Upper Cretaceous sequences and sea-level history, New Jersey Coastal Plain

Kenneth G. Miller² Department of Geological Sciences, Rutgers University, Piscataway, New Jersey 08854, USA Peter J. Sugarman New Jersey Geological Survey, P.O. Box 427, Trenton, New Jersey 08625, USA James V. Browning Department of Geological Sciences, Rutgers University, Piscataway, New Jersey 08854, USA Michelle A. Kominz Department of Geosciences, Western Michigan University, Kalamazoo, Michigan 49008-5150, USA Richard K. Olsson Mark D. Feigenson John C. HernaÂndez Department of Geological Sciences, Rutgers University, Piscataway, New Jersey 08854, USA

ABSTRACT pean sections, and Russian platform BACKGROUND outcrops points to a global cause. Because We developed a Late Cretaceous sea- backstripping, seismicity, seismic strati- Predictable, recurring sequences bracketed level estimate from Upper Cretaceous se- graphic data, and sediment-distribution by unconformities comprise the building quences at Bass River and Ancora, New patterns all indicate minimal tectonic ef- blocks of the stratigraphic record. Exxon Pro- Jersey (ODP [Ocean Drilling Program] Leg fects on the New Jersey Coastal Plain, we duction Research Company (EPR) de®ned a 174AX). We dated 11±14 sequences by in- interpret that we have isolated a eustatic depositional sequence as a ``stratigraphic unit tegrating Sr isotope and biostratigraphy signature. The only known mechanism composed of a relatively conformable succes- age resolution ؎0.5 m.y.) and then esti- that can explain such global changesÐ sion of genetically related strata and bounded) mated paleoenvironmental changes within glacio-eustasyÐis consistent with forami- at its top and base by unconformities or their the sequences from lithofacies and biofacies niferal ␦18O data. Either continental ice correlative conformities'' (Mitchum et al., analyses. Sequences generally shallow up- sheets paced sea-level changes during the 1977, p. 53); by implication, ``genetically re- section from middle-neritic to inner-neritic Late Cretaceous, or our understanding of lated'' refers to global changes in sea level paleodepths, as shown by the transition causal mechanisms for global sea-level (Vail et al., 1977). Christie-Blick and Driscoll from thin basal glauconite shelf sands change is fundamentally ¯awed. Compari- (1995) clari®ed the genetic connotation by (transgressive systems tracts [TST]), to son of our eustatic history with published recognizing sequence boundaries as unconfor- medial-prodelta silty clays (highstand sys- ice-sheet models and Milankovitch predic- mities associated with lowering of base level, .tems tracts [HST]), and ®nally to upper± tions suggests that small (5±10 ؋ 106 km3), including eustatic and tectonic mechanisms delta-front quartz sands (HST). Sea-level ephemeral, and areally restricted Antarctic Sequences have been recognized in diverse estimates obtained by backstripping (ac- ice sheets paced the Late Cretaceous global stratigraphic environments (e.g., ranging from counting for paleodepth variations, sedi- sea-level change. New Jersey and Russian siliciclastic to carbonate settings; see exam- ment loading, compaction, and basin sub- eustatic estimates are typically one-half of ples in Wilgus et al., 1988; de Graciansky et sidence) indicate that large (Ͼ25 m) and the EPR amplitudes, though this difference al., 1998) from the Proterozoic (e.g., Christie- rapid (K1 m.y.) sea-level variations oc- varies through time, yielding markedly dif- Blick et al., 1988) to the Holocene (e.g., Lock- curred during the Late Cretaceous green- ferent eustatic curves. We conclude that er et al., 1996) and have been related to global house world. The fact that the timing of Up- New Jersey provides the best available es- sea-level (eustatic) variations (Vail et al., per Cretaceous sequence boundaries in timate for Late Cretaceous sea-level 1977; Haq et al., 1987; Posamentier et al., New Jersey is similar to the sea-level low- variations. 1988). However, differences in accommoda- ering records of Exxon Production Re- tion (the sum of subsidence and eustatic search Company (EPR), northwest Euro- Keywords: eustasy, sequence stratigraphy, change) and sediment supply control sequence sea-level history, New Jersey Coastal Plain, architecture, including the nature and signi®- ²E-mail: [email protected]. Late Cretaceous, backstripping. cance of surfaces bounding and within se-

GSA Bulletin; March/April 2004; v. 116; no. 3/4; p. 368±393; doi: 10.1130/B25279.1; 13 ®gures; 1 table; Data Repository item 2004043.

For permission to copy, contact [email protected] 368 ᭧ 2004 Geological Society of America UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN

dle Eocene (Barron, 1983; Huber et al., 2002), the only known mechanism for producing the large (k10 m), rapid (Ͻ1 m.y.) global sea- level changes reported for this time (e.g., Haq et al., 1987; Hallam, 1992) is glacio-eustasy (see Donovan and Jones, 1979; Pitman and Golovchenko, 1983). Studies have begun to question the assumption of a completely ice- free world during the Cretaceous (e.g., Stoll and Schrag, 1996; Miller et al., 1999a, 2002; Price, 1999), suggesting that ice-volume changes may have been an important control on greenhouse sea-level changes. Evaluation of mechanisms of sea-level change requires a ®rm chronology and a means of extracting rates of change. Several passive-margin and epicontinental sea regions provide relatively high-resolution sequence- stratigraphic records for the Late Cretaceous. Deposits of the U.S. Western Interior Seaway provide a tephrochronology of sequences (Kauffman, 1977; Dean and Arthur, 1998), though these sections are complicated by com- pressional tectonics. The epicontinental and passive-margin sections of northwest Europe and the Russian platform provide excellent records of transgressive-regressive sequences (Hancock, 1993; de Graciansky et al., 1998; Sahagian et al., 1996). However, differing bio- zonal schemes complicate interregional cor- relations, and a ®rm global record of Upper Cretaceous sequences has proved to be elu- sive. Though workers at EPR produced a Late Cretaceous eustatic record (Haq et al., 1987), the database on which it is published is largely proprietary. De Graciansky et al. (1998) have provided public documentation of EPR's Me- sozoic sequences in northwest Europe. Where- as their Lower Cretaceous and older sequenc- es are reasonably well documented, the Upper Cretaceous sequences are poorly constrained in age. Pieces of the Upper Cretaceous Figure 1. Map showing the New Jersey Coastal Plain, the limit of Miocene and younger sequence-stratigraphic puzzle are falling into outcrop that approximates strike, and sites discussed here. place with publication of detailed studies of parts of the section (e.g., Gale et al. [2002] for the Cenomanian±early Turonian), but the quences (e.g., ¯ooding surfaces, transgressive shows that sequence boundaries from 42 to 8 overall picture of Upper Cretaceous sequences surfaces). In addition, accommodation and Ma correlate with global ␦18O increases, link- and rates of sea-level change is still blurry. sediment supply control the stacking patterns ing them with glacio-eustatic sea-level low- The New Jersey passive margin, particular- of facies successions within sequences (e.g., erings (Miller et al., 1996, 1998a). However, ly the onshore coastal plain (Fig. 1), has pro- systems tracts of Vail et al., 1977, and Posa- such a relationship has not been established vided a reference for Cenozoic sequences mentier et al., 1988) and the general three- for the ``greenhouse world'' of the pre±middle (Miller et al., 1996, 1998a) and potentially can dimensional arrangement of sequences (Rey- Eocene, in part re¯ecting the greater dif®culty provide similar records of Upper Cretaceous nolds et al., 1991). in obtaining and dating sequences older than sequences. The development of Late Creta- A clear relationship between sequence 42 Ma from passive margins throughout the ceous Sr isotope stratigraphy (Fig. 2; Burke et boundaries and glacio-eustatic sea-level low- world. al., 1982; Sugarman et al., 1995; McArthur et erings has been demonstrated for the ``ice- Mechanisms for sea-level change during the al., 1992, 1993, 1994; Howarth and McArthur, house world'' of the past 42 m.y. (Miller et greenhouse world are poorly understood. 1997) together with the application of Creta- al., 1996, 1998a). Comparing onshore and off- Though most investigators have assumed that ceous nannoplankton biostratigraphy (e.g., shore New Jersey sequences with ␦18O records Earth was largely ice free prior to the late mid- Bralower et al., 1995) provides the means of

Geological Society of America Bulletin, March/April 2004 369 MILLER et al.

Figure 2. Later Cretaceous Sr isotope age calibration for ODP (Ocean Drilling Program) Site 525 (open circles), U.S. Western Interior (open squares), and Kronsmoor, Germany (closed circles). A ®fth-order polynomial (thin gray line) and two linear ®ts (thick black lines) are shown.

reconciling correlations among regions and m paleodepth), prodelta silts (20±60 m), and ance on cuttings or discontinuous cores, or the evaluating the global signi®cance of regional outer-neritic glauconite sands (60±200 m). far updip locations of boreholes. sequences. Previous studies have documented at least Continuous coring and logging by onshore ®ve to eight Late Cretaceous cycles. Olsson ODP Leg 174AX at Bass River and Ancora, UPPER CRETACEOUS SEQUENCES IN (1963, 1975) recognized ®ve Late Cretaceous New Jersey, yielded thick, downdip Upper THE NEW JERSEY COASTAL PLAIN transgressive-regressive cycles in the New Jer- Cretaceous sections (Fig. 1) that allow dating sey Coastal Plain, whereas in the same area, of Upper Cretaceous (99±65 Ma) sequences Geologists have long noted cyclicity in Up- Owens and Sohl (1969) mapped ®ve to six (Fig. 5). For the Upper Cretaceous sediments per Cretaceous strata of the New Jersey Coast- Late Cretaceous cycles but inferred a tectonic al Plain (e.g., Lyell, 1845). Prominent middle- control on sedimentation. Owens and Gohn to outer-neritic glauconite beds (Figs. 3±5; (1985) recognized similar cycles throughout Figure 3. (A) Navesink±Mount Laurel for- Merchantville, Marshalltown, and Navesink the U.S. Atlantic Coastal Plain outcrops. By mational boundary and sequence bound- Formations) provide a visual key to recogniz- using outcrops and discontinuously sampled ary, Route 34, Matawan, New Jersey. SBÐ ing transgressive-regressive cycles (e.g., Wel- wells and boreholes, Olsson and coauthors sequence boundary; TSÐtransgressive ler, 1907). Upper Cretaceous facies in New (Olsson and Wise, 1987; Olsson et al., 1988; and surface; LSTÐhighstand systems tract; Jersey are generally deltaically in¯uenced Olsson, 1991) recognized eight transgressive- TSTÐtransgressive systems tract; HSTÐ marine-shelf facies (Fig. 4) with transgressive regressive cycles in New Jersey, identi®ed highstand systems tract. Paleodepth from shelf glauconite beds overlain by regressive them as sequences, and related them to sea- Miller et al. (1999a). (B) Sequence bound- prodelta silts and delta-front sands (Owens level change. Sugarman et al. (1995) integrat- aries observed in cores from Ancora, New and Sohl, 1969; Sugarman et al., 1995). The ed New Jersey Coastal Plain Sr isotope stra- Jersey. Wavy lines indicate sequence wave-dominated Niger Delta (Allen, 1970) tigraphy with nannofossil biostratigraphy to boundaries (unconformities). Depths in provides a useful analogue for these Late Cre- provide improved age estimates for these se- feet; cores generally represent 2 ft taceous environments (Fig. 4), as it is com- quences. Still, these previous studies were segments. posed of delta-front sands (inner neritic, Ͻ20 limited by poorly fossiliferous outcrops, reli- M

370 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN

Geological Society of America Bulletin, March/April 2004 371 MILLER et al.

Figure 4. Anatomy of a New Jersey Upper Cretaceous sequence. Shown at top is the facies model for the modern Niger River delta (Allen, 1970). On left is a diagrammatic model for a typical Upper Cretaceous marine sequence and gamma log (modi®ed after Sugarman et al., 1995). Pictures correspond to the three major facies observed in Upper Cretaceous marine sequences at Bass River (depths in feet are core depths); corresponding locations on modern facies model are indicated. of the New Jersey Coastal Plain, this paper Ancora boreholes in the site reports (Miller et Table 1 in Van Sickel et al. [2003] and Table presents sequence stratigraphy based on these al., 1998b, 1999b). On the basis of the data DR1 data for the Cenozoic are included be- boreholes that has million-year±scale age res- presented here (Figs. 5±11), we tentatively cause they were used in backstripping; see be- olution (Fig. 1). We provide (1) lithostrati- recognize three additional Upper Cretaceous low).1 For the neritic sections (primarily the graphic, facies, biostratigraphic, and Sr sequences for a total of 14. We informally Paleogene and Santonian±Maastrichtian), ben- isotope data for identifying sequences, docu- term sequences after their prominent basal thic foraminiferal biofacies studies provide the menting paleoenvironmental changes within (usually glauconite) lithostratigraphic units primary means of interpreting paleodepth and sequences (e.g., systems tracts), and correlat- largely named by Weller (1907) and recently recognizing inner (0±30 m; Fig. 4), middle ing sequences, and (2) age vs. depth diagrams updated by Owens et al. (1998). We discuss (30±100 m), and outer neritic (100±200 m) delineating the chronology of the Upper Cre- the Aptian to lowermost Cenomanian Potomac environments. Prodelta environments gener- taceous sequences as well as Late Cretaceous Formation (undifferentiated), Cenomanian± ally represent deposition in outer inner-neritic sedimentation rates. We use this sequence- lower Turonian Bass River sequences (Bass to inner middle-neritic (20±60 m) paleodepths stratigraphic framework to evaluate the con- River I, II, and III), upper Turonian±Coniacian (Fig. 4), and glauconite sands represent outer trolling mechanisms for the Upper Cretaceous Magothy sequences (tentatively divided into middle-neritic to outer-neritic paleodepths sequences in New Jersey and global sea level. Magothy I±III), Santonian Cheesequake se- (generally 60±200 m). The best-dated sections (e.g., outer neritic) often have the greatest er- METHODS quence, uppermost Santonian±middle Cam- panian Merchantville sequences (tentatively rors on absolute paleodepth (Ϯ50 m). How- ever, relative water-depth changes are better Sequence boundaries are recognized in divided into three sequences, Merchantville I± cores on the basis of physical stratigraphy in- III), the upper Campanian Marshalltown cluding irregular contacts (Figs. 3A and 3B), sequence, and the Maastrichtian±lower Paleo- 1GSA Data Repository item 2004043, sequence, reworking, bioturbation, major lithofacies cene Navesink sequence(s) (tentatively divid- age, paleodepth, paleodepth criteria for Ancora and changes (Fig. 5), gamma-ray patterns (Figs. 4 ed into Navesink I and II). Bass River for conservative age models and pre- Paleodepth and paleoenvironmental esti- ferred age models, and Sr isotope data, is available and 6±10), and age breaks (Fig. 11). Eleven on the Web at http://www.geosociety.org/pubs/ Upper Cretaceous to lowermost Tertiary se- mates are constrained by benthic foraminiferal ft2004.htm. Requests may also be sent to quences were identi®ed in the Bass River and biofacies (assemblages) and lithofacies (see [email protected].

