GSA Bulletin: Synthesis and revision of groups within the Newark

ABSTRACT

The Newark Supergroup currently includesnine stratigraphic groups, each of which ap-plies to part or all of the rock column of onlyone or a few basins. Because the group nomen-clature within the Newark Supergroup is nei-ther inclusive nor parallel in its concepts,nearly half of the strata within the Newark Su-pergroup lacks any group placement. A newsystem is proposed herein that (1) establishesunambiguous group boundaries, (2) places allNewark Supergroup strata into groups, (3) re-duces the number of group names from nine tothree, (4) creates parallelism between groupsand three major successive tectonic events thatcreated the rift basins containing the NewarkSupergroup, and (5) coincidentally providesisochronous or nearly isochronous groupboundaries. These proposed groups are (1) theChatham Group (Middle Triassic to basalLower Jurassic sedimentary rocks), (2) theMeriden Group (Lower Jurassic extrusive vol-canic and sedimentary rocks), and (3) the Aga-wam Group (new name) (Lower Jurassic sedi-mentary rocks above all early Mesozoicigneous intrusive and extrusive rocks).

This new rock classification system makesuse of the fact that a discrete interval of syn-chronous or nearly synchronous volcanismand plutonism occurred throughout the earlyMesozoic rift system of eastern North Amer-ica. The presence or absence of volcanic rocksprovides a powerful stratigraphic tool for es-tablishing regional groups and group bound-aries. The presence of sedimentary rocks in-jected by diabase dikes and sills, in the absenceof extrusive volcanic rocks, places Newark Su-pergroup rocks in the Chatham Group. Thepresence of extrusive volcanic rocks, interbed-ded with sedimentary rocks injected by dia-base dikes and sills, places Newark Super-group rocks in the Meriden Group. Thepresence of sedimentary rocks lacking both ex-trusive volcanic rocks and diabase dikes and

sills, places Newark Supergroup rocks in theAgawam Group. Application of this new re-gional group stratigraphy to the early Meso-zoic rift basins requires revision of the stratig-raphy of several basins to make formationboundaries match group boundaries.

INTRODUCTION

The Newark Supergroup (Fig. 1) is an inclu-sive stratigraphic term for all continental sedi-mentary and extrusive volcanic rocks of MiddleTriassic through Early Jurassic age that are pre-served in 29 rift basins exposed in eastern NorthAmerica (Olsen, 1978; Froelich and Olsen,1984). Age-equivalent rocks lie buried beneaththe Atlantic Coastal Plain, but these have notbeen added to the Newark Supergroup becausetheir age, areal extent, and lithologic compositionremain unknown (Luttrell, 1989, p. 2). The New-ark Supergroup is bounded by profound regionalunconformities, each representing 90 m.y. ormore. Although local internal unconformitieshave been documented, the Newark Supergroupregionally represents nearly continuous deposi-tion from Middle Triassic into Early Jurassic time(Olsen, 1986).

Formational nomenclature within the NewarkSupergroup generally is adequate (Luttrell, 1989,Plate 1), but group-level stratigraphy remains dis-organized and incompletely applied (Fig. 2).Four group names (Conewago, Lewisburg orLewisberry, Lisbon or Lisburn, and Manchester)have been proposed and subsequently abandoned(Luttrell, 1989); these are not discussed further.Among groups currently accepted in the litera-ture, two groups occur within one basin (Tucka-hoe and Chesterfield Groups in the Richmondbasin); one group encompasses the entire columnin several basins (Chatham Group in the Crow-burg, Wadesboro, Ellerbe, Sanford, and Durhambasins; Dan River Group in the Davie Countyand Dan River–Danville basins; Tuckahoe Groupin the Deep Run and Flat Branch basins; Cul-peper Group in the Barboursville and Culpeper

basins; and the Fundy Group in various areas ofthe Fundy basin), one group encompasses the en-tire column of a single basin (Hartford Group inthe Hartford basin), and one group encompassesonly part of the sequence in three basins (Bruns-wick Group in the Newark basin; Meriden Groupin the Pomperaug and Hartford basins). In otherearly Mesozoic basins, no groups have been pro-posed (Scottsburg, Randolph, Roanoke Creek,Briery Creek, Farmville, Taylorsville, Scottsville,Gettysburg, Cherry Brook, Deerfield, Northfield,and Middleton basins). Thus some groups corre-spond to one basin, some groups are present inmore than one basin, some groups include onlypart of one basin, and some basins have no groupassignment.

This approach to the group concept has unnec-essarily fragmented a relatively uniform regionallithostratigraphic and tectonostratigraphic pat-tern, because the depositional history of the entireNewark Supergroup can be divided into threediscrete and successive phases. In the first phase,separate basins formed and clay, silt, sand, andgravel were deposited in fluvial, lacustrine, andalluvial fan environments. This phase persistedabout 35 m.y., from the Middle Triassic to the be-ginning of the Jurassic (Olsen, 1986). A seconddepositional phase occurred within the EarlyJurassic, when large volumes of basaltic magmaspread as basalt flows across the basin floors.These flows are interbedded with sedimentaryrocks, formed from continental sediments thatcontinued to accumulate between basaltic erup-tions. On the basis of analysis of Milankovitch-type cycles, volcanic extrusions began synchro-nously within 21 k.y., ended synchronously, andlasted for 580 ± 100 k.y. (Olsen and Fedosh,1988; Olsen et al., 1996). The third phase com-menced after volcanism ceased and sedimenta-tion resumed without volcanic interruption in atleast two of the rift valleys.

Wherever the middle stage of the Newark Su-pergroup rock column is fully preserved, the in-terval of basaltic magmatism is represented bythree successive stratigraphic suites of multiple

195

Synthesis and revision of groups within the Newark Supergroup,eastern North America

Robert E. Weems U.S. Geological Survey, M.S. 926A, Reston, Virginia 22092

Paul E. Olsen Lamont-Doherty Earth Observatory of Columbia University, Palisades, New York 10964

GSA Bulletin; February 1997; v. 109; no. 2; p. 195–209; 9 figures.

WEEMS AND OLSEN

196 Geological Society of America Bulletin, February 1997

tholeiitic lava flows. Extensive local and regionalmapping throughout the early Mesozoic rift ba-sins has revealed no evidence for any extrusivevolcanic rocks below this interval or above it.Thus flow units are found in the northerly andwesterly early Mesozoic rift basins that still con-tain Jurassic rocks (Culpeper, Gettysburg, New-ark, Pomperaug, Hartford, Deerfield, and Fundy),and none are found in the southerly and easterlybasins where extensive palynological studiesshow that all preserved strata are Triassic in age(Dunay and Fisher, 1974; Cornet, 1977; Schaefferand McDonald, 1978; Olsen et al., 1982; Cornet

and Olsen, 1985; Traverse, 1986; Litwin et al.,1991). Although Early Jurassic diabase dikes andsills are common in the southerly and easterly riftbasins, any flows formerly associated with themhave been removed by erosion.

The geochemistry and petrology of the lavaflows have been extensively studied in theCulpeper, Gettysburg, Newark, Hartford, Deer-field, and Fundy basins (Smith et al., 1975; Pufferet al., 1981; Philpotts and Reichenbach, 1985;Tollo, 1988; Dostal and Greenough, 1992; Ho-zik, 1992; Puffer, 1992; Tollo and Gottfried,1992). The flows in the Pomperaug basin remain

unstudied. Tollo and Gottfried (1992) definedthree volcanic intervals, each with a distinctivechemistry probably derived from a separatemagma source. The lowest suite of flows (vol-canic interval I), which includes the correlativeMount Zion Church, Aspers (named below), Or-ange Mountain, Talcott, and North MountainBasalts (Fig. 3), consists of very similar high-ti-tanium quartz-normative (HTQ) basalts (Pufferet al., 1981; Tollo and Gottfried, 1992). The sec-ond suite of flows (volcanic interval II), which in-cludes the correlative Sander, Preakness, Holy-oke, and Deerfield Basalts and the slightly older

Figure 1. Areal distribution of theNewark Supergroup in easternNorth America. Basins in which theNewark Supergroup is preservedare listed by numbers (from Lut-trell, 1989).

NEWARK SUPERGROUP, EASTERN NORTH AMERICA

Geological Society of America Bulletin, February 1997 197

Hickory Grove Basalt (Fig. 3), consists of high-iron quartz normative/incompatible element-depleted (HFQ/IED) basalts and (in the Culpeperand Newark basins) rarer low-titanium quartz-normative (LTQ) basalts (Puffer et al., 1981;Tollo and Gottfried, 1992). The third and highestsuite of flows (volcanic interval III), which in-cludes the correlative Hook Mountain andHampden Basalts, is characterized by high-iron,high-titanium quartz normative/incompatible ele-ment-enriched (HFQ/IEE or HFTQ) basalts

(Puffer et al., 1981; Tollo and Gottfried, 1992).Local variations on these regional geochemicalpatterns justify different names for flow se-quences in different basins, but stratigraphic cor-relation between basins of these three major ex-trusive suites (each with a distinctive chemicalcomposition or range of compositions) seems tobe firmly established across the entire region(Tollo and Gottfried, 1992).

