In yet another approach to Paleozoic platforms, this study takes a restricted time slice and compares the cycle patterns in two different platforms: the Middle Appala- chians of Eastern North America, and the Great Basin on the Western side.

Shoal-water emergence cycles and ramp cycles are dif- ferentiated, and time-subsidence plots provide a way of comparing the histories. Forward computer-modelling can be designed to produce somewhat similar sequences. The facies patterns and cycle patterns of these platforms differ markedly from those of the Early Carboniferous and those of the Triassic.

The periodicities have not been established owing to the great uncertainties of stage durations. In addition, frequency ratios, so helpful in the Mesozoic, are not readi- ly applicable to these early Paleozoic times inasmuch as the precessional and obliquity frequencies were probably considerably higher than those prevailing now, whereas the eccentricity frequencies have probably not changed. Such differences might in the future allow a better ap- proach to the problem of deceleration of the earth's spin rate, but this will require cleaner cyclicity data than are now available.

RELATION OF EUSTASY TO STACKING PATTERNS OF METER-SCALE CARBONATE CYCLES, LATE CAMBRIAN, U.S.A.

DAVID OSLEGER Department of Earth Sciences

University of Cahfornia Riverside, California 92521

AND

J. FRED READ Department of Geological Sciences

Virginia Polytechnic Institute and State University Blacksburg, Virginia 24061

A~rRncr : An interbasinal study of Late Cambrian cyclic carbonate successions in the Appalachian and Cordilleran passive margins suggests that superimposed orders ofeustasy controlled the development of large-scale depositional sequences and the component peritidal to subtidal meter-scale cycles that comprise them. The focus of this paper is on the small-scale cyclicity, its probable control by Milankovitch-forced sea-level oscillations, and how stacking patterns of meter-scale cycles can be used to define internal com- ponents of larger-scale sequences and estimate variations in relative sea level.

Fining-upward peritidal cycles showing evidence of episodic emergence grade seaward into coarsening-upward subtidal cycles which lack evidence of emergence and form a continuum across the Cambrian carbonate platforms. Eustasy appears to exert the dominant control on the simultaneous development of peritidal and subtidal cycles on Late Cambrian carbonate platforms. Evidence for Milankovitch forcing of glacio-eustatic sea-level oscillations is provided by a 4:1 bundling of fifth-order meter-scale cycles (~ 96 ky) within fourth-order cycles spanning tens of meters (~ 440 ky) within the Big Horse Member of the Orr Formation in the House Range of Utah. The 4:1 bundling may manifest the short eccentricity to long eccentricity ratio ofthe Milankovitch astronomical rhythms.

Systematic changes in the stacking patterns of meter-scale cycles can be used in conjunction with Fischer plots to define long- term sea-level cycles. On Fischer plots ofperitidal cyclic successions, long-term relative sea-level rises are characterized by thick, subtidal-dominated cycles with thin laminite caps. Long-term relative sea-level falls are defined by stacks of thin, laminite-dominated cycles that show brecciated cycle caps and quartz sands toward the relative sea level lowstand. On Fischer plots of dominantly subtidal cyclic successions, long-term sea-level rise is characterized by storm-dominated, open marine carbonate cycles or thick, deep ramp, shale-based cycles. Falling segments of the Fischer plot are characterized by thin, shallow subtidal cycles composed of restricted lithofacies. Cycle stacking patterns (parasequence sets) provide the crucial link between the meter-scale cycles (parase- quences) and the larger scale sequences and their component systems tracts.

One- and two-dimensional models of pedtidal and subtidal cycle development indicate that, whereas peritidal cycle thickness is primarily controlled by accommodation space, deeper subtidal cycle thickness is primarily controlled by sedimentation rate. Lithofa- ties within peritidal cycles reflect the sedimentologic response to fluctuations in sea level; lithofacies within subtidal cycles may be related to fluctuations in the zones of fairweather and storm-wave reworking that oscillated in harmony with sea-level fluctuations and may have acted as a subtidal limit to upward aggradation. The 2-D modelling illustrates how stacked peritidal to deep subtidal carbonate cycles with thicknesses, compositions and stacking patterns similar to the Late Cambrian of North America can be generated by Milankovitch-driven composite eustasy.

INTRODUCTION

Hierarchies of stratigraphic cyclicity have long been recognized throughout the geologic record (e.g., Barrell 1917; Fischer 1964; Koersehner and Read 1989; Gold-

hammer et al. 1990; Borer and Harris 1991) and appear to be related to the combined effects of several orders of relative sea-level oscillations. Shallowing-upward, meter- scale carbonate cycles (parasequences) tend to be system- atically arranged within larger-scale successions (parase-

JOURNAL OF SEDIMENTARY PETROLOGY, VOL. 6 !, NO. 7, Dec., 199 l, P. 1225-1252 Copyright 1991, SEPM (Society for Sedimentary Geology) 0022-4472/91/0061-1225/$03.00

1226 DA V1D OSLEGER AND J. FRED READ

H(

LATE CAMBRIAN SEDIMENTARY FACIES

[~] Cratonal siliciclaslics

Shallow marine carbonates

Basinal sAiciclastics

Fio. ! .--Location map of sections measured in the study. Late Cam- brian base map modified from Palmer (1974) to show the inner and outer detrital belts and the middle carbonate belt.

quence sets). Stacking patterns of the meter-scale cycles (stratigraphic trends in cycle thickness and composition) can be used to identify large-scale sequences, their com- ponent systems tracts, and long-term relative sea-level changes. An interbasinal study of Late Cambrian peri- cratonic cyclic carbonates was conducted by logging me- ter-scale cycles of time-equivalent cyclic successions on separate platforms to evaluate various types of cycles, their stacking patterns, and potential mechanisms that may have controlled their origin.

The objectives of this this paper are to: 1) describe Late Cambrian peritidal to deep subtidal cycles and interpret the environmental conditions under which upward-shal- lowing occurred; 2) evaluate the controlling mechanisms of meter-scale cycle formation (specifically the connection with Milankovitch orbital variations); 3) illustrate char- acteristic stacking patterns of cycles that define rising and falling portions of sea-level curves using Fischer plots; and 4) use quantitative I -D and 2-D modelling to con- strain the probable conditions under which coeval peri- tidal and subtidal cycles were deposited.

STRATIGRAPHIC AND TECTONIC SETTINGS

Complete sections of Late Cambrian strata were mea- sured and logged bed-for-bed in the House Range of west central Utah and in the Appalachian Mountains in Ten- nessee, Virginia and eastern Pennsylvania (Fig. 1). Bio- stratigraphic control of the formations for each of the localities (Fig. 2) was obtained from published work

(Palmer 1965, 1971a, 1971b; Derby 1965; Rassetti 1965; Hintze 1974; Hintze and Palmer 1976; Hintze et al. 1980; Eby 1981; Taylor and Miller 1981; Miller et al. 1982; Orndorff 1988; Sundberg 1990). Primary estimates of subsidence rate and cycle duration were made using the DNAG time scale values for the duration of the Late Cambrian (Palmer 1983). However, considerable contro- versy exists regarding the total duration of Cambrian time (Cowie and Harland 1989), with new age dates (Benus 1988) supporting a much shorter time span. Therefore, a conservative 50% margin of error is incorporated into all calculations involving total Late Cambrian time.

Field locations were chosen on the basis of: 1) quality of exposure and absence of structural complications, 2) availability of biostratigraphic data (especially biomere boundaries), and 3) platform location along the transition between shallow-water carbonates and deeper-water fine- grained siliciclastics where intertongueing relations best define excursions in sea level. A total of 2200 m of section was logged and numerous other sections previously de- scribed in the Appalachians (Zadnik 1960; Markello 1979; Koerschner 1983; Demicco 1981) and in Utah-Nevada (Palmer 197 la; Kepper 1972; Lohmann 1976; Rees 1986) were field-checked. Hand samples of individual lithofa- cies were slabbed and thin sectioned to provide additional detail for paleoenvironmental interpretations. Details re- garding the exact locations of sections and logs of strati- graphic intervals can be found in Osleger (1990).

The Appalachian and Cordilleran passive margins orig- inated in response to breakup of a Late Proterozoic su- percontinent around 625 to 555 Ma (Bond et al. 1984). Both passive margins developed wedge-shaped prisms of post-rift subtidal to peritidal carbonates and interlayered siliciclastics during Late Cambrian time (Fig. 3). The Ap- palachian passive margin contains up to 1.6 km of Middle to Late Cambrian shallow water carbonates and intrashelf basin shale and siltstone (Read 1989). The Cordilleran passive margin of the western United States accumulated approximately 2.0 km of post-tiff Middle to Late Cam- brian carbonates and fine siliciclastics (Stewart and Poole 1974; Levy and Christie-Blick 1989).

SHALLOWING-UPWARD METER-SCALE CYCLES

A spectrum of meter-scale peritidal to deep subtidal carbonate cycles (1-15 m) can be recognized in Late Cam- brian strata of the two passive margins (Figs. 4, 7). Suc- cessions of fining-upward peritidal cycles (Wilson 1952; Chow and James 1987; Demicco 1985; Koerschner and Read 1989) extend entirely across the broad Late Cam- brian passive margin of the Appalachians but are re- stricted to a narrow zone near the Wasatch hinge line of the coeval Cordilleran passive margin (Palmer 1971a; Kepper 1972). Peritidal cycles in the Cordillera of Utah grade seaward into shallow to deep subtidal cycles show- ing an upward increase in grain size, bed thickness, and other indices of higher energy. Subtidal cycles are not capped by intertidal lithofacies, nor do the subtidal cycles exhibit exposure features such as microkarsting or vadose dissolution/cementation. The cycles form a continuum

EUS7"AS Y AND CYCLE STACKING PATTERNS OF LATE CAMBRIAN CARBONATES 1227

~ I '~0 ~ ~ i BIOMERE w

1

Z E P~CHA.~OtD

m ~

UJ CEPHAUlD

m "523 AM - X W

E MARJUMIID

TRILOBITE ZONE

MISSISSIOUOIA

SAUKIA

SARA TOGIA

TAENICEPHAL US

EL VJNIA

OUNDERBERGIA

APHEL ASPIS

CREPICEPHAL US

CEDARfA

BOLASPIDELLA

HOUSE RANGE, UTAH

LAVA DAM

'RED TOPS z u

HELLN- MARIA Z

SNEAK- OVER

CORSET SPRING ~

JOHNS WASH

CANDLAND

BIG HORSE

WEEKS FM.

