Qinglongopterus guoi, Lü & Unwin & Zhao & Gao & Shen, 2012

Lü, Junchang, Unwin, David M., Zhao, Bo, Gao, Chunling & Shen, Caizhi, 2012, A new rhamphorhynchid (Pterosauria: Rhamphorhynchidae) from the Middle / Upper Jurassic of Qinglong, Hebei Province, China, Zootaxa 3158, pp. 1-19 : 2-9

Qinglongopterus guoi gen. et sp. nov.

Etymology. The generic name is derived from ‘Qinglong’, the name of the Chinese administrative unit in which the holotype locality is to be found, and ‘pterus’ a declension of pteron, the Greek word for wing. The specific name is in honor of Chen Guo, who offered the specimen for scientific research.

Holotype. D3080, D3081 ( Figs 1, 2), an almost complete skull and postcranial skeleton preserved on a slab and counter-slab and housed in the Dalian Natural History Museum , Dalian, China.

Locality and horizon. Mutoudeng, Qinglong County, Hebei Province, China. Beds exposed at Mutoudeng have been assigned to the Tiaojishan Formation ( Bureau of Geology and Mineral Resources of Hebei Province, 1989) currently dated as latest Middle to earliest Upper Jurassic ( Zhang et al. 2008; Liu et al. 2010).

Diagnosis of genus and species by monotypy. Rhamphorhynchine distinguished by three autapomorphies: relatively short skull that forms only 28% of body length (skull+(cervical+dorsal+sacral vertebrae)); short, slender pteroid with a knob-like distal expansion; prepubis with relatively slender distal process.

Description. Preservation. The specimen is largely preserved in articulation and was likely complete when found, but appears to have lost the distal portion of the tail during collection ( Fig. 1, Table 1). The skeleton is preserved, for the most part, in dorsal view on D3080, and in ventral view on D3081. Unfortunately, the plane of separation between the two slabs was rather uneven, passing through the bones, rather than leaving the skeleton on one slab and impressions on the other. Consequently, fragments of each bone are preserved on both slabs. This, and in some cases, the poor preservation of bone, frequently hinder the identification of details, but by combining evidence from both slabs it is possible to determine the location, general shape and size of most skeletal elements.

The split through the skull appears to have been somewhat below half-height, leaving rather fragmented remains of the rostrum and cheek bones on D3080, overlying the mandibles ( Fig. 2). The position and shape of the main openings in the skull can be roughly determined and some of the individual cranial elements such as the quadrate can be observed. An almost complete sclerotic ring from the left orbit now lies just to the left of the cranium. In addition, the cranium appears to have become slightly disarticulated from the spinal column – possibly a preservational feature related to postburial compression of this markedly three-dimensional structure. Except for the tip of the tail, the spinal column is complete and fully articulated. Generally, the location and size of individual vertebrae can be determined, but cervicals #4 and #5 are poorly preserved and the sacrals are difficult to distinguish. The rib cage, consisting of ribs and gastralia, is remarkably well preserved and although dorsoventrally compressed, many of the ribs appear to retain their original location and orientation with regard to the spinal column and one another. Notes: -, inapplicable. Abbreviations: Wph, wing-phalanx.

Despite the considerable compression they must have experienced during burial, the scapula and coracoid of each shoulder girdle, while slightly displaced from each other, do not appear to have moved far from their original position in life, at least in dorsal view. The sternal ends of the coracoids are hidden beneath the spinal column. A broad, rather indistinct structure preserved beneath the spinal column level with the distal termination of the scapulae probably represents the sternum.

The forelimbs are both tightly folded at the elbow and at the wing-metacarpal–wing-finger joint. Indeed, in the latter case each wing-finger has rotated through a half circle and is now folded back against the wing-metacarpal, although on the right side wing-phalanx 1 has become slightly displaced from the roller joint on the distal end of the wing-metacarpal. Digits I–III of the left hand are largely obscured by the overlying skull, while the wing-finger, which curves medially beneath the trunk and the right hind limb, is almost complete, lacking only the distal portion of wing-phalanx 4, preserved as an impression. Phalanx 2 of the right wing-finger is broken, the distal fragment oriented at an obtuse angle to the larger portion and displaced somewhat toward the proximal end of the latter. The right wing-finger phalanx 3 is slightly displaced from, but continues the natural line of the distal fragment of wingphalanx 2, but wing-phalanx 4 is missing.

