch in situ organic growth. This last feature gives rise
to the two typical features of reef limestones, a massive appearance with no stratification (Fig. 4.45) and thepresence of organisms in growth position (Fig. 4.46).Many reef limestones would be classified as bound-stones, with the terms framestone, bindstone and baf-flestone used to describe specific types (see Section4.4); some reef-rocks do not have an obvious reefstructure on the hand-specimen or thin-section scaleand are a wackestone, lime mudstone or even a grain-stone. Primary cavities are a feature of many reefs, al-though they are commonly filled with skeletal debrisand cement. Synsedimentary cementation is a featureof many modern and ancient reefs (Section 4.7.1).
There are many factors controlling the growth of
Fig. 4.45 Devonian reef limestones in theCanning Basin, Western Australia. Thereef consists of massive unbeddedlimestone (central part of cliff); in front ofthe reef (to the left) occurs a talus slope ofreef debris, these fore-reef limestoneshave an original dip; behind the reef (tothe right) are flat-bedded limestones ofthe back-reef lagoon.
Fig. 4.46 In situ coral colony (Thecosmilia)from Triassic reef. Adnet, Austria.
Limestones 159
modern coral reefs and it is likely that these same fac-tors exerted an influence on coral and other reefs in thepast. For coral-reef growth, these factors are:1 water temperature —optimum growth occursaround 25°C;2 water depth —most growth takes place within 10mof the surface;3 salinity —corals cannot tolerate great fluctuations;4 turbidity and wave action —coral growth isfavoured by intense wave action and an absence of terrigenous silt and clay.
The majority of reefs occur along shelf margins, anagitated zone where waves and currents of the opensea first impinge on the sea floor. Smaller patch reefsdevelop in open lagoons behind shelf-margin reefs, oncarbonate ramps and epeiric platforms. Reefs, usuallyatolls, are also developed on submerged volcanic is-lands within the ocean basins. The configuration andmorphology of some present-day reefs is a reflection ofkarstic dissolution of earlier reefs during glacial low-stands of sea-level.
Many modern reefs along shelf margins show acharacteristic threefold division into fore-reef (reeffront/slope), the reef itself (reef-crest, reef-flat) and
back-reef. The reef front is a steep slope, vertical inplaces, with organisms constructing reef in the upperpart, passing down to a talus slope of coarse reef debris. Reef-derived carbonate breccias, debrites andturbidites may be present in the adjoining basin. Aprominent system of surge channels gives a spur andgroove morphology along the reef front, extending upto the reef flat, in some cases. The reef crest, covered byno more than 1–2m of water, is the site of prolific organic growth, of corals and algae on modern reefs.Behind the crest is the reef-flat, a pavement of mostlydead coral. The back-reef area consists of reef debrisadjacent to the reef-flat, passing shoreward to a quiet-water lagoon, where there may be patch reefs. Abroadly similar facies pattern is seen in many ancientbuild ups (Fig. 4.45). Classic examples include the Per-mian Capitan Reef in Texas (see Saller et al., 2000), the Devonian reefs of western Canada, Europe andAustralia, and the Triassic reefs of the Northern Calcareous Alps (Austria) and Dolomites (Italy).
Mud-mounds are massive accumulations of limemudstone and wackestone, tens to hundreds of metresacross, which pass laterally into well-bedded lime-stone. The mud-mounds contain only scattered fossils;those of some possible significance are crinoids, bryozoans, sponges and algae. Much of the mud ispeloidal. Mud-mounds typically occur in deeper wateron carbonate ramps and shelf slopes. The best devel-oped mud-mounds occur in the Palaeozoic, in the Carboniferous of northwest Europe, where some arereferred to as Waulsortian reefs, the Carboniferous ofNew Mexico and Montana, and the Devonian of NewYork State and Belgium (Fig. 4.47).
The origin of mud-mounds is still a problem.Processes that could be involved are (a) the precipita-tion of lime mud by microbes, (b) entrapment of mudthrough the baffling action of bryozoans and crinoidsand (c) concentration of mud into mounds by currents.Comparisons have been made with the mud banks ofFlorida and Shark Bay, where lime mud and skeletaldebris is baffled and trapped by sea-grass and algae(Bosence, 1995). The sediment itself is derived fromgreen-algal disintegration (Section 4.3.4) and thebreakdown of larger skeletal grains. Many papers on mud-mounds can be found in Monty et al. (1995).Perhaps the most spectacular examples occur in theAlgerian Sahara (Wendt et al., 1997).
