Middleton G.V. (Ed.) Encyclopedia of Sediments and Sedimentary Rocks
Подождите немного. Документ загружается.
STATISTICAL ANALYSIS Or SEDIMENTS AND SEDIMENTARY ROCKS 685
Figure S38 Subset of geological subjects, methods of data analysis,
iind lime in two-year periods lupperl; horizontal slice fo show time of
enlry of compuler applications in sedimentoloyy IKrumbein, 1969,
p, 254). see section on: William Chrislian Krumbein.
Applicatioti
of
QuittUitatiyc Methods to the Study
of
the Structure
and Hislory of
Rock.\
by Henry C. Sorby in 1908, which
presented results based on a mass of data accumulated
from stream experiments; W. Mackie described rounding of
particles in a 1897 paper. Ontltc
Lttu-.sthat Goycrnthe Rotttu/ing
ojFiirtiilt'sofSaud'. and .lohann A. Udden published his paper,
Meehitnicu! Cotupositioti of Clastic
Scdhnetits,
in 1914 (see
Merriam.
1981, Tor references). This early work on clastic
sediments was followed hy the contributions oi C.K. Wcnt-
worth.
William C. Krumbcin (arguably the most influential of
the numerical scdimentologists; see article in this volume).
A.B.
Vistelius. and J.C. Griffiths, who applied regression
analysis, time-series analysis, factor analysis, discriminant
analysis, and Markov chain analysis to scdimentologieal
problems (Figure S38).
Bill Krumbein and Francis Pettijohn's 19.18 textbook. Man-
ual
of Seditiwuiarv
Petrogntphy,
uas a primer on the analysis of
sediments and included two chapters on statistical methods, a
forerunner oi Krumbcin's later work. Robert
F""olk's
manual.
PctrologvofScditiicnlarv
Rocks
(1961) contains basic statistical
teehniques for the analysis of clastic rocks; Robert Miller and
.lames Kahn published a book on a general survey of statistical
methods in 1962; the Krumbein and Franklin Graybill book
on statistical models in 1965 was an introduction to the subject
witli emphasis on sediments; but .lohn CVdric Grifliths' Scictt-
tipc
Method
iti
Attalysi.s
of
Sediments
published in 1967 was at
the time the ultimate publieation on statistics applied to
sediments and sedimentary rocks. Griffiths' book was followed
by Frits Agtcrbcrg's tome on geomathematics (1974) and
Walther Schwarzacher's book on sedimentation models
(1975). both for the advanced student and professional.
Carbonate sediments have been treated differently than
clastic sediments because they are not amenable to disaggrega-
tion and particle analysis, .lohn Imbrie and Edward Purdy
were among the first to apply factor analysis to carbonates in
(heir 1962 paper.
Cla.ssilieationol
Moilcrti Baliaiitiaii
Carbonate
Sedinii'nt,s.
followed by lmbrie's Factor and
Vector Analy.sis
Pro-
gramsfor Analyzitig
Geological Data in 1963 and Purdy's
Recent
Calcium
Carhonaie Facies
of
ihe
Great
Bahamas
Bank in the
same year (see Merriam.
1981.
for rcferenees).
Statisticians, too. have applied their tools to problems in
the earth sciences since the turn of the 20th Century. Karl
Pearson studied the variation in homogeneous material;
A.N.
Kolmogorov used the lognormal law of distribution of
particle sizes to analyze crushed rock; Churchill Hisenhart was
interested in significance tests on lithological variation;
C.J.C. Ewing discussed operator variation in heavy-mineral
separation; S.S. Wilks explored statistical inference;
K.V. Mardia worked with directional data: G.S. Watson
contributed to statistics of orientation data; D.M. Hawkins
wrote on robust statistics: J.C. Gower was concerned with
sediment classification: M,l\ Dacey applied Markov chain
analysis; and J. Tukcy suggested that methods of confirmatory
data analysis, such as the jackknife approach, were logical and
useful for geological problems.
ln the early I96()s a French engineer, Georges Matheron.
introduced to the earth sciences a new branch of statistics
known as
geostatisiic/ues
(geostatistics), in his papers Traitede
geostatisiiquc
appliquve.
parts I (1962) and 2 (1963). For an
introduction in English, see Matheron (1963). The new school
is based on ihe idea oi regionalized variables or partially
dependent observations located in two- or three-dimensional
space,
Kriging is the name applied for methods of interpolat-
ing spatial geological daia that makes use of geostatistical
concepts.
Daniel F. Merriam
Bibliography
Agterberg.
F.P., 1974.
Geomathemuiies'.
Elsevier Scientific.
Cubilt. J.M., and Henley, S. (eds.). 1978. SUiiistiealAnalysisinGeohgy.
Bfnchmiirk Papers in Geology .^7, Stroudsburg, PA: Dowden,
Huicliinson, and Ross.
Davis.
J.C. 1986. Statistiesand
Data Analysis in
Geology.
2nd cdn. Ji>lin
Wiley & Sons.
Folk, R.L., 1961. Petrology of
.Sedimentary
Roek.s.
.Auslin, TX:
Hemphill's.
Gill.
n.. and Merriam. D.F. (eds.). 1979.
GeomathematiealandPetro-
physieal Shulies in
Sedimenlology.
Pergamon Press.
Griffiths. J.C.. 1967. Scientijie Method in
Analysi.s
oj Sediments.
McGi-:iw-HilL
Krumbcin.
W.C, 1969. The cotiipnler in geological perspective. In
McrriatTi.
D,F. (ed.).
Computer .-ipplicalions
in
the
luirth
S(ienees.
Plenttm Press, pp. 251-275.
Krumbcin.
W.C. and Pcltijohn. F.J.. 1938. Mamtal ot
Sedimentary
Petrography.
Applelon-Ceniiiry.
Krumbein,
W.C, and Gravbill. F.A., 1965. An
Introduction
lo Stalisti-
eal Models
in
Geology.
McGrau'-Hill.
Mars.il.
D., 1987.
Statistics for
Geoscieniists.
Pergamon Press.
Miitheron,
G., 1963. Prineiples ofgeostatislics.
KeonomicGeology.
58:
1246 1266.
Merriam.
D.F. (cd.). 1970,
Geo.stali.sticf,
aColliicimnni. Plenum Press.
Merriam,
DA. (cd.), 1972.
