[D] Confuciusornis sunia [~/~]
Describer
Hou, 1997
Time
Cretaceous Early Barremian
Classification
Saurischia Theropoda Tetanurae Coelurosauria Maniraptora Avialae Confuciusornithidae
Fossilsite
Chaomidianzi Formation ( previously referred to as the lower section of the Yixian Formation), Liaoning, China
Fall Under
Confuciusornis sanctus ~/~
Info
Confuciusornis sanctus ~/~ (Hou, Zhou, Gu and Zhang, 1995) > Confuciusornis chuonzhous ~/~ (Hou, 1997) > Confuciusornis sunia ~/~ (Hou, 1997) > Confuciusornis dui ~/~ (Hou, Martin, Zhou, Feduccia and Zhang, 1999) > Confuciusornis meidus ~/~ [Nomen Nudem] (Hou, 1997) > Confuciusornis shuzhi ~/~ [Nomen Nudem] (Hou, 1997)
Confuciusornis suniae* sp. nov.
Type: A complete skeleton provisionally housed at the Institute of Vertebrate Paleontology and Paleoanthropology under catalog #V11308. The specimen was provided for study by Mr. Shikuan Liang of the Chaoyang Municipal Paleontological Repository, Liaoning province.
Locality and stratigraphic position: Stratigraphically, the specimen occurs slightly higher than C. sanctus, being derived from the Late Jurassic white mudstones of the basal Yixian Fm. at the village of Sihetuncun, Chaoyang Co., Liaoning Province.
Etymology
“Sun” being pinyin romanization for the surname of the wife of the private collector, Mr. Shikuan Liang, who provided the specimen to IVPP for study. The nomenclature is in dedication of her virtuous character. The author stands in complete admiration of the couple’s sincere, faithful, and loving relationship, and concurs with colleagues that this new species of Confuciusornis should commemorate Ms. Sun through her family name.
Diagnosis
The rostral end of the premaxilla has a particularly open vacuity, the nasal process is elongated, external nares are large, frontals are short, and parietals are well developed. Cervical vertebrae are particularly flattened and broadened with small pleurocoels laterally; neural spines are narrow and extremely low anteroposteriorly. Thoracic vertebrae are long and thin with deep and long grooves within the pleurocoels. Lumbar diapophyses are fused. The posterior most three lumbar diapophyses are fused, sacral diapophyses are fused to the medial wall of the ilia, sacral neural spines are fused, and caudal vertebrae are basically fused. . Sternal body is narrow and long with deep elongated lateral recesses.
Description
This specimen is in an exceptional state of preservation due to the attention provided to it by Mr. Liang, who sent the specimen still embedded in its matrix to IVPP where it underwent an extended period of meticulous preparation on its dorsal side by Mr. Yutong Li. Detailed scrutiny of this specimen indicates that the merchant who sold it to Mr. Liang made some modifications to the right forelimb.
It is evident that a portion representing the ungual of the first digit to the first metacarpal does not belong to this specimen. This is determined by its coloration in addition to indications that there is a duplicated section of first metacarpal. The left forelimb, however, is completely preserved and indicates that the reconstructed first digit is basically consistent and will have no effect on interpretation of the specimen. At the time of burial, the cranial bones were disarticulated and shifted in position and the cervical vertebrae had also been shifted. An obvious feature is the poor degree of feather preservation; only few impressions are present. These are the only blemishes on an otherwise perfect specimen.
Skull
At the most anterior end of the premaxilla there is an extremely conspicuous Vshaped embayment, a feature that is not documented on any other specimen of Confuciusornis. It is obviously not an ontogenetic feature because the posterior margin of the embayment is completely fused and a medial suture line is unobservable. This character is also absent on all extant and other fossil species as well. There are relatively large grooves with additional delicate and short striations on each of the posterodorsolateral sides of the embayment. Ventral to the larger grooves the bone is flat, narrow, and smooth, lacking any further ornamentation. The nasal process of the premaxilla differs from that of Confuciusornis sanctus by being longer and extending posteriorly from the midpoint, with its posterior section being unfused. The lateral premaxilla forms the anteroventral wall of the external nares. The maxilla is anteriorly oblique, with its ventral portion overlying the posterolateral premaxilla.
The maxilla is a low triangular element with abundant irregular surficial pitting. Its anterior contact is with the premaxilla and posteriorly it gradually attenuates to become thin and narrow, to contact only a portion of the jugal and form the ventral wall of the external nares. A triangular septomaxilla forms the anterior wall of the external nares and lies between the base of the nasal process of the premaxilla and the anterior margin of the maxilla.
Although this element is interpreted to be present on the new specimen of Confuciusornis, its verification requires additional specimens. In the Crocodylia and synapsids, the septomaxilla is present at the posterior margin of external nares, and it is only on the phytosaurs, such as Machaeroprosopus, that it has migrated to the anterior region of the external nares. Obviously, this single character can not justify a close relationship between phytosaurs and birds, as a synthesis of numerous morphological characters is required.
The nasals and frontals are in contact but have been shifted in position. Lateral to the nasals are a pair of relatively large lacrimals which, although also displaced, display a distinct irregular morphology. The frontal is extremely avian in morphology with a broad and thick semicircular orbital rim which extends from the medial margin directly to the posterior margin of the orbit. The sclerotic ring has also been displaced with a portion of it being overlain by the nasal process of the premaxilla but the majority of it is well preserved, being nearly quadrate in morphology.
The anterior wing of the frontal extends directly to the midpoint of the nasals. Posterior to the orbit it is broad and spacious with two mound-shaped processes projecting from both sides of the midline that unite with the parietal suture to form a V-shaped configuration. The parietals are also well preserved, being completely fused at their midpoint with a small process projecting posteriorly and anterior portion being in contact with the frontal through an upside-down V-shaped depression.
The parietal contacts the occipital with a crest or ridge. With the exception of the occipital, the remaining occipital cranial elements have all been displaced, particularly the lateral acoustic process, which is exceptionally projected on the occipital region. The squamosal is compressed against the posterior cranium but is still quite distinct, overlying and compressing the quadrate on the left side. On the right side it is obscured by cranial elements such as the pterygoid. The foramen magnum is posteriorly positioned and has suffered dorsoventral compressional distortion.
The occipital condyle is rather laterally expanded. Although the specimen is prepared in dorsal perspective, a number of the palatal elements are exposed due to the disarticulation of the skull. A relatively acute and thin anterior end of the vomer is exposed on the right side of the lacrimal. A portion of the palatine is exposed between the premaxilla and the maxilla although its complete configuration is unclear. A pair of pterygoids is extremely well preserved with a slightly curved basal region which becomes thin as it extends directly laterally. The left posterior maxilla has been compressed upon the anterior portion of the mandible and the posterior portions of both mandibles lie lateral to two cervical vertebrae.
