Sequence stratigraphy of the Dakota Formation (Cenomanian

Sequence stratigraphy of the Dakota Formation (Cenomanian),southern Utah: interplay of eustasy and tectonicsin a foreland basin

DAVID ULICÏ NYÂ

Department of Geology, Charles University, Albertov 6, 128 43 Prague 2, Czech Republic(E-mail: [email protected])

ABSTRACT

The Dakota Formation in southern Utah (Kaiparowits Plateau region) is a

succession of ¯uvial through shallow-marine facies formed during the initial

phase of ®lling of the Cretaceous foreland basin of the Sevier orogen. It records

a number of relative sea-level ¯uctuations of different frequency and

magnitude, controlled by both tectonic and eustatic processes during the Early

to Late Cenomanian. The Dakota Formation is divided into eight units

separated by regionally correlatable surfaces that formed in response to relative

sea-level ¯uctuations. Units 1±6B represent, from bottom to top, valley-®lling

deposits of braided streams (unit 1), alluvial plain with anastomosed to

meandering streams (2), tide-in¯uenced ¯uvial and tide-dominated estuarine

systems (3A and 3B), offshore to wave-dominated shoreface (4, 5 and 6A) and

an estuarine incised valley ®ll (6A and 6B). The onset of ¯exural subsidence

and deposition in the foredeep was preceded by eastward tilting of the

basement strata, probably caused by forebulge migration during the Early

Cretaceous, which resulted in the incision of a westward-deepening

predepositional relief. The basal ¯uvial deposits of the Dakota Formation,

®lling the relief, re¯ect the onset of ¯exural subsidence and, possibly, a eustatic

sea-level rise. Throughout the deposition of the Dakota Formation, ¯exure

controlled the long-term, regional subsidence rate. Locally, reactivation of

basement faults caused additional subsidence or minor uplift. Owing to a

generally low subsidence rate, differential compaction locally in¯uenced the

degree of preservation of the Dakota units. Eustasy is believed to have been the

main control on the high-frequency relative sea-level changes recorded in the

Dakota. All surfaces that separate individual Dakota units are ¯ooding surfaces,

most of which are superimposed on sequence boundaries. Therefore, with the

exception of unit 6B and, possibly, 3B, most of the Dakota units are interpreted

as depositional sequences. Fluvial strata of units 1 and 2 are interpreted as low-

frequency sequences; the coal zones at the base and within unit 2 may

represent a response to higher frequency ¯ooding events. Units 3A to 6B are

interpreted as having formed in response to high-frequency relative sea-level

¯uctuations. Shallow-marine units 4, 5 and 6A, interpreted as parasequences

by earlier authors, can be divided into facies-based systems tracts and show

signs of subaerial exposure at their boundaries, which allows interpretation as

high-frequency sequences. In general, the Dakota units are good examples of

high-frequency sequences that can be misinterpreted as parasequences,

especially in distal facies or in places where signs of subaerial erosion are

missing or have been removed by subsequent transgressive erosion. Both low-

and high-frequency sequences represented by the Dakota units are stacked in

an overall retrogradational pattern, which re¯ects a long-term relative sea-level

rise, punctuated by brief periods of relative sea-level fall. There is a relatively

Sedimentology (1999) 46, 807±836

Ó 1999 International Association of Sedimentologists 807

major fall near the end of the M. mosbyense Zone, whereas the base of the

Tropic shale is characterized by a major ¯ooding event at the base of the

S. gracile Zone. A similar record of Cenomanian relative sea-level change in

other regions, e.g. Europe or northern Africa, suggests that both high- and low-

frequency relative sea-level changes were governed by eustasy. The high-

frequency relative sea-level ¯uctuations of »100 kyr periodicity and »10±20 m

magnitude, similar to those recorded in other Cenomanian successions in

North America and Central Europe, were probably related to Milankovitch-

band, climate-driven eustasy. Either minor glacioeustatic ¯uctuations,

superimposed on the overall greenhouse climate of the mid-Cretaceous, or

mechanisms, such as the ¯uctuations in groundwater volume on continents or

thermal expansion and contraction of sea water, could have controlled the

high-frequency eustatic ¯uctuations.

Keywords Cenomanian, foreland basin, sea-level, sequence, Western Interior.

INTRODUCTION

The sedimentary in®ll of a foreland basin primar-ily represents a record of orogen tectonics,¯exural subsidence of the foreland, rates ofweathering and sedimentation and eustatic sea-level change. Field studies as well as numericalmodels of relative sea-level changes in forelandbasins focus primarily on the relative roles oftectonic subsidence (or uplift) and eustasy incontrolling the accommodation and formation ofstratal packages and unconformities. Whereaslong-term cycles of relative sea-level change(2±10 Myr) are generally thought to be controlledby accretionary wedge growth and long-termerosion rates in the orogen, short-term sea-level¯uctuations (0á01±0á5 Myr) are generally attribut-ed to high-frequency eustasy, thrust events orchanges in intraplate stress levels (Peper, 1993).The driving mechanisms of the third- to ®fth-order relative sea-level ¯uctuations (sensu Vailet al., 1991) are a subject of ongoing controversyin sequence stratigraphy (cf. Christie-Blick &Driscoll, 1995).

The Dakota Formation in southern Utah is awell-exposed succession of ¯uvial and shallow-marine sediments formed during the initial phaseof ®lling of the Cretaceous foreland basin ofthe Sevier orogen. This paper focuses on thesequence-stratigraphic aspects of the DakotaFormation in the Kaiparowits Plateau region(including the Paria River Amphitheatre, Fig. 1).Compared with the westernmost part of the fore-land basin ®ll, adjacent to the orogenic front, thisregion was characterized by generally low rates ofsubsidence, and the depositional systems herewere very sensitive to even small-scale relativesea-level changes (Shanley & McCabe, 1995).

The facies associations and their boundingsurfaces are either physically traceable orcorrelatable throughout the region and offer anopportunity to study in detail the record ofrelative sea-level changes. The Dakota Formationin the Kaiparowits Plateau region has beenstudied most recently by Gustason (1989), amEnde et al. (1991) and Elder et al. (1994). Thispaper originated as part of a comparative study ofthe record of Middle and Late Cenomanian sea-level ¯uctuations in the US Western Interior andcentral Europe (preliminary results in UlicÏnyÂ,1994).

TECTONIC AND PALAEOGEOGRAPHICSETTING

During deposition of the Dakota Formation, theKaiparowits region lay at the western margin ofthe Western Interior Seaway, 140±200 km east ofthe thrust front of the Sevier orogenic belt (Fig. 1aand b). The Western Interior foreland basinformed through ¯exural subsidence resultingfrom loading of the cratonic lithosphere of NorthAmerica by low-angle thrust sheets of the Sevierorogen (e.g. Armstrong, 1968; Cross, 1986). Owingto the high ¯exural rigidity of the thermallymature western North American continentallithosphere, the foreland basin was relativelyshallow and broad (Angevine & Heller, 1987;Bally, 1989). In addition to ¯exural subsidence,however, sedimentation in the basin was affectedby normal faulting caused by load-induced reac-tivation of basement fault zones inherited fromthe Precambrian rifting and fragmentation ofthe North American continent (Picha, 1986;Schwans, 1995).

808 D. UlicÏnyÂ

Ó 1999 International Association of Sedimentologists, Sedimentology, 46, 807±836

Clastic material of the Dakota Formation insouthern Utah was shed generally south-east-wards from the NE- to SW-trending part of theSevier belt (Peterson, 1969; Gustason, 1989;Fig. 1). Wolfe (1989) showed that, near theArizona±New Mexico border, the sedimentsource during the mid-Cenomanian was fromthe Mogollon Highlands, a NW- to SE-trendinguplift of possible rift-¯ank origin (Bilodeau, 1986;Gustason, 1989). The V-shaped intersection of theSevier thrust belt and the Mogollon Highlandsformed a topographic control of the `GrandCanyon Bight' (Stokes & Heylmun, 1963), anembayment formed during the mid-Cretaceous

marine ¯ooding of the Western Interior basin(Fig. 1; cf. Gustason, 1989; Elder & Kirkland,1993). The Kaiparowits region was located at thenorth-western margin of this embayment, episod-ically ¯ooded and subaerially exposed during theLate Cretaceous.

STRATIGRAPHY AND NEWSUBDIVISION OF THE DAKOTAFORMATION

The Dakota Formation in the Kaiparowits regionrests with an angular unconformity on Jurassic

Fig. 1. (a) Position of the studyregion within the palaeogeographicframework of the Western InteriorSeaway. (b) A sketch map of majorCretaceous outcrops in the west-central Colorado Plateau region,showing also the approximatepositions of mid- to late-Cenomanianshorelines, as interpreted by Cobban& Hook (1984) and Wolfe (1989).(c) Outline of the Cretaceous outcrop ofthe Kaiparowits Plateau and the PariaRiver Amphitheatre. Major tectonicelements and locations of measuredsections are shown. (d) A schematic,approximately east±west cross-sectionof the Cenomanian and Lower±MiddleTuronian formations of southern andsouth-western Utah, modi®ed afterGustason (1989). The extent of thestudy area is shown by a dashedrectangle. No scale implied.

Sequence stratigraphy of Dakota Formation 809


rocks of the Entrada and Morrison formations,which are gently inclined to the east (Gustason,1989, his Fig. 2). Peterson (1969) divided theDakota Formation into three informal members:the conglomeratic `lower member' of ¯uvialorigin; the `middle member', also ¯uvial; andthe ùpper member' formed by shallow-marinesandstones of the Dakota Formation inter®ngeringwith tongues of the offshore marine Tropic Shale.Details of the history of stratigraphic terminologyand research are included in Gustason (1989) andam Ende et al. (1991). Correlation to ammonitestratigraphy is shown in Table 1.

The `lower member' of the Dakota Formation(unit 1 in this paper) yielded no fossils that wouldallow reliable dating. Peterson (1969) assignedthe lower member to the lower Cenomanian,whereas Gustason (1989) suggested an Albianage. Elder & Kirkland (1993) proposed early tomiddle Cenomanian age for the `lower member'.The `middle member' was dated by am Ende et al.(1991) as middle to late Cenomanian in age. Theùpper member' (units 4 to 6B here) belongs to theMetoicoceras mosbyense Zone. The top of theDakota Formation in the Kaiparowits region is amajor ¯ooding surface, overlain by distal offshore

Fig. 2. Correlation of units and their bounding surfaces of the Dakota Formation across the study area (location of themeasured sections shown in Fig. 1c). The sequence-stratigraphic interpretation includes only the main boundingsurfaces; details of subdivisions of individual units are discussed in text. Tectonic elements interpreted as synde-positionally active normal faults are indicated by bars at the top of the cross-section: note the incremental westwardincrease in thickness across most of them. The shift in datum from the base of the main body of the Tropic Shale inthe west to the base of unit 4 in the east is arbitrary and meant to prevent arti®cial westward plunge of boundingsurfaces, which would result from the choice of one ravinement surface (base of the Tropic) as a datum for the wholeregion. Note the in¯uence of differential compaction on the preservation of some units (especially above channels inunits 3A and 3B). Section 5 is modi®ed from Gustason (1989).

810 D. UlicÏnyÂ


shales of the Tropic Formation, belonging to thelower part of the Sciponoceras gracile ammonitezone (Gustason, 1989; Leithold, 1994). The strict-ly lithostratigraphic concept, which requires theoffshore mudstones of units 4 and 5 within theùpper member' to be termed `tongues of theTropic Shale', is not followed here: in the limitedstudy area, these units are treated as subunits ofthe Dakota Formation.

The sequence-stratigraphic study of the DakotaFormation presented here allowed a moredetailed subdivision of the Dakota rocks thanthat provided by Peterson (1969), Gustason (1989)and Elder et al. (1994). In this paper, the DakotaFormation is divided into six units (Table 1). Thebasis for the subdivision of the formation was therecognition of major, regionally traceable bound-ing surfaces (¯ooding surfaces, sequence bound-aries), which re¯ect the response of depositionalsystems to base level changes. With the exceptionof the valley-®lling unit 1, all units are present

throughout the study region and can be eitherphysically traced, or, where the exposure isdiscontinuous, correlated between the measuredsections presented here.

