Cretaceous Research (1999), 20, 1–27Article No. cres.1998.0133, available online at http://www.idealibrary.com on

Magnetostratigraphy, isotopic age calibrationand intercontinental correlation of the Red Birdsection of the Pierre Shale, Niobrara County,Wyoming, USA

*J. F. Hicks, †J. D. Obradovich and *L. Tauxe

*Scripps Institution of Oceanography, Geological Research Division, 9500 Gilman Drive, La Jolla,CA 92093-0220, USA†United States Geological Survey, Box 25046, Denver Federal Center, CO 90225, USA

Revised manuscript accepted 20 May 1998

The Red Bird section of the Pierre Shale in eastern Wyoming contains a relatively complete sequence of fine-grained marineclastics that were deposited between 81 and 69 million years ago in the Late Cretaceous epicontinental seaway of the USWestern Interior. These units not only contain a well-studied, high-resolution ammonite biostratigraphic sequence, by whichthe far-flung exposures of the seaway sediments are correlated across this region, but they are also isotopically well-dated dueto the presence of numerous sanidine-bearing volcanic ash layers. The magnetostratigraphy of the Red Bird section consistsof three geomagnetic reversals which can be independently calibrated by seven 40Ar/39Ar isotopic ages in an interval thatspans 12 million years of the Campanian and Maastrichtian stages. The magnetostratigraphic section can be confidentlycorrelated to that part of the geomagnetic polarity time scale (GPTS) that ranges from the base of subchron C33n to the baseof C31n. Linear interpolation and extrapolation from the isotopic ages gives the following age estimates for these reversalboundaries: C32n/C31r, 70.44&0.7 Ma; C31r/C31n, 69.01&0.5 Ma. The C33n/C32r reversal boundary cannot beidentified with complete confidence but it is certainly younger than the 74.62&1.2 Ma age interpolated for the reversal foundat the top of C33n. These age estimates make a significant contribution to the calibration of the GPTS for the CretaceousPeriod, which has previously relied heavily on interpolation between three or fewer calibration points that are widely spacedin age. In addition, the recognition of the chrons C33 through C31 in this section enables us to correlate the high resolutionammonite zonation of the US Western Interior directly to the time-equivalent European pelagic microfossil zonation basedon the magnetostratigraphic reference section at Gubbio in north-central Italy. ? 1999 Academic Press

K W: Cretaceous; Western Interior; Wyoming; Powder River Basin; magnetostratigraphy; isotopic age;chronistratigraphy; Pierre Shale; ammonite biostratigraphy; geomagnetic polarity time scale (GPTS).

1. IntroductionReconstruction of Earth’s history requires that eventsrecorded in rocks of various environments and agesbe placed into a common temporal framework. Oneof the most versatile and powerful tools for tyingtogether different geological sections is magneto-stratigraphy, in which a pattern of geomagneticpolarity zones is correlated to the geomagnetic polaritytime scale (GPTS). Magnetostratigraphy is versatilebecause it can be done in a variety of stratigraphicsettings and the GPTS is powerful because it serves asa central clearinghouse for a great store of age infor-mation derived from such diverse sources as bio-stratigraphy, isotopic decay, oxygen, carbon andstrontium isotope ratios and even the earth’s climaticresponse to orbital variations.

0195–6671/99/010001+27 $30.00/0

Since the ground breaking work of Heirtzler et al.(1968), there have been continual refinements to theGPTS. The single most important source of numeri-cal age calibration points is from isotopic datingmethods such as 40Ar/40K and 40Ar/39Ar. Theseisotopic calibration points have been tied to theGPTS both directly and indirectly, and the choice ofcalibration points is the principal difference amongthe time scales. Age information for the last tenmillion years of the GPTS has also come from tyingthe ‘astronomical’ time scale to the GPTS (e.g.,Shackleton et al., 1990; Hilgen, 1991), and thesetechniques have also been successfully applied toolder segments of the stratigraphical record (e.g.,Herbert & D’Hondt, 1990; Herbert et al., 1995).But the principal means of numerical calibration of

? 1999 Academic Press

2 J. F. Hicks et al.

104°W114°W

200 Kilometers0 100

112°W 110°W 108°W

48°N

44°N

42°N

46°N

106°W

Scale

Red Bird

Powder RiverBasin

Williston Basin

CANADA

Wyoming

North Dakota

South Dakota

Nebraska

Bighorn Basin

LaramieGreen River Basin

Idaho

Utah ColoradoDenver Basin

MontanaWyoming

Wind River Basin

Figure 1. General location of the Red Bird section in the Powder River Basin of easternmost Wyoming. For detailed sectionmaps and lines of section surveyed by the USGS, see Gill & Cobban (1966a).

the GPTS for the Late Cretaceous and Paleogene isby isotopic age data.

Adding to the database of calibration points fromwhich the GPTS is derived, we report here thediscovery of three geomagnetic reversals in the RedBird section of the Pierre Shale, which is locatedin eastern Wyoming (Figure 1). This sequence iswell known for the record it contains of the latestCretaceous ammonite biostratigraphy of the WesternInterior, and it is held to be an informal referencesection for the Campanian and Maastrichtian stages inthat region (Gill & Cobban, 1966a). The Pierre Shaleis a significant part of what is arguably one ofthe best dated marine sedentary sequences in theworld (Obradovich, 1993). During two intervals ofCretaceous time, peaking in the Cenomanian andagain in the Campanian-Maastrichtian, the WesternInterior foreland basin was inundated by ash falls froma magmatic arc developed along the western margin ofthe continent, and from volcanism associated with theemplacement of the Idaho, Boulder and Sierra Nevadabatholiths (Christiansen et al., 1994). The first geo-

chronologic time scale for this region was establishedby Obradovich & Cobban (1975), who isotopicallydated ash falls that ranged in age from latest Albian toearly Maastrichtian by the 40Ar/40K method, and ithas subsequently been revised using the more precise40Ar/39Ar laser fusion method (Obradovich, 1993).

The age control available in the Red Bird sectionallows us to unambiguously correlate the geo-magnetic reversals to the GPTS and estimate theirage by interpolation and extrapolation from datedlevels found both in section, or correlated biostrati-graphically from elsewhere. The biostratigraphicrecord of the Red Bird section contains 18 ammo-nite zones, which represents all but three of theammonite zones that are recognized in the PierreShale, and those three are cut out in a paraconform-ity whose stratigraphic position is very well known(Gill & Cobban, 1966a). This provides an importantindependent measure of the completeness of thesedimentary record preserved at Red Bird, and givesus confidence in our interpolations between thedated horizons.

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 3

The association of biostratigraphy, isotopic ages,and a geomagnetic reversal sequence in a single sur-face section presents a unique opportunity to calibratea part of the GPTS that is at present relatively poorlyconstrained. The first GPTS was derived from amagnetic profile from the South Atlantic that wasdated by simple extrapolation, and assumed relativelyconstant seafloor spreading rates (Heirtzler et al.,1968). This time scale has been updated many times,but the more recent revisions of the Upper Cretaceousinterval (Hallam et al., 1985; Kent & Gradstein, 1985;Haq et al., 1988; Harland et al., 1989; Cande & Kent,1992, 1995; Gradstein et al., 1994) have only twoor three dated calibration points from which tointerpolate.

An extensive revision of the GPTS has mostrecently been by Cande & Kent in 1992 and 1995.For the Late Cretaceous, the two Cande & Kent timescales differ in the age assigned to the K/T boundary(66.0 Ma in 1992 and 65.0 Ma in 1995). This dis-parity was caused by a change in the procedure forsample preparation which yielded the invalid age of66.0 Ma (Berggren et al., 1992), since corrected inSwisher et al. (1992) to 65.0 Ma. But there are still anumber of problems in the age assignments for thisinterval that need to be considered. The 1995 recali-bration of the GPTS is still restricted to just ninecalibration points for the most recent 83 millionyears of geologic time, with just three in the LateCretaceous (Cande & Kent, 1995). A perusal of theapproach taken to make up for the paucity of calibra-tion points reveals a number of shortcomings. Cande& Kent fit a natural cubic spline to pass exactlythrough the nine calibration points, and then calculatethe ages of the various chron boundaries. The ages forthe chron boundaries are reported to the nearestthousand years (3 decimal places) which gives theimpression of a precision that is not wholly warranted.The ages for the nine data points all carry analyticaluncertainties, plus the fact that the sea floor spreadinganomalies have uncertainties that range from 3% to asmuch as 17% for the various chron boundaries(Cande & Kent, 1995). However, these errors are nottaken into account in the cubic spline solution. Theresult is a very precise solution to a problem with veryimprecise parameters. The cubic spline solution hasanother drawback, in that when one changes a singledatum, the adjustments are not restricted to theinterval between the next oldest and youngest data.This change reverberates throughout the entireassemblage of splines. The reader is urged to comparethe two Cande & Kent (1992, 1995) latest Cretaceoustime scales that result from a simple change of asingular datum, the K/T boundary.

Another problem that has not been taken fully intoaccount is the possibility of analytical bias amongst thevarious sources of the data. For instance, Cande &Kent rely on the data of Obradovich (1988; andObradovich in Berggren et al., 1992) for the LateCretaceous, and then employ the data of Swisher et al.(1992) for the K/T boundary and the ‘minus 17 ash’of Denmark. However, no allowance was made for thedifference of some 500 000 years between the data ofObradovich & Swisher. This bias disappears at thenext younger calibration point of 33.7 Ma for theEocene/Oligocene boundary (Obradovich & Dockery,1995), but any solution between 34 and 80 Ma con-tains an error because of the combination of these twodata sets. Furthermore there are some discrepancies inthe ages that have been attributed to the Cretaceous-Tertiary (K/T) boundary. The K/T boundary levelhas been dated in a number of studies (Izett et al.,1991; McWilliams et al., 1992; Dalrymple et al.,1993). When all of the results are normalized to thecommon value for the monitor used in this study,the Taylor Creek Rhyolite, at 28.32 Ma (which isequivalent to 520.4 Ma for Mmhb-01; Samson &Alexander, 1987), an age of approximately 65.5 Ma isobtained. If the monitor ages are changed to alterna-tive values that are in common use by other lab-oratories, 27.92 Ma and 513.9 Ma for TCR andMmhb-1 respectively, then an age of 64.6 Ma results.None of these agree with the age obtained by Swisheret al. of 65.0 Ma. Gradstein et al. (1994) have thesame problems when generating their Mesozoic timescale. For the Late Cretaceous they utilize the data ofObradovich (1993) until they reach the K/T boundarywhere they then employ the data of Swisher et al.(1992), despite the studies cited above which wouldcall for a boundary older than 65.0 Ma when all thedata are normalized to the same monitor value.

This study begins to address a few of these prob-lems. We present a total of seven calibrating isotopicages within a single 12 million year interval of theCampanian and Maastrichtian. All but one of theseages are found either in direct association with thegeomagnetic reversals in the Red Bird section, or theyare obtained from other localities and correlated bio-stratigraphically. These ages were obtained from ashesthat were prepared and analysed in one laboratory,often in a single irradiation. This minimizes theuncertainty that arises from inter-laboratory andanalytical errors. It would appear that this would bethe ideal basis on which to estimate the ages of theCampanian and Maastrichtian reversal boundaries,but as this study shows, a very different set ofproblems and uncertainties is introduced.

4 J. F. Hicks et al.

2. Geologic setting

Throughout the Mesozoic, the Western Interior ofNorth America was occupied by a broad elongateasymmetric foreland basin that was submerged, atleast in part, beneath an epicontinental seaway fromthe late Aptian to the Paleocene. At its maximumdevelopment this seaway extended some 4900 kmfrom the Arctic Ocean to the Gulf of Mexico(Kauffman, 1977). Since at least the Jurassic, andthroughout the Cretaceous, the western edge of theNorth American plate was bounded by an activesubduction zone or trough that was in turn borderedby a melange wedge and a complex accretionarysystem of forearc basins and trenches (Hamilton,1978). To the east of the subduction zone was anemergent, tectonically active cordillera made up of aneastward migrating thrust belt (Elison, 1991) and amagmatic arc that formed in the core of a peninsulathat extended from the northeastern Asiatic platedown the western side of the North American plate(for overview, see Lehman, 1987). The western cor-dillera was the source for almost all of the fine-grainedclastic sediment that accumulated along the westernmargin of the foreland basin and in the epicontinentalseaway from the Campanian to the Maastrichtian.Extensive exposures of marine sediments that weredeposited in the foreland basin during the Campanianand Maastrichtian are found across Montana,Wyoming and the Dakotas. Originally named the FortPierre Group from outcrops in South Dakota alongthe Missouri River (Meek & Hayden, 1861), the namehas been shortened in general usage to Pierre Shale(Crandell, 1958), and changed to a formation of theMontana Group as defined by Gill & Cobban (1973).

A nearly continuous succession of the Pierre Shaleis exposed in surface section in Niobrara county, neara road junction that is named Red Bird on the maps ofeasternmost Wyoming. When traced to the west thesemarine shales are found to interfinger with wedges ofterrestrial and shallow marine strata (Figure 2) thatwere deposited in the coastal plain and nearshorepaleoenvironments that lay marginal to the seaway.The shoreline deposits were laid down in a broad beltthat trended north-south across western Montana andWyoming and fluctuated in response to periodic sea-level changes (Steidtmann, 1993) and tectonic activity(Weimer, 1984; Merewether & Cobban, 1986).Across the seaway as a whole, shoreline movementappears to have been controlled by a complex inter-action between eustatic sea-level change, thrustbelt tectonism and sediment loading and supply(Van Wagoner et al., 1990; Jordan, 1981; Jordan &Flemings, 1991; Vail et al., 1991).

From the late Santonian to the end of theMaastrichtian the broad pattern of Late Cretaceousshoreline evolution in the Western Interior was one ofa slow regional regression which was punctuated bya number of relatively rapid, local or subregionaloscillations in shoreline probably caused by tectonicmovements in the eastward migrating thrust belt (Gill& Cobban, 1966a, 1973; Merewether & Cobban,1986; Cobban et al., 1994). The stratigraphic extentand relationship of the wholly marine Red Birdsection to the terrestrial and nearshore facies to thewest, is shown in Figure 2. The Red Bird sectionspans a period marked by two major transgressions(locally named the Claggett and the Bearpaw, Figure3C), and two regressions (the Judith River and FoxHills; Lillegraven & Ostresh, 1990).

