A record of Late Jurassic vertebrates from Texas

Outcrops of Jurassic strata are rare in Texas. Exposures occur in the Malone Mountains in western Texas and a few isolated outcrops assigned to the Morrison Formation in the northern part of the Texas Panhandle. Jurassic vertebrate fossils were reported from the Malone Mountains (Cragin, 1905; Albritton, 1938), but no descriptions were published, and the location of any previously collected specimens is unknown. Thick sequences of Callovian salt and Oxfordian–Tithonian strata are well known in the subsurface of Texas. They record the opening of the Gulf of Mexico and development of a passive margin (Dickinson et al., 2010; Hudec et al., 2013). Regional Late Triassic–Jurassic uplift, across much of Texas, was associated with the early Gulf of Mexico extensional system. In the Texas Panhandle, exposures of Upper Triassic Dockum Group strata are commonly overlain unconformably by Cenozoic Ogallala Formation and younger strata. Therefore, the Malone Formation in western Texas is a very important potential record of Late Jurassic plants and animals (Fig. 1).

The oldest exposed rocks in the Malone Mountains belong to the Permian Briggs Formation and include dolomite, limestone, and evaporite. Unconformably overlying the Briggs Formation is a heterogeneous sequence of siliciclastic and carbonate strata of the Jurassic Malone Formation (Albritton and Smith, 1965; Berge, 1981). Cragin (1905) described a large number of invertebrate taxa from the Malone Formation, including echinoderms, pelecypods, gastropods, and cephalopods, from which he interpreted a Late Jurassic age. Albritton (1938), Albritton and Smith (1965), and Berge (1981) separated the Malone Formation into an upper and lower member. The lower member includes abundant siliciclastic strata with some limestone, whereas the upper member is dominantly limestone. The Malone Formation is overlain by the Cretaceous Torcer Formation that includes a basal clastic sequence followed by ~120 m of limestone. The conformable boundary with the underlying Malone Formation and a relatively diverse suite of invertebrate fossils led Albritton (1938) to interpret a Valanginian age for the Torcer Formation. In the Quitman Mountains, just south of the Malone Mountains, the Torcer Formation overlies the Malone Formation and is the basal unit of a very thick sequence of Lower Cretaceous strata (Huffington, 1943; Albritton and Smith, 1965). The Malone Formation dramatically changes thickness laterally and is absent in the Finlay Mountains approximately 5 miles north of the Malone Mountains. Albritton (1938) demonstrated thickness changes in the Malone Formation from 45 meters on the east side to 325 meters in the northwest part of the Malone Mountains. Huffington (1943) mapped a very thin interval (15–23 meters) of limestone, shale, and chert-pebble conglomerate that he interpreted as the upper member of the Malone Formation in the northwestern Quitman Mountains just south of I-10 (Fig. 2). Sedimentary facies and rapid lateral changes in thickness are consistent with deposition of the Malone Formation in an extensional setting along the northern margin of the Chihuahua trough.

No fossil vertebrates of Jurassic age have previously been described from Texas. Cragin (1905) and Albritton (1938) noted fossil vertebrate material and plants in their descriptions of the Malone Formation, but this material was never displayed as figures nor described. We have found no evidence that specimens were ever accessioned into a museum. After obtaining permission from local landowners and a paleontological use agreement from the State of Texas General Land Office, geologists and paleontologists from The University of Texas at Austin and Southern Methodist University visited the Malone Mountains in 2015 and 2016 to explore for fossil vertebrates and plants based on the previous reports. This report describes the first Jurassic vertebrate fossils from Texas.

Preparation. Deborah Wagner and Matthew Brown assisted with preparation on TMM 44038-1 using a zipscribe.

Photography. Specimens were photographed using an iPhone 11 or a Canon Power Shot D20 and processed using Adobe Photoshop. Thin sections were imaged using a Zeiss Axio Imager.A1 microscope with ZEN software.