These two posterior mandibles appear to represent a single side which is probably due to post burial deformation; otherwise they represent two distinct elements. The lateral side of the articular has a
dorsally projected process and its terminus represents one of the three elements of the glenoid fossa. The prearticular extends anteriorly to the medioventral surangular. An angular is well developed and extends directly anteriorly but becomes obscured by the premaxilla at its anterior end; thus only a portion of it is visible between the nasal process of the maxilla and the sclerotic ring. The right dentary has been displaced to lie posteromedial to the right maxilla where anteriorly it is acute and long and posteriorly it is narrow. The splenial is located quite distinctly at the center of the dentary.
Vertebrae
The entire vertebral column is preserved, and with the exception of the displacement of several cervicals, the sequence has not undergone much deformation. The first several anterior cervicals have been crushed together with the scattered cranial elements and thus it is regrettable that the atlas and axis cannot be described. There are at least nine extremely derived cervicals. The centra are amphicoelous and extremely broad and low, with an extremely broadened neural arch which is laterally high and medially concave.
The neural spine is low and narrow. Prezygapophyses are large with their articular facets placed anteriorly and projected laterally. An elongated trough lies laterally on each side of the neural arch and diapophyseal foramena are present. Postzygapophyses are extremely short and centra articular facets are extremely narrow. The thoracic vertebrae are large and circular, slightly flattened ventrally, are medially concave, neural arch is high, neural canal is circular, diapophyses extend anteriorly, and there are relatively deep and complex pleurocoels. Neural spines are extremely high, the anterior centrum articular face is large and expansive, and posterior articular facet is slightly smaller.
There are approximately ten thoracic vertebrae and worthy of note is that three in the lumbar region are fused in their neural spines and diapophyses. They have a relationship with the ilium, although they are still separated from it. The lumbars are also basically fused with the termini of the sacral diapophyses, four sacrals are also fused and there is a relatively large interval between the sacrals and lumbars, allowing a distinct recognition of the two series. Distal termini of the sacral diapophyses contact the medial wall of the ilium and neural spines are fused together with the lumbars. This fusion is more derived and well developed than noted on the Early Cretaceous Sinornis or Cathayornis species.
In the caudal region on this new species of Confuciusornis there are approximately 25 vertebrae with at least ten anterior caudals being unfused, or in natural articulation, and morphologically they resemble the thoracic vertebrae, although the centra are slightly smaller. Posterior to these are approximately 15 fused caudals which represent the first documentation of a pygostyle. Caudal diapophyses are narrow transverse plates, and neural spines are fused to form a single, high, linear medial crest that gradually becomes lower anteroposteriorly. Boundary lines delineating the centra are still quite conspicuous and perforations are still present between the neural spines. This caudal fusion is distinct and differs somewhat from the general condition of the other Early Cretaceous species.
Ribs
Almost none of the cervical ribs are preserved, because there is only a single right proximal end of a posterior rib present which displays a relatively short capitulum and relatively long tuberculum. Sternal ribs are extremely elongated, basically elliptical in cross-section and maintain a relatively long and shallow groove on their posterior side that extends from the proximal to distal end. There is no appreciable distinction between the morphology of a sternal and dorsal rib. On the right side there are four relatively short and thick rib segments but these are clearly not gastral ribs and it is difficult to accurately diagnose whether or not they are sternal ribs. Gastric ribs are distinct and randomly scattered on the block anterior to the fused lumbar vertebrae and lateral to the thoracic vertebrae. They are extremely slender and not as long as the sternal ribs.
Pectoral girdle
The pectoral girdle is extremely well preserved but the only element that may be completely described is the scapula. Only the right proximal end of the coracoid is visible, as the rest is overlain by the humerus. The left coracoid is completely obscured. The furcula is represented only by the two termini of the clavicle branches exposed at the lateral sides of the scapulae. The scapula is exposed dorsally as being completely fused with the coracoid, its relatively deep glenoid is situated at its proximoventral end, its proximodorsal margin is crescentic and relatively broadened, and slightly distal to the proximal end the scapular shaft is its narrowest. The dorsal margin of the scapula (the side approaching the vertebrae) is relatively rounded and from this point posteriorly the blade gradually decreases in thickness. The posterior blade of the scapula is thin and slightly expanded with a reduced terminus.
On the dorsal surface of the proximal shaft there is a shallow groove that runs posteriorly past the midsection where it then attenuates. The ventral margin of the scapula is also relatively rounded from its proximal end to the scapular blade but at the distal end it becomes thin. Consequently, the entire element is elongated with a thick proximal end and thin distal end. The coracoid basically resembles the morphology of C. sanctus, having an extremely expanded and thick proximal end and an extremely thick dorsal margin where it contacts the scapula and composes the anterior margin of the glenoid fossa. There is an expansive surface for articulation with the clavicle, the lateral margin is projected, and from the proximal point of its fusion with the scapula it forms a 90° curvature extending posteriorly and becoming thin. Its distal end is overlain by the humerus, obscuring its morphology. The clavicle branches of the furcula are extremely robust and its termini articulate with the large clavicle articular facets. The clavicle is the most robust element on the pectoral girdle.
Sternum
Morphology of this element is vague due to being obscured by the scapula, thoracic vertebrae, and ribs. But its right side appears to be planar, heart-shaped, relatively elongated with a rather acute terminus, and a lateral process is absent.
Forelimb
Compared to C. sanctus, the proximal pneumatic foramen on the humerus is smaller, indicating that it is gradually becoming lost, and therefore the depressed region surrounding the foramen has become more shallow and the proximolateral wall has become thin. In addition, the lateral tuberosity is extremely inconspicuous, and the medial tuberosity is not as well developed as on C. sanctus, although a medial crest is still present which extends from the proximal end to the humeral shaft. The shaft is straight, short, slightly laterally compressed, expands laterally toward the distal end, and there is a distinct lateral ectepicondylar process in the shape of a longitudinal prominence distal to which is a low and concave ectepicondyle. These latter features are absent in C. sanctus.
At the distal end of the humerus the medial margin of the entepicondyar process curves toward the midsection of the shaft but an actual entepicondyle is indistinct. Although there is no true olecranon fossa, there is a small depressed region. Distolaterally, there is a distinct autapomorphic ulnar condylar groove which is undocumented on any other fossil bird, although extant taxa possess a similar groove which is extremely shallow. The radius and ulna are relatively robust, and the straight-shafted radius being slightly thinner than the ulna. The proximal ulna is slightly laterally oblique and is expanded with a large humeral articular facet that is slightly concave at its midsection.