3. The Red Bird section

3.1. Introduction

The Pierre Shale is exposed at Red Bird in a con-tinuous 1100-m-thick section on the steep north-western limb of the Old Woman anticline, Niobraracounty, easternmost Wyoming (Figure 1). This well-studied exposure (Darton, 1901, 1918; Loomis,1915; Robinson et al., 1959; Gill & Cobban, 1961,1966a; Bergstresser & Frerichs, 1982; Bergstresser &Krebs, 1983; Bergstresser, 1983), is one of only a fewstratigraphic sections in the region where the PierreShale can be seen in its entirety. The section wasextensively studied by the US Geological Survey be-tween 1957 and 1962, and a detailed map, surveyedstratigraphic section and biostratigraphic zonationwere published (Gill & Cobban, 1966a). It has sinceserved as an informal reference locality for the latestCretaceous (Campanian to Maastrichtian) ammonitebiostratigraphy of the northern Great Plains region ofthe US Western Interior (although access is currentlyseverely restricted by the private landowner). Eighteenammonite zones have been recognized and establishedin this section (Figure 3B). Regionally the provincialammonite zonation defined at Red Bird is of greatstratigraphic importance and has been used to corre-late most of the latest Cretaceous marine sectionsacross the Western Interior (Gill & Cobban, 1965,1973; Shultz et al., 1980). The microfossil biostra-tigraphy of this interval has also been studied but,in comparison to the macrofossils, it is of inferiorresolution (Mello, 1969; Bergstresser & Frerichs,1982).

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 5

Har

ebel

lF

orm

atio

n

Bac

onR

idge

Fm

.

Lan

ce F

orm

atio

nR

ed B

ird

Sec

tion

Mee

teet

se F

orm

atio

n

Ju

dit

h R

iver

For

mat

ion

Soh

are

For

mat

ion

Fox

Hil

ls F

orm

atio

n

Tea

pot

San

dsto

ne

Par

kman

San

dsto

ne

Elk

Bas

in S

ands

ton

eCla

gget

t S

hal

e

Cod

y S

hal

e

Tel

egra

ph

Cre

ek

Bea

rpaw

Sh

ale

Campanian SantonianMaastrichtian

Am

mon

ite

Bio

zon

atio

nR

ed B

ird

Sec

tion

,N

iobr

ara

Co.

Wyo

min

g.G

ill &

Cob

ban

(19

66)

Bac

uli

tes

clin

olob

atu

s

Bac

uli

tes

gran

dis

Bac

uli

tes

bacu

lus

Bac

uli

tes

elia

si

Bac

uli

tes

jen

sen

i

Bac

uli

tes

rees

idei

Exi

telo

cera

s je

nn

eyi

Did

ymoc

eras

neb

rasc

ense

Bac

uli

tes

scot

tiB

acu

lite

s gr

egor

yen

sis

Bac

uli

tes

perp

lexu

s (l

ate)

Bac

uli

tes

gilb

erti

Bac

uli

tes

perp

lexu

s (e

arly

)B

acu

lite

s (s

moo

th)

Bac

uli

tes

aspe

rifo

rmis

Bac

uli

tes

mcl

earn

iB

acu

lite

s ob

tusu

s

Eag

le F

orm

atio

n

Nio

bra

raF

orm

atio

n

Vir

gell

e S

ands

ton

e

Pierre Shale

Fig

ure

2.S

ketc

hpa

neld

iagr

am,a

ligne

dbr

oadl

yw

est

toea

stac

ross

nort

hern

Wyo

min

g,sh

owin

gth

ege

nera

lstr

atig

raph

icre

lati

onsh

ips

ofth

efo

rmat

ions

ofth

eup

perm

ostC

olor

ado

and

Mon

tana

Gro

ups.

The

stra

tigr

aphi

cex

tent

and

appr

oxim

ate

posi

tion

ofth

eR

edB

ird

sect

ion

issh

own

bya

blac

klin

e.In

the

mar

ine

faci

esto

the

east

,the

Col

orad

oan

dM

onta

naG

roup

sar

ede

fined

litho

logi

cally

byth

eco

ntac

tbe

twee

nth

eP

ierr

eS

hale

and

the

chal

kyfa

cies

ofth

eN

iobr

ara

For

mat

ion.

Inth

ein

terc

alat

edm

arin

ean

dte

rres

tria

lfa

cies

toth

ew

est,

the

top

ofth

eC

olor

ado

Gro

uplie

sat

the

top

ofth

eC

ody

Sha

le,

and

the

Mon

tana

Gro

upex

tend

sfr

omth

eba

seof

the

Tel

egra

phC

reek

toth

eto

pof

the

Fox

Hill

sF

orm

atio

n.

6 J. F. Hicks et al.

3.2. Stratigraphy

The stratigraphic and biostratigraphic divisions of thePierre Shale at Red Bird are shown in Figure 3, partsA and B. For the detailed litho- and biostratigraphysee Gill & Cobban (1966a). The Pierre Shale at RedBird lies above the Niobrara Formation (Figure 2), acalcareous marine shale. The lowermost member ofthe Pierre Shale is the Gammon Ferruginous Member(labelled with an asterisk in Figure 3A), which con-sists of 10 m of dark grey, fissile and sideritic shales.The Pierre Shale has been divided into a number ofsequences, separated by unconformities, or sequencesurfaces that can be correlated to the relative sea-levelchanges (Van Wagoner et al., 1990). At the top ofthe Gammon Ferruginous Member lies a sequenceboundary, regionally known as the Pierre-Niobraraunconformity (DeGraw, 1975; Shurr & Reskind,1984; Weimer, 1984; Haq et al., 1988; Van Wagoneret al., 1990), which was formed by fluvial incision intoshelf sediments subaerially exposed during a fall in sealevel that occurred at about 80 Ma (Haq et al., 1988;Olsson, 1991). This sequence surface can be tracedacross the basin from the wholly terrestrial sedimentsof the Two Medicine Formation in northwesternMontana (Rogers et al., 1993), into the marine sedi-ments, where it lies at the base of the Claggett Shale(Hicks et al., 1995; Figure 2), to the Red Bird section,and out across the Dakotas and Nebraska (DeGraw,1975).

A rapid relative rise in sea level at the onset of theClaggett transgression drowned the incised anderoded topography (DeGraw, 1975) and 40 m of darkgrey, carbonaceous shale that comprises the regionallyextensive Sharon Springs Member (Elias, 1931) wasdeposited during the Claggett transgression (Figure3A, C). The Sharon Springs Member of the PierreShale has a complex geochemistry (Gill et al., 1972;Gautier, 1986) that may be related to upwelling in theseaway (Gautier & Parrish, 1987). The base of theSharon Springs is marked by a condensed sequencethat was formed during the onset of the Claggetttransgression (Figure 3C) and is composed of afossiliferous organic-rich fissile shale that contains fishscales and teeth. Also found in association with thecondensed sequence are several beds of yellowbentonite whose unique geophysical characteristicsenable this biostratigraphic level to be traced overmuch of the Western Interior (Gill et al., 1972, 1973;Crandell, 1950, 1958).

The second lowest bentonite bed in the SharonSprings Member is held, by convention, to be theArdmore bentonite, which is 80 cm thick at Red Bird(no. 1, B. obtusus, Table 2). This bentonite bed is

named after its type area at Ardmore in Fall RiverCounty, South Dakota, where it was at one timecommercially mined (Spivey, 1940). The Ardmorebentonite is not a traceable single discernible unit ofconsistent thickness, but rather the thickest ash layerthat is ‘‘. . . quarried on a commercial scale . . . isdesignated the Ardmore bed’’ (Spivey, 1940, p. 3).The bed designated the Ardmore can therefore varywidely based on this definition. It is more truthfullydefined as an interval of shale that contains severallevels of ash that can be traced over a wide area ofwestern South Dakota, central and eastern Montana,and eastern Wyoming and Colorado, in the lowermostpart of the Pierre Shale. The numbers of ash layersand their thicknesses vary widely across the region,ranging from a single tuffaceous unit composed ofmultiple stacked layers of ash and reworked ash in ElkBasin, Wyoming (Hicks et al., 1995) to 23 layers ormore in southwestern South Dakota (Spivey, 1940).The Ardmore is regionally found in the basal part ofthe Baculites obtusus Zone (Gill & Cobban, 1966a). Ithas been correlated with the middle unit of theElkhorn Mountains Volcanics in western Montana, asequence of 760 m of welded tuff and ash-fall crystaltuff (Gill & Cobban, 1973) that was formed during arelatively short lived volcanic event in the later stagesof the emplacement of the Boulder batholith and aredated at between 80 and 81 Ma (Robinson et al.,1968; recalculated using the modern decay constantsof Beckinsale & Gale; see Gale, 1982).

The Mitten Black Shale Member, which consists of285 m of medium to dark grey, carbonaceous, fissileand bentonitic shale, was deposited during the high-stand of the Claggett transgression (Figure 3A, C).During the ensuing Judith River regression, theshoreline moved basinward and a tongue of coarsersediment called the Red Bird Silty Member, a 185-m-thick sequence of silty shale, was deposited (Gill &Cobban, 1962). At the lowstand of the Judith Riverregression, fluvial incision on the coastal plain to thewest formed a sequence boundary beneath the TeapotSandstone (Reynolds, 1967; Gill & Cobban, 1996b;Van Wagoner, 1990; Figure 2).

This transgressive to regressive cycle repeats itselfwith the onset of the Bearpaw transgression in therange zone of Didymoceras nebrascense (Lillegraven& Ostresh, 1990; Figure 3B, C). The shift of theBearpaw shoreline landwards is reflected at Red Birdin the finer, darker and more organic rich sediments ofthe 220 m thick ‘‘lower unnamed shale member’’ (Gill& Cobban, 1966a; Figure 3A). Within the lowerunnamed shale member is a paraconformity withinwhich three ammonite range zones are missing. VanWagoner et al. (1990; see also Fox, 1993) have

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 7

80

1100 0

–80

Height above base of B. obtusus (m)

500

900

800

700

600

400

300

200

100

40A

ge (

Ma)

7468

7072

7678(8

)

–40

0

1000

AB

CD

EF

G

N3

R2

N2

R1

N1

C31

n

C31

r

C32

n

C32

r

C33

n

C33

r(1

)

Isot

opic

Age

s (T

able

1)

Can

de &

Ken

t (1

995)

Har

lan

d et

al.

(199

0)G

rads

tein

et

al. (

1994

)(7

)

(6)

(9)

(5)

(3)

(4)

(2)

Par

acon

form

ity

***

R9

T9

R8

T8

CLAGGETTBEARPAW

cf. T

rice

rato

psZ

one

Hop

losc

aph

ites

birk

elu

nd

i

Bac

uli

tes

clin

olob

atu

s

Bac

uli

tes

gran

dis

Bac

uli

tes

bacu

lus

Bac

uli

tes

elia

si

Bac

uli

tes

jen

sen

i

Bac

uli

tes

rees

idei

(9)

Exi

telo

cera

s je

nn

eyi

**D

idym

ocer

as n

ebra

scen

se

Bac

uli

tes

scot

ti

Bac

uli

tes

greg

orye

nsi

s

B. p

erpl

exu

s (l

ate)

Bac

uli

tes

gilb

erti

B. p

erpl

exu

s (e

arly

)

Bac

uli

tes

(sm

ooth

)

Bac

uli

tes

aspe

rifo

rmis

*B

acu

lite

s m

clea

rni

Bac

uli

tes

obtu

sus

Fox

Hil

lsF

orm

atio

n

Upp

eru

nn

amed

shal

em

embe

r

Kar

a B

ento

nit

ic M

br.

Low

eru

nn

amed

shal

em

embe

r

Red

Bir

dS

ilty

Mem

ber

Mit

ten

Bla

ckS

hal

e M

embe

r

Sh

aron

Spr

ings

Mbr

.

CAMPANIANMAASTRICHTIAN

80

Fig

ure

3.A

.S

trat

igra

phy

ofth

eP

ierr

eS

hale

.B

.T

hem

egaf

ossi

lbi

ostr

atig

raph

iczo

nati

on(b

oth

afte

rG

ill&

Cob

ban,

1966

a),

plot

ted

from

ada

tum

atth

eA

rdm

ore

bent

onit

ein

the

base

ofth

era

nge

zone

ofB

acul

ites

obtu

sus.

Pos

itio

nof

the

Cam

pani

an/M

aast

rich

tian

boun

dary

isaf

ter

Ken

nedy

etal

.(1

992)

.C

.The

tim

ing

oftr

ansg

ress

ion

and

regr

essi

onin

the

Cam

pani

anan

dM

aast

rich

tian

for

the

nort

hern

Gre

atP

lain

sre

gion

ofth

eW

este

rnIn

teri

orU

S,a

fter

Lill

egra

ven

&O

stre

sh(1

990)

.D

.V

GP

plot

ofth

eR

edB

ird

sect

ion.

VG

Pla

titu

des

are

plot

ted

vers

ushe

ight

inse

ctio

n.T

hebl

ack

dots

toth

eri

ght

ofth

edi

agra

mm

ark

the

stra

tigr

aphi

cpo

siti

onof

the

pale

omag

neti

csi

tes.

Dat

apr

oces

sed

asbe

st-fi

tlin

esar

esh

own

asci

rcle

s,F

ishe

rav

erag

esof

clus

ters

are

show

nas

open

tria

ngle

s.S

quar

esar

eac

cept

able

site

mea

nsco

nstr

aine

dus

ing

dire

ctio

nson

ly.

Sol

idtr

iang

les

are

atle

ast

part

ially

cons

trai

ned

usin

ggr

eat

circ

les.

E.

Mag

neto

stra

tigr

aphy

ofth

eR

edB

ird

sect

ion,

N1

toN

3,bl

ack/

whi

tesh

adin

g,no

rmal

/rev

erse

dpo

lari

tyin

terv

als.

Dia

gona

lha

shed

inte

rval

sre

pres

ent

eith

erin

terv

als

ofer

osio

n,or

inte

rval

sfr

omw

hich

nopa

leom

agne

tic

data

coul

dbe

obta

ined

(e.g

.,lo

wer

mos

tM

itte

nB

lack

Sha

leM

embe

r),t

hesh

aded

inte

rval

inR

2co

ntai

nsa

spur

ious

over

prin

tin

agl

auco

niti

cin

terv

al(s

eete

xt).

F.C

hron

olog

icco

rrel

atio

ndi

agra

mof

the

Red

Bir

dse

ctio

n,pl

otti

ngag

eve

rsus

heig

htin

sect

ion.

The

stra

tigr

aphi

cpo

siti

ons

ofth

eis

otop

icag

esob

tain

edin

this

stud

yar

enu

mbe

red

(1)

thro

ugh

(7)

and

corr

espo

ndto

Tab

le2

whi

chlis

tsth

esa

mpl

ing

loca

lity

and

bios

trat

igra

phic

inte

rval

ofea

chis

otop

icag

e.(7

)is

the

age

for

B.

clin

olob

atus

,af

ter

Obr

adov

ich

(199

3);

(8)

the

age

ofba

seof

C33

n,af

ter

Hic

kset

al.

(199

5).

The

isot

opic

ages

are

show

nin

rela

tion

toth

ees

tim

ated

geom

agne

tic

reve

rsal

ages

for

chro

ns33

thro

ugh

31fr

omth

ree

ofth

ere

cent

lypu

blis

hed

GP

TS

.(9

)P

arac

onfo

rmit

yde

scri

bed

inte

xt.