Malhemutical .Models
of
Sedimentary
Pro-
cesses.
Plcnimi Press.
686
STORM DEPOSITS
Merriam, D.F. (ed.), 1976. Qiiantilative
Teehnicjtie.s
lor the Analyst's o/
Sediments. Pergatnon Press.
Merriam. D.F. (ed.), 1981. Down-to-eitrth slalislics: solitUoiis looking
for geological problems. Syraeu.se L'niyersity Geological Conttilnt-
tion.s.
8: 97pp.
Milter. R.L., and Kalin, J.S., 1962. Statistical
Analysi,\
in
the
Geological
Seienees. John Wiley & Sons.
Rock, N.M.S., 19KK. \iitnerieal Geology. Leeliire Noles in Eiirlli
Sciences 18. Springei-Verlag.
Sorby, H.C. 1908. On llie appliealion of quanlilative methods to the
study of the structure and hislory of rocks. Quarterly Jourtial of the
GetilogiealSoeietyoJ Limdon. 44 (2): 171-233.
Sehuar7aeher, W., 1975. Sedimentation Models and Qtiantitatiye
Stratigraphy. Elsevier Scienlilic.
Cross-reference
Sedimenlologisls: William Clirisliaii Krnmbein
STORM DEPOSITS
Large storms have wide-ranging etlecls in a spectrum of
environments from outer shelves to coastal plains. The most
common deposits of storms in the rock record are sandstone
beds deposited under powerful uear-bottom vvaler motions
produced by waves and currctits. Early studies of the effects of
storms on modern shelves (e.g,. Hayes. 1967) sparked interest
in finding analogues in the rock record. This resulted in ihe
de\clopment of storm facies models, with partieular emphasis
on hummocky cross-stratification (HCS) (Walker. 1984). The
origin and dynamies of the depositing flows were considered
controversial (see Swift clal.,
198.1:
Walker, 1984) and initial
models relied on eross-shelf transport by turbidity currents
(Hamblin and Walker, 1979). Actualistic models of storm
dynamics subsequently constructed by oceanographers and
marine geologists largely focused on nearly shore-parallel
geostrophie currents. Duke (1990) suggested that aneient
siiallow-marine storm deposits (tempestites) were deposited
by geostrophie combined Rows and subsequent work indicates
that combined flows are in fact recorded in ancient sandy
tempestites (e.g., Midtgaard. 1996). However, many ancient
storm-deposited sandstone beds show greater thickness and
cross-shelf distribution than their modern counterparts ([_cckie
and Krystinik. 1989; Myrow and Southard, 1996). In addition,
the time between successive beds in many deposits in the rock
record may be many orders of magnitude too large to explain
as a simple series ol" storm events.
The stratilication preserved in tempestites. particularly
HCS.
has been studied e.\perimentally (Arnott and Southard.
1990) and through the analysis of grain fabric (Chcel. 1991).
Such studies, and those ol' modern shelves (e.g.. Amos elal.,
1996).
confirm that HCS fortns as a result of complex
oscillatory flow or combined flow (vva\es and currents). A
wide range of vertical stratification sequences is possible within
individual storm-deposited beds (Myrow and Southard, 1991).
Three processes operate during storms to produce the
wide range of tempestites encountered in the sedimentary
record: geostrophie eurrents, wave oscillations, and excess-
weight forces (density-induced flow) (Myrow and Southard,
1996).
The relative importance of the.se three processes may
change during individual events and between events both
within and between basins. Thus eonsiderable variability is
possible in the character of individual beds, although
most beds display graded bedding and evidence for flow
deceleration.
Geostrophie flows are near-shore-parallel flows that result
from Coriolis deflection of water masses that are driven
offshore in response to pressure gradients associated with
coastal setup. Combined flows result from the superposition of
storm-generated waves with these flows. Because these
geostrophie flows are oriented slightly obliquely offshore.
instantaneous bed shear stresses in the eombined-ilow bound-
ary layer show asymmetry in the ofTshore direction (Duke,
1990).
Sediment transport results from the nonlineai- addition
of the two eomponenis of flow iti the boundary layer, with the
net result that wave oscillations are stronger in the offshore
direction and thus preferentially move sediment in that
direetion. Whether this asymmetry in transport can account
for significant cross-shelf transport is unclear. Regardless of
transport mechanisms, most icmpeslile successions show
patterns of decreased bed thickness and evidenee for less
powerful currents from onshore to offshore. An exeeption to
this rule occurs in areas of nearshore sediment bypass (Myrow,
1992).
In many cases, ancient deposits contain evidenee for
powerful offshore-directed flow, which is at odds with
deposition under combined flows. Such flow is possibly
driven, at least in part, by excess-weight forces. These forces
lesult from high suspended sediment concentrations (SSCs)
near the bed. High SSCs can result from resuspension of
shoreline sand by storm processes. A number of studies
indicate that very high SSCs (up to 4 g/L) are produced on
shorel~aces and inner shelves and that this leads to powerful
seaward-directed sediment-laden flows (e.g., Wriglil et ai.
1994).
Plumes assoeiated with fluvial diseharge events may
also produce hyperpyenal (bottom-hugging) Hows, and such
flows are termed oeeanic Hoods (Wheatcroft. 2000). These
could be pre-ignited with regard to autosuspension. but this is
unnecessary for significant eross-shelf transport beeause
storm waves can provide the extra turbulence needed to
maintain flow. Results of both oeeanographie (e.g., Cac-
ehionc
etcd.,
1999) and modeling experiments of the northern
California coast indicate that gravity driven underflows
associated with ocean floods are an important factor for
cross-shelf transport of fine-grained sediment. Such flows
have high SSCs (up to 2,5 g/L) even at mid-shelf depths (50
m
to
70
m).
The sequence-slraiigiaphic cotitext of tempestites may
have a strong control on their abundance, nature, and
stratigraphic distribution. Transgressive to early highstand
conditions may have higher accommodation space and thus
greater potential for Coriolis effeet to produce geostrophie
flow, whereas late highstand conditions promote greater
bottom frietion and more offshore-directed flou (McKlc.
1994).
Late highstand to early lowstand systems traets of
third- or higher order sequenees tend to have abundant
tempestites due to tnobilization of coastal sand during relative
sea level fall (Einsele, 1996), Sediment supply by rivers, or in
situ earbonate production, has a first-order eontrol on
tempestite thickness and abundance. More work is needed to
better constrain the reeurrcnce intervals of ancient deposits.