A large spaceous cavity is present proximodorsally that extends as a distal groove and gradually narrows, but upon reaching the distal end it again broadens and finally, at the terminus of the ulna, again forms into a relatively deep cavity. At the distoventral ulna there is a projected and laterally expanded angle in addition to an irregularly shaped ulnare in articulation. The radius contacts the ulna with a proximal process, its proximal end is small, has an extremely narrow articular facet for the humerus, and a shallow groove is present proximally which extends and disappears by the midportion of the shaft. The cross-section of the shaft distal to the midportion is nearly circular but at its distal end it expands into a boot where it articulates with what appears to be a radiale.
There has been no displacement of the carpal elements, which is a rare condition. The carpal sequence is extremely well preserved and their morphology is distinct. The ulnare is irregular in shape and rather spherical with a large articular facet for the ulna but smaller articular facets for the other carpals and metacarpals. It predominantly articulates with MtIII. The radiale is large, boot-shaped, maintains a flat and elongated articular facet for the radius, and a large, deep, and concave facet for MtII. Between the ulnare and radiale are an additional two small spherical carpals.
The metacarpals are slightly distinct from those on C. sanctus. This is predominantly displayed in the more robust nature of MtII and the nature of its articulation with the carpals, as its dominant proximal articular facet is large and crescentic. The first metacarpal is slightly smaller and has been displaced distally. MtIII is laterally compressed to become thin and weak in addition to being slightly reduced. Its shaft is slightly curved approaching MtII, proximal articular facet is extremely small, and although the distal articular facet for the phalanges is slightly inflated, it is not expanded. MtI is ladle-shaped with a relatively thin distal end that maintains a relatively large and concave articular facet for the first phalanx.
The thin lateral edges of MtI are nearly as thin as those on MtII and it is in tight contact with the latter with the medial margin becoming thicker. This metacarpal’s proximal articular surface is extremely narrow, is not on the same plane as the proximal surface of MtII, and has actually lost an articular facet for the carpals. MtII is the most robust of these elements with a proximal surface as a large crescentically shaped trochlea, rather resembling the laterally surfaced morphology of extant taxa in which carpal articulation occurs proximolaterally. The lateral margins of the trochlea are expanded with a relatively deep medial depression that appears to represent a medial ligamental fossa. The shaft is robust with a short groove at its proximal end. Distally, the articular facet for the first phalanx of digit two is large with a process that extends to articulate with the expansive facet on the phalanx.
The articular surface at the proximal end of the first phalanx of digit I is convex and maintains a process that extends to facilitate a better articulation with the metacarpal trochlea. The ungual of digit I is the largest and most recurved in the series, resembling the condition of C. sanctus. Digit II has three phalanges the first of which is short and broad with a large proximal articular facet. The second phalanx is larger with a more robust proximal end and a distal end that gradually thins to articulate with an extremely small ungual. Digit III has five phalanges which are all short, with the exception of the ungual. The first and second phalanges of digit III are particularly short. Digit III is the most slender and weakest in addition to being slightly shorter than digit II, and ts ungual is relatively large and hooked.
Pelvic girdle
In dorsal perspective, the pelvic girdle is extremely well preserved with all three elements in natural articulation. The iliac body is elongated as is the preacetabular region which is also expanded with a relatively thick dorsal margin, an extremely thin ventral margin, and rather rounded anterior margin. The acetabular region is thickened, its anterior margin is projected to form a process, although it is unlike that on extant taxa, where there is a dorsal antitrochanter, and as such the position of the C. suniae process differs. Worthy of note is that dorsal to the acetabulum on the ilium there is a posterior, obliquely directed, long, and deep groove that nearly reaches the end of the ilium. The ilium posterior to the acetabulum is much shorter than the anterior portion and is narrow and rounded, with a slightly acute terminus. The ischium is extremely autapomorphic as there are three processes extending off the main body: The first is relatively short, broad, and forms the posterior wall of the acetabulum. The second process is plate-shaped, extends, and expands obliquely dorsally from the distomedial side of the ilium to the vertebral column, appearing as though it encircles the most posterior portion of the synsacral vertebrae. The third and largest process is one which extends posteriorly to the distal end. As previously stated, this morphology it is extremely autapomorphic with both ischia projecting to form broad and spacious arms that laterally encircle the unfused synsacral caudal vertebrae. They become broad and expansive at the terminus of the ilium and have a reflected angle at their distal ends that are medially constricted and a dorsal margin that is relatively thick. The ventral margin is linear with a terminus that extends directly to the ventral side of the synsacral caudal vertebrae. Whether the distal ends are fused or not is unclear as they are overlain completely by the caudals.
Pubis
The pubis is still the longest element of the three sacral elements but is more conservative and simple in morphology compared to the well developed anterior ilium and autapomorphic ischium posterior to it. It is present as a narrow, elongated, and thickened plate with its anterodorsal margin composing the ventral wall of the acetabulum. The dorsal section extends anteroventrally to surpass the anterior margin of the acetabulum, which represents the origin of the “pectineal process” of extant avian forms. The pubes extend posteriorly and at the lateral side of the ischium extend directly to the most posterior fused caudal centra but they do not expand. There is a very slightly projected, low dorsal ridge and the terminus is rounded.
Hindlimb
On the right hindlimb elements distal to the right tarsometatarsus are missing but the left side is exceptionally complete. This limb is extremely robust with a very slightly curved, relatively long, and robust femur. Although its head is rather enlarged, it nevertheless lacks a neck, resembling the morphology of its reptilian ancestors. Distinctive morphology of the femur includes the particularly well developed fourth trochanteric ridge proximally, which is larger than on many archaic reptiles. The fourth trochanter on the extant Alligator sinensis is relatively well developed and projects laterally but it is positioned relatively low on the shaft. The shaft of C. suniae is very slightly anteroposteriorly curved and the posterodistal end has an elongated and broadened trough initiating at the lower half of the shaft, and the lateral margins are thinned. The distal end maintains a deep trough, the position of which corresponds to the popliteal region on extant birds and thus it probably represents the attachment for Popliteus musculature. Distomedial and distolateral condyles are well developed, the intercondylar groove is narrow, a short process extends posteriorly off the lateral condyle, while another process on the fibular side facilitates the Flexor musculature. Between these processes there is a deep depression. The medial condyle is relatively broad and large with a particularly noticeable depression dorsal to it.