Det

ails

not

labe

lled

onfig

ure;

*Gam

mon

Fer

rugi

nous

Mem

ber;

**D

idym

ocer

asst

even

soni

;**

*See

text

Sec

tion

8.1

for

expl

anat

ion

ofam

bigu

ity

surr

ound

ing

posi

tive

iden

tific

atio

nof

C32

rat

Red

Bir

d.

8 J. F. Hicks et al.

correlated this paraconformity to a retrogradationalparasequence set found at the top of the TeapotSandstone, but its origin appears to be more complexthan that. Recent work by Izett et al. (1993, 1998) hascorrelated this same level to the Manson impactstructure, Iowa, on the basis of an 40Ar/39Ar age of74.11&0.09 Ma obtained from sanidine-bearing meltrocks in the structure itself (based on Mmhb-1 horn-blende age of 513.9 Ma; 75.17 Ma, Mmhb-1 horn-blende, 520.4 Ma, as used in this study; see fig. 5,p. 368, Izett et al., 1998). This is corroborated by thediscovery of shocked quartz (assumed to be part ofthe distal impact ejecta; Izett et al., 1993) and struc-tures attributed to an impact tsunami (Steiner &Shoemaker, 1996) in marine sediments of the CrowCreek Member of the Pierre Shale to the east of theBlack Hills Uplift, at the same biostratigraphic level asthe paraconformity at Red Bird.

The Kara Bentonitic Member (Figure 3C) isa 10-m-thick volcaniclastic unit consisting of abentonites and grey bentonitic shale. It is overlain bythe 210-m-thick ‘upper unnamed shale member’,which has a varied lithology of alternating layers ofclay-rich shale and more silty and sandy units thatwere deposited during the regression of the BearpawSea as the shoreline, represented by the Fox HillsFormation moved basinward (Figure 2). The FoxHills Formation lies in sharp contact with the under-lying marine shales with a prominent 60-cm-thickgreen bentonite at its base (B. clinolabatus, no. 7,Table 2). At Red Bird the Fox Hills is not exposed inits entirety but it is 90 m thick at Lance Creek (Dorf,1942), 40 km to the southwest. At Lance Creek theFox Hills is separated from the K/T boundary by800 m of fluvial sediments of the Lance Formation(Clemens, 1960).

4. Sampling methods

Between four and five separately oriented handsamples were taken for paleomagnetic analysis from52 sites in 1100 m of continuous section from the topof the Niobrara, through the Pierre Shale to the lowerpart of the Fox Hills Formation. The line of thesection that was sampled parallels Gill & Cobban’s(1966a) northern surveyed transect that runs fromsection 14 to 13, in T. 38 N, R. 62 W, NiobraraCounty, Wyoming. The sampling interval was con-trolled by the surface exposure and averaged 24 moverall, with a minimum interval of 1.6 m in the FoxHills Formation, and a maximum interval of 69.7 m inthe poor exposure of the Mitten Black Shale Member.

Each magnetostratigraphic sample site was placedstratigraphically to the nearest identifiable unit in the

original section description by Gill & Cobban (1966a,pp. A50–A62). The lithologic boundaries betweeneach of the members and formations, and the majorconcretion and bentonite horizons were identified andmarked in the field and compared to the originalsurveyed base map. The area surrounding eachsample site was prospected for fossils and a total of95 specimens were collected from six range zones.These fossil samples were identified by Karl Waage(Peabody Museum, New Haven) and Bill Cobban(USGS, Denver), and all the specimens were found tobe typical of the zone to which they were attributedbased on their stratigraphic position. This indepen-dently confirms the correlation of the paleomagneticsites to the stratigraphy and biostratigraphy of thesurveyed reference section.

5. Paleomagnetic analysis

A minimum of three and a maximum of five samplesfrom each site were prepared into paleomagneticspecimens by dry sanding into cubes nominally 2 cmon a side. A single sample from every other site wasinitially selected to undergo step-wise thermal demag-netization (room temperature, 75)C, 100)C, 125)C,150)C, 200)C, 250)C, 300)C, 350)C, 400)C, 450)C).Schonstedt thermal and single-axis AF demagnetizerswere used. All demagnetization and measurement wascarried out in the magnetically shielded room at theScripps Institution of Oceanography, which has anambient field of 200 nT. The magnetic measurementswere made on a CTF three-axis cryogenic magnetom-eter. The results of this initial run indicated that thesamples contained a secondary goethite componentwith an unblocking temperature of 150)C. In no casewas there more than a small fraction of the naturalremanent magnetization (NRM) remaining after350)C and in many cases the data became scatteredafter treatment to about 250 to 300)C. The scatterprobably results from alteration of magnetic mineral-ogy during thermal demagnetization. In all cases therewas a dramatic change in NRM by the 150)C demag-netization step, causing us to suspect the presence of agoethite component. Thus, in order to avoid thepossible contribution of a spurious goethite com-ponent (which would be difficult to demagnetize usingalternating field treatment alone) and to avoid thescatter caused by heating to high temperatures, weadopted the following measurement protocol for allremaining specimens.

The natural remanent magnetization (NRM) ofeach of the 2–4 specimens remaining from each sitewas first measured. Then each specimen was heatedto 150)C and remeasured. Subsequently, each was

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 9

South

East, Down

North

West, Up

(a)

South

East, Down

North

West, Up(b)

South

East, Down

North

West, Up(c)

South

East, Down

North

West, Up

(d)

270 90

0

180

270 90

0

180

Figure 4. Representative samples showing (a) left vector end-point diagram of a quasi-linear demagnetization path, showingthe calculated best-fit line trending towards the origin. Each demagnetization step is plotted as a pair of the vector’scartesian components: North (N), East (E) and Down. Solid circles, N, E, open squares, N, Down; (a) right,corresponding equal-area plot (open/black circles, upper/lower hemisphere); (b) left (plot same as a, above) demagnetiz-ation path that follows a great circle path to a stable (reversed) direction; (c) ‘hovering’ behavior where sample mean wascalculated by Fisher average of a cluster of points; (d) completely random sample, discarded from study.

subjected to step-wise alternating field (AF) demag-netization in 7 to 9 steps (2.5 mT, 5.0 mT, 7.5 mT,10.0 mT, 15.0 mT, 20.0 mT, 30.0 mT, 40.0 mT,50.0 mT). Examples of the results from this treatmentregimen are shown in Figure 4a, b for a normal andreversed specimen.

In general, the demagnetization data from thermaltreatment alone, and alternating field/thermal treat-ment protocols are quite similar, with the latter beingsignificantly less scattered than the former. Thedemagnetization data were analysed in the followingmanner. Specimens having a rather simple quasi-linear characteristic component (e.g., Figure 4a) that

display a final component trending to the origindefined by at least 4 consecutive demagnetizationsteps, were analysed using principal componentanalysis (PCA) (Kirshvink, 1980). Lines were deemedacceptable if they had a maximum angle of deviation(MAD) of less than 25).

Some specimens behaved in a manner similar tothose shown in Figure 4c. The polarity is not in doubt,but the demagnetization data are clustered and abest-fit line (PCA) cannot be calculated that wouldmeet the minimum criteria listed above. In thesecases, we calculated a mean direction using Fisherstatistics (Fisher, 1953), from a subset of at least 4

10 J. F. Hicks et al.

consecutive demagnetization points (usually manymore) that clustered around a stable direction. Thesedirections met our minimum criteria if the circularstandard deviation (CSD) was less than 35). Sitemeans were calculated using standard Fisher statistics(Fisher, 1953), from sites that contained at least threedirections estimated by either PCA or Fisher analysis.Site means with CSDs less than 35) were retained andmake up our highest quality, or á sites, and all otherswere discarded.

Most of the specimens were amenable to eitherPCA or Fisher analysis. However, the demagnetiz-ation data of specimens from a number of sites did notlie along a line or cluster in a single direction, but laywithin a plane. The vector endpoint diagram of anexample of this behavior is shown in Figure 4b, withan accompanying equal-area plot showing the trace ofthe great circle that best fits these data. The equal-areaplot of the great circles and other data used tocalculate the site mean from such a site is shownin Figure 5. This is a plot of the data from fourspecimens from a site in the uppermost part of R2,just below the reversal with N3 (Figure 3E. The traceof the best fit planes from three of the specimens areshown as great circles (Kirshvink, 1980), along with

the data from one specimen to demonstrate the pathof the demagnetization steps. This site also had onespecimen that allowed calculation of a Fisher meandirection from a cluster of data points (the squaresymbol in Figure 5. The three great circles and singledirected line can be combined using the method ofMcFadden & McElhinny (1988) to give a meandirection (shown as an open triangle) and a cone of95% confidence (á95, shown as a dotted circle). Foursites analysed in this manner were deemed acceptableusing an arbitrary cut-off of 35) for the á95. The meandirections of sites with at least one great circlespecimen are plotted as VGPs denoted by an invertedfilled triangle in Figure 3D. It is worth noting that (a)both normal and reversed significant great circle sitemeans are shown in Figure 3D; and (b) none of thesites calculated by the great circle method lies adjacentto a reversal boundary, and therefore they do not havean overly significant effect on the interpreted mag-netostratigraphy. They do, however, contribute sig-nificantly to the ‘body’ of data that makes up thereversals, and make up the second tier of data, thesites. A few specimens failed to meet any of the abovecategories; the data were too scattered for meaningfulanalysis and were ignored. Such a specimen is shownin Figure 4d.

The site means calculated from directions only (ásites) are plotted as circles in the equal projectionshown in Figure 6. Those at least partially constrainedby great circles (â sites) are plotted as triangles. Thereversed mode is Fisher distributed and has a mean,and á95 could be calculated using Fisher statistics(Fisher, 1953; Figure 6). The normal mode is notFisher distributed and the 95% confidence ellipsewas estimated using the bootstrap method of Tauxeet al. (1991). The mean normal and reversed direc-tions with associated confidence ellipses are plottedin Figure 6b. Also shown is the direction of thepresent dipole field (PDF, shown as a star) and theexpected direction for the late Cretaceous (shown asa diamond) calculated from the reference pole ofDiehl et al. (1983). The normal component of mag-netization is significantly different from both the PDF,and the expected Cretaceous direction, but none-theless the polarity of the individual sites is not indoubt.

270 90

0

180

Figure 5. Representative example of a site mean partiallyconstructed by great circles. Reversed site where thedata from three samples (plot as for Figure 4) do notdefine a single component, but lie along a great circle.The data points for one sample shown with calculatedbest fit great circle path [open (closed) symbols areupper (lower) hemisphere projections]. Open squaresymbol is sample that yielded an acceptable Fishermean. The three best fit great circles and one directiongive an average direction shown as a triangle with theassociated á95 plotted. These were calculated using themethods of McFadden & McElhinny (1988).

6. Magnetostratigraphy

All the significant specimens and site mean directionswere converted to virtual geomagnetic poles (VGPs).The calculated latitude of each site mean reflects thepolarity (positive being normal and negative beingreversed), and are plotted vs height in section in

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 11

Figure 3D. Data based on lines calculated by PCA areplotted as open circles and those based on Fisheranalysis of clusters are plotted as open triangles. Sitemeans based on directional data only are shown asfilled squares and those at least partially constrainedby great circles are shown as inverted filled triangles.The magnetostratigraphic interpretation of the data isshown in Figure 3E.

The magnetostratigraphy of the lowermost part ofthe Red Bird section could not be reliably interpretedbecause of the geochemistry of the rocks. TheGammon Ferruginous, Sharon Springs and the basalpart of the Mitten Black Shale Members are verycarbonaceous and in places contain secondaryferruginous minerals that make them paleomagneti-cally unreliable. This interval is shown by the diagonalshading at the base of Figure 3E.

The following section summarizes the bio-, litho-and magnetostratigraphic data. The numbers with theprefix YPM, are Yale Peabody Museum specimennumbers. The lithologic unit numbers and the USGSsamples sites are derived from the reference section ofGill & Cobban (1966a).

(c)(b)

(a) Total N = 34

Dec,333.8 169.3 η, 4.67 13.90

Mode: 1st: 2nd: Mode: 1st: 2nd:

Inc, 33.5 –42.1 dec, 218.46 163.05

α95, 6.4 25.0 inc, 32.93 47.75

N, 25 9 ζ, 6.64 34.32

κ,21.5. dec, 96.54 256.50

Fisherian ?noyesinc39.233.13

Figure 6. (a) Fisher statistics for the data shown in the equal-area plot below. (b) Equal-area projections in stratigraphiccoordinates. Circles are acceptable site means constrained only by direction (á sites), and triangles are partiallyconstrained by great circles (â sites); open/upper, filled/lower hemisphere projections. (c) Mean normal and reverseddirections with associated confidence ellipses. Present dipole field (shown as a star). Late Cretaceous expected direction(shown as a diamond) after Diehl et al. (1983).

6.1. Polarity interval N1

The base of the measurable polarity sequence lies atthe bottom of a 525-m-long interval of normal polarity(N1; Figure 3E) that ranges from the lowermostpart of the Mitten Black Shale Member (Baculitesasperiformis Zone) through the Red Bird SiltyMember, to the lower middle part of the ‘lowerunnamed shale member’ (Exiteloceras jenneyi Zone;Figure 3A, B). Ammonite samples were taken from

12 J. F. Hicks et al.

this interval that broadly delimit the range zones ofBaculites perplexus (early form) (YPM32719), Baculitesgilberti (e.g., YPM32718), Baculites perplexus (lateform) (YPM32684) and Baculites gregoryensis(YPM32686).

6.2. Polarity interval R1

A 25-m-interval of reversed polarity (R1), spans threesites that lie entirely within the zone of Exitelocerasjenneyi as expressed at Red Bird. A paraconformitywhich cuts out the top of E. jenneyi and the threeammonite zones of D. cheyennense, Baculites compressusand Baculites cuneatus, is known to lie at the top of thisrange zone (Figure 3B, E). This paraconformity,labelled in Figure 3F, is clearly identified in theoriginal stratigraphy of the Red Bird section andhas been traced regionally across Wyoming (Gill &Cobban, 1966b; Van Wagoner et al., 1990). Thestratigraphic position of this paraconformity at RedBird has been precisely fixed with respect to thepaleomagnetic sample sites using both the litho-logic description of the reference section, the identi-fication of characteristic bentonite layers and thebiostratigraphy.

6.3. Polarity interval N2

Above the level of the paraconformity is a 210-m-thicknormal polarity interval (N2) which runs from themiddle of the ‘lower unnamed shale member’(Baculites reesidei), through the Kara BentoniticMember to the lower part of the ‘upper unnamedshale member’ (Baculites baculus). Baculites andscaphites specimens were taken from this normalpolarity interval that clearly delimit the boundaries ofthe range zones of Baculites reesidei (6 specimens, e.g.,YPM32727), Baculites jenseni (YPM32692), Baculiteseliasi (25 specimens, e.g., YPM32679) and Baculitesbaculus (3 specimens, e.g., Jeletzkytes plenus,YPM32700).

