CD Home

Copyright 1962 - American Association of Petroleum Geologists. All rights reserved.

CLASSIFICATION OF CARBONATE ROCKS ACCORDING

TO DEPOSITIONAL TEXTURE1
ROBERT J. D W A M P
Houston, Texas
ABSTRACT
Three textural features seem especially useful in classifying those carbanate rocks that retain their
depositional texture (1) Presence or absence of mrhnate mud, which differentiates muddy carbonate
from g r a k s h t ; (2) abundance of gains, which allows muddy carbonates to be subdivided into t~tdsiwss,
wackarlc-ne, and patkstonc; and (3) presence of signs of binding during deposition, which characterizes
bwmhiose. The distinction between grain-support and mud-support differentiates packstone from
w a c k e s l o n ~ k s t o n eh full of its particular mixture or grains, wxkestone is not. Rocks retaining tw
little of their epositional texture to be classified are set aside as crysfdlim corbonades.

terpretation of depositional environment, and it

will provide the convenience of class names based
only on depositional texture.
Ideally, one could confidently divide unaltered
dolomite retain their depositional texture, in a
carbonate
rocks into two main groups, which
more or l e s ghost1y fashion, despite diagenesis. A
descriptive classification based on depositional Gtabau (1904) termed clastic and biogenic. According to the ideal, one group is controlled
texture is thus a generally helpful adjunct to other
classifications, particularly to those based on mainly by hydraulic conditions. Origin of particles
genetic kind of particles and to those based on is relatively insignificant, and subdivision is acmincralogic composition. It will not substitute for cording to particle size. Calciruditc thus is disfurther description and classification, nor will it tinguishcd from calcarenite. The other group is
produce ready-made interpretations, but it will controlled mainly by the biologic or biochemical
focus attention on whichever few textural proper- processes responsible for producing carbonate.
ties are chosen as particularly significant for in- Here, partide size is relatively insignificant, and
subdivision is made according to origin of parti' Part of a symposium arranged by the Research cles. Crinoidal limestone thus is distinguished
Committee, and presented at Denver, Colorado, April from coral l i m ~ t o n e As
. it turns out, most sam27, 1961, under joint auspices of the Assoriation and the
Society of Economic Paleontologists and Mineralogists. ples are neither clearly in one group nor clearly in
Manuscript received, April 23,1961.
the other. Instead, they fall in the middle ground
Shell Develoament Com~anv. Emloration and where hydraulic conditions and biolo~icor bioProduction ~ i v i G o n .I am dkepfy indebted t o R. L. chemii
procmes are in joint control-Folk, who shared his knowledge of carbonates and
Focusing attention on whether or not the
classification prior to publicration, and repcaredly
tributed to the evolution of the present concepts; to laraer
or
- articles are hans~orted(fragmented
.
R. N. Ginsburg, whose studim of Recent carbonates disarticulated) does not ;e-esta,,lkh
the
clarified pmblems in ancient carbonates; and to D. L.
Arnsburv. K. 1. WBU. S. I). Kerr. 0. P. Maiewske. P. F. boundary between the two FJouPs. Too often One
,
Moore, R . ~ . - ~ u r r aJ. M. l'kks, B. F: ~ e r k h s G.
isforced to conclude that &bonate sediment
Rittenbouse, C. I. {hith, B. W. Wilson, and J. L. W= t r m s ~ o r t e da little but not far, or that some
%vason,who mve m,-ouraarment and
cri ticisin.
components were transported and others were not,
INTRODUCTION

The increasing use of thin sections and oled

slabs has shown that most limestone and much

Looses packing would require mud-support. Sam les were impregnated on the beach in such a way as to
were then reirnpregnatE8.~n the laboratory and sliced. All samples were from Cayo

preserve natural packing and

Centro, Cayos Arcas, Campeche Banks, Mexico.

a. Coral lime gravel, X 1, reflected light. South side of island, kC.64.
B. Coral-red algae lime sand, X 25, transmitted cross-polarized light. Northeast side of island, AC-47.
G. Red algae lime sand, XS, transmitted cross-pokrited light. Southeast side of island, AC-83.