There is a small triangular bone which has been modified from tendon situated lateral to the distomedial condyle that should represent a patella. The fibula is extremely slender and exceeds the length of the tibiotarsus by three-quarters. Its proximal end is relatively broad with a convex articular surface. On the medial side of the shaft there is a ridge to facilitate the tibiotarsus. The tibiotarsus is robust and is slightly longer than the femur. Both termini are not very expanded, the proximal articular facet is extremely planar and articulates principally with the medial condyle of the femur. From an anterior perspective, the proximal end appears to maintain a lateral cnemial crest which resembles extant Aves by being a relatively large semi-circular structure. On the posterolateral shaft there is a relatively long fibular crest. Extant Aves possess a proximomedial cnemial crest but on this specimen it is represented only by a small process. The shaft is slightly compressed anteroposteriorly. Distally, genuine medial and lateral condyles are absent and instead there are two differentially sized articular facets, the medial of which is relatively long and broad while the lateral facet is relatively short. There is a depression dorsal to these articular facets which represents the location for the subsequent development of the supratendinal bridge. On the lateral side of the distolateral articular surface, there is a relatively pronounced bulge, suggesting the presence of an ectepicondylar depression. Whether or not a depression is present on the entepicondylar region is still unknown due to lack of specimen preparation.
The tarsometatarsus is the shortest element on the hindlimb, being relatively broad. and exposed on its posterior side. Its proximal end is the only portion of it that is fused, which is distinct from other specimens, but resembles the Early Cretaceous birds from Western Liaoning Province. MtIII and MtIV are nearly equivalent in length, MtV is only visible in proximal cross section as the majority of it, including the proximal articular surface, is missing, but in lateral perspective it appears to be relatively flat. In cross-section, the shaft of MtIII appears to be hollow. MtI is located posterodistal to MtII with the two being in tight association. The distal trochleas on all the metatarsals are relatively broad with that of MtIII being the largest. There is an extremely narrow groove on the posterior MTIV which becomes deeper and broader approaching its distal end. In posterior perspective the proximal ends of MtII and MtIII are spaciously broad, MtIV is only visible at the lateral margin, and only toward the distal end is its shaft noted to also expand and constitute a major portion of the tarsometatarsus. Distally MtIII and MtIV are equivalent in length and MtII is the shortest with its trochlea situated proximal to that of MtIII.
Confuciusornis suniae measurements (mm).
Premaxilla length 39.0
Premaxilla to anterior margin of nares 12.0
Preserved maxilla length 26.0
Maximum breadth of preserved parietal 26.0
Pterygoid length 24.0
Breadth of mandible articulation 7.0
Breadth of cervical 6 9.0
Length of cervical 6 8.0
Length of cervical 6 neural spine 2.5
Scapula length 42.5
Proximal breadth of coracoid 7.5
Distance between clavicle termini 21.0
Sternum length 43.0
Approximate breadth of sternum 24.0
Approximate length of sternal rib 49.0
Approximate breadth of sternal rib 1.7
Humerus length 51.0
Approximate breadth of proximal humerus 28.0
Length of proximal humeral pneumatic foramen 6.0
Humerus breadth at midshaft 10.5
Breadth of distal humerus 1.05
Ulna length 46.0
Ulna breadth at midshaft 4.5
Radius length 44.0
Radius breadth at midshaft 3.0
Radiale length 6.5
Ulnare length 5.5
Metacarpal I length 8.0
Metacarpal I breadth 5.5
Metacarpal II length 25.5
Metacarpal II proximal breadth 7.5
Metacarpal II midshaft breadth 4.0
Metacarpal III length 18.0
Phalanx of manus digit I length 19.0
Ungual of manus digit I length (including sheath) 20.0
First phalanx of manus digit II length 18.5
First phalanx of manus digit II breadth 6.0
Second phalanx of manus digit II length 20.0
Ungual of manus digit II length 8.0
First phalanx of manus digit III length 5.0
Third phalanx of manus digit III approximate length 14.0
Fourth phalanx of manus digit III length 15.0
Ungual of manus digit III length 16.0
Length of thoracic 8 5.0
Breadth of thoracic 8 anterior articular surface 4.0
Synsacral lumbar vertebrae approximate length 13.0
Synsacral lumbar vertebrae diapophyseal breadth 11.0
Synsacral sacral vertebrae length 18.0
Synsacral sacral diapophyseal breadth 14.0
Synsacral caudal vertebrae length 32.0
Synsacral caudal vertebrae breadth 8.0
Ilium length 33.5
Ilium preacetabular length 23.0
Anterior ilium approximate breadth 8.0
Preserved length of ischium 22.5
Transverse process of ischium length 10.0
Pubis length 47.0
Pubis terminal fusion breadth 2.0
Femur length 45.0
Femur proximal breadth 10.0
Femur distal breadth 7.0
Tibiotarsus length 55.0
Tibiotarsus proximal breadth 8.0
Tibiotarsus distal breadth 7.0
Fibula approximate breadth 41.0
Fibula proximal breadth 4.5
Tarsometatarsus length 26.0
Tarsometatarsus proximal breadth 6.0
Tarsometatarsus distal breadth 7.0
Pes digit I length including ungual 15.0
Pes digit II length including ungual 22.5
Pes digit III approximate length without ungual 18.0
Pes digit III ungual length 10.0
Pes digit IV length without ungual 19.0
Pes digit IV ungual length 12.0
Phalanges and unguals
The manus and pes digits are all relatively short but the unguals are all relatively large and recurved. Pedal digit I is extremely short but the articular surface for its ungual is extremely broad and thus from a ventral perspective, the distal end is broader than the proximal end and the ventral surface of the shaft is slightly concave. The ungual on pes digit I is the smallest in the series. Digit II has two phalanges, each of which are longer than the pes digit I phalanx, the articular surfaces at each end are expanded, the ventral surface is relatively concave and there is a longitudinal groove present. Lateral grooves on the ungual are particularly well developed. Digit III has three phalanges, the first two of which are relatively long and robust and the third phalanx is relatively short although its proximal end is expanded. The ungual of digit III is equivalent in size to the ungual on digit II. Digit IV has four relatively short and thick phalanges although its length is nearly equivalent to digit III and its ungual morphology is also nearly equivalent to that of digit III. The last phalanx is the longest in the sequence.