The hind limbs are folded beneath the tail. In both cases the femur projects laterally from the pelvis region while the crus is folded back sharply at the knee, almost perpendicular to the femur, and no longer in natural articulation. Both the left and right metatarsus and pes were originally complete and except for some displacement of the fourth and fifth digits of the right pes, are preserved in articulation.

Skeletal anatomy. Seen from above the skull has a subtriangular profile reaching its greatest transverse width at the level of the mandibular articulation. The skull is relatively broad, its maximum width attaining almost 50% skull length. By contrast in other rhamphorhynchids where this value can be measured, for example ‘ Parapsicephalus ’ purdoni and Rhamphorhynchus muensteri , it is 35% or less, although in juveniles of the latter, which have relatively broader skulls, it may reach 40% ( Wellnhofer 1975b; Bennett 1995). The skull is relatively short and forms only 28% of body length (skull + vertebral column excluding the tail) compared to a minimum value of 34% for juveniles of Rhamphorhynchus muensteri ( Wellnhofer 1975b) and similar or higher values for other rhamphorhynchines. The same relationship is to be found with regard to limb elements: for example, the skull of Qinglongopterus guoi is only 1.28 times the length of the humerus, whereas in other rhamphorhynchines including Dorygnathus banthensis and juveniles of Ramphorhynchus the minimum value for this relationship is 1.70 ( Wellnhofer 1975b; Padian 2008a). Anurognathids also have relatively short skulls, but their morphology is markedly different from that of Qinglongopterus guoi and other basal pterosaurs ( Wellnhofer 1975b; Bennett 2007).

The tip of the rostrum is drawn out into a slender, prow-like process. The prenarial portion of the rostrum occupies approximately one third of total skull length, while the orbits are relatively large, their anterior-posterior diameter exceeding one third the entire length of the skull. The nares appear to be elongate, slit-like, and lie close to the midline of the skull. The antorbital opening is seemingly more oval in shape and lies posteroventral to the narial opening. The narial and antorbital openings are of similar length to each other, but, as in Rhamphorhynchus ( Wellnhofer 1975a) , much smaller than the orbit.

The jugal is of slender construction, with lacrimal and postorbital processes at an obtuse angle, leaving the ventral region of the orbit open. The quadrate appears to have been inclined somewhat posteriorly with regard to the horizontal axis of the skull. The lower temporal fenestra is relatively large with a subtriangular profile, and thus more similar to the condition in Scaphognathus ( Wellnhofer 1975b) and Dorygnathus ( Wild 1971; Padian 2008a), rather than Rhamphorhynchus ( Wellnhofer 1975a) . The mandibles appear to have united in a symphysis anteriorly, the tip of which tapers to form an elongate, pointed and slightly upcurved prow, one of the defining characteristics of Rhamphorhynchinae . The mandibular rami are narrow and lath-like, lacking the distinct coronoid eminence present in basal pterosaurs such as dimorphodontids and campylognathoidids ( Unwin 2003).

The dentition consists of at least eight dental alveoli on each side of the rostrum, the posteriormost preserved tooth alveolus located below the anterior end of the orbit. There are at least six alveoli in each mandible. The spikelike teeth are long, slender, gently curved and oval in cross-section with sharp tips. Tooth diameter increased toward the mid part of the tooth row and teeth were relatively well spaced, the gaps between them far exceeding the base diameter of teeth bounding each gap, as is typical for rhamphorhynchids, and unlike the more crowded dentitions of basal pterosaurs such as Eudimorphodon ( Wild 1978) . The first few pairs of teeth were both procumbent and divergent, projecting forward and outward from the dental margins as in the rhamphorhynchines Dorygnathus ( Wellnhofer 1991; Padian 2008a), Rhamphorhynchus ( Wellnhofer 1975a, 1991; Bennett 1995), Angustinaripterus ( He et al.1983) and Sericipterus ( Andres et al. 2010) . Toward the posterior end of the tooth row, the teeth become more vertical.