6.3. Reversed polarity interval R2

From the lower part of the ‘upper unnamed shalemember’ (Baculites baculus) to 10 m below the baseof the Fox Hills Formation (Baculites clinolobatus) isa well defined 110-m-long reversed polarity interval(R2). Ammonite specimens from the range zone ofBaculites grandis were taken from this reversed polarityinterval (e.g., YPM32664).

In the uppermost Pierre Shale and the lowermostpart of the Fox Hills Formation, is an interval ofhighly questionable normal polarity that ranges fromthe upper part of the B. clinolobatus Zone into the

Hoploscaphites birkelundi Zone (originally defined atRed Bird as the Sphenodiscus (coahuilites) Zone by Gill& Cobban, 1966a, redesignated by Landman &Waage, 1993). This normal polarity interval is shownas a shaded zone within R2 in Figure 3E. No ammo-nite fossils were recovered from this interval. Abovethis is a 40-m-thick reversed interval comprising threeclosely spaced sites that lie in an unfossiliferoussequence of interbedded siltstone and fine sandstonein the Fox Hills Formation.

6.4. Normal polarity interval N3

The uppermost three sites in the section are of normalpolarity and lie more than 60 m above the base ofthe Fox Hills Formation, within the Hoploscaphitesbirkelundi Zone (Landman & Waage, 1993).

7. Isotopic age dating

Due to the 39Ar recoil effects (Hess & Lippolt, 1987),biotites extracted from bentonites often yield uninter-pretable heating spectra or total fusion ages that aretoo old when compared to sanidine phenocrysts fromthe same unit. The ages used in this study thereforewere obtained by the 40Ar/39Ar laser fusion dating ofsanidine crystals (York et al., 1981). The bentoniteswere disaggregated and sieved to retain the materialcoarser than 0.1 mm (+140 mesh), processed in aFrantz magnetic separator, and the sanidine concen-trates recovered from the non-magnetic fraction byusing heavy liquids. These were then washed with adilute solution of HF (approximately 12%) andrinsed. The largest crystals (+100 mesh or coarser)were then examined with a petrographic microscopeusing the central focal masking technique (Wilcox,1983) to check sample purity and then under crossednicols to look for the presence of detrital microcline.All occluded and altered grains were removed by handpicking, and the individual grains for study wereselected. The sanidine crystals were then encapsulatedin aluminium packets and stacked in a quartz tubewith every three unknowns surrounded by a monitor(Taylor Creek rhyolite, Dalrymple & Duffield, 1988),normalized for an age of 520.4 Ma for the McClureMountain standard (Samson & Alexander, 1987).The unknowns and standards were then irradiated for30 hours in the central thimble of the USGS TRIGAreactor (Dalrymple et al., 1981).

Samples were analysed using the GLM continuouslaser system at the USGS Isotope facility at MenloPark (Dalrymple, 1989). After irradiation, thesamples were placed into a high vacuum extractionsystem and baked out overnight at 260)C. Several

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 13

crystals of sanidine, both monitors and unknowns,along with vacuum-degassed zero-aged basalt as aflux, were then fused with a six watt argon-ion laser.Laser fusion was carried out on individual grainswherever possible, but sometimes multiple sanidinegrains had to be used because of the relatively finegrain size of the sanidine phenocrysts recovered fromsome of these bentonites. The 40Ar/39Ar laser total-fusion data for these ages are listed in Table 1. Theexpelled gases were cleaned of their reactive compo-nents by two SAES getters and the argon isotopiccomposition (Table 1) determined using a MAP 216high sensitivity rare gas mass spectrometer operated inthe static mode. A J curve (which depicts the variationof neutron flux with vertical position of the unknownsin the quartz tube) was then determined for the entirepackage (Table 1). Nominally five individual determi-nations are made for each monitor level. Errors for theneutron flux at each level are quadratically combinedwith the uncertainties for the individual unknowns.The final results are reported with the uncertaintiesexpressed at the 95% confidence level for the error ofthe mean employing Student’s t factor (Taylor, 1982)for small sample populations (Table 1).

In Table 2 are listed the locations and provenanceof the ash levels whose 40Ar/39Ar isotopic ages wereused to estimate the ages of the reversal boundariesin the Red Bird section. The relative stratigraphicposition of each ash bed is shown in Figure 3F. Anisotopic age for Baculites compressus is plotted in thediagram, but the B. compressus zone is missing at thelevel of the paraconformity at the top of E. jenneyi, andthe age was not used in the calibration of the reversals.

The age for the range zone of Baculites asperiformis(Figure 3F, labelled 8) is based on an extrapolatedage for the C33n/C33r reversal boundary obtainedfrom isotopic ages from bentonites sampled from theClaggett Shale and Judith River Formation in ElkBasin in the northern Bighorn Basin, Wyoming(Hicks et al., 1995). This polarity boundary has beenreported as lying in the range zone of Baculitesasperiformis in southeastern Alberta (Lerbekmo,1989). The extrapolated age of the C33r/C33nreversal is 79.35 Ma, which agrees well with theinterpolated age of 79.5 Ma for Baculites asperiformisobtained from the isotopic time scale of Obradovich(1993). Unfortunately the position of the C33r/C33nreversal boundary could not be directly measuredpaleomagnetically at Red Bird due to the verycarbonaceous shale lithologies at the base of thesection in the Mitten Black Shale Member and theSharon Springs.

8. Calibrating the Red Bird geomagneticpolarity sequence

The magnetostratigraphy of the Red Bird section canbe uniquely correlated to that part of the GPTS thatranges from C33n to C31n, with isotopic agesobtained from the section itself, and by biostrati-graphic correlation with dated bentonite horizonssampled throughout the region (Table 2). Theseisotopic ages are plotted on a correlation diagram inFigure 3F. Also plotted for comparative purposes arethe age estimates for the polarity reversals C33r toC31n from the time scales of Cande & Kent (1995;hereafter referred to as CK95), Gradstein et al. (1994;MTS94) and Harland et al. (1989; PTS89).

Where the isotopic age was obtained from a sampletaken from the section at Red Bird, as for 1, 6 and 7(Figure 3F), then the stratigraphic position of the ageis known precisely. However, the remaining ages, 2, 3,5 and 8, have been correlated to the section bio-stratigraphically, and their exact positions within theammonite zones to which they are attributed is notknown with any precision. For the purposes of thisstudy, such ages are placed stratigraphically in themiddle of the zone as it is defined in the published RedBird section (Gill & Cobban, 1966a). The relativeprecision of this biostratigraphic correlation is simplya function of the thickness of the biozone in thereference section, and is shown as a vertical error bar.Calculation of the age of a reversal boundary can bemade with a simple interpolation formula using strati-graphic thickness measurements, with the assumptionthat no major unconformities or rapid changes insedimentation rate occur between the two datedsamples (see Kowallis et al., 1995). Calculating theuncertainty associated with such an age estimate is alittle more difficult, as at Red Bird the exact strati-graphic level of the bounding isotopic ages is notalways known.

8.1. Calibration of the N1/R1 reversal boundary(top of C33n)

The age of normal polarity interval N1 (Baculitesasperiformis to E. jenneyi) is controlled by bracketingisotopic dates that range from 80.04 to 74.31 Ma(Figure 3F, nos 2 and 3). Compared to the publishedGPTS, shown in Table 3, this age range correlates N1with C33n. The reversal at the base of C33n wasnot found at Red Bird, so we use the estimateof 79.34&0.6 Ma from Hicks et al. (1995; no. 8 inFigure 3F) which is correlated to the range zone ofBaculites asperiformis (Lerbekmo, 1989). At Red Bird,the reversal at the top of C33n is projected to lie near

14 J. F. Hicks et al.

T

able

1.40A

r/39A

rla

ser

tota

l-fu

sion

data

.

Am

mon

ite

zone

,S

ampl

eIr

rad.

no.

Min

eral

JL

ab.

no.

40A

r/39A

r37A

r/39A

r36A

r/39A

r%

40A

r*40A

r*/3

9A

r kA

ge(M

a)1

Sig

ma

Bac

ulite

scl

inol

obat

usJD

06-1

3S

anid

ine

0.00

6951

91Z

1241

5.68

219

0.00

2690

0.00

0007

99.8

15.

6713

669

.75

0.24

91-O

-14

91Z

1242

5.67

173

0.00

6820

0.00

0068

99.5

05.

6431

469

.41

0.25

91Z

1243

5.70

191

0.00

3320

0.00

0041

99.6

35.

6811

169

.87

0.25

91Z

1244

5.66

542

0.00

4020

0.00

0030

99.6

95.

6479

069

.47

0.24

91Z

1245

5.66

440

0.00

3000

0.00

0057

99.5

45.

6385

869

.36

0.22

69.5

70.

37B

acul

ites

gran

dis

GL

N12

-S

anid

ine

0.00

7064

94Z

0148

5.74

477

0.15

1072

0.00

0352

98.2

45.

6443

570

.53

0.54

92-O

-32

94Z

0149

6.18

429

0.65

0385

0.00

2690

87.8

35.

4337

767

.95

0.77

94Z

0150

6.12

762

0.58

0398

0.00

1950

91.2

05.

5908

069

.88

0.57

94Z

0151

5.76

147

0.14

0844

0.00

0487

97.5

45.

6203

370

.24

0.42

94Z

0152

5.80

351

0.21

0036

0.00

0669

96.7

25.

6142

370

.16

0.55

70.1

50.

65B

acul

ites

rees

idei

JD01

0-A

San

idin

e0.

0080

2293

Z03

175.

1475

30.

1041

520.

0002

0398

.82

5.08

696

72.1

50.

3392

-O-3

093

Z03

185.

1302

40.

0781

000.

0001

9598

.82

5.06

993

71.9

20.

3572

.02

Bac

ulite

sco

mpr

essu

sJD

06-1

7S

anid

ine

0.00

6915

91Z

1251

6.06

965

0.00

8050

0.00

0093

99.4

16.

0337

673

.74

0.24

90-O

-15

91Z

1252

6.06

012

0.00

5843

0.00

0167

99.0

46.

0020

273

.36

0.32

91Z

1253

6.03

953

0.00

5330

0.00

0099

99.3

76.

0015

573

.36

0.29

91Z

1254

6.06

417

0.00

8320

0.00

0116

99.3

06.

0216

173

.60

0.24

91Z

1255

6.05

284

0.00

5260

0.00

0098

99.3

86.

0153

173

.52

0.26

73.5

20.

39E

xite

loce

ras

jenn

eyi

GL

N12

-S

anid

ine

0.00

7078

94Z

0137

5.98

865

0.00

6890

0.00

0000

99.8

65.

9801

174

.79

0.22

93-O

-16

94Z

0138

5.99

030

0.00

7620

0.00

0212

98.8

15.

9192

174

.04

0.31

94Z

0139

5.95

274

0.00

8370

0.00

0001

99.8

65.

9441

474

.35

0.29

94Z

0140

5.97

397

0.00

8200

0.00

0113

99.3

05.

9321

774

.20

0.28

94Z

0141

6.00

208

0.00

7470

0.00

0192

98.9

15.

9369

174

.26

0.31

94Z

0142

5.96

669

0.00

8300

0.00

0091

99.4

15.

9313

074

.19

0.26

74.3

10.

43B

acul

ites

scot

tiJD

018-

San

idin

e0.

0080

4993

Z03

315.

3908

40.

0166

000.

0001

3899

.10

5.34

242

75.9

50.

4192

-O-1

393

Z03

325.

3622

90.

0215

780.

0000

6799

.50

5.33

533

75.8

50.

2493

Z03

335.

3982

00.

0283

000.

0000

4999

.61

5.37

705

76.4

30.

2693

Z03

345.

3644

70.

0043

100.

0000

0299

.83

5.35

516

76.1

30.

2393

Z03

355.

3742

90.

0423

000.

0000

7199

.50

5.34

783

76.0

30.

2976

.07

0.51

Bac

ulite

sob

tusu

sJD

018-

San

idin

e0.

0063

3794

Z05

997.

2711

60.

0167

000.

0001

5499

.27

7.21

790

80.6

80.

29A

rdm

ore

Ben

toni

te94

Z06

007.

2072

30.

0114

000.

0000

9199

.51

7.17

232

80.1

80.

28R

B92

-15

94Z

0601

7.16

481

0.01

7100

0.00

0066

99.6

27.

1377

079

.81

0.29

94Z

0602

7.15

476

0.01

4800

0.00

0045

99.7

07.

1335

479

.76

0.24

94Z

0603

7.18

055

0.01

8800

0.00

0010

99.8

57.

1699

580

.16

0.24

94Z

0604

7.17

979

0.01

2400

0.00

0142

99.3

07.

1296

879

.72

0.25

80.0

40.

40

Not

e:40A

r*,

Rad

ioge

nic

argo

n.A

ges

and

unce

rtai

ntie

ssh

own

inbo

ldnu

mbe

rsre

pres

ent

the

unw

eigh

ted

mea

nan

der

ror

ofth

em

ean

atth

e95

%co

nfide

nce

leve

l,ex

cept

for

the

resu

ltfo

rB

acul

ites

rees

idei

whi

chis

sim

ply

the

mea

n.C

aan

dK

corr

ecti

ons

and

deca

yco

nsta

nts:

(36A

r/37A

r)C

a=

2.69

&0.

24#

10"

4;

(39A

r/37A

r)C

a=

6.79

&0.

051#

10"

4;

(40A

r/39A

r)k=

9.1&

5.4#

10"

3;

ë c+

ë c1=

0.58

1#

10"

10

yr"

1;

ë b=

4.96

2#

10"

10

yr"

1;

40K

/K=

1.16

7#

10"

10

atom

/ato

m.

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 15

Table 2. Ages and locations of bentonites used in this study to calibrate the geomagnetic reversals observed in the Red Birdsection.

Range zone Age (Ma) Sample location

Baculites clinolobatus (7) 69.57&0.37 30 cm bentonite bed from unit 112, Red Bird section (Gill & Cobban, 1966a),sec. 14, T. 38 N, R. 62 W, Niobrara County, Wyoming. Collector Jason Hicks.

Baculites grandis (6) 70.15&0.65 1.40 m bentonite bed from unit 97, Red Bird section (Gill & Cobban, 1966a),sec. 14, T. 38 N, R. 62 W, Niobrara County, Wyoming. Collector Jason Hicks.

Baculites reesidei (5) 72.50&0.44 No sample location or precision available. Precision given by C. C. Swisher,pers. comm. (1992). Age referenced in Lerbekmo et al. (1979). Revised age byBaadsgaard et al. (1993).

Baculites compressus (4) 73.52&0.39 22-cm-thick bentonite 1.5 m above base of zone in the Bearpaw Shale, NE, NE,sec. 14, T. 1 N, R. 33 E, Big Horn County, Montana. Collectors, E. A.Merewether and W. A. Cobban.

Exitelocerus jenneyi (3) 74.31&0.43 Bentonite bed in upper part of the Mancos Shale, NE corner, SE, SE, sec. 17,T. 48 N, R. 7 W, Montrose County, Colorado.