SW VIRGINIA NE TENN.

COPPER RIDGE I

CONOCO- CHEAGUE

( MAYNARD.

VILLE

8

ELBROOK

EASTERN PENN.

ALLENTOWN DOLOMITE

FIG. 2.--Biostraligraphic chart of Late Cambrian strata in the Cordilleran and Appalachian passive margins.

across the carbonate platforms and are genetically linked to one another by shared lithofacies (Fig. 4) (Osleger 1991). These asymmetric, meter-scale cycles are the parase- quences of sequence stratigraphic terminology in that they are "relatively conformable successions of genetically re- lated beds bounded by marine flooding surfaces" (Van Wagoner et al. 1987).

The vast majority of meter-scale cycles recognized on both passive margins are asymmetric with relatively thin basal lithofacies recording abrupt drowning and relatively thick upper lithofacies recording gradual shoaling. To- ward the outer platform of both passive margins, some cycles exhibit subequal amounts of deepening and shal- lowing lithofacies. These symmetric cycles are relatively rare, however, and are restricted to deeper water positions on each platform. No deepening-upward cycles were rec- ognized.

Estimations of average cycle duration are complicated by errors in the absolute time scale, the effects of com- paction, and assumptions of constant sedimentation rates. Acknowledging these potential sources of error, average cycle durations for non-decompacted Late Cambrian cy- cles range from roughly 40 to almost 150 ky. Taking a conservative 50% margin of error into account, this range of durations may extend from about 20 to 225 ky, the normal range expected for meter-scale cycles (Algeo and Wilkinson 1988).

Vertical and Lateral Consistency of Cycles

Late Cambrian meter-scale cycles of the Appalachians are extremely rhythmic vertically in outcrop with only minor variations in the arrangement of component litho- facies (Demicco 198 l; Koersehner and Read 1989). Peri-

tidal cycles comprise successions of hundreds of stacked cycles but are difficult to correlate laterally in the Ap- palachians because of the distance between outcrops, the lack of marker beds and the lack of precise biostratigraph- ic control. However, groups of time-equivalent Late Cambrian cycles with distinct stacking patterns (fourth- and third-order scale) can be correlated along the Ap-

LATE CAMBRIAN PLATFORM MORPHOLOGIES

APPALACHIAN REEF - RIMMED PLATFORM NW SE

UTAH - NEVADA DISTALLY - STEEPENED RAMP W E

FIG. 3.--Late Cambrian platform morphologies of the Appalachian and Cordilleran passive margins. Formation and group names super- imposed on lithologic symbols.

1228 DA VID OSLEGER AND J. FRED READ

GRADATION OF CYCLE "TYPES ACROSS A LATE CAMBRIAN SHALLOW TO DEEP RAMP

LAMINITE-CAPPED PERmDAL CYCLE

THROMBOLJTE BIOHERM, SHALLOW SUBTIDAL CYCLE

SL m

1 o -~.~- o o o o oOnOoOoO/.

--~- l

~ID GRAI~T~E, SHALLOW SUBTIDAL CYCLE

~ RYPTALGAL LAMINITE

~ ] I ~ SKELETAL'PELLETAL THICK I~:-~-~-' PACKSTONE WITH LAMINITE I ' I STORM BEDS

/ , , ' t~- c r ^- .

SKELETAL PACKSTONE, MID-RAMP CYCLE

FWWB

t ~ = t "

SPICUUT1C WACKESTONE, DEEP RAMP CYCLE

~ RIBBON ~ BURROWED ROCK WACKESTONE

[~] THROMBOUTIC ~ PELOIDAL WACKESTONE BOUNDSTONE PACK, STONE

700IDJNTRACLAST ~ ARGILLACEOUS GRAIN, STONE NODULAR WACKESTONE FIG. 4.--Arrangement of peritidal to deep subtidal cycle types across a hypothetical Late Cambrian platform. Note the location of the zones

of fairweather and storm-wave reworking and their relation to cycle types.

palachians from Virginia to Pennsylvania using Fischer plots (Read 1989; Osleger and Read, unpublished data).

Subtidal cycles of the Utah Cordillera are repetitive over 15 to 40 successive cycles before gradually changing to a different cycle type. In the House Range, cycles can be tracked as subparallel bands for many kilometers along the mountain flank and meter-scale deep subtidal cycles can be correlated between outcrops greater than 45 km apart (Fig. 5), indicating that the subtidal cycles are not local facies mosaics. No up-dip peritidal cycles exist that can be directly correlated with down-dip subtidai cycles in the House Range.

Lithofacies and Depositional Environments

Peritidai Cycles.--Laminite-capped cycles (0.4-7.0 m) of the Appalachian Late Cambrian are composed of a basal ooid-intraclast grainstone lag deposit overlain by either ribbon carbonates or thrombolite boundstones (Ta- ble 1; Fig. 4). The cycles are capped by mudcracked thick laminites and/or cryptalgal laminites; quartz arenites or carbonate clast breccias may cap some cycles, particularly during long-term relative sea-level fall. The cycles exhibit abrupt upper and lower boundaries but have gradational internal boundaries between iithofacies. Peritidal cycles extend over much of the Appalachian reef-rimmed shelf (Zadnik 1960; Reinhardt 1977; Demicco 1981; Read

1985) and are recognized within the Elbrook, Copper Ridge, Conococheague and Allentown Formations.

The basal ooid-intraclast sandy lag deposit migrated onto the underlying tidal fiat cap from shallow offshore wave-agitated shoals during initial rapid transgression. Hardgrounds developed on the lag deposit as the trans- gressive rise of sea level outpaced sediment production. As the rate of relative sea-level rise decreased, throm- bolites locally established themselves on marine-ce- mented grainstone lags and grew to sea level. Ribbon rocks accumulated adjacent to bioherms in shallow sub- tidal to lower intertidal conditions. The rippled peloidal silts/fine sands were laid down during storms with drapes of lime mud settling out during the waning stages (De- micco 1983).

Progressive shallowing and progradation is reflected in the upward transition into increasingly mudcracked rib- bon rocks, SH and LLH stromatolites and thick laminites. Centimeter-scale thick laminites are mechanically-de- posited couplets of fine pelodial silts and mud drapes laid down on the intertidal fiats by storm and tidal currents (Hardie and Ginsburg 1977). This lithofacies often caps cycles or grades up into cryptalgal laminites. Lack of bur- rowing, abundant mudcracks, silicified evaporite nodules and windblown quartz sand within laminite lithofacies indicate hypersaline and semiarid conditions. Cratoni- cally-derived quartz sands (Wilson 1952; Koerschner and Read 1989) were probably brought in during long-term

E USTAS Y ,tND CYCLE STACKING PA TTERNS OF LATE CAMBRIAN CARBONATES 1229

falls in relative sea level and were incorporated into cycle caps during short-term regression or reworked by the suc- ceeding marine transgression into the basal lag deposit of the succeeding cycle.

Relatively few characteristics of supratidal conditions have been observed in peritidal cycles of the Appalachi- ans (Hardie and Shinn 1986; Koerschner and Read 1989). The absence of bedded evaporites (other than isolated silicified nodules), the relative scarcity of dissolution breccias and the lack of erosional relief along the sharp top of laminite caps may indicate erosional removal dur- ing the formation of a deflation surface down to paleo- watertables. Supratidal evaporites may have been dis- solved during the occasional rains that occurred in the semi-arid climate. Lack of land plants would have not favored caliche development, and the prevailing desic- cating conditions might have inhibited cementation, making erosional removal of supratidal sediment by eo- lian action a tenable mechanism for forming the planar surface at the tops of most laminite caps. However, some evidence of exposure, non-deposition and erosion is ex- hibited. Relatively scarce, irregular veneers of carbonate clast breccia fill solution-enhanced lows at the top of some caps and are probably thin regoliths that developed on the non-vegetated Late Cambrian exposed flats. Fossil molds and geopetally-filled leach voids in subtidal lime- stones indicate flushing by undersaturated meteoric wa- ters (Koerschner and Read 1989) during relative sea level falls.

Shallow Subtidal Cycles.--Shallow subtidal cycles are defined by the interpreted paleowater depth of the cycle cap and include cycles capped by thrombolite bioherms, ooid grainstones and skeletal packstones. Paleowater depths are based upon modem analogs of the lithofacies and their sedimentary structures.

Cycles capped by thrombolite bioherms (1.5-12.0 m) consist of a basal dark gray peloidal packstone overlain by stacked thrombolite-stromatolte bioherms and later- ally equivalent light gray cross-bedded peloidal-oncolitic grainstone (Table 2; Fig. 4). These cycles record progra- dation of shallow subtidal bioherms and associated high energy grainstones over slightly deeper subtidal peloidal packstones of a restricted shelf. More than thirty of these cycles are recognized within the upper Hellnmaria Mem- ber of the Notch Peak Formation throughout the House Range of west central Utah.

These cycles were initiated by onlap of peloidal pack- stones/wackestones onto thrombolite-stromatolite bio- herrns. Horizontal to low angle cross-lamination, lack of recognizable skeletal material, and dark gray bioturbated textures suggest restricted, quiet water (but not necessarily deep) deposition. With slowing of the rate of short-term relative sea-level rise, thrombolitic bioherm complexes were able to establish themselves on hardgrounds or other stable substrates. The bioherms are laterally discontin- uous, suggesting development as isolated, shallow sub- tidal "patch reefs" on top of the basal peloidal veneer. Continued slow rates of relative sea-level rise are indi- cated by the stacking of individual bioherms up to 12 m thick without intervening bedded lithofacies. Many of the

STEAMBOAT PASS, S. HOUSE RANGE

ORR RIDGE, N. HOUSE RANGE

I " I I

I I

i

"1 J ,

! 1 ~ _ _

i i ' .,;I wP

PTEROCEPHAL I ID

. B IOMERE - - ~-

- " , - -I i!~! ~ i:ii!ii!ii ii~ii,, / 1 " ! " | ' l

," ,J I I i V l "

/ I " - I |

FIG. 5.--Correlation of deep subtidal cycles within the Sneakover Pass Member, Orr Formation, across 45 km within the House Range. Correlation based upon the Orr/Notch Peak and Steamboat Spring/ Sneakover formational boundaries.

thrombolitic bioherms have stromatolitic laminae out- lining the outer surface of the mound, and some stro- matolitic biohermal layers are interbedded within the dominantly thrombolitic complex (Fig. 4). This implies either episodic shallowing to intertidal depths or perhaps variations in salinity (and associated grazing and boring epifauna) related to periodically restricted conditions on the platform (Aitken 1967; Kennard and James 1986).