Comparison of these data to those from the Modern will help
STROMATACTIS
687
to establish ihe degree to which aetualistic models apply to the
aneient.
Paul Myrow
Bibliography
Amos. C.L., Li. M.Z.. and Choang, K.-S., 1996. Storm-generated,
hummocky siralificiitioti oti the outer-Scot inn shelf. Geo-Marine
Letters. 16: 8? 94.
Ariioll.
R.W., and Suuihard, J.B., I'M). Exploratory flow-duel
cxpcrjiiienis on cotiibincd-tlow bed conflgurations, and sotne
impliualioiis lor iiiicrpreiing storm-event stratification. Journal o)
Sedimentary
Petrology.
60: 211-219.
Cacehione, D.A., Wiberg, P.L.. Lynch,
J.,
Irish,
.1.,
and Traykovski, P..
1999.
Estimates oi' suspendL'd-sedrmenl llux and bcdform aelivily
on the inner portion ot" ihe liel conlinenlal shelT.
Mctrine Geologw
154: S3 47
(.heel. R.J.. 1991. Grain f;ihric in hummoeky eross-slratilicd sioriii
beds:
gcnelie impliealions. Jottrmd o/Seilimenlary
Petrology.
61:
102 IK).
Duke,
W.L., 1990. Geoslrophiccireitlalion or shallow marine lurhidily
currents'.' ihc dilemtiia ol" paleollow patterns in siorin-inllueneed
prograding shoreline systems. Journal ot
Sedimenttiry
Petrology.
60:
87U 88.1.
Einsele. G., 1996. Eveni deposits: the role of setiimcnl supply and
relative se;i-level changes -overview.
Sedimentary
Geology,
104:
11 .17.
liamblin,
A.P.. and Walker. R.G., 1979. Storm-dominated shallow
marine deposits: the Femie-Kootenay (Jurassic) transition, soulh-
ern rocky mountains.
CimadianJournalofEurthSeienees.
16: \(f!y
l(i90.
Leekic, D.A.. and Kr>»itinik, L.F., 1989. Is there evidence lor
geoslrophic currents preserved in the sedimentary reeord of inner
to middle-sheir deposits?
J<mrnal
of
Sedimentary
Petrology.
59:
862-870.
McKie.
T.. 1994. Geostrophie versus IViction-dominated slorm ilow:
paleoctirrenl evidenee from Ihe Lale Permian Brotherton Ibrma-
lion,
lingland.
Sedimentary
Geology.
93: 73-84.
Midlgaaid,
H., 1996. inner-shelf lo lower shoreface Htimmockv
sandstone bodies wiih evidenee for geostrophic-intluenced com-
bined flow, lower Cretaceous, west Greenland. Journalof Sedinten-
lary
Researeh.
66;
.14.1
.153.
Myrow, P.M., 1992. By puss-zone lempesiiie facies model and
pro.\imality trends for an aticient muddy shoreline and shelf. Joto-
ualo!
Sedimentary
Petrology.
62: 99- i 15.
Myrow, P.M., inui Southard. J.B..
I99I.
Combined-How model for
vertical stratification sequences in shallow marine storm-deposited
beds.
Journal ot Sedimentary
Petnilogy.
61:
302-210.
Myrow. P.M., and Southard, J.B., 1996. Tempeslite deposition.
Jour-
tialof
Sedimetttary
Researeh.
66: K75 8K7.
Swift. D.J.P.. Figueiredo, A.G. Jr, Freeland, G.L.. and Oertel, G.F.,
1981.
Hiimnux:ky cross-stratiiieation and megaripples: a geological
double standatdV
JournaloJSedimentary
I'elrolimy,
47: 1242-1260.
Walker. R.Ci., I9S4. Shelf and shallow marine sands. In Walker. R.G.
(ed.),
Faeies Models, 2nd edn. Toronto: Geoscience Canada.
Reprint Series I, pp. 141 170.
Wheatcroi't. R.A., 2000. Oeeanic flood seditnentation: a new perspec-
tive.
Contitienlal
Shelf
Re.search.
20: 2059 2066.
Wt-ight, L.D.. Xu, J.P., iind Madsen. O.S.. 1994. Aeross-slicif heiitluc
transports on the inner shelf of the Middle Atlantic Bight during
the "Halloween storm" ol' 1991. Marine
Geology.
118: 61-77.
Cross-references
Bedding and Internal Structures
Coastal Sedimentary Faeies
Faeies Models
(iutter and Gutter Casts
Hutiimock_\ and Swaley Cross-Stratificaiion
Paleocitrrent Atialysis
Sands, Citavels, and their Lithitied Equivalents
STROMATACTIS
Strornataciis is a spar body common in Paleozoic carbonate
nutd-nioiiiids but much less common in Mesozoic ones.
Although it has been regarded originally as a fossil (Dtipont,,
1881), and later as a permineralized unfossilizable organism or
as recrystallized patehes of the host carbonate, most carbonate
workers now agree that it is the result of centripetal
cementation in a eavity system, regardless of the origin of
the cavity
(e.g..
Bathurst. 1982; Lees and Miller. ^1985;
Bourque and Boulvain, 1993: several papers in Monty
vial..
1995). Because it is a diagenetie strttcture, and not a fossil,
stromatactis should not be capitalized, nor italicized. There is
no consensus on the definition of stromatactis and the origin of
the cavities into whieh it developed.
Some workers define stromatactis in an all-inclusive manner
and eonsider that stromatactis is all spar bodies having a flat
base and digitate top resulting from the cement filling of
cavities, no tnatter if the spar bodies, and hence the eavities.
are connected or isolated in the host rock
(e.g.,
Monty, in
Figure S39 Sfromalactis <ind reUited structure trom Devonian
(Frasnian) mud-mounds in the Belgi<3n Ardennes. (A) Typic.il
stmm,i!<ittis limestone showing tross-section in a stromatactis
network. Although the spar bodies look as individuals, they are
interionnected in the third dimension. Les Maquettes Quarry. Width of
picture is 35cm. IB) Cross-section of filling of layered sediment in the
original cavity network that has inhii>ited stromatactis development;
Croisette Quarry. This kind of structure has been misidentilied
occasionally as stromatolite. Width of picture is 9
cm.