Baculites scotti (2) 76.07&0.51 15 cm bentonite in the Lewis Shale (possibly equivalent to the Huerfanitobentonite marker bed in the subsurface) CW, sec. 11, T. 23 N, R. 1 W, RioArriba County, New Mexico. Collectors, C. M. Molenaar, D. Nummendal andW. A. Cobban.

Baculites obtusus (1) 80.04&0.45 Ardmore bentonite bed in the base of the Pierre Shale at the Red Bird section(Gill & Cobban, 1966a), sec. 14, T. 38 N, R. 62 W, Niobrara County,Wyoming. Collector Jason Hicks.

Table 3. Summary table of the geomagnetic reversals found in the Red Bird section, comparingthe age estimates from Red Bird to three of the more recently published geomagnetic polarity timescales: CK95, Cande & Kent (1995); MTS94, Gradstein et al. (1994); PTS89, Harland et al.(1989). *The top of C32r must be older than B. reesidei, so this is a conservative constraint. **Thebase of C32r is no older, and probably younger than this estimate. See Section 8.1 in text forsummary of ambiguity surrounding the correlation of R1 with C32r. ***Age for the base of C33nfrom Hicks et al. (1995).

(In text) Geomagnetic reversal This study RB97 CK95 MTS94 PTS89

R2/N3 C31r/C31n 69.01 68.737 68.8 68.13N2/R2 C32n/C31r 70.44 71.071 71.0 70.14R1/N2 C32r/C32n (>72.5)* 73.004 73.7 72.35N1/R1 C33n/C32r? <74.60** 73.619 74.5 73.12Missing C33r/C33n 79.34*** 79.075 79.8 79.09

the base of E. jenneyi (Figure 7), where it is bracketedbetween isotopic ages of 76.07&0.51 in Baculitesscotti and 74.31&0.43 Ma in E. jenneyi (Table 2).Biostratigraphically correlated to the middle of therange zone, these isotopic ages lie only 87 m (and1.86 Myr) apart. By simple linear interpolation thisgives a calculated sedimentation rate of about 47 m ofcompacted sediment per million years, and an age forthe top of C33n of 74.60 Ma (Figure 7).

However, the relative precision of this interpolatedage is very poor due to the large number of underlyingassumptions upon which it is based. Figure 7 shows indetail the stratigraphic position of the paleomagneticsample sites, the ammonite range zones and the

calibrating isotopic ages that bound the reversal at thetop of C33n.

Our first assumption is that the isotopic ages corre-late to the middle of the corresponding ammoniterange zone. But in the reference section at Red Birdthe precision to which the stratigraphic position of theammonite range zone is known is limited to theoccurrence of fossiliferous concretionary horizons inthe Red Bird section. If we consider, for example, thecase of Baculites scotti, the density of fossiliferous layersis remarkably good (compared to many other sectionsin this region), but the actual stratigraphic range inwhich Baculites scotti is found is some 43 m thick, andbounded by unfossiferous zones 9 m thick above, and

16 J. F. Hicks et al.

Figure 7. Detailed age correlation diagram of part of the Red Bird section from Baculites scotti to B. reesidei, showing theammonite zonation, isotopic ages and magnetostratigraphy. A. The actual sampled range of ammonite biozone (thickbar) and its inferred range. Baculites scotti has been sampled over 65% of its inferred biostratigraphic range, E. jenneyi over80%; B. The stratigraphic difference between the midpoint of the actual sampled range and the inferred biostratigraphicrange; C. The biostratigraphic range of the corresponding biochronologic age; D. The polarity transition zone (NACSN,1983; the stratigraphic distance between the reversed and normal polarity paleomagnetic sites that bracket the reversal),representing 30 m of section. The polarity-reversal horizon is projected to lie midway between these two sites; E. Theanalytical precision of the isotopic age; F. The maximum age and, G the minimum age of the reversal boundary, giventhe uncertainties of stratigraphic placement and the precision of the isotopic ages; H. The interpolated age (this study)of the C33n/C32r polarity reversal; I. 2.3-Myr interval that represents the relative precision of the interpolated age of thegeomagnetic reversal boundary given D, the stratigraphic uncertainty in the placement of the reversal boundary.Interpolated age is 74.62 Ma, with a minimum and maximum age range (I) of 73.61 to 75.87 Ma.

6 m thick below. Some 26% of the range zone isdepauperate. The limits of its inferred biostratigraphicrange as defined by Gill & Cobban (1966a) thereforelie within the adjacent barren zones. The actualsampled range of Baculites scotti at Red Bird is shownby the thick black bar adjacent to A in Figure 7, and

the inferred stratigraphic range of the biozone isshown by the thinner lines. The difference betweenthe mid-point of the actual sampled range and theinferred range of Baculites scotti is labelled B, and islittle more than a meter, so small that it may bediscounted. The isotopic age of Baculites scotti could,

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 17

in fact, be correlated anywhere in the 58-m-thickinferred range zone, an interval that defines thelimits of stratigraphic uncertainty, shown by thevertical error bars labelled C in Figure 7. A furthercomplication is introduced in the projection of theE. jenneyi isotopic age to the middle of the range zoneas an indeterminate amount of that zone may be cutout by the overlying paraconformity (Figure 7).

Our second assumption is that the stratigraphicposition of the polarity reversal itself is projected to liemidway between the bounding normal and reversedpaleomagnetic sample sites (labelled R and N,reversed/normal, in Figure 7). Based on these criteriait could fall anywhere in the 30 m interval (labelled D)that separates the adjacent paleomagnetic sites, aninterval that encompasses three ammonite rangezones, D. nebrascense, D. stevensoni, and E. jenneyi(Figure 7). The nature of the exposure in a drystream bed of low relief makes closer sampling animpossibility in this interval.

In order to estimate the age of the reversal bound-aries we make our third, and in some ways, overridingassumption; namely, that there has been continuousand uniform sedimentation throughout the timeperiod that is bracketed or bounded by the isotopicages, an assumption which cannot be directlyaddressed except at levels where unconformities areknown to exist (e.g., Figure 3, no. 9). If we assumecontinuous sedimentation and combine the precisionof the isotopic ages (labelled E in Figure 7) with thestratigraphic uncertainty associated with both the agesthemselves (C), and the inferred position of the geo-magnetic reversal (D), then we can define a maximum(labelled F, 75.87 Ma), and minimum (labelled G,73.61 Ma) age range within which the age for thereversal (labelled H, interpolated at 74.62 Ma) maylie.

Although the reversal appears at first glance to bewell dated by bracketing isotopic ages, it can be seenthat the two secondary assumptions (a, that the actualchron position lies midway between the two nearestpaleomagnetic sites, and b, that the biostratigraphi-cally correlated isotopic ages lie in the middle of thebiozones), give rise to considerable uncertainties. Thechron boundary may be anywhere within plus orminus 15 m of its stated position, the ‘biochronologic’or biostratigraphically correlated isotopic age may beas much as &30 m, and the isotopic ages themselveshave an analytical precision of some &0.5 Myr.Therefore we conclude: if sedimentation rate was evenand uninterrupted, if the isotopic ages approximatethe age of the middle of the range zone, and if thereversal lies near the base of Exiteloceras jenneyi whereit is projected, then the age of the reversal boundary at

the top of C33n is interpolated at 74.62 Ma, with anestimated precision that ranges from a minimum of73.6 Ma to a maximum of 75.9 Ma (a total rangeof some 2.3 Myr).

The interpretation that this reversal is the C33n/C32r boundary is complicated by age estimates fromthe San Juan Basin (Fassett & Obradovich, 1986;Fassett & Steiner, 1997; Fassett et al., 1997) whichsuggest that the top of C33n is somewhat youngerthan our interpolation indicates. Additionally, thereare indications that some short reversals that areunrecognized in the marine anomaly sequence of theGPTS may occur in the upper part of C33n. Lindsayet al. (1982) reported a short reversal in the samestratigraphic sequence in which Fassett et al. (1997)found two reversed intervals at the top of C33n.Either of these polarity reversals may potentially becorrelated to the reversal R1 at Red Bird. The over-lying paraconformity close to the top of C33n makesit impossible to resolve this issue in the Red Birdsection, and the identification of the reversal R1 mustremain ambiguous. In subsequent discussion in thetext this reversal is referred to as C32r, but ourconservative interpretation is that the top of C33nis probably younger than our interpolated age of74.62 Ma for the base of R1 (Figure 7, H), but iscertainly not older than our maximum age estimate of75.9 Ma (Figure 7, F).

Problems such as these can only be overcome by thedetailed study of a surface section with no knownunconformities, where the bentonite ages have aknown, precise stratigraphic position close to thechron boundaries, and exposure and lithologies thatallow paleomagnetic sites to be sampled at very shortintervals across the reversal boundary. In this regionthese conditions have so far only been met in theJudith River Formation at the base of C33n (Hickset al., 1995).

8.2. The R1/N2 reversal boundary

The isotopic age of 74.31&0.43 in the range zone ofExiteloceras jenneyi, correlates the R1 reversed polarityinterval, shown in Figure 3, to at or near the top ofC33n (Table 3). The top of the reversed interval liesin contact with the paraconformity shown in Figure 7,within which three ammonite range zones are missing.One of the missing zones is Baculites compressus, whichhas been dated at 73.52&0.39 Ma (Table 1). If R1is indeed C32r, which is some 0.6 to 0.9 Myr induration as suggested by the GPTS, then our ageof 74.31 Ma for the lower part of the reversal inExiteloceras jenneyi indicates that the top of the reversalprobably lies near the Baculites compressus range zone.

18 J. F. Hicks et al.

Above the paraconformity, the range zone ofBaculites reesidei, dated at 72.5 Ma (Lerbekmo, 1989;Baadsgaard et al., 1993) lies at the base of the normalpolarity interval N2, which can in turn be correlatedwith the lower part of C32n (Table 3).

8.3. Calibration of the N2/R2 reversal boundary(C32n/C31r)

The upper reversal boundary of N2 lies within 27 m ofa bentonite at the base of Baculites grandis that hasbeen dated at 70.15&0.65 Ma (Table 2). This corre-lates reasonably well with the GPTS estimates for thetop of 32.1 n (Figure 3F; Table 3). The short reversedinterval C32.1r appears to be missing from some-where within the middle to upper part of Baculitesbaculus, which coincides with a sampling gap, due topoor exposure, of nearly 60 m.

The base of reversed polarity interval R2 (Figure 3,E) lies in the lower part of Baculites grandis, approxi-mately at the level of unit 96 in the reference section ofGill & Cobban (1966a), and can be correlated toC31r on the basis of the isotopic ages that bracket theN2/R2 reversal (Tables 2, 3). The C32n/C31r geo-magnetic reversal is shown in detail in Figure 8,and lies between an isotopic age of 72.5&0.40 Ma(labelled A), projected to the mid-point of theBaculites reesidei Zone, and the age from Red Birdin the Baculites grandis Zone of 70.15&0.65 Ma(labelled B). The position of the upper bentonite isknown precisely, and the position of the Baculitesreesidei isotopic age is projected to the mid-point of thebiozone (Figure 8A). However, there is an unquanti-fiable error introduced into this projection, as thelower part of C32n, and an unknown thickness ofBaculites reesidei, is missing at the level of the para-conformity (C). However, because the C32n/C31rreversal lies so close to the upper isotopic age (B), anydiscrepancy in the age and position of A will have onlyresulted in a small change in the interpolated age ofthe reversal. For example, a change in A of 1.0 Myrwould only alter the interpolated age of the overlyingreversal by some 0.1 Myr.

Following the procedures outlined in the previoussection, interpolation between A and B in Figure 8gives an age for the C32n/C31r reversal of 70.44 Ma,with a minimum/maximum age range (D), based onthe relative precision of the isotopic ages and theirstratigraphic uncertainty, of 71.09 to 69.79 Ma, arange of 1.29 Myr. The range of D should beincreased by at least 0.1 Myr to account for thepossible discrepancies in the stratigraphic correlationof A to the Red Bird section. The relative precision ofthis interpolated age is far better than the age estimate

for the reversal at the top of C33n, primarily becausethe position of the reversal is known to within a fewmeters due to extensive resampling in good exposure(see Figure 3D for the location of the paleomagneticsites). Most of the uncertainty in this age estimateis derived from the relatively poor precision of theuppermost isotopic age (B). There is little differencein the age estimate of the C32n/C31r reversal bound-ary whether we interpolate between A and B, orbetween A and E, the isotopic age at the base of theFox Hills in B. clinolobatus. The difference in age ofaround 0.2 Myr is labelled F in Figure 8.

8.4. Calibration of the R2/N3 reversal boundary(C31r/C31n)

A short normal polarity interval found in R2 (shadedgrey in Figure 3E) is composed of two sites thatstraddle the contact between the Pierre Shale and theFox Hills Formation. After the first samples from thisinterval had been analysed it was re-sampled, but ofthe five paleomagnetic sites collected in total, only twowere found to contain statistically significant direc-tions. This normal appears to be lithologically con-trolled as it corresponds almost exactly with theoccurrence of glauconite over an interval of 50 m inthe base of the Fox Hills, from unit 110 to unit 115 inthe reference section of Gill & Cobban (1966a). Ofthe three statistically rejected sites in this interval, themost visibly glauconite samples were truly random inorientation, so it is regarded as an anomalous over-print controlled by the mineralogy of the iron-richglauconite clays.

Above the level of the glauconitic beds lies a 40-m-thick reversed polarity interval, composed of threeclosely spaced sites sampled from an interbeddedsiltstone and fine sandstone sequence in the Fox Hillsthat displays a strong and consistent magnetic direc-tion. Above this level the topmost three sites in thesection are of normal polarity and make up theinterval N3 (Figure 3E). The top of the Fox HillsFormation is not found in this section, so assumingthat the total thickness of the Fox Hills is approxi-mately the same as that exposed at Lance Creek(90 m, Dorf, 1942), then the C31r/C31n reversalboundary lies in the upper part of the formation,about 60 m above the base of the Pierre/Fox Hillscontact, in the zone of Hoploscaphites birkelundi(Landman & Waage, 1993).