METHODS

The Dakota Formation was studied in a number ofvertical measured sections between the PariaAmphitheatre and the north-western margin ofLake Powell. The sedimentological observationsincluded description of lithologies, sedimentaryand biogenic structures, palaeocurrent measure-ments and fossils. Locally, the measured sectionswere supplemented by a study of architectures ofparticular sedimentary bodies. Attention waspaid to the recognition of major, regionallyextensive, bounding surfaces formed in responseto base level/sea-level changes. The NW±SEcorrelation line (Fig. 2), connecting most of the

Fig. 2. (Contd.)



studied sections, provided a two-dimensionalcontrol of thickness and lithofacies variation.The sequence-stratigraphic terminology used inthe paper follows the Èxxon' concepts as pre-sented by Van Wagoner et al. (1988, 1990), withthe exception of the parasequence concept, asdiscussed in the text. The division of a forelandbasin system into particular depozones followsthe terminology of DeCelles & Giles (1996).

DESCRIPTION AND INTERPRETATIONOF UNITS 1±6Unit 1

Description

Unit 1 in®lls the relief of the basal Cretaceousunconformity in southern Utah. Peterson (1969)found that the palaeovalleys underlying unit 1are oriented uniformly towards the south-east.This was con®rmed by a detailed palaeocurrentanalysis (Gustason, 1989). In spite of the dis-continuous occurrence of this unit, a systematicwestward increase in thickness of the valley ®lls

is observed in the study region and further west.The maximum thickness in the study region is11 m in the westernmost sections (Sections2 and 3 in Fig. 2). Gustason (1989) described a50-m-thick, 8- to 10-km-wide valley ®ll in thesouth-western Markagunt Plateau region,»40 km west of the western margin of the studyregion.

Unit 1 comprises interbedded coarse-grainedsandstone and conglomerate. The most commonlithology is cross-bedded, moderately to poorlysorted, coarse-grained, pebbly sandstone. Thecross-set thickness ranges between 0á1 and 3 m;trough cross-bedding generally prevails over pla-nar cross-bedding. Conglomerates dominate inthe lower part of unit 1 and range from poorlysorted, massively bedded, cobble conglomeratesto trough and planar cross-bedded granule con-glomerates. Thick sets of inclined strati®cation,where well-preserved, ®ne upward from con-glomerate-dominated `toesets' to medium-grainedsandstone in the `topset' strata (Fig. 3a). Clasts areoccasionally up to 10 cm in diameter, especiallyin the lowermost parts of the unit. Grey to

~94á5b

~95á8b

Table 1. Summary of the Dakota Formation units de®ned in this paper, with interpreted depositional environments,chronostratigraphic correlation and a comparison with earlier subdivisions of the Dakota Formation in southernUtah.

Unitno.

Main depositionalenvironment Subdivisions of earlier authors

Biozones; Ar/Ar dates (Myr);chronostratigraphy

6B Estuarine±uppershoreface

Upper member (Peterson, 1969;Gustason, 1989) included inParasequence 2 (Elder et al., 1994) Metoicoceras

mosbyensezone

LA

TE

(part

)

CE

NO

MA

NI

AN

6A Estuarine; offshoreto shoreface

Upper member (Peterson, 1969;Gustason, 1989) included inParasequence 2 (Elder et al., 1994)

5 Offshore to shoreface Upper member (Peterson, 1969;Gustason, 1989) Parasequence 2(Elder et al., 1994)

4 Offshore to shoreface Upper member (Peterson, 1969;Gustason, 1989) Parasequence 1(Elder et al., 1994)

3B Estuarine,tide-dominated

Middle member (Peterson, 1969;Gustason, 1989) (upper part)

Equivalentto Calycocerascanitaurinumzone3A Tide-in¯uenced, ¯uvial Middle member (Peterson, 1969;

Gustason, 1989) (upper part)

2 Fluvial, anastomosed Middle member (Peterson, 1969;Gustason, 1989) (lower part) MIDDLE

1 Fluvial, braided Lower member (Peterson, 1969;Gustason, 1989) ? ?EARLY

Sources of Ar/Ar dates marked by superscripts: (a) Obradovich (1993); (b) estimates based on Bohor (1991) and M. A.Kirschbaum (pers. comm., 1997).

93á9 � 0á72a

~94á3a

812 D. UlicÏnyÂ


dark-grey, carbonaceous and also greenish toyellowish mudstone interbeds occur rarelythroughout unit 1 and are preserved as erosional

relicts. Individual mudstone beds, up to 20 cmthick, have a limited lateral extent, generally lessthan 5 m. Mudstone intraclasts occur in the

Fig. 3. Units 1 and 2. (a) unit 1: a scour-based set of inclined strati®cation, 2á3 m thick, probably representing theresult of combined lateral and downstream accretion of a gravel bar; Section 6. (b) unit 1: a close-up view of cross-bedded sandstone and conglomerate capped with a coarse, massive, poorly sorted conglomerate at the top of unit 1;Section 7. (c) unit 1: interbeds of carbonaceous mudstone, deposited as an in®ll of a temporarily abandoned channel;Section 12. (d) Contrast in depositional style of units 1 and 2: conglomerates of unit 1 with a scoured base (arrow) areoverlain by mudstones and coals of unit 2; Section 13. (e) unit 2: shallow channel ®lled by inclined heterolithiclaminites showing lenticular and ¯aser bedding, overlying and partly incised into the coal zone 2 at the base of unit 2;between the base of the coal (arrow) and the immediately underlying Jurassic basement is »0á1 m of matrix-supportedconglomerate, equivalent to unit 1; Section 15.



conglomerates and, locally, root traces occur inmudstone and sandstone beds. Palaeocurrentindicators show a south-eastward mean ¯owdirection, in accordance with previous observa-tions (Peterson, 1969; Gustason, 1989).

The fossiliferous limestone and chert clasts inthe conglomerates indicate derivation from UpperPalaeozoic rocks then exposed in Sevier thrusts inwestern Utah and Nevada (Gustason, 1989).Silici®ed wood clasts occur commonly as frag-ments several centimetres long but, exceptionally,logs several metres long and about 1 m indiameter have been found. They probably comefrom the weathering residue of the JurassicMorrison Formation. Clasts of cross-bedded,®ne-grained sandstone were reworked directlyfrom the underlying Entrada Formation.

Unit 1 generally ®nes upwards. Thin beds ofpoorly sorted conglomerate form unit 1 outsidethe major valley ®lls. Gustason (1989) reported awidespread, poorly sorted, matrix-supported con-glomerate or conglomeratic pavement occurringin the topmost parts of unit 1 and interpreted it asa product of pedimentation and formation of anunconformity at the top of unit 1. This lithologywas recognized in some, although not all, sec-tions in the present study (e.g. Sections 3 and 5 inFig. 2; Fig. 3b). However, in Sections 6 and 7(Fig. 2), up to 3á5 m of erosional relief occurs ontop of unit 1. This relief is ®lled by whitish,carbonate-cemented, ®ne- to medium-grainedsandstone.

Environmental interpretation

On the basis of the lithofacies and sedimentarygeometries, unit 1 is interpreted as having beendeposited by braided streams characterized byhighly ¯uctuating discharge, leading to rapidshifting of channels. Gravel and sand bedformsmigrated in the channels, some forming trans-verse bars (downstream accretion mesoforms) andlinguoid or compound bars (lateral accretionmesoforms sensu Miall, 1985). Rapid channelmigration permitted only limited accumulation ofmudstone in abandoned channels or scour pits.This interpretation is in agreement with Gustason(1989), am Ende et al. (1991) and Kirschbaum &McCabe (1992). The braided streams of unit 1transported sediment from the orogenic front in asouth-eastwards direction. Peterson (1969)showed that the valley-®ll trends coincide withthe orientation of present-day structural features(Fig. 1c), indicating that basement fractures con-trolled the palaeovalley trends.

Unit 2

Description

Unit 2 is a succession dominated by mudstones,sandstones and coals. It corresponds to the lowerand middle portions of the `middle member' ofthe Dakota Formation (Table 1; the `lower coalzone' and a part of the `middle alluvial interval' ofGustason, 1989). It is either underlain by unit 1,commonly also separated by a coal zone (coalzone 1 of Kirschbaum & McCabe, 1992) or restsdirectly on Jurassic bedrock (Figs 3d and 4). Thetop of unit 2 is an erosional surface underlyingtide-in¯uenced deposits of units 3A and 3B. Inthe study region, unit 2 shows a distinct west-ward increase in thickness from »10 m in the eastto »30 m in the west (Fig. 2).

Unit 2 is dominated by mudstones, mostlymassive, pale grey to dark grey, locally carbona-ceous. Pedogenic features such as rooted zonesand slickensides occur commonly throughoutunit 2. Sharp-based, thin, sheet-like sandstonebeds occur encased in the mudstones. Thesesandstones commonly ®ne upwards into silt-stones and normally show small-scale cross-bed-ding or ripple cross-lamination. Siderite concre-tions are found locally within the mudstones.

Isolated channel ®lls commonly display basalintraformational conglomerates, consisting mostlyof mudstone and coal clasts and vertebrate bonefragments. The channel-®ll lithology is dominat-ed by well-sorted, ®ne-grained, trough cross-bedded sandstones, which may grade upwardsinto ripple cross-laminated sandstones with mud-stone interbeds. Mudstone channel ®lls arerelatively rare. Climbing ripples are common inthe ®ner grained sandstone facies. Channelsimmediately overlying the basal coal zone in theeasternmost sections (Fig. 3e) are ®lled withinclined heterolithic strata characterized by ¯a-ser±lenticular bedding. Commonly, channel ®llsin unit 2 show a high degree of synsedimentarydeformation, which includes recumbent cross-beds and disharmonic folding (cf. Gustason, 1989;Kirschbaum & McCabe, 1992). Mean palaeocur-rent direction within unit 2 is east-south-east(Gustason, 1989), but the dispersal of palaeocur-rents is high.

Architecturally, the channel-®ll bodies includelateral accretion, vertical and concentric accre-tion, and are documented in detail by Kirsch-baum & McCabe (1992). A part of the channel ®llsdescribed by these authors from the Henrievillearea is, however, now included in unit 3A andwas found to ®ll a topography incised into unit 2

814 D. UlicÏnyÂ


(Figs 4 and 5). Gustason (1989) described anapparent change in ¯uvial architecture acrossthe region, from isolated channel-®ll bodiesencased in mudstones in the west, towards wide,laterally persistent, lateral-accretion lithosomesin the east. Re-examination of relevant sections bythe author showed that most of the laterallyextensive channel systems in the `middle Dakota'are, in fact, incised channels of units 3A and 3Band cannot be correlated to unit 2 in the westernpart of the study region.

A conspicuous component of unit 2 is coal,occurring in distinct assemblages of seams (coalzones) ranging in thickness from 5 cm to 2á5 m.Unit 2 comprises coal zones 1±3 of Kirschbaum &McCabe (1992). Coal zones 1 and 2 correspond tothe Bald Knoll coal of Doelling & Graham (1972).Coal zone 1 is the least extensive, whereas coalzone 2 is a major group of thick coals that areregionally correlatable. Dispersed, thin coals oflimited lateral extent within unit 2 correspond tocoal zone 3 of Kirschbaum & McCabe (1992). Thecoal zones 4 and 5 of the `middle member of the

Dakota' are genetically related to units 3A, 3Band, possibly, 4.


Unit 2 is interpreted as representing an ana-stomosed river system, characterized by isolatedchannels encased in ¯oodplain and swampdeposits (cf., e.g. Smith, 1986). The sheet-likesandstones interbedded with the mudstones andcoals represent crevasse-splay deposits. Thisinterpretation of the depositional system is inagreement with that of Gustason (1989) andKirschbaum & McCabe (1992), but it applies onlyto the lower portion of their `middle Dakota'. Thewide palaeocurrent dispersal reported by Gusta-son (1989) is partly caused by the sinuosity of theanabranches, and partly because the measure-ments of Gustason (1989) also included units 3Aand 3B, commonly characterized by intensemeandering.

Kirschbaum & McCabe (1992, their Fig. 13)pointed out the apparent similarities between

Fig. 4. Contrasts in depositional architectures of units 2, 3A and 3B, Henrieville-East (Section 4). (a) Interpretation ofthe photomosaic in (b); 4a±d in (a) are the locations of measured sections shown in Fig. 5. Unit 2, based by theprominent coal zone 2 (cz 2) is dominated here by isolated channels encased in mudstone and coal-dominated¯oodplain deposits. The base of unit 3A is characterized by thinner sandstone in®lls of a channel system incised intounit 2; unit 3B shows distinct lateral accretion geometry throughout and local channel ®lls. CH, channel; s, sand-stone; ihs, inclined heterolithic strati®cation; cz, coal zone; cs, crevasse splay.



the Dakota anastomosed rivers and the Okavango¯uvial system in Botswana and interpreted theDakota system as `some form of large inland deltaintroducing clastics into a widespread mire'.However, the occurrence of inclined heterolithicstrati®cation, locally rhythmic and containing¯aser to lenticular bedding, close to the base ofunit 2 in the eastern part of the study region isinterpreted here as a record of tidal in¯uence on¯uvial deposition (Fig. 3e; cf. Thomas et al.,1987; Shanley et al., 1992). Therefore, it is sug-gested here that the ¯uvial system of unit 2 wasnot a `large inland delta', as hypothesized byKirschbaum & McCabe (1992), but may have beenconnected to an estuarine system east of theKaiparowits region.