Comparison and discussion
This new Late Jurassic primitive avian specimen should be assigned to the genus Confuciusornis based upon following characters: the jaws are edentulous; and have grooved surficial ornamentation and a keratinous beak, the proximal humerus maintains a pneumatocoel, the ungual on digit I is particularly enlarged, the ungual on digit 2 is particularly small, and the sternum is extremely well developed. Justification for the erection of the new species C. suniae is based upon characters distinct from the other known species of Confuciusornis including the autapomorphic “V” shaped vacuity on the premaxilla, elongated nasal process of the premaxilla, transversely expanded cervical vertebrae, extremely small neural spines, sternum heart-shaped and lacking lateral processes, fusion of the lumbar and synsacral vertebrae, and fusion of the posterior caudals. C. suniae corresponds to the size of C. sanctus and as described above, differs from the latter in its extremely distinct “V” shaped surface of the premaxilla, the function of which is still not determined.
If, as in the extant family Anseriformes, it is an adaptation for aquatic habitats, then the rostrum would be spaciously expanded and not have a bifurcated orifice. Therefore, the autapomorphic character of the beak causes one to also consider the corresponding morphology of the hindlimb, in which the fourth digit is as long as the third digit and maintains the largest ungual among the series. Metatarsals III and IV are also similar in length indicating that at very least the species is adapted to riparian habitats and possibly to fully aquatic habitats. Other significant autapomorphies include the arcuate orbital region with an expanded and thickened margin which reinforces the frontal and provides it with an angled edge, which is a rarely documented feature on primitive birds. Whether it represents a conservative function, facilitates a cranial crest, or perhaps represents a combination of the two is unknown. Extant migratory birds have more substantial orbital regions compared to predators, song birds, or ground dwelling birds. And although wading birds have particularly thickened and broadened orbital regions their morphology is unlike that of C. suniae. On the new specimens of C. sanctus, cervical vertebrae are poorly preserved but it is possible the centra are relatively broad.
The cervicals of C. suniae, however are unexpectedly distinct. Firstly, the diapophyses are broadened and there is a lateral projection. Secondly, the centra are thin and expanded. Thirdly, the neural spine is low and small. Fourthly, the anterior diapophyseal articular facets are positioned anterolateral to the centrum articulation and project anteriorly, which implies increased mobility between the cervicals, unlike the simplified general condition of the postzygapophyes being posteriorly compressed upon the prezygapophyses of its counterpart. Moreover, the anterior and posterior articulations are multiple, complex, and not in tight articulation. This may possibly be a function of supplementing articulation because the centra are not sufficiently modified (saddle shaped) and would thus provide more flexibility in the cervical region as opposed to the inflexible and rigid mode of its reptilian ancestors. This specimen’s dorsal vertebrae still retain primitive characters such that individual dorsal vertebrae recovered in the field would be diagnosed as belonging to a small member of the Reptilia and not Aves. The fusion of the lumbar and sacral vertebrae is another apomorphic character for this species and to date is the first documentation of this character in a specimen from the Early Cretaceous of Western Liaoning.
Its occurrence here indicates that not only is this character contemporaneous with the other members of the fauna, but also that the degree of evolutionary development differed between taxa. That is to say that the environment was selecting for this phenomenon. The pneumatocoel in the proximal humerus is conspicuously more reduced than in C. sanctus, which is regarded as an apomorphic character. It was previously predicted that Confuciusornis would possess pneumatocoels independent from the hollow limb shaft, representing a character for flight adaptation. C. suniae slightly postdates C. sanctus and as such the documentation of a conspicuously smaller humeral pneumatocoel contradicts the concept of reducing skeletal weight to adapt to flight, unless the organism is capable of making the pneumatocoels confluent in order to attain the same objective.
A comparison of these two species indicates that both have apomorphic forelimb morphology in which the third metacarpal is fused to the second metacarpal, particularly at their proximal ends, the shafts have become more slender and slightly curved, and there is a tendency to fuse the carpals. Some of the carpals have become fused with the proximal second metacarpal while the proximal end of the second metacarpal has developed a relatively large trochlea, which distinctly resembles later avian forms. Obviously, the forelimb of Confuciusornis is still not a well developed wing and its flight stroke is determined to be very weak because the basic characters determining a wing morphology lie in the reduction and incipient fusion of the carpals while digits become reduced or approach fusion in addition to the extreme reduction in ungual phalanges by the retention of only one or two. Additionally, the length of the humerus still slightly exceeds that of the radius and ulna, further indicating a weak power stroke. An autapomorphic character of C. suniae lies in the relatively robust long bones of the forelimb and hindlimb. This is particularly noticeable in the hindlimb where there is hardly any distinction of robusticity between the femur, tibiotarsus, and the tarsometatarsus. The tibiotarsus is only slightly longer than the femur and the tarsometatarsus does not attain one-half the length of the tibiotarsus.
The short tarsometatarsus is a symplesiomorphic character among primitive birds and is perhaps due to the organisms being not fully adapted to their habitats. However the degree of habitat adaptation may be more fully reflected in the measurement index of the pedal elements. The hindlimb of C. suniae has one additional prominent plesiomorphic character: the fourth trochanter is extremely well developed for facilitating the attachment of the Caudofemoralis and laterally the Caudofemoralis brevis. But there is an extremely well developed fibular groove on the distolateral femoral condyle which is not documented on any other Late Jurassic bird. Another apomorphic character of the hindlimb lies in the well developed fibular crest on the lateral side of the tibiotarsus, which is not documented on C. sanctus or Archaeopteryx. The remaining characters of the hindlimb, however, are all relatively primitive. In general, this new specimen of Confuciusornis is in many respects more derived than other specimens. Undoubtedly, this is due to its occurrence in a higher stratigraphic level and also because it has a greater adaptation for flight, which evolved relatively rapidly and involved the morphological modification of several organs. The degree of modification on this specimen exceeds that for intraspecific variation. Within the evolutionary process, after the appearance of Confuciusornis in western Liaoning, there occurs regional speciation, particularly between different habitats, such that appropriate conditions in isolated habitats would lead to the genesis of species while within a single geographic region, and stimulated by a number of factors (environmental fluctuation for instance), the inheritance of modified fundamental genetic codes would also provide speciation.
C. chuonzhous has a particularly small ungual on the first digit of the pes and the remaining pedal unguals are not very curved, differing distinctly from C. suniae. In 1995, global interest was stimulated after the publications in Nature and Chinese Science Bulletin regarding the discovery of C. sanctus within the Late Jurassic basal mudstones of the Yixian Fm. in the Shangyuan region of Beipiaoshi, western Liaoning. This was followed by subsequent discoveries of numerous specimens, some of which were in pristine condition. The attention provided to Confuciusornis was primarily due to three factors: First, this was the first well preserved specimen of a primitive bird discovered in over a hundred years, since the excavations of Archaeopteryx in Germany, and it was contemporaneous with the latter. Second, despite the numerous characters shared with Archaeopteryx, the Chinese genus is edentulous, representing the oldest member of Aves with a keratinous beak replacing the dentition. Moreover, the humerus possesses the autapomorphic character of a pneumatocoel and the sternum is relatively well developed, differing widely from that of Archaeopteryx. Third, the autapomorphic nature of abundant new Confuciusornis specimens provokes workers into renewed debate regarding the origin of Aves.