Computed Tomography (CT). All CT imaging was performed at the High-Resolution X-ray CT Facility at The University of Texas at Austin using a North Star Imaging micro-computed tomography scanner with a Fein Focus High Power source and a Perkin Elmer detector. TMM 47259-1 was scanned at 350 kV, 0.5 mA, with voxel size of 189.2 mu. TMM 47257-1 was scanned at 140 kV, 0.15 mA, with voxel size of 48.7 mu.

Detrital Zircon U/Pb Geochronology. Detrital zircons were analyzed using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) at the Arizona LaserChron Center (www.laserchron.org) utilizing methods described by Gehrels et al. (2006, 2008).

SMU, Southern Methodist University. TMM, Texas Vertebrate Paleontology Collections, The University of Texas at Austin. Specimens are curated in the Texas Vertebrate Paleontology Collections. Locality numbers are formatted TMM ##### and specimen numbers include the locality number, followed by a dash and the catalog number, i.e., TMM ##### - #. A locality is a specific point on a map. An area, such as the Jackson Ranch area, may have multiple localities.

Based on a preliminary collection of invertebrate fossils, Cragin (1897) recognized the significance of the Malone Formation as a rare occurrence of Jurassic strata exposed in Texas. Cragin (1905) provided a comprehensive description of the invertebrate fauna from the Malone Formation and noted the rare occurrence of vertebrate fossils as well. He concluded that the invertebrate fauna indicated a Jurassic age for the Malone Formation with strongest affinities to the Kimmeridgian and Tithonian. In his faunal list for the Malone Formation, Cragin (1905) included pycnodont fish tooth, cycloid fish scales, and Enaliosaur “(fragments of bones, indicating animal of considerable size)”, at least some of which were noted from a small outcrop on the east side of the Malone Mountains (Fig. 2). The vertebrate fossils were never described and no indication of accession into a museum collection has been found.

Albritton (1937) described fossil foraminifera from the lower member of the Malone Formation that he interpreted to indicate a Late Jurassic age, although both taxa were new species. Albritton (1938) documented an extensive list of invertebrates, many of which were previously described by Cragin (1905) from the Malone Formation, including 56 species of pelecypods, 18 gastropods, and 15 cephalopods. He interpreted the ammonites to be the best evidence for correlation of the Malone Formation with Jurassic strata in Mexico and, like Cragin, concluded that the Malone Formation was Kimmeridgian–Tithonian in age. Based on numerous measured sections and extensive field mapping, he described the Malone Formation as a sequence of conglomerate, sandstone, sandy shale, siltstone, gypsum, and limestone. He recognized a lower member with a higher abundance of thin-bedded, clastic strata, and an upper member characterized by the predominance of limestone.

Albritton (1938) recorded “large bone fragments” in the description of his measured Section #7 in the northern Malone Mountains. His field samples are housed at the Shuler Museum of Paleontology, SMU, and were inspected in 2019, but no evidence of vertebrate fossils was found. Albritton’s correspondence, from the time he worked in the Malone Mountains, is housed at the DeGolyer Library at SMU. Letters by Albritton, E.H. Sellards, and Marland Billings provide context for Albritton’s decision to study the Malone Mountains. His correspondence to his family provides some insight into the logistical challenges of field work in west Texas, but no field notes were found in the collection at the DeGolyer Library.

Imlay (1940) described fossil pelecypods from Mexico and compared them to the fauna from the Malone Formation. He agreed with Cragin (1905) and Albritton (1937, 1938) that the fauna was most likely of Late Jurassic age and correlated the formation with the La Casita Formation of the Saltillo-Torreon area of Mexico. Young (1969) identified the ammonite Kossmatia kingi from the upper Malone Formation and correlated it with the middle Tithonian (Kossmatia rancheriaensis zone).