Units 3A and 3B

Units 3A and 3B share similar facies andarchitectural characteristics: both are markedlyheterolithic, show signs of tidal in¯uence in thedepositional environment and occupy incised

channel systems. The units are separated fromone another by a palaeosol capped with adistinctive coal bed containing a thin bentoniteinterbed, belonging to coal zone 4 of Kirschbaum& McCabe (1992). However, channels at thebase of unit 3B commonly incise into unit 3A,removing the bounding palaeosol and coal andthus making it dif®cult to distinguish betweenthe two units. For this reason, UlicÏnyÂ (1996)treated them as a single unit, and their similarityis re¯ected in their designation as units 3Aand 3B.

Description

Unit 3A is 0±10 m thick and occupies channelsincised into the underlying unit 2 (Fig. 2). Itcomprises trough cross-bedded sandstone andheterolithic sandstone±mudstone laminites andgrades upwards into silty mudstones, commonlycarbonaceous, capped by a palaeosol and a coalbed. The channel ®lls show lateral accretionarchitecture (Figs 4 and 6a), commonly with

Fig. 5. Measured sections and palaeocurrent data for units 2 to 3B (Sections 4a±d, locations shown in Fig. 4a). Thelocations of features illustrated in Figs 6 and 7 are shown. Symbols as in Fig. 2.

816 D. UlicÏnyÂ


transitions between inclined heterolithic strati®-cation (sensu Thomas et al., 1987) and cross-bedded sandstones forming the upper and lowerportions of the channel ®lls. Sandstone-dominatedchannel ®lls, with abundant lags of mud and coalclasts, are characterized by trough cross-sets withcommon mud drapes and, locally, mud coupletson foresets (Fig. 6). The drapes are commonlyformed by carbonaceous debris instead of mud.Bottomsets show ¯aser lamination. Synsedimen-tary deformation is common locally. Heterolithiclaminites (Fig. 6c) show a typical continuum of¯aser to lenticular, `tidal' bedding, locally rhyth-mic. A channel ®ll in Section 2 (Fig. 6a) containedcorbulid bivalves; elsewhere, bivalve shell mouldsoccur in sandstones. Palaeocurrent indicators,where measured, show eastward sediment trans-port (Fig. 5).

Unit 3B is 0á8 to more than 9 m thick and is ahighly heterolithic facies succession. It is sepa-rated from unit 3A by a coal bed (Fig. 4; coal zone4 of Kirschbaum & McCabe, 1992), except forcases where the coal has been removed bysubsequent erosion. Unit 3B was identi®ed inmore sections than the underlying unit 3A.Although unit 3B does not occupy such markedlyincised channels as unit 3A, regional correlationshows that it changes thickness across relativelyshort distances (Fig. 2), and local incision intounderlying units is common (Fig. 7)

Unit 3B is most commonly characterized by a®ning-upward trend, from ®ne-grained, well-sorted, planar and trough cross-bedded sand-stones at the base, through heterolithic laminitesto black, carbonaceous mudstones at the top,rooted and capped with a coal bed (Fig. 7b and c).

Fig. 6. (a) Unit 3A: part of a laterally accreted channel ®ll formed by heterolithic laminites containing corbulidbivalves; Section 2. (b) Mud-draped foresets of small-scale cross-sets, Section 4. (c) Rhythmic laminites interpreted asneap±spring cycles revealed by alternating lenticular and ¯aser bedding, Section 4 (thickness shown is »0á5 m).(d) Double mud drapes preserved in cross-bedded sandstone above the base of unit 3A, Section 4. For stratigraphiccontext, see Fig. 5.



Large-scale depositional architecture, where suf-®ciently exposed (over tens to hundreds ofmetres), is dominated by lateral accretion, in boththe cross-bedded sandstones and the laminites(Fig. 4). The sandstone foresets are draped bymud and carbonaceous debris, commonlygrouped to form tidal bundles (Fig. 7; cf. Visser,1980; Nio & Yang, 1991). Some cross-sets displaysigmoidal-shaped foresets and, locally, herring-bone cross-bedding. Palaeocurrents are directedpredominantly west to north-west (Fig. 5 and 7).The interlaminated sandstones and mudstonesshow various types of `tidal bedding', locallymarkedly rhythmic (Fig. 7a). In places, the unitcontains extremely abundant corbulid bivalves.Arenicolites and Ophiomorpha occur in thesandstones. In places where both units 3A andB are absent or very thin, a thick palaeosol isdeveloped under unit 4 (Sections 5 and 6 inFig. 2).


The occurrence of coupled mud drapes on fore-sets, sigmoidal-shaped foresets, herringbone

cross-bedding and ¯aser- to lenticular-beddedlaminites indicates clearly that units 3A and 3Bformed in tide-in¯uenced settings (Figs 5, 6 and7; cf. Nio & Yang, 1991). In Section 4, sampled forpalynological analysis, marine microplanktonwas found in the coals capping both unit 3Aand unit 3B, as well as in the mudstones withinunit 3B (M. SvobodovaÂ, pers. comm., 1996). Also,the occurrence of corbulid bivalves attests tobrackish water in¯uence in both units.

The two units differ primarily in the abundanceof tidal indicators, which are more common inunit 3B than in unit 3A, and in the lateral extentof their deposits. Although both units are charac-terized by an order of magnitude variation inthickness across the study area, which resultsfrom their occurrence as channelized sedimentbodies, the channel systems of unit 3A are muchmore localized than those of unit 3B (Fig. 2).

Unit 3A is interpreted as deposited in a tide-in¯uenced ¯uvial environment, based on thecriteria listed above and on comparisons withother tide-in¯uenced ¯uvial systems (e.g. Shan-ley et al., 1992). The ¯uvial system occupiedchannels and shallow valleys incised into unit 2.

Fig. 7. Unit 3B. (a) Neap±spring tide cycles preserved in laminites in®lling a temporarily abandoned estuarinechannel (Section 4d in Fig. 5). (b and c) Photograph and measured section of a typical ®ning-upward succession ofsubtidal to supratidal deposits in®lling an estuarine tidal channel; Section 10. (d and e) Photograph and measuredsection of a subtidal channel ®ll showing neap±spring tide cyclicity recorded in mud-draped foresets of ¯ood-oriented dunes; Section 14; arrows point at clusters of neap tide laminae.

818 D. UlicÏnyÂ


Individual channels meandered within theincised valleys. The in¯uence of brackish water,documented by corbulids, as well as the probableneap±spring tide rhythmicity recorded in thelaminites (Fig. 6c) suggest that the environmentof deposition was close to the junction of the¯uvial system and the upper estuary (Rahmani,1988; Shanley et al., 1992).

The deposition of unit 3B indicates the estab-lishment of an estuarine, tide-dominated deposi-tional environment. The more widespread natureof this unit relative to unit 3A probably resultsboth from a larger area of deposition and possiblyfrom erosion of unit 3A by scour in subtidalchannels (cf. Oertel et al., 1991; UlicÏnyÂ &SÏpicÏaÂkovaÂ, 1996). Herringbone cross-bedding in-dicates a strongly asymmetrical tidal ellipsecaused by con®nement of ¯ow in channels; thisis also supported by the abundance of muddrapes (cf. Yang & Nio, 1989). The lateral accre-tion architecture of both the cross-bedded sandsand heterolithic strata re¯ects lateral migration ofgently dipping sand bodies, interpreted as tidalpoint bars, within estuarine channels (Thomaset al., 1987; Ainsworth & Walker, 1994). Thinsuccessions dominated by tidal laminites,commonly ®ning upwards, represent tidal ¯atdeposits outside the main channel courses.Palaeocurrent data show consistently a SE±NWtrend of the channels, similar to the drainagepattern of underlying ¯uvial units. Local accu-mulations of corbulid bivalves in areas where theunit is thin seem to re¯ect low local rates ofsedimentation (e.g. Section 2 in Fig. 2). Locally,the laminites show rhythmic thinning and thick-ening of the two lithologies, re¯ecting neap±spring tide rhythmicity (Fig. 7; cf. Kvale &Archer, 1990; Nio & Yang, 1991).

A trend of increasing tidal in¯uence in aneastward direction, supported by the abundanceof sand-dominated channels ®lled by subtidaldunes in the Big Water area, indicates that thewestern part of the study area represents themiddle estuary, with heterolithic estuarine pointbars and tidal ¯ats, whereas the eastern part wasclose to the estuary mouth (Fig. 7d and e). Theposition of the estuary-open marine transition isunclear, but a thin, coarsening-upward, shorefacesuccession that occurs in the easternmost part ofthe study region (Fig. 2), tentatively interpretedas coeval to tidal deposits of unit 3B, suggests thatthis transition may be preserved in the south-eastern tip of the Kaiparowits region. Precisecorrelation is dif®cult because of discontinuousexposure.

Units 4 to 6B

Units 4 to 6B represent the ùpper member' of theDakota Formation in the study area (Table 1). Afeature common to units 4, 5 and 6A is anupward-coarsening facies succession interpretedas offshore to shoreface deposits. Speci®c toindividual units are occurrences of other lithofa-cies, including cross-bedded sandstones,heterolithic laminites, coals and others. Theirrelevance to sequence-stratigraphic interpretationnecessitated separate descriptions of these faciesassociations for individual units. Units 6A and 6Bare locally dif®cult to distinguish because of verysimilar facies associations.

Gustason (1989) and Elder et al. (1994) dividedthe ùpper member' in the study area into twoshallowing-upward parasequences and correlatedthem to limestone±shale couplets in the distalfacies of the Western Interior. It will be demon-strated below that the sequence stratigraphy of theùpper member' is signi®cantly more complicated.

Description

Unit 4 is 0±6 m thick and comprises an upward-coarsening succession of shallow-marine mud-stones and sandstones (Fig. 8a and c). It thinsboth downdip and updip from the zone ofmaximum thickness between Sections 2 and 9.Downdip, it pinches out near Section 14 (Fig. 2).

The basal surface of unit 4 is sharp, commonlyoverlain by a fossiliferous sandstone, 0á1±1 mthick, rich in bivalve shells, mostly small oysters(Fig. 8). This deposit, commonly cemented bysiderite and interbedded with coal debris, locallyoverlies a thin (up to 1 m thick) succession ofmudstones rich in corbulid bivalves or a thin,lenticular-bedded, heterolithic succession.

The main body of unit 4 consists of grey marinemudstones, which alternate with ®ne-grained,well-sorted, subarkosic to arkosic sandstonesfrom the middle part of the unit upwards. Thesandstone beds thicken upwards. From bottom totop, a succession of wave-ripple lamination,undulating parallel lamination and hummockyto swaley cross-strati®cation occurs in the sand-stones. Synsedimentary deformation is common,especially in sections at the western edge of theKaiparowits Plateau (Sections 8 and 9 in Fig. 2),and is dominated by convolute bedding and ball-and-pillow structures (Fig. 8c).

Unit 5, up to 12 m thick, is an upward-coarsening, shallow-marine succession similarto that of unit 4. The boundary between units 4



and 5 is either sharp, overlain by an oyster lag,commonly sideritic, or it is marked by shallowscours and, locally, channels up to 4 m deep.The channels are ®lled with cross-beddedsandstone with coal debris and very abundantoysters (Fig. 8d and e). The top of unit 5 is anerosional surface, locally penetrated by roots,overlain by various facies assemblages of units6A and 6B.

In the uppermost parts of unit 5, trough cross-bedded sandstones are thicker than analogousfacies of unit 4, but are not continuously pre-served. To the east of Section 9, gutter casts beginto occur in the middle portion of the unit. Furthereastwards, the thickness of unit 5 is reduced to afew metres because of erosion at the base ofoverlying unit 6A.

Unit 6A, up to 8 m thick, comprises twodifferent facies associations, here called proximaland distal. The distal part of unit 6A is anupward-coarsening succession, analogous to theunderlying units 4 and 5, of dark, silty mud-stones, wave-rippled sandstones and thin HCSbeds. It occurs east of Section 12 and, in theCoyote Creek area (Section 11), it passes updipinto the proximal assemblage. The top of unit 5 isoverlain by a distinctive carbonaceous layer, upto 40 cm thick, formed largely by redepositedcarbonaceous debris with silty and sandy admix-ture and locally abundant thin-shelled bivalvesand gastropods (Fig. 9f and g). The topmost partsof unit 6A pass upwards into heterolithic lam-inites and carbonaceous mudstones of the prox-imal facies assemblage.