The C31r/C31n reversal boundary has been datedin the GPTS at between 68.737 and 68.13 Ma (Table3). These estimates compare favorably to our isotopicage of 69.57&0.37 Ma from the top of the PierreShale in Baculites clinolobatus, which lies some 72 m

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 19

Figure 8. Detailed age correlation diagram of part of the Red Bird section from Exiteloceras jenneyi to Hoploscaphitesbirkelundi, showing the ammonite zonation, isotopic ages and magnetostratigraphy. A. The biostratigraphic range of thecorresponding biochronologic age, label adjacent to isotopic age for Baculites reesidei; note that the base of the range forthe age is in contact with the paraconformity (shown by hashed interval); B. The analytical precision of the isotopic agefor Baculites grandis; C. Paraconformity; D. The interpolated age of the C32n/C31r polarity reversal bracketed by themaximum and minimum age estimates given the uncertainties; E. The isotopic age from the base of the Fox Hills/top ofBaculites clinolobatus; F. The difference in the interpolated age of the C32n/C31r reversal resulting from projecting bothfrom A to B and A to E; G. Range of age estimate for the interpolated age for C32n/C31r reversal boundary. Interpolatedage of 69.01 Ma assigned by this study is in the middle of range G; H. Difference in interpolated age of C31r/C31n inreversal resulting from projecting both from A to B and A to C.

below the top of the reversed interval. In Figure 8,the age of the C31r/31n boundary is estimated byextrapolation from using both the lowermost age inBaculites reesidei (A), and Baculites grandis (B). Thisgives us an averaged estimate for the age of thereversal that is not unduly affected by changes in the

sedimentation rate which probably occurred as thehigher energy shoreline environment approached RedBird and the Western Interior seaway regressed fromthe region. Note that the line of projection from A toE passes through the error bars in the age of B. Thedifference between the two estimates (G) is only some

20 J. F. Hicks et al.

0.23 Myr, so we have taken the mid-point, 69.01 Ma,as the age of the reversal, and suggest that the relativeprecision is about &9.60 Myr, or the precision of theunderlying age (E), &0.37 Myr, combined with theuncertainty of F (Figure 8). The age range labelled Hin Figure 8, represents the maximum difference in theage estimate obtained by projecting from A through E,up to the C31r/C31n reversal boundary. The agerange of H is only about 0.60 Myr, so our estimatedrelative precision of &0.60 Myr for the C31r/C32nreversal is suitably conservative.

9. Sea-level change and measuredsedimentation rates

In the Western Interior US, the widespread exposureof Cretaceous sediments, in combination with theabundant ammonite fossils, accurately record thechanges in sea level, as the westward extent ofthe ammonite biozones can be correlated to theadjacent nearshore and coastal plain sediments, togive a very detailed picture of the migration of theshoreline during this time (Lillegraven & Ostresh,1990; Cobban et al., 1994). The correlation diagramshown in Figure 3F clearly indicates that at Red Birdthe changes in sedimentation rate moved in concertwith the rise and fall of relative sea level shown inFigure 3C, labelled T8 to R9 (after Lillegraven &Ostresh, 1990).

This relationship is best seen when the relativechange in sea level (Figure 3C) is compared to thecorrelation diagram of the Red Bird section (Figure3F). The sedimentation rate (data summarized inTable 4), calculated from the Ardmore Bentonite tothe projected base of C33n, is 96 m of compactedsediment per million years (m/Myr). This periodcorresponds biostratigraphically to the Claggett trans-gression (Figure 3C, labelled T8 after Lillegraven &Ostresh, 1990) which ranges from Baculites obtusus tothe middle of Baculites asperiformis. This is undoubt-edly a high overall average sedimentation rate, be-cause at the base of this interval is a condensedsequence in the lowermost part of the Sharon SpringsMember (Figure 3A; unit no. 13 in the referencesection of Gill & Cobban, 1966a). This was formedduring a period of very low sediment supply betweenthe onset and the peak of the Claggett transgression.Such condensed sequences are formed as the shore-line transgresses and the loci of terrigenous depositionmoves landward, effectively starving the shelf anddeeper parts of the basin of terrigenous material. Theterrigenous sediments are deposited in a succession oflaterally extensive nearshore deposits (Loutit et al.,1988), which in this case corresponds chronologically

to the Claggett shale and the westernmost outcrops ofthe shoreface sandstones of the Parkman Sandstone(Figure 2). The condensed interval that formsbasinward is characterized by abundant planktonicfossil assemblages, authigenic minerals, organicmatter, and bentonites (Loutit et al., 1988). In theSharon Springs Member, the condensed interval isrepresented by a fissile, dark grey, carbonaceous shale,with a complex geochemistry of abundant limonite,jarosite, and gypsum, regionally traceable thinbentonite layers, and a concentrated layer of fishscales and bones (Gill & Cobban, 1966a, 1972).

As the rate of relative sea-level rise decreased andthen fell with the onset of the Judith River regression(Figure 3C, labelled R8), the locus of depositionprograded basinward, and the terrigenous contentincreased up-section. In the Red Bird section, thisrelationship is seen in the gradual lithologic shiftthroughout R8, from carbonaceous shales in the baseof the Mitten Black Shale Member (Figure 3A), to thesilty units in the Red Bird Silty Member that wereformed as the higher energy shoreline environment,represented by the Parkman Sandstone (Fox, 1993),drew closer to Red Bird (Figure 2). At the same time,the average sedimentation rate for the interval fromBaculites asperiformis to Baculites scotti rose by nearly50% to 139 m/Myr (Table 4).

At the top of the Red Bird Silty Member, there isa marked lithologic shift back to clay-rich,carbonaceous shale in the base of the ‘lower unnamedshale member’, in the Didymoceras nebrascense Zone(Figure A3). This lithologic change marks the onset ofthe Bearpaw transgression (Figure 3C, labelled T9)and is reflected by a sharp decrease in sedimentationrate (about one third of its previous rate) of 47 m/Myrbetween Baculites scotti and Exiteloceras jenneyi (Table4). The peak of transgression coincides with thesurface of non-deposition or erosion at Red Bird(paraconformity, Figure 3F). Across the para-conformity between Exiteloceras jenneyi and Baculitesreesidei, the measured sedimentation rate decreasesfurther to an average of 23 m/Myr.

The peak of the Bearpaw transgression occurred inthe range zones of Baculites reesidei when a largemarine embayment occupied western Montana. Asthe final regression of the Cretaceous seaway began inthe latest Campanian, the Fox Hills shoreline began toprograde rapidly across the basin (Figure 2), and thesedimentation rate increased fourfold to an average of93 m/Myr in the first part of R9 (Table 4; Figure 3C)and then peaked at 162 m/Myr for the remainder ofthe Pierre Shale deposition when the shorelineregressed across eastern Wyoming in the zone ofBaculites clinolobatus (Cobban et al., 1994).

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 21

The detailed chronostratigraphy of the Red Birdsection for the Late Campanian and Maastrichtianshows not only an excellent correlation between sea-level changes and sedimentation rate, as predicted anddemonstrated by numerous workers in this region(for example, Van Wagoner et al., 1990), but it alsoprovides one of the first detailed quantifications of theprocess.

10. Global correlation of the Red Birdmagnetostratigraphic section

Table 4. Summary table showing the compacted sediment accumulation rate in the Red Birdsection, calculated for the intervals between the age calibration points listed in Table 2. The datashow the relationship between sedimentation rate and the fluctuating transgressive and regressivecycles of the Cretaceous seaway during the Late Campanian and Maastrichtian (Lillegraven &Ostresh, 1990). The percent change column simply shows the change in sedimentation rate as apercentage increase or decrease relative to the preceding time interval.

Biostratigraphic intervalTime period

(Myr)Sedimentation rate

(m/Myr) T/R % Change

B. obtusus to B. asperiformis 0.69 96 T8B. asperiformis to B. scotti 3.28 139 R8 +45%B. scotti to E. jenneyi 1.86 47 T9 "66%E. jenneyi to B. reesidei 1.71 23 T9 "51%B. reesidei to B. grandis 2.35 93 R9 +304%B. grandis to B. clinolobatus 0.58 162 R9 +74%

10.1. Correlation of Red Bird to Gubbio, Italy

The Red Bird magnetostratigraphic sequence thatranges from C33n to C31n (Figure 3G) can bedirectly and simply correlated to the time-equivalentsequence in the reference section at Gubbio, Italy(Alvarez et al., 1977; Roggenthen & Napoleone, 1977;Lowrie & Alvarez, 1977), and thereby the planktonicforaminiferal zonation scheme of the European pel-agic realm can be correlated directly to the WesternInterior provincial ammonite zonation (Figure 9). Inthis diagram the foraminiferal biostratigraphy ofGubbio (Premoli-Silva, 1977) was correlated on thebasis of the strict proportionality of the stratigraphicposition of the biozone boundaries within theassociated geomagnetic interval.

In this study there is a degree of uncertaintysurrounding the correlation of R1 at Red Bird with theC32r polarity reversal, but we have shown the range ofGlobotruncana calcarata in Figure 9 plotted at between0.7 and 0.81, or some four-fifths of the way up in theC33n interval, which is where it is found at Gubbio.Globotruncana calcarata has been reported fromnear the top of the Annona Chalk of Arkansas inassociation with a bentonite dated at 75.2&0.5 Ma

(Obradovich et al., 1990). Chronologically, thisisotopic age suggests that the range zone ofGlobotruncana calcarata is probably younger and cor-relates to an interval that ranges from B. scotti to E.jenneyi (Obradovich et al., 1990; see also Figure 7).

The Campanian/Maastrichtian boundary as definedat Gubbio on the basis of planktic foraminifera lies atthe last occurrence of Globotruncana calcarata, whichwe correlate paleomagnetically to B. gregoryensis in theammonite zonation of the Western Interior (Figure 9).This would appear to be anomalously old, as Kennedyet al. (1992) place the top of the Campanian in theBaculites eliasi Zone, based on the occurrence ofNostoceras (N.) hyatti and Jeletzkytes nodosus (Owen)in the Baculites jenseni Zone in the Pierre Shale ofColorado. These forms are latest Campanian in age,and are found in a short stratigraphic interval inPoland below the first occurrence of Belemnitellalanceolata, whose appearance is held to mark theCampanian/Maastrichtian stage boundary (Birkelundet al., 1984). This correlation of the Campanian/Maastrichtian boundary is in part supported by theseawater 87Sr/86Sr curve for the Cenozoic andCretaceous (Koepnick et al., 1985; Hess et al., 1986).McArthur et al. (1992) analysed the 87Sr/86Sr ratiosfrom ammonites collected from the Western Interiorand compared them with values from the firstappearance of Belemnitella lanceolata, and inferred thatthe boundary lies within the zone of Baculites jenseni.Both of these placements of the stage boundary cor-respond to the lower to middle part of C32n (Figure3), which in turn correlates to the upper part of theGlobotruncana tricarinata zone at Gubbio (Figure 9).

The magnetostratigraphic correlation of the LateCretaceous of the Western Interior of the US with theLate Cretaceous of Europe does not solve any of the

22 J. F. Hicks et al.

problems encountered in fixing the Campanian/Maastrichtian biostratigraphically in the type sections(Eaton, 1987; Kennedy, 1995). But it is a means ofcorrelation that can potentially cut through theproblems of biostratigraphic diachroneity, and thediscrepancies between the zonation schemes based onmacro- and microfossils (Hambach et al., 1995).

Figure 9. Chronologic correlation diagram of the magnetostratigraphy of the Red Bird section and the correspondingWestern Interior ammonite biozonation, directly correlated with the reference section of Alvarez et al. (1977) fromGubbio, and the corresponding planktonic forminiferal zones of Premoli-Silva (1977).

10.2. Correlation of Red Bird to Red Deer Valley, Canada

A ground-breaking magnetostratigraphic study of thelatest Cretaceous sediments of the Red Deer Valleywas carried out by Lerbekmo & Coulter in 1984. Theyplaced the Campanian/Maastrichtian boundary at thelevel of Baculites baculus in a magnetostratigraphic

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 23

section at the Red Deer Valley, in the fluvio-deltaicsequence of the Edmonton Group. The top of thepolarity sequence was controlled by the reversal 29n/29r within which the K/T boundary was fixed palyno-logically and paleontologically (Lerbekmo et al.,1979). The polarity sequence below this level containsthe zones Baculites grandis to Baculites reesidei. Onthe basis of Jeletsky’s (1968) placement of theCampanian/Maastrichtian boundary at the base ofBaculites baculus (dated at 71.0 Ma; Lerbekmo &Coulter, 1984), and on the first appearance of thepalynofloral genus Wodehouseia, Lerbekmo (1989)correlated this level to the stage boundary at Gubbio,which is placed in the upper part of C33n (Alvarezet al., 1977). This led to the mis-identification ofC32n as C33n. Comparison of the composite mag-netostratigraphic section at Red Deer Valley with RedBird shows that the ammonite zones Baculites reesideithrough to the lower part of Baculites grandis lie in anormal polarity interval in both sections. This studyidentifies this polarity interval as C32n for the reasonsoutlined in the previous sections.

10.3. Correlation of the Western Interior seaway to thePacific slope province

To the west of the Western Interior foreland basin, onthe Pacific side of the Late Cretaceous peninsulaadjacent to the convergence zone, anomaly 32r hasbeen identified in the deep-sea fan deposits of thePoint Loma Formation at La Jolla and Point Lomain southern California (Bannon et al., 1989). TheCampanian/Maastrichtian boundary (as fixed byAlvarez et al., 1977, at the extinction of Globotruncanacalcarata), was placed just below the range zone ofBaculites occidentalis, in the uppermost part of C33n(Bannon et al., 1989). Assuming uniform sedimenta-tion in the Point Loma Formation, Baculites occiden-talis extends from approximately four-fifths of the wayup in C33n, to the C33n/C32r reversal, or from about75.3 Ma to 74.6 Ma. If the N1/R1 reversal at RedBird is held to lie close to the top of C33n then we cannow correlate the range of Baculites occidentalis (asdefined in the Point Loma Formation) from approxi-mately the uppermost part of Baculites scotti to at leastthe top of Exiteloceras jenneyi, thereby establishinga direct correlation between the Western Interiorseaway and the Pacific slope strata.

11. Conclusions

Isotopic ages are critical in calibrating the oceanicmagnetic anomalies that are the template for theGPTS, and it is the GPTS that can then be used as a

measure to estimate the ages of the geologic stageboundaries in regions where isotopic ages are notavailable. In this way a well-calibrated GPTS can bethe central link that correlates isotope stratigraphy andbiostratigraphy with isotopic ages. Theoretically, theprecision of the GPTS ‘link’ is limited only by theanalytical precision of the calibrating isotopic agesthemselves and the precision with which the width ofthe ocean floor anomalies can be measured (approxi-mately 7%). This chronostratigraphic analysis of thePierre Shale in eastern Wyoming makes a number ofcontributions to the calibration of the GPTS, andto the broad geologic and paleontologic frameworkof the Campanian and Maastrichtian. These aresummarized below:(1) Within the limitations that we have outlined,this study correlates the geomagnetic reversals of themiddle Campanian and lower Maastrichtian to thehigh resolution ammonite zonation of the PierreShale. In doing so we have significantly improvedthe isotopic age constraints of three of the latestCretaceous reversals in the interval from C33n toC31n. Within the uncertainties that we have clearlydefined, our age estimates are entirely in accordancewith those already published for the GPTS. Theresults of this study indicate that further chronostrati-graphic analyses of latest Cretaceous strata in thisregion have the potential to produce multiple inter-polations across individual reversal levels. This shoulddecrease our current levels of uncertainty to thepoint where the interpolated ages of the geomagneticreversals begin to approach the precision of the cali-brating isotopic ages, although it should be noted thatthe isotopic age assignments for even a well-knownand intensively studied level, such as the K/T bound-ary, which has been the subject of numerous studies,still range over more than one million years. This is aprecision of no better than between one and twopercent, and a far cry from the published analyticalprecision of the isotopic ages themselves.(2) Correlation of the magnetostratigraphy of thePierre Shale to the GPTS has allowed the first directchronologic correlation between the Western Interiorammonite biozonation scheme and: (a) the referencesection at Gubbio in Italy; (b) the Point LomaFormation on the Pacific slope. This is a very incom-plete sample of numerous correlations that are alsopossible, but cannot be addressed here (for asummary, see Opdyke & Channell, 1996; Table 11.1,pp. 184, 185).(3) Dating this marine sequence at multiple levels hasenabled us to quantify the changes in sedimentationrate that occur over time in a foreland basin, osten-sibly in response to relative changes in sea level. The

24 J. F. Hicks et al.

causal mechanisms behind these sea-level fluctuationsare clearly allocyclic, and temporally and spatiallycomplex, controlled as they are by episodic thrustingin the west, followed by rapid changes in sedimentsupply and subsidence. But the fluctuations in terri-genous input to the more distal part of the basinappear to be predictable and move in synchronicitywith migration of the strandline over time.

