LateCretaceous (Lancian) dinosaurs from the McRae Formation,Sierra County, New Mexico byRichard P. Lozinsky, Adrian P. Hunt, and Donald L. Wolberg,New Mexico Bureau of Minesand Mineral Resources, Socorro,NM, and Spencer G Lucas,University of New Mexico, Albuquerque, NM Introduction prospectinghas yielded additional dinosaur- in the Cutter sag and Jornada del Muerto Identifiable skeletal elements of Late Cre- bearing localities(Fig. 1). In this paper we (Fig.2). Bushnell(1953, 1955) recognized two taceous dinosaurs are known from three for- describethe dinosaurand fossil-plantoccur- McRaemembers, the lower JoseCreek and mations in New Mexico: the Fruitland rences,we review previous studies of the the upper Hall Lake, and he suggestedthat Formation and Kirtland Shale of the SanJuan McRaeFormation, and we discussthe ageof the McRaeFormation may approach1,000 m Basin, and the McRae Formation in and the McRae and its oossible correlation to in thickness(Fig. 3). Becausepart of the sec- around the Elephant Butte Reservoir area. stratigraphicunits in the WesternInterior. tion is submergedbeneath the waters of El- This is the second published description of eohant Butte Reservoirand becauseof dinosaur remains from the McRae Forma- McRae Formation pronounced faulting, measurement of a tion. The first report was published by Lee Although recognizedas a discretestrati- completesection is not possible. (1905). Recent discoveries of McRae dino- graphic unit since the early twentieth cen- The fose Creek Member restsunconform- saurs began in 1981when Lozinsky observed tury (e.g., Lee, 1905,1.907), the McRae ablvuoon rocksof the MesaverdeGroup and the presence of dinosaur-bone fragments Formation was first named by Kelley and consisisof a sequenceof sandstone,shale, duriig geologic studies in the Elepharit Butte Silver (1952).This unit consistsof a thick se- conglomerate,and an unusual brecciacon- area (Lozinsky, f982). Subsequent fossil quenceof nonmarine clasticrocks exposed glomerate (Bushnell, 1953;Lozinsky, 1982, 1984).This memberis 120m thick at its type locality. The Hall Lake Member conformablyover- lies the JoseCreek and is distinguished by its abundanceof purple and maroon shales. The boundary between the two membersis marked by a basalconglomerate. When the conglomerateis absent, the boundary is marked by the first purple or maroon shale. The top of the Hall Lake Member is absent in all areasand is unconformablyoverlain by variousCenozoic units. Paleobotany Megascopicplant remainsappear to be re- stricted to the JoseCreek Member and are most common in the finer grained sand- stonesand siltstones.Fossil leaves also have been reported from the underlying upper CrevasseCanyon Formation (Table 1; Loz- insky, 1982;Wallin, 1983). Despite difficulties that arisefrom comparing CrevasseCanyon and fose Creek fossil floras to floras else- where, the assignmentof a Late Cretaceous age to the CrevasseCanyon and McRae is consistentwith available data as indicated below. In addition to fossil leaves,abundant pet- rified wood fragments and occasionalfrag- mentary logs and in situ stumpsoccur in the JoseCreek Member. Some of the stumps are of substantialsize ( t 1 m in diameter)and appearto be in growth position. Vertebratepaleontology Lee (1905)reported dinosaur bonesin rocks at or near Elephant Butte. Two additional collectionswere made in the areaduring the next eight years (Kelley and Silver, 1952), and these fossils were referred to the cera- topsian genusTricuatops by I . W. Gidley (Lee, 1907;LulI,1933). The fossilmaterial collected during the 1905-1913period was included in the collectionsof the U.S. National Mu- seum. Today, this material consists of two FIGURE 1-Generalized geologic map of the Elephant Butte Reservoir area; dinosaur-bearing localities dinosaur vertebral centra and other bone are highlighted in blue. November 1984 Nm Mexico Geology - o _c. C) ()3 o .9 a FIGURE4-Anterior viewsof USNM2437 (x tlz\, t- dorsalcentra referred to Triceratops,possibly by S\8 J. W. Gidley (Lee,1907; Lull, 1933;bar is 10 cm long). USNM 2437 collection contains no cranial + material and cannot provide a realistic basis 6ptslr'7:\ \ for generic identification. However, the two centra of USNM 2437 are large and within i*l \te\ s\ the size range of Triceratops(Hatcher and others, 1907,pl.8). Torosaurus,a Late Cre- taceousceratopsian as large as Triceratops,is known only from cranial remains (Colbert and Bump, 1947).It is not possibleto deter- mine whether USNM 2437rcpresentskicer- / o 25mi ntops,Tbrosaurus, or a new ceratoPsiantaxon. ', - Between1913 and 1981no dinosaurfossils o 25km were collected from the McRae Formation, although no less than seven geologic map- FIGURE 2-Tectonic setting of the Elephant Butte area. Hatched lines are