The proximal part of unit 6A is exposed west ofSection 11 (Fig. 2), where it rests on unit 5. Itslower bounding surface is locally burrowed andpenetrated by roots and shows a pronouncederosional relief (Fig. 9b and h; Section 8 inFig. 2). The proximal facies association includesplanar and trough cross-bedded, ®ne- to medium-grained sandstones; massive, intensely bioturbat-ed, ®ne-grained sandstones; heterolithiclaminites; carbonaceous mudstones and coals.Cross-bedded sandstones occur as channel ®lls,commonly several metres thick, interbedded withheterolithic laminites. Common trace fossilsinclude Ophiomorpha, Thalassinoides andArenicolites. The massive sandstone facies containsabundant trace fossils, which show that intensebioturbation caused the loss of original physicalstructures. The heterolithic laminites show acontinuum of ¯aser, wavy and lenticular beddingand commonly ®ne upwards, capped by coals orcarbonaceous mudstones (Fig. 9b). Distinct

cracks occur in the heterolithics and coals of theproximal part of unit 6A in sections near the EastKaibab monocline, ®lled by overlying sands andshowing compactional folding (Fig. 9b).

Unit 6B overlies an erosional surface on top ofthe proximal facies of unit 6A. The boundary ismarked by tree stumps locally preserved underunit 6B or by coals at the top of unit 6A, witherosional features such as gullies and channels(Fig. 9a±e). In the Rimrocks area (Figs 2 and 9a),the channels incised into the coals are ®lled withdark-coloured, bioturbated sandstone and silt-stone with extremely abundant carbonaceousdebris. Other facies include cross-beddedsandstones with bidirectional (WNW±ESE)palaeocurrent patterns, mudstone drapes on fore-sets, sigmoidal foresets, locally well-developedmudstone couplets forming tidal bundles andbundle sequences (Fig. 9i±l). Unit 6B is easilyidenti®ed to the east of Section 5, where its mostconspicuous facies, sandstone with accumula-tions of thick-shelled Exogyra oysters, mostlypreserved in situ as òyster banks', occurs belowthe basal surface of the Tropic Shale (Fig. 9d ande). This facies is commonly underlain by ®ne-grained, bioturbated sandstone with scatteredbivalves.

Although it is dif®cult to correlate unit 6Bphysically to the west of Section 5 because of

Fig. 8. Units 4 and 5. (a) Typical appearance of units 4and 5 as simple shallowing-upward `parasequences' inSection 10; see text for discussion and (e) for a mea-sured section. (b) Lower part of unit 4, interpreted as abackbarrier transgressive systems tract, overlain by awave ravinement surface marked by a shell lag (arrow);Section 12. (c) A body of convolute-bedded sandstone(arrow), interpreted as a submarine slump, thickeningtowards the right from 0á7 to 2á5 m, occurs in theuppermost part of unit 4; Section 8. (d) Cross-beddedsandstone with oyster coquinas on foresets, ®lling achannel incised into unit 4 and interpreted as a tidalinlet ®ll, part of a transgressive systems tract of unit 5;Section 2. (e) Selected measured sections illustratingthe internal complexity of units 4 and 5, interpreted ashigh-frequency sequences; note that where the trans-gressive systems tracts (TSTs) are not preserved(Section 10), both units have the appearance of para-sequences. The systems tracts de®ned are facies based:RST, `regressive systems tract' (regressive deposits inwhich the boundary between the highstand (HST) andfalling stage (FSST) systems tracts cannot be identi-®ed); SB, sequence boundary; FS, ¯ooding surface;RSME?, tentatively interpreted regressive surface ofmarine erosion, de®ned by the sharp base of shorefacesandstones passing downdip into a gradual upward-coarsening succession. See Discussion for comments.

820 D. UlicÏnyÂ




Fig. 9. Units 6A and 6B. (a) A channel at the base of unit 6B (arrow) incised into the coals of the proximal part of unit6A and ®lled with siltstones containing a high proportion of carbonaceous debris; the lenticular shape of the channel®ll is caused by high compaction of coal underlying the channel ¯anks; Section 9. (b) A crack in coal and underlyingheterolithic strata of unit 6A, ®lled with overlying bioturbated sandstone and showing compactional folding; Section9; note also slumped beds overlying the channelized 5/6A boundary. (c) The 6A/6B bounding surface showingerosional scour in coal, ®lled by bioturbated sandstone of unit 6B; Section 9. (d) Base of the main body of the Tropicshale, marked by a lag deposit dominated by Pycnodonte newberryi oysters, overlying the Exogyra sandstone of unit6B; Section 9. (e) Casts of tree roots and stumps preserved beneath the 6A/6B boundary; note the Exogyra sandstonedirectly overlying the scoured surface; arrows show branching roots; Section 7. (f) The 5/6A boundary marked byaccumulated carbonaceous debris (arrow) overlying an erosional surface at the base of the distal facies of unit 6A;Section 13. (g) Close-up view of the base of unit 6A shown in (f).

822 D. UlicÏnyÂ


Fig. 9h±l. (h) Units 6A and 6B ®lling a channel incised into unit 5; note the lateral accretion surfaces in the sandstones (arrow);Section 2. (i) Neap±spring tidal cyclicity preserved in tidal bundles of subtidal channel of unit 6B; Section 2, see measuredsection in (j). (j) Lithofacies and palaeocurrents of units 6A and 6B; Section 2. (k) Sandstone-®lled channel incised into thedistal facies of unit 6A and interpreted as a tidal inlet ®ll related to the 6B transgression; Section 12; see Depositional history fordetails. (l) Measured section of the channel ®ll shown in (k).

Seq

uen

cestra

tigrap

hy

of

Dakota

Form

atio

n823

Ó1999

Inte

rnatio

nal

Asso

cia

tion

of

Sed

imen

tolo

gists,

Sed

imen

tolo

gy,

46,

807±836

discontinuous exposure, several channel ®llsconsisting of cross-bedded sandstone with tidalbundles and heterolithic laminite interbeds at thewestern margin of the study area are interpretedas belonging to unit 6B (Fig. 9i and j). Thiscorrelation, however, is tentative, based on se-quence-stratigraphic reasoning explained furtherbelow.

Unit 6B is overlain by a thin, bioturbated,fossiliferous sandstone rich in Pycnodonte new-berryi oysters, commonly forming laterallyextensive coquinas, which is the basal depositof the main body of the Tropic Shale (Fig. 9d).


The upward-coarsening successions of units 4 and5 and the distal part of unit 6A are typicalsuccessions formed in offshore to wave-dominat-ed lower to upper shoreface environments by theprogradation of a wave-dominated strandplain(for example, cf. Walker & Plint, 1992 and refer-ences therein). Orientation of wave-ripple crests isgenerally north±south (Fig. 8e), which corre-sponds to the shoreline strike during the deposi-tion of these units. Gutter casts found in units 4and 5 were oriented essentially perpendicular to

the shoreline. Mudstones with abundant corbulidbivalves, found locally in the basal parts of units4 and 5, as well as heterolithic laminites, areinterpreted as preserved relicts of back-barrier,lagoonal/estuarine deposits, and channel-®llingsandstones found locally at the base of unit 5 areinterpreted as relicts of tidal inlet ®lls (Fig. 8e).

The proximal part of unit 6A is interpreted asdeposits of a tide-dominated estuarine setting.Gustason (1989) also identi®ed the tidal deposi-tional regime in this part of the Dakota Formation,but he interpreted the tide-in¯uenced deposits aslagoon deposits formed behind a progradingbarrier island system. Although the fauna of unit6 is similar to that described by FuÈ rsich &Kirkland (1986) in a late Cenomanian brackishlagoon deposit from the Black Mesa, Arizona,muddy facies that could be interpreted as back-barrier lagoon-®ll sediments are not common inunit 6. Strong tidal currents are recorded in theabundant cross-bedded sandstones ®lling chan-nels and representing subtidal sand bars, as wellas in tidal laminites, formed on tidal ¯ats.Carbonaceous mudstones and coals were depos-ited in supratidal marshes. This facies assemblageis very similar to cases described by Devine(1991), Leckie & Singh (1991) and Ainsworth &Walker (1994) from tide-dominated, transgressiveestuarine deposits ®lling channels incised intowave-dominated shoreface deposits.

Unit 6B was deposited in environments similarto the proximal part of unit 6A. Whereas thecross-bedded sandstones and heterolithic lam-inites are interpreted as tidal channel and tidal¯at facies, the sandstones with Exogyra accumu-lations correspond to a high wave energyenvironment and are interpreted as foreshoredeposits. A special case among various tidalchannel ®lls of unit 6B is a large channel complexin the Wahweap Creek area. The channel, up to300 m wide, »10 m deep and dominated bybidirectionally cross-bedded sandstones is in-cised into the offshore±shoreface deposits of thedistal part of unit 6A (Fig. 9k and l). It mayrepresent either a tidal inlet ®ll, as argued byGustason (1989), or a transgressive estuarinechannel occupying inherited erosional topogra-phy of a sequence boundary. A combined inter-pretation is also possible, because it is commonfor tidal inlets within transgressive barrier sys-tems to form in pre-existing channels and furtherenhance their relief (e.g. Oertel et al., 1991). Inany case, the channel is considered a relict of atransgressive barrier system, not a progradingone, as argued by Gustason (1989) and followed

Fig. 10. Schematic cross-sections, not to scale, show-ing the hypothetical succession of main events causingthe deposition of unit 1. (a) Uplift of a forebulge causesincision of valleys deepening to the west. (b) Onset of¯exural regime in the region, presumably during theEarly Cenomanian, causes the backtilting of the surface,¯attening of ¯uvial gradients and ®lling of the valleysby orogen-derived clastics (unit 1). (c) Deposition ofunit 2 under conditions of continuing ¯exural subsi-dence. Comments in text.

824 D. UlicÏnyÂ


by Elder & Kirkland (1993), because barrier-lagoon systems form essentially during transgres-sions (cf. Devine, 1991; Swift et al., 1991a).

DEPOSITIONAL HISTORY

Unit 1

The depositional history of unit 1, complicated bythe lack of reliable age data, is closely related tothe preceding events responsible for the incisionof the topography underlying unit 1. This erosio-nal episode probably coincided with the eastwardtilting of the Jurassic bedrock, which occurredduring the Early Cretaceous (the `Neocomian'regional uplift; Peterson, 1969). The eastwardtilting and resulting westward-deepening erosionof the Jurassic strata are explained here as aconsequence of the rise of a forebulge formed inresponse to initial thrusting to the west of themid-Cretaceous position of the Sevier thrust front(Fig. 10). A similar explanation has beenproposed recently by Currie (1998) for thepalaeovalley systems underlying the BuckhornConglomerate member of the Cedar MountainFormation in north-eastern Utah. Other, alterna-tive interpretations, such as a thermal uplift (cf.Heller & Paola, 1989) or ¯uctuations in compres-sional intraplate stress (Heller et al., 1993; Peper,1993), are not preferred here (for detaileddiscussion, see Currie, 1998). The forebulge upliftwould be consistent with the hypothesis of anEarly Cretaceous foredeep developed in front ofthe Pavant and Canyon Range thrust sheets inwestern Utah (Royse, 1993; Schwans, 1995).

The data of Currie (1997, 1998) indicate that,during this time, eastward migration of theforebulge shifted the NE Utah region from the`back-bulge depozone' of the foreland basin sys-tem (as de®ned by DeCelles & Giles, 1996) to thezone of forebulge uplift. In the Kaiparowitsregion, situated further east of the Sevier orogenicfront, this shift may have occurred later. Thesubsequent eastward propagation of the Sevierthrust front during the Cenomanian caused theonset of the typical ¯exural regime of the foredeepdepozone: the backtilting of the Jurassic basementin southern Utah, possibly combined with aeustatic rise, led to deposition of the unit 1 valley®lls. Although detailed research is clearly neces-sary to test the hypothesis of a forebulge-relatedincision, the depth of erosional relief preserved(»50 m maximum) and the width of the areaaffected by erosion (»150 km along section inFig. 1d, normal to the thrust front) are within the

range of values estimated for the height andwavelength of a forebulge, given the ¯exuralrigidity expected of the Western Interior base-ment (cf. DeCelles & Giles, 1996 and referencestherein).