Figure 3. Close-up of the skull: Qinglongopterus guoi gen. et sp. nov. with some of the main elements labelled. Abbreviations: bt, basipterygoid; lrm, left ramus of mandible; ltf, lower temporal fenestra; pr, prow; q, quadrate; rrm, right ramus of mandible; sl, sclerotic ring; t, teeth. Scale bar = 1cm.

The neck is approximately 60% the length of the dorsal+sacral series (DSV), which falls within the range known for many basal pterosaurs. Assuming that the atlas is buried beneath the skull, there appear to have been nine cervicals, the standard count for basal pterosaurs ( Bennett 2004). The third to seventh vertebrae are relatively short and blocky, the midline length only slightly exceeding the minimum width. Slender ribs project posteriorly from beneath the postzygapophyses. Cervicals eight and nine are somewhat shorter, attaining only two thirds the length of the mid-series of the cervicals. The ninth cervical supported well developed ‘free’ ribs.

There are at least 15 dorsal vertebrae, the first twelve bearing ribs, but not the last three. The first rib, articulating with the first true dorsal (at position ten in the vertebral column) is the most robust. The following ribs show a small, but steady, decline in thickness posteriorly, the last dorsal rib (articulating with dorsal 12) reaching only half the width of the first dorsal rib. Typically, the proximal half of each dorsal rib shows only a very gentle degree of curvature, while the distal half is much more markedly curved. There is an increase in rib length from the first to the sixth, followed by a steady decrease in length, the last rib reaching less than 60% the length of the longest rib.

The sacrum is estimated to have contained four sacrals and was followed by a further four ‘free’ caudals, subsequent to which the caudals became increasingly elongate and were sheathed in long, stiff, thread-like extensions of the prezygapophyses, postzygapophyses and chevrons as in many other basal pterosaurs including Campylognathoides liasicus ( Wellnhofer 1974) , Rhamphorhynchus muensteri ( Wellnhofer 1975a) , Darwinopterus modularis ( Lü et al. 2010a) and Fenghuangopterus lii ( Lü et al. 2010b) .

The element identified as the sternum has a broad sternal plate with bevelled posterolateral margins. Insofar as comparisons can be made it is similar in shape and relative size to the sternum of juvenile individuals of Rhamphorhynchus muensteri ( Wellnhofer 1975a: fig. 8f). The scapula and coracoid are not coossified, indicating that this individual was not osteologically mature. The length of the coracoid is slightly longer than that of the scapula. The proximal end of the coracoid supports a short, rounded process that was interpreted by Padian (1983) as the acrocoracoid process. There is no evidence of a brachial tubercle. The proximal end also bears a large subtriangular buttress of bone that projects at right angles to the shaft, supported part of the glenoid and articulated with the scapula. The scapula is dorsoventrally compressed and exhibits a gentle medial curvature. It tapers along the entire length of the shaft, except at the distal end which is rounded, a little asymmetric and slightly expanded. It might not be coincidental that the distal end of each scapula lies close to the fourth dorsal (= 13 th vertebra), the neural spine of which bears a facet that articulated with the distal end of the scapula in those pterodactyloids that had a notarium.

Forelimb length (sum of principal elements) is approximately 167.1 mm, leading to an estimated maximal wingspan of 344.2 mm. The humerus is relatively robustly constructed, as in many other basal pterosaurs. The proximal articular surface is delimited laterally by a low eminence and continues medially where it is supported by the flange-like medial crest. The deltopectoral crest is located at the proximal end of the humerus, as in other basal pterosaurs, but unlike pterodactyloids where it is displaced distally to varying degrees ( Wellnhofer 1978; Unwin 2003). The deltopectoral crest, which is more or less perpendicular to the long axis of the humerus, is subrectangular in outline with proximal and distal margins that curve into the shaft of the humerus and a gently convex anterior margin. The shaft of the humerus is slightly waisted, reaching its narrowest at about mid-length. The distal termination is strongly expanded and markedly asymmetric, the buttress supporting the radial condyle projecting well above the level of the shaft.