372 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN

Figure 5. Updip (Ancora) to downdip (Bass River) comparison of lithostratigraphic units and sequences for the Upper Cretaceous sections. Wavy lines indicate unconformities. Shown are cumulative percent of clay-silt, ®ne sand, medium to coarse sand, glauconite, shells, and mica. Legend given in Figure 6. constrained within sequences. For example, cedure. Errors range from a few meters in benthic foraminiferal biofacies (Kominz and paleodepths clearly shoal up-section in se- nearshore sections, to Ϯ15 m as is typical for Pekar, 2001; Pekar and Kominz, 2001), and quences above maximum-¯ooding surfaces. inner middle-neritic zones), to Ϯ50 m for future drilling in the coastal plain is designed The paleodepth estimates were used to model outer-neritic zones (see Data Repository). to allow two-dimensional backstripping. eustatic variations by using backstripping Two-dimensional ¯exural backstripping pro- Lithofacies were recognized by using grain (see below); however, paleodepth remains the vides a means for re®ning water-depth esti- size, general lithology, bedding, and sedimen- greatest uncertainty in the backstripping pro- mates by using paleoslope modeling of the tary structures. Cumulative-percent plots of

Geological Society of America Bulletin, March/April 2004 373 MILLER et al.

374 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN the sediments in the cores were computed crop (Fig. 3A), and the lower Marshalltown though age is the independent variable (Fig. from samples taken at ϳ5ft(ϳ1.5 m) inter- sequence described here. 2). A ®fth-order polynomial was ®t to the 87Sr/ vals. Each sample was dried and weighed be- Age control for the Upper Cretaceous se- 86Sr data under the assumption that age is the fore washing, and the dry weight was used to quences is provided by integration of Sr iso- independent variable (Fig. 2). On the basis of compute the percentage of sand vs. silt and tope stratigraphy and biostratigraphy. Calcar- an in¯ection seen in the ®fth-order ®t, the data clay. The sand fraction was dry sieved through eous nannoplankton data provided by D. set was then broken into two groups, and two a 250 ␮m sieve, and the fractions were Bukry for Bass River and L. de Romero for linear regressions were ®t to the data (t ϭ weighed to obtain the percent of very ®ne and Ancora (Miller et al., 1998b, 1999b) yield ex- age): t ϭ 31,908.531372Ð(87Sr/ 86Sr ϫ ®ne vs. medium and coarse sand. The sand cellent biostratigraphic control, aided by sev- 44,984.801888), applicable from 73.5 to 65 fractions were examined by using a micro- eral key planktonic foraminiferal datum levels Ma and t ϭ 39,104.339163Ð(87Sr/ 86Sr ϫ scope, and a visual estimate was made of the (R.K. Olsson, this study). Pollen biostrati- 55,154.82511), applicable from 86.0 to 73.5 relative percentages of quartz, glauconite, car- graphic data provided by G.J. Brenner (in Ma. bonate (foraminifers and other shells), mica, Miller et al., 1998b, 1999b; and 2002, person- By using a similar late Campanian± and other materials contained in the sample. al commun.) yield age constraints on the non- Maastrichtian regression, Sugarman et al. As discussed above, Upper Cretaceous lith- marine to marginal-marine Magothy and Po- (1995) conservatively estimated age errors of ofacies at Bass River and Ancora generally tomac Formations. Ϯ1.9 m.y. at the 95% con®dence interval for follow a predictable, repetitive transgressive- Sr isotope±based age estimates were ob- one Sr isotope analysis; age errors for this in- regressive sequence pattern (Fig. 4): (1) basal tained from mollusk and foraminiferal shells terval are purportedly one order of magnitude unconformities (lower sequence boundaries); (Table DR2). Approximately 4±6 mg of shell better according to Howarth and McArthur (2) generally thin, lower, in situ glauconite or foraminiferal tests were cleaned ultrasoni- (1997). We estimate that the maximum age sands that are assigned to the TST of Posa- cally and dissolved in 1.5N HCl. Sr was sep- resolution using Sr isotope ratios for this in- mentier et al. (1988); basal glauconite sands arated by using standard ion-exchange tech- terval is Ϯ1 m.y. (i.e., the external precision are often absent in marginal- to shallow- niques (Hart and Brooks, 1974) and analyzed of ϳ0.000020 divided by the slopes of the re- marine sequences; (3) a coarsening-upward on a VG sector mass spectrometer at Rutgers gressions of ϳ0.000020/m.y.) succession of regressive medial silts deposited University. Internal precision on the sector for We integrated biostratigraphic and Sr iso- in prodelta environments (lower HST); (4) the data set averaged 0.000009, and the exter- tope ages on age vs. depth diagrams (Fig. 11). further regressive upper quartz sands of the nal precision is approximately Ϯ0.000020 We used biostratigraphic time-scale correla- upper HST; and (5) the upper sequence bound- (Oslick et al., 1994). NBS 987 is measured for tions of Bralower et al. (1995) and Erba et al. ary. The upper HST often includes reworked, these analyses at 0.710255 (2␴ standard de- (1999) tied to the Gradstein et al. (1994) geo- detrital glauconite recognized by covariance viation ϭ 0.000008, n ϭ 22) normalized to magnetic polarity time scale to obtain a ®rm with quartz and brown to yellow-green glau- 86Sr/88Sr of 0.1194. chronology. We derived ages of sequences conite grains. Sequences often yield a distinc- Late Cretaceous ages (Tables DR2A±D) from these plots and discuss them to one sig- tive gamma log signature with a hot zone at were assigned by using two new linear re- ni®cant decimal place (e.g., 71.2 Ma) to main- sequence boundaries, high values in glauco- gressions developed for upper Coniacian tain consistency within and between sites, nite sands, moderate values in silty clays, and through Maastrichtian sections (Fig. 2). We though resolution is typically no better than low values in the quartz sands (Fig. 4). Be- constructed this new reference section by us- Ϯ0.5 m.y. cause the TSTs are thin, the maximum- ing Sr isotope age data for sections in the U.S. We estimated eustatic amplitudes by using ¯ooding surfaces (MFS) are dif®cult to dif- Western Interior (McArthur et al., 1994), one-dimensional inverse models termed ferentiate from unconformities; both are often Kronsmoor, Germany (McArthur et al., 1992), ``backstripping'' (Watts and Steckler, 1979; associated with shell beds and gamma-ray and South Atlantic ODP Site 525 (Sugarman Bond and Kominz, 1984; Bond et al., 1989). peaks. Flooding surfaces, particularly MFSs, et al., 1995). Sr isotope ratios were plotted vs. Backstripping removes the effect of sediment may be differentiated from sequence bound- age to determine the evolution of seawater loading from observed basin subsidence. aries by the association of erosion and rip-up 87Sr/ 86Sr through time (Fig. 2). As in previous Backstripping studies have shown that simple clasts, age breaks discerned from Sr isotope studies that follow linear-regression tech- thermal subsidence, sediment loading, and stratigraphy and biostratigraphy, lithofacies niques developed for radiocarbon calibration compaction are the dominant causes of sub- successions, and benthic foraminiferal biofa- (Draper and Smith, 1981), 87Sr/ 86Sr ratios sidence in the New Jersey Coastal Plain cies. The transgressive surface (TS), marking measured at each depth are dependent vari- (Kominz et al., 1998; Van Sickel et al., 2003). the top of the LST, represents a change from ables and age estimates (based upon magne- By assuming thermal subsidence on a passive generally regressive to transgressive facies; tochronology at Site 525; based on biostratig- margin, the tectonic part of subsidence can be because LSTs are generally absent, these sur- raphy and radiometric dates in the Western removed and a eustatic estimate obtained. By faces are generally merged with the sequence Interior and Kronsmoor) are independent var- using forward modeling, Steckler (1981) dem- boundaries. Notable exceptions include the iables (Miller et al., 1991; Oslick et al., 1994). onstrated that accommodation in the New Jer- base of the Navesink Formation, seen in out- Sr isotope ratios were plotted on the abscissa, sey Coastal Plain is dominated by the ¯exural

Figure 6. Potomac and Bass River sequences in the Ancora and Bass River boreholes. Shown are core depth in feet; generalized lithologic description; downhole gamma log; core depth in meters; cumulative percent of clay-silt, ®ne sand, medium to coarse sand, glauconite, shells, and mica; biostratigraphic age control; sequence boundaries (wavy lines); and paleoenvironments. FSЯooding surface. Thick gray lines between sites indicate correlation of sequences. N

Geological Society of America Bulletin, March/April 2004 375 MILLER et al. Figure 7. Magothy sequences. Caption and legend as in Figure 6.

376 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN Figure 8. Santonian and middle to lower Campanian sequences. Caption and legend as in Figure 6.

Geological Society of America Bulletin, March/April 2004 377 MILLER et al. Figure 9. Upper Campanian sequences. Caption and legend as in Figure 6.

378 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN

response to sediment loading of the stretched crust seaward of the basement hinge zone (i.e., offshore beneath the modern shelf). Steckler (1981) also showed that coastal-plain subsi- dence is thermal in form beginning 15±20 m.y. after rifting. Our data set begins at 100 Ma, ϳ25 to 50 m.y. after subsidence began beneath the coastal plain (Olsson et al., 1988) and ϳ70 m.y. after rifting ceased (Klitgord et al., 1988); therefore, the subsidence generated by ¯exure in the coastal plain is thermal in form (Kominz et al., 1998; Van Sickel et al., 2003). Bond and Kominz (1984) showed that Airy backstripping of sediment loaded on rig- id lithosphere, such as beneath the New Jersey Coastal Plain, will exhibit a curvature that is identical to that of the true tectonic subsi- dence. Thus, the difference between observed subsidence and a best-®t theoretical thermal curve (termed R2 for second reduction; Bond and Kominz, 1984) is the result of either sea- level change or any subsidence unrelated to two-dimensional passive-margin subsidence (Kominz et al., 1998).

RESULTS

Rifting and thermal subsidence began off- shore in the Baltimore Canyon Trough be- tween 180 and 165 Ma, but deposition did not begin onshore until ca. 120±110 Ma when suf®cient plate rigidity was attained, initiating thermo¯exural subsidence in the coastal plain (Sheridan and Grow, 1988). The Potomac For- mation is the basal unit of the northern Atlan- tic Coastal Plain and comprises largely non- marine clays, silty clays, sands, and gravel (Glaser, 1969). This group is poorly exposed in outcrop in New Jersey and was only sam- pled in Leg 174AX boreholes at Ancora (Fig. 6). The Potomac Formation has been dated primarily by using pollen (generally pollen Zone III and older; lowermost Cenomanian± Albian; Doyle and Robbins, 1977). Age con- trol on the Potomac Formation is otherwise lacking except for marine intercalations at An- cora that yield nannofossil assignment to Zone CC9 (upper Albian to lower Cenomanian; de Romero in Miller et al., 1999b) and a pollen assignment to the pollen Zone III/IV transition (older than 97 Ma; lowermost Cenomanian; G.J. Brenner in Miller et al., 1999b). At least two apparent unconformities occur in the up- permost Potomac Formation at Ancora (Fig. 6), and there are numerous erosional surfaces found deeper in the Potomac Formation (Doyle and Robbins, 1977). The regional and global signi®cance of erosional surfaces (un- conformities vs. autocyclical shifts) in the

Figure 10. Maastrichtian sequences. Captiondominantly and legend as in Figure 6. nonmarine Potomac Formation is

Geological Society of America Bulletin, March/April 2004 379 MILLER et al.

generally uncertain. The Potomac Formation is overlain by a series of primarily marine Up- per Cretaceous units; we provide sea-level es- timates by analyzing the sequence stratigraphy and paleo±water depths for these units.

Sequences, Facies, Paleoenvironments

Bass River Sequences (Cenomanian±Early Turonian) Cenomanian to lower Turonian sequences in New Jersey are part of the Bass River For- mation (Petters, 1976). In the downdip Bass River borehole, this formation (1806.4 ft to total depth of 1956.5 ft [550.59±596.34 m]) consists of one sequence of gray, shelly, fos- siliferous, micaceous (chloritic), clayey silt and silty clay with occasional sandy silts, which becomes slightly sandy at the top (Fig. 6). The base of the sequence (Bass River III; see below) was not sampled at Bass River and the sequence shallows up-section from middle to inner neritic. Three middle Cenomanian±lower Turonian sequences (Bass River I, II, and III) are found at Ancora, with sequence boundaries at 1061.9, 1082.5, 1110.9, and 1148.1 ft (323.67, 329.95, 338.60, and 349.94 m; Fig. 3). The lower Bass River I sequence at Ancora has a basal neritic glauconite sand (TST), medial- prodelta silt and clay (lower HST), and upper- prodelta shelly clays and silts with thin sands (upper HST). The medial Bass River II se- quence consists of (1) neritic glauconitic clay to clayey glauconite sand (TST); (2) sandy, micaceous clay deposited in a prodelta envi- ronment (lower HST); and (3) a ®ne quartz sand that coarsens up-section and was depos- ited in a delta-front environment (upper HST). The upper Bass River III sequence consists of micaceous silty clay deposited in a prodelta environment; it is slightly glauconitic at the base, re¯ecting inter®ngering between open- shelf and prodelta environments, with mica and shells near the top. Lithostratigraphic and biostratigraphic correlations show that only the upper of the three sequences (Bass River III, uppermost Cenomanian±lower Turonian) is represented in the Bass River borehole (Fig. 6) and that the lower sequences were not pen- etrated at this site (i.e., the hole bottomed in the Bass River III). The Bass River sequences at Bass River and Ancora were deposited in inner- to middle-neritic paleodepths, although some Figure 11. Late Cretaceous age vs. depth plots for sequences at Bass River (top) and siltier, less shelly laminated intervals are in- /Ancora (bottom). Sr isotope ages are from Table DR2 with ؎1 m.y. error bars; open terpreted as prodelta and sands as delta front circles are diagenetically altered samples. Conservative age models indicated in thick black nearshore deposits. Benthic foraminifera in- line; preferred age models indicated in gray. Horizontal thin lines indicate sequence dicate that the Bass River sequences at Bass boundaries. CCÐnannofossil zonesÐopen triangles; XÐplanktonic foraminiferal age es- River and Ancora were deposited predomi- timates. Gray-shaded blocks at bottom indicate time represented in each borehole. Thin vertical dashed lines are drawn at 5 m.y. increments.