In their USGS Professional Paper on the geology of the Sierra Blanca region, Albritton and Smith (1965) reviewed the stratigraphic and paleontological evidence for the age of the Malone Formation and concluded that it was an Upper Jurassic sequence consisting of two members. The contact between these two members was described by Albritton and Smith (1965) as conformable and transitional. They recognized that the two members were, in part, age-equivalent facies. The ammonites Idoceras and Kossmatia from the lower and upper members respectively, were interpreted to indicate Kimmeridgian and Tithonian ages for these strata. Fossil plants were simply described as “driftwood”. Cragin (1905) identified the boring clam Martesia maloniana in some of the plant fossils, consistent with the interpretation that plant material had been transported into a marine environment prior to burial.

The most recent detailed mapping of the Malone Mountains was by Berge (1981). He interpreted the lower member of the Malone Formation to represent a marginal marine, fan-delta system and recognized “sheetflood” (clastsupported conglomerate), “debris-flow” (matrix-supported conglomerate), “sandstone” and “distal” facies. Berge (1981) interpreted limestone within the Malone Formation to represent subtidal through supratidal facies. He did not report fossil vertebrates but commented on the locally abundant fossil wood.

The Malone Formation records sedimentation along the northern edge of the Chihuahua trough, an elongate northwest–southeast basin primarily located in northeast Mexico. Various authors, including Goldhammer and Johnson (2001), have interpreted a long-lived history of the Chihuahua trough. Beginning in the Late Triassic, this area was back-arc to the Cordilleran convergent margin, but also linked to early extension in the Gulf of Mexico system. A significant phase of Late Jurassic extension is recorded by the locally derived boulder conglomerates and rapid thickness variations in the Malone Formation. Haenggi and Muehlberger (2005) interpreted a regional event of subsidence in the Chihuahua trough associated with oblique extension in the Late Jurassic. Haenggi (2002, his fig. 1) interpreted the margin of the Chihuahua trough to be immediately north of the Malone Mountains, consistent with the absence of the Malone Formation in the Finlay Mountains.

The northern margin of the Chihuahua trough experienced northeast-directed shortening during the Late Cretaceous–Paleogene Laramide interval. This tectonism is recorded by the northeast-vergent fold-and-thrust belt that involves strata as young as Early Cretaceous in the Malone Mountains (Berge, 1982; Haenggi, 2002) and as young as early Late Cretaceous in the Quitman Mountains (Albritton and Smith, 1965). Although most mapped thrust faults are detached within Permian–Cretaceous strata, some authors have suggested that Late Jurassic basinbounding normal faults may have been at least partially inverted during the Laramide orogeny (Haenggi, 2002). Basin-and-Range style extensional faulting characterized the late Tertiary tectonic history.

The dominant structural style exposed in the Malone Mountains is a northeast-vergent fold- and-thrust belt with northwest-striking thrust faults and northwest-trending hinge lines. The fold belt is locally dissected by a series of younger, relatively small-throw normal faults. The best exposed thrust faults are detached within the Permian strata, although Berge (1981) also interpreted a series of deeper thrust faults to explain the easternmost folds. The style of contractional deformation, with locally overturned bedding exposed in tight, northeast-vergent folds, is consistent with fault-propagation folding. In the Malone Mountains, this deformation involves both the Upper Jurassic Malone Formation and Lower Cretaceous Torcer Formation. A number of additional Lower Cretaceous formations have been mapped in the Quitman Mountains representing shallow marine to marginal marine carbonate and siliciclastic strata (Huffington, 1943). The general trend in this sequence is that it becomes thinner and more siliciclastic from south-to-north (Albritton and Smith, 1965). The Quitman thrust places overturned Cretaceous strata in the hanging wall over Cretaceous strata in the footwall (Albritton and Smith, 1965). Strata as young as Cenomanian–Turonian are involved in the large-scale contractional deformation. Upper Eocene intrusive and volcanic rocks cut the contractional structures in the Quitman Mountains (Henry and Price, 1984). Both detached and basement-involved structural styles of Laramide age have been described at various locations along the Chihuahua trough (Carciumaru and Ortega, 2008).