Home

PUSE 1,-Grain-support

i n beach deposits.

Home
Home

PLATE
11.-Grain-support

influenced t)y grain shzpe.

Home

CLASSIFICATION ACCORDING TO D EYOSITIONAL TEXTURE

or that degree of transportation is indeterminate.
Indeed, the outstanding differencesbetween landderived sediment and carbonate sediment stem
from the simple fact that, as a rule, Isnd-derived
sediment is produced by destruction far fmm the
site of deposition, whereas carbonate sediment i s
produced by destruction and construction at or
near the site of deposition, The recognized usefulness of ~Iassificationsbased on origin of particles
attests ta the idea that carbonate grains are unlike land-derived sediment in distance of transport; ii grains did not generally remain near
where they were produced, such dassiiications
would lack the environmental significance they
are known to have. Possibly, the ease with which
carbonate grains are destroyed, or cemented, accounts lor their reIatively short transport.
SIGNIFICANCE OF MUD

The distinction between sediment deposited in

calm water and sediment deposited in agitated
watcr is fundamental. Evidence bearing on this

111

problem thus deserves to be incorporated in class

names. This can be accomplished in several ways,
One is to focus attention on average or predominant size, which erroneously assumes that all
sizes in a sample are equally significant hydraulically. Another is to focus attention on the size,
abundance, and condition of the coarse materia1
brought to the site of deposition. This emphasis
on what might be called currents of delivery has
long been successFu1 in dealing with landderived
sediments, but does not work well in lime sediment because of the local origin of many coarse
grains. A third way is to focus attention on the
fine material that was able to remain a t the site of
deposition. This emphasis on what might be called
currents of removal seems advisable if we wish to
characterize carbonate sediment systematicalIy
in terms of hydraulic environment.
Inasmuch as d m water is characterized by
mud being able to settle to t h e bottom and remain
there, it seems that the muddy rocks deserve to be
contrasied with mud-free rocks, scgardIess of the

E X P L A N AOF
~ NPLATEI1
h
r packing would require mud-support. Aggregates were made by sedirnenting wet graina in quiet water
onto a bed of reviously deposited pains ( H d h d u plates visible at base of aggregate), drying, impregnating with
clmr plastic,
then staining. Plate l I 4 wa?l sedimentcd dry.
Per cent graina i s expressed in two ways, both values being rounded to the nearest 5
cent. Per cent grainsolid, which is measured by water displacement, refers to the vdurne of solid matter In t e grains divided by the
volume of the qgrepte. Per cent grain-hulk, which is measured by point-counting, differs in that opening and
deep indentations in grains are conventionally counted as grain instead of as pore.
a. Well ellrted, well rounded, highly spherical porcelain balls, X I , 65 per cent grain-bulk, 60 per cent p i n -

seeing,

~r

solid.
b. High-spired snails, X I), 45 per cent grain-bulk, 25 per cent grain-solid.
c. Small clams, X 1b, 30 per cent grain-bulk, 25 per cent grain-solid.
d. Corn W e s , X I , 30 per cent grain-bulk.
B. Rose corals, X 3, 70 per cent grain-bulk, 20 per cent grain-solid.
f. Branching red algae broken to $-inch lengths, X I #,20 per cent grain-bulk, 15 per cent pin-solid.

Home

PLATE 111.-Indications

of grain-support.