380 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN

nantly in inner-neritic paleodepths (Epistomina- biofacies), with deeper water (middle neritic) at ¯ooding surfaces (Fig. 6). A higher-order cyclicity is apparent in the benthic foraminif- eral biofacies that show several distinct para- sequences bounded by ¯ooding surfaces with- in the Bass River III sequence at Bass River (Fig. 6; Sugarman et al., 1999). Total water- depth variation within this sequence was rel- atively small (ϳ20 m). Nannofossils and planktonic foraminifera indicate that the Bass River sequences at Bass River and Ancora are Cenomanian to lower Turonian. The Cenomanian/Turonian bound- cernible time gap, and this is the best age estimate ary (CC10/CC11) at Bass River is placed at ϳ1935.5 ft (589.94 m; Sugarman et al., 1999). This placement is consistent with the highest occurrence of the planktonic foraminiferal ge- nus Rotalipora at 1945 ft (592.84 m) (Sugar- man et al., 1999) and an assignment to pollen 98.3 UZA2.3 95.5/99.5 NR 99.7

(Ma) new age Europe platform Frohlich (2002) Zone IV (upper Cenomanian to Turonian; G.J. Ͼ

Hiatus age EPR Seq. EPR old/ NW Russian Matthews and Brenner in Miller et al., 1998b). At Ancora, nannofossils indicate that the section below 1074.5 ft (327.51 m) is Cenomanian; the sec- 9.6 tion above this is barren of nannofossils. (m/m.y.) Average Sed. rate Planktonic foraminifers tentatively suggest that the section above 1074.5 ft (327.51 m) is the lower Turonian Whiteinella archeocreta- (m)

Thickness cea Zone. The Bass River II sequence is as- signed to Subzones CC10b and CC10a partim, whereas the Bass River I sequence is assigned (m) Ancora Depth to CC10a partim and CC9 (Fig. 6). Previous studies only recognized one

(ft) Cenomanian±lower Turonian sequence in Depth New Jersey in the coeval outcropping Raritan Formation (Owens and Gohn, 1985; Olsson, 1991). The Bass River I±III sequences are Age (Ma) equivalent to depositional sequence 2 of Owens and Gohn (1985) and KCE1 of Olsson (1991). 11.2 Average (m/m.y.) Sed. rate Magothy Sequences (Late Turonian± Coniacian) Two upper Turonian±Coniacian sequences (m) at Bass River are assigned to the Magothy Thickness Formation (Fig. 7). The lower sequence (Ma- gothy I; 1749.0±1806.4 ft [533.10±550.59 m]) (m) TABLE 1. AGES OF SEQUENCES AT BASS RIVER AND ANCORA COMPARED TO EPR AND EUROPEAN ESTIMATES Depth consists of (1) a basal micaceous, lignitic, silty clay deposited in a prodelta environment (TST

(ft) and lower HST), and (2) an upper lignitic silty Depth sand and ®ne sand deposited in a delta-front environment that coarsens up-section to a peb-

Age bly, moderately sorted sand deposited in a ¯u- (Ma) vial setting (upper HST). A sequence bound- ary occurs in an interval of no recovery between 1745 and 1749.0 ft (531.88±533.10 m). The upper sequence (Magothy III;

: NRÐnot resolved/recovered; alt.Ðalternate (conservative) age model (in gray in Fig. 11 where only one number is given for hiatus age, there is no dis 1709.5±1745 ft [521.06±531.88 m]) consists of (1) a kaolinitic white and red clay with red Note Sequence New JerseyPaleoceneNavesink II?Navesink INav I±II (alt.)Marshalltown 64.5±66Marshalltown (alt.)Englishtown 67.1±69 1270 64.5±69Englishtown 71.2±75.7 (alt.) BassMerchantville 387.2 River III? 1441 1295 1295Merchantville II? 1259 75.8±76.7 439.2 394.7 77.8±81Merchantville 394.7 I 383.8 1473 3.4Merch 83.1±83.7 I±III (alt.) 1653Cheesequake 44.5 449.0 1674 10.9 7.5 83.7±83.9Magothy III 503.8 76.9±83.9 510.4Magothy 1683 II 2.3 84.3±85.2 1683Magothy 9.8 I 513.2 54.8 9.7Bass 1709 2.4 River 513.2 3.7 III 6.6 64.5±66Bass 86.7±87.8 River 521.1 IIBass River 10.9 2.8 64.5±68.7 1749 I 64.2 73±75 17.1Potomac 634 67±68.7 92.1±93.5 NR 533.2 13.1 651.3 7.9 90±91.4 1950 651.3 77.8±80.8 193.3 757.4 198.6 76±76.7 14.0 1806 83.2±83.5 198.6 NR 12.1 594.5 230.9 9.2 909 792.3 550.7 NR 11.8 928 83.5±83.8 6.6 241.6 8.8 77.8±83.8 277.1 43.8 32.3 5.3 NR 945.3 17.5 11.0 282.9 945.3 84.3±85.2 10.6 288.2 35.6 2.8 288.2 612.5 4.4 86.7±87.8 16.2 31.3 957.4 71.8±76 5.8 186.7 3.1 12.5 291.9 987 5.3 15.2 46.6 92.8±93.5 11.5 69±71.2 66±67 757.4 75±76 1083 300.9 76±77.8 69±73 19.3 230.9 76.7±77.8 3.7 NR 81.0±83.1 17.6 792.3 330.0 UZA4.4 7.8 UZA4.3 TA1.1 9.0 UZA4.5 32.3 UZA4.1 241.6 83.5 77.5/78.5 4.1 84±84.3 83.9±84.3 75/75 6.1 68/67 71/70.5 79/80 10.6 UZA3.4 75.5±79 71.2±74 NR 8.2 7.7 UZA3.5 85.2±86.7 66±67 88.3±89.8 7.6 NR UZA3.3 85/84 80/82.5 87.8±88.3 1063 5.9 NR NR NR 87.5/85.5 NR 93.5±94.6 76 UZA3.2 323.9 82±83 NR 1109 94.6±95.8 93.5 NR 85±85.5 88.5/87.5 77.8 1148 75.54 70.69 23.0 338.1 84 85.5 77.15 97±98.3 ?UZA2.6 67.45 NR 350.0 NR 79.98 8.1 91/92 15.3 11.9 86.85 87.5 84.83 82 89.8±90 NR 7.3 9.9 91.4±92.1 UZA3.1 88.87 UZA2.7 95.8±97 NR 94.6 90/90.5 90.5/91 UZA2.4 NR UZA2.5 93.72 NR 94/97 93/95 89.8 92 NR NR 96.6 91.7 NR 95.74 of the unconformity). sandstone pebbles that was deposited in a del-

Geological Society of America Bulletin, March/April 2004 381 MILLER et al. ta plain/inter¯uvial environment; (2) an over- relation does not agree with pollen data (Fig. tween the Cheesequake and Merchantville lying coarser unit, consisting of well-sorted, 7). On the basis of the pollen data, we tenta- Formations. Pollen (pollen Zone VI at 955.1 lignitic interbedded sandy silty clay and clay- tively suggest that there are three Magothy se- ft [291.11 m]; G.J. Brenner, 2002, personal ey very ®ne sand that coarsens up-section to quences; Magothy I (pollen Zone IV, Turoni- commun.) provides the only age-diagnostic medium to ®ne quartz sand; these sands were an) and Magothy III (pollen Zone V, upper fossils for the Cheesequake sequence at An- deposited in a delta-front environment; and (3) Coniacian) are represented at Bass River, cora, which is correlated with the Cheese- micaceous laminated clay and interbedded and Magothy II (pollen Zone V, uppermost quake sequence at Bass River on the basis of clay and clayey, pebbly medium to coarse Turonian±lower Coniacian) and Magothy III stratigraphic position and lithologic similarity. quartz sand deposited in marginal-marine en- are represented at Ancora (Fig. 7). The Cheesequake Formation was described vironments (as indicated by the presence of The Magothy Formation in outcrop consists in outcrop by Litwin et al. (1993) and for- marine nannofossils). This retrogradational of sands and silty clays deposited in complex mally named by Owens et al. (1998) as a succession is unusual for coastal-plain se- nonmarine to marginal-marine environments slightly glauconitic clayey silt assigned to quences because they otherwise invariably (Owens and Gohn, 1985). Regionally, this Santonian to lowermost Campanian pollen shallow up-section; the Magothy III situation unit thickens to over 300 m toward the Long Zone VI, similar to Ancora. Owens et al. suggests only the TST is represented and the Island platform and thins toward Delaware (1998) noted that the Cheesequake Formation HST is truncated. where it all but disappears (Perry et al., 1975; in the Toms River and Freehold boreholes Two upper Turonian±Coniacian sequences Olsson et al., 1988). Previous chronostrati- (Fig. 1) is late Santonian (nannoplankton (Magothy II and III) are also identi®ed at An- graphic correlations of the Magothy Forma- Zones CC16±CC17) at the base to earliest cora (Fig. 7). The lower sequence, Magothy II tion have proved to be challenging. Olsson Campanian at the top, spanning the Santonian/ (990±1062.5 ft [301.75±323.85 m]), is a mas- (1991) noted a Coniacian KCE1 sequence in Campanian boundary within the formation. In sive, lignitic, medium to coarse sand with a offshore wells, but was not able to date the contrast, we note it to be slightly older few silty clay interbeds within ®ning-upward Magothy Formation onshore as equivalent. (CC15±CC16) at Bass River. It is thus uncer- successions. The lower sequence boundary Owens and Gohn (1985) identi®ed the Ma- tain whether the sequences identi®ed at the (1062.5 ft [323.85 m]) is a sharp contact be- gothy Formation as their depositional se- Bass River and Ancora boreholes correlate tween coarse sediments of the Magothy above quence 3, though they correlated it as Conia- with the Cheesequake Formation previously Bass River Formation clays, whereas the up- cian, Santonian, and lower Campanian; our identi®ed by Owens et al. (1998) and Litwin per sequence boundary occurs in a coring gap studies establish the Magothy as older than the et al. (1993). Nevertheless, the data from Bass (985±990 ft [300.23±301.75 m]), and the se- Santonian, and we assign it to the upper River and Ancora establish that there is one quence boundary is placed at the gamma log Turonian±Coniacian. lower to middle Santonian sequence downdip, in¯ection (987 ft [300.84 m]; Figs. 3 and 7). and it is likely that this is equivalent to the The environment of the Magothy II could be Cheesequake Sequence (Santonian) poorly dated Cheesequake Formation updip. ¯uvial or marginal marine (e.g., tidal delta); A distinct sequence at Bass River and An- the paucity of ®ne sediments and extensive cora lies between unconformities at the top of Merchantville Sequences (Latest burrowing suggests deposition in a nearshore- the Magothy and the base of the Merchantville Santonian±Middle Campanian) marine setting. The upper sequence, Magothy Formations. This sequence is tentatively cor- A thick lower Campanian sequence in out- III (957.4±987 ft [291.82±300.84 m]), consists related with the Cheesequake Formation and crop consists of glauconite sand at the base of (1) interbedded slightly sandy silty clay and sequence in outcrop. At Bass River, the (Merchantville Formation), a medial, very lignitic sand deposited in a delta-front or es- Cheesequake sequence (1683.2±1709.2 ft thick micaceous clay (Woodbury Formation), tuarine setting that comprises the TST and (2) [513.04±520.96 m]) consists of glauconite and an upper clayey sand (lower Englishtown poorly sorted coarse to very coarse quartz clay (TST) that coarsens up-section to clayey Formation). At Ancora (Fig. 8), clayey glau- sand deposited in a ¯uvial environment (HST; glauconite-quartz sand (HST). At Ancora, the conite sands to glauconitic (40%±60%) clays Fig. 7). Cheesequake Formation and sequence (957.4± assigned to the Merchantville Formation ap- The Magothy sequences at Ancora are both 945.3 ft [291.82±288.13 m]) consists of slight- pear above a distinct, irregular sequence assigned to upper Turonian±Santonian pollen ly micaceous, slightly glauconitic, clayey, boundary at 945.3 ft (288.13 m; Fig. 3B). Zone V of Christopher (1982), correlating silty quartz sand (TST) that coarsens up- Carbonate-rich glauconitic clay near the top of with the upper Magothy sequence at Bass Riv- section above an MFS (953.2 ft [290.54 m]) the Merchantville Formation represents the er (pollen Zone V) and parts of the Magothy to lignitic, silty, ®ne to medium, slightly peb- MFS (Fig. 8; 904.4 ft [275.66 m]). The con- Formation in outcrop (pollen Zone V; Chris- bly quartz sand (HST). tact with the glauconitic clays of the lower- topher, 1982). The upper Magothy sequence At Ancora, the Cheesequake sequence was most Woodbury is gradational; glauconite de- (III) at Bass River is tentatively assigned to deposited in inner-neritic environments; clay, creases up-section. The thick (ϳ106 ft [32.31 nannofossil Zone CC14 (late Coniacian to ear- mica, and glauconite decrease up-section, in- m]) Woodbury Formation consists of laminat- ly Santonian) and pollen Zone V. The lower dicating shallowing up-section. At Bass River, ed to slightly burrowed, very micaceous, lig- Magothy sequence at Bass River is assigned a glauconite clay at the base of the unit is nitic, slightly shelly, very dark gray clay with to pollen Zone IV (upper Cenomanian to Tu- interpreted as a thin TST; mixed glauconite occasional pyrite nodules. An indurated zone ronian), suggesting that it is older than the se- and quartz at the top result from reworking of between 797.6 and 797.7 ft (243.11 and quences at Ancora or in outcrop; it may ac- glauconite in the HST. 243.14 m) marks the contact with the shelly, tually correlate with the Raritan Formation in The Cheesequake at the Bass River bore- micaceous, glauconitic quartzose silty sands outcrop. The simplest interpretation is to cor- hole is Santonian (calcareous nannoplankton of the lower Englishtown Formation. A se- relate the two Magothy sequences at Ancora Zones CC15 and CC16) with a late Santonian quence boundary at 792.3 ft (241.49 m; Fig. with the two at Bass River; however, this cor- hiatus associated with the unconformity be- 8) separates the lower Englishtown Formation