The Jackson Ranch area, in the northwest Malone Mountains, includes a relatively simple, southwest-dipping (20–30 degrees) panel of Permian through Jurassic strata exposed above a northeast-vergent thrust fault that places Permian rocks over the lower Malone Formation (Berge, 1981) (Figs. 3, 4).

The local structural geology is more complex on the east side of the Malone Mountains. Although the general structure is an upright syncline, with a northwest–southeast-trending hinge line, the southern portion of the outcrop exposes a series of closely spaced, secondary folds with north–south-oriented hinge lines (Fig. 4). Beds of the lower Malone Formation, near TMM 47258, dip easterly at 30–50 degrees, but rapidly become shallower to the southeast where they return to a gentle westerly dip.

The nature of the outcrop and complexities of the structural geology make it very difficult to propose a detailed stratigraphic correlation between TMM 44038 and TMM 47258. Based on the stratigraphic distance below the base of the upper Malone unit, it is likely that both localities are from within the upper half of the lower Malone Formation.

In 2015, geologists and paleontologists from the University of Texas and Southern Methodist University made a reconnaissance trip to the Malone Mountains to investigate previous claims of bone fragments and fossil plants in the Malone Formation (Cragin, 1905; Albritton, 1938). The fieldwork focused on two areas, one in the northwestern Malone Mountains and one on the eastern side of the range. The Jackson Ranch area in the northwestern Malone Mountains is located on private land adjacent to Texas State land. We graciously received permission from the current landowners to access their property and explore for fossils. The Jackson Ranch area includes fossil localities TMM 44038 and TMM 47257 and is located near measured Section #7 of Albritton (1938) where he noted “large bone fragments” and plant material as well as fish scales and abundant invertebrate fossils (Fig. 5).

The localities in the eastern Malone Mountains (TMM 47258, TMM 47259) are near where Cragin (1905) noted invertebrate fossils, fossil plant material, a fish tooth, and “bone fragments” (Cragin #1 area). Near the southern end of these exposures, we observed abundant invertebrate fossils (pelecypods, gastropods, cephalopods), plant material, and rare vertebrate bone fragments. The vertebrate material was discovered in gray, sandy limestone and limestone concretions (Fig. 6). Collection of fossils from this area was sanctioned under a paleontological use agreement with the State of Texas General Land Office (Lease # SD20150013).

Both areas were revisited during 2016 and additional material was recovered including the vertebra described below from TMM 44038 (Fig. 7). Additional outcrops were also explored on the west end of Finlay Spur (Fig. 4). Although the lower Malone Formation is reasonably well exposed, vegetation is denser than at the nearby Jackson Ranch area. Fossil invertebrates and plants were relatively abundant, but no vertebrate fossils were recovered.

Two samples of calcareous sandstone from the lower Malone Formation were collected from the Jackson Ranch area. U-Pb ages were obtained from detrital zircon grains at the University of Arizona LaserChron Center. MM002 was collected from calcareous sandstone approximately 25 feet above the base of the Malone Formation. Of the 100 detrital zircon grains analyzed, all U-Pb ages are Permian and older (Fig. 8) with 19% being Paleozoic, 79% Proterozoic and only 2% Archean. The most abundant age population yields Grenville ages (1000–1100 Ma) with a statistically significant peak at 1040 Ma. MM001 was collected from calcareous sandstone approximately 55 feet stratigraphically above MM002 in the lower Malone Formation (Fig. 5). Of the 99 detrital zircon grains analyzed, 8 grains are Mesozoic in age and the youngest statistically significant age peak was observed at 247 Ma (Triassic). Three grains produced Jurassic ages (151 ± 2, 167 ± 3, 174 ± 2 Ma) (Fig. 8). Although only a single grain, the 151 ± 2 Ma age is consistent with the paleontologically constrained depositional age of the lower Malone Formation described above. Paleozoic ages were recovered from 13% of the detrital zircon grains, while 75% were Proterozoic, and only 2% were Archean. As with MM002, the most abundant zircon population is Grenville in age with a statistically significant age peak of 1033 Ma (Fig. 8).