The ulna is slightly more robust than the radius, and in their general morphology these elements correspond quite closely to those of Rhamphorhynchus . Notably, the ulna and radius are relatively elongate, reaching 159% the length of the humerus, as in other rhamphorhynchids ( Unwin 2003), distinguishing them from more basal clades such as Dimorphodontidae and Campylognathoididae where the forearm is relatively short (111–144%). The distal termination of the radius is markedly convex, rounding into the ventral margin of the shaft ventrally, but meeting the dorsal margin of the shaft at a sharp angle dorsally. The distal termination of the ulna has a well developed dorsal condyle.

The carpus consists of multiple unfused elements in both the proximal and distal row, further illustrating the osteological immaturity of this individual. The proximal row contains two narrow tablet-like bones, the more medial of which is likely the radiale, while the lateral is probably the ulnare. The distal row contains a large lateral element that contacts the wing-metacarpal, and several medial elements. The pteroid ( Fig. 2b) is relatively short, only 22.5% the length of the humerus. It has a strongly expanded proximal articulation and a convex, knob-like distal expansion that appears to have been an articular condyle. Consequently, the pteroid looks remarkably similar to intermediate phalanges of the manual claw digits, but based on their location, orientation and the completeness of the manus in this individual, it is certainly not a displaced digital phalanx. As interpreted here, the pteroid morphology implies the presence of a second pteroid element that was more distal to the first. There is no direct evidence for such an element, although it is possible that this was unossified, or lost during collection. Rhamphorhynchus is the only other pterosaur known, at present, in which the pteroid consists of two elements although in this case the basal element is relatively shorter ( Wellnhofer 1975a: fig 12a, b).

Metacarpals I–III contact the distal carpal row, as in all other non-pterodactyloids, extend to the distal end of the metacarpus, and are all the same length, as in other breviquartossans ( Unwin 2003). The fourth (wing) metacarpal is a stout bone, with a proximal articulation that is strongly expanded, especially ventrally and with a well developed bicondylar articulation distally. At just over half the length of the humerus, the relative size of the wingmetacarpal is typical for rhamphorhynchids, but somewhat shorter than in Rhamphorhynchus where in individuals of comparable dimensions it averages a little over 60% the length of the humerus ( Wellnhofer 1975b, Table 2).

The manus exhibits the standard phalangeal formula of 2.3.4.4. As in other pterosaurs the penultimate phalanges in digits I–III are highly elongate while other intermediate elements (II1, III2) are short and nubbin-like. The claws are relatively large, at least twice the size of those in the pes. The proximal half of each ungual is relatively deep with a larger flexor tubercle, while the distal half is strongly recurved and sharply tipped.

The wing-finger forms a relatively large proportion of forelimb length (67%), an autapomorphy of Rhamphorhynchinae ( Unwin 2003) . Wing-phalanges 1 and 2 are of similar length, while phalanx 3 is somewhat shorter and phalanx 4 is even shorter but lacks its distal tip and may originally have been longer. These proportions are unusual within non-pterodactyloids, occurring in Nesodactylus hesperius ( Colbert 1969) , but not Dorygnathus banthensis ( Padian 2008a) and some, but not all, examples of Rhamphorhynchus muensteri ( Wellnhofer 1975b: Übersichtstabelle). These unusual proportions are also observed in Darwinopterus modularis ( Lü et al. 2010a) and some basal pterodactyloids such as Pterodactylus kochi ( Wellnhofer, 1970) . Wing-phalanges are straight or slightly curved and seem to be rounded in cross-section, lacking a deep groove in the posterior margin, as described in Rhamphorhynchus ( Wellnhofer 1975a) and Nesodactylus ( Colbert 1969) , although this feature may be obscured by burial or compression. A large rounded epiphysis that forms part of the extensor tendon process on the proximal termination of wing-phalanx 1 is not fused to the phalanx.

The preacetabular process of the ilium was elongate. The anterior third has a gentle medial curvature. The pubis has a straight, vertical anterior profile, while the ventral half of the posterior margin is deeply excavated showing that there was only a relatively short contact between the pubis and ischium immediately ventral to the acetabulum. The ischium has a quadratic outline with a slightly concave dorsal margin, while the remaining margins are gently convex. The prepubis ( Fig. 2c) consists of a gently curved spar of bone. The proximal half is relatively broad, narrowing at about the mid point into a more slender distal process. At this same point the main spar gives rise to a small, short, lateral process.