Shallow subtidal conditions for the thrombolites are supported by the laterally equivalent light gray, crossbed- ded peloidal-oncolitic-oolitic-intraclastic grainstones

1230 DA V1D OSLEGER AND J. FRED READ

TABLE 1.--Peritidal lithofacies

Intraclast Breccia (5-20 cm) Bedding Characteristics: Laterally discontinuous veneers of angular to

elongate intraclasts in a calcrete matrix within mudcracked laminite lithofacies; often abundant quartz sand; intraclasts often poorly sorted with no preferred orientation or grading; irregular upper surface usu- ally overlain by oolitic grainstone.

Internal Composition/Texture: Angular elasts composed of LLH stro- matolites and cryptalgal laminites and have corrodcd edges and doio- mitic rinds; quartz sand grains are moderately sorted, subrounded and frosted; vuggy voids common.

Cryptaigal Laminite (0.2-3.0 m) Bedding Characteristics: Dolomite; mm-scale planar and crinkly lam-

inations; mudcracks, deep prism cracks, tepees and silicified evaporite nodules common; thin flat pebble conglomerates and mud chip in- traclast layers; occasional 1-3 grain thick quartz sand stringers; grades upward from ribbon rock or more commonly thick laminite; occa- sionally capped by irregular cherty breccias but more typically over- lain by intraclastic transgressive lag of overlying cycle; some cycles are reversing with the cryptalgal laminite coarsening upward into less mudcracked thick laminite.

Internal Composition/Texture: Laminar couplets composed of basal silt- size peloidal packstones grading up into mudstone laminae; some low-angle cross-lamination and micro-scoured bases in peloidal silts; some laminoid fenestrae.

Thick Laminite (0.2-3.0 m) Bedding Characteristics: Dolomite; cm-scale laminations of silt and

mud couplets; laminations planar to wavy to discontinuous; common cross-laminated scour fills and current ripples; some short mudcracks and silicified evaporite nodules; commonly overlie ribbon rocks; usu- ally grade up into cryptalgal laminite but sometimes will cap incom- plete cycles.

Internal Composition/Texture: Couplets consist of peloid-quartz silt packstone that grades up into dolomitic mudstone; some well-round- ed quartz sand laminae 1 or 2 grains thick; some laminoid fenestrae.

Ribbon Carbonate (0.5-4.0 m) Bedding Characteristics: Alternating irregular layers of peloidal lime-

stone and dolomitic mud; discrete burrowing; some gutter scours with cross-laminated peloidal fill; shallow mudcracks become more abun- dant upward; common flat pebble conglomerate beds with internal scours, hardgrounds, and mud drapes; often flank and overlie throm- bolite bioherms; fine upward into thick laminites.

Internal Composition/Texture: Peloidal packstones grade upward into argillaceous dolomite caps; peloidal laminae contain minor quartz silt and skeletal debris and show occasional scoured bases and low- angle cross-lamination; flat pebble beds composed of imbricate dis- coidal clasts of laminated peloidal packstone or doiomitic mudstone; matrix between clasts consists of sand-size intmclasts, trilobite and echinoderm debris, pellets, and minor quartz silt.

Stromatolite Bonndstone (0.5-2.5 m) Bedding Characteristics: SH and LLH stromatolites typically encrust

tops of thrombolite bioherms or basal grainstone lags; fingers often coalesce into fan-like forms or crenulated sheets; club shapes common toward the tops of bioherm complexes; commonly surrounded lat- erally by intraclast-peloidal packstones or ribbon rocks which onlap and smother the stromatolites.

Internal Composition/Texture: Alternating mm-scale laminae of dolo- mitic peloidal silts and muds; some irregular and laminoid fenestrae; some thin quartz silt laminae and coarser laminae of intraclasts.

Thrombolite Boundstone (0.5-2.5 m) Bedding Characteristics: Globose to upward-widening flat-topped bio-

herms to coalescent biostromes; individual bioherrns often stacked on top of one another (up to 12 m thick) without intervening contin- uous bedded lithologies; digitate fingers (1--6 cm high x 1-2 cm wide) have erosional edges, are grouped in clusters and often overlie massive cores of thrombolites; lime sand interhead fill flank and locally onlap and blanket bioherms; bi-directional and high-angle cross-bedding, scours and multiple hardgrounds common in interhead fill; lime sands fine upward into ribbon carbonates; thrombolites typically nucleate on underlying intraclastic grainstone lag.

Internal Composition/Texture: Clotted micritic textures; commonly grain-rich reflecting the composition of the interhead fill; small in- traclasts, skeletal debris, ooids and pellets are dominant components; fingers show traces of Girvanella, thin stringers of micfite and mi- crospar cements; Renalcis recognized toward the base of many of the bioherrns; common irregular fenestrae with geopetal fillings.

which resemble modem, high energy, non-skeletal grain- stones enveloping growing stromatolitic bioherms in ti- dal channels in the Bahamas (Dill et al. 1986). Irregularly laminated oncolites, coated peloids, and high-angle tabular erossbedded gra ins tones that a l te rnate w i th hor i zonta l ly bedded packstones reflect variable energy conditions. The lack of open marine fauna within the inter-bioherm grain- stones may reflect either elevated salinities on the re- stricted platform or intermittent high wave or tidal en- ergies on mobile sandy substrates that precluded the establishment of grazing organisms. Only the robust mol- lusks Mathevia and Matherella are found associated with the bioherms, supporting the case for high energy con- ditions.

Cycles capped by ooid grainstone (0.5-4.2 m) consist of burrowed wackestone/packstone grading up into on- colite-skeletal packstone/grainstone capped by oolitic grainstone (Table 2; Fig. 4). The succession of lithofacies record progradation of oolitic shoals over deeper ramp lithofacies and occur in the Big Horse Member, Orr For- mation of the House Range of Utah (Lohmann 1976).

Burrowed wackestone/packstones are subtidal facies deposited below fairweather wave base under normal ma- rine conditions. Pervasive bioturbation (ichnofabric in- dex 3-5; Droser and Bottjer 1986), bioclastic debris, and clusters of pellets suggest an active infauna. Laterally dis- continuous skeletal packstone lenses with erosional bases and burrowed tops are rapidly deposited storm beds that escaped homogenization by burrowers. The abundant quartz silt may have been transported from the craton across the inner detrital belt (Palmer 197 la) and onto the carbonate platform through a west-trending subtidal channel that debouched near the House Range (Lohmann 1977).

With shoaling, skeletal sand sheets migrated across the burrowed wackestones and were reworked by storm and wave currents. Megarippled units suggest emplacement as sand waves and shallow bars. The upward transition from open marine skeletal packstones to oncolitic-peloi- dal grainstones indicates increasingly shallow, restricted conditions (Enos 1983), perhaps peripheral to active ooid shoals (Hine 1977).

I:'US7;1SY AND ('YCLE ST.4C'KING P.4TTERNS OF LATE CAMBRIAN CARBONATES 1231

Ooid Grainstone (0.1 -1.5 m) Skeletal-Oncolitic Grainstone (0.5-2.5 m) Bedding Characteristics: Thin to medium bedded, common high angle

crossbedding; random stacked hardgrounds; gradationally overlie skeletal-oncolitic grainstones; abruptly overlain by burrowed (ii3-ii5) wackestone/packstone lithofacies.

Internal Composition/Texture: 90-95% well-sorted ooids; common oo- litic intraclasts and random subangular quartz sand grains; ooids are concentrically laminated and have echinoderm-trilobite-quartz silt nuclei; isopachous marine cement rims and fine to medium equant spar cements; some dolomitized ooids.

Peloidal Grainstone (1.0-4.0 m) Bedding Characteristics: Thin to medium-bedded dolomite; light to

medium gray cross-bedded grainstones; laterally equivalent to (and gradationally onlap) thrombolitic-stromatolitic biohenns; abruptly overlain by either thin-bedded, evenly-laminated dark gray peloidal dolomites or thrombolitic bioherms.

Internal CompositioniTexture: Fine to medium crystalline dolomite; dominantly peloidal grainstones with variable numbers of large on- colites, ooids and intraclasts (all seen as ghosts); flat pebble conglom- erates in lenses and scours; robust Mathevia and Matherella mollusk shells; some internal mm-scale gentle cross-lamination of peloids.

Bedding characteristics: Thin to medium bedded; megarippled, high angle crossbeds; commonly coarsen upward from skeletal packstone up to interbedded oncolitic and skeletal grainstones; random in- terbedded lenses of burrowed (ii3-ii5) wackestone low in lithofacies and occasional interbedded short digitate stromatolite fingers high in lithofacies; gradationally overlies burrowed wackestone lithofacies and underlies oolitic grainstone lithofacies.

Internal CompositioniTexture: Cm-size, well-rounded oncolites with large intraclast cores; some multigenerational oncolites; moderate sorting; random peloids and muddy intraclasts; micritic and abundant turbid marine cements. Skeletal packstone/grainstones composed of abundant trilobite and echinoderm debris and common pellets and rounded elongate muddy intraclasts; some alignment and imbrication of allochems; some crystal silt-filled geopetal voids and shelter po- rosity.

The crossbedded oolitic grainstone cap resulted from progradation of ooid shoal complexes as migrating spill- over lobes that formed in response to storm or tidal cur- rents (Hine 1977; Harris 1979). Rounded oolitic intra- clasts indicate early marine cementation and the lack of leached ooids or other vadose features suggests continual submergence.

Cycles capped by skeletal packstone (1.0-7.5 m) are composed of basal nodular argillaceous wackestone over- lain by burrowed, storm-deposited wackestone/pack- stone coarsening upward into a skeletal packstone cap. (Table 2; Fig. 4). They occur in the lower Big Horse Mem- ber (Orr Formation) of the House Range. These cycles developed on the mid-ramp at intermediate water depths above the zone of storm wave reworking seaward of ooid grainstone shoals. The succession of lithofacies record gradually increasing storm influence as the platform shal- lowed to skeletal shoal depths.

The basal nodular, argillaceous wackestone is a distal storm facies deposited on the middle ramp between bur- rowed wackestones and packstones and deeper water sil- iciclastic muds (Aigner 1985). The low angle cross-lam- inated rnicropeloidal and quartz silty layers within the modular limestones were probably transported seaward from the shallow ramp during periodic storms. Nodules may have formed early by submarine lithification under weak bottom currents (Mullins et al. 1980) or may be the result of late pressure solution and compaction. Argilla- ceous muds fell out of suspension during waning storm activity. The two sediment types were mixed by the bur- rowing infauna.