A chronologic study in this region has to cope witha number of factors that greatly complicate whatshould be the relatively simple task of calibratingmagnetic reversals with isotopic ages that are foundtogether within a common biostratigraphic frame-work. All the necessary components are, after all,found together in the same outcrop at numerouslocalities in the Western Interior US. But the work ishindered by a number of practical difficulties. The ashlevels are numerous, but only a relative few will yieldsufficient sanidine to be datable, and the levels theydate may not be the most critical ones. Ammonites arelocally abundant, but a perusal of many of even thebest correlated sections in this region (e.g., Schultzet al., 1980), show that there are large gaps in thesequence where ammonites are not preserved. Thepresence (or absence) of the magnetic reversals isinfluenced by lithology, the completeness of surfaceexposure and the presence of unconformities in thesequences. A more precise calibration of the latestCretaceous hinges on finding the few sections in thisregion where these nullifying factors are minimised.

Acknowledgements

This study was carried out with the support ofthe following. To JFH: Sigma Xi, Grants-in-Aidof Research (1989, 1990); Geological Society ofAmerica Research Grants (1989, 1990, 1992); RockyMountain Coal Geology Scholarship; ColoradoScientific Society Seven Oriel Memorial Fund;Yale-Peabody Museum. Acknowledgement is alsomade to the donors of the Petroleum Research fund,administered by the ACS, for partial support of thisresearch (to LT, ACS-PRF, 30442-AC8). To JDO:Pecora Fellowship (1995–1997) and support from theNational Geologic Mapping Program of the USGS.Fieldwork was carried out with the assistance ofDavid Clark, Lisa Maislich, and Matthew Carrano.We gratefully acknowledge the help and advice ofDr William A. Cobban, Dr Karl Waage, Dr Leo J.Hickey, and Dr Kirk R. Johnson, at whose suggestionthis study was initiated.

References

Alvarez, W., Arthur, M. A., Fischer, A. G., Lowrie, W., Napoleone,G. & Silva, I. P. 1977. Upper Cretaceous-Paleocene magnetic

stratigraphy at Gubbio Italy V. Type section for the LateCretaceous-Paleocene geomagnetic reversal time scale. GeologicalSociety of America, Bulletin 88, 383–389.

Baadsgaard, H., Lerbekmo, J. F., Wilbrans, J. R., Swisher III, C. C.& Fanning, M. 1993. Multi-method radiometric age for abentonite near the top of the Baculites reesidei Zone ofsouth-western Saskatchewan (Campanian-Maastrichtian stageboundary?). Canadian Journal of Earth Sciences 30, 769–775.

Bannon, J. L., Bottjer, D. J., Lund, S. P. & Saul, L. R. 1989.Campanian/Maastrichtian stage boundary in southern California:resolution and implications for large-scale depositional patterns.Geology 17, 80–83.

Berggren, W. A., Kent, D. V., Obradovich, J. D. & Swisher, C. C.1992. Towards a revised Paleogene geochronology. In Eocene-Oligocene climatic and biotic evolution (eds Prothero, D. R. &Berggren, W. A.), pp. 29–45 (Princeton University Press,Princeton, NJ).

Bergstresser, T. J. 1983. Radiolaria from the Upper CretaceousPierre Shale, Colorado, Kansas, Wyoming. Journal of Paleontology57, 877–882.

Bergstresser, T. J. & Frerichs, W. E. 1982. Planktonic Foraminiferafrom the Upper Cretaceous Pierre Shale at Red Bird, Wyoming.Journal of Foraminiferal Research 12, 352–361.

Bergstresser, T. J. & Krebs, W. N. 1983. Late Cretaceous(Campanian-Maastrichtian) diatoms from the Pierre Shale,Wyoming, Colorado and Kansas. Journal of Paleontology 57,883–891.

Birkelund, T., Hancock, J. M., Hart, M. B., Rawson, P. F.,Remane, J., Robaszynski, F., Schmid, F. & Surlyk, F. 1984.Cretaceous stage boundaries – proposals. Bulletin of the GeologicalSociety of Denmark 33, 3–20.

Cande, S. C. & Kent, D. V. 1992. A new geomagnetic polaritytime scale for the Late Cretaceous and Cenozoic. Journal ofGeophysical Research 97, B10, 13 917–13 951.

Cande, S. C. & Kent, D. V. 1995 Revised calibration of thegeomagnetic polarity time scale for the Late Cretaceousand Cenozoic. Journal of Geophysical Research 100, B4,6093–6095.

Christiansen, E. H., Kowallis, B. J. & Barton, M. D. 1994.Temporal and spatial distribution of volcanic ash in Mesozoicsedimentary rocks of the Western Interior: an alternative recordof Mesozoic magmatism. In Mesozoic System of the RockyMountain Region, USA (eds Caputo, M. V., Peterson, J. A. &Franczyk, K. J.), pp. 73–94 (Rocky Mountain Section SEPM,Society for Sedimentary Geology, Denver, CO).

Clemens, W. A. 1960. Stratigraphy of the type Lance Formation.International Geological Congress, 1st Session Report 2, 7–13.

Cobban, W. A., Merewether, E. A., Fouch, T. D. & Obradovich,J. D. 1994. Some Cretaceous shorelines in the Western Interiorof the United States. In Mesozoic System of the Rocky MountainRegion, USA (eds Caputo, M. V., Peterson, J. A. & Franczyk,K. J.), pp. 393–414 (Rocky Mountain Section, SEPM Society forSedimentary Geology, Denver, CO).

Crandell, D. R. 1950. Revision of the Pierre Shale of central SouthDakota. American Association of Petroleum Geologists, Bulletin 34,2337–2346.

Crandell, D. R. 1958. Geology of the Pierre area, South Dakota.US Geological Survey, Professional Paper 307, 83 pp.

Dalrymple, G. B. 1989. The GLM continuous laser system for40Ar/39Ar dating: description and performance characteristics.US Geological Survey, Bulletin 1890, 89–96.

Dalrymple, G. B., Alexander Jr, E. C., Lanphere, M. A. & Kraker,G. P. 1981. Irradiation of samples for 40Ar/39Ar dating using theGeological Survey TRIGA reactor: US Geological Survey,Professional Paper 1176, 55 pp.

Dalrymple, G. B. & Duffield, W. A. 1988. High 40Ar/39Ar datingof Oligocene rhyolites from the Mogollon-Datil volcanic fieldusing a continuous laser system. Geophysical Research Letters 15,463–466.

Dalrymple, G. B., Izett, G. A., Snee, L. W. & Obradovich, J. D.1993. 40Ar/39Ar age spectra and total-fusion ages of tektites from

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 25

Cretaceous-Tertiary boundary sedimentary rocks in the BelocFormation, Haiti. US Geological Survey, Bulletin 2065, 20 pp.

Darton, N. H. 1901. Preliminary description of the geology andwater resources of the southern half of the Black Hills andadjoining areas in South Dakota and Wyoming. US GeologicalSurvey, 21st Annual Report 4, 489–599.

Darton, N. H. 1918. The structure of parts of the central GreatPlains. US Geological Survey, Bulletin 691-A, 1–26.

DeGraw, H. M. 1975. The Pierre-Niobrara unconformity in west-ern Nebraska. In The Cretaceous System in the Western Interiorof North America (ed. Caldwell, W. G. E.), Geological Associationof Canada, Special Publication 13, 589–606.

Diehl, J. F., Beck Jr, M. E., Beske-Diehl, S., Jaconson, D. & HearnJr, B. C. 1991. Paleomagnetism of the Late Cretaceous-earlyTertiary north-central Montana alkalic province. Journal ofGeophysical Research 96, 11 723–11 740.

Dorf, E. 1942. Flora of the Lance Formation at its type locality,Niobrara County, Wyoming, pt. 2 of Upper Cretaceous floras ofthe Rocky Mountain region. Carnegie Institution Washington,Publication 508, 70–159.

Eaton, J. G. 1987. The Campanian-Maastrichtian boundary in theWestern Interior of North America. Newsletters on Stratigraphy 18,31–39.

Elias, M. K. 1931. The geology of Wallace County, Kansas. KansasGeological Survey, Bulletin 18, 43–131.

Elison, M. W. 1991. Intercontinental contraction in western NorthAmerica: continuity and episodicity. Geological Society of America,Bulletin 103, 1226–1238.

Fassett, J. E. & Obradovich, J. D. 1986. A high precision 40Ar/39Arage-spectrum plateau of 74 Ma from the uppermost, UpperCretaceous Fruitland Formation, Kirtland Shale, identifiespaleomagnetic chron 33-normal in the San Juan basin, NewMexico. Geological Society of America, Abstracts with Programs 18(6), 598.

Fassett, J. E. & Steiner, M. B. 1997. Precise age of C33n-C32rmagnetic polarity reversal, San Juan Basin, New Mexico andColorado. In Mesozoic structure and paleontology of the FourCorners area, New Mexico Geological Society, 48th Field ConferenceGuidebook (in press).

Fassett, J. E., Cobban, W. A. & Obradovich, J. D. 1997. Biostrati-graphic and isotopic age of the Huerfanito bentonite bed ofthe Upper Cretaceous Lewis Shale at an outcrop near Regina,New Mexico. In Mesozoic structure and paleontology of the FourCorners area, New Mexico Geological Society, 48th Field ConferenceGuidebook (in press).

Fisher, R. A. 1953. Dispersion on a sphere. Proceedings of the RoyalSociety A217, 295–305.

Fox, J. E. 1993. Stratigraphic cross-sections M-M ‘through R-R’showing electric logs of Upper Cretaceous and older rocks,Powder River Basin, Wyoming. US Geological Survey, Oil and GasInvestigations Chart OC-137.

Gale, N. H. 1982. The physical decay constants. In Numericaldating in Stratigraphy (ed. Odin, G. S.), pp. 107–122 (John Wileyand Sons, Chichester).

Gautier, D. L. 1986. Cretaceous shales from the Western Interiorof North America: sulfur/carbon ratios and sulfur-isotopecomposition. Geology 14, 225–228.

Gautier, D. L. & Parrish, J. T. 1987. Sharon Springs Memberof Pierre Shale: test of the upwelling hypothesis. GeologicalAssociation of Canada/Mineralogical Association of Canada,Abstracts with Programs 12, 45.

Gill, J. R. & Cobban, W. A. 1961. Stratigraphy of lower and middleparts of the Pierre Shale, northern Great Plains. US GeologicalSurvey, Professional Paper 424-D, D185–D191.

Gill, J. R. & Cobban, W. A. 1962. Red Bird Silty Member of thePierre Shale, a new stratigraphic unit. US Geological Survey,Professional Paper 450-B, B21–B24.

Gill, J. R. & Cobban, W. A. 1965. Stratigraphy of the Pierre Shale,Valley City and Pembina Mountain areas, North Dakota. USGeological Survey, Professional Paper 392-A, 20 pp.

Gill, J. R. & Cobban, W. A. 1966a. The Red Bird section ofthe Upper Cretaceous Pierre Shale in Wyoming. US GeologicalSurvey, Professional Paper 393-A, 73 pp.

Gill, J. R. & Cobban, W. A. 1966b. Regional unconformity in LateCretaceous Wyoming. US Geological Survey, Professional Paper550-B, B20–B27.

Gill, J. R. & Cobban, W. A. 1973. Stratigraphy and geologic historyof the Montana Group and equivalent rocks, Montana, Wyomingand North and South Dakota. US Geological Survey, ProfessionalPaper 776, 37 pp.

Gill, J. R., Cobban, W. A. & Schultz, L. G. 1972. Stratigraphy andcomposition of the Sharon Springs Member of the Pierre Shale inWestern Kansas. US Geological Survey, Professional Paper 728,50 pp.

Gradstein, F. M., Agterberg, F. P., Ogg, J. G., Hardenbol, J., vanVeen, P., Thierry, J. & Huang, Z. 1994. A Mesozoic time scale.Journal of Geophysical Research 99, B12, 24 051–24 074.

Hallam, A., Hancock, J. M., LaBreque, J. L., Lowrie, W. &Channell, J. E. T. 1984. Jurassic to Paleogene: Part 1. Jurassicand Cretaceous geochronology, and Jurassic to Paleogene mag-netostratigraphy. In The chronology of the geological record (ed.Snelling, N. J.), Geological Society of London, Memoir 10, 118–140.

Hambach, U., Hiss, M. & Peryt, D. 1995. Magnetostratigraphyof Santonian to Maastrichtian strata from NW Germany andPoland. Subcommision on Cretaceous Stratigraphy, Second Inter-national Symposium on Cretaceous Stage Boundaries. Institut Royaldes Sciences Naturelles de Belgique, Abstracts Volume, Brussels 1995,p. 51.

Hamilton, W. 1978. Mesozoic tectonics of the Western UnitedStates. In Mesozoic paleogeography of the Western United States (edsHowell, D. G. & McDougall, K. A.), pp. 33–70 (Pacific Section,Society of Economic Paleontologists and Mineralogists).

Harland, W. B., Armstrong, R. L., Cox, A. V., Craig, L. E., Smith,A. G. & Smith, D. G. 1989. A geologic time scale 1989, 263 pp.(Cambridge University Press, Cambridge).

Haq, B. U., Hardenbol, J. & Vail, P. R. 1988. Mesozoic andCenozoic chronostratigraphy and cycles of sea-level change. InSea-level changes – an integrated approach; Special Publication,Society of Economic Paleontologists and Mineralogists 42, 71–108.

Heirtzler, J. R., Dickson, G. O., Herron, E. M., Pitman, W. C. &Le Pichon, X. 1968. Marine magnetic anomalies, geomagneticfield reversals, and motions of the ocean floor and continents.Journal of Geophysical Research 73, 2119–2136.

Herbert, T. D. & D’Hondt, S. L. 1990. Precessional climatecyclicity in Late Cretaceous-early Tertiary marine sediments: ahigh resolution chronometer of Cretaceous-Tertiary boundaryevents. Earth and Planetary Science Letters 99, 263–275.