Home

CLASSIFICATION ACCORDING TO DEPOSITIONAL TEXTURE

amount and size of included coarse material. Folk
(1959) and Brarnkamp and Powers (1958) in their
new classihcations of carbonate rocks have much
the same viewpoint. The terms mud and grains
are used variously by different authors. Usage
here is based on particle size, grains being larger
than 20 microns and mud being smaller than 20
microns. The distinction thus parallels the distinction between matrix and grains in sandstone (Pettijohn, 1957,p. 284). Freedom from mud is taken
to mean virtual a b s e n c e l e s s than 1 per cent.
GRAIN-SUPPORT AND MUD-SUPPORT

Rocks bearing carbonate mud constitute the

bulk of many carbonate sequences and therefore
require subdivision. The m a t useful textural subdivision of such rocks seems to be on the basis of

abundance of grains. Such subdivision allows

mapping of gradients in rate of production of
grains relative to rate of accumulation of mud.
Three degrees of abundance can be recognized.
In the most common case, grains are abundant
enough to be prominent, say more than 10 per
cent, but are not so abundant as t o support one
another. In such a texture, the grains are sometimes said to be "floating." Here they will be
called "mud-supported." Rocks in which grains
are less abundant than I 0 per cent constitute a
second category. Rocks in which grains are so
abundant as to support one another, just as they
do in mud-free rocks, which are necessarily
"grain-supported," are a third category.
The distinction between mud-support and
grain-support seems to be more meaninglul than

Floored interstices, sheller effects, embayed contacts, and overly dose packing indicate grain-support.
ad-Floored interstices are pducerl by fine sediment filtering into coarser sediment, and by m n c u m t
deposition of h e and coarse Sediment. Marine infiltering reaches deeply only if the size difference is large, and
even then does not fill interstices to the root.
a. Oyster shells infiltered by lime mud in the laboratory, X 1, reflected lipht. Mud floors (Fi)are layered, and
drape over irregularities. Shrinkage during drying r e d u d the amount of mud in the intentim, but voids (Lr) were
beneath the shells hefort! drying.
b. Gravel. H d i m d a , and dliths infiltered by very fine sand in tbe lahratory. XI. rdected baht. Sand
largely fills interstices in thc p v c l la er, rwhespart ktythrough the Hdimlmeda layer, and do= notknter the
d t h layer. The voids ( V ) bzneath t e pebbles are original shelter effects-the sediment did r o t shrink during
drying.
c. Beach sand infiltered by lime mud in the laboratory, X5, reflected light. Xndltering (M) is negligible except
in the top layer of grains. Note the shelter effects at the large p i n (S).

d. FIwred interstices (Fd) produced by concurrent deposition ot otilithn and HdiwPneda in the laboratory, in
volume ratio of 3: 1, X3, reflected light. Oijliths making h o r s (Fi)in tower quarter of picture filtered into interstices in previously deposited bed of pure Hulimeda.
e. Embnved contacts. X20. transmitted unnolarized tight. Lake Vallev limestone. Mississi~~ian.
, Sacramento
Mountains, ~ e Mexico,
w
AGN.'
f. Overly close packing, X 50, transmitted u n p o l a M light. Mission Canyon formation, Mississippian, Shell
Richey NNP-I, Richey field, D a m n County, Montana, N f 1-7312.
Ga

..

Home

114

ROBERT J. DUNHAM

any aIternative percentage boundary. A gainsupported rock is full of its particular mixture of
grains, whereas a mud-supported rock is not. In
landderived sediments, the figures 65 per cent
grains, 35 per cent porosity or interstitial debris,
have been used to mark the boundary between
grain-support (frameu-ork) and mud-support (d isrupted framework). If all carbonate grains were
as equidimensional as oiiljths, a percentage
boundary rnight equally well be substituted for
visual distinction between mud-support and
grain-support in carbonate rocks; but carbonate
gains commonly are shaped like potato chips and
twigs instead of like marbles. Because of this, a
grain-supported rock whose grains are ImmotriQ,a
platy alga shaped rather like a cornflake, or
Haliaeeda, would contain a far smaller percentage
of grains than would a grain-supported rock whose
gains are crinoid ossicles. A 65 per cent boundary,
or a 50 per cent boundary, would put the Iwnolba
rock in one group and the ainoid rock in another
group; yet both rocks would contain as much of
their particular mixture of grains as their volume
allows them to hold.
Objection may arise that differentiating mudsupported from grain-supported rocks is impossibly subjective because of the need to envision a
three-dimensional arrangement of irregular shapes