Available biostratigraphic data from the inter-layered sedimentary strata support geochemical

correlations of basalts. Cornet (1977) and Cornetand Olsen (1985) showed that the lowest HTQflow in the Culpeper, Gettysburg, Newark, Hart-ford, and Fundy basins all occur about 50 mabove the base of the Corollina meyeriana paly-nofloral zone (after Cornet, 1977). Thus the initi-ation of basaltic extrusion in all of these basinswas essentially synchronous within present lim-its of stratigraphic resolution. Olsen (1988)showed that the sedimentary cycles from the topof volcanic interval II in the Newark and Hartford

Figure 2. Currently accepted group nomenclature for the Newark Supergroup. Unnumbered portions of stratigraphic columns are not placed in anygroup. 1—Chatham Group, 2—Dan River Group, 3—Tuckahoe Group, 4—Chesterfield Group, 5—Culpeper Group, 6—Brunswick Group, 7—Meri-den Group, 8—Hartford Group, 9—Fundy Group. Data compiled from Luttrell (1989) except as follows: Hartford Group from Lorenz (1988) and Hu-bert et al. (1992). Jurassic stage assignments based on work by Padian (1989). Work by Litwin and Weems (1992) and Litwin and Ash (1993) take thecolumns of the Wadesboro, Sanford, Durham, Dan River, and Taylorsville basins well into the Norian; the top of the Richmond basin column is raisedto match correlative Taylorsville basin strata. Position of the columns for the Scottsburg, Randolph, Roanoke Creek, Briery Creek, Farmville, andScottsville basins are based partly on work by Robbins (1985) in the Farmville basin, Olsen (cited in Schaeffer and McDonald, 1978), and on extensivefieldwork by Weems. The base of the Barboursville and Culpeper columns are raised because no direct evidence exists for Carnian strata; the top of theCulpeper column is lowered on the basis of evidence discussed in text. Unconformity at the Norian-Hettangian boundary for the Deerfield basin is basedon work by J. P. Smoot, mentioned in text. The top of the Pomperaug and Deerfield columns are lowered on the basis of the work of Huber and Mc-Donald (1992) and evidence discussed in text. Location of column of the Middleton basin is based on similarity with basal rock types in nearby basins.

WEEMS AND OLSEN


basins (the Preakness and Holyoke Basalts) up tothe base of volcanic interval III (the Hook Moun-tain and Hampden Basalts) are time equivalent.This indicates that the end of the middle flow se-quence and the beginning of the upper flow se-quence in those basins also were essentially syn-chronous, despite slight differences between thebulk chemistry of the individual flow units withinthe second flow sequence in each basin (Puffer etal., 1981). Limited work on the cyclic stratigra-phy of the Waterfall Formation in the Culpeperbasin (Olsen, 1997, p. 381) suggests correlationof that unit with the Towaco and East Berlin For-mations of the Newark and Hartford basins ratherthan with the Boonton and Portland Formationsof the Newark and Hartford basins, as previouslyassumed (Lee and Froelich, 1989; Luttrell,1989). Therefore, available biostratigraphic datasupport the interbasinal correlation of volcanicintervals I through III made on the basis of theirgeochemical characteristics. Together, both sets

of data indicate that the succession of magmaticcompositions and the temporal duration of extru-sive volcanism were the same in all early Meso-zoic rift basins that still preserve basalt flows.

The widespread and unique occurrence withinthe Newark Supergroup of the volcanic flow–bearing interval makes it a regionally useful rock-stratigraphic marker horizon. The lithic charac-teristics of the flow-bearing interval, wherever itis preserved, permit a ready and natural divisionof the entire Newark Supergroup into three re-gionally recognizable groups, defined by the topand base of the flow-bearing interval and the re-gional unconformities that mark the top and baseof the entire Newark Supergroup.

This approach to the group ranking of theNewark Supergroup has numerous advantages.(1) It clusters rocks into entirely sedimentary orsedimentary and volcanic lithostratigraphic groupsthat formed during each of three major tectonicstages through which the early Mesozoic rift

basins evolved (nonvolcanic-volcanic-nonvol-canic). (2) It provides unambiguous boundaries forthese groups throughout the geographic range ofthe Newark Supergroup. (3) For the first time, it in-clusively places all Newark Supergroup strata intogroups. (4) It reduces the number of group namesfrom nine to three. (5) Coincidentally, it providesisochronous or nearly isochronous boundaries forthese groups, and thus helps make lithostrati-graphic and time-stratigraphic concepts paralleland synchronous.

The chief disadvantage to the group rankingsproposed here is that it would supersede (andthus displace) the existing group nomenclaturefor the Dan River–Danville, Davie County, Rich-mond, Deep Run, Flat Branch, Barboursville,Culpeper, Newark, Hartford, and Fundy basins(10 basins). However, extensive literature doesnot exist for four of these basins (Deep Run, FlatBranch, Barboursville, and Davie County). Theproposed group rankings would not change the

Figure 3. The distribution offormations in basins that includestrata higher than the ChathamGroup. Formational nomencla-ture in the Culpeper and Gettys-burg basins are shown as modi-fied herein. In other NewarkSupergroup basins not shownhere, all formations belong to theChatham Group (Fig. 5). Verticalscale does not reflect relativestratigraphic thicknesses, al-though horizontal scale does re-flect stratigraphic equivalence.HMDN BSLT—Hampden Ba-salt, GRANBY TUFF—GranbyBasaltic Tuff, TLCT BSLT—Tal-cott Basalt, HITCHCOCK VLC-NCS—Hitchcock Volcanics,FM—Formation.

group rankings recognized in the Crowburg,Ellerbe, Wadesboro, Sanford, Durham, and Pom-peraug basins (6 basins), and would place inclu-sive group names for the first time in the Scotts-burg, Randolph, Roanoke Creek, Briery Creek,Farmville, Taylorsville, Scottsville, Gettysburg,Cherry Brook, Deerfield, Northfield, and Mid-dleton basins (12 basins). We believe that a re-gionally applicable group nomenclature, inte-grating global-scale tectonic history withstratigraphic nomenclature, produces positive re-sults that outweigh the changes required. Namesand definitions for the three proposed inclusivelithostratigraphic groups of the Newark Super-group are as follows.

REVISIONS TO GROUP NAMESWITHIN THE NEWARK SUPERGROUP

Chatham Group (Middle Triassic to LowerJurassic, Anisian to Hettangian Stages)

This name, proposed by Emmons (1857) asthe Chatham Series, is the oldest stratigraphicname applied to rocks of the lowest part of theNewark Supergroup. Olsen (1978) was the firstto use the term Chatham Group. The type area ofthe Chatham Group is in the Sanford basin, butEmmons also applied the term to the basal stratain all of the basins known in his day (Emmons,1857, p. 1–4, 19–98). Figure 4 shows horizonsand basins to which Emmons explicitly appliedthis name. In all cases, the term was applied torocks below the level of the lowest lava flow. Al-though Emmons (1857, p. 19–29) guessed theupward extent of this group, he acknowledgedthat “the upward termination of the series … isnot clearly defined.” (p. 19). Because the distinc-tions between basalt flows and diabase sills werenot well established then, it was not obvious tohim that basalt flows could serve as stratigraphicboundaries in the various basins.

Emmons attempted to distinguish his ChathamSeries from overlying rocks partly on the basis ofplant megafossils. This approach has proven tobe inaccurate, because plant megafossils in theNewark Supergroup more often reflect deposi-tional environments than stratigraphic horizons.However, fish, reptile, and footprint remains(with which Emmons also recognized his group),as well as palynomorphs, document fauna andflora in rocks of the Chatham Group that are dis-tinctly different and older than any known fromrocks between and above the lava flows in theCulpeper, Newark, Hartford, Deerfield, andFundy basins. The most sweeping faunal and pa-lynofloral turnover documented within the New-ark Supergroup has been placed about 50 m be-low the lowest lava flows within the Newark

Supergroup (Cornet and Olsen, 1985; Fowell,1993; Fowell and Olsen, 1993; Fowell et al.,1994). This turnover occurs at or above the high-est occurrence of fish and reptiles listed by Em-mons (1857, p. 34–96) as characteristic of theChatham and below the lowest occurrence of fishand reptiles listed by Emmons (1857, p. 99–149)as characteristic of strata above the Chatham. Byplacing the upward limit of the Chatham Groupat the base of the first lava flow, we do not con-tradict the original lithologic concept which Em-mons proposed, and his biostratigraphic conceptalso remains largely intact.