As the depositional surface shallowed with aggradation, the argillaceous content of the sediment decreased, grain size and skeletal content generally increased, and the abundance of storm beds with hummocky cross-stratifi- cation increased. Storm-deposited skeletal packstones (5- 20 cm) are laterally discontinuous, fine upward and, from

base to top, consist of: 1) sharp, scoured bases, 2) skeletal debris with peloids and mud perched above shells with pendant bladed marine cements extending down into now- occluded shelter pores, 3) hummocky cross-stratified pe- loidal packstones and 4) bioturbated calcisiltite caps. This allochthonous debris was transported by storm-generated currents and then reworked by oscillatory shear currents (Aigner 1985). Upward within individual cycles, skeletal material becomes more abundant and storm deposits ap- pear amalgamated with numerous wavy beds of subtly graded skeletal debris. Common platy ginanellid crusts are imbricated and suggest that the packstone cycle cap may have formed within the photic zone (Pfeil and Read 1980).

Deep Ramp/Intrashelf Basin Cycles.-Deep subtidal cycles are characterized by sedimentary structures within the cycle cap indicative of deposition below the zone of storm wave reworking. These cycles are commonly shaly and may be capped by spiculitic wackestone, skeletal storm beds or flat-pebble conglomerates.

Cycles capped by spiculitic wackestone (0.7-3.1 m) are composed of basal nodular argillaceous mudstone over- lain by burrowed spiculitic wackestone with upward-in- creasing skeletal packstone lenses (Table 3; Figs. 4-6). These cycles occur in the Sneakover Member (Orr For- mation) and in the Hellnmaria and Lava Dam Members (Notch Peak Formation) of the House Range. Very sim- ilar deep ramp cycles have been described in the Upper Muschelkalk of the South-German Basin (Aigner 1985) and the Catalan Basin of Spain (Calvet and Tucker 1988). Spiculitic wackestone-capped cycles of the House Range developed on the deep ramp very near the base of storm wave reworking seaward of the skeletal-packstone capped cycles. The abundant bioturbation (ii3-ii5) and trilobite and echinoderm debris within storm beds attest to well- oxygenated, normal marine conditions.

Shaly cycles capped by skeletal storm beds (2.5-1 5.0 m)

1232 DA VID OSLEGER AND J. FRED READ

TABLE 3.--Deep subtidal lithofacies

Thin-bedded Peioidal Packstones (1.0-4.0 m) Bedding Characteristics: Dolomitized, dark gray, thin to medium-bed-

ded peloidal packstones; random cm-scale horizontally laminated horizons; abruptly overlie light gray cross-bedded peloidal facies and thrombolite bioherms; occasionally differentially compacted beneath biohermal mounds.

Internal CompositionFFexture: Consist internally of dolomitized dark gray peloids and patchy micrite; evenly laminated; probable burrow- ing traces; lack any of the sedimentary features and grain types ex- hibited in the distinct cross-bedded peloidal grainstone lithofacies.

Nodular Argillaceous Waekestone (1.0-8.0 m) Bedding Characteristics: Discontinuous, nodular, wavy and thin-bed-

ded; recessive weathering; argillaceous seams with concentrations of ferroan dolomite separate platy mudstone/wackestoae; quartz silt lenses common; random very low-angle cross-laminated lenses and nodules; some ram-scale horizontal laminations; common horizontal burrows on tops of bedding planes.

Internal Comoosifion/Texture: Variable intermixed textures from mud- stone through packstone with wackestone dominant; micropeioids and quartz silt dominant along with finely comminuted skeletal debris including sponge spicules; patchy lime mud; occasional floating elon- gate trilobite fragments; peloids and quartz silt alternate in low angle mm-scale cross-laminations with divergent dip angles (hummocky cross-stratification?)

Burrowed Wackestone/Packstone (0.5-4.0 m) Bedding Characteristics: Nodular to wavy thin beds ofdoiomitic, mot-

tled limestone; thin quartz siltstone lenses abundant; occasional skel- etal packstone lenses; pervasively bioturbated (ii4-iiS); gradational contact with overlying oncolitic-skeletal packstones; abrupt lower contact with ooid grainstones of underlying cycle.

Internal Com0osition/Textare: Silt-size finely comminuted grains dom- inant with floating, randomly oriented echinoderm-trilobite debris common; clusters of peloids; abundant subangular quartz silt to fine sand; discrete burrows commonly dolomitized (ferroan); medium equant dolomite/calcite void-filling cements.

Spiculitic Wackcstone (0.7-2.5 m) Bedding Characteristics: Ledge-forming, medium to thick-bedded; mot-

tled medium to dark gray; black chert in discontinuous lenses and nodules; thoroughly bioturbated (ii5-ii6) with ferroan dolomitized horizontal burrows; skeletal packstone lenses become more common toward top of lithofacies; gradual lower contact with underlying nod- ular mudstones and abrupt upper contact with overlying nodular mudstones.

Internal Composition/Textare: Alternating cm-scale mudstones/wacke- stones and subordinate lenses and wavy beds of skeletal packstones; wackestones are strongly burrow-homogenized (ii5-ii6); sponge spic- ules dominate with common trilobite and echinoderm debris; pellets associated with burrows; some admixed quartz silt; packstone lenses have erosive scoured bases and are composed of unbroken elongate skeletal debris and muddy intraclasts with occluded shelter porosity and perched peloidal muds; some grading is evident with lenses with mud drapes at the top; some gentle micro-cross-laminations ofpeloids with bidirectional orientations.

consist of a thick basal shale abruptly overlain by upward- coarsening skeletal wackestone/packstones (Table 4; Fig. 7). These cycles occur in the Candland Shale, Corset Spring Shale and Steamboat Pass Members of the Orr Formation of Utah. These cycles commonly characterize long-term rises in relative sea level that cause onlap of deep outer ramp siliciclastic facies onto shallow ramp carbonate fa- cies.

The thick basal shale formed below the zone of storm wave reworking. Siliciclastic clays accumulated in a dys- aerobic environment, as indicated by the olive green to dark gray color, mildly bioturbated laminae, and sparse trilobite and phosphatic brachiopod fauna. The clays were probably derived from the craton and were transported across the carbonate belt (perhaps through the House Range Embayment trough) and onto the deep ramp as dilute clouds or bottom-hugging nepheloid layers (Board- man and Neumann 1984).

The uppermost carbonate beds of these shale-domi- nated cycles reflect rapid shallowing from shale to bio- turbated wackestones up into skeletal packstones. A few of these cycles shallow up to large (1.5 x 1.5 m) throm- bolitic bioherms that nucleated on flat-pebble conglom- erate storm beds. The abrupt transition in paleowater- depths between the deep, quiet water shales (perhaps water depths of > 40 to 60 m) and the shallow, clear water carbonates (perhaps water depths between 5 to 20 m) suggests that these cycles probably did not form by simple aggradation, which would provide a maximum of only 15 m of shallowing, but rather experienced a relative sea level rise (shales) followed by relative sea level fall (car-

bonates) (Osleger 1991). No evidence of subaerial ex- posure of the skeletal carbonates or the bioherms is rec- ognized, indicating that sea level never fell below the platform. With renewed relative short-term sea level rise, carbonate sedimentation ceased and the skeletal sands or bioherms were abruptly covered with clays deposited be- low the zone of storm wave reworking.

Shaly cycles capped by flat-pebble conglomerates (0.8- 5.5 m) consist of a basal calcareous green-brown shale grading upward into cross-laminated peloidal grainstones and quartz siltstones. The cycles are capped by amalga- mated flat-pebble conglomerate beds (Table 4; Fig. 7). They occur in the Nolichucky Formation, Virginia and Tennessee (Markello and Read 1982), and in Late Cam- brian strata of central Texas (Osleger 1990), Montana (Sepkoski 1982), and the southern Canadian Rockies (Aitken 1978).

The Nolichucky cycles record deposition above and below a fluctuating zone of storm wave reworking in a shallow intrashelf basin. The Conasauga basin was ad- jacent to the craton and derived its siliciclastic sediment from distant deltas (Hasson and Haase 1988; Read 1989). The base of storm wave reworking may have been shallow due to the barrier effect of the peritidal Elbrook platform to seaward (Markello and Read 1982). Progressive shal- lowing within individual cycles is indicated by an increase in grain size and in storm-generated sedimentary struc- tures. The peloidal grainstone/quartz siltstone lithofacies was deposited under the influence of oscillatory shear currents as indicated by parallel lamination and micro- hummocky cross-stratification.

EUSTASY AND CYCLE STACKING PA TTERNS OF LATE CAMBRIAN CARBONATES 1233

FIo. 6.-- Deep ramp cycles with spiculitic wackestone caps, Sneakover Member, Orr Formation, House Range. Basal lithofacies is composed of argillaceous nodular wackestones and exhibits a recessive weathering pattern. Overlying ledge-forming cap consists of spiculitic wackestone with upward-increasing storm-deposited packstone lenses composed of open marine skeletal debris.

The flat-pebble conglomerate caps of the cycles were deposited during severe storms that eroded the underlying semi-l ithified peloidal grainstone and redeposited the rounded, elongate clasts within tabular to lenticular beds (Sepkoski 1982). Multi-generational clasts and thin mud drapes that separate conglomerate beds within amalga- mated units were formed by mult iple storm events. Di- verse skeletal debris within the matrix between clasts re- flects normal marine conditions.

MECHANISMS CONTROLLING METER-SCALE CYCLE DEVELOPMENT

Mechanisms proposed to explain the genesis of meter- scale carbonate cycles have focused on perit idal cycles common throughout the rock record. The recognition of shallow to deep subtidal cycles that formed simultaneous- ly with perit idal cycles requires some modif ication of the mechanisms proposed for the perit idal cycles. Three models have been suggested to explain the origin of shal- lowing-upward, meter-scale cycles: 1) autocyclicity, 2) ep- isodic subsidence, and 3) high-frequency oscil lations in eustatic sea level. Each of the proposed mechanisms must explain the upward shallowing of individual cycles, the repetit ive stacking of similar cycles throughout a vertical sequence, and the simultaneous development of tidal flat- capped cycles and subtidal cycles across a carbonate plat- form.