The detrital zircon age distributions suggest that the source area for the Malone Formation sandstones was dominated by Grenville age rocks, or rocks that contained abundant Grenville age zircon grains. The pattern of early Paleozoic and Precambrian zircon ages, in the lower Malone Formation samples, is similar to the age distribution reported from the Cambrian Van Horn Formation exposed east of the Malone Mountains (Spencer et al., 2014). The pattern is also similar to age populations reported from the Upper Jurassic Morrison Formation in New Mexico (e.g., CP49) (Dickinson and Gehrels, 2008). The Morrison Formation sediments were likely shed north from the Mogollon-Burro rift shoulder that separated the Morrison basin from the Chihuahua trough. The appearance of Mesozoic zircon grains in the stratigraphically higher sample (MM001) indicates a linkage to the Mesozoic magmatic arc. Middle and Late Jurassic age igneous rocks are widespread in southern Arizona and northern Sonora near the northern end of the Chihuahua trough (Mauel et al., 2011).

As part of our exploration for Jurassic vertebrates in Texas, we also visited the limited outcrops north of Dalhart in the Texas Panhandle that had been mapped as questionable Morrison Formation by Eifler et al. (1984). Richmond et al. (2020) show Morrison Formation thicknesses between 0 and 100 meters in the subsurface northwest of Dalhart, however those strata are represented by very limited outcrops. Detrital zircon geochronology from a sandstone in this area produced abundant Late Jurassic grains but no Cretaceous grains. Although not unique, this pattern is consistent with a Morrison Formation correlation as concluded by Myers, et al. (2018). Unfortunately, no vertebrate fossils were found at this locality. However, fossil vertebrates have long been known from outcrops of the Morrison Formation in the nearby Oklahoma Panhandle (Stovall, 1938; Richmond, 2020).

Plants

Figure 9 

Plant fossils in the lower Malone Formation are relatively abundant, although not well preserved. Most of the plant material represents woody stems ranging from centimeter-scale fragments up to meter-scale permineralized gymnosperm logs preserved in a variety of modes (Fig. 9). Teredolites is common in the larger wood fragments consistent with the interpretation that wood was washed into shallow marine environments, adjacent to the lower Malone Formation fan-delta systems and drifted prior to burial. The fossil plants are currently being studied in detail and will be described in a separate paper.

Sauropterygia Owen, 1860 

Plesiosauria Blainville de, 1835 

Pliosauroidea Seeley, 1874; Welles, 1943 

Pliosauridae Seeley, 1874 

cf. Pliosauridae

Figure 10 

Localities: TMM 44038.

Referred Specimens: TMM 44038-1 vertebra.

Description: TMM 44038-1 includes a partial centrum, an associated neural arch and a smaller secondary fragment of the neural arch (Fig. 10). Very little of the exterior surface of the centrum is preserved so the original geometry of the vertebra is uncertain. Both the posterior and anterior articular surfaces appear to be platycoelous. The centrum is oval in shape, when viewed from the anterior, with a maximum preserved dorsoventral height of 92.9 mm and mediolateral width of 111.9 mm. Based on small patches of exterior bone surface, the original dimensions were probably not greater than 95 mm x 120 mm. The maximum preserved antero-posterior length of the centrum is 53 mm. There is no unequivocal fit of the neural arch to the centrum, but the two specimens were found within centimeters of each other, are consistent in size, and it is reasonable to conclude that they are from the same vertebra. The maximum antero-posterior length of the primary neural arch fragment is 58 mm. The neural canal is oval, slightly elongate laterally (30 mm x 27 mm) and narrows from 28.9 mm to 23.7 mm, consistent with anteroposterior narrowing observed in some pliosaurids. A median cleft is present on the dorsal surface of the neural arch along the antero-posterior midline of the neural spine (Fig. 10C). A partial chevron facet is preserved on the ventrolateral portion of the centrum (Fig. 10D). The facet is crescent-shaped with a maximum preserved dorsoventral length of 25 mm. There is no evidence of foramina on either the dorsal or ventral surface of the vertebra. The lack of subcentral foramina and the ventrolateral chevron facet are consistent with the identification of this specimen as a caudal vertebra. However, some Early Cretaceous pliosaurids (Stenorhynchosaurus munozi, Kronosaurus-Eiectus and Brachauchenius lucasi) lack subcentral foramina on their cervical vertebrae (Fischer et al., 2023).