Olsen (1978) used the name Chatham Group toinclude all rocks in the Wadesboro, Sanford, andDurham basins, and the North Carolina Geologi-cal Survey (1985) added to these the rocks in theCrowburg and Ellerbe basins. These definitions ofthe Chatham are more restrictive geographicallythan Emmons intended, but they do not contradictthe rock or time stratigraphic concept that Em-mons proposed. Similarly, this recent usage is notcontradicted by the broader rock stratigraphicsense of the group name as we propose here, nordoes the recent usage include any units that wouldbe excluded by the definitions we propose here.The chief difference in recent usage and the usageproposed here is that our terminology is applied toa wider area, including the entire rock-strati-graphic columns preserved in the Crowburg,Wadesboro, Ellerbe, Sanford, Durham, Scotts-burg, Randolph, Roanoke Creek, Briery Creek,Farmville, Richmond, Flat Branch, Deep Run,Taylorsville, Davie County, Dan River–Danville,Scottsville, Barboursville, Cherry Brook, North-field, and Middleton basins (Fig. 5). In addition,the lower units in the Culpeper, Gettysburg,Newark, Pomperaug, Hartford, Deerfield, andFundy basins are also within the definition of thisgroup as proposed herein (Fig. 3). However, if Lu-cas and Huber (1993) are correct in suggestingthat the Lower Economy beds of the WolfvilleFormation (Fundy basin) are separated from therest of the Wolfville by an unconformity repre-senting about 10 m.y., those beds may deserve tobe named as a separate formation and that forma-tion possibly should be excluded from the defini-tion of the Chatham Group. The thickest sectionof the Chatham Group occurs in the Newarkbasin, where the Stockton, Lockatong, and Pas-saic Formations together are about 6000 m thick(Witte et al., 1991).

One stratigraphic problem must be overcomebefore the term Chatham Group can be applied tothe Gettysburg basin. The currently defined Get-tysburg Formation mostly includes strata thatshould be assigned to the Chatham Group as de-fined here. However, the highest part of the Get-tysburg Formation includes the basalt at Aspers

(Cornet, 1977) and an overlying sequence of sed-imentary rocks (Luttrell, 1989). These two units(basalt and overlying sedimentary rocks) do notfall within our definition of the Chatham Group(Fig. 3). Therefore the Gettysburg Formation, ascurrently defined, would cross the boundary be-tween two groups in violation of the North Amer-ican Stratigraphic Code (North American Com-mission on Stratigraphic Nomenclature, 1983).For this reason, the term Gettysburg Formation isrestricted here to the rocks below the basalt at As-pers. The basalt at Aspers of Cornet (1977) is for-malized as the Aspers Basalt, and sedimentaryrocks above the Aspers are placed in the Bender-sville Formation. These units are defined in theStratigraphic Revision of the Gettysburg Forma-tion section herein. The Aspers and Bendersvillebelong in the Meriden Group, which is also de-fined herein.

Meriden Group (Lower Jurassic,Hettangian Stage)

The Meriden Formation was proposed by Kry-nine (1950) for lava flows and interbedded sedi-mentary rocks in the Hartford basin. The flowsand interbedded sedimentary rocks were namedindividually by Lehmann (1959), and Sanders(1968) clustered these new units by raising theMeriden to group rank. It is proposed that thisgroup name be applied throughout the basins ofthe Newark Supergroup for correlative portions ofthe rock column that include extrusive lava flowsand interbedded sedimentary rocks. As here de-fined, the Meriden Group occurs in the Culpeper,Gettysburg, Newark, Pomperaug, Hartford, Deer-field, and Fundy basins (Fig. 3). The MeridenGroup achieves its greatest known thickness(more than 2000 m) in the Culpeper basin (Leeand Froelich, 1989).

Stratigraphic work by Olsen et al. (1996) indi-cates that the entire Meriden Group was de-posited over a time interval of 550 ± 50 k.y.shortly after the beginning of the Jurassic Period.This places the Meriden Group entirely withinthe Hettangian stage of the Jurassic. Unfortu-nately, the absolute age of the group is not knownprecisely. All of the flow basalts have been moreor less hydrothermally altered (Seidemann et al.,1984), possibly at about 175 Ma (Sutter, 1988).The alteration has caused the flows to yield ascattering of apparent ages somewhat youngerthan their true age. The best age estimates ob-tained so far are U-Pb dates of 201 ± 1 Ma de-rived from the Palisades and Gettysburg sills(Dunning and Hodych, 1990), 40Ar/39Ar plateaudates around 201 ± 1 Ma from diabases in theCulpeper basin (Sutter, 1988), and 40Ar/39Arplateau dates around 196 ± 1 Ma derived from



WEEMS AND OLSEN


K-feldspar cements formed from hydrothermalfluids associated with Meriden age-equivalent ig-neous activity (Kunk et al., 1995).

In the Newark and Hartford basins, both thebottom and top of the Meriden Group are markedby basaltic lava flows (Fig. 3). In the Culpeper,Gettysburg, Pomperaug, and Fundy basins, thebase of the Meriden Group is marked by the baseof the lowest basaltic lava flow and the top ismarked by the modern erosional surface. Thehighest stratigraphic unit in the Culpeper basin,the Waterfall Formation, in the past has been cor-related with units above the Meriden Group (Leeand Froelich, 1989; Luttrell, 1989), but recentgeochemical work indicates that its stratigraphicposition is lower than previously assumed. For

this and other reasons, detailed herein in a sepa-rate section, all of the upper Culpeper basin col-umn is assigned to the Meriden Group. Similarly,although the highest strata in the Pomperaugbasin were formerly correlated with the PortlandFormation (Luttrell, 1989), recent work by Huberand McDonald (1992) indicates that the highestbeds are instead correlative with the East BerlinFormation and thus should be retained within theMeriden Group.

In the Deerfield basin, the lower part of theMeriden Group appears to be missing at an un-conformity. Although Luttrell (1989) assumedthat the Triassic Sugarloaf Formation in the Deer-field basin extended upward continuously to thebase of the Deerfield Basalt, the Sugarloaf ap-

pears to be restricted to rocks below a local basin-wide unconformity (Fig. 3) that truncates the topof the Chatham Group (J. P. Smoot, cited in Ol-sen, 1997, p. 380). The lowest part of the overly-ing Meriden Group is also missing. Beds that liebetween this local basinwide unconformity andthe Deerfield Basalt have been termed the FallRiver beds by Olsen et al. (1992). Above theDeerfield Basalt, the relationship between theTurners Falls Sandstone and the Mount Toby For-mation was originally described as intertonguing(Willard, 1951, 1952) and later as unconformable(Cornet, 1977, p. 214–221; Robinson and Lut-trell, 1985). Olsen et al. (1992, p. 528) docu-mented interfingering relationships between Tur-ners Falls and Mount Toby rocks, which indicate

Figure 4. Rock columns (dark horizontal pattern) specifically placed in the former Chatham Series by Emmons (1857). Formal uses of the ChathamGroup in other specific parts of the Newark Supergroup by later workers (Olsen et al., 1982; North Carolina Geological Survey, 1985) are shown byvertically ruled areas of columns. The smaller basins were poorly known or unknown in Emmons’day, thus the absence of commentary by him can-not be construed to mean that he would have excluded rocks in those basins from his Chatham Series.



that no unconformity exists between the twounits. The combined thickness of the TurnersFalls Sandstone and the Mount Toby Formationmay be only about 250 m (Willard, 1951, 1952),and this is less than the thickness of the EastBerlin Formation that occupies the same strati-graphic position to the south in the Hartford basin(Colton and Hartshorn, 1966). Because the upperpart of the Deerfield column appears to be thin,and because there is no hint of the HampdenBasalt or Granby Tuff, which are well developedonly 20 km south of the Deerfield basin, it may bethat no Deerfield basin strata are as young as thePortland Formation as suggested by Luttrell(1989). Therefore, both the Turners Falls Sand-stone (type area) and the Mount Toby Formationhere are retained within the Meriden Group.Above the Turner Falls Sandstone type area, a

covered interval occurs below other cyclic lakestrata that are either fault-repeated portions of theTurner Falls Sandstone or strata that correlatewith the Portland Formation of the AgawamGroup (Olsen, 1997, p. 380). If these strata belongwith the Agawam Group, they need to be namedand a lower boundary established.

The definition of the Meriden Group cannotinclude the numerous early Mesozoic dikes andsills within the Newark Supergroup basins or inthe country rock around them, because the NorthAmerican Stratigraphic Code (North AmericanCommission on Stratigraphic Nomenclature,1983) makes no provision for combining plu-tonic igneous terminology with extrusive vol-canic and/or sedimentary terminology. There-fore, names proposed for different compositionsof diabase (York Haven, Rossville, and Quarry-

ville) by Smith et al. (1975) cannot be properlyincluded within the Meriden Group, even thoughthe York Haven type of diabase has been associ-ated strongly with the lavas of volcanic interval Iof the Meriden Group, and the Rossville type hasbeen associated strongly with the lavas of vol-canic interval II (Puffer, 1992; Tollo and Got-tfried, 1992). Although the Quarryville type ofdiabase does not have a chemistry that matchesthe chemistry of any flow yet known from an ex-posed basin (Puffer, 1992), there seems little rea-son to doubt that all of the early Mesozoic dia-base dikes and sills were formed during the samemagmatic episode that produced the flows.Therefore, the dikes and sills appear to be genet-ically connected with and time correlative withthe Meriden Group (Olsen et al., 1996), eventhough they cannot be placed formally within it.