382 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN from glauconitic silty clay above (upper En- 11). However, assuming that the base of Zone [241.49 m]; Fig. 3), glauconitic silty clay de- glishtown sequence). CC19 is correct (Fig. 11), then Merchantville posited in inner middle-neritic paleodepths At Bass River (Fig. 8), the Merchantville sedimentation rates would have been excru- continues up-section to an MFS (789.5 ft Formation appears above a disconformity at ciatingly slow (0.25 cm/k.y. at Bass River and [240.64 m]) associated with a minor gamma- 1683.2 ft (513.04 m) and consists of biotur- 0.35 cm/k.y. at Ancora, under the assumption ray peak (Fig. 8). Here, there is an abrupt fa- bated, shelly, lower glauconitic clays and of a duration of 3.5 m.y. for these three cies change to shelly, slightly sandy, mica- clayey glauconite sands (up to 70% glauco- zones). This unreasonably slow rate prompted ceous, laminated silts deposited in a prodelta nite) and upper glauconitic foraminiferal us to reexamine the Merchantville glauconite environment. These grade up-section to un- clays. Peak abundances of foraminifers mark sand. consolidated, poorly sorted, lignitic, ®ne to an MFS (1660 ft [505.97 m]). The thick An alternative interpretation of the bio- medium quartz and (reworked) glauconite (ϳ166 ft [50.60 m]) Woodbury Formation ap- stratigraphic, limited Sr isotope, and gamma sand deposited in a nearshore environment. pears above this MFS and consists of mica- log data suggests that the Merchantville glau- The upper sequence boundary is at 757.2 ft ceous, fossiliferous, bioturbated, lignitic silty conite sand consists of two sequences and the (230.79 m). clay to clay that becomes slightly sandy up- TST of a third sequence. An unconformity is At Bass River, shelly glauconite sand and section. The contact of the Woodbury For- tentatively placed at a sharp gamma log in- overlying burrowed glauconitic clays depos- mation with the overlying lower Englishtown crease at 928 ft (282.85 m; within Zone ited in middle-neritic paleodepths comprise Formation is gradational. The lower English- CC17) at Ancora. The lithologic expression of the TST between the sequence boundary town Formation (1490±1472.6 ft [454.15± this sequence boundary is reasonably clear: (1472.6 ft [448.85 m]) and the MFS (1467.4 448.85 m]) consists of micaceous, very lig- black, clayey, glauconite sand overlies an in- ft [447.26 m]). Glauconite progressively de- nitic, cross-bedded, silty ®ne sand. The upper terval of yellowish-green glauconite sand that creases up-section above the MFS, where the contact separating the lower and upper En- comprises the HST of the underlying sequenc- section consists of sandy, clayey silt and silty glishtown Formation is a sequence boundary es. A similar gamma increase at Bass River at clay deposited in a prodelta environment (low- (1472.6 ft [448.85 m]; Fig. 8). 1674 ft (510.24 m; between Zones CC16 and er HST) at inner-neritic paleodepths. Clay de- Overall the Merchantville±Woodbury± CC17) appears to correlate well with the 928 creases up-section to very micaceous, slightly lower Englishtown section comprises a ft (282.85 m) gamma-ray peak at Ancora. A glauconitic, bioturbated, in some places transgressive-regressive succession, though second unconformity within the Merchantville shelly, clayey ®ne sand deposited in a delta- we tentatively break the Merchantville For- Formation at Ancora is tentatively placed at front environment (upper HST). mation into two additional sequences (see be- ϳ909 ft (277.06 m) at a sharp gamma log kick A sample near the top of the Englishtown low). The Merchantville Formation was de- within Zone CC18. The lithologic expression Formation (758.5 ft [231.19 m]) at Ancora is posited in middle- to outer-neritic paleodepths of this sequence boundary is subtle: slightly tentatively assigned to Zone CC20 on the ba- and represents the TST; carbonate-rich glau- clayey, black glauconite (Ͼ50%) sand overlies sis of the occurrence of Ceratolithoides acu- conitic clay near the top represents the MFS. a sulfur-rich interval, with more clay-rich, leus. On the basis of the absence of Cerato- Benthic foraminifers indicate that the Wood- slightly quartzose glauconite sand below. A lithoides aculeus, the interval from 763.5 to bury Formation was deposited primarily in in- very similar gamma log pattern occurs at 792.2 ft (232.71±241.46 m) was tentatively ner middle-neritic paleodepths (ϳ40±80 m); 1652.5 ft (503.68 m) at Bass River also within assigned to Zone CC19 (Miller et al., 1998b), the laminated, lignitic, very micaceous clays Zone CC18, and we are reasonably con®dent though the absence of this taxon may be due indicate a prodelta environment. Clean, cross- that this is a regional unconformity contained to scarce and poorly preserved nannofossils. bedded sands of the lower Englishtown For- within the cryptic Merchantville glauconite The assignment to Zone CC19 is contradicted mation were deposited in inner-neritic paleo- sands. These sequences are dif®cult to recog- by three Sr isotope age estimates that suggest depths (Ͻ30 m). The mixture of glauconite nize by using lithology alone because they correlation to Biochron CC20 (see below). At and quartz sand in the upper Englishtown For- consist of TST glauconite overlying HST re- Bass River, neither nannofossils nor plankton- mation (Fig. 8) is typical of reworking of worked glauconite sands. ic foraminifera yield diagnostic zonal assign- glauconite in HSTs. Thus, we interpret that the thick (40.9 ft ments from this sequence; the underlying se- The Merchantville sequence(s) are upper- [12.47 m] at Ancora; 29.2 ft [8.90 m] at Bass quence is assigned to undifferentiated Zones most Santonian to lower Campanian. This se- River) Merchantville glauconites are com- CC19 to CC20 and the overlying sequence to quence has been widely recognized both in posed of two thin sequences (Merchantville I CC21a. New Jersey (the KC1/Merchantville± and II) as well as being the transgressive sys- In outcrop, the Englishtown Formation con- Woodbury±Englishtown sequence of Olsson tems tract of the thick Merchantville± sists of a lignitic, slightly glauconitic, cross- (1991) and elsewhere in the Atlantic Coastal Woodbury±lower Englishtown (ϭ Merchant- bedded sand deposited in a delta-front envi- Plain (depositional sequence 4 of Owens and ville III sequence; Fig. 8). Further studies at ronment (Owens and Sohl, 1969; Owens et al., Gohn, 1985). The Woodbury and lower En- additional locations are needed to test our in- 1998). Owens et al. (1998) informally divided glishtown Formations are assigned to calcar- terpretation of three sequence boundaries as- the subsurface Englishtown Formation into eous nannoplankton Zone CC19 at Ancora sociated with the Merchantville Formation. upper and lower lithologic successions; their and Zone CC19 to CC20 undifferentiated at lower Englishtown is the upper HST of the Bass River. The Merchantville Formation is Englishtown Sequence (Middle Campanian) Merchantville III sequence at Bass River and assigned to calcareous nannoplankton Zones The upper Englishtown Formation at An- Ancora, whereas their upper Englishtown CC16, CC17, and CC18 at both holes. It was cora and Bass River is a sequence that consists forms a complete sequence (Fig. 8) at both originally assumed that the Merchantville For- of a lower glauconite sand or glauconitic clay, sites. The relative thinness of the upper En- mation represented a concatenation of these medial silt, and upper sand (Fig. 8). Above a glishtown sequence (35.1 ft [10.70 m] at An- zones (e.g., the conservative age model, Fig. sequence boundary at Ancora (792.3 ft cora; 32.1 ft [9.78 m] at Bass River) has rel-

Geological Society of America Bulletin, March/April 2004 383 MILLER et al. egated it to obscurity (e.g., it was not sediments, yielding a hot gamma-ray signature At Bass River, the Navesink/Mount Laurel differentiated from the lower Englishtown by for these sands. The sequence boundary with unconformity (1294.5 ft [394.56 m]) is over- Olsson [1991] or Owens and Gohn [1985]), the overlying Navesink Formation (1294.5 ft lain by a 0.3-m-thick contact zone, with re- though it is certainly a distinct and important [394.56 m]; Fig. 3) is a disconformity with worked Mount Laurel sands and phosphorite middle Campanian sequence. It was mapped extensive reworking. pellets mixed with Navesink clayey glauconite as the Kwc2 cycle of Owens et al. (1998) and Benthic foraminifers indicate deposition of sand. Quartzose, slightly clayey glauconite Ks3 of Gohn (1992). the Marshalltown and Wenonah Formations in sand at the base of the sequence becomes middle-neritic paleodepths and the Mount more clay rich and less quartzose up-section Marshalltown Sequence (Late Campanian) Laurel Formation in inner middle-neritic pa- to an MFS associated with a gamma log in- The Marshalltown is a thick upper Cam- leodepths (Fig. 9). Detailed benthic foraminif- crease (1288 ft [392.58 m]) and a change to panian sequence consisting of glauconite sand eral paleobathymetric estimates from Bass glauconitic, carbonate-rich, clay. Carbonate at the base (Marshalltown Formation), a me- River (Fig. 9; Skinner, 2001) show only ϳ35 decreases and clayey, fossiliferous, very bio- dial silty clay (Wenonah Formation), and an m of shallowing within the Marshalltown se- turbated glauconite sands continue to a gra- upper ®ne to coarse quartz sand (Mount Lau- quence at Bass River (from ϳ75minthe dational contact, marking the top of the Na- rel Formation) at the top. At Ancora, the En- Marshalltown to 60 m in the Wenonah and vesink Formation. The overlying New Egypt glishtown/Marshalltown contact is a distinct lower Mount Laurel, to 40 m in the upper Formation consists of brownish-gray shelly sequence boundary (Figs. 3B and 9; 757.2 ft Mount Laurel). This modest shallowing re- clay that continues to a 6-cm-thick spherule [230.79 m]) with a shelly, micaceous glauco- sulted in distinct facies changes. layer, the base of which marks the K/T bound- nitic clay above (Fig. 3B). We interpret a sur- The Marshalltown±Wenonah±Mount Laurel ary (1260.4 ft [384.17 m]; Olsson et al., face and sharp facies change from glauconitic sequence has been widely recognized both in 1997). Above the spherule layer, a return to clay below to glauconite sand above (752.7 ft New Jersey (the KC2 sequence of Olsson, glauconitic clay or clayey glauconite sand [229.42 m]) as a transgressive surface (TS) 1991) and elsewhere in the Atlantic Coastal marks the Paleocene Hornerstown Formation. and the basal Marshalltown Formation as an Plain (depositional sequence 5 of Owens and The extremely slow sedimentation rates for LST; this is one of the few lowstand deposits Gohn, 1985). In other New Jersey boreholes the Maastrichtian (Table 1) Navesink se- recognized in the coastal plain (see Discus- it has been dated as late Campanian by using quence (Ͻ0.3 cm/k.y. at both sites) prompted sion). Burrowed clayey glauconite sands typ- calcareous nannofossils (CC20 to CC21 [un- us to reconsider the continuity of this section. ify the Marshalltown Formation; a peak in differentiated] in the Marshalltown CC22b in We tentatively identify an additional sequence carbonate represents the MFS (746 ft [227.38 the Mount Laurel) and Sr isotope stratigraphy boundary within the Maastrichtian section at m]). The contact between the Wenonah and (Sugarman et al., 1995). At Bass River, the both Ancora and Bass River. At Ancora, we the Marshalltown Formations is gradational. Marshalltown Formation is assigned to Zone tentatively place a sequence boundary (634 ft The Wenonah Formation consists of slightly CC21a, the Wenonah Formation to Zones [193.24 m]) just above a peak in clay and be- glauconitic, micaceous, shelly, woody, clayey CC21 to CC22 [undifferentiated], and the low a maximum in glauconite expressed as an silty sand to sandy clayey prodelta silts, with Mount Laurel Formation to Zones CC21 to increase in downhole gamma-ray values (Fig. decreasing glauconite up-section (Fig. 9). The CC22 [undifferentiated] at the base to CC23a 10); this sequence boundary separates the Na- contact with the Mount Laurel Formation is at the top. Previous assignments of this se- vesink I from the overlying Navesink II se- also gradational, with silty clay and mica de- quence to the early Maastrichtian (Olsson, quence. At Bass River, we tentatively place a creasing up-section in the transitional interval. 1991) resulted from differences in time scales. sequence boundary at the Navesink/New The Mount Laurel Formation generally coars- This sequence is late Campanian according to Egypt Formation contact (1270 ft [387.10 m]). ens up-section from glauconitic silty ®ne sand the Gradstein et al. (1994) time scale; the hi- These inferred sequence boundaries require to a slightly clayey, glauconitic (ϳ20%) ®ne atus associated with the Navesink/Mount Lau- veri®cation from other locations. to medium quartz sand, with common ovoid rel contact spans the Campanian/Maastrichtian The Maastrichtian Navesink sequence(s) at phosphate grains near the top of the formation. boundary. both boreholes were deposited primarily in A major disconformity caps the sequence middle-neritic paleodepths. Maximum paleo- (651.3 ft [198.52 m]; Figs. 3B and 9). Navesink Sequence(s) (Maastrichtian) depths were ϳ80 m at Bass River and ϳ60 m At Bass River, a sequence boundary occurs The Maastrichtian at Ancora and Bass Riv- at Ancora; the deposits sampled in both bore- at the base of the Marshalltown Formation er consists of at least one, and possibly two, holes shallow up-section to paleodepths of (1440.5 ft [439.06 m]; Fig. 9). The Marshall- sequences, and deposition was continuous ϳ30±50 m (Olsson et al., 2002). There was town Formation at Bass River comprises fos- across the Cretaceous/Tertiary (K/T) bound- little paleodepth change across the K/T bound- siliferous, clayey, shelly, glauconite sands and ary. At Ancora, a sequence boundary (651.3 ary; only general shallowing occurred in the glauconitic clays. A peak in foraminiferal ft [198.52 m]; Fig. 3B) associated with a layer last 0.5 m.y. of the Cretaceous and the ®rst abundances represents the MFS (1430 ft of phosphate pebbles separates carbonate-rich, 0.5 m.y. of the Tertiary (Olsson et al., 2002). [435.86 m]). Glauconite decreases gradually foraminiferal glauconitic clay of the Navesink The Navesink I sequence is dated as early up-section above this level, and the section Formation from underlying quartz sand of the to middle Maastrichtian. At Ancora and Bass ®nes to micaceous silts of the Wenonah For- Mount Laurel Formation. Carbonate content River, it is assigned to Zone CC25 and lower mation. The contact with the overlying Mount increases in the lower Navesink, peaking at Zone CC26. The Navesink II at Bass River is Laurel Formation is also gradational; sandy the MFS (647 ft [197.21 m]) and decreasing assignable to the uppermost Maastrichtian silt, ®ne sand, and medium to coarse quartz up-section. Quartz sand decreases up-section (calcareous nannofossil M. prinsii Subzone of sand become successively dominant up- to the MFS, suggesting deepening up-section CC26; the last 0.5 m.y. of the Cretaceous) to section. Phosphate grains are common at the in the TST (i.e., no LST is preserved, unlike lowermost Danian (Zones P0, P␣, P1a, and top of the formation, similar to the Ancora the situation present in outcrop; see below). P1b partim). The Navesink II continues across