by looking at a two-dimensional view. The difticulty is real but not so bad as it at first seems.
(The errors are not much greater than those encountered in attempting to measure ot estimate
per cent g a i n s in rocks having hollow or indented
grains.) Experience gained in examining mud-free
carbonates, which are necessarily grain-supported
is an aid in determining what the kind of support
is in muddy carbonates. Other aids are the
floored interstices, embayed grains, overly close
packing, and sheltering effects seen in gain-supported rocks and not in mud-supported rocks.
Illustrations of these and of grainsupported
natural and artificial aggregates are shown in

Plates 1-111.
The phenomenon of grain-support has the
added importance of bearing on the postdcpasitional history of the sediment. Soluble grains that
a r t grain-supported are in contact with each
other, and thus have a chance of forming a connected network of molds (or incomplete molds)
unlike the relatively disconnected molds in mudsupported rocks. Furthermore, compaction affects
grain-supported sediment differently than it does
mud-supported sediment. The grains in grainsupported sediment carry the weight of the overburden. This tends to cause weak grains to brecciare. I t also tends to protect from compaction

(XS, transmitted unpolariaed light)

a. Coral lime borrRdsJona. Toronto limestone, Pennsylvanian VirgiE, Greenwwd County, Kmms, L38-8b.
b. Laminated lime h d s J o n c (stromntotite). Cabbage-head structure in West Spring Creek limestone,

Ordovician, Arbuckle Mountains, ADN-I .

G. F b d lime boundstme. Floored opening (P)constructed by intergmwn complex (C) of encrusting Foraminifera, algae, and hydro man^? Pernuan Wollcamp. Shell State ETA-5, Townsend field, L a County, New
Mexico, ETAS-103M.2.

Home

PLATEIV.-Minding

in boundstone.

PLATEV.-Mudstone and wackestone.

Home

CLASSIFICATION ACCORDING TO DEPOSITIONAL TEXTURE

DEPOSITIONAL

TEXTURE

Cradrrswported

Mud-su~wrted
Mae thm
10 percell1 qratns

Mudstone

Wnckestone

Original cwnponents
were bwnd topther
durlns depasltlon..
as rtmnn 5y i n t e w w n
skeletal matter
Imlfl"tlon m n t r a l to k v r ! ~ ,
m red~ment~loored
cavllres that
ae mokd ever by wganac 4
questiol~ablyorga111cmdrer m d
are l m law to be IfltersLlces.

Contamr mud

I p d c l e s d chny and iine silt slze )

Less han
10 percent grams

OEPOSITIOWRL TEXTURE
NOT RECOGNIZABLE

RECOGNIZABLE

Orlglnal Cmpomnts Not Bound Togelller 3 v l r q Depsltlon

Packslent

Lackt mud
and i s
-

aaln-suppc++ed

Grarnstone

117

CrysLalllne

Carbonate

I Subdluide accordlos to
class~f~catruos
deslgnpd t~ bezr
on fiysncal texture or dlalenesrs.1

Boundstar~c

any sheltered mud beneath strong grains, perhaps

making the mud more susceptible to leaching.
Collapse brccciation improved permeability in

during deposition are scarce, but are worth special

attention. Three signs of binding during depositian are recognized (Pl. IV). One i s interconnected
muddy grain-support4 l i m ~ t o n e composed skeletal matter, such as occurs where colonial
mostly of Iwnouia in the Paradox formation of the corals or encrusting Foraminifera grow one on the
Desert Creek field, Utah (Murray, 1960, p. 66). other. Another is lamination contrary to gravity,
Leaching oi mud from interstices in muddy grain- such as occurs in the crinkly lamination of
supported limestone composed mostly of crinoids strornatulites. The lhird is sediment-floored cavicaused the porosity in some of the Devonian ties that are too large to be interstices and are
reservoirs in Andrews County, Texas (F.J. I ~ c i a , roofed over hy organic or questionably organic
matter, such a
personal communication).
q the large and smalI tunnels and
grotto- in coral reefs.
BINDING

hfost carbonate rocks retaining their depositional texture are ljthified sedin~cnt made of
clearly discrete and origimlIy loose particles. The
rocks that differ by showing signs of being bound