The hind limb (femur+tibia+metatarsal III) is relatively short, reaching less than 90% the length of the DSV. This is typical for rhamphorhynchines such as Rhamphorhynchus and Dorygnathus which also have relatively short legs, but unlike other non-pterodactyloids where the hind limb exceeds the length of the DSV.

The femur is lightly constructed and relatively short, only 68.5% the length of the humerus, a proportion that is almost the lowest for non-pterodactyloid pterosaurs ( Unwin 2003: table 2). The slightly waisted neck between the femoral head and the shaft is directed at about 145° to the shaft. The articular condyle is gently convex with strongly convex margins, and the greater trochanter is relatively small and poorly developed. The shaft is slightly constricted, reaching a minimum width at about midlength. The distal end is bicondylar, but the condyles are only weakly developed.

The tibia is straight and tapers evenly from the proximal to the distal end. In anterior aspect, the proximal termination has a gently convex profile. The elongate splint-like fibula is closely appressed, but seemingly not fused, to the tibia. The fibula tapers distally, terminating at approximately 65% the length of the tibia. The astragalus and calcaneum, which do not appear to be fused to the distal end of the tibia, are rather indistinct and seemingly not fully ossified.

Metatarsals I–IV are long, slender, straight, tightly bundled proximally, and show only a low degree of divergence from one another distally. At 46% the length of the tibia, metatarsal III is relatively elongate, as in other rhamphorhynchines, whereas this value is somewhat less in many other non-pterodactyloids. Metatarsal I is slightly shorter than metatarsals II and III, which are the same length, while metatarsal IV is markedly shorter than these (<85% length of mt III). Metatarsal V is short and stubby with a hemispherical proximal end and a blunt distal end.

The pedal phalangeal formula is 2-3-4-5-2. The second and third digits are of similar length and slightly longer than the first and fourth, which are also of a similar length to one another. The relative lengths of the phalanges correspond almost exactly to those of Rhamphorhynchus ( Wellnhofer 1975a, fig 17). Penultimate phalanges are highly elongate in digits I to III, but shorter than the basal phalanx in digit IV. Phalanx two of digit V is gently curved and slightly longer than the first phalanx which is straight and tapers, being significantly broader proximally, than distally. Pedal unguals are small, slender and only very slightly recurved.

Ontogenetic status. D3080/1 exhibits several features which suggest that this individual was a juvenile rather than a subadult, or an adult: relatively large orbit; separate scapula and coracoid; poorly ossified sternum; lack of fusion between carpals; absence of fusion between the extensor tendon epiphysis and wing-phalanx I; pubis and ischium not fused to each other, or to the sacral ribs; astragalus and calcaneum not fused to each other, or to the tibia. Comparison with a closely related form, Rhamphorhynchus muensteri (see below), which is represented by a nearly complete postnatal ontogenetic series (see Wellnhofer 1975b; Bennett 1995), and details of ontogeny in other pterosaurs ( Wellnhofer 1970; Bennett 1993, 1996) further suggest that D3080/1 was a juvenile. While composite elements such as the syncarpals and tibiotarsus are not fused, most individual elements, including carpals and tarsals, are well ossified and have clear distinct outlines. This differs, for example, from the smallest (and presumably youngest) individuals of R. muensteri in which the tip of the dentary is unossifed and other bones are incompletely ossified and have simple shapes ( Bennett 1995). In addition, external bone surfaces appear to be fairly well ossified and do not exhibit the coarse, fibrous texture evident in the rostrum and limb bones of embryos ( Chiappe et al. 2004; Ji et al. 2004; Wang and Zhou 2004) or hatchlings (Unwin unpub data). These features suggest that D3080/1 was a mid-term or possibly even late stage juvenile, rather than a hatchling or very young juvenile.

That D3080/1 was not a subadult is shown by the complete lack of fusion between individual bones, and the similarity in some key proportions (orbit length/skull length = 0.36; wing-finger length/forelimb length = 0.67; wingfinger less than seven times humerus length) to juvenile individuals of Rhamphorhynchus muensteri ( Bennett 1995) . The morphology of the sternum, humerus and pelvis of D3080/1 also compares more closely to juvenile examples of R. muensteri , rather than larger, adult individuals.