Arrows indicate
sedimentary polarity.
688
STROMATOLITES
Monty el al.. 1995). Others consider that too wide-open a
definition of stroniataetis has led to a misuse of the term, and
that stromatactis is a specific fabric that should he more
restrictively defined (e.g.. Bathurst, 19t^2, 1998). Based on
original material deseribed by Dupont (1881) in the Belgian
Ardennes, and following Bathurst (1982) who drew attention
to the occurrence of stromatactis in reticulate swarms.
Bourque and Boulvain
(1993,
p.608) proposed to restrict fhe
term stromatactis to "a spar network, whose elements have flat
to undulose smooth lower surf"aees and digitate upper surfaees.
made up principally of isopachous crusts of centripetal eement
and embedded in frnely erystalline limestone".
Stressing the network nature of stromatactis brings into
relief that stromatactis is one of a two end-members
continuum of a specific cavity network filling controlled by
two antagonistic processes, internal sedimentation versus
centripetal marine cementation. When the rate of internal
sedimentation is lower than the rate of marine eementation,
centripetal cementation of the eavity network forms stro-
mataetis (Figure S39(A)); when the rate of internal sedimen-
tation is higher than the rate of marine cementation,
layered sediment inhibits stromataetis development and fills
entirely the cavity network (Figure S39(B)) (see Neuweilcr
elal.^ 2001, for discussion on the processes involved and their
causes).
Several scenarios were proposed to account for the
formation of the cavity system, among which: (1) winnowing
of noneohesive mud sandwiehed between layers of early
cemented sediment crusts (Bathurst, 1982) or lithified micro-
bial mats (Pratt. 1982) on the sea-floor; (2) collapse of muddy
sediment bridged by eohesive mierobiai mats on the sea-floor
{Lees and Miller, //; Monty elal.. 1995); (3) fiuid escape after
the collapse of decayed microbially bound sediment in the
shallow burial environment (Monty, in Monty er ai, 1995);
and (4) collapse of unccmenicd material in an early cementing
sponge/mud tnixture in the shallow subsurface environment
(Bourque and Boulvain, 1993).
Pier re-And re Bourque
Bibliography
Biitiiursi. R.C'.G., 1998. Tlie world's mosi spectactilar carbonate mud
tnoLinds (Middle Devonian. Algerian Sahani) Discussion. Jour-
nalof Sedimentary Re.search. 68: IUf>l.
Bathurst, R.G.C, 19K2. Genesis of stromalactis cavities between
stibtnarini; crtisis in Palaeozoic carbonate nuid buildups, JottmaloJ
the
Geological
Society of London. 139: 165 181.
Bourque, P.-A., and Bouivaiti, I".,
I9')3.
A model for the origin and
pctrogcncsis of the red siromatactis limestone of Paleozoic
carbonate mounds. Jinirmdot Sediiueniury
Pclrology,
63: 607-619.
Dupont, E., 1881. Siir j'origine des calcaiics devoniens de la Belgiquc.
Bulleliii de
I'.ieademie
Royale des Seienees de Belgiipie. Third Series,
2:
264-280.
Lees,
A., and Miller. J., 1985. Facies variation in Waulsortian
buildups. Part 2: Mid-i;)inaiuian btiildups from Furope and North
America.
Geological
Journal, 20: 159 ISO.
Monty, C.L.V., Bosence. D.W.J., Bridges. P.M.. atid Piatt, B.R. (eds.),
!995.
Carhonaie Mud-Mounds. Their
Origin
and Lyohition. Inierna-
tional Association of Sedimentologists, Special Ptiblication, 2^.
Netiweiler, F., Bourque, P.-A.. and Boulvain. F..
20(11.
Why
is stromatactis so rare in Mesozoic carbonaie tnud mounds.'
TerraNoya, 13: 333 337.
Pratt, B.R., !9S2. Siromatolitic fntmcwork of carbonate mud-tnounds.
Journal of Sedimentar\ Petrology. 52: 1203-1227.
Cross-references
Bacleriii in Sediments
Carbonate Mud-Mounds
Cemenis and Cementalion
Diagetiesis
Diagenetic Structm^es
STROMATOLITES
Stromatolites are critical geological feaitircs as they are the
only biogenic sedimentary structures that have an age range
extending fully fYom the early Archean to the Recent, thereby
demonstrating a ftmdanientally unifbrmitarian phcnotnenon.
First deseribed from the Cambrian of New York State by J. H.
Steel in 1825, and already widely recogni/ed by the time the
term "stromatolite"" was coined by E. Kalkowsky in 1908 lor
Triassie examples, it was C. D. Walcoit in 1914 who noted
their predominance in Precambrian strata and first dedticed a
microbial origin (Riding, 1999). M. Black was the first to
document, in 1933, the stromatolite-forming activities of living
microbial mats, on tropical tidal flats of Andros Island,
Bahamas. The discoveries of modern analogs of columnar
forms in Shark Bay, Western Atistralia in 1961 (Figure S40(A),
(B)) and at Lee Stocking Island. Bahamas in 1986 are
milestones in the earth sciences (see Golubic, 1991). 3.5Ga
stromatolites from Western Australia represent the oldest
physical evidence of biological activities, most certainly
photosynthctic, on Earth. In turn, stromatolites signpost the
seareh for evidence of life on other planets.
Definition
Even though the evidence for prior mierobiai aetivity is
tistially circumstantial, fhe definition of stromatolite has
evolved as the composition and metabolism of benthie
mierobiai mats and biofilms in various settings and the wide
variety of resulting struetures and their evolutionary context
have beeome better understood, Mueilagenous mats and
biofilms arc complex communities dominated by bacteria
and, in the photic zone, by eyanobaeteria, although other
microorganisms may contribute. While shallow-marine and
lacustrine calcareous stromatolites are the most commonK
recognized (Riding, 2()()()), other deposits that can be embraced
by the term include many tufas or travertines, phosphate
erusts.
siliea sinters, and manganese nodules, for which a
microbial control or influenee ean be determined or inferred.
Thus.
Riding (1999; see also Krumbein, 1983) advoeated a
broad approaeh: a stroma(olitc is simply a laminated henihic
tnicrohial deposit. Although there is no constraint of seale,
strotnatolites are typieally centimeter- to meter-sized, and
laminae microtiicters to millimeters in thiekness. Spheroidal
counterparts that rolled around during formation arc oncoids
(or oncolites). Non-laminated struetures similar to stromato-
lites are thrombolites (Pratt, 1995).