Figure 5. Group terminology proposed here for the Newark Supergroup. Compare with Figure 2 for previous group usages. Shading patternsreflect group placements as defined in this paper.

Agawam Group (New Name) (LowerJurassic, Hettangian and Sinemurian? Stages)

This name is proposed to include all NewarkSupergroup formations stratigraphically abovethe highest Newark Supergroup lava flows. Theincluded units are the Boonton Formation (New-ark basin), and the Portland Formation (Hartfordbasin) (Fig. 3). No stratigraphic name exists thatincorporates the concept of the regional deposi-tional event that occurred after volcanism andplutonism ceased. Therefore, the name AgawamGroup is proposed to include the two formationsthat lie directly above the extrusive volcanicrocks included within the Meriden Group. TheAgawam type area is in the northern part of theHartford basin in Agawam Township, HampdenCounty, Massachusetts. Only the strata that lieabove (east of) the Hampden Basalt are included(Fig. 6). Exposures along the Westfield River eastof the Hampden Basalt outcrop, especially theexcellent exposures east and west of BridgeStreet along the north border of the town of NorthAgawam, constitute a reference section for thisunit. The location of this reference section wasshown in Colton and Hartshorn (1966). The topof the Agawam Group is marked everywhere bythe modern erosional surface or by a thin cap ofsurficial Quaternary deposits. The greatest re-ported thickness of the Agawam Group (about2500 m) is in the Hartford basin (Hubert et al.,1978), but there could be as much as 5000 m ofsection present in that basin.

The Agawam Group is Early Jurassic in age.Cornet (1977) suggested that the upper part ofthis group could be as young as Middle Jurassic(Bathonian), but such an extensive age range wasbased on very tenuous evidence. A more recentestimate of the age of the Boonton Formationplaces it entirely within the Hettangian Stage(Olsen and Kent, 1996). Pollen, vertebrate fos-sils, and ichnofossils from the lower to middlepart of the Agawam Group (specifically the Port-land Formation) are very similar to fossils fromthe lower to middle part of the Moenave Forma-tion of the Glen Canyon Group on the ColoradoPlateau (Olsen and Padian, 1989). The KayentaFormation, which probably lies unconformablyabove the Moenave, so far has yielded no pollenbut contains many vertebrate fossils and ichno-fossils (such as Dilophosauripus, Hopiichnus,and Kayentapus hopii) (Welles, 1971; Clark andFastovsky, 1989; Padian, 1989), most of whichare unknown from the Agawam Group. Some ofthese Kayenta fossils clearly are derived in theircharacter states relative to related (but more prim-itive) fossil forms found in the Agawam Group.Thus there is no compelling evidence for corre-lating the Agawam Group with strata on the Col-orado Plateau any higher than the Moenave For-

mation. Worldwide comparison of vertebrate re-mains from the overlying Kayenta Formation in-dicates that it is Sinemurian to Pliensbachian inage (Padian, 1989). This in turn constrains theage of the Moenave Formation (and the AgawamGroup) to probably no younger than Sinemurian,as shown in Olsen (1997, p. 341).

The above defined groups and their distributionin the various early Mesozoic rift basins are

shown in Figures 3 and 5. This system of threegroups is much simpler to use than the existingsystem of nine groups; it is fully inclusive of theentire Newark Supergroup and has more preciselyand easily defined boundaries than the groupnomenclature currently in use. The proposedgroup nomenclature mirrors the fundamentalthreefold tectonostratigraphic and lithostrati-graphic stages through which the Newark Super-

WEEMS AND OLSEN


Figure 6. Map of the NewarkSupergroup basins in the Con-necticut Valley, showing the loca-tion of the Northfield, Deerfield,and Hartford basins, AgawamTownship, Massachusetts (verti-cal rule), and the distribution ofthe Lower Jurassic AgawamGroup (light gray). In the Hart-ford basin,Agawam rocks are thePortland Formation. The typearea for the Agawam Group isthe light gray dot-patterned por-tion of Agawam Township east of(stratigraphically directly above)the Hampden Basalt. Obliquelyruled areas are underlain byrocks of the Chatham and Meri-den Groups.



group evolved, associating a lower package of al-luvial fan–fluvial–lacustrine continental depositsthat are locally injected by diabase and thermallymetamorphosed, a medial package of thin to thickextrusive volcanic basalt flows that are inter-spersed with volumetrically subordinate packagesof alluvial fan–fluvial–lacustrine continental de-posits locally injected by diabase and thermallymetamorphosed, and an upper package of alluvialfan–fluvial–lacustrine continental deposits unaf-fected by volcanism. Our proposed system ofgroup names is simple, inclusive, defined by read-ily observable and mappable boundaries, and ac-curately reflects one of the most important tec-tonic events recorded by the Newark Supergroup.Because our group names incorporate the rockcolumns included in the Dan River Group, Tucka-hoe Group, Chesterfield Group, Culpeper Group,Brunswick Group, Hartford Group, and FundyGroup, these terms are abandoned herein.

STRATIGRAPHIC REVISION OF THEGETTYSBURG FORMATION

Gettysburg Formation

The Gettysburg Formation was defined as allstrata in the Gettysburg basin from the top of theNew Oxford Formation to the top of the pre-served basin column (Stose and Bascom, 1929).The upper boundary of the Gettysburg Formationhere is moved downward to the base of the basaltat Aspers of Cornet (1977), thereby excluding thebasalt and the sedimentary rocks overlying thebasalt from the Gettysburg Formation. The lowerboundary of the Gettysburg Formation and itscomposite type section remain unchanged. Stratanow excluded from the Gettysburg Formation arenamed formally below, although they are not yetapproved by the North American Commission onStratigraphic Nomenclature.

Aspers Basalt (New Name) (Lower Jurassic)

Part of this unit was recognized as a flow byStose and Bascom (1929), but its full areal extentand geochemistry were not established until thework of Smith et al. (1975, p. 948), who referredthis unit to their York Haven type diabase. Al-though the Aspers Basalt is geochemically equiv-alent to York Haven diabase, the type locality ofthe York Haven is a plutonic body. Therefore, al-though York Haven is the valid name for earlyMesozoic diabase dikes and sills that have a high-titanium quartz normative composition, the NorthAmerican Code of Stratigraphic Nomenclature(North American Commission on StratigraphicNomenclature, 1983, p. 848, art. 22, art. 31) doesnot provide for extending a lithodemic rock namesuch as York Haven diabase to an extrusivebasaltic flow in a sedimentary sequence, even if

Figure 7. Areal distribution ofthe Aspers Basalt and Bender-sville Formation in the Gettysburgbasin of Pennsylvania. Type areasof both units lie in the small syn-cline immediately west of Aspers.Dot pattern—pre-Triassic rocks(Late Proterozoic-lower Paleo-zoic); dashed horizontal rule—New Oxford Formation (UpperTriassic); white—Gettysburg For-mation (Lower Jurassic and Up-per Triassic); oblique rule (Jdy)—York Haven type diabase dikesand sills (Lower Jurassic); cross-hachured (Jdr)—Rossville typediabase dikes and sills (LowerJurassic); vertical rule (Ja)—As-pers Basalt (Lower Jurassic);black (Jb)—Bendersville Forma-tion (Lower Jurassic).

WEEMS AND OLSEN


both came from a common magma source. There-fore, the term York Haven is unavailable as a for-mal stratigraphic name for any basalt flow unit,and the name Aspers Basalt, derived from the in-formal name “basalt at Aspers” used by Cornet(1977), is here adopted for the basalt-flow rem-nants west of Heidlersburg. The type area is l kmwest of Aspers, and the unit is estimated to beabout 60 m thick (Cornet, 1977). The lowerboundary of this unit is defined as the contact be-tween the lowest basalt and the uppermost (ther-mally metamorphosed) sedimentary rocks of theGettysburg Formation as restricted above. Its up-per boundary is defined as the contact between thehighest basalt and the overlying sedimentaryrocks of the Bendersville Formation, definedbelow.

On the basis of its position (about 50 m abovethe base of the Corollina meyeriana palynofloralzone of Cornet, 1977), and on its high-titaniumquartz normative composition, the Aspers Basalt islaterally equivalent to the Mount Zion ChurchBasalt in the Culpeper basin and the Jacksonwald(= Orange Mountain) Basalt in the Jacksonwaldsyncline of the Newark basin (Fig. 3). It is inadvis-able to assign these flow remnants to either theMount Zion Church or Jacksonwald and/or Or-ange Mountain Basalts, because of the great dis-tance between the Aspers basalt and its correla-tives in adjacent basins and because each basin hasa different stratigraphic column for all other units.