Reefs are important hydrocarbon reservoirs withthe most porous facies generally occurring in the upper
Fig. 4.47 Lower Carboniferous mud-mound (muleshoe mound),Sacremento Mountains, New Mexico. Photograph courtesy ofDave Hunt.
160 Chapter 4
fore-reef and reef-framework facies. However, marinecementation is most prevalent in this zone and soporosities can be reduced from high primary values.The talus at the toe of reef slopes also makes goodreservoirs. Many reefs are dolomitized, and also thisenhances their reservoir qualities. Examples includethe Devonian reefs of western Canada, such as theLeduc and Golden Spike, and the Cretaceous rudistreefs of the Gulf Coast Rim, Mexico and the MiddleEast. Rudist bivalves have very porous skeletons butmuch of this intraskeletal porosity is not connected.The best reservoirs are thus located in rudist debrisbeds, rather than in the reefs themselves (which tend tobe muddy) or in rudist limestones that have sufferedmuch mechanical compaction during burial.
4.10.7 Pelagic limestones
Where water depth is too great for benthic organismsto flourish, in excess of some 50–100m, then carbon-ate sediments composed of pelagic organisms will ac-cumulate in the absence of clay. The maximum depthof accumulation is controlled by the rate of carbonatedissolution. In low latitudes, the ocean is saturatedwith respect to CaCO3 in the upper few hundred me-tres and below this it becomes undersaturated, firstwith respect to aragonite, and then calcite. Below a fewhundred metres CaCO3 begins to dissolve, but it is notuntil greater depths that the rate of CaCO3 dissolutionincreases substantially (this depth is the lysocline) (seeFig. 4.48). The depth at which the rate of dissolution isbalanced by the rate of supply is known as the carbon-ate compensation depth (the CCD). This depth variesin the oceans, its position being controlled by calcare-ous plankton productivity, which itself depends largely on nutrient supply, and water temperature. Inthe tropical regions of the world’s oceans, the CCD forcalcite is between 4500 and 5000m; the CCD for arag-onite is about 2000m less. The CCD shallows intohigher latitudes and seawater is undersaturated in re-spect of CaCO3 in temperate and polar waters. Cal-careous oozes can accumulate on the sea floor, which isshallower than the CCD; siliceous oozes and red claysare present below this depth. Fluctuations in the CCDback into the Cenozoic and Mesozoic are now welldocumented.
Modern pelagic carbonates are composed ofpteropods (aragonitic), coccoliths (Fig. 4.9) andforaminifers (both calcitic), and are found on outer
continental shelves, continental slopes and oceanfloors starved of terrigenous clay, and on submarinerises, drowned reefs and volcanoes (seamounts andguyots) rising from the ocean floor. Ancient pelagiclimestones occur in the Mesozoic of the Alpine region,the Ammonitico Rosso, Maiolica, Biancone andScaglia, for example, and in the Devonian and Car-boniferous of Hercynian Europe, the Cephalopoden-kalk and Griotte. Characteristic features of pelagiclimestones, apart from a dominantly pelagic fauna(e.g. Fig. 4.17), are their condensed nature and evi-dence for synsedimentary cementation in the form ofhardgrounds, lithoclasts, sheet cracks and neptuniandykes. Many pelagic limestones are nodular (Fig.4.49) and some contain ferromanganese nodules andcrusts (Section 6.7). Resedimentation of pelagic sedi-ment is common too, particularly in slumps anddebrites.
The Cretaceous chalks of northwest Europe andsouthern USA are composed largely of coccoliths andare thus pelagic limestones. Deposition took place atdepths of around 50–150m, so that there is a signifi-
Increasing rate of dissolution
Dissolution
Sediment supply
Lysocline
Calcareous oozesCCD
Siliceous oozesand red clay
0
1
2
3
4
5
6
Dep
th (
km)
Fig. 4.48 Carbonate dissolution in the ocean. Below thecarbonate compensation depth (CCD), carbonate sediment doesnot accumulate.