Morphology and microstructure
Stromatolites include beds and domes with planar, wavy, or
convex latnination, and rccfai masses composed of eonieal or
STROMATOIITFS 689
Figure S40 Meter-sized stromatolites.
lAl
inlertidal
flal,
Holocetie, Shark Bay, Weslcrn Auslrali^i. (Photograph courtesy
of R. A.
Wood).
iBl
("o.islal
t'xposLire. Pelil Uirdin Formation (Upper Cambrian), norlhweslern Newfoundland.
Figure
S41
ukrups and polished 5ur1at
uiul,
C.oUeni.-]
form. Helena Formalion
l,
soulhwestern Alberia.
(B)
Cunicil columns, Conophyton form.
1-5
member, Atar Group (Neoproterozoic), Mauritania.
(Cj Branchinjj columns. Ciss Fjord Formation (Upper Cumbrian), Devon Island, (Canadian Arctit.
(D)
Branching columns,
Bdlccilid
form.
1-5
member, Atar Group. (E) Wavy lamination, Lockport Formalion (Lower Silurian}, northern Ohio.
(F)
Laminae
(.oating
Porites
coral rubhk'.
Forercet'iHolocene), northern
convcxly laminated colutnnar and briinching forms (sec
examples in Walter, 1976) (Figure S41(A) (E)). In addition,
Phanerozoic marine stromatolites are commonly intergrown
with sessile invertebrate skclelons (Figure S4I(F)). Beginning
in the 19th century, many fossil stromatolites were given
linnean binomial names. This practiee has declined, although a
number of recurring "form-genera" arc still commonly used as
descriptors.
690
STYLOIITFS
Figure S42 Stromatolite microslructure, photomicrographs of ihin
sections. (A) Laminae of mitrite and clotted micrile, separated by
finely crystalline cak ile cemenl. 1-5 member, Atar Group
(Neoprolero7oic}, Mauritania. (B) Laminae of mi{Tite and
recrystallized micrite. Peg Formation (PaleoproterozoicI, Northwest
Territiories.
Lamination consists of rhythmic combinations of micro-
bially triggered and sometimes inorganie preeipitates, permi-
neraiized microbes, physieally trapped sedimentary particles,
and primary pore space that is subsequently cemented (see
examples in Riding and Av^'ramik. 2000) (Figure S42{A). (B)).
Calcareous stromatolites are especially vulnerable to dia-
genetie modification, such as recrystallization, dolomitization.
and silieitieation. The last process may fortuitously preserve
mierobiai remains before substantial degradation (e.g.. Knoll
et al.. 1991). In general, some degree of synsedinientary
lithifieation is a prerequisite to preservation.
Evolution
Strotnatolites are conspicuous in younger Archean and
Proterozoic shallow-tnarine earbonate successions (e.g.,
Walter (7i//., 1992; Hofmann. 2000). A Precambrian biostrati-
graphy based on rnorphology and mierostructure has been
attempted but has not proved eonvincing. Still common in the
Cambro Ordovician. they also oecur sporadically in younger
strata. The apparent decline in abundance of marine stroma-
tolites has been attributed to competition for substrate spaee
by enerusting invertebrates and ealeareous algae, the effects of
grazing and burrowing animals, chemieal ehanges in sea-
water, and evolving sedimentologieal conditions (Pratt, 1982;
Grotzinger and Knoll, 1999). Evolutionary aspeets of stroma-
tolites from other settings are uueertain ai present.
Frontiers
Microbial biofilms and mats are complex systems, constantly
in a state of flux between growth and decay, and mueh remains
to be learned about microbially induced diagenesis and the
proeess and temporal significance oi lamina formation and
overall growth rales in both aneient and moderti examples of
all types. A thorough assessmetit of stromatolitie structures of
non-earbonate mineralogy and from non-marine settings, such
as hot springs, is slill in progress. Lastly, if present, the remains
of past microbial activities on other planets—astrobiology -
may come from stromalolitic rocks, the cue for continued
research on Earth's own reeord.
Brian R. Pratt
Bibliography
Golubic, S., 1991. Modern stromiitolilcs; a rcviL'w. In Riding, R.,
iitid Awiamik. S.M. (eds.). Calcareous Algae and Stromatoliles.
Springer-Verlag, pp. .^141 561.
Grolzingcr, J.P.. and Knoll. A.H., 1999. Strotnalolitcs in Ptecatnbrian
carbonates: cvoltitionary mileposts or environmental dipsticks.'
Anntiat Reyiens of Earth and Planetary Sciences. 27:
? 13
358.
Ilolhiann. H.J., 2(K)(). Archean stromatoiiles as microhial ;irchives. In
Riding. R., and Awraniik, S.M.
(eds.},
Mierohial Seditnenls.
Springer-Verlag, pp. 315 327.
Knoll, A,H.. Suetl, K., and Mark. ,1,. 1991. Paieobioiogy of a
Neoproicrozoic tidal tlat/lagoon complex: the Draken Conglo-
merate formation. Journal of Paleontology, 65: 531 570.
Kiitmbein, VV.E., 1983. Stromatolites -ihe challenge of a term in spaee
and time.
Preectmhrian
Research. 20:
493-531.
Pritti, B.R., 1982. Strotiialolite decline—a reconsideration. Geologw
10:
512-515.
Priitt, B.R., 1995. The origin, biota and evolution of deep-water mud-
nionnds. In Monty. C.L.V., Bosence, D.. Bridges. P.H.. and Pratt,
B.R. (eds.I, Carhonaie Mud-numnds: Their Origin and Fvo/ittion.
International Association of Sedimcntoloaists, Special Publieation,
23.
pp. 49-123.
Riding, R,. 1999, The term stromalolite: towards an essential
ddinition. Lethaia. 32: 321-330.
Riding, R,. 2()()(J. Microbial earbonalcs: the geological reeord of
caicifted baeterlal-alcal mats and biofttms. Sedimenlology. 47 (I):
179-214.
Riding, R., and Auramik. S.M. (eds.), 2()(.H). Microhicd Sediments.
Springer-Verlag.
Walter, M.R.
(ed.},
1976. Stromatolites. Developments in Sedimentol-
ogy 2(1. Else\ier.