The Ar/Ar dates from bentonites in the lowerpart of unit 2 suggest that the youngest possibleage of unit 1 is Early or early Middle Cenomanian(based on Bohor et al., 1991; see below). On thebasis of lithologic similarity between unit 1 andthe Lower Cretaceous conglomerates found in theWestern Interior (e.g. the Buckhorn Conglomer-ate), Gustason (1989) interpreted the age of unit 1as Early Cretaceous (Albian). However, the LowerCretaceous conglomerates found in the forelandof the Sevier belt and interpreted formerly assynorogenic conglomerates recording activethrusting are now considered to be older thanthe ¯exural subsidence phase in the Cretaceousforeland basin (cf. Heller & Paola, 1989; Currie,1998). The incision of the valleys in the Kaiparo-wits region most probably provided clastics forAlbian gravels, studied by Heller & Paola (1989),further eastward in the Sevier foreland. Theincision might have been enhanced by the effectsof Early Cretaceous eustatic ¯uctuations, asinterpreted by Aubrey (1992) for the sub-Dakotasurface in New Mexico, but this possibility has tobe tested by improved stratigraphic correlation.

The evidence of erosion and pedogenesis at thetop of unit 1 suggests a gradient change causedeither by uplift (favoured by Gustason, 1989) orby a sea-level fall. Although signs of subaerialerosion were recognized on bounding surfaces ofmany of the units of the Dakota Formation, thelack of physical correlation to coeval paralicstrata, as well as the lack of reliable chronostrati-graphic data for unit 1, preclude an unequivocalexplanation of the cause of this erosional event.

Unit 2

The contrast in ¯uvial style between units 1 and 2(Fig. 3d), as well as the internal organization ofunit 2, suggest a signi®cant change in streampro®les and a major decrease in regional sedimentinput rate in the Dakota ¯uvial system. Such achange in depositional style, emphasized by theoccurrence of coal on top of unit 1 over most ofthe area, can be attributed either to an increasedrate of ¯exural subsidence in the foreland basinleading to accelerated backtilting of the streampro®les and trapping of coarse sediment close tothe orogenic front, or to a relative sea-level risecausing transgression on the coast east of the



Kaiparowits area and resultant ¯attening of thestream pro®les. The latter explanation is favouredhere because of the occurrence of probably tide-in¯uenced strata immediately above the base ofunit 2 in the eastern part of the study area.

Further considerations of the depositional his-tory of unit 2 depend on the accuracy of its ageestimates, further complicated by the lack ofbiostratigraphic data. Bohor et al. (1991) datedaltered volcanic ashes from the Cenomanian ofthe Kaiparowits area, including the coal zones 2and 3; although the absolute ages are »3 Myryounger than comparable Cenomanian dates ofObradovich (1993), the internal consistency of theBohor et al. (1991) data set allows an estimate ofthe time span represented by the strata betweenthe two zones as »1á4 Myr. The corrected age forthe coal zone 2 is close to 95á8 Myr, the Early±Middle Cenomanian boundary (Table 1; based onObradovich, 1993).

Coal seams in ¯uvial strata deposited oncoastal plains have been interpreted recently astime-equivalents of ¯ooding surfaces at coevalcoastlines, re¯ecting a rise in water table on thecoastal plain, as well as the decrease in rate ofsediment input resulting from base-level rise(Flint et al., 1995 and references therein). Basedon the characteristics of the coal zones and theclastic deposits between them, unit 2 is inter-preted as recording a long-term cycle of base-level change. The discontinuous coal zone 1 isinterpreted as re¯ecting initial ¯ooding, whereasthe laterally continuous, thick coal zone 2 maycorrelate to a maximum ¯ooding interval. Theclastic deposits between these coal zones aredominated by mudstones and are tentativelyinterpreted here as a transgressive systems tract.The deposits overlying the coal zone 2 arecharacterized by an upward increase in theabundance of channels and by thin, discontinu-ous coal seams (coal zone 3 of Kirschbaum &McCabe, 1992), which may represent a relativebase-level highstand (cf. Wright & Marriott, 1993;Olsen et al., 1995). The erosional surface under-lying unit 3A is interpreted as a sequenceboundary (see below). It must be kept in mind,however, that numerous palaeosols and coalswithin unit 2 may re¯ect a number of higherfrequency base-level ¯uctuations superimposedon the long-term trend.

Units 3A and 3B

The occurrence of the tide-in¯uenced unit 3Aabove the ¯uvial deposits of unit 2 clearly indicates

a relative sea-level rise. However, the erosionalrelief underlying unit 3A and thick palaeosolsdeveloped where channel-®ll sandstones of unit3A are absent (e.g. Section 6 in Fig. 2) are regardedas evidence for a preceding base-level fall, and thusthe base of this unit is interpreted as a sequenceboundary coinciding with a ¯ooding surface. The®ning-upward pattern within unit 3A is a responseto ®lling of accommodation, terminated bydeposition of mudstones with pedogenic featuresand coals at the top of unit 3A.

Erosional topography also exists at the base ofunit 3B, locally forming channels up to 8 m deep(Fig. 7), which formed either by subaerial erosionas a result of a base-level fall or by autogenicchannel incision by tidal currents in the estuarineenvironment. Such `tidal ravinement' (Allen,1991) is common in estuarine settings and maycause some ¯ooding surfaces to resemblesequence boundaries super®cially (UlicÏnyÂ &SÏpicÏaÂkovaÂ 1996). Erosion by tidal currents seemslikely because the deepest incision is observed inthe eastern part of the study area, where evidenceexists for vigorous current action in subtidalchannels of the `lower estuary' (Fig. 7d and e).On the other hand, in the western part of the area,coal zone 4 on top of unit 3A is overlain bysandstones and heterolithics of the `middleestuary' portion of unit 3B, without signi®canterosion. It is also possible that the coal bedunderlying the sandstones and heterolithicdeposits of unit 3B in the western part of thestudy area re¯ects mire formation on top of unit3A, in response to early sea-level rise.

The depositional history of units 3A and 3B issummarized as follows: (i) after a relative sea-level fall, incision of shallow channel systemsinto unit 2 occurred, accompanied by pedogene-sis in the inter¯uves; (ii) a relative sea-level risecaused ®lling of the incised topography by tide-in¯uenced ¯uvial deposits of unit 3A; (iii) after astillstand or a minor relative sea-level fall, depo-sition of unit 3B occurred in response to signi®-cant ¯ooding of the area and onset of depositionin a tide-dominated, estuarine system, whichcovered a much larger area than the localizedchannels of unit 3A. The carbonaceous mudrocksand coals in the topmost parts of unit 3B formedin a supratidal environment in response to ®llingof the accommodation. The possibility of asigni®cant relative sea-level fall terminating de-position of unit 3B can be neither ruled out norcon®rmed by existing data, but the marine strataof unit 4 clearly overlie a subaerially exposedsurface.

826 D. UlicÏnyÂ


Units 4 and 5

The base of unit 4 is a record of the ®rst marine¯ooding of the Kaiparowits area, at the beginningof the M. mosbyense Zone (cf. Elder et al., 1994).Locally preserved mudstones with brackish waterbivalves and heterolithic laminites of intertidalorigin in the lower part of unit 4 (e.g. Sections 6, 8and 12; Figs 2 and 8b and e) probably representremnants of back-barrier deposits formed in atransgressive barrier±lagoon depositional system.Most of the deposits of this system weredestroyed by subsequent wave ravinement, whichresulted in the deposition of an oyster lag eitheron top of the underlying unit 3B or on top of therelicts of back-barrier deposits. This trangressivelag is overlain by the shallowing-upward off-shore±shoreface succession, which forms thebulk of unit 4. No direct signs of subaerial erosionwere found at the top of unit 4.

The base of unit 5 is another prominent marine¯ooding surface, in many ways analogous to thatof unit 4. The back-barrier and tidal-inlet faciesfound in the westernmost part of the area repre-sent relict transgressive deposits of a barriersystem similar to that of unit 4. The shell lags atthe base of the offshore muds of unit 5 overlie awave ravinement surface, which marks the max-imum ¯ooding during deposition of the unit. Theoverall thickness of unit 5, compared with that ofunit 4, indicates a relatively larger increase inaccommodation. The gradual ®lling of accommo-dation, represented by the bulk of unit 5, wasterminated by subaerial exposure and valleyincision. Based on the data available, it is dif®cultto speculate about the position of the boundarybetween the highstand and falling stage deposits(sensu Plint, 1996) within the shallowing-upwardsuccession of unit 5.

Both unit 4 and unit 5 super®cially resembleoffshore±shoreface parasequences, as de®ned byVan Wagoner et al. (1990), and were interpretedas such by Elder et al. (1994). The internalstructure of these units, however, allows inter-pretation of both transgressive and regressiveportions (systems tracts) of depositionalsequences within both units and is discussedfurther below.

Units 6A and 6B

A signi®cant basinward shift in facies occurs atthe top of unit 5, an erosional surface with up to

10 m relief, traceable throughout the study regionand locally penetrated by plant roots, whichjusti®es the interpretation of a sequence boundaryunderlying unit 6A. Additional supportingevidence is the separation of the distal part ofunit 6A from the underlying deposits by a bedcontaining reworked coal material. This bed isinterpreted as a lag deposit overlying a waveravinement surface, i.e. a surface that was initiallyexposed subaerially and covered by a mire. Thisinterpretation means that the marine lowstanddeposits overlying the sequence boundary weredeposited east of the Kaiparowits area. A tidalchannel below the base of unit 6A near Section 15(Fig. 2) may correspond to this lowstand, but thecorrelation is tentative because of discontinuousexposure. Apart from the major sequence boun-dary separating units 5 and 6A, the evidenceprovided by incised channels in coals at the topof unit 6A and tree stumps overlain by theExogyra sandstones of unit 6B suggest an episodeof subaerial exposure before the deposition ofunit 6B.

The 5/6A sequence boundary correlateschronostratigraphically with a more deeplyincised sequence boundary, of up to 20 m relief,in the upper part of the Twowells Sandstonein north-western New Mexico (Mellere, 1994;UlicÏnyÂ, 1994). The evidence for a relative sea-level fall contradicts earlier interpretations byGustason (1989) and Elder & Kirkland (1994) thatthe uppermost parts of the Dakota formed duringa short-term sea-level stillstand.

The deposition of unit 6 took place in severalstages: (i) initial transgression, causing the estab-lishment of a mire on the sequence boundary; (ii)marine transgression leading to destruction of themire and deposition of tide-dominated, estuarinefacies in the western part of the area (unit 6A,proximal) and, later, a highstand or falling stageleading to the deposition of wave-dominated,strandplain deposits (unit 6A, distal); (iii) minorrelative sea-level fall, which caused local subae-rial erosion of unit 6A; and (iv) ¯ooding leadingto the deposition of unit 6B.

The formation of the wave ravinement surfaceat the base of the offshore Tropic Shale markedthe major ¯ooding event of the early S. gracileZone (cf. Gustason, 1989; am Ende et al., 1991).The low-diversity molluscan assemblage formedduring conditions of low sediment supply andpossibly high in¯ux of nutrients, indicatingabrupt deepening.



DISCUSSION

Sequences vs. parasequences

Units 1±6 of the Dakota Formation have beende®ned on the basis of the interpretation of theirbounding surfaces: all unit boundaries wereshown to represent ¯ooding (and correlative)surfaces. Most of them formed initially duringbase-level falls, i.e. as sequence boundaries,sensu Van Wagoner et al. (1988). The interpreta-tion of the individual units as sequences, notparasequences, as thought earlier by Elder et al.(1994) and UlicÏnyÂ (1994), stems from the evi-dence for relative sea-level falls discussed inprevious sections. The ambiguity of interpretat-ion of the basal surfaces of units 3B, 4 and 5emphasizes the problems concerning the distinc-tion between parasequences and high-frequency

sequences. Especially where evidence of subaeri-al erosion is equivocal, or where it disappears in abasinward direction, it is dif®cult to distinguishbetween the two categories (Figs 8 and 11).