A utocycl icity

The autocyclic model (Ginsburg 1971; Wilkinson 1982; Hardie et al. 1991) depends upon the periodic progra- dation of tidal flats over the subtidal carbonate factory to restrict the size of the carbonate source area, effectively

SHALE-BASED CYCLES INTRASHELF BASIN

i 3WB/

FIAT-PEBBLE CONGLOMERATE SHALY CYCLE

DEEP SHALY RAMP

KEY TO LITI'-IOLOGIES '~ -~ ,SKELETN.

PACK.STONE

~ FLAT PEBBLE CONGLOMERATE

l ~ PELOIDAL PACK/GRNNSTONE

BURROWED WACKESTONE

GREEN-BROWN SHALE

SL B

iiii!iiiiiiSWB !i!i!ii

SKELETAL PACKSTONE SHALY CYCLE

FIG. 7. - - Late Cambrian shaly cycles of the Conasauga intrashelfbasin of the Appalachians and of the Cordilleran deep ramp of Utah. Siliciclastic shales are abruptly overlain by "clear-water carbonates" with storm-deposited caps. Note the possible shallower position of storm-wave base in the protected intrashelf basin.

1234 DA bTD OSLEGER AND J. FRED READ

TABLE 4.--Shaly deep ramp/intrashe~'basin lithofacies

Flat Pebble Conglomerate (0.1-0.6 m) Bedding Characteristics: Amalgamated irregular thin beds and lenses

cap coarsening-upward cycles; scoured bases common into underlying peloidal grainstones and quartz siltstones; elongate clasts imbricated to edgewise to random orientations; mud drapes separate individual beds within amalgamated units; matrix includes skeletal debris and glauconite; abruptly overlain by shale (Noliehueky) or peloidal silt- stone (Point Peak) lithofacies.

Internal Composition/Texture: Elongate rounded clasts typically com- posed of laminated peloidal grainstone and quartz siltstone of un- derlying lithofacies; clasts less commonly composed of skeletal pack- stone (Nolichucky) or micritic-spiculitic (Point Peak); many clasts have iron-stained rinds and are often bored; matrix between clasts consists of peloids and skeletal debris (abundant brachiopod-trilobite- echinoderm in Point Peak); quartz silt common; some occluded shel- ter porosity and perched peloidal muds.

Peloid Grainstone/Quartz Siltstone (0.4-2.0 m) Bedding Characteristics: Interlaminated calcareous siltstones and silty

peloidal grainstones in thin irregular beds; internal parallel and hum- mocky cross-lamination; typically grades from quartz silty toward the base to dominantly peloidal grainstones at the top of the lithofacies; lower contact with shale lithofacies (Nolichucky Fm only) begins with very thin siltstone beds intercalated within the shale eventually be- coming pure calcareous peloidal siltstone; this facies forms the base of the cycles in the Point Peak where they are platy bedded and. less well-cemented but otherwise identical to the Nolichucky laminated peloidal quartz siltstones; Cruziana trace fossils; coarsens upward into amalgamated flat pebble conglomerates or, less commonly, skeletal- ooid packstones.

Internal Composition/Texture: Peloids and quartz silt are dominant grain types with subordinate finely-comminuted skeletal debris; some laminae show very fine normal grading; fine skeletal debris--domi- nantly echinoderms and trilobites.

Calcareous Shale (0.5-10.0 m) Bedding Characteristics: Olive green to dark gray fissile shale; breaks into small chips upon separation suggesting bioturbation; random, very

thin lime mudstone beds increase in frequency upward in the lithofacies; overlain by calcareous quartz siltstone lithofacies (Nolichucky) or thin nodular wackestone (Candland and Corset Springs); abruptly overlie flat pebble eonglomerate/skeletal-ooid grainstone lithofacies or thrombolitic bioherms.

Internal Composition/Texture: Composed of clay-sized micas and associated clay minerals as well as calcareous micropeloids; occasionally fine trilobite and phosphatic brachiopod fragments.

shutting down carbonate production until tectonic sub- sidence recreates broad shoal-water areas. Implicit in the model are the assumptions of static sea level over tens to hundreds of thousands of years and complete shoaling to tidal levels. Weaknesses in this model are the inordi- nately long lag times (> 20 ky) necessary for the creation of water depth sufficient to resume carbonate production and the assumption of complete non-deposition over tens of thousands of years (Grotzinger 1986b; Koerschner and Read 1989; Read et al. 1991). Perhaps the biggest draw- back to autocyclic control is the simultaneous develop- ment of purely subtidal cycles that, by definition, have no progradational tidal fiat cap that could influence the shrinking of the carbonate factory (Grotzinger 1986b). The inability of the autocyclic model to explain incom- plete shallowing of subtidal cycles that develop seaward of peritidal cycles precludes it as a potential controlling process on Late Cambrian cycle development.

Other autocyclic models invoke 1) the lateral migration of tidal channels to produce shallowing-upward peritidal cycles (Cloyd et al. 1990) and 2) autocyclic responses to "sediment production, tidal variations, and wave and storm activity" to explain the lack of lateral correlatability of Cambro-Ordovician pefitidal cycles in eastern Ten- nessee (Kozar et al. 1990). As with the progradational model (Ginsburg 1971), variations in sediment accu- mulation and redistribution cannot explain the origin of regional subtidal cycles, but may contribute to variability in the internal composition of individual cycles (Osleger 1991). Autocyclic mechanisms may only be viable as an explanation of stratigraphic "noise" within individual cy- cles but probably do not control the development of re- petitive stacks of cycles or the synchronous development of pefitidal and subtidal cycles on Late Cambrian plat- forms.

Episodic Subsidence

Repeated pulses of downfaulting have been proposed (Hardie et al. 1986; Cisne 1986) to generate abruptly the accommodation potential for asymmetric cycle devel- opment. If the stress limits between faulting episodes were rhythmic based on some threshold value, then this model could conceivably explain the coexistence ofpefitidal and subtidal cycles. However, the lateral extent of such events would be limited and could not explain the widespread nature of carbonate cycles across entire platforms (e.g., Demicco 1985; Grotzinger 1986a; Hardie and Shinn 1986). Additionally, modern examples of tectonic pulsing (Yeats 1978; Bull and Cooper 1986; Atwater 1987) are restricted to tectonically active settings, poor analogs for ancient mature passive margins such as existed during Late Cambrian time. Other tectonic mechanisms such as intraplate stress (Cloetingh 1986; Karner 1986) are too slow (0.01-0.1 m/ky) and non-periodic to produce high- frequency meter-scale cycles. It seems hard to conceive of repeated tectonic pulses (each 20 to 200 ky duration) over millions of years to produce repetitive cycles (all within a fairly narrow range of thicknesses) on mature passive margins.

Eustatic Oscillations

High-frequency oscillations in sea level, probably con- trolled by fluctuations in glacial ice volume, provide the simplest explanation for the origin of meter-scale peritidal and subtidal cycles (Fischer 1964; Matthews 1984; Good- win and Anderson 1985; Goldhammer et al. 1987; Koerschner and Read 1989; numerous others). Consid- ering the evidence for eustatic control on third-order se- quence development (Vail et al. 1977; Haq et al. 1987;

EUSTAS Y AND CYCLE $724 CKING PATTERNS OF LATE CAMBRIAN CARBONATES 1235

Ross and Ross 1988; Osleger and Read, unpublished data), it seems likely that higher frequency sea-level fluctuations were superimposed on the longer-term sea level events and, by association, were also eustatic in origin. Super- imposed orders of eustatic sea-level oscillations (com- posite eustasy, Goldhammer et al. 1990) provide the best explanation for the upward shallowing of individual cy- cles, stacking patterns with cyclic successions, and the simultaneous development of peritidal cycles and subti- dal cycles across carbonate platforms.

Although it seems clear that sea level fluctuated eustati- cally to generate individual meter-scale cycles as well as stacked cyclic successions, the forcing mechanism behind high-frequency sea level oscillations is far from certain. It has been proven that Plio-Pleistocene sea levels fluc- tuated in response to variations in global ice volume as a function of changes in solar insolation forced by Mil- ankovitch astronomical rhythms (e.g., Hays et al. 1976; Berger 1977). Temporal association with continental gla- ciations has made glacio-eustasy and Milankovitch or- bital forcing probable as a cause of the Permo-Carbon- iferous cyclothems (Wanless and Shepard 1936; Heckel 1986). It has been more difficult to make a case for Milan- kovitch control on stratigraphic cyclicity in ancient rock sequences deposited during times of more equable cli- mates and no known major glaciations. Van Houten (1964), Olsen (1986) and Anderson (1986) used varve- calibrated sedimentation rates to show Milankovitch pe- riodicities for rocks of Triassic and Permian age. Schwarz- acher and Fischer (1982), Schwarzacher and Haas (1986), and Goldhammer et al. (1987) used a 5: ! recurrence ratio of meter-scale cycles within megacycles, representing the precession signal modulated by the short eccentricity sig- nal, as evidence for Milankovitch control of Mesozoic cycles. Borer and Harris ( 1991) recognized a 4:1 ratio for Permian cycles and suggested that the bundling mani- fested 100 ky short eccentricity cycles superimposed with- in the 400 ky long eccentricity cycle.

Other attempts at showing a Milankovitch influence on ancient cyclic sequences have depended upon the av- erage periodicities of cycles that roughly coincide with the range of Milankovitch periods of 19-23 ky, 41 ky, 95-123 ky, or 413 ky. It has been recognized that the periods of precession and obliquity signals have changed through geologic time due to changing earth-moon rela- tionships, whereas the long and short eccentricity cycles probably have remained constant through time since they are based on interplanetary gravitational forces (Walker and Zahnle 1986; Berger et al. 1989). The ranges of vari- ance (the 21 ky precession signal may have approached 17 ky and the 41 ky obliquity signal may have approached 28 ky during the Early Paleozoic) are insignificant when compared to the large errors associated with absolute age dates for Early Paleozoic rocks. As cautioned by Hardie and Shinn (1986) and Algeo and Wilkinson (1988), cal- culations of average cycle period within the Milankovitch band are to be expected for meter-scale cycles and are insufficient evidence for orbital control on cycle forma- tion.