Both genera possess plesiomorphic characters for the class Aves, including the presence of feathers, absence of a postorbital, large orbit, a large antorbital fenestra, and a well developed sternum. Furthermore, both maintain cranial elements that are relatively thick and not fully pneumaticized, the nasal process of premaxilla is extremely short, carpals are unfused, digits are separated, there are three large and acute unguals, vertebral centra are amphicoelous, and the hallux is in opposition to the other metacarpals. Both genera share fundamental plesiomorphic characters in addition to distinctly derived characters. That is to say, the two are noticeably distinct by the presence of full dentition on one and the earliest occurrence of a keratinous beak replacing the dentition on the other. In addition to the dentition, the crania of the two genera are also distinct in many characters: The margins of the premaxilla, maxilla, and mandible of Confuciusornis are ornamented with long longitudinal striations between which are troughs, a condition which is absent on Archaeopteryx, and may be interpreted as the genesis of the keratinous beak. Also, after the divergence from its reptilian ancestors, Confuciusornis inhabited ecological niches that differed from Archaeopteryx and its ancestors. However, there is still an extremely primitive character retained on both taxa, represented by the presence of the extremely short process of the premaxilla, which is a character inherited from reptilian ancestors. In addition, the two genera share apomorphic characters, indicating that they lay within the same evolutionary phase (or level). The nasal process of the premaxilla only attains the anterior margin of the external nares, a condition also documented in numerous archosaurs such as the Pseudosuchia, Crocodylia, and several members of the Dinosauria. An additional shared character includes the presence of an exceptionally large splenial, the only difference between the two lies in its slightly more elongated and acutely triangular morphology on Archaeopteryx whereas on Confuciusornis it is long, thin, and plate-shaped with a rounded terminus.
Despite the relatively short premaxilla on Confuciusornis, the nasals and anterior frontals are narrow and long, indicating a more narrow and elongate rostral region than on Archaeopteryx, which has a more broadened anterior frontal. The quadrate morphology on Confuciusornis is basically similar to that of Archaeopteryx although on the latter the quadrate contact with the braincase is a matter of significance as its contact is posterodorsal and because of the absence of the squamosal there is controversy regarding its veracity. The quadrate-squamosal contact in Confuciusornis is distinct. On the ancestral reptilian forms there is an intimate relationship between the quadrate and quadratojugal and on the majority of archosaurs the mandible articulates with both these elements, but it is only in the ornithischian dinosaurs and other related diapsids that the quadratojugal has become reduced and migrated anteriorly to become a portion of the jugal and isolated from contact with the mandible. On extant Aves the quadratojugal and jugal are completely fused and it is only on juvenile specimens that the quadratojugal is independent and in contact with the anteroventral portion of the quadrate. On both Confuciusornis and Archaeopteryx the quadratojugal is still independent and is basically present as a triangular element that is in contact with the posterolateral side of the jugal and fused with the ventrolateral side of the quadrate, but still does not compose part of the jaw articulation.
The posterior portion of the Confuciusornis skull differs from Archaeopteryx as the former is not constricted or narrow and instead is relatively broad. It also differs from the vast majority of coelurosaurian dinosaurs and is more similar to the Crocodiliformes. Postcranially, Confuciusornis is extremely distinct from Archaeopteryx, primarily in its vertebral morphology. To date, there has been no specialized study of Archaeopteryx vertebrae, but current data indicates that the cervical centra are elongated and a ventral keel is absent, whereas in Confuciusornis the centra are relatively short, the neural arch is particularly enlarged, and there is a relatively well developed ventral keel which are characters shared with several thecodonts and Crocodylians. The sternum of Confuciusornis also differs by the presence of a long and relatively deep elliptical groove on its lateral side (it is reported that several small theropods also possess this character) which may represent a character for weight reduction, and also slightly resembles the morphology on Crocodiliformes. Confuciusornis caudal vertebrae are also distinct from Archaeopteryx by having reduced centra with rather elongated diapophyses and displaying a general developmental direction of the gradual and complete loss of the tail, whereas on Archaeopteryx the caudal centra are extremely elongated and nearly all the centra lack diapophyses, with the exception of several centra proximal to the sacrum.
Not only is there an absence of the trend toward tail reduction, but conversely, the tail appears to be functional. The forelimb of Confuciusornis is most characteristic: its proximal humerus is extremely inflated, particularly transversely expanded, and maintainins a pneumatocoel. This morphology is absolutely autapomorphic, although it does resemble the humerus on some Crocodiliformes even to the extent of sharing several characters with extant crocodiles such as Alligator sinensis. The humerus of Sphenosuchus from the Late Triassic-Early Jurassic of South Africa is also particularly broadened with an extremely thickened proximal margin and extremely well developed deltoid process. Obviously, the pneumatocoel on the Confuciusornis humerus is for the reduction of weight and thereby the facilitation of flight. On Archaeopteryx, the proximal humerus is only slightly expanded, resembles a thin rectangular plate, and its humeral head is extremely small and inconspicuous.
On the anconal aspect of the Confuciusornis proximal humerus lateral to the pneumatocoel, there is a relatively long conspicuous crest for facilitating the Posterior latissimus dorsi, which is nearly absent in Archaeopteryx and the coelurosaurs, but is relatively well developed on the Crocodiliformes, where it also facilitates the attachment of the Teres major. The expansion of the proximal humerus on Confuciusornis facilitates the further development of the Biceps which in turn is intimately related to a more flexible pectoral girdle. Additional forelimb distinction lies in the large unguals on its first and third digits, particularly on digit I which is extremely large and recurved, while the ungual on the second digit is reduced. These are functional adaptations and extremely significant as the second digit phalanges are extremely robust, representing the primary element for facilitating flight as opposed to the grasping of branches or tree trunks.
Thus the facility for grasping lay in the first and third digits. Contrary to this, the manus on Archaeopteryx is constructed with the ungual on the third digit being the smallest and the ungual on the second digit retained as an implement for grasping, crawling, and climbing, despite the second digit being more robust as the primary flight mechanism. The caudal regime on both taxa are not modified for flight and it appears as though the feathers functioned only for ornamentation. There are extensive distinctions between sternum morphology of the two genera. Previously, Archaeopteryx specimens either did not preserve a sternum or it was poorly preserved, until the description of A. bavarica (Wellnhofer, 1993), which describes the element as a transversely expanded plate with a slightly crescentic terminus.