A parasequence (Van Wagoner et al., 1990) is a`relatively conformable succession of geneticallyrelated beds or bedsets bounded by marine¯ooding surfaces or their correlative surfaces', inwhich the vertical facies succession re¯ects agradual decrease in water depth. Kamola & VanWagoner (1995, p. 31) state explicitly that `thereare no ¯ooding surfaces nor unconformitieswithin a parasequence'. The shallow-marineunits 4, 5 and 6A, however, show characteristicscontradicting the parasequence de®nition. Units4 and 5 show a more complicated internalorganization of facies assemblages than a simpleshallowing upward associated with parasequenceprogradation. The locally preserved transgressive

Fig. 11. Conceptual, simpli®ed cross-sections along depositional dip, illustrating the formation of the transgressiveand regressive portions of high-frequency sequences exempli®ed by units 4 to 6A. (a) Main components of atransgressive barrier system, with accommodation created in the backbarrier region and wave ravinement reworkingpart of the previous sequence as well as older transgressive deposits. (b) Minor, short-term, relative sea-level (RSL)fall superimposed on a long-term rise (e.g. unit 5, see inset) result in a long-term retrogradational stacking pattern ofthe high-frequency sequences. (c) Transgressive deposits ®ll a pronounced relief of the underlying sequence boun-dary; the regressive portion of the sequence is almost detached from the transgressive systems tract because the high-frequency relative sea-level rise is superimposed on a low-frequency fall (exempli®ed by unit 6A, inset), resulting ina long-term progradational stacking. Symbols in insets show positions of underlying sequence boundaries andposition of the time slices (b) and (c) in the relative sea-level curve. The systems tracts de®ned are facies based. TST,transgressive systems tract; RST, `regressive systems tract' (regressive deposits in which the boundary between thehighstand and falling stage systems tracts cannot be identi®ed); SB, sequence boundary; TS, transgressive surface;LST, lowstand systems tract. No scale implied.

828 D. UlicÏnyÂ


back-barrier and inlet deposits are interpreted as atransgressive systems tract, whereas the offshore±shoreface successions form the highstand tofalling stage portion of a sequence (Figs 8e and11), which is called here informally the regressivesystems tract because of dif®culties in ®eldrecognition of the boundary between the high-stand and falling stage deposits. The systemstracts are de®ned on the basis of facies and keysurfaces, not on the basis of internal geometricarrangement (stacking pattern) of depositionalunits, which is typical of high-frequencysequences at a scale similar to that of the Dakotaunits (cf. Swift et al., 1991a,b; Plint, 1996). Thus,units 4 and 5 can be termed high-frequencysequences, although some of their boundingsurfaces cannot be shown unequivocally to rep-resent sequence boundaries (for instance, at thetop of unit 4, any evidence of subaerial erosionmay have been destroyed by subsequent trans-gressive erosion). On the contrary, unit 6A isbracketed by relatively well-constrained sequenceboundaries. Its proximal facies assemblage isinterpreted as estuarine deposits of the transgres-sive systems tract, whereas most of the distalportion of this unit is interpreted as the `regres-sive' systems tract, based on the spatial distribu-tion of facies. The division of the sequences intothe facies-based, transgressive and regressiveportions is conceptually very close to the depo-sitional couplets of Devine (1991).

Although a more general discussion of thesequence vs. parasequence de®nitions is beyondthe scope of this paper, the examples from theùpper Dakota' illustrate that the distinctionbetween high-frequency sequences and parase-quences commonly depends primarily on theavailability of evidence and the amount of data(Christie-Blick & Driscoll, 1995; Swift et al.,1991b). At the scale of the high-frequencysequences observed in the Dakota Formation, itis considered more practical to use the term`parasequence' for a special case of a depositionalsequence (cf. Swift et al., 1991b), rather than toamend the Van Wagoner et al. (1988, 1990)parasequence de®nition by including transgres-sive deposits (Arnott, 1995).

Periodicity and magnitude of relativesea-level changes

Good geochronological and biostratigraphicalconstraints are available for units 4±6B (Table 1),but the accuracy of dating decreases downsection,

which affects the estimates of relative sea-levelchange periodicity. In the case of unit 1, forwhich reliable information on age is absent, norelationship between the deposition and somehypothetical period of base-level change can beestablished.

The long-term (up to 1á4 Myr) cycle of base-level change, recorded by the deposition of unit 2,is an order of magnitude longer than the cyclesrecorded in the overlying units (Table 1). The factthat the thickness of unit 2 is commonly close tothat of units 4±6B combined (total time ofdeposition »400 kyr; Table 1 and below) illus-trates the importance of compaction in modifyingthe preserved thickness. Considering compactionratios and peat growth rates, Kirschbaum &McCabe (1992) concluded that coals and carbo-naceous muds in strata equivalent to unit 2 mayrepresent up to 90% of the time of deposition. Anumber of relative base-level changes of higherfrequency may have in¯uenced the deposition ofunit 2, but their understanding will require adetailed correlation of the numerous coals andpalaeosols within the unit.

The M. mosbyense biozone, which comprisesunits 4±6B, represents 0á4 Myr or less, based onObradovich (1993). The simplistic assumptionthat units 4±6B each represent equal incrementsof time gives a maximum average frequency ofrelative sea-level change of 100 kyr per unit.Similar periodicities, suggesting the involvementof Milankovitch forcing in relative sea-levelchange, were found by Plint (1996) in the Earlyto Middle Cenomanian Dunvegan Formation ofWestern Canada (»150 kyr) and by UlicÏnyÂ &SÏpicÏaÂkovaÂ (1996) in the Middle to Late Cenoman-ian of Bohemia (100±140 kyr). Based on the Ar/Arages from coal zone 3 (»94á5 Ma), units 3A and 3Brepresent a similar frequency of relative sea-levelchanges (Table 1).

The magnitudes of the relative sea-level ¯uctu-ations are dif®cult to determine exactly, but thecomparison of palaeobathymetric interpretationsof facies juxtaposed across the bounding surfacesof units 3A to 6B allows an estimate of 10±20 m,possibly up to 30 m at the 5/6B sequence boun-dary.

Regional correlation of Dakota units

Elder et al. (1994) established regional correlationbetween carbonate-shale couplets in the pelagicrhythmites of the central part of the West-ern Interior and the nearshore `parasequences'



in southern Utah (including the ùpper Dakota'of the Kaiparowits area). Such correlation seemsto support the above hypothesis of Milanko-vitch forcing of relative sea-level changes. How-ever, part of the correlation by Elder et al.(1994) is based on recognition of only two`parasequences' in the ùpper Dakota' in theKaiparowits region: units 6A and 6B are con-sidered parts of their `parasequence 2' and, inthe SE Kaiparowits area, marine mudstones ofunit 5 are treated as a downdip continuation ofunit 4 (their `parasequence 1'). Also, Elder &Kirkland (1993) do not recognize a relative sea-level fall around the M. mosbyense/S. gracileboundary.

Two linked problems arise when comparing the`parasequences' of Elder et al. (1994) and theunits 4±6B presented in this study. First, thisstudy shows that more than two units separatedby ¯ooding surfaces occur in the ùpper Dakota'in the southern Kaiparowits area, and thus thecorrelation of individual surfaces to concretionand limestone beds across some 200 km betweenthe Kaiparowits area and the nearest offshoresection in Elder et al. (1994), as well as thenumber of cycles correlated, should be revised.Secondly, the evidence for sequence boundaries,especially 5/6A and 6A/6B, presented hereimplies that the theoretical concept of correlatingthe nearshore `parasequence' ¯ooding surfaces todistal basinal rhythms should be modi®ed toinclude also the response to relative sea-level falland lowstand deposition in the distal part of thebasin (e.g. considering the role of skeletal lime-stones; cf. Sageman, 1996).

Regional and local tectonic controlson depositional style

The main control on the depositional style of theDakota Formation was regional ¯exural subsi-dence within the foreland basin. The overallnorth-westward increase in total thickness of theDakota Formation re¯ects the position of the studyregion within the foredeep depozone of theforeland basin system (sensu DeCelles & Giles,1996) during the Cenomanian, where it remaineduntil the Santonian (cf. Eaton & Nations, 1991;Leithold, 1994). In the region west of the Pau-nsaugunt Fault, the total thickness of strata coevalto units 1±6B reaches more than 300 m. Gustason(1989) reported an incremental westwardincrease in thickness of the Dakota across prom-inent fault zones, such as the Paunsaugunt Fault,

and concluded that pre-existing basement frac-tures, approximately parallel to the orogenicfront, were reactivated as normal faults and addeda local component of subsidence to the regional¯exural pattern. Schwans (1995) termed theportion of the foreland basin characterized bythis subsidence pattern a `proximal zone' of theforeland basin.

The study region, however, is located east ofthe Paunsaugunt Fault and belongs to the `distalzone' of the foredeep depozone sensu Schwans(1995), where similar rapid lateral changes inthickness are far less pronounced. In the Hen-rieville area, abrupt changes in thickness of mostunits and local erosion of most of unit 4 wererecorded in sections spaced 700 m apart acrossthe NNE-striking Dry Creek valley, which sepa-rates Sections 3 and 4. This valley probablyfollows a basement fault roughly parallel to thePaunsaugunt and other major faults. The erosionof unit 4 in Section 4 (Fig. 2) is interpreted as theresult of a coincidence of transgressive erosionand minor local uplift along a reactivated base-ment fault. Elsewhere, similar local changes inthickness of individual units are associated withdifferential compaction, namely of units 2, 3Aand 3B, where channel-®ll sandstones laterallyinter®nger with mudrocks and coals. On shorttime scales, compaction rates may have beenmore important in creating accommodation thanthe relatively slow tectonic subsidence.

Abundant deformational structures in unit 2have been interpreted by Kirschbaum & McCabe(1992) as sediment liquefaction resulting fromseismic activity. However, these phenomenaoccur throughout unit 2, and it is dif®cult to linkthem to particular faults; a more likely explana-tion of liquefaction is high sedimentation rates inthe ¯uvial channels. Along the East KaibabMonocline, sandstone bodies generated byslumping on the shoreface of unit 4 are interpre-ted as recording local seismic activity (Figs 2 and8c). Also, the clastic dykes in unit 6A suggestinvolvement of seismic shocks in this area(Fig. 9b); they are downward injected, do notform polygonal patterns and penetrate variouslithologies, which implies that coal compactionalone was probably not suf®cient to cause theirformation (cf. discussion in Tanner, 1998, andreferences therein). The precursor of today's EastKaibab Monocline was apparently a fault parallelto and with a similar history to the Paunsaugunt,Sevier and Hurricane Faults in southern Utah (cf.Picha, 1986).

830 D. UlicÏnyÂ


Causes of high- and low-frequency sea-levelchanges

Based on the periodicity estimates above, theDakota units fall into two distinct groups: (i) units1 and 2, representing depositional sequencesformed in ¯uvial environments and recordingpresumably long-term (1á5 Myr for unit 2,unknown for unit 1) base-level changes; and (ii)units 3A to 6B, interpreted as high-frequencysequences formed in response to »100 kyr relativesea-level ¯uctuations. The overall stacking of theDakota units results from a combination of low-and high-frequency relative sea-level ¯uctuations(Fig. 12). However, no attempt is made here toestablish hierarchies of sequence-stratigraphicsubdivisions such as sequence and parasequencesets of various orders sensu Mitchum & VanWagoner (1991).

For unit 1, the onset of ¯exural subsidence wasthe most important causal mechanism, although apossible involvement of eustasy in the base-levelrise causing the ®lling of valleys cannot be ruledout. As long as the duration of depositionremains unknown, the 1/2 sequence boundarymay be explained as a response either to aeustatic fall or to crustal relaxation after atemporary cessation of thrusting activity. Themarked base-level rise recorded at the base ofunit 2 was probably linked to a eustatic rise, butit may have been assisted by accelerated back-

tilting of the coastal plain, which helped to trapthe coarser clastics closer to the source during thebeginning of deposition. If the eustatic explana-tion is adopted, unit 2 would correspond to along-term cycle analogous to the `third-order'cycles of Vail et al. (1991). Individual wide-spread coal zones and palaeosols within unit 2may represent a higher frequency signal of base-level changes.

For units 3A to 6B, eustasy is believed to havebeen the cause of the relative sea-level ¯uctua-tions of »10±20 m magnitude. All units occurconsistently across the study area, »65±70 kmalong depositional dip, although their thicknessesvary in response to regional changes in subsi-dence rate, local syndepositional faulting, localerosion during sea-level falls and differentialcompaction. The estimated periodicity of therelative sea-level changes, »100 kyr, falls withinthe range of the Milankovitch-band orbital cycles,the eccentricity cycle being the closest match,which favours climate-driven eustasy as themechanism responsible for the relative sea-levelchanges.

Some authors interpret ¯ooding events ofsimilar frequency as re¯ecting periods of punc-tuated thrusting (Kamola & Huntoon, 1995). Peper(1993) also includes changes in intraplatestress level in the group of mechanisms poten-tially causing high-frequency sea-level changes.However, the fact that the thicknesses of all

Fig. 12. Summary of the depositional history and sequence stratigraphy of the Dakota Formation in the Kaiparowitsregion, southern Utah. Comments in text.



high-frequency units are of similar orders ofmagnitude suggests that the mechanism drivingthe sea-level changes was cyclic, which supportsthe eustatic forcing, rather than thrust events orchanges in compressional stress.