Herbert, T. D., D’Hondt, S. L., Premoli-Silva, I., Erba, E. &Fischer, A. G. 1995. Orbital chronology of Cretaceous-earlyPaleocene marine sediments. In Geochronology, time scales andstratigraphic correlation (eds Berggren, W. A., Kent, D. V., Aubry,M. & Hardenbol, J.), Special Publication, Society of EconomicPaleontologists and Mineralogists 54, 81–94.

Hess, J., Bender, M. L. & Schilling, J.-G. 1986. Evolution ofthe ratio of strontium-87 to strontium-86 in seawater fromCretaceous to present. Science 231, 979–984.

Hess, J. C. & Lippolt, H. J. 1987. Interpretation of 40Ar/39Arspectra of biotites: evidence from hydro-thermal degassing exper-iments and TEM studies. Chemical Geology (Isotope GeoscienceSection) 66, 137–149.

Hicks, J. F., Obradovich, J. D. & Tauxe, L. 1995. A new calibrationpoint for the Late Cretaceous time scale: the 40Ar/39Ar isotopicage of the C33r/C33n geomagnetic reversal from the Judith RiverFormation (Upper Cretaceous), Elk Basin, Wyoming, USA.Journal of Geology 103, 243–256.

Hilgen, F. J. 1991. Extension of the astronomically calibrated(polarity) time scale to the Miocene/Pliocene boundary. Earthand Planetary Science Letters 107, 349–368.

Izett, G. A., Dalrymple, G. B. & Snee, L. W. 1991. 40Ar/39Ar age ofCretaceous-Tertiary boundary tektites from Haiti. Science 252,1539–1542.

26 J. F. Hicks et al.

Izett, G. A., Cobban, W. A., Obradovich, J. D. & Kunk, M. J. 1993.The Manson impact structure 40Ar/39Ar age and its distal impactejecta in the Pierre Shale in southeastern South Dakota. Science262, 729–732.

Izett, G. A., Cobban, W. A., Dalrymple, G. B. & Obradovich, J. D.1998. 40Ar/39Ar age of the Manson impact structure, Iowa, andcorrelative impact ejecta in the Crow Creek Member of thePierre Shale (Upper Cretaceous), South Dakota and Nebraska.Geological Society of America, Bulletin 110, 361–376.

Jeletsky, J. A. 1968. Microfossil zones of the marine Cretaceous ofthe Western Interior of Canada and their correlation with zonesand stages of Europe and the Western Interior of the UnitedStates. Geological Survey of Canada, Paper 67, 66 pp.

Jordan, T. E. 1981. Thrust loads and foreland basin evolution,Cretaceous, Western United States. American AssociationPetroleum Geologists, Bulletin 65, 2506–2520.

Jordan, T. E. & Flemings, P. 1991. Large-scale stratigraphicarchitecture, eustatic variation and unsteady tectonism: atheoretical approach. Journal of Geophysical Research 96,6681–6699.

Kauffman, E. G. 1977. Geological and biological overview:Western Interior Cretaceous basin. The Mountain Geologist 14,75–99.

Kennedy, W. J. 1995. Ammonite definitions of Late Cretaceousstage boundaries: progress since 1984. Subcommision onCretaceous Stratigraphy, Second International Symposium onCretaceous Stage Boundaries. Institut Royal des Sciences Naturelles deBelgique, Abstracts Volume, Brussels 1995, p. 65.

Kennedy, W. J., Cobban, W. A. & Scott, G. R. 1992. Ammonitecorrelation of the uppermost Campanian of Western Europe, theUS Gulf Coast, Atlantic Seaboard and Western Interior, andthe numerical age of the base of the Maastrichtian. GeologicalMagazine 129, 497–500.

Kent, D. V. & Gradstein, F. M. 1985. Cretaceous andJurassic geochronology. Geological Survey of America, Bulletin 91,1419–1427.

Kirschvink, J. L. 1980. The least-squares line and plane andthe analysis of paleomagnetic data. Royal Astronomical Society,Geophysics Journal 62, 699–718.

Koepnick, R. B., Burke, W. H., Denison, R. E., Hetherington,E. A., Nelson, H. F., Otto, J. B. & Waite, L. E. 1985. Construc-tion of the seawater 87Sr/86Sr curve for the Cenozoic and theCretaceous: supporting data. Chemical Geology (Isotope GeoscienceSection) 58, 55–81.

Kowallis, B. J., Christiansen, E. H., Deino, A. L., Kunk, M. J. &Heaman, L. M. 1995. Age of the Cenomanian-Turonian bound-ary in the Western Interior of the United States. CretaceousResearch 16, 109–129.

Landman, N. H. & Waage, K. M. 1993. Scaphitid ammonites ofthe Upper Cretaceous (Maastrichtian) Fox Hills Formation inSouth Dakota and Wyoming. Bulletin of the American Museum ofNatural History 215, 257 pp.

Lehman, T. M. 1987. Late Maastrichtian paleoenvironments anddinosaur biogeography in the Western Interior of North America.Palaeography, Palaeoclimatology, Palaeoecology 60, 189–217.

Lerbekmo, J. F. 1989. The stratigraphic position of the 33-33r(Campanian) polarity chron boundary in southeastern Alberta.Bulletin of Canadian Petroleum Geology 37, 43–47.

Lerbekmo, J. F., Evans, M. E. & Baadsgaard, H. 1979. Magneto-stratigraphy, biostratigraphy and geochronology of Cretaceous-Tertiary boundary sediments, Red Deer Valley. Nature 279,26–30.

Lerbekmo, J. F. & Coulter, K. C. 1984. Late Cretaceous to earlyTertiary magnetostratigraphy of a continental sequence: RedDeer Valley, Alberta, Canada. Canadian Journal of Earth Sciences22, 567–583.

Lillegraven, J. A. & Ostresh, L. M. 1990. Late Cretaceous (earliestCampanian/Maastrichtian) evolution of western shorelines of theNorth American Western Interior Seaway in relation to knownmammalian faunas. Geological Society of America, Special Paper243, 1–8.

Lindsay, E. H., Butler, R. F. & Johnson, N. M. 1982. Testing ofmagnetostratigraphy in late Cretaceous and early Tertiarydeposits, San Juan Basin, New Mexico. Geological Society ofAmerica, Special Paper 190, 449–453.

Loomis, F. B. 1915. A new mosasaur from the Fort Pierre.American Journal of Science, 4th Series 39, 233, 555–566.

Loutit, T. S., Hardenbol, J., Vail, P. R. & Baum, G. R. 1988.Condensed sections: the key to age determination and correlationof continental margin sequences. Society of Economic Paleontol-ogists and Mineralogists, Special Publication 42, 183–213.

Lowrie, W. & Alvarez, W. 1981. One hundred million years ofgeomagnetic polarity history. Geology 9, 392–397.

McArthur, J. M., Kennedy, W. J., Gale, A. S., Thirwell, M. F.,Chen, M., Burnett, J. & Hancock, J. M. 1992. Strontium isotopestratigraphy in the Late Cretaceous: intercontinental correlationof the Campanian/Maastrichtian boundary. Terra Nova 4,385–393.

McFadden, P. L. & McElhinny, M. W. 1988. The combinedanalysis of remagnetization circles and direct observations inpaleomagnetism. Earth and Planetary Science Letters 87, 161–172.

McWilliams, M. O., Baksi, A. K., Bohor, B. F. & Murali, A. V.1992. High precision relative ages of K/T boundary events inNorth America and Deccan Trap volcanism in India. EOS,Transactions of the American Geophysical Union 73 (14), p. 363.

Meek, F. B. & Hayden, F. V. 1861. Descriptions of new lowerSilurian, (Primordial), Jurassic, Cretaceous and Tertiary fossilscollected in Nebraska territory: with some remarks on the rocksfrom which they were obtained. Proceedings of the Academy ofNatural Sciences, Philadelphia, 415–447.

Mello, J. F. 1969. Foraminifera and stratigraphy of the upper partof the Pierre Shale and lower part of the Fox Hills Sandstone(Cretaceous) north-central South Dakota. US Geological Survey,Professional Paper 611, 121 pp.

Merewether, E. A. & Cobban, W. A. 1986. Biostratigraphic unitsand tectonism in the mid-Cretaceous foreland of Wyoming,Colorado, and adjoining areas. In Paleotectonics and sedimentationin the Rocky Mountain region, United States (ed. Peterson, J. A.),American Association of Petroleum Geologists, Memoir 41, 443–468.

NACSN (North American Commission on Stratigraphic Nomen-clature) 1983. North American Stratigraphic Code. AmericanAssociation of Petroleum Geologists, Bulletin 67, 841–875.

Obradovich, J. D. 1988. A different perspective in glauconite as achronometer for geologic time scale studies. Paleoceanography 3,757–770.

Obradovich, J. D., Bryant, W. A., Cobban, W. A. & Kennedy,W. J. 1990. An isotopic age for the late Campanian Globotruncanacalcarata planktonic foraminifera from the Annora Chalk ofSW Arkansas and its relation to the western interior ammonitezones. Geological Society of America, Abstracts with Programs 22 (7),A234.

Obradovich, J. D. & Cobban, W. A. 1975. A time-scale for the LateCretaceous of the Western Interior of North America. GeologicalAssociation of Canada, Special Paper 13, 31–54.

Obradovich, J. D. 1993. A Cretaceous time scale. In Evolution of theWestern Interior Basin (eds Caldwell, W. G. E. & Kauffman,E. G.), Geological Association of Canada, Special Paper 39,379–396.

Olsson, R. K. 1991. Cretaceous to Eocene sea-level fluctuations onthe New Jersey margin. Sedimentary Geology 70, 195–208.

Opdyke, N. D. & Channell, J. E. T. 1996. Magnetic stratigraphy,346 pp. (Academic Press, San Diego).

Premoli-Silva, I. 1977. Upper Cretaceous-Paleocene magneticstratigraphy at Gubbio, Italy. II. Biostratigraphy. GeologicalSociety of American, Bulletin 88, 371–374.

Reynolds, M. W. 1967. Physical evidence for Late Cretaceousunconformity, south-central Wyoming. US Geological Survey,Professional Paper 575-D, D24–D28.

Robinson, C. S., Mapel, W. J. & Cobban, W. A. 1959. Pierre Shalealong western and northern flanks of Black Hills, Wyoming andMontana. American Association of Petroleum Geologists, Bulletin 43,101–123.

Magnetostratigraphy, isotopic age calibration and intercontinental correlation 27

Robinson, G. D., Klepper, M. R. & Obradovich, J. D. 1968.Overlapping plutonism, volcanism and tectonism in the Boulderbatholith region, western Montana. In Studies in volcanology – amemoir in honor of Howel Williams (eds Coats, R. R., May, R. L. &Anderson, C. A.), Geological Society of American, Memoir 116,557–576.

Rogers, R. R., Swisher, C. C. & Horner, J. R. 1993. 40Ar/39Ar ageand correlation of the non-marine Two Medicine Formation(Upper Cretaceous), north-western Montana. Canadian Journalof Earth Sciences 30, 1006–1075.

Roggenthen, W. M. & Napoleone, G. 1977. Upper Cretaceous-Paleocene magnetic stratigraphy at Gubbio, Italy IV. UpperMaastrichtian-Paleocene magnetic stratigraphy. GeologicalSociety of America, Bulletin 88, 378–383.

Samson, S. D. & Alexander Jr, E. C. 1987. Calibration of theinterlaboratory 40Ar-39Ar dating standard, MMhb-1. ChemicalGeology (Isotope Geoscience Section) 66, 27–34.

Schultz, L. G., Tourtelot, H. A., Gill, J. R. & Boerngen, J. G. 1980.Composition and properties of the Pierre Shale and equivalentrocks, northern Great Plains region. US Geological Survey,Professional Paper 1064-B, 114 pp.

Shackleton, N. J., Berger, A. & Peltier, W. R. 1990. An alternativeastronomical calibration of the lower Pleistocene timescale basedon ODP Site 677. Transactions of the Royal Society of Edinburgh,Earth Sciences 81, 251–261.

Shurr, G. W. & Reskind, J. 1984. Stratigraphic framework of theNiobrara Formation (Upper Cretaceous) in North and SouthDakota. In The Mesozoic of middle North America (eds Stott, D. F.& Glass, D. J.), Canadian Society of Petroleum Geologists, Memoir9, 205–219.

Spivey, R. C. 1940. Bentonite in southwestern South Dakota. SouthDakota Geological Survey, Report of Investigations 36, 56 pp.

Steidtmann, J. R. 1993. The Cretaceous foreland basin and itssedimentary record. In Geology of Wyoming (eds Snoke, A. W.,Steidtmann, J. R. & Roberts, S. M.), Geological Survey ofWyoming, Memoir 5, 250–271.

Steiner, M. B. & Shoemaker, E. M. 1996. A hypothesized Mansonimpact tsunami: paleomagnetic and stratigraphic evidence in theCrow Creek member, Pierre Shale. Geological Society of America,Special Paper 302, 425–432.

Swisher, C. C., Grajales-Nishimura, J. M., Montanari, A.,Margolis, S. V., Claeys, P., Alvarez, W., Renne, P. R., Cedillo-Pardo, E., Maurrasse, F. J. M. R., Curtis, G., Smit, S. &McWilliams, M. O. 1992. Coeval 40Ar/39Ar ages of 65.0 millionyears ago from Chicxulub crater melt rock and Cretaceous-Tertiary boundary tektites. Science 257, 954–958.

Tauxe, L., Kylstra, N. & Constable, C. 1991. Bootstrap statisticsfor paleomagnetic data. Journal of Geophysical Research 96,11 723–11 740.

Taylor, J. R. 1982. An introduction to error analysis, 270 pp.(University Science Books, Mill Valley, CA).

Vail, P. R., Audemard, F., Bowman, S. A., Eisner, P. N. &Perez-Crus, C. 1991. The stratigraphic signatures of tectonics,eustasy and sedimentology – an overview. In Cycles and events instratigraphy (eds Einsele, G., Ricken, W. & Seilacher, A.),pp. 617–659 (Springer-Verlag, Berlin).

Van Wagoner, J. C., Mitchum, R. M., Campion, K. M. &Rahmanian, P. 1990. Siliciclastic sequence stratigraphy in welllogs, cores, and outcrops: concepts for high-resolution correlationof time and facies. American Association of Petroleum Geologists,Methods in Exploration Series 7, 55 pp.

Weimer, R. J. 1984. Relation of unconformities, tectonics andsea-level changes, Cretaceous of Western Interior U.S.A. InInterregional unconformities and hydrocarbon accumulation (ed.Schee, J. S.), American Association of Petroleum Geologists, Memoir36, 7–35.

Wilcox, R. E. 1983. Refractive index determination using thecentral focus masking technique with dispersion colors. AmericanMineralogists 68, 1226–1236.

York, D., Hall, C. M., Yanase, Y., Hanes, J. A. & Kenyon, W. J.1981. 40Ar/39Ar dating of terrestrial materials with a continuouslaser. Geophysical Research Letters 8, 1136–1138.