The sternum on Confuciusornis is rather well developed with a projected arcuate anterior margin in addition to two well developed anterior processes and a relatively expanded distal end with a slightly acute and small process at its midpoint. Both have obviously yet to develop a carina and are unlike the morphology of later birds; instead the sternum resembles more closely the condition in the Crocodylia and several squamates. Thus, it is extremely significant that the pectoral girdles of these two genera are very similar, such that if one were to consider the morphology of the pectoral girdle as the sole criterion, then both genera appear to be similar enough to represent a single generic rank. This also indicates the two actually share the same basic plesiomorphic condition of primitive members of Aves.
Another significant character shared between the two genera is the fusion of the distal pubes, which on Confuciusornis is not inflated but on Archaeopteryx is inflated to form a pubic boot. But despite their morphological distinction, their derived nature is equivalent, although they have yet to initiate the process of developing the genuine avian “open” pelvic girdle. The fused termini of the Confuciusornis pubes have a suture line represented by a shallow groove. This is quite distinct from the Early Cretaceous avian specimens that have separated pubi. The description of Confuciusornis morphological characters, its resemblance to, and distinction from Archaeopteryx, indicate that these two most primitive birds in the world are extremely distinct from each other, but that they also share numerous plesiomorphic characters which reflect their chronology and also imply that they shared a common reptilian ancestor. The morphology of Confuciusornis indicates that the avian ancestors were not dinosaurs, as is currently fashionable to believe.
This hypothesis is substantiated by the following two points: The avian forelimb (wing) was derived from the Reptilia and transformed into a flight aparatus through 50 adaptive modification and without undergoing reduction. On the contrary, in order to facilitate flight, the radius, ulna, and carpals have become elongated. Thus the reptilian ancestors could not have been those with short forelimbs. The early members of Dinosauria, and particularly the coelurosaurs, all had particularly short forelimbs. Secondly, the avian fusion of the metatarsals differs completely from the dinosaurian mode which is principally represented by the second and fourth metatarsals whereas on birds the principle metatarsal is MtIII. Furthermore, it may be inferred from the morphology of Confuciusornis that more archaic and primitive forms should be present in Middle to Early Jurassic sediments of China and that extensive exposures of Triassic terrestrial sediments should produce genuine avian ancestors.
The Italian genus Megalancosaurus should be subject of more extensive study as its extremely well developed braincase, large orbits, acute rostrum, elongated external nares, and a scapula that resembles the avian condition provoke extreme interest. Perhaps its extremely elongated cervical region was also an adaptation toward an avian mode of life and it is very possible that it represents an avian ancestor. But prior to its formal recognition as such, there is still much more research required. For instance, several workers have suggested that a clavicle and sternum are absent on this genus. This small archosaur has relatively elongated forelimbs; humerus, radius, and ulna are nearly equivalent in length, are gracile, and distinctly rather dexterous. Within the Reptilia this is the only taxon with morphology approaching the avian condition and its age is appropriate for the genesis of the class. If a similar archosaur can be recovered from the Triassic sediments of China then the longstanding controversies regarding the origin of the class Aves will be solved.
Source: Polyglot Paleontologist
Hou, 1997
Time
Cretaceous Early Barremian
Classification
Saurischia Theropoda Tetanurae Coelurosauria Maniraptora Avialae Confuciusornithidae
Fossilsite
Chaomidianzi Formation ( previously referred to as the lower section of the Yixian Formation), Liaoning, China
Fall Under
Confuciusornis sanctus ~/~
Info
Confuciusornis sanctus ~/~ (Hou, Zhou, Gu and Zhang, 1995) > Confuciusornis chuonzhous ~/~ (Hou, 1997) > Confuciusornis sunia ~/~ (Hou, 1997) > Confuciusornis dui ~/~ (Hou, Martin, Zhou, Feduccia and Zhang, 1999) > Confuciusornis meidus ~/~ [Nomen Nudem] (Hou, 1997) > Confuciusornis shuzhi ~/~ [Nomen Nudem] (Hou, 1997)
Confuciusornis suniae* sp. nov.
Type: A complete skeleton provisionally housed at the Institute of Vertebrate Paleontology and Paleoanthropology under catalog #V11308. The specimen was provided for study by Mr. Shikuan Liang of the Chaoyang Municipal Paleontological Repository, Liaoning province.
Locality and stratigraphic position: Stratigraphically, the specimen occurs slightly higher than C. sanctus, being derived from the Late Jurassic white mudstones of the basal Yixian Fm. at the village of Sihetuncun, Chaoyang Co., Liaoning Province.
Etymology
“Sun” being pinyin romanization for the surname of the wife of the private collector, Mr. Shikuan Liang, who provided the specimen to IVPP for study. The nomenclature is in dedication of her virtuous character. The author stands in complete admiration of the couple’s sincere, faithful, and loving relationship, and concurs with colleagues that this new species of Confuciusornis should commemorate Ms. Sun through her family name.
Diagnosis
The rostral end of the premaxilla has a particularly open vacuity, the nasal process is elongated, external nares are large, frontals are short, and parietals are well developed. Cervical vertebrae are particularly flattened and broadened with small pleurocoels laterally; neural spines are narrow and extremely low anteroposteriorly. Thoracic vertebrae are long and thin with deep and long grooves within the pleurocoels. Lumbar diapophyses are fused. The posterior most three lumbar diapophyses are fused, sacral diapophyses are fused to the medial wall of the ilia, sacral neural spines are fused, and caudal vertebrae are basically fused. . Sternal body is narrow and long with deep elongated lateral recesses.
Description
This specimen is in an exceptional state of preservation due to the attention provided to it by Mr. Liang, who sent the specimen still embedded in its matrix to IVPP where it underwent an extended period of meticulous preparation on its dorsal side by Mr. Yutong Li. Detailed scrutiny of this specimen indicates that the merchant who sold it to Mr. Liang made some modifications to the right forelimb.
It is evident that a portion representing the ungual of the first digit to the first metacarpal does not belong to this specimen. This is determined by its coloration in addition to indications that there is a duplicated section of first metacarpal. The left forelimb, however, is completely preserved and indicates that the reconstructed first digit is basically consistent and will have no effect on interpretation of the specimen. At the time of burial, the cranial bones were disarticulated and shifted in position and the cervical vertebrae had also been shifted. An obvious feature is the poor degree of feather preservation; only few impressions are present. These are the only blemishes on an otherwise perfect specimen.