Firm evidence for the above interpretation of aeustatic control on relative sea-level changeswould, of course, be inter-regional or globalcorrelation. Correlation of high-frequency sea-level changes in the Cenomanian is still verycomplicated but, for certain intervals of the LateCenomanian, the eustatic nature of low-frequencycyclicity is well documented by biostratigraphiccorrelation. The 5/6 sequence boundary in theKaiparowits region correlates to a marked sea-level fall at the end of the M. mosbyense biozone(Calycoceras naviculare in Europe) documented,for example, from the Twowells Sandstone, NewMexico (Mellere, 1994), British Chalk (Owen,1996), Tunisia (Robaszynski et al., 1993) or Cen-tral Europe (the `sub-Plenus' sea-level fall; Voigtet al., 1992; UlicÏnyÂ et al., 1997). Although stack-ing patterns in siliciclastic successions are sensi-tive to local rates of sediment input, the overallretrogradational stacking of high-frequency se-quences between unit 2 and the base of the TropicShale shows a marked similarity to the record oflate Middle±early Late Cenomanian, long-termsea-level rise in central Europe (UlicÏnyÂ & SÏpicÏaÂ-kovaÂ, 1996), punctuated by short-term relativefalls in sea level.

It is not the purpose of this paper to discuss thepossible driving mechanisms of high-frequencyeustatic sea-level changes, but it is worth notingthat the global mechanisms causing eustatic¯uctuations of such a short periodicity duringthe Middle±Late Cenomanian, a time intervalclose to the mid-Cretaceous greenhouse climaticpeak (Jenkyns et al., 1994), are still controversial.Although Cenomanian glacioeustasy producing10±?30 m of sea-level change is still dif®cult toaccept for many researchers (cf. discussion inChristie-Blick & Driscoll, 1995), some recentcomputer models produce seasonal or evenpermanent continental ice close to the SouthPole and thus raise the possibility of limitedglacioeustasy even during the Cenomanian. Al-ternative mechanisms could be changes in thevolume of groundwater stored on continents,which could be responsible for sea-level changesof up to 20 m over periods of 10±100 kyr, or oceansteric (thermohaline) volume changes (Revelle,1990).

CONCLUSIONS

1 The tectonic controls on the sedimentation ofthe Dakota Formation included:

(a) eastward migration of a forebulge causingthe tilt of pre-Cretaceous basement strata and theincision of a westward-deepening valley systeminto them;

(b) ¯exure-dominated subsidence regime of theforedeep depozone, which probably started dur-ing the early Cenomanian and controlled thelong-term subsidence rate and its dip-parallelaccommodation pro®le;

(c) reactivation of basement faults, which led toan abrupt westward increase in subsidence rateacross some faults, as well as minor uplifts,which caused local erosion and submarineslumping.

2 Eustatic sea-level changes had the followingeffects:

(a) formation of regionally extensive ¯oodingsurfaces, mostly coinciding with previouslyformed sequence boundaries, which separate theDakota units;

(b) the preserved thicknesses of individualDakota units vary because of regional and localsubsidence rate variations as well as because oflocal incision on sequence boundaries, but allunits occur consistently across the study region;

(c) the ¯uvial strata of units 1 and 2 are cappedby erosional surfaces, interpreted as updip equiv-alents of sequence boundaries; the coal zones atthe base and within unit 2 may represent aresponse to higher frequency ¯ooding events;

(d) the shallow-marine units, interpreted asparasequences by earlier authors, can be dividedinto facies-based systems tracts and show signs ofsubaerial exposure on their boundaries, whichallows them to be interpreted as high-frequencysequences;

(e) the overall stacking pattern of the high-frequency sequences is retrogradational, re¯ect-ing a long-term relative sea-level rise, punctuatedby brief periods of relative sea-level falls, includ-ing a relatively major fall near the end of theM. mosbyense Zone; this is similar to the recordof Cenomanian relative sea-level change in otherregions, e.g. central Europe or northern Africa,suggesting that eustasy also played an importantrole in governing the low-frequency relative sea-level changes.3 The driving mechanisms of the eustatic changesare controversial. The high-frequency relative


832 D. UlicÏnyÂ

sea-level ¯uctuations of »100 kyr periodicity and»10±20 m magnitude were probably related toMilankovitch-band, climate-driven eustasy. Ei-ther minor glacioeustatic ¯uctuations, superim-posed on the overall greenhouse climate of themid-Cretaceous, or mechanisms such as the¯uctuations in groundwater volume on conti-nents, or thermal expansion and contraction ofseawater, may have controlled the high-frequencyeustatic ¯uctuations.

ACKNOWLEDGEMENTS

Special thanks are due to Jeff Eaton for havingintroduced me to the ®eld area, for generouslogistical support that made the project possible,for fruitful discussions and camaraderie duringthe ®eldwork periods. Criticisms and suggestionsby Gary Nichols helped to improve the early draftof the paper, and Gus Gustason is thanked for aninformal discussion of different views on Dakotasequence stratigraphy. I am grateful to thereviewers, Peter Schwans and Mark Kirschbaum,and to Sedimentology editor Guy Plint forincisive reviews and suggestions that improvedthe paper signi®cantly. M. Kirschbaum is thankedfor providing unpublished data on bentonite Ar/Ar ages. Marcela SvobodovaÂ provided unpub-lished palynological data. Rita UlicÏnaÂ was themost patient and enthusiastic ®eld assistant. The®eldwork was supported ®nancially by GSAFoundation grant no. 5071-92 and by donors tothe Czechoslovak Charta 77 Foundation. Thisstudy was initiated during part of the author'sdoctoral study at the University of Illinois,Champaign-Urbana.

REFERENCES

Ainsworth, R.B. and Walker, R.G. (1994) Control ofestuarine valley-®ll deposition by ¯uctuations ofrelative sea-level, Cretaceous Bearpaw-HorseshoeCanyon transition, Drumheller, Alberta, Canada. In:Incised-Valley Systems: Origin and SedimentarySequences (Ed. by R.W. Dalrymple, R. Boyd and B.A.Zaitlin), SEPM Spec. Publ., 51, 159±174.

Allen, G.P. (1991) Sedimentary processes and facies inthe Gironde estuary: a recent model for macrotidalestuarine systems. In: Clastic Tidal Sedimentology(Ed. by D.G. Smith, G.E. Reinson, B.A. Zaitlin and R.A.Rahmani), Can. Soc. Petrol. Geol. Mem., 16, 29±40.

Angevine, C.L. and Heller, P.L. (1987) Quantitativebasin modelling. Geol. Soc. Am. Short Course Notes,pp. 1±80.

Armstrong, R.L. (1968) Sevier orogenic belt in Nevadaand Utah. Geol. Soc. Am. Bull., 79, 429±458.

Arnott, R.W.C. (1995) The parasequence de®nition ±are transgressive deposits inadequately addressed?J. Sedim. Res., B65, 1±6.

Aubrey, W.M. (1992) New interpretations of the stra-tigraphy and sedimentology of uppermost Jurassic tolowermost Cretaceous strata in the San Juan Basin ofnorthwestern New Mexico. US Geol. Survey Bull.,1808J, 1±17.

Bally, A.W. (1989) Phanerozoic basins of North Amer-ica. In: The Geology of North America ± an Overview(Ed. by A.W. Bally and A.R. Palmer), Chap. 15, Vol.A, The Geology of North America, pp. 397±443.Geological Society of America, Boulder, CO.

Bilodeau, W.L. (1986) The Mesozoic Mogollon High-lands, Arizona: an early Cretaceous rift shoulder.J. Geol., 94, 724±735.

Bohor, B.F., Dalrymple, G.B., Triplehorn, D.M. andKirschbaum, M.A. (1991) Argon/Argon dating oftonsteins from the Dakota Formation, Utah. Geol.Soc. Am., Abstracts with Programs, 23, A85.

Christie-Blick, N. and Driscoll, N.W. (1995) Sequencestratigraphy. Ann. Rev. Earth Planet. Sci., 23,451±478.

Cobban, W.A. and Hook, S.C. (1984) Mid-Cretaceousmolluscan biostratigraphy and paleogeography ofsouthwestern part of Western Interior, United States.In: Jurassic±Cretaceous Biochronology and Paleogeo-graphy of North America (Ed. by G.E.G. Westerm-ann), Geol. Assoc. Can. Spec. Paper, 27, 257±271.

Cross, T.A. (1986) Tectonic control of foreland basinsubsidence and Laramide style deformation, westernUnited States. In: Foreland Basins (Ed. by P.A. Allenand P. Homewood), IAS Spec. Publ., 8, 15±39.

Currie, B.S. (1997) Sequence stratigraphy of nonmarineJurassic±Cretaceous rocks, central Cordilleran fore-land-basin system: Geol. Soc. Am. Bull., 109,1206±1222.

Currie, B.S. (1998) Upper Jurassic±Lower CretaceousMorrison and Cedar Mountain Formations, NE Utah±NW Colorado: relationships between nonmarinedeposition and early Cordilleran foreland-basin de-velopment. J. Sedim. Res., 68, 632±654.

DeCelles, P.G. and Giles, K.A. (1996) Foreland basinsystems. Basin Res., 8, 105±123.

Devine, P.E. (1991) Transgressive origin of channelledestuarine deposits in the Point Lookout Sandstone,northwestern New Mexico: a model for Upper Cre-taceous, cyclic regressive parasequences of the U.S.Western Interior. Am. Assoc. Petrol. Geol. Bull., 75,1039±1063.

Doelling, H.H. and Graham, R.L. (1972) SouthwesternUtah coal ®elds: Alton, Kaiparowits Plateau andKolob-Harmony. Utah Geol. Mineral. Survey Monogr.Series, 1, 1±333.

Eaton, J.G. and Nations, J.D. (1991) Introduction: tec-tonic setting along the margin of the CretaceousWestern Interior Seaway, southwestern Utah andNorthern Arizona. In: Stratigraphy, DepositionalEnvironments, and Sedimentary Tectonics of theWestern Margin, Cretaceous Western Interior Seaway



(Ed. by J.G. Eaton and J.D. Nations), Geol. Soc. Am.Spec. Paper, 260, 1±8.

Elder, W.P. and Kirkland, J.I. (1994) Cretaceous Paleo-geography of the southern Western Interior region.In: Mesozoic Systems of the Rocky Mountain Region,USA (Ed. by M.V. Caputo, J.A. Peterson and K.J.Franczyk), SEPM Rocky Mountain Section, 415±440.

Elder, W.P., Gustason, E.R. and Sageman, B.B. (1994)Correlation of basinal carbonate cycles to nearshoreparasequences in the Late Cretaceous Greenhornseaway, Western Interior, USA. Geol. Soc. Am. Bull.,106, 892±902.

am Ende, B. (1991) Depositional environments,palynology, and age of the Dakota Formation, south-central Utah. In: Stratigraphy, DepositionalEnvironments, and Sedimentary Tectonics of theWestern Margin, Cretaceous Western Interior Seaway(Ed. by J.D. Nations and J.G. Eaton), Geol. Soc. Am.Spec. Paper, 260, 65±83.

Flint, S., Aitken, J. and Hampson, G. (1995) Applica-tion of sequence stratigraphy to coal-bearing coastalplain successions: implications for the UK CoalMeasures. In: European Coal Geology (Ed. by M.K.Whateley and D.A. Spears), Geol. Soc. Spec. Publ.,82, 1±16.

FuÈ rsich, F.T. and Kirkland, J.I. (1986) Biostratinomyand paleoecology of a Cretaceous brackish lagoon.Palaios, 1, 543±560.

Gustason, E.R. (1989) Stratigraphy and Sedimentologyof the Middle Cretaceous (Albian-Cenomanian)Dakota Formation, Southwestern Utah. PhD Thesis,University of Colorado, Boulder.

Heller, P.L. and Paola, C. (1989) The paradox of LowerCretaceous gravels and the initiation of thrusting inthe Sevier orogenic belt, United States western inte-rior. Geol. Soc. Am. Bull., 101, 864±875.

Heller, P.L., Beekman, F., Angevine, C.L. and Cloe-tingh, S.A.P.L. (1993) Cause of tectonic reactivationand subtle uplifts in the Rocky Mountain region andits effect on the stratigraphic record. Geology, 21,1003±1006.

Jenkyns, H.C., Gale, A.S. and Cor®eld, R.M. (1994)Carbon- and oxygen-isotope stratigraphy of theEnglish Chalk and Italian Scaglia and its palaeocli-matic signi®cance. Geol. Mag., 131, 1±34.