Bendersville Formation (New Name)(Lower Jurassic)

The Bendersville Formation is named for thetown of Bendersville, Pennsylvania (Biglerville7.5-minute quadrangle), which is about 1 kmnorthwest of a small syncline 1.5 km west of As-pers that contains the type area of this unit. TheBendersville Formation includes all sedimentaryrocks directly above the Aspers Basalt (definedabove), both in the type area syncline near Aspersand along the western basin border fault 9 km tothe southwest of Aspers (Arendtsville 7.5-minutequadrangle) (Fig. 7). The top of this unit is trun-cated by the modern erosional surface. Only about230 m of stratigraphic section remain (Cornet,1977). On the basis of its stratigraphic position justabove volcanic interval I, the Bendersville Forma-tion is laterally equivalent to the basal MidlandFormation in the Culpeper basin and the basalFeltville Formation in the Newark basin. The greatdistance (>100 km) between this part of the Get-tysburg basin and both of the laterally equivalent,previously named units precludes obvious assign-ment to either one. Near the northwestern borderfault of the Gettysburg basin, the Bendersville is

composed of fanglomerates with quartzitic tometarhyolitic clasts. These rock types originallywere mapped as part of the Arendtsville Fanglom-erate Lentil by Stose and Bascom (1929), but theconglomerates above the Aspers Basalt have beenexcluded from the definition of the Arendtsville(Luttrell, 1989, Plate 1). We also exclude the Ben-dersville conglomerates from the ArendtsvilleLentil, so that each of these units remains withingroup boundaries. Away from the border fault, theBendersville Formation grades southeastward intopollen-bearing, olive-green, thickly bedded, clay-ey siltstones (Cornet, 1977) that apparently formedin lacustrine to paludal depositional environments.

REINSTATEMENT OF THE BULL RUNFORMATION

The “Bull Run Shales” were named as a strati-graphic unit in the Culpeper basin by Roberts(1928). He did not map out the areal extent of thisunit, but he did establish a type area along “BullRun, a small stream between Prince William andFairfax counties. Bull Run battlefield, namedfrom this stream, lies about 9.5 km (6 miles) duewest of Manassas. Almost the only rocks out-cropping over this region are the Bull Runshales” (Roberts, 1928, p. 39; Fig. 8). Robertsalso listed supplementary outcrops for the BullRun, “the most northern being on the bluffs of thePotomac River 31/2 miles [about 5.5 km] east ofLeesburg.…” Lee (1977) later used the term BullRun Formation north of its type area, but he alsodefined a new stratigraphic unit, the Balls BluffSiltstone, which he named for the same bluff thatRoberts listed as the northernmost exposure ofthe Bull Run shales. Lee (1977) stopped mappingjust north and east of the type area of the BullRun Formation, but projection of his Balls BluffSiltstone along strike places it across the entiretype area of the Bull Run Formation. Because theBalls Bluff Siltstone as mapped by Lee includesthe type area of the Bull Run Formation of Rob-erts, and because Lee only nominally retained thename Bull Run Formation for beds not correla-tive with the type area of that unit, Lee’s strati-graphic revisions were not valid according to theNorth American Code of Stratigraphic Nomen-clature (North American Commission on Strati-graphic Nomenclature, 1983, article 22c). Ac-cordingly, Lindholm (1979) abandoned the BallsBluff Siltstone and used the term Bull Run For-mation for the entire stratigraphic column fromthe top of the Manassas Sandstone up to the baseof the lowest basaltic lava flow exposed in theCulpeper basin (Lindholm, 1979, Fig. 1). Subse-quently, Lee and Froelich (1989) abandoned theBull Run Formation altogether, reinstating the

Balls Bluff Siltstone for the lower part of the BullRun column and creating the Catharpin CreekFormation to include its upper part.

The rocks included within the Catharpin CreekFormation lie above the type area of the Bull RunFormation and are dominantly sandstones ratherthan shales or siltstones. Therefore, it was rea-sonable for Lee and Froelich to exclude thesestrata from the Bull Run Formation. However,they were not justified in replacing the Bull RunFormation in its type area and outcrop belt withthe Balls Bluff Siltstone, because the term BullRun was applied validly there and has priority by60 years. Thus abandonment of the Bull Run For-mation was unjustified, and the name herein is re-instated as the valid name for the stratigraphic in-terval between the Manassas Sandstone and theCatharpin Creek Formation (Fig. 9). The BullRun Formation is composed of the following fivemembers.

Balls Bluff Member

The Balls Bluff Siltstone of Lee (1977) is re-tained, but it is reduced in rank to a member andrestricted to rocks within the Bull Run Formationthat are similar to those of the Balls Bluff typesection (Figs. 8 and 9). Because the Balls BluffMember includes significant volumes of shaleand fine-grained sandstone, especially near thewestern border faults of the Culpeper and Bar-boursville basins, “siltstone” is not retained aspart of the unit name. This member typically isthick-bedded or massive, and probably formed insluggish fluvial and overbank environments ofdeposition (Sobhan, 1985). The Balls Bluff is thelowest unit of the Bull Run Formation throughoutthe Culpeper and Barboursville basins. In mostareas it underlies the Groveton Member (definedbelow), but in the northern part of the Culpeperbasin the Balls Bluff intertongues with theGroveton and underlies the Leesburg Member(Fig. 9). The maximum thickness of this unit isabout 900 m.

Groveton Member (New Name)

The Groveton Member, herein named forGroveton in the Manassas National BattlefieldPark (Gainesville 7.5-minute topographic map),includes that part of the Bull Run Formation con-taining prominent, laterally persistent, graycyclic lacustrine sequences. Its type area is theManassas National Battlefield Park, excludingareas underlain by intrusive diabase (Fig. 8).Good exposures in the Manassas National Battle-field Park occur along U.S. Highway 50 and Vir-ginia State Road 234, and along the south bank of



Bull Run Creek along the northern edge of thepark. The Groveton Member is characterized byrelatively thin (1–10 m) sequences of gray shalesinterbedded in a highly rhythmic pattern withmuch thicker (6–60 m) sequences of red shales,siltstones, and occasionally sandstones. Theserocks apparently formed in lacustrine to playaflat depositional environments (Sobhan, 1985;Gore, 1988a, 1988b; Smoot and Olsen, 1988).The Groveton Member is dominated by shales

and siltstones, but where it intertongues with fan-glomerates along the western border fault of thebasin it locally includes sandstones and con-glomerates within the red intervals between thegray shales. The base of the Groveton is the baseof the lowest cyclic lacustrine red or gray shalepresent in the Bull Run column; its top is at thetop of the highest gray lacustrine shale separatedby no more than 60 m of red siltstone from thegray shales below it. Toward the northern end of

the Culpeper basin, the Groveton grades laterallyin its lower part into the Balls Bluff Member andin its upper part into the Leesburg Member(Fig. 9). This unit may be 2500 m thick.

Leesburg Member

The Leesburg Limestone Conglomerate Mem-ber was defined by Lee (1977), shortened to theLeesburg Conglomerate Member by Lindholm

Figure 8. The type area of the Bull Run Formation and the Groveton Member of the Bull Run Formation in the Culpeper basin of Virginia.Dashed vertical rule—Manassas Sandstone, Poolesville Member (Upper Triassic); dot pattern—Bull Run Formation, Balls Bluff Member (Up-per Triassic); white—Bull Run Formation, Groveton Member (Upper Triassic); oblique rule—quartz normative, high-titanium diabase (YorkHaven type, Lower Jurassic); cross-hachured—quartz normative, low-titanium diabase (Rossville type, Lower Jurassic); horizontal rule—Up-per Triassic rocks thermally metamorphosed by injected Lower Jurassic diabase.

WEEMS AND OLSEN


(1979), and shortened further to the LeesburgMember by Lee and Froelich (1989). The unit ishere retained in the Bull Run Formation asmapped by Lindholm (1979) and Lee and Froe-lich (1989). The unit only occurs in the northernportion of the Culpeper basin. The LeesburgMember, dominated by clasts of limestone anddolostone, is as thick as 1100 m (Lee and Froe-lich, 1989); it overlies the Balls Bluff Memberand interfingers laterally with the upper part ofthe Groveton Member (Fig. 9).

Cedar Mountain Member

The Cedar Mountain Member of the Bull RunFormation was named by Lindholm (1979) fordominantly greenstone conglomerates that occurin the southwestern portion of the Culpeperbasin in the vicinity of Cedar Mountain, 14 km

southwest of Culpeper. This same body of con-glomerates was renamed the Mountain RunMember by Lee and Froelich (1989), eventhough the name Cedar Mountain Member wasvalidly applied by Lindholm (1979) and has pri-ority. Therefore, the name Mountain Run Mem-ber is abandoned and the name Cedar MountainMember is reinstated as a member of the BullRun Formation. The Cedar Mountain Member isas thick as 640 m.

Haudricks Mountain Member

The Haudricks Mountain Member, hereplaced in the Bull Run Formation, was named byLee and Froelich (1989) for conglomerates in theBarboursville basin in the vicinity of HaudricksMountain that have sandstone, quartzite, andfine-grained metasiltstone clasts interbedded

with fine- to coarse-grained arkosic sandstone.The unit grades laterally into the Balls BluffMember of the Bull Run Formation (Fig. 9). TheHaudricks Mountain Member can be as thick as500 m (Lee and Froelich, 1989).