Limestones 161
cant benthic macrofauna of echinoids, bivalves andbrachiopods. Hardgrounds are common within thechalks and these are mineralized with phosphate andglauconite. In the Chalk of the North Sea, resedimen-tation is widespread, with major slides and slumps,debrites and turbidites, all greatly expanding thethickness of the formation. The North Sea Chalk is amajor hydrocarbon reservoir and depositional as wellas diagenetic factors were important in the acquisitionof reservoir qualities. The most porous horizons arecommonly those that were resedimented, rather thanthe pelagic chalks. Fracturing of the chalk through up-doming of Zechstein evaporites has greatly improvedthe permeability. Early oil entry and overpressuring inhibited burial compaction and cementation. SeeGlennie (1998) for more information on North Sea petroleum geology.
Papers on pelagic limestones are contained in Hsü &Jenkyns (1974); also see the reviews of Scholle et al.(1983), Tucker & Wright (1990) and Clari & Martire(1996).
4.10.8 Resedimented deep-water limestones
Shallow-water carbonate sediment is transported intothe deep sea by the same processes as discussed for sili-ciclastics (see Section 2.11.7): slides, slumps, debrisflows, turbidity currents and modified grain flows. Inthe slope area between shelf and basin, slump-foldedlimestones and slides are common, especially in pelagic–hemipelagic slope facies. Megabreccias, in-cluding very large blocks of shallow-water limestone,
are common off major shelf-margin reefs (e.g. Fig.4.50), and many have formed through sea-level falls(Spence & Tucker, 1997). Carbonate debrites are veryvariable in texture, commonly forming a spectrumfrom unsorted and chaotic breccias, to inversely andnormally graded breccias, perhaps with associatedgraded grainstones (A to E in Fig. 2.78). Such beds areprominent in the Devonian of the Canadian Rockiesand Cambro-Ordovician Cow Head Group of New-foundland. In the basins themselves, limestone tur-bidites are usually interbedded with hemipelagic dark shales (Fig. 4.51). Sole structures, graded andplanar bedding and cross-lamination are all developedin limestone turbidites, just as in siliciclastic examples,although in many, good Bouma divisions (see Section2.11.7) are not present.
Carbonate turbidite basins are usually fed from aline source, i.e. the whole platform margin, rather thanone (or several) point sources(s) as in the case of manysiliciclastic turbidite basins. Thus the submarine fanmodel is not appropriate for most limestone turbiditeformations, and in fact very few ancient, carbonatesubmarine fans have been described. Resedimentedlimestones are usually ascribed to either the slopeapron (see Fig. 4.52), or base-of-slope apron models,depending on whether the resedimented limestones interfinger with the shallow-water shelf limestones(the first model) or whether the upper slope is bypassedand sediment accumulates at the toe of the slope (thesecond model). Ancient examples of slope aprons arewell developed in the Triassic of the Dolomites(Bosellini, 1984), the Permian Capitan Reef of Texas
Fig. 4.49 Pelagic limestone: a micritic, nodular limestone thatcontains a pelagic fauna. Pressure dissolution effects are commonand result in the clay seams (flasers) between nodules. Devonian.Montagne Noire, France.
Fig. 4.50 Block of reef-rock in deep-water hemipelagicmudstones, which also contain limestone turbidites of shallow-water skeletal debris. Triassic. The Dolomites, Italy.
162 Chapter 4
and the Devonian Canning Basin reefs of Western Aus-tralia (Fig. 4.45), where large-scale, quite steeply dip-ping sheets of shallow-water debris extend up to theplatform margin. These megascale cross-beds arecalled clinoforms, a term coming from seismic stratig-raphy for dipping reflectors. Base-of-slope apron de-posits are known from the Lower Palaeozoic of theAppalachians and Western Cordillera in North
America (Mullins & Cook, 1986). Resedimented Cre-taceous rudist limestones are important oil reservoirsin the Golden Lane Atoll of Mexico (Enos & Stephens,1993). They were deposited in a base-of-slope apronwith debris coming from shelf-margin rudist reefs.Basin-margin, resedimented carbonates are com-monly the sites of syngenetic mineralization. Hydro-thermal fluids ascending along faults have
Fig. 4.51 Limestone turbidites composedof crinoidal and other skeletal grainsinterbedded with hemipelagic mudrocks.Succession is inverted. This is the typicalappearance of turbidites (both carbonateand siliciclastic) in laterally continuousbeds with interbedded deep-watermudrocks. Devonian. Cornwall, England.