Walter, M.R., Grotzinger. J.P., and
Schopf.
J.W.. 1992. Prolerozoie
stromatolites. In
Schopf.
J.W.. and Klein, C.
(eds.}.
The
Prolerozoie
Biosphere: A Midtidisciplinary Study. Cambridge L!niversity Press,
pp.
253-260.
Cross-references
Algal and Bacterial Carbonate Sedimenis
Carbonate Mud-Monnds
Diagenesis
Mierobially Indttecd Sedimentary Structures
Neritic Carbonate Depositional Environments
Phosphorites
Reefs
Siromataclis
Tidal Flats
Ttifas and Traverline.s
STYLOLITES
Introduction
Stylolites are macroscopic serrate subplanar surfaces at whieh
mineral material has been removed by pressure dissoluiion.
resulting in a decrease in total volume of rock. Stylolites are
STYLCJLITfS
691
/ A\ Maximurn \ ,[• ,
\f^) compression * J;' \[ !•"•
(S)
Typical early stylolite
^ ' in grainstone
Typicai early stylolite
in mudstone
Insoluble material
—— Column
Relative motion
Typical late stylolite
in grainstone
Typical late stylolite
in mudstone
Figure S43 (A) Components of a stylolite. (Bl Typical profiles of
styloliles in different carbon.ile lithologies lafter Andrews and
K.iilsb.uk,1997, figure 15l,
made readily visible by the accumulation of residues of
relatively insoluble minerals, such as pyrite, oxides, clays,
and other silieates. If no sueh residue has aeeumulated, a
stylolile may be recognizable only by its truncation ofelements
of the host rock's preexisting fabrie.
Stylolites commonly consist of iiiterfitting striated eolumns
capped by subptanar laminae of insoluble material
(Figure S43(A)). The columns form parallel lo ma.ximuni
eompressive stress, as parcels of rock on each side of the
stylolile are pushed past each other. The eolumns are
commonly perpendicular to the general plane of the stylolite.
but they can be inclined relative to it. The eolumns are
t\pieally parallel to the direction of maximum compression
during their formation.
Stylolites take their name from these columns, in that
"stylolite" is derived from the Greek word "stylos" (for
column or pillar). In older works and some recent Eurasian
literature, individual columns are sometimes called "stylo-
lites",
but more recent publications generally use that term for
the entire surface delincd by the columns as a grotip. Thus a
hori/oiual stylolite in modern American usage consists of
many vertical stylolites as delined in some older or Eurasian
usage.
Stylolites are common in many limestones and dolostones,
can be common in evaporites and deeply buried sandstones.
and have been reported in igneous rocks (Golding and
ConoUy. 1962). Stylolites parallel to bedding of sedimentary
rocks are commonly thought to result from compression by
overburden, whereas stylolites perpendicular to bedding
(iransverse stylolites) are usually attributed to teetonie
compression. Trans\erse stylolites can predate or postdate
bedding-parallel styloiites in the same rock (Andrews and
Railsback, 1997),
Development
Development of a stylolite requires a solution into which
minerals ean dissolve and a pore network through which
dissolved solids can diffuse or advect from the developing
stylolite. Porosity also enhances stylolite development by
localizing stress on nonpore areas and thus increasing stress
there. Some models of stylolite development have therefore
suggested that bedding-parallel stylolites form in zones of prior
high porosity (Merino
(•/(;/..
1983). and that most if not all
transverse stylolites form along preexisting fractures.
Because its formation involves movement of two roek
\oiumes toward each other, development ofa stylolite requires
a previously lithilied host rtick. Compression of unlithilied
(A)
Figure S44 (A) Three-dimensional image of a styiolite (after Smith,
2()00.
figure 7e). (Bl One two-dimensional profile of stylolite shown in
A, Nunit)ers idt-nlify like positions in the two
grainy sediments commonly results in intergranular, rather
than whole-rock, compaction, and serrate intergranular
contacts can form. Such contacts are microscopic stytolites,
the lateral extent of which is limited to ihe area of contact
between two grains. Either intergranular compaction and
suturing or (more commonly) cementation ean lilhify a
sediment sui'liciently for a true macroscopic stylolite to form.
Stylolites vary from serrate to undulatory, but even most
undulatory surfaces ("dissolution seams") are serrate at
microscopic scales. Development ofa serrate surface requires
dissolution that is alternately localized on opposite sides of the
stylolite. In limestones, stylolites in grain-rieh roeks (grain-
stones and packstones) are generally more serrate than those in
waekestones and micstones. presumably because juxtaposition
of various grain types and o\' intergranular pores localizes
destruction of rock volume (Railsback. 1993) (Figure S43(B)).
Stylolites that formed early in a rock's history are commonly
more serrate than those formed later, presumably because
more extensive lithifieation reduces variability of rock fabries
and of solubility on opposite sides ofa stylolite (Andrews and
Railsback. 1997) (Figure S43(B)).
Problems and progress
Many studies o\' st\lolites have examined their host rocks as a
means to determine lithologies or rock characteristics favoring
formation of stylolites. Such studies are potentially flawed,
however, because the lithology in which the stylolite formed
has inevitably been destroyed during that stylolite"s growth.
The lithotogy (or lithologies) in which a stylolite is presently
observed mav in fact have characteristics that halted, rather
692
SUBMAKINb FANS AND CHANNELS
tlnin protnoted, stylolite development. Otherwise, the present
host lithology would have been destroyed by ftirther develop-
ment of the stylolite.
Traditional studies ol' stylolites have examined planar
sections cut parallel to the colutnns. Such studies have used
the resulting two-ditnenslonal images to classify stylolites
according to their general appearance {Park and Schot, 1968;
Buxton and Sibley, 1981), quantify their morphology
tising as many as five paratiielers (Railsback, 1993; Andrews
and Railsback. 1997), or measure their fractal dimension
(Drummond and Sexton. 1998). More recently, sequential
sectioning has been used to generate ihree-dimensional images
of stylolites (Smith. 2000), which detnonsirate that the sides of
columns form an anastomosing interconnected surface that
can serve as a pathway for fluids (Figure S44).