384 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN the K/T boundary into the lower Paleocene below ϳ1470 ft (ϳ450 m) at Bass River (Fig. ly, at Ancora (Fig. 13, Table 1). The ages of where there is an unconformity: Zone P1b is 11). Nannofossil biostratigraphy constrains the Magothy II and Magothy III sequence not represented at Bass River, and a break oc- the age of the Magothy III sequence to the late boundaries are estimated as 90±89.8 and curs within this zone at Ancora. Coniacian. At both sites, the upper Turonian± 88.3±87.8 Ma, respectively. The Magothy I, The basal Navesink unconformity is found Santonian nonmarine Magothy I and II se- Magothy II, and Magothy III sequences ap- throughout the eastern United States at the quences are dated with pollen biostratigraphy pear to correlate with the UZA 2.7, UZA3.1, base of Owens and Gohn's (1985) deposition- to the stage level. Moderate (Ϯ1 m.y.) reso- and UZA3.2 sequences of EPR (Haq et al., al sequence 6. The unconformity in the out- lution is provided by biostratigraphy alone for 1987), respectively. crop on Route 34 in Matawan, New Jersey, is the Cenomanian±lower Turonian sections at The lower to middle Santonian Cheese- particularly interesting (Fig. 3A): (1) the un- both sites. The ages of the sequences at Bass quake sequence is separated from the Mago- derlying Mount Laurel Formation is a well- River agree with Ancora within better than 1 thy III sequence by an unconformity and a 1.5 sorted medium sand with Ophiomorpha and m.y. (Fig. 11). m.y. hiatus (86.7±85.2 Ma). It is separated Asterosoma burrows and tabular planar cross- The break between the lower Cenomanian from the overlying Merchantville sequences bedding, indicating onshore-offshore current upper Potomac Formation (pollen Zone III/IV by a short hiatus (84.3±83.9 Ma). The Cheese- directions on the lower shoreface (Martino transition and nannofossil Zone CC9) and the quake sequence appears to correlate with the and Curran, 1990); (2) the sequence boundary Cenomanian±lower Turonian Bass River se- UZA3.3 sequence of Haq et al. (1987). is marked by a distinct erosional surface and quence I is a regional unconformity (Owens The upper Santonian Merchantville I (84.3± a phosphorite layer that is commonly iron ce- and Gohn, 1985) with a hiatus of at least ϳ1 84.0 Ma), uppermost Santonian to lower Cam- mented at its base (Miller et al., 1999b); (3) a m.y. in duration (97.0±95.8 Ma). This se- panian Merchantville II (84.0±83.5 Ma), and heterolithic lag unit consists of sand pods of quence boundary correlates with the major lower Campanian Merchantville III (81.0± reworked Mount Laurel sands (some of which middle Cenomanian sea-level lowering of 77.8 Ma) sequences had moderate to high sed- is cemented into calcarenitic clasts), rip-up Gale et al. (2002) and the UAZ2.4/2.3 se- imentation rates (13±17 m/m.y. at Bass River, clasts, and shelf silts; this unit shows a re- quence boundary of EPR (Haq et al., 1987). 12±19 m/m.y. at Ancora). There is no discern- gressive facies pattern up-section and is inter- Three middle to late Cenomanian (Bass ible hiatus between the Merchantville I and II preted as an LST [Miller et al., 1999b]); (4) a River I±III) sequences have been identi®ed at sequences (83.5 Ma) and an ϳ2 m.y. hiatus cemented erosional surface at the top of the Ancora; the basal sequence boundaries for (83.1±81 Ma) between the Merchantville II lowstand section is interpreted as a TS (Miller Bass River II and III are dated as 94.6 and and Merchantville III sequences. The Mer- et al., 1999b); and (5) a clayey glauconite sand 93.5 Ma, respectively (hiatuses are not dis- chantville I, II, and III sequences appear to (the Navesink Formation) deposited in ϳ60 m cernible). The Bass River I and II sequences correlate with the UZA3.4, UZA3.5, and paleodepth (Olsson, 1991). are below the TD at Bass River, but three of UZA4.1 sequences of EPR (Haq et al., 1987), In outcrop, the Navesink Formation is over- us (Sugarman, Miller, and Browning) have respectively. lain by silts and sands of the Red Bank For- identi®ed these sequences at the borehole The age of the middle Campanian English- mation (HST; Sugarman et al., 1995). In the drilled in 2002 at Millville, New Jersey, sug- town sequence is poorly constrained (ca. subsurface, the HST sands largely disappear, gesting that they are regional in extent. Sedi- 76.8±76 Ma). The hiatuses between the En- and the Maastrichtian section is dominated in mentation rates for the middle Cenomanian to glishtown and underlying Merchantville updip locations such as Ancora by clayey lower Turonian sequences at Ancora (7±10 m/ (77.8±76.7 Ma) and overlying Marshalltown glauconite sands assigned to the Navesink m.y.) are about average for the Late Creta- sequences (76±75 Ma) are poorly resolved. Formation (Fig. 10). Farther downdip at Bass ceous (9.6 and 11.2 m/m.y. at Ancora and The Englishtown sequence appears to corre- River, the Navesink lithology is overlain by Bass River, respectively). However, the Bass late with the UZA4.3 sequence of EPR (Haq glauconitic clay assigned to the New Egypt River III sequence is expanded at Bass River et al., 1987); we do not see evidence for the Formation (Olsson, 1960). where sedimentation rates exceeded 31 m/ UZA4.2 sequence. m.y., the most rapid of any Upper Cretaceous The upper Campanian Marshalltown se- Chronology and Sedimentation Rates sequence in New Jersey (Table 1; Fig. 12). quence (ϳ75.7±71.2 Ma) had high sedimen- The Bass River I, II, and III sequences cor- tation rates at Ancora (16.2 m/m.y.) and av- Integration of Sr isotope stratigraphy and relate with the UZA2.4, UZA2.5, and UZA2.6 erage sedimentation rates at Bass River (9.7 biostratigraphy on age vs. depth diagrams sequences of EPR (Haq et al., 1987), m/m.y.). The Marshalltown sequence appears (e.g., Fig. 11) provides age resolution of about respectively. to correlate with the UZA4.4 sequence of Haq Ϯ0.5 m.y. for the middle Campanian to ear- A major middle Turonian sequence bound- et al. (1987). A major hiatus (ϳ2.2 m.y.) sep- liest Tertiary (ca. 80±64.5 Ma). The chronol- ary (hiatus 92.1±91.4 Ma) separates the Ma- arates the Marshalltown sequence from the ogy is less certain for the Santonian±early gothy I sequence from the lower Turonian Maastrichtian Navesink I sequence (69±67 Campanian (ca. 85.7±80 Ma), as illustrated by Bass River III sequence. The upper Turonian± Ma). The Navesink I sequence appears to cor- one of two possible age models for this inter- Coniacian Magothy Formation appears to rep- relate with the UZA4.5 sequence of EPR (Haq val at both sites (Fig. 11). We prefer the less resent three sequences, though differentiation et al., 1987). continuous age models for each site (gray of the Magothy II sequence at Ancora from The inferred sequence boundary between lines, Fig. 11; Table 1) to the more conser- the Magothy I at Bass River relies solely on the Navesink I and II sequences may be as- vative age models that assume no hiatuses pollen correlations (pollen Zone IV for the sociated with an ϳ1 m.y. late Maastrichtian across major unconformities. Biostratigraphy former, V for the latter). Limited data indicate hiatus (67±66 Ma). This hiatus requires veri- alone provides age control on the early Cam- that sedimentation rates were at least 12 m/ ®cation; it may correlate with the TA1.1/ panian and older sequences at Bass River be- m.y. for the Magothy I and III and 15 and 8 UZA4.6 sequence boundary of Haq et al. cause of diagenetic alteration of Sr isotopes m/m.y. for the Magothy II and III, respective- (1987). No matter which age model is used

Geological Society of America Bulletin, March/April 2004 385 MILLER et al.

386 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN

(continuous or discontinuous; Fig. 11), it is in¯uenced shoreline model for Upper Creta- to the northeast. In the case of the Magothy I clear that there was a dramatic drop in sedi- ceous facies styles in the New Jersey Coastal sequence, the high supply of sand and low sea mentation rates between the Campanian and Plain developed by Owens and Sohl (1969), level overwhelms the sequence-stratigraphic the Maastrichtian (Fig. 13). Owens and Gohn (1985), and Sugarman et al. signature. The Magothy II and III sequences At both sites, deposition was continuous (1995) is valid in these downdip locations, show a predicted pattern of upper HST sands. across the Cretaceous/Tertiary (K/T) bound- and the systems tracts of Posamentier et al. This difference suggests peak delivery of sand ary, within sequence Navesink II (66±64.5 (1988) are applicable to the coastal plain both to the Ancora±Bass River transect during the Ma). A sequence boundary is associated with updip (Sugarman et al., 1995) and downdip as late Turonian deposition of the Magothy I se- an early Danian hiatus (Biochron P1b; 64.5± shown here. HSTs are sandy and thick, where- quence and reduced supply thereafter. Sand 63.0 Ma). as TSTs are thin and generally composed of supply also increased in the late Campanian The sequence boundaries at the bases of the glauconite in the deeper marine sequences. during deposition of the Mount Laurel For- Bass River I, Magothy I, Magothy III, Cheese- Lowstand deposits are rare in the coastal mation. Comparisons with outcrop and other quake, Merchantville I, upper Englishtown, plain; Upper Cretaceous LSTs are only pre- subsurface data suggest multiple sources of Marshalltown, and Navesink sequences are re- served in the Marshalltown sequences at An- sand during deposition of the Mount Laurel gional in extent, occurring not only in both cora and the outcrop of the Navesink Forma- sequence because the formation shows a dis- boreholes (Fig. 11), but also in other New Jer- tion in Matawan (Miller et al., 1999b). tinct along-strike pattern of thickening and sey sites (these are equivalent to the eight se- Lowstand deposits are not expected landward thinning of sands (Owens et al., 1970; Martino quences of Olsson, 1991) and throughout the of prograding clinoform in¯ection points ex- and Curran, 1990). Atlantic Coastal Plain (e.g., Owens and Gohn, cept in incised valleys. Available seismic pro- Comparison of the Ancora and Bass River 1985). The Bass River I, Bass River II, Mer- ®les have not revealed a clinoform geometry sequences provides insights into updip vs. chantville II, and Merchantville III sequence for the Upper Cretaceous section of New Jer- downdip patterns of sedimentation. The An- also appear to be regional, whereas the re- sey, though progradation can be inferred by cora and Bass River cores constitute a dip pro- gional signi®cance of the Magothy II and Na- comparing the timing of appearance of medi- ®le (Fig. 1). These sites are 33 km apart, vesink II sequences is uncertain. um to coarse sands within the Mount Laurel which translates into ϳ66 m of paleodepth Sedimentation rates were high during at Formation at Ancora (74 Ma) to the downdip variation under the assumption of a gradient least three periods during which the Magothy occurrence at Bass River (72.5 Ma). Thus, we of 1:500 (Steckler et al., 1999). Every Creta- II (ca. 89±88 Ma), Merchantville/Woodbury infer that the preservation of LSTs in the Up- ceous sequence found at the downdip Bass (ca. 81±78 Ma), and Mount Laurel (ca. 75±72 per Cretaceous section was restricted to in- River site is also represented at the updip An- Ma) units were deposited (Fig. 13). High sed- cised valleys in regions behind clinoform in- cora site, and the facies pattern persists updip imentation rates may also have been associ- ¯ection points. to downdip (Fig. 5). This depositional style is ated with deposition of the upper Englishtown The deltaically in¯uenced shoreline model unexpected because Miocene sequences in unit (ca. 77±76 Ma; Fig. 13). These periods (Fig. 4) predicts sandy HSTs in most sequenc- New Jersey show a distinct pattern in which re¯ect times of increased in¯ux of siliciclastic es, a prediction that is generally upheld in Up- more sequences are preserved in downdip input. per Cretaceous sequences at Ancora and Bass boreholes (Miller et al., 1997). We attribute River. However the major aquifers (Mount this contrast in depositional styles to deposi- DISCUSSION Laurel and Magothy Formations) are thick tion on a ramp during the Cretaceous com- sand accumulations that thicken and thin clos- pared to deposition of thick (hundreds of me- Facies Changes Within Sequences er and farther away from point sources, re- ters), prograding clinoforms on the Miocene spectively. For example, the Magothy sands shelf (Steckler et al., 1999). The updip Ancora Continuous cores through the thick Upper thicken dramatically toward the Long Island site has higher amounts of medium to coarse Cretaceous sections at Ancora and Bass River platform, but thin along strike to the southeast quartz sand than the downdip Bass River site, provide insights about models for sedimenta- of the Ancora±Bass River dip pro®le (Owens as expected because Ancora is more proximal tion within sequences. The basic deltaically and Gohn, 1985), indicating a primary source to the source. Glauconite is also more com-

Figure 12. Comparison of Late Cretaceous data. Thin horizontal dashed lines are drawn at 5 m.y. increments (Sant.ÐSantonian). (A) Deep-sea benthic foraminiferal ␦18O records (ODP [Ocean Drilling Program] Site 463, Barrera and Savin, 1999; Site 690, Barrera and Savin, 1999; Site 511, Huber et al., 1995; Sites 1049, 1050, Huber et al., 2002) and planktonic foraminiferal ␦18O records (Site 463, Barrera and Savin, 1999). Arrows are drawn through the in¯ection points of the isotopic records. Temperatures are computed by

؊1.2½ with respect to PDB [Peedee belemnite]; Shackleton and Kennett, 1975), with Nuttallides values؍ assuming an ice-free world (␦w of؊0.76½ relative to equilibrium (Pak and Miller, 1992), and the paleotemperature equation in Barrera and Savin (1999). (B) New Jersey composite sequences (derived from Fig. 1). Gray boxesÐtime represented; white areasÐhiatuses; and thin white linesÐinferred hiatuses. (C) Backstripped R2 eustatic estimates for Bass River (black, thin discontinuous lines) and Ancora (gray, thick discontinuous lines) for both the conservative and preferred age models. (D) Our best estimate of eustatic changes for the New Jersey Coastal Plain derived from the R2 curves (black continuous lines indicates parts of the curve constrained by data, dashed lines indicate parts inferred). (E) The relative sea-level curve from northwest Europe (gray continuous line; Hancock, 1993) and the backstripped record from the Russian platform (black continuous heavy line; Sahagian et al., 1996). Arrows indicate positive ␦18O in¯ections (inferred cooling and/ or ice-volume increases). (F) The EPR eustatic estimate (Haq et al., 1987). TA and UZA and numbers on the left side of the panel refer to sequences de®ned by EPR (Haq et al., 1987). N

Geological Society of America Bulletin, March/April 2004 387 MILLER et al.

greater thickness of the Merchantville For- mation compared with other glauconite sands can be explained by the fact that this forma- tion is at least two or three different sequences concatenated together. It is clear that the high abundance of glauconite throughout the Na- vesink sequence(s) at Ancora is due to a very low siliciclastic input in the central part of the New Jersey Coastal Plain where the Navesink lithology is dominant (e.g., at Ancora). In the northern part of the coastal plain, local sands developed from small point sources (e.g., the Shrewsbury Member of the Red Bank For- mation; the Tinton Formation). Downdip at Bass River, a slowly accumulated clay pre- dominates (New Egypt Formation). The up- permost Cretaceous±lower Paleocene green- sands (the Hornerstown cycle of Olsson (1991); the Navesink II sequence here) re¯ect very little siliciclastic input, and deposition was almost exclusively of authigenic glauco- nite. These facies relationships can be ex- plained by high late Campanian siliciclastic input, its reduction during the Maastrichtian, and its virtual shutdown during the early Pa- leocene. This pattern is widespread, occurring throughout the New Jersey Coastal Plain, and must be ascribed to regional changes in sedi- ment supply. Olsson et al. (2002) showed that deposition was continuous across the K/T boundary at Bass River and that there was a minimal change in sea level associated with the K/T boundary. In the subsurface we show that the K/T boundary occurred during deposition of the Navesink II sequence (ca. 66±64.5 Ma; Figs. 11 and 12). The sequence shallows up- section during the last 0.5 m.y. of the Creta- ceous, culminating in a shallowing beginning in the last 100 k.y. of the Cretaceous (Olsson et al., 2002). However, there is no sequence boundary associated with the K/T boundary; a sequence boundary occurs in Biochron P1b Figure 13. Average sedimentation rates within sequences for Ancora (black) and Bass with a hiatus from ca. 64.5 to 63 Ma (Figs. River (thick gray lines) derived from Table 1. The Bass River III sequence at Bass River 11 and 12). is not shown (off scale, 31 m/m.y.). KÐCretaceous; San.ÐSantonian; Con.ÐConiacian. Sequences and Sea Level mon at the updip site, suggesting peak glau- common in the Bass River, Magothy, and The Ancora and Bass River boreholes ex- conite deposition on the middle shelf and Woodbury Formations downdip (Fig. 5). tend the dated record of sequences on the New smothering by clay deposition on the outer Sea level and sediment supply were impor- Jersey Margin from the Cenozoic (Miller et shelf. Recycling of glauconite (shown by tant constraints on sequence development. al., 1996, 1998a) back to nearly 100 Ma. Pi- weathered brown to yellow-green grains com- Thick, nonmarine to marginal-marine deposits oneering work of Olsson (1963, 1975), Owens pared to in situ, authigenic green-black glau- of the upper Turonian±Coniacian Magothy and Sohl (1969), and Owens and Gohn (1985) conite) is indicated by the covariance of glau- Formation were in¯uenced by high sediment established that there were at least ®ve Upper conite and quartz sands in the HST. Recycling supply and a generally lower sea level (ϳ30 Cretaceous sequences in the New Jersey of glauconite in the HSTs is common in both m lower than the Bass River sequences). The Coastal Plain. Here we not only document the Upper Cretaceous sections, though more re- Merchantville Formation was deposited dur- ages of these sequences fully, but also rec- cycling occurred at the updip site. Mica is ing a general peak in sea level during the latest ognize and date six to eight additional more common at the updip site, but it is still Santonian±early Campanian (Fig. 12). The sequences.