Upper slope
Slope apron
slumps,debrites,
few blocks,coarse turbidites Shelf margin
reefs and sands
Clinoforms ofshelf-margindebris
Lower slope
few debrites,medium – fine
turbidites,pelagics
Basin floor:pelagic andhemipelagicdeposits andthin turbidites
Fig. 4.52 Facies model for a carbonateslope apron where slope faciesinterdigitate with platform margin facies.In the base-of-slope apron model, notshown here, sediment mostly bypasses theupper slope and is deposited at the toe ofthe slope. From Tucker & Wright (1990).
Limestones 163
precipitated base-metal sulphides in the porous limestones in a number of cases (see Anderson & McQueen, 1982).
4.10.9 Carbonate sequences
Carbonate sedimentation responds readily to relativechanges in sea-level and within many carbonate for-mations, large-scale and small-scale cycles can be iden-tified. As noted in Chapter 1, relative sea-level changesoperate on many scales. For carbonates, the long time-scale second- (107 years) and third- (106 years) ordersea-level changes are responsible for the developmentof the carbonate platforms themselves and the deposi-tion of hundreds of metres of limestone. Within theseplatforms, depositional sequences can be identified, byrecognizing the key surfaces, and their component sys-tems tracts, from the large-scale arrangement of faciespackages, in terms of their onlap (retrogradation),aggradation and offlap (progradation) or from the fa-cies successions, reflecting relative sea-level changes.Rimmed shelf margins, where sedimentation rates arehigh, are strongly affected by relative sea-level changesand stationary (aggradational), offlapping, onlap-ping, back-stepping, drowned and emergent typeshave all been described. In the concepts of sequencestratigraphy, most carbonate deposition takes placewithin the transgressive and highstand systems tracts,especially the latter when large quantities of sedimentare deposited in the basin (highstand shedding) and
clinoforms are well developed. During sea-level low-stands, carbonate platforms are usually emergent andso subject to extensive meteoric diagenesis and karsti-fication. However, collapse of the margin may takeplace at this time to generate megabreccias. Figure4.53 shows a sequence stratigraphic model for arimmed shelf. For papers on carbonate sequencestratigraphy see Loucks & Sarg (1993) and Harris et al. (1999).
Metre-scale cycles (or parasequences) are commonwithin platform carbonate sequences. These are the result of depositional processes and fourth- and fifth-order relative sea-level changes, operating on a 104–105 years scale. Many of these cycles display ashallowing-upward trend. There are many types but acommon one consists of shallow subtidal sedimentspassing up into tidal-flat facies, with evidence of emer-gence at the top (Fig. 4.54). The latter may take theform of a palaeokarstic surface or palaeosoil. Mixedsiliciclastic–carbonate cycles (see Fig. 2.68) andevaporite–carbonate cycles (see Figs 5.16 & 5.17) alsooccur. The parasequences may be repeated many times
Reef Grainstone Talusgrainstone 20–30° slope
mudstone 5–10° slope
Packstone/Mudstone
sbmfs
Retrogradational TST
Progradational(offlapping)highstandclinoforms
HST
HST
Lowstand megabreccia (LST)
Condensedsection (TST)
AggradationalTST
mfssb/ts
Onlap
sb/ts
Fig. 4.53 Sequence stratigraphic model for a rimmed shelf.There is much possible variation in the response of carbonatesedimentation to relative sea-level change. Here, the firstsequence shows the case of reefal sedimentation keeping upwith the sea-level rise during the TST so that the sedimentsaggrade, whereas in the second sequence the sea-level riseexceeds carbonate production in the TST and the shelf-marginsands back-step, i.e., retrograde. Abbreviations as for Fig.2.86.
164 Chapter 4
in a succession, and they can show systematic varia-tions in thickness and character upwards. Upward-increasing thickness and proportion of subtidal faciesin each successive cycle generally indicate an overalltransgressive trend, and upward-decreasing thicknessof cycles and increasing proportion of tidal-flat faciesreflect an overall regressive trend. Many parase-quences have formed through tidal-flat progradationand the lateral, seaward migration of facies belts.However, there has been much argument over thecauses of the repetition of the cycles. Purely sedimen-tary processes have been invoked in a migrating tidal-island model and loss of carbonate source area. Atectonic mechanism of periodic fault movement orjerky subsidence has also been suggested (e.g. Satter-ley, 1996). However, it is currently very popular to invoke sea-level changes though ocean-water volumechanges, especially via ice-caps, brought about by orbital forcing in the Milankovitch band (20000–400000years). Statistical analysis and computer mod-elling of carbonate cycles have lent support to an as-tronomic control. See the reviews and papers byTucker & Wright (1990), Wilkinson et al. (1997),Lehrmann & Goldhammer (1999) and Lehmann et al.(2000).