cf. Plesiosauria

Figures 11, 12 

Localities: TMM 47259

Referred Specimens: TMM 47259-1, propodial

Description: TMM 47259-1 (Figs. 11, 12) is contained within a limestone concretion and includes both dorsal and ventral halves of a limb bone possibly representing the propodial of a small plesiosaurian. Plesiosaurian femora and humeri have a limited suite of distinctive characters and the fragmentary nature of TMM 47259-1 makes identification even more challenging. Therefore, we identify this specimen as a propodial from a small, or possibly juvenile, plesiosaurian. The distal portion of the propodial was at least 30 mm thick. The preserved thickness of the bone, at the proximal end of the broken shaft, is 25 mm (minimum). Although only the distal end of the propodial is preserved, the specimen is 118 mm in length with a maximum width of approximately 77 mm (Fig. 12). The specimen displays a broad, dorsally concave, distal end that narrows to a more equant shaft at the proximal end. The distal termination of the bone has two relatively straight facets. The ventral surface is relatively flat. Buchy (2007) recognized at least three different taxa of pliosaurid plesiosaurians from Kimmeridgian strata in northeast Mexico and noted rare elasmosaurids as well, although identification of the latter group is very uncertain. Elasmosaurids are generally thought to have appeared in the Early Cretaceous (Benson and Druckenmiller, 2014).

The concretion containing TMM 47259-1 was reconstructed in an attempt to image the bone geometry using a high-resolution CT scanner. Unfortunately, the density contrast between bone and limestone matrix was too small to successfully differentiate the bone.

An interesting detail of TMM 47259 is the presence of both the trace and body fossil of a boring bivalve in the bone (Fig. 13). A matrix-filled, semi-circular burrow, approximately 8 mm in diameter at the base and 20 mm in length, extends roughly perpendicular from the margin of the bone. This burrow contains at least part of the boring bivalve at its base. Trace fossils of boring bivalves have been commonly referred to the ichnogenus Teredolites, if the substrate is wood, and to Gastrochaenolites, if the substrate is rock or shell (Kelly, 1988, Donovan and Ewin, 2018). As mentioned above, Teredolites are common in fossil wood from the lower Malone Formation. The presence of such a bivalve in bone might be referred to Gastochaenolites, but Donovan and Ewin (2018) have argued that substrate is a poor basis for identification. In either case, it is clear that the propodial was exposed underwater in a relatively stable setting for the boring bivalve to have taken up residence. As in many boring bivalves, the burrow appears to widen with depth with a bulbous portion at the base and a narrower segment towards the outer surface of the bone. This has been attributed to the growth of such bivalves as they excavated the burrow. Some boring bivalves are xylophagous, that is they eat the wood into which they bore and digest the cellulose. Others are filter feeders and use the burrow as a mechanism to attach to a substrate. The boring bivalve observed in TMM 47259-1 is interpreted as a filter feeder, as it is found in bone and has a relatively short burrow.

Cragin (1905) named the boring bivalve “Martesia maloniana”, based on specimens of borings in fossil wood from the Malone Formation. Subsequent revisions to the systematic paleontology of the Pholadidea (boring bivalves), recognized this taxon as Opertochasma mallonianum (Haga and Kase, 2011). Haga and Kase (2011) interpreted Opertochasma mallonianum as a filter feeder based on burrow morphology. They further suggested that the occurrence of O. mallonianum in the Malone Mountains is anomalous as this would place it outside of what they interpreted as the Late Jurassic “warm” climatic zone. The paleogeographic map they used for their analysis is more appropriate for the Bathonian–Oxfordian and depicts the Malone Mountains location incorrectly as having been part of the Sundance seaway. The Malone Formation is correlative with the Morrison Formation, rather than the Sundance Formation, and was deposited on the northern edge of the Chihuahua trough associated with both the Gulf of Mexico and Pacific Ocean systems.