388 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN

The ages of the New Jersey sequences are virtually the same (if not slightly higher) eu- (Sheridan et al., 1991); however, there is no remarkably similar to the global compilation static estimate than the Merchantville (ϩ82 m evidence of faulting younger than the Ceno- of EPR (Haq et al., 1987) and of Late Creta- for the Bass River I vs. 80±78 m for the Mer- manian Potomac Formation (Olsson, 1991). ceous events in northwest Europe (Hancock, chantville) even though the water depths were 4. Our drilling results (e.g., Miller et al., 1993) and Russia (Fig. 12; Sahagian et al., much less (inner-neritic vs. middle neritic). 1997) document that Cretaceous to middle 1996). Of 16 eustatic lowerings reported by This result is consistent with observations on Miocene sedimentation in New Jersey was EPR (Haq et al., 1987), 14 show correlative other margins that show peak sea level in the more continuous than in many other coastal- events (within Ϯ0.5 m.y.) in New Jersey (Fig. Cenomanian±Turonian (e.g., the Russian plat- plain regions such as the Cape Fear Arch and 12). As noted by Hancock (1993), ®ve to six form, Fig. 12; Sahagian et al., 1996). the South Carolina Coastal Plain (e.g., Brown later Cretaceous (85±65 Ma) sequences in Backstripping of the Ancora and Bass River et al., 1972; Sohl and Owens, 1991). northwest Europe appear to correlate with se- sections con®rms that Late Cretaceous sea- 5. Sedimentation in the New Jersey Coastal quences in New Jersey (Fig. 12); limited age level changes were large (tens of meters) and Plain displays a simple pattern of increasing control, time-scale problems, and the lack of rapid (Ͻ1 m.y.), as purported by Haq et al. preservation of sequences downdip for the backstripping in the northwest European data (1987) and documented for the early Late Cre- Neogene (Miller et al., 1997) and widespread preclude closer comparison. An early Late taceous by backstripping of Russian platform sequences from the Late Cretaceous (Olsson, Cretaceous eustatic estimate from the Russian sections (Fig. 12; Sahagian et al., 1996). Such 1991, this study) to Eocene (Miller et al., platform (Sahagian et al., 1996) provides an large, rapid changes in global sea level can 1997). Both patterns argue against major tec- excellent comparison; six events correlate only be explained by glacio-eustasy (Donovan tonic changes in the coastal plain. with New Jersey, but the Bass River II and III and Jones, 1979; Pitman and Golovchenko, Thus, our backstripping results combined events are not discernible in the Russian plat- 1983). However, noneustatic mechanismsÐ with seismicity, seismic stratigraphic data, and form record. The correspondence among these such as large, rapid variations in subsidenceÐ distribution patterns of sediments all indicate records indicates a global control on Upper could explain the patterns we observe; rapid minimal tectonic effects on the Late Creta- Cretaceous sequence boundaries: eustatic variations of in-plane stress could account for ceous to Tertiary New Jersey Coastal Plain. change. large, rapid variations in subsidence (e.g., Having eliminated horizontal and vertical tec- Backstripping of coastal-plain boreholes Saurborn et al., 2000). Karner (1986) modeled tonics as a source of these events, the remain- provides eustatic estimates that can be com- the impact of in-plane stress on passive mar- ing mechanism is glacio-eustasy. pared from site to site to evaluate internal con- gins. Excess subsidence can generate an ap- Although the timing of EPR eustatic low- sistency. Van Sickel et al. (2003) and Miller parent sea-level highstand, such as we ob- erings may be more or less correct, both the et al. (2003) used the Ancora and Bass River serve. In his model, Karner (1986) found that New Jersey and Russian results show that the records to provide the ®rst fully backstripped for an old plate (in our case, 70±100 m.y. EPR curve cannot be used as a valid Late Cre- eustatic estimate for the entire Upper Creta- postrifting), subsidence landward of the hinge taceous eustatic record because the amplitudes ceous section in New Jersey (Fig. 12). Back- zone is generated by compressive stress and of the major EPR eustatic lowerings were too stripping of Upper Cretaceous onshore New decreases landward. Thus, the Ancora R2 es- high by a factor of at least two (Fig. 12). Al- Jersey sequences yields sea-level amplitude timates would be expected to be lower than though we do not capture the full amplitude changes of greater than 25 m in less than 1 the Bass River R2 estimates. In general, the of change, this limitation is not suf®cient to m.y. (Fig. 12). We do not capture the full eu- fact that we see the opposite situation (Fig. explain the very large differences in amplitude static amplitude across major hiatuses (dashed 12) suggests that these events were not caused between EPR and New Jersey/Russian esti- lines, Fig. 12) because mostly TSTs and HSTs by in-plane stress. mates (Fig. 12). In addition, the amplitude dif- are preserved and LSTs are largely missing; Active normal and reverse faulting has also ferences between New Jersey and EPR vary therefore, the actual lowstands may be lower been cited in the Atlantic coastal regions through time, yielding markedly different- than our estimates. The most prominent fea- (Prowell, 1988). In particular, broad (40±300 looking eustatic curves (Fig. 12). For exam- ture of our eustatic estimate are three major km), Tertiary, tectonic uplift and subsidence ple, the extremely large middle Turonian and rises at ca. 69, 76, and 84 Ma; these represent of the South Carolina Coastal Plain has been middle Maastrichtian events reported by EPR major ¯ooding events expressed by the devel- mapped by Weems and Lewis (2002). We are muted in both backstripped records, opment of widespread glauconite deposition think that Late Cretaceous±Tertiary tectonics whereas the major ¯ooding events at 69, 76, on this and other passive margins (e.g., north- is an unlikely cause for our R2 events for a and 84 Ma in New Jersey are less important west Europe). number of reasons: in the EPR record (Fig. 12). We conclude that Backstripping often yields results that are 1. Backstripping documents that the only it is time to abandon the use of the EPR record counterintuitive because water-depth varia- discernible tectonic effect during the past 100 for the Late Cretaceous and suggest that the tions do not necessarily equate to sea-level m.y. on coastal-plain subsidence is thermo- New Jersey and Russian platform back- changes. Backstripping of the Ancora and ¯exural (Kominz et al., 1998; Van Sickel et stripped records provide the best substitute. Bass River sections shows that major marine al., 2003). Having eliminated tectonics as a cause for transgressions associated with the Merchant- 2. Southern New Jersey has not been the our sea-level estimates, the only mechanism ville, Marshalltown, and Navesink Formation site of the large number or magnitude of earth- that can explain the size and rapidity of our glauconites are all eustatic highstands (Fig. quakes seen near Charleston, South Carolina, eustatic estimates is glacial growth and decay. 12). Although relative water depths were and other areas of active faulting in the Atlan- If ice-volume changes drove Late Cretaceous greater during the deposition of the glauco- tic Coastal Plain (Seeber and Armbruster, sea-level changes, then foraminiferal ␦18O rec- nites than during the Cenomanian±Turonian 1988). ords should show increases associated with se- sequences, backstripping shows the Bass Riv- 3. Seismic lines from southern New Jersey quence boundaries. Such a link has been es- er I middle Cenomanian sequence actually has image faults in the New Jersey Coastal Plain tablished for the late middle Eocene to

Geological Society of America Bulletin, March/April 2004 389 MILLER et al.

Miocene (Miller et al., 1996, 1998a; Brown- ephemeral, and paced by Milankovitch forc- changes during this time or our understanding ing et al., 1996; Pekar et al., 2001). However, ing. Modeling evidence (DeConto and Pol- of causal mechanisms for global sea-level Late Cretaceous ␦18O records have not at- lard, 2003) indicates that a 5±10 ϫ 106 km3 change is fundamentally ¯awed. tained the resolution needed to test their re- ice sheet (Fig. 3B of DeConto and Pollard, lationship with sequence boundaries. Previous 2003, shows a 10 ϫ 106 km3 ice-sheet sce- SUMMARY AND FUTURE WORK studies have linked the ca. 71 Ma basal Na- nario) could have developed when atmospher- 18 vesink I sequence boundary with ␦ O increas- ic CO2 fell below a threshold. For the Oligo- We dated 11±14 Upper Cretaceous sequenc- es in deep-sea benthic and low-latitude plank- cene continental con®gurations, this threshold es at the Bass River and Ancora sites and cor- tonic foraminifera; on this basis, Miller et al. was estimated as three times that of the pres- related the sequence boundaries with sea-level (1999a) suggested that the 20±40 m eustatic ent; Cretaceous thresholds would have dif- lowerings of EPR, northwest European, and lowering of sea level at 71 Ma resulted from fered, but the modeling results illustrate Russian sections, establishing a global cause. the growth of a transitory ice cap equivalent greenhouse ice-sheet and sea-level dynamics. Backstripping of the Bass River and Ancora to 25%±40% of the volume of the present-day This ice sheet would not have reached the records provides a eustatic estimate for the East Antarctic ice cap. Scarce ␦18O data have Antarctic coast, hence explaining the relative Late Cretaceous that differs in amplitude and limited previous comparison of links between warmth in coastal Antarctica, but it would shape from the EPR record but agrees with Late Cretaceous global ␦18O records and have signi®cantly in¯uenced sea level by as backstripped records from the Russian plat- sequences. much as ϳ25 m and global ␦18O by as much form. The large, rapid eustatic changes require New benthic foraminiferal ␦18O data (Huber as 0.25½. Application of modeling results either growth and decay of ice sheets ϳ25%± et al., 2002) allow preliminary comparisons suggests that a maximum of 25% of the ϳ1½ 50% of the size of the modern East Antarctica between Upper Cretaceous sequence stratig- ␦18O increases at ca. 96, 93±92, and 71.2 Ma ice sheet during the supposedly ice-free Late raphy and ␦18O records (Fig. 12). Coverage for may be attributed to ice (ϳ25 m of eustatic Cretaceous or an unidenti®ed mechanism that the Coniacian±lower Campanian is still lim- lowering); ϳ75% would be attributed to deep- controlled sea-level change at this time. Stable ited to the deeply buried records from ODP water cooling of 3±4 ЊC (which by itself isotope data suggest a glacio-eustatic cause for Site 511, and there are large data gaps in the would cause only 3±4 m of eustatic lowering; the Campanian/Maastrichtian boundary (ca. lower to middle Campanian and upper Turon- Jacobs and Sahagian, 1993). Unlike the Oli- 71.2 Ma) lowering and are consistent with a ian to Coniacian. Despite these limitations, gocene and younger icehouse world, these glacio-eustatic cause for older lowerings. ␦18O comparisons are intriguing (Fig. 12), fur- Late Cretaceous ice sheets probably only ex- However, additional stable isotope data from ther suggesting the presence of small ice isted during short intervals of peak Milankov- deep-sea and onshore sections are needed to sheets in this alleged greenhouse world: (1) A itch forcing, and the continent was ice free test this link. major middle Cenomanian sequence boundary during much of the greenhouse Late Creta- We integrate our interpretation of sea level, (see also Gale et al., 2002) between the Po- ceous to middle Eocene. ages, environments, and sedimentation rates tomac and Bass River I sequences (hiatus at Milankovitch forcing paced the develop- into an overview of Late Cretaceous sedimen- ca. 97±95.8 Ma) correlates with a major ment of the ephemeral ice sheets in Antarcti- tation in the New Jersey Coastal Plain. The (Ͼ1½) ␦18O increase (Fig. 12), and (2) a mid- ca. Modeling studies of Matthews and Froh- New Jersey Coastal Plain formed as a result dle Turonian sea-level lowering associated lich (2002) predicted glacio-eustatic falls from of thermo¯exural subsidence some 50 m.y. af- with the Bass River III/Magothy contact (92± Milankovitch orbital solutions that are similar ter rifting. Though accommodation was large- 91.5 Ma) may correlate with a major increase to those we obtained from the New Jersey ly due to ¯exural subsidence of the coastal in benthic foraminiferal ␦18O values (ϳ1.0½), margin (Table 1). This convergence of model plain, the form of onshore subsidence is ther- though additional data are needed to deter- predictions (Matthews and Frohlich, 2002) mal (Kominz et al., 1998). From ca. 120 to 97 mine the precise timing of the increase (Fig. with our sea-level history is remarkable (Table Ma, deposition was largely ¯uvial and/or 12). Several other Coniacian±Campanian ␦18O 1). The alternative to invoking Late Creta- alluvial-plain clays (inter¯uves and overbank increases (dashed arrows, Fig. 12) may be re- ceous ice sheets is that global sea-level chang- environments) and sands (channel, point-bar, lated to sequence boundaries, but the data are es were paced by as-yet-unde®ned mecha- crevasse-splay, and ¯uvial-lacustrine environ- too sparse to provide a ®rm correlation. nisms, because none of the other hypothesized ments) of the Potomac Formation. Pollen pro- Our comparisons link several Upper Creta- mechanisms (temperature effects, storage in vides the primary age control for these red ceous sequence boundaries to ␦18O increases lakes, deep-water changes, groundwater, or beds that span the Early Cretaceous/Late Cre- that are major deep-water (hence high- sea ice; Jacobs and Sahagian, 1993) can ex- taceous boundary. Coring downdip at Ancora latitude) cooling events and possibly ice vol- plain the observed 20±30 m changes in Ͻ1 recovered marine intercalations in the Poto- ume events; the increases must be ascribed m.y. mac Formation. Future drilling may yield suf- primarily to cooling because the ϳ1½ in- With the exception of the ca. 71 Ma and ®cient marine beds and improved pollen stra- creases cannot be totally due to ice-volume or perhaps 96 Ma events, Late Cretaceous com- tigraphy that may allow deciphering of the salinity variations. Nevertheless, the link be- parisons between sequence boundaries and Potomac Formation. tween ␦18O increases and eustatic lowerings ␦18O increases are not compelling, and future Following a major middle Cenomanian eu- (Fig. 12) implies that at least a part of the Late work must generate more detailed ␦18O rec- static lowstand (97±95.8 Ma), marine deposi- Cretaceous ␦18O signature was due to devel- ords for the Late Cretaceous. Nevertheless, tion predominated in downdip locations (the opment of ice sheets. our backstripping results require that large, Bass River Formation at Bass River and An- Miller et al. (2003) explained the presence rapid sea-level variations occurred in the Late cora). This major marine incursion occurred of ice sheets in the greenhouse world of the Cretaceous greenhouse world, and we must in response to a general rise of sea level of Late Cretaceous by proposing that the ice conclude that either small- to medium-sized ϳ20 m, punctuated by two eustatic lowerings sheets were restricted in area in Antarctica, (5±10 ϫ 106 km3) ice sheets paced sea-level (94.6 and 93.5 Ma). A major middle Turonian