An objective way of determining cycle periods is by

spectral analysis of cyclic successions where dominant periodicities can be extracted and ratios between the pe- riods can be used to establish Milankovitch control (e.g., Schwarzacher and Fischer 1982; Herbert and Fischer 1986; Kominz and Bond 1990). However, spectral anal- ysis is particularly difficult for shallow platform carbon- ates of Early Paleozoic age for the following reasons. 1) Peaks on the power spectra are difficult to calibrate since long-term accumulation rates used to convert thickness per cycle to time per cycle are dependent upon the ra- diometric time scale with its large uncertainties (Fig. 2). 2) The assumption of constant sediment accumulation throughout the duration of the cyclic succession is un- likely due to differential sedimentation rates for different lithofacies (Kominz and Bond 1990) and the effects of long-term changes in sea level. 3) "'Missed beats" are a common phenomenon of shallow platform carbonates (Hardie and Shinn 1986; Koerschner and Read 1989; Goldhammer et al. 1990), resulting in a noisy spectrum. 4) Peritidal cyclic successions are a poor proxy for time, because much of the cycle period is taken up by non- deposition (Read et al. 1986; Read et al. 1991). Spectral analysis of Early Paleozoic shallow platform carbonates may only be viable in conjunction with techniques for deriving better time series such as gamma analysis (Ko- minz and Bond 1990), a method that deserves further testing.

Late Cambrian Eustasy and Milankovitch Rhythms

Evidence from the Cordilleran passive margin suggests that Milankovitch orbital variations may have controlled low-amplitude glacio-eustatic fluctuations during the Late Cambrian. In the Big Horse Member of the Orr For- mation of the House Range, meter-scale fifth-order cycles are stacked into shallowing-upward successions at the fourth-order scale as well as at the third-order scale (Fig. 8). The Big Horse Member comprises the upper portion of one long-term third-order shallowing-upward sequence (220 m thick; approximately 4.8 m.y. duration). The long- term sequence is composed of stacked deep ramp cycles in the lower portion gradually shoaling up to stacked shallow ramp cycles with large thrombolite bioherrns marking the top of the sequence. This third-order se- quence has superimposed within it 11 fourth-order de- positional cycles (l 5-45 m thick; average of ~ 440 ky) that are typically composed of three to four fifth-order cycles (0.5-8.0 m thick; average of ~ 96 ky). The low- ermost fifth-order cycle within each fourth-order bundle is typically the thickest and is dominated by deeper water lithofacies. The fifth-order cycles gradually thin upward and shallow upward within each fourth-order bundle.

The 4:1 bundling is illustrated in a mirror plot of de- viations from average cycle thickness within the Big Horse Member (Fig. 9). The mirror plot is simply a tracing of a Fischer plot of the Big Horse Member (discussed below) and was created to accentuate the bundled nature of the cycles. Assuming that the estimate of long-term accu- mulation rate (0.046 m/ky) derived from the DNAG time scale is reasonable, the 4:1 bundling may manifest the

1236 DA VID OSLEGER AND J. FRED READ

250

On"

I ,~'] I o . j ~o ,

BIG HORSE MEMBER Big Horse Member ORR FORMATION House Range, Utah

180 o OoO . ~ - -

,o~oo- ~ j 4

170-, '~" [ , ~1 ! A! - _' 2. " M, ~o*O=O

- - B

- _~ ? ~ "~ 3O

Fie. 9.--Mirror plot of deviations from average cycle thickness in the Fio. 8.--Hierarchy of cycles within the Big Horse Member, Orr For-

mation, House Range, Utah. Column on the left shows long-term third- order shallowing evident from the storm-influenced deep ramp cycles with open marine faunas in the lower Big Horse progressively giving way to shallow subtidal cycles characterized by restricted lithofacies upward in the Big Horse Member. Dashes to the right of the left column denote generalized fourth-order cycles that are shown in derail in the columns on the right. Composition of the fourth-order cycles suggests rapid deepening in the basal cycle followed by progressively shallower conditions toward the upper cycles. Note the 4:1 bundling of fifth-order cycles within fourth-order sets.

short eccentrically (95-123 ky) to long eccentricity (413 ky) ratio. The bundles exhibiting a 3:1 ratio may simply have missed a cycle beat, perhaps as a low ampl i tude sea- level event oscil lated above the platform with no apparent sedimentologic response.

No evidence can be recognized within the fifth-order eccentricity cycles for superimposed cyclicity that may represent the precession or obl iquity signals. Because the Cordil leran passive margin extended essentially E -W at about 10 to 15*N during the Late Cambrian (Scotese and McKerrow 1990), the lack of a 41 ky obliquity cycle is to be expected because the effect of changing axial tilt is

Big Horse Member, Orr Formation, House Range. The plot is simply a tracing of the Fischer plot shown in Figure 13 and oriented vertically. Fourth-order bundles are marked by the abrupt appearance of thick basal cycles over thinner cycles below.

minimal toward low latitudes (Berger 1978). The appar- ent lack of a precession signal may be due to a number of interrelated factors. 1) The relative ampl itudes of the precession-generated sea level signals may have been low compared to those of the dominant eccentricity cycle. 2) Perhaps the deposit ional setting o f the Cordi l leran shal- low to deep ramp was not a sensitive enough recorder of each individual sea level event. Only the higher-ampli- tude ~ 100 ky sea-level event may have caused a sedi- mentologic response on the carbonate platform in the vicinity of the House Range. 3) Phase relations of the interacting Mi lankovitch frequencies may have sup- pressed the precession signal due to destructive interfer- ence. It seems likely that at certain t imes in the geologic past, constructive interference has acted to enhance the individual Mi lankovitch frequencies and, conversely, de- structive interference has acted to mask the Mi lankovitch frequencies. Phase relations may provide a partial expla- nation of cyclic successions with no evidence of bundling

EUSTASY AND CYCLE ST.4CKING PATTERNS OF LATE CAMBRIAN CARBONATES 1237

or for weakly cyclic intervals within an overall strongly cyclic section.

Glacio-Eustasy During the "Non-Glacial" Late Cambrian

The connection between Milankovitch orbital varia- tions, the shrinkage and growth of continental ice sheets and eustasy has been well-documented (e.g., Berger et al. 1984). However, a direct link between changes in solar insolation related to Milankovitch astronomical rhythms and changes in sea level and sedimentation during glob- ally warm periods of Earth history has yet to be found (Barron et al. 1985). To account for the low to moderate amplitude (perhaps 15-25 m based on 2-D modelling) sea-level oscillations proposed to simultaneously generate Late Cambrian peritidal and subtidal cycles, a sink for the storage and release of moderate volumes of seawater needs to be identified.

Paleogeographic reconstructions for the Late Cambrian place most continental land masses between 60N and S latitudes (Scotese and McKerrow 1990). Only Baltica and the southern margin of Gondwana extend into higher southern latitudes where climates may have been signif- icantly cooler than the generally warm global climate. Ziegler et al. (1981) have suggested that the paleogeo- graphic configuration of the continents during the Cam- brian facilitated a latitudinal zonation of prevailing winds and ocean currents within the high latitudes that may have reduced the absorbtion of solar radiation, enhancing the possibility of cooler Cambrian climates than previ- ously believed. Additionally, climate modelling of pre- sumably warm periods of Earth history suggest that the interiors of mid- to high latitude continents may have had subfreezing temperatures and that no global climate is truly "equable" (Sloan and Barton 1990). Even though no major large-scale continental glaciers existed during the Late Cambrian, diamictites and striated cobbles have been reported in lower Tremadocian strata of Argentina and Bolivia (Erdtmann and Miller 1981) which were lo- cated in a part of Gondwana believed to have experienced cool climates during the Late Cambrian-Early Ordovician (Scotese and McKerrow 1990). Alpine glaciers may have been present in ancestral mountain belts of continental interiors of major land masses and provide a possible sink for small portions of the 20 (_+5)-meter sea-level oscillations estimated for the Late Cambrian meter-scale cycles. However, the apparent absence of a reservoir large enough for the rapid storage and release of moderate vol- umes of seawater remains a major weakness in the con- nection between Milankovitch orbital variations and Late Cambrian meter-scale cyclicity.

STACKING PATTERNS OF METER-SCALE CYCLES

Characteristic meter-scale fifth-order cycles system- atically change upward within third- and fourth-order sequences and define distinct stacking patterns. Cycle stacking patterns provide the crucial link between the

20

1

og lo ._>

Average Cycle Duration Path of Relative

I , . ~ C, hange in Sea

Subsidence

100 200 300 ~0 T~e (ky)

FiG. 10.--Explanatory diagram of the Fischer plot technique. The horizontal scale of the plot represents time and the vertical scale is the cumulative cycle thickness in meters. For each cycle the amount of accommodation space provided by linear subsidence is plotted over the duration of the average cycle period. Cycle thickness is plotted vertically. The net difference can be interpreted to define the change in accom- modation space through time.

meter-scale cycles and the larger scale sequences and their component systems tracts. Stacks of genetically related cycles are the parasequence sets of sequence stratigraphic terminology.

Fischer plots (Fig. 10) illustrate deviations from av- erage cycle thickness throughout a stratigraphic interval. They can be interpreted as graphic displays of relative changes in accommodation space through time (Fischer 1964; Goldhammer et al. 1987, 1990; Read and Gold- hammer 1988; Read 1989). Each fifth-order cycle is as- signed an average cycle duration by dividing the total estimated duration of the cyclic succession by the number of meter-scale cycles. This average cycle duration is mere- ly a device for assigning time per cycle and does not imply that each cycle was actually deposited over the same du- ration. The horizontal axis could just as easily be divided into equivalent units that equal "cycle number". If the plot was constructed so that cycle thickness equaled time, the resulting plot would define a horizontal line. Thus it is necessary to assign a constant time of deposition to each cycle to generate relative rises and falls on the plot.

Interpretation of individual Fischer plots should only be made in unison with temporally equivalent Fischer plots that show similar patterns of rises and falls (Koerschner and Read 1989; Read et al. 1991). Fischer plots of Late Cambrian cyclic strata have been correlated between the Cordilleran and Appalachian sections and provide excellent evidence for eustatic control on third- order sequence development (Osleger 1990; Osleger and Read, unpublished data). The correlated plots suggest that stacks of thick cycles plot as positive slopes and are pre- sumed to have formed under conditions of increased ac- commodation space provided by relative sea-level rise. Stacks of thin cycles plot as negative slopes and are pre- sumed to reflect reduced accommodation space during

1238 DA I,'ID OSLEGER AND J. FRED READ

m

15

14.1 > W - J

Ut W >

5

CONOCOCHEAGUE FORMATION KEY TO LITHOFACIES ~- - - - - CRYPTALGAL LAMINITE

WYT.EWt.LE. VIRG,NIA M,N,T I I ~ 'X I"" ~ ~7 RIBBON ROCK ,~I ~ ~ l~7.~"~,~q THROMBOLITE BOUNDSTONE

O-

10

5 t '~ , ,O ,1

-:2 ~ _ Om

m

L_J

- - -W'--w--

- -7 - - - v

,X~, :o , ' , . oJQ=

I ~ ~s~' l - -

0 m -~.~. .~. .~_~j __ 0m

A) SUBTIDAL-DOMINATED B) LAMINITE-DOMINATED C) PERITIDAL CYCLES PERITIDAL CYCLES PERITIDAL CYCLES WITH QUARTZ SAND

F]G. 11 .--Fischer plot of the Conococheague Formation constructed from the Wytheville, Virginia section (from data in Koerschner and Read 1989). Cycles containing quartz sand are black. Stacking patterns of representative cycles are shown pulled out from their position on the Fischer plot. Note the difference in scales between the three columns of cycles and how the subtidal-dominated cycles are considerably thicker than the peritidal-dominated cycles.

relative sea level fall. The method seems to be best suited for peritidal cycles or subtidal cycles that shallow to near sea level.