Additional material from TMM 47258 may represent fragments of marine reptile bones. Some of the specimens are thin (5–15 mm) bone fragments. This might be consistent with pectoral and limb bones of plesiosaurians but could also represent fragments of bones from a variety of other Late Jurassic vertebrates. It may represent cortical bone that has separated from larger bone fragments. TMM 47258-1 includes a subtriangular element (Fig. 14A) with cancellous interior fabric, as well as a number of thin fragments with parallel-fibered grain (Fig. 14B). The fibrous grain runs parallel to the long axis of the bone and to the exterior bone surface. Parallel-fibered bone is commonly associated with reptiles (Huttenlocker et al, 2013). A thin section of TMM 47258-1, cut approximately orthogonal to this fibrous grain, reveals the bone structure with aligned fibers arranged in a loosely lamellar geometry. Primary osteons are relatively tightly packed and are approximately 300–400 μm in diameter (Fig. 15). Although not unique, the bone histology is consistent with other examples from plesiosaurians (Fleischle et al., 2018).

The vertebra from the Upper Jurassic lower Malone Formation represents a large marine reptile. The size and shape of the partial centrum is generally consistent with caudal vertebrae of plesiosaurians. The shape of the centrum, the position of the chevron facet, and the lack of ventral foramina are all consistent with this identification. Wintrich et al. (2017) used CT imaging and histological sections to study the internal structure of cervical vertebra from plesiosaurians. They found the large foramina to be commonly paired, symmetrical and to extend from the ventral side of the centrum through to the floor of the neural canal. After careful examination of TMM 44038-1 under microscopy, there is no evidence for such throughgoing foramina. Caudal vertebra in pliosaurid plesiosaurians commonly lack these foramina.

The size and morphology of TMM 44038-1 is similar to plesiosaurians known from age equivalent strata in northeastern Mexico (Frey and Stinnesbeck, 2014). Buchy (2007) reviewed the record of marine reptiles from the Upper Jurassic La Casita and La Caja Formations of northeast Mexico, including crocodilians, ichthyosaurs, and plesiosaurians. Buchy et al. (2006) and Buchy (2007) recognized at least three different taxa of pliosaurids from the Kimmeridgian of northeast Mexico, as well as rare elasmosaurid specimens. The dimensions listed in Buchy (2007) for elasmosaurid vertebrae are significantly smaller than TMM 44038-1. A cervical vertebra described in Buchy (2007), assigned to Pliosauridae indet., is similar in size to TMM 44038-1. Pliosauarids are known to range through the Jurassic and Cretaceous. Contrary to the identification of Late Jurassic elasmosaurids in northeastern Mexico, elasmosaurids are generally thought to appear in the Early Cretaceous (Benson and Druckenmiller, 2014). Based on morphology and considering the fossil record from correlative strata in northern Mexico, the vertebra from the Malone Formation probably represents a pliosaurid plesiosaurian, although identification is uncertain.

Identification of the partial propodial is even less certain. TMM 47259-1 is fragmentary and likely represents a partial limb element of a marine reptile. It may represent a small plesiosaurian, but no morphologic characters are definitive. Additional specimens are too fragmentary to be identified at this time although bone histology is consistent with that of marine reptiles.