CLASSES AND NAMES

The concepts outlined abovc allow five textural

classes to be recogi~hed(TaMe I). The names
tentatively attached to these classes are fairly

(X25, transmitted unpolarjd light)

a. Ostmod lime m&ione, less than 1 per cent pin-bulk. Dupemw formation, Late Devonian, Brown's
Gulch section, Rlaine County, Montana, BG-17.

6. Mixed-fossil doIomite wuck&wre, 20 per cent pin-bulk. The rock contains no calcite. San Andreas Permian,
Shell D. Roberts No. 15, 5,124 ieet, Wasson field, Ymkum County, Texas, 1-300.
G. Ostracod-Lithiclast lime m b s I w , slightly dolomitized, 40 per cent grain-bulk. Duperow formation, Late
Devo~an,Shell Richey NP-1, 8,900 feet, Richey field, McCone County, Montana, 1.562.

Home

ROBERT J.

118

short, meaningful English nouns that apply only ate rocks are termed packstone (Pl. VI). Grainto texture. They can be combined with names for support is generally a property of rocks deposited
grain-kind classes and mineralogic classes, as in agitated water, and muddiness is generally a
shown in the figure captions, ant1 they also can property of rocks deposited in quiet water. A rock
stand alone where mineralogy and grain-kind are exhibiting both properties is peculiar, and it is
well to have i t isolated for further study. It may
nat at issue.
Mdsbone.--Muddy carbonate rocks containing record simple compaction of wackestone, as is
less than 10 per cent grains (10 per cent grain-bulk suggested where interstices are completely filIed
as defined in P1.11) are termed mudstone PI. V). with mud. It may record early or late infiltering
T h e name mudstone is synonymous with calcilu- of previously deposited mud-free sediment, or
tite, except that it does not specify mineralogic pro1iL production of grains in calm water, as is
composition and thus avoids such ambiguities as suggested where interstices are floored with mud.
dolomite calcilutite, and it does not specify that It may record mixing by burrowers or incomplete
the mud is of clastic origin. The significance of winnowing or partial leaching of mud, as is sugmudstone, aside from the implication of calm gested by patchily distributed mud.
Crainsdme.-Mud-free carbonate rocks, which
water, is the apparent inhibition of grain-producing organisms.
are necessarily grain-supported, are termed
Wackalwee,--Mud-supported carbonate rocks grainstone (PI. VII). Grainstonesare not all of the
containing more than 10 per cent grains (10 per same hydraulic significance. Some are current
cent grain-bulk) are termed wackestone (PI. V).
laid; some are the product of mud being bypassed
T h e name has much against it, but it does have while locally produced grains accumulate, or of
the advantage of calling t.o mind a mixture of mud mud bcing winnowed from previously deposited
and grains simikr to that seen in some sandstones, muddy sediment; and some, conceivably, are the
and it is less awkward than expressions such as product of locally produced grains accumulating
calcarenitic calcilutite or calearenitic limestone. too rapidly to be contaminated by mud. ComPacksto~c.-Grain-support4 m u d d y carbon- monly, origin cannot be definitely known from