Fischer Plots and Peritidal Successions

The relationship between cyclic peritidal carbonates of the Conococheague Formation of southwestern Virginia and long-term relative sea-level events defined by its Fi- scher plot (Koerschner and Read 1989) is shown in Figure 11. The Conococheague Formation is composed of hun- dreds of stacked peritidal cycles that record periodic, high- frequency fluctuations in relative sea level (Demicco 1985; Koerschner and Read 1989). The Fischer plot defines a major relative sea-level rise and fall within this portion of the Conococheague Formation. Stacking of thick, sub- tidal-dominated cycles with thin laminite caps occurs during the rising portions of the plot (Fig. 11, column A).

Stacking of thin, laminite-dominated cycles occur during falling portions (Fig. 1 1, column B). Brecciated cycle caps and quartz sands become common toward the troughs on the plot (Fig. l 1, column C).

Similar cycle stacking patterns are exhibited in the peri- tidal Allentown Formation of eastern Pennsylvania (Fig. 12). During long-term rises on the plot, cycles are thick with oolitic bases and thin stromatolitic caps (Fig. 12, column B). Ooids have dropped cores and cycle caps are brecciated indicating meteoric diagenesis during episodic short-term emergence. During long-term falls on the plot, cycles show thin oolitic transgressive lags overlain by thrombolites that grade up into LLH stromatolites and cryptalgal laminites (Fig. 12, columns A and C). The caps o f many cycles are marked by regolithic breccias devel- oped on the emergent tidal flat. Toward the troughs on the plot (Fig. 12, column C), erosionally-capped cycles contain quartz sand.

E I~S'TAS Y AND ( 'YCLE STACKING PA TTERNS OF LATE CAMBRL4N CARBONATES 1239

>

<

<

KEY TO LITHOFAGIES ALLENTOWN FORMATION ~ R~.~X~ e^P .

I" "- ' " ; " "I OOMOG~SrONE /~I

" N ""s

' _ . . .

B) SUBTIDAL-DOMINATED PERmDAL CYCLES

o-

lo-

20-

A) TIDAL FLAT-DOMiNATED PERmDAL CYCLES

\

3- T:~ " "~~.~ -

. . . . " " ' ' ' ' " O

". :4 : . I Om C) PERmDAL CYCLES

WITH QUARTZ SAND

FIG. 12.--Fischer plot of the lower Allentown Formation constructed from the Easton, Pennsylvania section with cycle slacking patterns expanded from their position on the plot. Small dots below individual cycles on Fischer plot denote regolithic cycle caps. Note the variation in scales between the three intervals and the relative thicknesses of the component cycles. Oolitic grainstone bases of cycles on the rising portions of the Fischer plot are considerably thicker than those on the falling portions. Tidal flat caps are considerably thinner on cycles that formed during the relative sea-level rise but dominate in the cycles that formed on the relative sea level fall.

The Conococheague and Allentown Fischer plots il- lustrate the significant difference in cycle thickness and lithofacies composition between stacks of cycles gener- ated during rising and falling relative sea level. Peritidal cycle thickness is controlled by the total amount of ac- commodation space provided by subsidence and eustasy. For these peritidal cycles, stacks of thicker cycles were formed during relative sea level rise that generated ac- commodation space beyond that provided by subsidence. Stacks of thinner cycles were formed during relative sea level fall that reduced accommodation space provided by subsidence. Quartz sands were brought in and brecciated laminite caps were developed during relative sea-level lowstands that exposed craton interiors and the inner platform. Assuming relatively constant tectonic subsi- dence, the control on the long-term changes in relative

sea level is believed to be eustasy, on the basis of cor- relation of the Fischer plots above with equivalent sec- tions in the Appalachians and Utah (Osleger and Read, unpublished data).

Fischer Plots and Subtidal Successions

Stacking patterns of dominantly subtidal cyclic succes- sions and their relationship to Fischer plots are shown on Figure 13. The Fischer plot of the upper Big Horse and Candland Shale Members of the Late Cambrian Orr Formation of the House Range shows stacks of deeper subtidal cycles on the rising segments of the plot. These stacks of genetically-related cycles are characterized by storm-dominated carbonate cycles (Column A) or thick, deep ramp, shale-based cycles (Column C). Falling por-

1240 DAVID OSLEGER AND J. FRED READ

ORR FORMATION, HOUSE RANGE, UTAH

20-

. J uJ > lO-

w W 0-

~ -10 W

-20

10 ~ ~ I O ' - -

Om ,

A) STORM-DOMINATED DEEP RAMP CYCLES

(BIG HORSE)

KEY TO LffHOFACIES * * ~ ' ,~)OOID-ONCOLITE GRST

I ~ ~/ SKELETALPKST

[ . ~ BURROWED WKST Lil~mlml/ A RGI LLACEOUS WKST ~ O L I V E GREEN SHALE

= T IME >

tO - e .~. . __ . .~ . . " J

E m

o o o~

/ 0m

B) SHALLOW SUBTIDAL CYCLES WITH OOID GRAINSTONE CAPS (BIG HORSE)

C) SHALY CYCLES WITH SKELETAL STORM BED CAPS (CANDLAND SHALE)

Fxo. 13.--Fischer plot of the upper Big Horse and Candland Shale Members of the Orr Formation, House Range, Utah. The scale is the same for all three stacks of cycles; note the thin shallow subtidal restricted cycles versus the substantially thicker, deeper subtidal, open marine carbonate and shaly cycles.

tions of the Fischer plot are characterized by thin, oolite grainstone-capped cycles (Column B). Common lithofa- cies within these shallow ramp cycles are oncolitic pack- stones-grainstones, ooid grainstones, thrombolite bio- herms and SH and LLH stromatolites, all indicative of shallow, restricted conditions.

Fourth-order cycles on the Fischer plot of the Big Horse Member occur as bundles of one thick cycle followed by two to three thinner cycles (Figs. 8, 9 and 13). The mirror plot of Figure 9 was created from the Fischer plot of Figure 13 and better illustrates the 4:1 bundling of meter-scale cycles within the Big Horse Member. Thicknesses for these fourth-order cycles range from 15 to 45 m and their average duration is ~ 440 ky. Cycles on the fourth-order rises are consistently composed of thick cycles dominated by deep subtidal lithofacies whereas fourth-order falls are consistently composed of thin cycles dominated by shal- low subtidal lithofacies (Fig. 8). This is essentially the same stacking pattern recognized within the third-order sequence only repeated over shorter time increments.

The systematic arrangement of similar subtidal cycles on rising and falling limbs of Fischer plots suggests that,

like the peritidal cycles, they record changes in accom- modation space generated by third-order relative sea-lev- el fluctuations. This suggests that time-equivalent suc- cessions of meter-scale peritidal or shallow subtidal cycles may be correlated using Fischer plots and combined to define third-order, and perhaps fourth-order, sea-level events. I fa good degree ofcorrelatibility can be attained between geographically distinct sections, then the long- term fluctuations may be considered to have been eustatic in origin (Osleger and Read, unpublished data).

Effect o f Sedimentation Rate on the Form o f Fischer Plots

Some deeper subtidal successions of cyclic carbonates show Fischer plots whose trend is opposite to that ex- pected from the above examples. Within the Notch Peak Formation of Utah, a thick succession of shallow subtidal cycles capped by thrombolite bioherms shallows to tidal depths before grading up into a series of deeper water cycles capped by spiculitic wackestone (Fig. 14). Using stacking patterns established from other similar cycle

EUSTASY AND CYCLE STACKING PA TTERNS OF LA TE CAMBRIAN CARBONATES 1241

,< W Q. ~

O z ~ -

i - -

~ - ~ C ~ _

A A A ~

: t - - DEEP RAMP, m SPICULITIC -- WACKESTONE -- CYCLES

PERITIDAL _ CYCLES 7

- 7 r

- \ \ \ \

\ \

THROMBOLITE- ! _ STROMATOLITE

BOUNDSTONE / CYCLES "3

WJ

m

DEEP TIDAL SUBTIDAL FLAT \

\

/ /

c \ 45 30 15 0'm RELATIVE "~

WATER / CUMULATIVE CYCLE THICKNESS DEPTH

FIG. 14.--Vertically-oriented Fischer plot of the upper Notch Peak Formation, House Range, Utah. Dashes to right of strat column indicate cycle tops. Groups of like cycles are noted with arrows. Key horizons are connected to the Fischer plots by the dashed lines; they are not perfectly horizontal because the stratigraphic column is in thickness and the Fischer plot is in time. Interpreted paleowater depth curve to the fight is shown for comparison.

1242 DAVID OSLEGER AND J. FRED READ

A) MODEL SEA LEVEL CURVE \ ^

'r

m 0 100 200 Q

I- EMERGENT .~ WATER SUPRATIDAL SURFACE

DEPTHS / AGGRADING I EO,MENT ,.SEA EVE

~ ' ~ , ~ \ \ ' ~ " ~SHALLOW SUBTIDAL " ~ ' ~ . . . ~ " 'TIDAL FLAT I:ACiES . . . . FACIES

~" -SUBSIDENCE I I

1 O0 200 TIME (KY)

FIG. 15A.--Explanatory diagram of the 1-D modeling. The sea level curve is composed of in-phase symmetrical 20 and 40 ky periods and asymmetrical 100 ky periods superimposed on a long-term sea-level rise/fall. Any combination of amplitudes of sea level cycles can be input and define the vertical axis. Sloping lines to lower right represent linear subsidence of deposited sediments through time. The lines sloping to the upper right represent the aggrading sedment surface and changes in slope reflect differing sedimentation rates of water depth-dependent lithofacies. The period of non-deposition following drowning is a pre- determined lag time.

types, one would expect the Fischer plot of the restricted thrombolitic cycles to be associated with a long-term fall in sea level, whereas the plot of the deep water cycles would be expected to form a long-term rise in sea level. However, the thick thrombolitic cycles plot as a positive slope on the Fischer plot, whereas the thin spiculitic wackestone cycles plot as a negative slope.