The Malone Formation is correlative with at least part of the La Casita and La Caja Formations in northern Mexico and these formations were all deposited along the northern margin of the Chihuahua trough. Zell et al. (2014) reviewed the stratigraphy of the La Casita and noted the overall transgressive nature of this sequence of shallow marine clastic and carbonate strata. Marine vertebrates from the La Casita include fish, ichthyosaurs, crocodilians, and plesiosaurians (Buchy et al., 2006). Paleogeographic isolation has been inferred as a reason for perceived endemism in the Late Jurassic marine vertebrates of northern Mexico (Buchy, 2007; Stinnesbeck and Frey, 2014). However, the taxonomy of many of these vertebrates is still uncertain and the invertebrates indicate exchange with both the boreal Pacific and the Gulf of Mexico (Zell and Stinnesbeck, 2015). Late Jurassic paleogeography along the Chihuahua trough was likely complex with multiple emergent areas in an active back-arc setting. However, marine connectivity was maintained to the Paleo-Pacific and to the Atlantic Gulf of Mexico (Mauel et al., 2011) (Fig. 16). Rapid lateral thickness changes in the Malone Formation and the presence of coarse conglomerates are consistent with active rifting in the Late Jurassic along the northern margin of the Chihuahua trough.

Fossil wood indicates forested areas occurred adjacent to drainages that fed the lower Malone Formation fan-delta systems. Paleogeographic reconstructions suggest that such drainages would have been short with headwaters in nearby higher elevations, along the flanks of the Mogollon-Burro rift shoulder (Mauel et al., 2011, Lawton et al., 2020). Zell et al. (2014) noted an abundance of fossil wood interpreted as driftwood in the La Casita Formation in southern Coahuila. Rivers, draining the northeast flank of the Mogollon-Burro topographic high, fed the Morrison Formation basin. The Morrison Formation includes a very diverse assemblage of non-marine flora and fauna (Chure et al., 2006).

The Late Jurassic Chihuahua trough was perhaps similar to the Gulf of California today in terms of both geologic setting and biologic diversity with some limited endemism, while having biogeographic connectivity to the open ocean. The modern Gulf of California exhibits remarkable diversity and abundance of marine vertebrates, in part due to the high rate of primary productivity, and the variety of habitats controlled by bathymetric and oceanographic complexity (Urbán, 2010). This complexity supports tropical and temperate species from coastal and oceanic environments.

The first description of fossil vertebrates from Jurassic strata in Texas includes a possible pliosaurid plesiosaurian from the lower Malone Formation in the Malone Mountains. This taxon is represented by a single vertebra, but additional bone fragments have been recovered from marginal-marine to shallow-marine facies. Identification of taxa from these bone fragments is uncertain, but may include additional plesiosaurian material. Abundant invertebrate fossils constrain the age of the Malone Formation to Kimmeridgian–Tithonian (Albritton and Smith, 1965). Extensive exposures of the lower Malone Formation in the Malone Mountains await further prospecting. The Malone Formation is correlative with the La Casita and La Caja Formations in northern Mexico that have also produced marine reptiles including crocodilians, ichthyosaurs, and plesiosaurians. These strata, and their associated fossils, were deposited during the Late Jurassic along the northern margin of the Chihuahua trough. This feature was a tectonically and bathymetrically complex series of linked extensional basins behind an active convergent margin to the west and possibly also associated with the opening of the Gulf of Mexico. Detrital zircon geochronology from the lower Malone Formation is consistent with both the age and paleogeographic setting of these Upper Jurassic strata.

We would like to thank the Jackson Family for permission to access their property in the Malone Mountains. D. Walker graciously shared his knowledge of the area and provided much appreciated logistical support for our field work. K. Creitz, at the Texas General Land Office, was very helpful in establishing the appropriate Paleontological Use Agreements for our work on state lands. M. Colbert and J. Maisano scanned and helped visualize data at The University of Texas at Austin High Resolution X-ray Computed Tomography (CT) Facility. C. Sagebiel assisted with curation of the specimens at the Texas Vertebrate Paleontology Collections, The University of Texas at Austin. The original manuscript was much improved through critical reviews by T. Adams, P. Druckenmiller, R. Hunt-Foster, M. Loewen, and M. May. We would also like to thank R. Frost, B. Orr, and A. Snoke for assistance with the editorial process.