( X 5, transmitted
o. Csinoid lime ~ k s 6 m coverlying lens of lime

unpolarid Iiht)
mudstone. Overly close packing (75 per C?!,

pin-bulk)

that g r a i n - s u ~ w
t a s acquired during compaction. Mission Canyon fomatlon, Mlrumpplan, Shell
Richey NP-1, Richey ~eld,Dawson County, Montana, NP 1-7312.
b. Hydroman? lime packstom. Floored interstices ( F ) indimte that gmin-suppart is original and suggests
in6ltering. Permian Woifcamp, SheU Hilburn No. 1, Townsend field, Lea Cwnty, New Mexico1 SH-1-10,556.
c. Coated-grain lime p a c k s l m . Abraded nuclei of grains indicates turbulent water; presence of mud indicates

sugwts

quiet water; patchy distribution of mud suggests that burrowers mixed originally interbedded sand and mud
together. Toronto limestone, PennsylvanianVirgil, Kansas, Bed A.

Home

Home

Home
CLASSIFICATION ACCORDING TO DEPOSITIONAL TEXTURE
the single sample that is being classified. The
class name thus denotes merely the absence of
mud, and, of course, the corollary that the grains
are supported by each other. If the grains were
not self-supporting, the tormer presence of nowrecrvstallized mud would be indicated. Subdivjding grainstone, so as to include more evidence in
class names, is a temptation. Many possible subdivisions suggest themselves. Grain size is, of
course, one possibility. Sorting is another, for the
welI-sorted grainstones are scarce enough to be
noted. Wear is another. Names such as calcrditc,
sortedstonc, and wornstma are useful in studying
some suites of rocks, but, at present, none of Lhese
properties seems of wide enough application to
warrant being incorporated in a general classification.
Boundslone.-Carbonate rocks showing signs of
being bound during deposition are termed
boundstone (Pl. IV). The signs of binding are
specific, and they occur within the sample being
classified. Except f& that, the concept of boundstone plays rht same role as the recioid carbon te,
biohermal carbonate, constructed hiogenic car-

bonate, klintite, and biolithite of other classifica-

tions.
Crys.!aZZiw carbona~a.-Inasmuch as this classification is concerned with depositional environments, the rncks retaining toolittle of their depositional texture to be classified must be set aside.
Such rocks are here termed crystalline carbonates
(specifically, crystalline dolomite, crystalline limestone). Although depositional texture is lacking,
relics or ghosts of grains commonly d o w dassification according to origin of grains; for example,
crinoid-bearing crystalline dolomite.
REFERENrnS

B+amp,
R. A., and Po~vem,R. W., 1958, Classifimt~onof Arabian carbonate rocks: Geol. Soc. America
Bull., v. 69,no. 10, p. 1305-1318.
Folk, R. L., 1959, Practical w t r o p p h i c classification
of limestones: Am. h c . Petroleum Geologists Bull.,
v. 43, no. 1, p. 1-38.
Grabau, A. W., 1904, On the classi6cation of wdimentary rocks: Am. Genlogist, v. 33, p. 22S247.
Murray, R. C., 1960, Origin of prosity in mrhnnte
rocks: Jour. Sd. Petrology, v. 30, no. 1, p. 59-84.
Petf john, F. P., 1957,Sedimentary rocks, 2d cd.: New
York, lhrper and Bros.

EWLANATXON
OF PLATE
VII
(X5, transmitted unpolarized light)
a. Oijlith lime gaimhne (elite). Note the sheltering done by the coatecl brachiopd shell (S)near the center
~ a i n are
s smaller and more numerous above the sheH than below the shell; compare Plate ID-6. Wapanucka
limestone, Pennsylvanian Marrow, Bromide,Oklahoma, W-23.
b. Pusulinid lime graksdm, dm=-cemented. Seemingly lwse acking is due to fusulinids being unusually
elongate. Uniform orientation (long axis perprndicular to page) and Ermlcage indicate that the grains are currentlaid. Back-reef limestone o i Capitan reef, Permian Gwdnlupe, Pinery Canyon, Guadalupe Mountains, Texas,
PN-&A.
E. Lithiclast lime gmimtoac. Note the sand-bred intentices I F ) ; compare PIate IIM. El Paso limestone,
Ordovician, Franklin Mountains, Texas, AGE.