Tibbstown Formation (Abandoned)

When Lee and Froelich (1989) named theTibbstown Formation, they assumed that therocks assigned to this unit had a northerly strikeand lay above rocks here assigned to the BullRun Formation. However, detailed mapping inthe Culpeper East quadrangle by R. E. Weems re-veals that Tibbstown rock types strike northeastto eastward and grade laterally into rocks typicalof the Balls Bluff and Groveton Members of theBull Run Formation. The outcrop pattern is verycomplex and impractical to map (Fig. 9). There-

Figure 9. Diagrammatic repre-sentations of the stratigraphy ofstrata cropping out within theCulpeper and Barboursville ba-sins above the Manassas Sand-stone and (in the Culpeper basin)below the Catharpin Creek For-mation. Nomenclature of Lee andFroelich (1989) is shown at bot-tom, nomenclature adopted hereis shown at top. Conglomeraticmembers are shown by dark zig-zag pattern. Location of the aban-doned Tibbstown–Balls Bluffboundary is shown in the uppercolumn by a dashed line.

fore, the name Tibbstown Formation is aban-doned and its constituent parts are incorporatedinto the Bull Run Formation.

AGE AND GROUP ASSIGNMENT OFTHE WATERFALL FORMATION

The Waterfall Formation, highest stratigraphicunit in the Culpeper basin, formerly was corre-lated with the Boonton and Portland Formations(Lee and Froelich, 1989; Luttrell, 1989) of theAgawam Group. This correlation was based inpart on a comment by Cornet (1977, p. 260) thathis Corollina torosa palynozone (defined fromthe middle and upper beds of the Portland For-mation) “could be represented … in the youngeststrata of the Culpeper Group.” Although Cornet(1977, p. 167) had a sample locality in thesestrata (MBK at Millbrook Quarry), he listed nopollen or spore taxa from this locality and in Ap-pendix I only noted that it “produces poorly pre-served palynomorphs” (p. 447). Elsewhere hestated, “particularly significant would be the ab-sence of Redfieldius in the Millbrook Quarry fishbed if that part of the section were younger thanthe highest stratigraphic occurrence of Redfield-ius spp. Unfortunately, the palynoflorules fromthe two fish beds do not provide any solution tothis problem” (p. 133), and that “the Corollinatorosus palynoflora, may also be represented inthe Culpeper and Greenfield Groups, based onstratigraphic relationships” (p. 247). Thus Cornetprovided no palynological evidence to supporthis suggestion that sample MBK might representhis Corollina torosa palynozone and implied intwo places that his assignment was based on evi-dence other than pollen or spores. Therefore,none of the Waterfall Formation has been demon-strated to be as young as the Corollina torosapalynozone.

Another argument for correlating the WaterfallFormation with the Agawam Group was based onthe assumption that the Sander Basalt (the thirdflow sequence recognized in the Culpeper basin)correlates with the Hook Mountain and Hamp-den basalts (the third and highest flow sequencesin the Newark and Hartford basins). However,geochemical work on these basalts (Tollo andGottfried, 1992; Hozik, 1992) demonstrates thatthe Sander Basalt correlates with the Preakness,Holyoke, and Deerfield Basalts in the Newark,Hartford, and Deerfield basins, and not with theHook Mountain and Hampden Basalts. This im-plies that the Waterfall Formation likely corre-lates with the Towaco Formation, the East BerlinFormation, and the Turners Falls Sandstone and/or Mount Toby Formation (Fig. 3), rather thanwith the Boonton and Portland Formations aspreviously assumed. Moreover, preliminary work

on the cyclic stratigraphy of the Waterfall Forma-tion (Olsen, 1997, p. 381) supports this view, be-cause the cyclicity within the Waterfall Forma-tion is most similar to that of the Towaco and EastBerlin Formations.

The Waterfall Formation is cut in its lower partby an olivine normative diabase dike (Gottfried etal., 1991, U.S. Geological Survey core holeOpal #1, p. 15 and plate 1), demonstrating thatdeposition of the lower part of this unit precededthe end of early Mesozoic igneous activity. In ad-dition, Hentz (1985) mapped two fine-grainedbasaltic bodies along the western basin borderfault near Beulah Church and Broad Run (Thor-oughfare Gap 7.5-minute quadrangle), which heinterpreted as lava flows capping the youngeststrata of the Waterfall Formation. The body nearBroad Run is stratigraphically discordant (Hentz,1985, Fig. 2), suggesting that it might be a dike,but either interpretation places volcanic rocksthroughout the column of the Waterfall Forma-tion and thus excludes the Waterfall Formationfrom the definition of the Agawam Group.

An alternative explanation of the two areas ofbasaltic rock near Thoroughfare Gap was offeredby Lee and Froelich (1989, p. 27), who inter-preted them as slivers of the Sander Basalt re-peated by faulting within the western border faultcomplex of the Culpeper basin. If this proves tobe true, then the highest preserved strata of theWaterfall Formation are not capped by basalts asHentz thought, and diabase or basalt are not dem-onstrably present in the highest strata of the Wa-terfall Formation. Although plausible, this argu-ment requires an impressive throw of 1500 malong the obscure fault that separates the basaltslivers in question from the adjacent upper Wa-terfall Formation. Such an impressive throw onan obscure fault tends to imply that other suchfaults probably exist, which in turn implies thatthe estimated thickness of the Waterfall Forma-tion may be much too great.

In the absence of geochemical data, either in-terpretation remains plausible. The dike cited byGottfried et al. (1991) still documents igneousrocks cutting the older parts of the Waterfall For-mation, and this still supports correlation of thelower part of the Waterfall Formation with theEast Berlin and Towaco Formations. BecauseHentz (1985) described angular unconformitiesnear the top of the Waterfall section, it remainspossible that the stratigraphic horizon equivalentto the Hampden and Hook Mountain Basalts islost within one of these unconformities. If so, thehighest strata of the Waterfall Formation may be-long in the Agawam Group and may need to bemapped and named separately. At our presentlevel of understanding, however, this is specula-tive. Therefore, we retain all of the Waterfall For-

mation in the Meriden Group until paleontologi-cal, geochemical, and/or stratigraphic studiesshow convincingly that any Agawam-equivalentstrata are present in the Culpeper basin.

SUMMARY

The rock columns of the early Mesozoic riftbasins of eastern North America collectively havebeen termed the Newark Supergroup, because theyshare a common tectonic history that is distinctlydifferent from other North American rock se-quences. In contrast, disparate parts of this super-group have been combined into nine groups thatneither provide an inclusive stratigraphic frame-work for the entire supergroup nor cluster the in-cluded formations in a manner that optimally re-flects their lithic contrasts and similarities. For thisreason, we propose a stratigraphic reorganizationof the groups within the Newark Supergroup thatemphasizes stratigraphic and tectonic elementscommon to the rock columns of all basins. Twoexisting group names are retained: the ChathamGroup, which is expanded to apply to strata in allbasins from the base of each stratigraphic columneither to the top of the preserved column, to thebase of the lowest preserved lava flow, or to an un-conformity encompassing the basal flow horizon;and the Meriden Group, which is retained in itsoriginal sense but expanded to apply to all litho-logically similar (and age-equivalent) Newark Su-pergroup lava flows and strata interbedded be-tween flows in other basins. Thus defined, theMeriden Group now includes the lavas and in-terbedded sedimentary rocks contained in theCulpeper, Gettysburg, Newark, Pomperaug, Hart-ford, Deerfield, and Fundy basins. The name Aga-wam Group is proposed here to include all stratahigher than the flow-bearing Meriden interval inthe two basins where such strata are preserved(Newark and Hartford basins). Seven other groupnames (Brunswick Group, Chesterfield Group,Culpeper Group, Dan River Group, HartfordGroup, Fundy Group, and Tuckahoe Group), pre-viously applied in local areas of the eastern NorthAmerican early Mesozoic rift basins, are hereinabandoned as stratigraphically redundant. As heredefined, the Chatham Group ranges in age fromMiddle Triassic to earliest Early Jurassic. TheMeriden Group and the Agawam Group are bothEarly Jurassic in age.

Unlike the ninefold group nomenclature previ-ously applied to parts of the Newark Supergroup,our threefold group nomenclature places all New-ark Supergroup strata into groups and establishesunambiguous boundaries between groupsthroughout all of the eastern North American earlyMesozoic rift basins. Our rock classification sys-tem makes use of the fact that a discrete episode of



WEEMS AND OLSEN


synchronous or nearly synchronous volcanism oc-curred throughout this early Mesozoic rift system.Thus the presence or absence of extrusive volcanicrocks (and their genetically related diabase dikesand sills) provides a powerful stratigraphic tool forestablishing regionally applicable groups andgroup boundaries, allows us to create parallelismbetween stratigraphic groups and some of the ma-jor tectonic events that created the eastern NorthAmerican rift basins, and coincidentally providesisochronous or nearly isochronous group bound-aries within the Newark Supergroup.