Kamola, D. and Huntoon, J.E. (1995) Repetitive stratalpatterns in a foreland basin sandstone and theirpossible tectonic signi®cance. Geology, 23, 177±180.

Kamola, D. and Van Wagoner, J.C. (1995) Stratigraphyand facies architecture of parasequences with exam-ples from the Spring Canyon Member, BlackhawkFormation, Utah. In: Sequence Stratigraphy of Fore-land Basin Deposits (Ed. by J.C. Van Wagoner and G.T.Bertram), Am. Assoc. Petrol. Geol. Mem., 64, 27±54.

Kirschbaum, M.A. and McCabe, P.J. (1992) Controls onthe accumulation of coal and on the development ofanastomosed ¯uvial systems in the Cretaceous Da-kota Formation of southern Utah. Sedimentology, 39,581±598.

Kvale, E.P. and Archer, A.W. (1990) Tidal deposits asso-ciated with low-sulfur coals, Brazil Fm. (Lower Penn-sylvanian), Indiana. J. Sedim. Petrol., 60, 563±574.

Leckie, D.A. and Singh, C. (1991) Estuarine deposits ofthe Albian Paddy Member (Peace River Formation)and lowermost Shaftesbury Formation, Alberta,Canada. J. Sedim. Petrol., 61, 825±849.

Leithold, E.L. (1994) Stratigraphic architecture at themuddy margin of the Cretaceous Western InteriorSeaway, southern Utah. Sedimentology, 41, 521±542.

Mellere, D. (1994) Sequential development of an estu-arine valley ®ll: the Twowells Tongue of the DakotaSandstone, Acoma Basin, New Mexico. J. Sedim.Res., B64, 500±515.

Miall, A.D. (1985) Architectural-element analysis: anew method of facies analysis applied to ¯uvial de-posits. Earth-Sci. Rev., 22, 261±308.

Mitchum, R.M. and Van Wagoner, J.C. (1991) High-frequency sequences and their stacking patterns:sequence-stratigraphic evidence of high-frequencyeustatic cycles. Sedim. Geol., 70, 131±160.

Nio, S.-D. and Yang, C.-S. (1991) Diagnostic attributesof clastic tidal deposits: a review. In: Clastic TidalSedimentology (Ed. by D.G. Smith, G.E. Reinson,B.A. Zaitlin and R.A. Rahmani), Can. Soc. Petrol.Geol. Mem., 16, 3±28.

Obradovich, J.D. (1993) A Cretaceous time scale. In:Evolution of the Western Interior Basin (Ed. byW.G.E. Caldwell and E.G. Kauffman), Geol. Assoc.Can. Spec. Paper, 39, 379±396.

Oertel, G.F., Henry, V.J. and Foyle, A.M. (1991) Impli-cations of tide-dominated lagoonal processes on thepreservation of buried channels on a sediment-starvedcontinental shelf. In: Shelf Sand and Sandstone Bod-ies: Geometry, Facies and Sequence Stratigraphy (Ed.by D.J.P. Swift, G.F. Oertel, R.W. Tillman and J.A.Thorne), IAS Spec. Publ., 14, 379±394.

Olsen, T., Steel, R., Hogseth, K., Skar, T. and Roe, S.-L.(1995) Sequential architecture in a ¯uvial succession:sequence stratigraphy in the Upper CretaceousMesaverde Group, Price Canyon, Utah. J. Sedim.Res., B65, 265±280.

Owen, D. (1996) Interbasinal correlation of the Cen-omanian stage; testing the lateral continuity ofsequence boundaries. In: High Resolution SequenceStratigraphy: Innovations and Applications (Ed. byJ.A. Howell and J.F. Aitken), Geol. Soc. Spec. Publ.,104, 269±293.

Peper, T. (1993) Tectonic Control on the SedimentaryRecord in Foreland Basins. PhD Thesis, Vrije Univ-ersiteit, Amsterdam.

Peterson, F. (1969) Cretaceous sedimentation and tec-tonism in the southeastern Kaiparowits region, Utah.USGS Open-®le Report.

Picha, F. (1986) The in¯uence of preexisting tectonictrends on geometries of the Sevier orogenic belt andits foreland in Utah. In: Paleotectonics and Sedi-mentation in the Rocky Mountain Region, UnitedStates (Ed. by J.A. Peterson), Am. Assoc. Petrol. Geol.Mem., 41, 309±320.

Plint, A.G. (1996) Marine and nonmarine systemstracts in fourth-order sequences in the Early-Middle Cenomanian, Dunvegan Alloformation,northeastern British Columbia, Canada. In: HighResolution Sequence Stratigraphy: Innovations and


834 D. UlicÏnyÂ

Applications (Ed. by J.A. Howell and J.F. Aitken),Geol. Soc. Spec. Publ., 104, 159±191.

Rahmani, R.A. (1988) Estuarine tidal channel andnearshore sedimentation of a Late Cretaceousepicontinental sea, Drumheller, Alberta, Canada. In:Tide-In¯uenced Sedimentary Environments andFacies (Ed. by P.L. de Boer et al.), pp. 433±471. D.Reidel Publishing, Dordrecht.

Revelle, R. (ed.) (1990) Sea Level Change. NationalResearch Council, Studies in Geophysics. NationalAcademy Press, Washington, DC.

Robaszynski, F. et al. (1993) Sequence stratigraphy in adistal environment: the Cenomanian of the KalaatSenan region (central Tunisia). Bull. Centres Rech.Explor.-Prod. Elf Aquitaine, 17, 395±433.

Royse, F. (1993) Case of the phantom foredeep: EarlyCretaceous inwest-centralUtah.Geology,21,133±136.

Sageman, B.B. (1996) Lowland tempestites: deposi-tional model for Cretaceous skeletal limestones,Western Interior basin. Geology, 24, 888±892.

Schwans, P. (1995) Controls on sequence stacking and¯uvial to shallow-marine architecture in a forelandbasin. In: Sequence Stratigraphy of Foreland BasinDeposits (Ed. by J.C. Van Wagoner and G.T. Bertram),Am. Assoc. Petrol. Geol. Mem., 64, 55±102.

Shanley, K.W. and McCabe, P.J. (1995) Sequence stra-tigraphy of Turonian±Santonian strata, KaiparowitsPlateau, southern Utah, USA. Implications forregional correlation and foreland basin evolution. In:Sequence Stratigraphy of Foreland Basin Deposits(Ed. by J.C. Van Wagoner and G.T. Bertram), Am.Assoc. Petrol. Geol. Mem., 64, 103±136.

Shanley, K.W., McCabe, P.J. and Hettinger, R.D. (1992)Tidal in¯uence in Cretaceous ¯uvial strata fromUtah, USA: a key to sequence-stratigraphic interpr-etation. Sedimentology, 39, 905±930.

Smith, D.G. (1986) Anastomosing river deposits, sedi-mentation rates and basin subsidence, MagdalenaRiver, northwestern Colombia, South America.Sedim. Geol., 46, 177±196.

Stokes, W.L. and Heylmun, E.B. (1963) Tectonic historyof southwestern Utah. In: Guidebook to the Geologyof Southwestern Utah (Ed. by E.B. Heylmun), Inter-mountain Assoc. Petrol. Geol., 12th Annual FieldConference, pp. 19±25.

Swift, D.J.P., Phillips, S. and Thorne, J.A. (1991a)Sedimentation on continental margins, IV: lithofa-cies. In: Shelf Sand and Sandstone Bodies: Geome-try, Facies and Sequence Stratigraphy (Ed. by D.J.P.Swift, G.F. Oertel, R.W. Tillman and J.A. Thorne),IAS Spec. Publ., 14, 89±152.

Swift, D.J.P., Phillips, S. and Thorne, J.A. (1991b)Sedimentation on continental margins, V: parase-quences. In: Shelf Sand and Sandstone Bodies:Geometry, Facies and Sequence Stratigraphy (Ed. byD.J.P. Swift, G.F. Oertel, R.W. Tillman and J.A.Thorne), IAS Spec. Publ., 14, 153±188.

Tanner, P.W.G. (1998) Interstratal dewatering origin forpolygonal patterns of sand-®lled cracks: case studyfrom late Proterozoic metasediments of Islay, Scot-land. Sedimentology, 45, 71±90.

Thomas, R.G., Smith, D.G., Wood, J.M., Visser, J.,Calverley-Range, E.A. and Koster, E.H. (1987)Inclined heterolithic strati®cation ± terminology,description, interpretation and signi®cance. Sedim.Geol., 53, 123±179.

UlicÏnyÂ, D. (1994) Mid- to Late-Cenomanian high-fre-quency sea-level change: correlation between theWestern Interior of the US and the Bohemian Creta-ceous Basin, central Europe. In: Abstract Volume,High Resolution Sequence Stratigraphy: Innovationsand Applications, pp. 101±102. University of Liver-pool, Liverpool.

UlicÏnyÂ, D. (1996) Interplay of eustasy and tectonics in aforeland basin: sequence stratigraphy of the DakotaFormation (Cenomanian), southern Utah. In: Ab-stracts, 5th International Cretaceous Symposium and2nd Workshop on Inoceramids, Freiburg.

UlicÏnyÂ, D. and SÏpicÏaÂkovaÂ, L. (1996) Response to high-frequency sea-level change in a ¯uvial to estuarinesuccession: Cenomanian palaeovalley ®ll, BohemianCretaceous Basin. In: High Resolution SequenceStratigraphy: Innovations and Applications (Ed. byJ.A. Howell and J.F. Aitken), Geol. Soc. Spec. Publ.,104, 247±268.

UlicÏnyÂ, D., HladõÂkovaÂ, J., Attrep, M., CÏ ech, S., HradeckaÂ,L. and SvobodovaÂ, M. (1997) Sea-level changes andgeochemical anomalies across the Cenomanian±Turonian boundary: PecõÂnov quarry, Bohemia. Pal-aeogeogr. Palaeoclimatol. Palaeoecol., 132, 265±285.

Vail, P.R., Audemard, F., Bowman, S.A., Eisner, P.N.and Perez-Cruz, C. (1991) The stratigraphic signa-tures of tectonics, eustasy and sedimentology ± anoverview. In: Cycles and Events in Stratigraphy (Ed.by G. Einsele, A. Seilacher and W. Ricken), pp. 617±659. Springer-Verlag, Berlin.

Van Wagoner, J.C., Posamentier, H.W., Mitchum, R.M.,Vail, P.R., Sarg, J.F., Loutit, T.S., and Hardenbol, J.(1988) An overview of fundamentals of sequencestratigraphy and key de®nitions. In: Sea-LevelChanges ± an Integrated Approach (Ed. by C.K.Wilgus et al.), Soc. Econ. Palaeont. Mineral. Spec.Publ., 42, 39±44.

Van Wagoner, J.C., Mitchum, R.M., Campion, K.M. andRahmanian, V.D. (1990) Siliciclastic sequencestratigraphy in well logs, cores, and outcrops. Am.Assoc. Petrol. Geol. Methods in Exploration Series, 7,1±55.

Visser, M.J. (1980) Neap-spring cycles re¯ected inHolocene subtidal large-scale bedform deposits: apreliminary note. Geology, 8, 543±546.

Voigt, T., Pohl, T. and TroÈger, K.-A. (1992) Geologicalevidence for the sub-Plenus regression in Saxony. In:Abstracts of the 4th International Cretaceous Sym-posium, Hamburg.

Walker, R.G. and Plint, A.G. (1992) Wave and storm-dominatedshallow-marinesystems. In:FaciesModels:Response toSea-LevelChange. (Ed.byR.G.WalkerandN.P. James), pp. 219±238. Geol. Assoc. of Canada, St.John's.

Wolfe, D.G. (1989) The Stratigraphy and Paleoenvi-ronments of Middle Cretaceous Strata along the



Central Arizona±New Mexico Border. MSc Thesis,University of Colorado, Boulder.

Wright, V.P. and Marriott, S.B. (1993) The sequencestratigraphy of ¯uvial depositional systems: the roleof ¯oodplain sediment storage. Sedim. Geol., 86,203±210.

Yang, C.S. and Nio, S.D. (1989) An ebb-tide delta de-positional model ± a comparison between the mod-

ern Eastern Scheldt tidal basin (southwestNetherlands) and the lower Eocene Roda Sandstonein the southern Pyrenees (Spain). Sedim. Geol., 64,175±196.

Manuscript received 11 November 1996;revision accepted 13 January 1999.


836 D. UlicÏnyÂ

Documents

Sequence stratigraphy of the Dakota Formation (Cenomanian