Significance
Stylolites are significant in several fields. Styloliles are
important petrologically because they modify rock fabrics
and because they generate dissolved solids that precipitate
elsewhere as cement. Stylolites are significant to stratigraphy
because weathering of stylolites generates apparent bedding in
many stratigraphic sections and because loss of material along
slylolites can cause significant stratigraphic shortening un-
related to erosion. Hydrologically, stylolites have been cited as
barriers to fluid flow and. in other settings, as conduits for fluid
flow. Stylolites are also commonly used as records of
compressive stress in tectonic studies, and development of
transverse styiolites can contribute to crustal shortening
parallel to the direction of their columns.
L. Bruce Railsback
Bibliography
Andrews. L.M., and Railsback, L.B., 1997. Controls on slylolitc
development: tnorphologic. lithologic, and temporiil cvidctice from
bedding-parallel and transverse slylolites from tlic US Appala-
eliians. Jmirnal of Geology. 105:
59-73.
Buxton. T.M.. and Sibley, D.F.. 1981. Pressure solution features in a
shallou buried limestone. Journal of Sedinienlarv Petrolology, 51:
19-26.
Drummond, C, and Sexton. D.. 1998. Fraetal strueturo of stylolites.
Journal of Sedimentary Researeh, 68: K- 10.
Golding. H.G., und Conolly, J.R., 1962. Stylolites in volcanic rocks.
Journaloj Sedintentary Petrology, 32: 5.'^4-53S.
Merino, E., Ortuleva. P.. and Striekholm, P., 1983. Generation of
evenly-spaeed pressure-solution seams during (late) diagenesis: a
kinetie iheory. Contrihiiiions lo Mineralogy and
Petrology.
82: 360
370.
Park. W.C. and Sehot. E.K., 1968. Styloliles: their nattuv and origin.
Journal of Sedimentary Petrology, 38: 175 191,
Railsbaek. L.B., 1993. Lithologic eonlrols on morphology of pressure-
dissolulion surfaces (.styloliles and dissolution seams) in Paleozoie
roeks from the Mideastern United Stales. Journal of Sedimentarv
Petrology. 63: 513-522.
Smith, J.V.,
200(1.
Three-dimensional morphology and connectivity
of styloliies hyperaetiviited during veining. Journal oj Strueiural
Geology, 22: 59-64.
Cross-references
tompaetion (Consolidation) of Sediments
Diagenesis
Diagenetic Structures
Pressure Solution
SUBMARINE FANS AND CHANNELS
Like alluvial fans Iq.v.]. submarine fans are cniie-like
accumulations of sediment, developed at a change of slope,
generally below a single major feeder channel (though a few
fans may have more than one channel, generally they are not
all active at the same time). Submarine fans and channels,
liowcvcr, are much larger and tnore cotnmon than subaerial
fans,
and have been constructed mainly by turbidity ctirrents
and olher sediment gravity flows (see Gravity-Driyen Ma.ss
Flons).
The coarser sediment is largely confined to the
distributary channels, and much of the fan may be composed
of silt or mud deposited by overflows on levees (Figure S45).
The need for a single major feeder channel to produce a fan
means that, although carbonate tnrbidites are common (see
Titrhtdites).
carbonate fans are rare. The large size of many
submarine fans often produces a geometry that is constrained
by preexisting topography (which may. in turn be related to
active tectonics) so that a typical fan shape is not well
developed. In addition, examples of thick siliciclastic turbidites
constructed at the base of subtnarine slopes but lacking fan
geometry have also been described (as submarine ramp
deposits: Heller and Dickinson,, 1985; Reading and Richards,
1994).
Modern submarine fans-channel systems were first de-
scribed from offshore California in the lafe 195()s and
recognition oi small ancient submarine fans in the strati-
graphic record followed shortly after (see historical review in
Pickering
etaf,
1989, pp. I 4). The facies model of Mufti and
Ricci Lucchi (1972; see figure S47) esfablished a pattern for
interpretation that persisted, wifh variafions, for the next 20
years.
Subsequent mapping of the ocean basins revealed thai
submarine fans include some of the largest geomorphic
feafures on fhe earfh's surface, reaching lengths of l,500kms
and widths of almost KOOOktns (Indtis fan: Kolla and Coumes,
1987;
McHargtie and Webb, I9S6; and see McHargue in
Weinier and Link, 1991). Such fans are much too large to be
recognizable in exposed ancient rocks, and the deposits would
be difficult to recognize even in subsurface basin studies. Large
submarine channels have been described from fhe deep sea,
with channel depths of more than a km, and widths of more
than 10 kills, flanked by levees that rise fens of" meters above
the surrounding surface and extend many kms away fVoni the
channel. Such channels are also difficulf to recognize in ancient
rock outcrops, though a few of comparable size have been
identified in stibsurface studies (Walker, 1992).
Large fans are generally composed oi finer sediment fhan
the "classical'" small fans described by Mufti and Ricci Lucchi.
and attempts have recently been made to describe alternative
fan models characteristic of finer grained lurbidite systems
(Reading and Richaids, 1994; Boutna and Stone. 2000; sec
Figure S51). It is claimed that large \olumes of sediment and
the presenee of tnud increases the "efficiency" of sand
transporf.
thereby allowing more sand fo be carried through
the midfan channels lo fhe disfal fan and basin plain. Bouma
(in Bouma and Stone. 2000) suggests that fine grained fans are
typical of passive margin fecfonic seftings. with long fluvial
tt"ansport leading to deltas, and from them to the head of the
main distribtitary channel. The sand in ihem is concentrated af
fhe base of the slope, and/or is transported efficiently ihrough
SUBMARINE FANS AND CHANNELS
693
Submarine
Canyon
Mass Movement Deposits
Figure S4.'i SurtVicc morphology ot .1 submarine fan. From Nurmark vt .il {^^')^, their Figure (> on
channels to distal sandy lobes. Coarse grained fans, in
confrasf.
are fypical of more tectonically active regions, with
only short fluvial feeder systetns, and with no prominent delta;
tnuch sand may be supplied into fhe feeder channel by
longshore
driff.
Much of the coarse sediment is deposited in
fhe midfan region. Though large fans may be dominated
overall by filler sediment, many do contain substantial
atncHints of sand, for example the Amazon fan (Piper and
Normark. 2001: see figure S48).
Not all submarine channels are directly related to submarine
fans:
fhe best known exeeption being the Northwest Atlantic
Mid-ocean channel, described by Hesse and Rakofsky (1992).