390 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN

eustatic lowering (92.1±91.4 Ma) separated Cretaceous into the early Paleocene. The Late oceans, in Johnson, C.C., and Barrera, E., eds., Evo- lution of the Cretaceous ocean-climate system: Boul- deposition of the Bass River Formation from Cretaceous Epoch ended in New Jersey with der, Colorado, Geological Society of America Special that of the thick, nonmarine to marginal- the delivery of impact-related spherules from Paper v. 332, p. 245±282. marine Magothy Formation. Chicxulub, Mexico (Olsson et al., 1997). Barron, E.J., 1983, A warm equable Cretaceous: The nature of the problem: Earth Science Reviews, v. 19, High rates of sediment supply from a north- Thick, continuously cored sections at An- p. 305±338. east source and a generally lower sea level cora and Bass River have provided new in- Bond, G.C., and Kominz, M.A., 1984, Construction of tec- tonic subsidence curves for the early Paleozoic mio- characterized late Turonian to Coniacian Ma- sights into Late Cretaceous facies and sea- geocline, southern Canadian Rocky Mountains: Im- gothy deposition, punctuated by two eustatic level history, though important issues remain plications for the subsidence mechanisms, age of lowerings (90±89.8 and 88.3±87.8 Ma). Peak unresolved by drilling only two holes. The breakup, and crustal thinning: Geological Society of America Bulletin, v. 95, p. 155±173. delivery of sand to the Ancora±Bass River number and signi®cance of Magothy, Mer- Bond, G.C., Kominz, M.A., Steckler, M.S., and Grotzinger, transect occurred during late Turonian Mago- chantville, and Navesink sequences require J.P., 1989, Role of thermal subsidence, ¯exure and thy I deposition; the sand supply was reduced veri®cation and have global implications if eustasy in the evolution of early Paleozoic passive- margin carbonate platforms: Society of Economic Pa- thereafter, as shown by the fact that the sands our eustatic estimate is to be used in place of leontologists and Mineralogists Special Publication, of the Magothy II and III sequences are re- the EPR record. The Magothy issue of two vs. v. 44, p. 39±61. stricted to the HSTs. three sequences (and attendant aquifer sands) Bralower, T.J., Leckie, R.M., Sliter, W.V., and Thierstein, H.R., 1995, An integrated Cretaceous microfossils A Coniacian eustatic lowering (86.7±85.2 has local and regional hydrogeologic impor- biostratigraphy: Society of Economic Paleontologists Ma) was followed by deposition of the thin tance. Though outcrop-subsurface correlations and Mineralogists Special Publication 54, p. 65±79. lower to middle Santonian marine Cheese- and chronology of sequences are both greatly Brown, P.M., Miller, J.A., and Swain, F.M., 1972, Structural and stratigraphic framework, and spatial distribution quake sequence and a subsequent late Santon- improved, uncertainties remain about the re- of permeability of the Atlantic Coastal Plain, North ian eustatic lowering (84.3±83.9 Ma). Perva- lationship and age of critical units (e.g., age Carolina to New York: U.S. Geological Survey Pro- fessional Paper 796, 79 p., 59 plates. sive glauconite deposition during a peak in sea of the Englishtown sequence). Drilling at Browning, J.V., Miller, K.G., and Pak, D.K., 1996, Global level began in the latest Santonian to early Millville, New Jersey (Fig. 1), provides a Cre- implications of Eocene greenhouse and doubthouse Campanian with the Merchantville Formation. taceous section intermediate in dip position sequences on the New Jersey Coastal Plain: The ice- house cometh: Geology, v. 24, p. 639±642. The greater thickness of the Merchantville between Ancora and Bass River, though the Burke, W.H., Denison, R.E., Hetherington, E.A., Koepnick, Formation vs. other glauconite sands is the re- Upper Cretaceous section thins toward the R.B., Nelson, H.F., and Otto, J.B., 1982, Variation of sult of concatenation of three different se- southeast. Future drilling between Ancora and seawater 87Sr/86Sr throughout Phanerozoic time: Ge- ology, v. 10, p. 516±519. quences; eustatic lowerings occurred across Bass River is needed to resolve the Magothy Christie-Blick, N., and Driscoll, N.W., 1995, Sequence stra- the Santonian/Campanian boundary (83.5 Ma) issue, whereas drilling in the northwest adja- tigraphy: Annual Review of Earth and Planetary Sci- and in the early Campanian (83.1±81 Ma). cent to the coast is needed to penetrate the ences, v. 23, p. 451±478. Christie-Blick, N., Grotzinger, J.P., and von der Borch, High sedimentation rates (Ͼ17 cm/k.y.) dur- thickest marine Upper Cretaceous section pos- C.C., 1988, Sequence stratigraphy in Proterozoic suc- ing the middle Campanian deposition of the sible while allowing correlation to offshore cessions: Geology, v. 16, p. 100±104. Woodbury±lower Englishtown units represent seismic pro®les. Drilling was scheduled along Christopher, R.A., 1982, The occurrence of the Complexiopollis-Atlantopolis Zone (palynomorphs) in high deltaic input. Following a middle Cam- the coast in fall 2003 near Sea Girt, New Jer- the Eagle Ford Group (Upper Cretaceous) of Texas: panian eustatic lowering (77.8±76.7 Ma) and sey (Fig. 1). Journal of Paleontology, v. 56, p. 525±541. deposition of a thin, neritic glauconite sand Dean, W.E., and Arthur, M.A., eds., 1998, Stratigraphy and paleoenvironments of the Cretaceous Western Interior (now the lower part of the upper English- ACKNOWLEDGMENTS Seaway, USA: SEPM (Society for Sedimentary Ge- town), siliciclastic input continued to be high ology), Concepts in Sedimentology and Paleontology This paper is dedicated to the memory of the late during deposition of the middle Campanian Series, no. 6, 255 p. James Owens, who was committed to understanding DeConto, R.M., and Pollard, D., 2003, Rapid Cenozoic gla- Englishtown sequence. A late Campanian eu- the New Jersey Coastal Plain; this paper is a testa- ciation of Antarctica induced by declining atmospher- static lowering (76±75 Ma) was followed by ment to his support of onshore drilling and coastal- ic CO2: Nature, v. 421, p. 245±249. deposition of the glauconite sand bodies now plain studies. We thank W. Harris, M.E. Katz, and de Graciansky, P.C., Hardenbol, J., Jacquin, T., Farley, M., and Vail, P. R., eds.,1998, Mesozoic±Cenozoic se- forming the late Campanian Marshalltown se- G. Karner for reviews and the members of the Coastal Plain Drilling Project who are not listed quence stratigraphy of European basins: SEPM (So- quence; quartz sand input continued to be high ciety for Sedimentary Geology) Special Publication, here, especially those who supplied critical pub- 786 p. during the deposition of the HST of this se- lished Late Cretaceous data sets for nannofossils (D. Donovan, D.T., and Jones, E.J.W., 1979, Causes of world- quence (in the Mount Laurel Formation). Bukry, L. de Romero) and pollen (G.J. Brenner). wide changes in sea level: Geological Society [Lon- A eustatic lowering spanning the Campan- The New Jersey Geological Survey (H. Kasabach, don] Journal, v. 136, p. 187±192. K. Muessig, and R. Dalton) supplied materials, per- Doyle, J.A., and Robbins, E.I., 1977, Angiosperm pollen ian/Maastrichtian boundary (71.2±69 Ma) was sonnel, and logging support, funded all drilling zonation of the continental Cretaceous of the Atlantic followed by a secondary peak in sea level and costs for Bass River, and provided partial support Coastal Plain and its application to deep wells in the a return to glauconite deposition in the Maas- for drilling costs at Ancora. Supported by National Salisbury Embayment: Palynology, v. 1, p. 43±78. trichtian Navesink Formation. Siliciclastic in- Science Foundation grants OCE 0084032, EAR97± Draper, N.R., and Smith, H., 1981, Applied regression anal- 08664, EAR99±09179 (all to Miller), and EAR98± ysis (2nd edition): New York, Wiley, 407 p. put was greatly reduced as glauconite depo- Erba, E., Premoli Silva, I., Wilson, P.A., Pringle, M.S., Sli- 14025 (to Kominz), the New Jersey Geological Sur- sition reigned during the Maastrichtian in ter, W.V., Watkins, D.K., Arnaud Vanneau, A., Bra- vey, and the Ocean Drilling Program. lower, T.J., Budd, A.F., Camoin, G.F., Masse, J.-P., many parts of the state, whereas local sands Mutterlose, J., and Sager, W.W., 1999, Synthesis of of the Red Bank Formation constitute the stratigraphies from shallow-water sequences at Sites HSTs in other parts. There may have been a REFERENCES CITED 871 through 879 in the western Paci®c Ocean, in Hag- gerty, J.A., Premoli Silva, I., Rack, F., and McNutt, late Maastrichtian eustatic lowering (67±66 Allen, J.L.R., 1970, Sediments of the modern Niger delta: M.K., eds., Proceedings of the Ocean Drilling Pro- Ma), but there was minimal change in sea lev- A summary and review: Society of Economic Pale- gram, Scienti®c results, Volume 144: College Station, el and no sequence boundary associated with ontologists and Mineralogists Special Publication 15, Texas, Ocean Drilling Program, p. 873±885. p. 138±151. Gale, A.S., Hardenbol, J., Hathaway, B., Kennedy, W.J., the K/T boundary at Bass River and Ancora. Barrera, E., and Savin, S.M., 1999, Evolution of late Young, J.R., and Phansalkar, V., 2002, Global corre- Siliciclastic input was minimal from the latest Campanian±Maastrichtian marine climates and lation of Cenomanian (Upper Cretaceous) sequences:

Geological Society of America Bulletin, March/April 2004 391 MILLER et al.