ACKNOWLEDGMENTS

We thank Tyler Clark (North Carolina Geolog-ical Survey), Jack B. Epstein (U.S. GeologicalSurvey [USGS]), Rodger T. Faill (PennsylvaniaGeological Survey), Pamela J. W. Gore (DeKalbCollege), Charles W. Hoffman (North CarolinaGeological Survey), Phillip Huber (University ofBridgeport), Stanley S. Johnson (Virginia Divi-sion of Mineral Resources), Elizabeth D. Kooz-min (USGS), Don Monteverde (New Jersey Geo-logical Survey), Randall C. Orndorff (USGS),Ronald A. Parker (USGS), Gene Rader (VirginiaDivision of Mineral Resources), James P. Reger(Maryland Geological Survey), Charles L. Rice(USGS), Eleanora I. Robbins (USGS), Joseph P.Smoot (USGS), C. Scott Southworth (USGS),and Richard P. Tollo (George Washington Univer-sity) for their insightful reviews and comments.

REFERENCES CITED

Clark, J. M., and Fastovsky, D. E., 1989, Vertebrate biostratigraphy of theGlen Canyon Group in northern Arizona, in Padian, K., ed., Thebeginning of the age of dinosaurs; faunal change across the Trias-sic-Jurassic boundary: New York, Cambridge University Press,p. 285–301.

Colton, R. B., and Hartshorn, J. H., 1966, Bedrock geologic map of theWest Springfield quadrangle, Massachusetts and Connecticut:U.S. Geological Survey Geological Quadrangle Map GQ-537,scale 1:24 000.

Cornet, B., 1977, The palynostratigraphy and age of the Newark Super-group [Ph.D. dissert.]: University Park, Pennsylvania State Uni-versity, 504 p.

Cornet, B., and Olsen, P. E., 1985, A summary of the biostratigraphy ofthe Newark Supergroup of eastern North America with commentson early Mesozoic provinciality, in Weber, R., ed., Simposio sobrafloras del Triasico Tardio su fitogeografia y paleoecologia, III Con-gresso Latinoamericano de Paleontologia, Mexico: UniversidadNacional Autle Map GQ. Tollo (George Washington University)for 67–81.

Dostal, J., and Greenough, J. D., 1992, Geochemistry and petrogenesisof the early Mesozoic North Mountain basalts of Nova Scotia,Canada, in Puffer, J. H., and Ragland, P. C., eds., Eastern NorthAmerican Mesozoic magmatism: Geological Society of AmericaSpecial Paper 268, p. 149–160.

Dunay, R. E., and Fisher, M. J., 1974, Late Triassic palynofloras of NorthAmerica and their European correlatives: Reviews in Palaeob-otany and Palynology, v. 17, p. 179–186.

Dunning, G. R., and Hodych, J. P., 1990, U/Pb zircon and baddeleyiteages for the Palisades and Gettysburg sills of the northeasternUnited States: Implications for the age of the Triassic/Jurassicboundary: Geology, v. 18, p. 795–798.

Emmons, E., 1857, American geology (part 6): Albany, New York,Sprague and Company, 152 p.

Fowell, S. J., 1993, Palynology of Triassic/Jurassic boundary sectionsfrom the Newark Supergroup of Eastern North America—Impli-cations for catastrophic extinction scenarios [Ph.D. dissert.]: New

York, Columbia University Press, 133 p.Fowell, S. J., and Olsen, P. E., 1993, Time-calibration of Triassic/Jurassic

microfloral turnover, Eastern North America: Tectonophysics,v. 222, p. 361–369.

Fowell, S. J., Cornet, B., and Olsen, P. E., 1994, Geologically rapid LateTriassic extinctions: Palynological evidence from the Newark Su-pergroup, in Klein, G. D., ed., Pangea—Paleoclimate, tectonicsand sedimentation during accretion, zenith and break-up of a su-percontinent: Geological Society of America Special Paper 288,p. 197–206.

Froelich, A. J., and Olsen, P. E., 1984, Newark Supergroup, a revision ofthe Newark Group in eastern North America: U.S. Geological Sur-vey Bulletin 1537-A, p. A55–A58.

Gore, P. J. W., 1988a, Paleoecology and sedimentology of a Late Trias-sic lake, Culpeper basin, Virginia, U.S.A.: Palaeogeography,Palaeoclimatology, Palaeoecology, v. 62, p. 593–608.

Gore, P. J. W., 1988b, Late Triassic and Early Jurassic lacustrine sedi-mentation in the Culpeper basin Virginia, in Manspeizer, W., ed.,Triassic–Jurassic rifting; continental breakup and the origin of theAtlantic Ocean and passive margins, Part A: New York, Elsevier,Developments in Geotectonics 22, p. 369–400.

Gottfried, D., Froelich, A. J., and Grossman, J. N., 1991, Geochemicaldata for Jurassic diabase associated with early Mesozoic basins inthe eastern United States: Culpeper Basin and vicinity, Virginiaand Maryland: U.S. Geological Survey Open-File Report91-322-F, p. 1–51, 1 plate.

Hentz, T. F., 1985, Early Jurassic sedimentation of a rift-valley lake;Culpeper basin, northern Virginia: Geological Society of AmericaBulletin, v. 96, p. 92–107.

Hozik, M. J., 1992, Paleomagnetism of igneous rocks in the Culpeper,Newark, and Hartford/Deerfield basins, in Puffer, J. H., and Rag-land, P. C., eds., Eastern North American Mesozoic magmatism:Geological Society of America Special Paper 268, p. 279–308.

Huber, P., and McDonald, N. G., 1992, Revised stratigraphy and paleon-tology of the early Mesozoic Pomperaug basin (Newark Super-group), western Connecticut: Geological Society of America Ab-stracts with Programs, v. 24, no. 7, p. A357–A358.

Hubert, J. F., Reed, A. A., Dowdall, W. L., and Gilchrist, J. M., 1978,Guide to the redbeds of central Connecticut; 1978 field trip, east-ern section of S.E.P.M.: University of Massachusetts, Departmentof Geology and Geography, Contribution 32, 129 p.

Hubert, J. F., Feshbach-Meriney, P. E., and Smith, M. A., 1992, The Tri-assic–Jurassic Hartford rift basin, Connecticut and Massachusetts:Evolution, sandstone diagenesis, and hydrocarbon history: Amer-ican Association of Petroleum Geologists Bulletin, v. 76,p. 1710–1734.

Krynine, P. D., 1950, Petrology, stratigraphy, and origin of the Triassicsedimentary rocks of Connecticut: Connecticut Geological andNatural History Survey Bulletin 73, 247 p.

Kunk, M. J., Simonson, B. M., and Smoot, J. P., 1995, 40Ar/39Ar con-straints on the age of K-feldspar cementation in nonmarine sedi-ments of the Newark Gettysburg and Culpeper basins: GeologicalSociety of America Abstracts with Programs, v. 27, no. 1, p. 62.

Lee, K. Y., 1977, Triassic stratigraphy in the northern part of the Culpeperbasin, Virginia and Maryland: U.S. Geological Survey Bulletin1422-C, p. C1–C17.

Lee, K. Y., and Froelich, A. J., 1989, Triassic–Jurassic stratigraphy of theCulpeper and Barboursville basins, Virginia and Maryland: U.S.Geological Survey Professional Paper 1472, 52 p.

Lehmann, E. P., 1959, The bedrock geology of the Middletown quadran-gle: Connecticut Geological and Natural History Survey Quadran-gle Report 8, 40 p.

Lindholm, R. C., 1979, Geologic history and stratigraphy of the Trias-sic–Jurassic Culpeper basin, Virginia: Geological Society AmericaBulletin, v. 90, part 2, p. 1702–1736.

Litwin, R. J., and Ash, S. R., 1993, Revision of the biochronology of theChatham Group (Upper Triassic), Deep River basin, North Car-olina: Reviews of Palaeobotany and Palynology, v. 77, p. 75–95.

Litwin, R. J., and Weems, R. E., 1992, Re-evaluation of the age of Trias-sic strata (Doswell Formation) of the Taylorsville basin, Va.: Vir-ginia Journal of Science, v. 43, p. 265.

Litwin, R. J., Traverse, A., and Ash, S. R., 1991, Preliminary palynolog-ical zonation of the Chinle Formation, southwestern U.S.A., andits correlation to the Newark Supergroup (eastern U.S.A.): Re-views of Palaeobotany and Palynology, v. 68, p. 269–287.

Lorenz, J. C., 1988, Triassic–Jurassic rift-basin sedimentology: historyand methods: New York, Van Nostrand Reinhold Company, 315 p.

Lucas, S. G., and Huber, P., 1993, Revised internal correlation of the New-ark Supergroup Triassic, eastern United States and Canada, in Lu-cas, S. G., and Morales, M., eds., The nonmarine Triassic: NewMexico Museum of Natural History and Science Bulletin, v. 3,p. 311–319.

Luttrell, G. W., 1989, Stratigraphic nomenclature of the Newark Super-group of eastern North America: U.S. Geological Survey Bulletin1572, 136 p.

North American Commission on Stratigraphic Nomenclature, 1983,North American stratigraphic code: American Association of Pet-roleum Geologists Bulletin, v. 67, p. 841–875.