normal faults with hatchures ping projects included the McRae (Doyle, on the downthrown side; the lines are dashed where faults are buried or inferred. Solid lines delineate 1951;Kelley and Silver,1952; Bushnell, 1953; major uplifts (after Woodward (1982, and others, 1978).Shaded portion is the area coveredby Lozinsky 1951; Melvin, 1984). McCleary,'1960;Thompson, 1.963;Warren, 1978).However, in 1968,John Hawley and William Seager(now at the New Mexico Bureau of Mines and Mineral Re- sources and New Mexico State University, TABLE 1-CnEvessl CaNvoNeNo Iosr Cntlr ruones respectively) were notified of a partial cer- atopsianskull discoveredat a time of a very TOP ERODEDAND NOT Taxa Rock unit Butte Res- SEEN,OVERLAIN BY low water level in the Elephant this locality has not QUATERNARY-TERTIARY JoseCreek Member Geinitziacf. formosa ervoir. Unfortunately, (r of McRae ?Cannamagnifolia been exposedsince then. BASALTFLOWS AND Formation Sabalmontana The studies initiated bv Lozinskv in 1981 F ALLUVIUM (Bushnell, 1953; Plryllitescf. ratonensis E, haveresulted in the discoveryof l2"localities lrl Kottlowski, et al., Salixsp. (Fig. 1) that have yielded fossil matedal rep- 1956;Lozinsky, sp. F Cinnamomum taxa(another locality 1982) Sequoiasp. resentingfour dinosaur lies just outside the map area to the north). ro HALL LAKE Seauoiareichenbach: tr) Sabatitessp. Additionally, in 1983 a yachtsman discov- o, -zJSBEf-r-'(-884m) ' Sabalitesmontana ered a locality on the east bank of the res- Ficusplanicostata ervoir that yielded an incomplete jaw of (D C Ficustrinerius Tyrannosaurusrex. This very significant spec- E o Viburnummarginatum imen is in the collections of the New Mexico Araucar it es longifo lia Museum of Natural History and has been made available to us through the courtesy of a CrevasseCanyon Sequoiamontana recognized its sig- l Formation Ficusolaniostata David Gillette, who first o nificance.A detaileddescription of this spec- lrJ (Lozinsky, 1982; cf . Dryophyllum sp. o ASH CANYON Wallin, 1983) cf . Inurophyllum wardiana imen will be undertaken by the New Mexico MEMBER(24O m) cf . Dillenitescleburn: Museum of Natural History and the New F rO UJ @ Cercidiphyllumsp. Mexico Bureau of Mines and Mineral Re- E. () BARRENMEMBER cf. Quercusaiburnit'olia sourcesin a separatepaper. i fossilsprobably camefrom low ( 328 m) THE USNM in the Hall Lake Member. Localities discov- 3 COAL-BEARING fragments,all cataloguedas USNM 2437.lt ered recentlyoccur at or near the boundary MEMBER(137 m) is likely that thesespecimens are all that re- between the Hall Lake and JoseCreek Mem- main of the 1905-1913collections. Figure 4 bers.Fossil bone is often associatedwith con- shows the two centra that have been referred glomeraticor shaly beds. Discussionof the FIGURE 3-Generalized stratigraphic column for to Triceratops. most significant specimens found by us fol- the Elephant Butte area based on Lozinsky (1982, Current criteriafor defining ceratopsiantaxa lows. 1984)and Wallin (1983). are basedmainly on skull characteristics.The New Mexico Geolo3y November 1984 FIGURE5-TKM001, dorsaltheropod centrum:(A) anterior,(B) lateral,and (C) dorsalviews; TKM007,Alamosaurus sp. humerus: (D) anterior and (E) posterior views; TKM002, ceratopsiancoronoid process:(F-G) lateral views; TKM020, ceratbpsiancrest fragment:(H) dorsaland (I) ventral views; TKM01l, ankylosaurarmor fragment:(f-K) lateraland (L) basalviews. The solid bars are all 5 cm in Iength exceptD-E, which is 10 cm. November7984 NewMexico Geology Centrum TKM001(Fig. 5, A-C) is an incomplete,but relativelywell preserved,dorsal centrum. It is strongly opisthocoelousand ovoid in out- preserved anteroventrally, but it is likelv to have been similarly devetopedand to have {ormed a deep concavity between the two lipJike structures. The strongly opisthocoelouscharacter, rel- atively large size, presence of a pleurocoel and secondaryforamina on the pedicelsof the neura! arch, and the excavaiednature betweenanterior and posterioroverhanging FIGURE6-Sauropod femur (TKM009), still partially encasedin sandstone of the fose Creek Member lips lead us to believe that TKM001 re-prel matrix,is more than 168cm long. sentsa large theropod, possibly a tyranno- still saurid. partially encasedin matrix, extendsdi- Ceratopsian crest fragment TKM020 (Fig. 5, H-I) is a bone fragment Humerus 125mm x 115 mm that varies in thickness from 34 mm at its broken edge to 8 mm at an almost intact margin. The cortex is can- cellous and bounded above and below by dense, laminar bone. The surfacesof th-e representsa sauropod.The most reasonable fragment are slightly concave and slightly hypothesisis that this specimen be referred convex, respectively. Both surfaces exhibit to Alamosaurus.However, there is a strong narrow linear and arcuate