Evidence for Milankovitch control on sea level: Ge- surfaces and sequence boundaries: Comparisons be- Olsson, R.K., 1963, Latest Cretaceous and earliest Tertiary ology, v. 30, p. 291±294. tween observations and orbital forcing in the Creta- stratigraphy of New Jersey Coastal Plain: American Glaser, J.D., 1969, Petrology and origin of Potomac and ceous and Jurassic (65±190 Ma): GeoArabia, v. 7, Association of Petroleum Geologists Bulletin, v. 47, Magothy (Cretaceous) sediments, middle Atlantic p. 503±538. p. 643±665. Coastal Plain: Maryland Geological Survey Report of McArthur, J.M., Kennedy, W.J., Gale, A.S., Thirlwall, M.F., Olsson, R.K., 1975, Upper Cretaceous and lower Tertiary Investigations, no. 11, 101 p. Chen, M., Bunett, J., and Hancock, J.M., 1992, Stron- stratigraphy of New Jersey Coastal Plain: New York, Gohn, G.S., 1992, Preliminary ostracode biostratigraphy of tium isotope stratigraphy in the Late Cretaceous: In- Petroleum Exploration Society of New York, Second subsurface Campanian and Maastrichtian section of tercontinental correlation of the Campanian/Maas- Annual Field Trip Guidebook, 49 p. the New Jersey Coastal Plain, in Gohn, G.S., ed., Pro- trichtian boundary: Terra Nova, v. 4, p. 385±393. Olsson, R.K., 1991, Cretaceous to Eocene sea-level ¯uc- ceedings of the 1988 U.S. Geological Survey work- McArthur, J.M., Thirlwall, M.F., Gale, A.S., Chen, M., and tuations on the New Jersey margin: Sedimentary Ge- shop on the geology and geohydrology of the Atlantic Kennedy, W.J., 1993, Strontium isotope stratigraphy ology, v. 70, p. 195±208. Coastal Plain, U.S. Geological Survey Circular 1059, in the Late Cretaceous: Numerical calibration of the Olsson, R.K., and Wise, S.W., 1987, Upper Paleocene to p. 15±21. Sr isotope curve and intercontinental correlation for middle Eocene depositional sequences and hiatuses in Gradstein, F.M., Agterberg, F.P., Ogg, J.G., Hardenbol, J., the Campanian: Paleoceanography, v. 8, p. 859±873. the New Jersey Atlantic Margin: Cushman Foundation Van Veen, P., Thierry, J., and Huang, Z., 1994, A Me- McArthur, J.M., Kennedy, W.J., Chen, M., Thirlwall, M.F., for Foraminiferal Research Special Publication 24, sozoic time scale: Journal of Geophysical Research, and Gale, A.S., 1994, Strontium isotope stratigraphy p. 99±112. v. 99, p. 24,051±24,074. for the Late Cretaceous: Direct numeric calibration of Olsson, R.K., Gibson, T.G., Hansen, H.J., and Owens, J.P., Hallam, A., 1992, Phanerozoic sea-level changes: New the Sr isotope curve for the U.S. Western Interior Sea- 1988, Geology of the northern Atlantic Coastal Plain: York, Columbia University Press, 266 p. way: Palaeogeography, Palaeoclimatology, Palaeoe- Long Island to Virginia, in Sheridan, R.E., and Grow, Hancock, J.M., 1993, Transatlantic correlations in the cology, v. 108, p. 95±119. J.A., eds., The Atlantic Continental Margin, U.S.: Campanian±Maastrichtian stages by eustatic changes Miller, K.G., Wright, J.D., and Fairbanks, R.G., 1991, Un- Boulder, Colorado, Geological Society of America, of sea-level: Geological Society [London] Special locking the ice house: Oligocene±Miocene oxygen Geology of North America, v. I-2, p. 87±105. Publication 70, p. 241±256. isotopes, eustasy, and margin erosion: Journal of Geo- Olsson, R.K., Miller, K.G., Browning, J.V., Habib, D., and Haq, B.U., Hardenbol, J., and Vail, P.R., 1987, Chronology physical Research, v. 96, p. 6829±6848. Sugarman, P.J., 1997, Ejecta layer at the Cretaceous- of ¯uctuating sea levels since the Triassic (250 million Miller, K.G., Mountain, G.S., the Leg 150 Shipboard Party, Tertiary boundary, Bass River, New Jersey (Ocean years ago to present): Science, v. 235, p. 1156±1167. and Members of the New Jersey Coastal Plain Drilling Drilling Program Leg 174AX): Geology, v. 25, Hart, S.R., and Brooks, C., 1974, Clinopyroxene-matrix Project, 1996, Drilling and dating New Jersey p. 759±762. partitioning of K, Rb, Cs, and Ba: Geochimica et Cos- Oligocene±Miocene sequences: Ice volume, global sea Olsson, R.K., Miller, K.G., Browning, J.V., Wright, J.D., mochimica Acta, v. 38, p. 1799±1806. level, and Exxon records: Science, v. 271, and Cramer, B.S., 2002, Sequence stratigraphy and sea Howarth, R.J., and McArthur, J.M., 1997, Statistics for p. 1092 ±1094. level change across the Cretaceous-Tertiary boundary strontium isotope stratigraphy: A robust LOWESS ®t Miller, K.G., Browning, J.V., Pekar, S.F., and Sugarman, on the New Jersey passive margin, in Koeberl, C., and to the marine Sr-isotope curve for 0±206 Ma, with P.J., 1997, Cenozoic evolution of the New Jersey MacLeod, K.G., eds., Catastrophic events and mass look-up table for the derivation of numerical age: Coastal Plain: Changes in sea level, tectonics, and sed- extinctions: Impacts and beyond: Boulder, Colorado, Journal of Geology, v. 105, p. 441±456. iment supply, in Miller, K.G., and Snyder, S.W., eds., Geological Society of America Special Paper 356, Huber, B.T., Hodell, D.A., and Hamilton, C.P., 1995, Mid- Proceedings of the Ocean Drilling Program, Scienti®c p. 97±108. to Late Cretaceous climate of the southern high lati- results, Volume 150X: College Station, Texas, Ocean Oslick, J.S., Miller, K.G., Feigenson, M.D., and Wright, tudes: Stable isotopic evidence for minimal equator- Drilling Program, p. 361±373. J.D., 1994, Oligocene±Miocene strontium isotopes: to-pole thermal gradients: Geological Society of Miller, K.G., Mountain, G.S., Browning, J.V., Kominz, M., Stratigraphic revisions and correlations to an inferred America Bulletin, v. 107, p. 392±417. Sugarman, P.J., Christie-Blick, N., Katz, M.E., and glacioeustatic record: Paleoceanography, v. 9, Huber, B.T., Norris, R.D., and MacLeod, K.G., 2002, Deep Wright, J.D., 1998a, Cenozoic global sea-level, se- p. 427±443. sea paleotemperature record of extreme warmth dur- quences, and the New Jersey transect: Results from Owens, J.P., and Gohn, G.S., 1985, Depositional history of ing the Cretaceous: Geology, v. 30, p. 123±126. coastal plain and slope drilling: Reviews of Geophys- the Cretaceous series in the U.S. Atlantic Coastal Jacobs, D.K., and Sahagian, D.L., 1993, Climate-induced ics, v. 36, p. 569±601. Plain: Stratigraphy, paleoenvironments, and tectonic ¯uctuations in sea level during non-glacial times: Na- Miller, K.G., Sugarman, P.J., Browning, J.V., Olsson, R.K., controls of sedimentation, in Poag, C.W., ed., Geolog- ture, v. 361, p. 710±712. Pekar, S.F., Reilly, T.J., Cramer, B.S., Aubry, M.-P., ic evolution of the United States Atlantic Margin: Karner, G.D., 1986, Effects of lithospheric in-plane stress Lawrence, R.P., Curran, J., Stewart, M., Metzger, J.M., New York, Van Nostrand Reinhold, p. 25±86. on sedimentary basin stratigraphy: Tectonics, v. 5, Uptegrove, J., Bukry, D., Burckle, L.H., Wright, J.D., p. 573±588. Feigenson, M.D., Brenner, G.J., and Dalton, R.F., Owens, J.P., and Sohl, N., 1969, Shelf and deltaic paleoen- Kauffman, E.G., 1977, Geological and biological overview: 1998b, Bass River Site, in Miller, K.G., Sugarman, vironments in the Cretaceous±Tertiary formations of Western Interior Cretaceous basin: Mountain Geolo- P.J., Browning, J.V., et al., Proceedings of the Ocean the New Jersey Coastal Plain, in Subitsky, S., ed., Ge- gist, v. 14, p. 75±99. Drilling Program, Scienti®c results, Volume 174AX: ology of selected areas in New Jersey and eastern Klitgord, K.D., Hutchinson, D.R., and Schouten, H., 1988, College Station, Texas, Ocean Drilling Program, Pennsylvania and guidebook of excursion: New U.S. Atlantic continental margin: Structural and tec- p. 5±43. Brunswick, New Jersey, Rutgers University Press, tonic framework, in Sheridan, R.E., and Grow, J.A., Miller, K.G., Barrera, E., Olsson, R.K., Sugarman, P.J., and p. 235±278. eds., The Atlantic Continental Margin, U.S.: Boulder, Savin, S.M., 1999a, Does ice drive early Maastrichtian Owens, J.P., Minard, J.P., Sohl, N.F., and Mello, J.F., 1970, Colorado, Geological Society of America, Geology of eustasy? Global ␦18O and New Jersey sequences: Ge- Stratigraphy of the outcropping post-Magothy Upper North America, v. I-2, p. 19±55. ology, v. 27, p. 783±786. Cretaceous formations in southern New Jersey and Kominz, M.A., and Pekar, S.F., 2001, Oligocene eustasy Miller, K.G., Sugarman, P.J., Browning, J.V., Cramer, B.S., northern Delmarva Peninsula, Delaware and Mary- from two-dimensional sequence stratigraphic back- Olsson, R.K., de Romero, L., Aubry, M.-P., Pekar, land: U.S. Geological Survey Professional Paper 674, stripping: Geological Society of America Bulletin, S.F., Georgescu, M.D., Metzger, K.T., Monteverde, 60 p. v. 113, p. 291±304. D.H., Skinner, E.S., Uptegrove, J., Mullikin, L.G., Owens, J.P., Sugarman, P.J., Sohl, N.F., Parker, R., Hough- Kominz, M.A., Miller, K.G., and Browning, J.V., 1998, Muller, F.L., Feigenson, M.D., Reilly, T.J., Brenner, ton, H.H., Volkert, R.V., Drake, A.A., and Orndorff, Long-term and short-term global Cenozoic sea-level G.J., and Queen, D., 1999b, Ancora Site, in Miller, R.C., 1998, Bedrock geologic map of central and estimates: Geology, v. 26, p. 311±314. K.G., Sugarman, P.J., Browning, J.V., et al., Proceed- southern New Jersey: U.S. Geological Society Mis- Litwin, R.J., Sohl, N.F., Owens, J.P., and Sugarman, P.J., ings of the Ocean Drilling Program, Scienti®c results, cellaneous Investigations Series Map I-2540-b, scale 1993, Palynological analysis of a newly recognized Volume 174AX (supplement): College Station, 1:100,000, 4 sheets. Upper Cretaceous marine unit at Cheesequake, New Texas, Ocean Drilling Program, p. 1±65 [online]: Pak, D.K., and Miller, K.G., 1992, Paleocene to Eocene Jersey: Palynology, v. 17, p. 123±135. http://www-odp.tamu.edu/publications/174AXSIR/ benthic foraminiferal isotopes and assemblages: Im- Locker, S.D., Hine, A.C., Tedesco, L.P., and Shinn, E.A., VOLUME/CHAPTERS/174AXS࿞1.PDF. plications for deepwater circulation: Paleoceanogra- 1996, Magnitude and timing of episodic sea-level rise Miller, K.G., Sugarman, P.J., Browning, J.V., Kominz, phy, v. 7, p. 405±422. during the last deglaciation: Geology, v. 24, M.A., Hernandez, J.C., Olsson, R.K., Wright, J.D., Pekar, S.F., and Kominz, M.A., 2001, Two-dimensional pa- p. 827±830. Feigenson, M.D., and Van Sickel, W., 2003, Late Cre- leoslope modeling: A new method for estimating wa- Lyell, C., 1845, Notes on the Cretaceous strata of New taceous chronology of large, rapid sea-level changes: ter depths for benthic foraminiferal biofacies and pa- Jersey, and other parts of the United States bordering Glacioeustacy during the greenhouse world: Geology, leo shelf margins: Journal of Sedimentary Research, the Atlantic: Quarterly Journal of the Geological So- v. 31, p. 585±588. v. 71, p. 608±620. ciety of London, v. 1, p. 55±60. Mitchum, R.M., Vail, P.R., and Thompson, S., 1977, The Pekar, S.F., Christie-Blick, N., Kominz, M.A., and Miller, Martino, R.L., and Curran, H.A., 1990, Sedimentology, ich- depositional sequence as a basic unit for stratigraphic K.G., 2001, Evaluating the stratigraphic response to nology, and paleoenvironments of the Upper Creta- analysis: American Association of Petroleum Geolo- eustasy from Oligocene strata in New Jersey: Geolo- ceous Wenonah and Mt. Laurel Formations, New Jer- gists Memoir, v. 26, p. 53±62. gy, v. 29, p. 55±58. sey: Journal of Sedimentary Petrology, v. 60, Olsson, R.K., 1960, Foraminifera of latest Cretaceous and Perry, W.J., Jr., Minard, J.P., Weed, E.G.A., Robbins, E.I., p. 125±144. earliest Tertiary age in the New Jersey Coastal Plain: and Rhodehamel, E.C., 1975, Stratigraphy of Atlantic Matthews, R.K., and Frohlich, C., 2002, Maximum ¯ooding Journal of Paleontology, v. 34, p. 1±58. Coastal Margin of United States north of Cape

392 Geological Society of America Bulletin, March/April 2004 UPPER CRETACEOUS SEQUENCES AND SEA-LEVEL HISTORY, NEW JERSEY COASTAL PLAIN

HatterasÐBrief survey: American Association of Pe- Atlantic seaboard of the U.S.: Intraplate neotectonics M.D., 1995, Uppermost Campanian±Maestrichtian troleum Geologists Bulletin, v. 59, p. 1529±1548. and earthquake hazard, in Sheridan, R.E., and Grow, strontium isotopic, biostratigraphic, and sequence Petters, S.W., 1976, Upper Cretaceous subsurface stratig- J.A., eds., The Atlantic Continental Margin, U.S.: stratigraphic framework of the New Jersey Coastal raphy of Atlantic Coastal Plain of New Jersey: Amer- Boulder, Colorado, Geological Society of America, Plain: Geological Society of America Bulletin, v. 107, ican Association of Petroleum Geologists Bulletin, Geology of North America, v. I-2, p. 565±582. p. 19±37. v. 60, p. 87±107. Shackleton, N.J., and Kennett, J.P., 1975, Paleotemperature Sugarman, P.J., Miller, K.G., Olsson, R.K., Browning, J.V., Pitman, W.C., III, and Golovchenko, X., 1983, The effect history of the Cenozoic and the initiation of Antarctic Wright, J.D., de Romero, L.M., White, T.S., Muller, of sea-level change on the shelf edge and slope of glaciation: Oxygen and carbon isotope analysis in F.L., and Uptegrove, J., 1999, The Cenomanian/Tu- passive margins: Society of Economic Paleontologists DSDP Sites 277, 279, and 281, in Kennett, J.P., Houtz, ronian carbon burial event, Bass River, NJ, USA: Geo- and Mineralogists Special Publication, v. 33, R.E., eds., Initial reports of the Deep Sea Drilling chemical, paleoecological, and sea-level changes: p. 41±58. Project, Volume 29: Washington, D.C., U.S. Govern- Journal of Foraminiferal Research, v. 29, p. 438±452. Posamentier, H.W., Jervey, M.T., and Vail, P.R., 1988, Eu- ment Printing Of®ce, p. 743±755. Vail, P.R., Mitchum, R.M., Jr., Todd, R.G., Widmier, J.M., static controls on clastic deposition IÐConceptual Sheridan, R.E., and Grow, J.A., eds., 1988, The Atlantic Thompson, S., III, Sangree, J.B., Bubb, J.N., and Ha- framework: Society of Economic Paleontologists and Continental Margin, U.S.: Boulder, Colorado, Geolog- tlelid, W.G., 1977, Seismic stratigraphy and global Mineralogists Special Publication, v. 42, p. 109±124. ical Society of America, Geology of North America, changes of sea level: American Association of Petro- Price, E.D., 1999, The evidence and implication of polar v. I-2, 610 p. leum Geologists Memoir, v. 26, p. 49±212. ice during the Mesozoic: Earth Science Reviews, Sheridan, R.E., Olsson, R.K., and Miller, J.J., 1991, Seismic Van Sickel, W.A., Kominz, M.A., Miller, K.G., and Brown- v. 48, p. 183±210. re¯ection and gravity study of proposed Taconic su- ing, J.V., 2004, Late Cretaceous and Cenozoic sea- Prowell, D.C., 1988, Cretaceous and Cenozoic tectonism ture under the New Jersey Coastal Plain: Implications level estimates: Backstripping analysis of borehole on the Atlantic Coastal Margin, in Sheridan, R.E., and for continental growth: Geological Society of America data, onshore, New Jersey: (in press). Grow, J.A., eds., The Atlantic Continental Margin, Bulletin, v. 103, p. 402±414. Watts, A.B., and Steckler, M.S., 1979, Subsidence and eus- U.S.: Boulder, Colorado, Geological Society of Amer- Skinner, E.S., 2001, The upper Campanian Marshalltown tasy at the continental margin of eastern North Amer- ica, Geology of North America, v. I-2, p. 557±564. sequence of the New Jersey Coastal Plain, Bass River ica: American Geophysical Union Maurice Ewing Se- Reynolds, D.J., Steckler, M.S., and Coakley, B.J., 1991, and Ancora Boreholes [M.S. thesis]: New Brunswick, ries, v. 3, p. 218±234. The role of the sediment load in sequence stratigra- New Jersey, Rutgers University, 83 p. Weems, R.E., and Lewis, W.C., 2002, Structural and tec- phy: The in¯uence of ¯exural isostasy and compac- Sohl, N.F., and Owens, J.P., 1991, Cretaceous stratigraphy tonic setting of the Charleston, South Carolina, region: tion: Journal of Geophysical Research, v. 96, of the Carolina Coastal Plain, in Horton, J.W., Jr., and Evidence from the Tertiary stratigraphic record: Geo- p. 6931±6949. Zullo, V.A., eds., The geology of the Carolinas: Car- logical Society of America Bulletin, v. 114, p. 24±42. Sahagian, D., Pinous, O., Olferiev, A., Zakaharov, V., and olina Geological Society ®ftieth anniversary volume: Weller, S., 1907, A report on the Cretaceous paleontology Beisel, A., 1996, Eustatic curve for the Middle Knoxville, University of Tennessee Press, p. 191±220. of New Jersey: Trenton, New Jersey, Geological Sur- Jurassic±Cretaceous based on Russian platform and Steckler, M.S., 1981, Thermal and mechanical evolution of vey of New Jersey Paleontology Series, v. 4, 871 p. Siberian stratigraphy: Zonal resolution: American As- Atlantic-type margins [Ph.D. thesis]: New York, Co- Wilgus, C.K., Hastings, B.S., Kendall, C.G.St.C., Posamen- sociation of Petroleum Geologists Bulletin, v. 80, lumbia University, 261 p. tier, H.W., Ross, C.A., Van Wagoner, J.C., 1988, Sea- p. 1433±1458. Steckler, M.S., Mountain, G.S., Miller, K.G., and Christie- level changes: An integrated approach: Society of Saurborn, M.J., Karner, Garry, D., Christie-Blick, N., Plint, Blick, N., 1999, Reconstruction of Tertiary prograda- Economic Paleontologists and Mineralogists Special A.G., and Scholz, C.H., 2000, Accommodation chang- tion and clinoform development on the New Jersey Publication, v. 42, 407 p. es in Upper Cretaceous sediments of the Alberta fore- passive margin by 2-D backstripping: Marine Geolo- land basin: Implications for the role of changes in in- gy, v. 154, p. 399±420. MANUSCRIPT RECEIVED BY THE SOCIETY 17 OCTOBER 2002 plane force on unconformity development: Geological Stoll, H.M., and Schrag, D.P., 1996, Evidence for glacial REVISED MANUSCRIPT RECEIVED 29 MAY 2003 Society of America Abstracts with Programs, v. 32, control of rapid sea level changes in the Early Cre- MANUSCRIPT ACCEPTED 31 JULY 2003 no. 7, p. 156. taceous: Science, v. 272, p. 1771±1774. Seeber, L., and Armbruster, G., 1988, Seismicity along the Sugarman, P.J., Miller, K.G., Bukry, D., and Feigenson, Printed in the USA

Geological Society of America Bulletin, March/April 2004 393