North Carolina Geological Survey, 1985, Geologic map of North Caro-lina: Raleigh, 1 sheet, scale 1:500 000.

Olsen, P. E., 1978, On the use of the term Newark for Triassic and EarlyJurassic rocks in eastern North America: Newsletters on Stratigra-phy, v. 7, p. 90–95.

Olsen, P. E., 1986, A 40-million year lake record of early Mesozoic or-bital climatic forcing: Science, v. 234, p. 789–912.

Olsen, P. E., 1988, Continuity of strata in the Newark and Hartfordbasins, in Froelich, A. J., and Robinson, G. R., Jr., eds., Studies ofthe early Mesozoic basins of the Eastern United States: U.S. Geo-logical Survey Bulletin 1776, p. 6–18.

Olsen, P. E., 1997, Stratigraphic record of the early Mesozoic breakup ofPangea in the Laurasia-Gondwana rift system: Annual Reviews ofEarth and Planetary Science, v. 25, p. 337–401.

Olsen, P. E., and Fedosh, M. S, 1988, Duration of the early Mesozoic ex-trusive igneous episode in eastern North America determined byuse of Milankovitch-type lake cycles: Geological Society of Amer-ica Abstracts with Programs, v. 20, no. 1, p. 59.

Olsen, P. E., and Kent, D. V., 1996, Milankovitch climate forcing in thetropics of Pangaea during the Late Triassic: Palaeogeography,Palaeoclimatology, Palaeoecology, v. 122, p. 1–26.

Olsen, P. E., and Padian, K., 1989, Earliest records of Batrachopus fromthe southwestern United States, and a revision of some EarlyMesozoic crocodylomorph ichnogenera; in Padian, K., ed., Thebeginning of the age of dinosaurs; faunal change across the Trias-sic-Jurassic boundary: New York, Cambridge University Press,p. 259–273.

Olsen, P. E., McCune, A. R., and Thomson, K. S., 1982, Correlation ofthe early Mesozoic Newark Supergroup by vertebrates, principallyfishes: American Journal of Science, v. 282, p. 1–44.

Olsen, P. E., McDonald, N. G., Hober, P., and Cornet, B., 1992, Stratig-raphy and paleoecology of the Deerfield rift basin (Triassic–Juras-sic, Newark Supergroup), Massachusetts, in Robinson, P., andBrady, J. B., eds., Guidebook for field trips in the Connecticut Val-ley region of Massachusetts and adjacent states: Amherst, Univer-sity Massachusetts, Department of Geology and Geography Con-tribution 66, New England Intercollegiate Geological Conference84th Annual Meeting, p. 488–535.

Olsen, P. E., Schlische, R. W., and Fedosh, M. S., 1996, 580 ky durationof the Early Jurassic flood basalt event in eastern North Americaestimated using Milankovitch cyclostratigraphy, in Morales, M.,ed., The continental Jurassic: Museum of Northern Arizona Bul-letin 60, p. 11–22.

Padian, K., 1989, Presence of the dinosaur Scelidosaurus indicates Juras-sic age for the Kayenta Formation (Glen Canyon Group, northernArizona): Geology, v. 17, p. 438–441.

Philpotts, A. R., and Reichenbach, I., 1985, Differentiation of the Meso-zoic basalts of the Hartford basin, Connecticut: Geological Soci-ety of America Bulletin, v. 96, p. 1131–1139.

Puffer, J. H., 1992, Eastern North American flood basalts in the contextof the incipient breakup of Pangea, in Puffer, J. H., and Ragland,P. C., eds., Eastern North American Mesozoic magmatism: Geo-logical Society of America Special Paper 268, p. 95–118.

Puffer, J. H., Hurtubise, D. O., Geiger, F. J., and Lechler, P., 1981, Chem-ical Composition and stratigraphic correlation of the Mesozoicbasalt units of the Newark basin, New Jersey, and the Hartfordbasin, Connecticut: Geological Society of America Bulletin, v. 92,part 1, p. 155–159.

Robbins, E. I., 1985, Palynostratigraphy of coal-bearing sequences inearly Mesozoic basins of the eastern United States, in Robinson,G. R., and Froelich,A. J., eds., Proceedings of the second U.S. Geo-logical Survey workshop on the early Mesozoic basins of the East-ern United States: U.S. Geological Survey Circular 946, p. 27–29.

Roberts, J. K., 1928, The geology of the Virginia Triassic: Virginia Geo-logical Survey Bulletin 29, 205 p.

Robinson, P., and Luttrell, G. W., 1985, Revision of some stratigraphicnames in central Massachusetts: U.S. Geological Survey Bulletin1605-A, p. A71–A78.

Sanders, J. E., 1968, Stratigraphy and primary sedimentary structures offine-grained, well-bedded strata, inferred lake deposits, Upper Tri-assic, central and southern Connecticut: Geological Society ofAmerica Special Paper 106, p. 265–305.

Schaeffer, B., and McDonald, N. G., 1978, Redfieldiid fishes from the Tri-assic–Liassic Newark Supergroup of Eastern North America: Bul-letin American Museum of Natural History, v. 159, p. 131–173.

Seidemann, D. E., Masterson, W. D., Dowling, M. P., and Turekian, K.K., 1984, K-Ar dates and 40Ar/39Ar age spectra for Mesozoicbasalt flows of the Hartford Basin, Connecticut, and the NewarkBasin, New Jersey: Geological Society of America Bulletin, v. 95,p. 594–598.

Smith, R. C., Rose, A. W., and Lanning, R. M., 1975, Geology and geo-chemistry of Triassic diabase in Pennsylvania: Geological Societyof America Bulletin, v. 86, p. 943–955.

Smoot, J. P., and Olsen P. E., 1988, Massive mudstones in basin analysisand paleoclimatic interpretation of the Newark Supergroup, inManspeizer, W., ed., Triassic–Jurassic rifting: Continental breakupand the origin of the Atlantic Ocean and passive margins: New York,Elsevier, Developments in Geotectonics 22, Part A, p. 249–274.

Sobhan, A. N., 1985, Petrology and depositional history of Triassic redbeds, Bull Run Formation, eastern Culpeper basin, Virginia [M.S.thesis]: Washington, D.C., George Washington University, 224 p.

Stose, G. W., and Bascom, F., 1929, Fairfield-Gettysburg folio, Pennsyl-

vania: U.S. Geological Survey Geological Atlas of the UnitedStates, no. 225, 8 maps, 22 p.

Sutter, J. F., 1988, Innovative approaches to the dating of igneous eventsin the early Mesozoic basins of the eastern United States, in Froe-lich, A. J., and Robinson, G. R., Jr., eds., Studies of the early Meso-zoic basins of the eastern United States: U.S. Geological SurveyBulletin 1776, p. 194–200.

Tollo, R. P., 1988, Petrographic and major-element characteristics ofMesozoic basalts, Culpeper basin, Virginia, in Froelich, A. J., andRobinson, G. R., Jr., eds., Studies of the early Mesozoic basins ofthe eastern United States: U.S. Geological Survey Bulletin 1776,p. 105–113.

Tollo, R. P., and Gottfried, D., 1992, Petrochemistry of Jurassic basaltfrom eight cores, Newark basin, New Jersey: Implications for the

volcanic petrogenesis of the Newark Supergroup, in Puffer, J. H.,and Ragland, P. C., eds., Eastern North American Mesozoic mag-matism: Geological Society of America Special Paper 268,p. 233–260.

Traverse, A., 1986, Palynology of the Deep River basin, North Carolina,in Gore, P. J. W., ed., Depositional framework of a Triassic riftbasin: The Durham and Sanford sub-basins of the Deep Riverbasin, North Carolina: Tulsa, Oklahoma, Society of Economic Pa-leontologists and Mineralogists field guidebook, SoutheasternUnited States, Third Annual Midyear Meeting, 1986, Raleigh,North Carolina, field trip 3, p. 66–71.

Welles, S. P., 1971, Dinosaur footprints from the Kayenta Formation ofnorthern Arizona: Plateau, v. 44, no. 1, p. 27–38.

Willard, M. E., 1951, Geologic Map of the Mount Toby quadrangle,

Massachusetts: U.S. Geological Survey Geologic QuadrangleMap GQ-8, scale 1:31 680.

Willard, M. E., 1952, Geologic Map of the Greenfield quadrangle, Mass-achusetts: U.S. Geological Survey Geologic Quadrangle MapGQ-20, scale 1:31 680.

Witte, W. K., Kent, D. V., and Olsen, P. E., 1991, Magneto-stratigraphyand paleomagnetic poles from Late Triassic–earliest Jurassic strataof the Newark basin: Geological Society of America Bulletin,v. 103, p. 1648–1662.

MANUSCRIPT RECEIVED BY THE SOCIETY AUGUST 7, 1995REVISED MANUSCRIPT RECEIVED MAY 13, 1996MANUSCRIPT ACCEPTED MAY 29, 1996



Printed in U.S.A.

GSA Bulletin: Synthesis and revision of groups within the Newark

Documents

Transcript of GSA Bulletin: Synthesis and revision of groups within the Newark