For an ancient large-scale canyon/channel compiex, not
obviously related fo a fan, see Brtihn and Walker (1997).
Ancient subtnarine fans include many large petroletim
reservoirs: they have become increasingly important as
exploration has moved into deeper water, offshore. For a
review of ihe special problems posed by exploration of such
fans see Weimer etal. (2000).
Canyons, channels, and depositional lobes
Canyons
The upper parts
ol"
submarine channels, particularly those that
incise the edge of the continental shelf are generally called
canyons. They have been known since the end of the
nineieenih century, ln most cases, they are clearly erosional
feattires. with some even eroded through hard igneous or
metatnorphic rocks. The eroding agcnl remained controversial
luitil the 195()s. Though the main agent oi erosion is now
thought to be turbidity currents, there is still debate about the
role played by other agents, including rivers (dtu"ing lower
stages of sea level), mass-wasting processes (including slides,
slumps, and debris flows), sand flows, earthquakes, and
structural or tectonic features. At lower sea levels (especially
dtning the last ice age) rivers flowed across the continental
plains (the shelves at higher sea-level stands), and entered what
are.
at higher stands, the heads of submarine canyons. Then,
because of the lower water temperatures and higher suspended
A. Eroslonal
B. Eroslonat/deposltlonal (mixed)
C. Deposftional
Figure S46 Frosion.il and Hepnsilional thannels, as inclitated by
seismiL profiles. From Clark and Pickering (199b, ihcir Figure
1
on
p.195, based on the earlier work ol' Normark).
sediment loads in the rives, the river flows plunged below the
surface of the sea to form "hyperpicnal Hows:" that is.
turbidity currents formed directly from river flows, without fhe
deposifion and resuspension fhaf is generally necessary foday
(Mulder and Syviiski. 1995).
During the greater part of the existence ofa canyon, there is
only temporary sforage of
sedimenf.
but filling of fhe canyon,
and therefore its preservation in the geologieal reeord, can
certainly take place. Many, but not all, modern submarine
canyons are related, if nof indeed directly connected with
major rivers. A good compiiaiion of the early lilcralurc is given
in Whiiaker (1976) and tlie more recent literature is reviewed
hy Pickering c/i//. (1989).
Channels
The term channel is reserved l"oi* features underlain by nearly
contemporaneous sediment, on fan surfaces, or in deep-sea
basins. Channels are of all sizes, and may be either erosional or
deposilional in character (Figure S46), as clearly indicated by
694
SUBMARINE FANS AND CHANNFI.S
seismic profiling, for both modern and ancient channels (Clark
and Pickering. 1996). Most active channels are directly
connected with the main feeder canyon or channel, but some
in the mid to lower fan have local "headwaters" on submarine
slopes or on the Hanks of large levees. In the midfan region, the
position of the active channel changes IVequently due to
avulsion. The scale of submarine channels is generally larger
than that of subaerial river channels, because they are adjusted
to transmit intermittent turbidity current tlows. many of which
are large events, and most of which travel more slowly, on a
given subtnarine slope, than a river of comparable size would
on a eotnparablc subaerial slope.
Generally, many of the flows exceed the capacity of the
channel, so most submarine channels have very large
constructional levees, compared with rivers. In large fans, the
thickness of levee deposits may reach hundreds of meters
(Figure S49). Coriolis effects are proportionately larger for
turbidity currents than for rivers, causing more frequent
overbank Ilow on one side of submarine channels, so that one
levee is generally higher Uian ihe other. In other respects,
submarine channels show many features in common with
subaerial channels, including internal terracing, braiding, and
meandering.
Though many flows overflow channels, particularly in the
middle and tower parts of fans, most of the coarser sediment
(sand, and especially gravel) is carried in the lower part of the
flow, so is confined to the channels, and only spread out a
sheet-like deposits where the flows emerge from the distal end
of the channel (Figure S50).
On submarine Tans, well-developed channels are character-
istic of the middle region of the fan. While the channels are
active, there are generally sites of sediment by-passing, or
accumulation only of the coarsest lag deposits. Aggradation
and catastrophic events cause channels to change position
frequently on the fan surfaces, and abandonned channels are
filled by smaller flows carrying finer sediment, so in the
geological record they can frequently be recognized by their
basal erosion surface and lag. by amalgamation surfaces, and
by lining and thinning upward sequences in the sandstone
filling, even when the channel geometry cannot be clearly seen
in the outcrop. For further details see Clark and Pickering
(1996).
Depositional lobes
Depositional lobes, with only minor channels, and with a
generally convex-up profile can be mapped beiow (distal to)
distributary channels on some modern fans. Their presence is
inferred in ancient fans, by the existence of "packets"" of
relatively coarse turbidite beds (some tneters or tens of meters
thick),
separated from olher packets by finer sediments.
Though some authors have claimed that these "packets" show
thickening and coarsening upward sequences (supposedly
produced by progradation of the lobes) statistical studies have
failed to support this as a general trend. Attempts to relate the
statistical trends actually exhibited by turbidite sequences to
fan environments have been reviewed by Carlson and
Grotzinger (2001).
Facies models for small, coarse fans
Many authors have proposed models (almost all of them
variations of the original model of Mutti and Ricci Lucchi,
1972) to aid in the identiiication oi ancient fans and in the
description and interpretation of the observed faeies. For
succinct accounts, see Walker (1992) and Normark elal.
(1993;
and in Weimer and Link, 1991). and for a more extended
account see Pickering etaL (1989). Besides the use of trends in
bed thickness and grain size, these schemes use the frequency
SCAR
CLASS CLASS CLASS CLASS CLASS
A B C D.E F
" "
T
NNEI
FAN
INNER
FAN
MIDDLE
FAN
OUTER
FAN
BASIN
PLAIN
Figure S47 The Ian model (it" Mulli arid Ritti Lucchi (1472) <is redrawn by Pickering ct ,il. 11989, their Figure 7.2 on p.l62i. "Class A", etc. refer
to t'acies: A are conglomerates, B are coarse sandstones, bolh generally massive and anialj^amated. C are "classical turbidites" of the proximal
type,
ihat can be described by the "Bouma sequence" of internal structures ifound mainly un the ouler fans and basin floor). D and E are thin
alternations of mud and sand. F are chaolic facies. FU refers In fining and thinninj; iipwrirH, CU to coarsening and thickening upward.