Relative sea-level changes are a major factor in thediagenesis of limestones as well as their sedimentation.Long periods of relative sea-level rise lead to the pump-ing of seawater through carbonate sediments, and this can result in extensive marine cementation anddolomitization. Extended periods of relatively low
sea-level allow meteoric water to penetrate a carbon-ate formation, and, under a humid climate, this maygenerate dissolutional porosity, surface and sub-surface karst, calcite cementation and mixing-zone-related dolomitization. If the climate is more arid, thenevaporite precipitation and reflux dolomitization arelikely early diagenetic processes. See the reviews ofRead & Horbury (1993) and Tucker (1993).
Further reading
Crevello, P., Sarg, J.F., Read, J.F. & Wilson, J.L. (Eds)(1989) Controls on Carbonate Platform and BasinDevelopment. Special Publication 44, Society of Eco-nomic Paleontologists and Mineralogists, Tulsa, OK,405 pp.Harris, P.M., Saller, A.H. & Simo, J.A. (Eds) (1999)Advances in Carbonate Sequence Stratigraphy: Appli-cation to Reservoirs, Outcrops and Models. SpecialPublication 63, Society of Economic Paleontologistsand Mineralogists, Tulsa, OK, 421 pp.Insalaco, E., Skelton, P.W. & Palmer, T.J. (Eds) (2000)Carbonate Platform Systems: Components and Inter-actions. Special Publication 178, Geological Society ofLondon, Bath, 240 pp.Loucks, R.G. & Sarg, J.F. (Eds) (1993) Carbonate Se-quence Stratigraphy. Memoir 57, American Associa-tion of Petroleum Geologists, Tulsa, OK, 545 pp.Moore, C.H. (1989) Carbonate Facies, Diagenesisand Porosity. Elsevier, Amsterdam, 338 pp.Morse, J.W. & Mackenzie, F.T. (Eds) (1990) Geo-
Graphic log Sediments
Palaeosoil – palaeokarsticsurface ± calcrete
Interpretation
Fenestral biopelmicrites/wackestones + stromatolites
Biopel- & oosparites/grain-stones, + cross-bedding (etc.)
Local bioherms + biostromes
Biopelmicrites/wackestones± terrigeneous clay, muchbioturbation, storm beds
Basal intraformationalconglomerate
Supratidal/emergence
Tidal flat
Low intertidal–shallow, agitatedsubtidal
Deeper-watersubtidal
Reworking duringtransgression
Fig. 4.54 Shallowing-upward limestonecycle (a parasequence): there are manyvariations on this general theme, dependinglargely on the energy level of the shorelineand on the climate. Such cycles are typically1–10 m thick.
Limestones 165
chemistry of Sedimentary Carbonates. Elsevier, Amsterdam, 696 pp.Purser, B.H., Tucker, M.E. & Zenger, D.H. (Eds)(1994) Dolomites: a Volume in Honour of Dolomieu.Special Publication 21, International Association ofSedimentologists, Blackwell Science, Oxford, 451 pp.Scoffin, T.P. (1987) Carbonate Sediments and Rocks.Blackie, Glasgow, 274 pp.Scholle, P.A., Bebout, D.G. & Moore, C.H. (Eds)(1983) Carbonate Depositional Environments. Memoir 33, American Association of Petroleum Geologists, Tulsa, OK, 708 pp.
Tucker, M.E. & Bathurst, R.G.C. (1990) CarbonateDiagenesis. Reprint Series 1, International Associa-tion of Sedimentologists, Blackwell Science, Oxford,312 pp.Tucker, M.E. & Wright, V.P. (1990) Carbonate Sedimentology. Blackwell Science, Oxford, 482 pp.Tucker, M.E., Wilson, J.L., Crevello, J.R., Sarg, J.R. &Read, J.F. (Eds) (1990) Carbonate Platforms: Facies,Evolution and Sequences. Special Publication 9, Inter-national Association of Sedimentologists, BlackwellScience, Oxford, 328 pp.