One reason for this counterintuitive result may be that the relative thicknesses of subfidal cycles are controlled by sedimentation rate rather than by sea-level-determined accommodation space. Thick thrombolitic cycles may have accumulated rapidly within the zone of optimal car- bonate productivity, rapidly filling to near sea level. In contrast, the thinner, argillaceous, deeper water cycles may simply have accumulated slowly. The resulting trend on the Fischer plot is an apparent long-term rise and fall in sea level generated by water-depth-dependent sedi- mentation rates of the cycles rather than by sea-level- controlled accommodation space.

A second possible interpretation is that the thickness of subtidal cycles may indeed be controlled by accom- modation space but that the upper limit to vertical ag- gradation may be the base of normal fairweather or storm- wave reworking rather than tidal level (Osleger 1991). Subtidal cycle thickness could be limited by reworking and redistribution of sediment when the depositional sur- face intersects an energy barrier associated with a zone of active wave or storm-current winnowing. The deep ramp cycles of the upper Notch Peak are capped by thin lenses of skeletal packstone storm beds, which suggests

that the base of storm reworking may have precluded any further vertical aggradation. Consequently, the resulting cycles are thin and define an anomalous negative slope on the Fischer plot of Figure 14.

This example illustrates that Fischer plots of mixed cycle types must not be interpreted alone but should be used in conjunction with time-equivalent plots to deter- mine their value. Caution must be exercised when inter- preting the form of individual Fischer plots of subtidal cycles without looking at the internal composition of the cyclic succession as well as at other plots of coeval inter- vals.

MODELLING OF CYCLE STACKING PATTERNS

One- and two-dimensional computer modelling are valuable techniques for assessing the effects of controlling variables on the generation of cyclic sequences and for testing the feasibility of models related to the origin of meter-scale cycles. One-dimensional models (Read el al. 1986) graphically track the simultaneous interaction of eustatic sea level, the sediment surface, and long-term subsidence to generate synthetic stratigraphic columns. Two-dimensional models (Koerschner and Read 1989; Read et al. 1992) integrate a more sophisticated set of parameters to produce synthetic geologic cross-sections that simulate the vertical and lateral facies distribution of cycles and the internal geometry of longer-term de- positional sequences. The model types can be used to complement one another by showing the simpler concepts with the one-dimensional modelling and then reproduc- ing a more detailed, more realistic simulation of actual stratigraphic data using the two-dimensional modelling.

The one-dimensional models (Read et al. 1986) incor- porate linear long-term subsidence, simplified sea-level curves composed of high-frequency in-phase oscillations superimposed on a longer-term rise/fall, water-depth-de- pendent sedimentation rates of lithofacies, and lag time after flooding to produce synthetic stratigraphic columns (Fig. 15). The two-dimensional models (Read et al. 1992) (Fig. 16) incorporate many of the same variables as the one-dimensional models but with significant refinements. Antecedent topography and platform slope is digitized before the program runs and is constrained by modern analogs of carbonate platform morphologies. The model divides the platform into 200 localities whose increment width varies with the pre-determined length of the plat- form. Tectonic subsidence is separated into regional and rotational components, and isostatic subsidence is cal- culated for each time slice to account for sediment and water loading. The synthetic cross-sections produced are truly two-dimensional because the isostatic response to sediment and water loading at single localities affects ad- jacent localities along the elastic beam 200 km on either side (Read et al. 1991).

The form of the eustatic sea-level curve can be gen- erated by any combination of high-frequency, asymmet- ric or symmetric sine waves superimposed on a long-term sine wave or digitized curve. The input values for the sea-level curve allow for any combination of cycle periods

EUSTASY AND CYCLE STACKING PA TTERNS OF L4 TE CAMBRIAN C4RBONA TES 1243

PERITIDAL CYCLES ~. 40- ~! E J 30- UJ > LU 20- .2

m 10" '~" . . . . . . . . . . ' - 03 I . . . . . . . . . . . ~---

1244

0rn A

DA V1D OSLEGER AND d. FRED READ

PERITIDAL PLATFORM B

_ _ I

SU BTIDAL PLATFORM

20 t V.E.=5000

40 Okra 160 260

C

D I

E 30 v ._1 I..U > 20- uJ

< 10- uJ

0 200 400 600 800

TIME (k.y.)

KEY TO LITHOFACIES I~1 TIDAL FLAT I I SHALLOW SUBTIDAL

DEEP SUBTIDAL DEEPEST SUBTIDAL

FIo. 16A.--2-D model ofa peritidal to subtidal transition across a hypothetical platform. Water depths and sedimentation rates of facies are the same as in the i-D models of Figure 15. Amplitudes of the sea level oscillations are also the same as in the I-D models but the periods have been input as 19, 23, 41 and 100 ky and allowed to interfere to produce the complex sea level curve in the inset. Initial slopes on the peritidal platform are < 0.01 m/kin and are ~ 0.04 m/km on the subtidal platform, comparable to modem carbonate platforms. The apparent abrupt break in slope around 550 km is an artifact of the vertical exaggeration (5000) and translates to 0.2 m/kin or a fraction of a degree. Rotational subsidence at the outer edge of the platform is 0.015 m/ky. Duration of the run is 800 ky and time lines are denoted every 200 ky.

posed upon a third-order "driver" (Goldhammer et al. 1990). The lithologic composition of the synthetic cycles was highly irregular and totally unlike Late Cambrian cycles observed in the field. Narrowing the range of ran- dom periods alleviated the problem to an extent, but stratigraphic trends of thickening and thinning cycles were absent. The experiments suggest that periods of Late Cambrian sea-level oscillations were constrained within relatively narrow ranges.

Previous discussion of the origin of meter-scale cycles has shown composite eustasy to be the most likely mech- anism controlling cycle development. The 4:1 bundling of Late Cambrian cycles in the Big Horse Member sug- gests sea-level control by Milankovitch orbital variations. On the assumption of Milankovitch-forced glacio-eu-

stasy, sea-level fluctuations with cycle periods of 19, 23, 41 and 100 ky were used in the modelling.

Amplitudes of Relative Sea Level Oscillations. -- Am- plitudes of relative sea-level fluctuations that generated the Late Cambrian meter-scale cycles must account for the simultaneous development of peritidal cycles as well as deep ramp shaly cycles that formed on different parts of the Late Cambrian platforms. For peritidal cycles where the uppermost datum is known to be tide level, the av- erage cycle thickness may be a minimum approximation of the total amount of accommodation space created by the combined effects of subsidence and sea level (Grot- zinger 1986a; Goldhammer et al. 1987; Koerschner and Read 1989).

Stratigraphic thickness of subtidal cycles cannot be used

FiG. 16B.--Columns of stacked synthetic cycles generated in the 2-D model of Figure 16A. Column A is from the inner peritidal platform; column B is from the outer peritidal platform; column C is from the inner subtidal platform; column D is from the outer subtidal platform. Actual stacked cycles of Late Cambrian cyclic successions are aligned below the synthetic cycles for comparison. Key to lithofacies is the same as in Figures 15 and 16A.

F.U,S'TASY AND CYCLE STAUK1NG PATTERNS OF LATE CAMBRIAN CARBONATES 1245

A) INNER B) OUTER C) INNER D) OUTER PERITIDAL PERITIDAL SUBTIDAL SUBTIDAL PLATFORM PLATFORM PLATFORM PLATFORM

48-

25-

0m- ALLENTOWN FORMATION

PERITIDAL CYCLES

SNEAKOVER MBR. ORR FORMATION SUBTIDAL CYCLES

52 ~ ' -

II II

]

D

m

25

0m CANDLAND SHALE MBR.

ORR FORMATION DEEP RAMP SHALY CYCLES

1246 DA V1D OSLEGER AND J. FRED READ

to approximate the amplitudes of relative sea level os- cillations. The only way to estimate the amplitudes of the high-frequency relative sea-level oscillations that gener- ated subtidal cycles is to use the difference in water depths between estimated storm-deposited and fairweather-re- worked lithofacies. The base of storm wave reworking may be defined geologically by the first appearance of hummocky cross-stratified carbonate packstones and grainstones with interbedded wackestones above suspen- sion-settled carbonate mudstones and siliciclastic shales. The base of fairweather wave reworking may be defined geologically by the transition from storm-influenced sandy muds upward into winnowed carbonate sands. Geological estimates of fairweather- and storm-wave base are spec- ulative and must be based on oceanographically-defined modern analogs.

Fairweather wave base has been estimated at 10 to 20 m on the Yucatan shelf(Logan et al. 1969) and 8 to 20 m in the Persian Gulf(Purser and Evans 1973). The Cor- dilleran passive margin may have fronted a semi-enclosed ocean basin (Stewart and Suczek 1977) where storm wave base may have been approximately 60 m minimum, based upon the semi-enclosed Yucatan platform (Logan et al. 1969).

Potential amplitudes of short-term sea-level oscilla- tions can be estimated using the ranges between fair- weather and storm wave base. Estimating the maximum range to be about 50 m (60 m storm wave base minus 10 m fairweather wave base) and the minimum range to be about 10 m (30 m storm wave base minus 20 m fair- weather wave base), a reasonable mid-range might be about 30 m. Several reasons exist for even this mid-range value of about 30 m to be too high. One constraint on the total amplitude of the high-frequency sea level oscil- lations is provided by the composition of Appalachian peritidal cycles. The extent of exposure ofperitidal cycles is dependent on the amplitude, and therefore the rate of fall, of the sea-level fluctuation. Rapid sea-level falls would preclude the development of thick tidal flat caps (Koer- schner and Read 1989) and result in dominantly subtidal, disconformity-capped cycles similar to the Plio-Pleisto- cene of the Bahama platform (Beach and Ginsburg 1980) and the Quaternary of south Florida (Perkins 1977). Sea- level fall rates had to have been reasonably slow to allow for the accumulation of tidal fiat caps that average be- tween 1 to 2 m in thickness. Koerschner and Read (198