Skull
At the most anterior end of the premaxilla there is an extremely conspicuous Vshaped embayment, a feature that is not documented on any other specimen of Confuciusornis. It is obviously not an ontogenetic feature because the posterior margin of the embayment is completely fused and a medial suture line is unobservable. This character is also absent on all extant and other fossil species as well. There are relatively large grooves with additional delicate and short striations on each of the posterodorsolateral sides of the embayment. Ventral to the larger grooves the bone is flat, narrow, and smooth, lacking any further ornamentation. The nasal process of the premaxilla differs from that of Confuciusornis sanctus by being longer and extending posteriorly from the midpoint, with its posterior section being unfused. The lateral premaxilla forms the anteroventral wall of the external nares. The maxilla is anteriorly oblique, with its ventral portion overlying the posterolateral premaxilla.
The maxilla is a low triangular element with abundant irregular surficial pitting. Its anterior contact is with the premaxilla and posteriorly it gradually attenuates to become thin and narrow, to contact only a portion of the jugal and form the ventral wall of the external nares. A triangular septomaxilla forms the anterior wall of the external nares and lies between the base of the nasal process of the premaxilla and the anterior margin of the maxilla.
Although this element is interpreted to be present on the new specimen of Confuciusornis, its verification requires additional specimens. In the Crocodylia and synapsids, the septomaxilla is present at the posterior margin of external nares, and it is only on the phytosaurs, such as Machaeroprosopus, that it has migrated to the anterior region of the external nares. Obviously, this single character can not justify a close relationship between phytosaurs and birds, as a synthesis of numerous morphological characters is required.
The nasals and frontals are in contact but have been shifted in position. Lateral to the nasals are a pair of relatively large lacrimals which, although also displaced, display a distinct irregular morphology. The frontal is extremely avian in morphology with a broad and thick semicircular orbital rim which extends from the medial margin directly to the posterior margin of the orbit. The sclerotic ring has also been displaced with a portion of it being overlain by the nasal process of the premaxilla but the majority of it is well preserved, being nearly quadrate in morphology.
The anterior wing of the frontal extends directly to the midpoint of the nasals. Posterior to the orbit it is broad and spacious with two mound-shaped processes projecting from both sides of the midline that unite with the parietal suture to form a V-shaped configuration. The parietals are also well preserved, being completely fused at their midpoint with a small process projecting posteriorly and anterior portion being in contact with the frontal through an upside-down V-shaped depression.
The parietal contacts the occipital with a crest or ridge. With the exception of the occipital, the remaining occipital cranial elements have all been displaced, particularly the lateral acoustic process, which is exceptionally projected on the occipital region. The squamosal is compressed against the posterior cranium but is still quite distinct, overlying and compressing the quadrate on the left side. On the right side it is obscured by cranial elements such as the pterygoid. The foramen magnum is posteriorly positioned and has suffered dorsoventral compressional distortion.
The occipital condyle is rather laterally expanded. Although the specimen is prepared in dorsal perspective, a number of the palatal elements are exposed due to the disarticulation of the skull. A relatively acute and thin anterior end of the vomer is exposed on the right side of the lacrimal. A portion of the palatine is exposed between the premaxilla and the maxilla although its complete configuration is unclear. A pair of pterygoids is extremely well preserved with a slightly curved basal region which becomes thin as it extends directly laterally. The left posterior maxilla has been compressed upon the anterior portion of the mandible and the posterior portions of both mandibles lie lateral to two cervical vertebrae.
These two posterior mandibles appear to represent a single side which is probably due to post burial deformation; otherwise they represent two distinct elements. The lateral side of the articular has a
dorsally projected process and its terminus represents one of the three elements of the glenoid fossa. The prearticular extends anteriorly to the medioventral surangular. An angular is well developed and extends directly anteriorly but becomes obscured by the premaxilla at its anterior end; thus only a portion of it is visible between the nasal process of the maxilla and the sclerotic ring. The right dentary has been displaced to lie posteromedial to the right maxilla where anteriorly it is acute and long and posteriorly it is narrow. The splenial is located quite distinctly at the center of the dentary.
Vertebrae
The entire vertebral column is preserved, and with the exception of the displacement of several cervicals, the sequence has not undergone much deformation. The first several anterior cervicals have been crushed together with the scattered cranial elements and thus it is regrettable that the atlas and axis cannot be described. There are at least nine extremely derived cervicals. The centra are amphicoelous and extremely broad and low, with an extremely broadened neural arch which is laterally high and medially concave.
The neural spine is low and narrow. Prezygapophyses are large with their articular facets placed anteriorly and projected laterally. An elongated trough lies laterally on each side of the neural arch and diapophyseal foramena are present. Postzygapophyses are extremely short and centra articular facets are extremely narrow. The thoracic vertebrae are large and circular, slightly flattened ventrally, are medially concave, neural arch is high, neural canal is circular, diapophyses extend anteriorly, and there are relatively deep and complex pleurocoels. Neural spines are extremely high, the anterior centrum articular face is large and expansive, and posterior articular facet is slightly smaller.
There are approximately ten thoracic vertebrae and worthy of note is that three in the lumbar region are fused in their neural spines and diapophyses. They have a relationship with the ilium, although they are still separated from it. The lumbars are also basically fused with the termini of the sacral diapophyses, four sacrals are also fused and there is a relatively large interval between the sacrals and lumbars, allowing a distinct recognition of the two series. Distal termini of the sacral diapophyses contact the medial wall of the ilium and neural spines are fused together with the lumbars. This fusion is more derived and well developed than noted on the Early Cretaceous Sinornis or Cathayornis species.
In the caudal region on this new species of Confuciusornis there are approximately 25 vertebrae with at least ten anterior caudals being unfused, or in natural articulation, and morphologically they resemble the thoracic vertebrae, although the centra are slightly smaller. Posterior to these are approximately 15 fused caudals which represent the first documentation of a pygostyle. Caudal diapophyses are narrow transverse plates, and neural spines are fused to form a single, high, linear medial crest that gradually becomes lower anteroposteriorly. Boundary lines delineating the centra are still quite conspicuous and perforations are still present between the neural spines. This caudal fusion is distinct and differs somewhat from the general condition of the other Early Cretaceous species.
Ribs
Almost none of the cervical ribs are preserved, because there is only a single right proximal end of a posterior rib present which displays a relatively short capitulum and relatively long tuberculum. Sternal ribs are extremely elongated, basically elliptical in cross-section and maintain a relatively long and shallow groove on their posterior side that extends from the proximal to distal end. There is no appreciable distinction between the morphology of a sternal and dorsal rib. On the right side there are four relatively short and thick rib segments but these are clearly not gastral ribs and it is difficult to accurately diagnose whether or not they are sternal ribs. Gastric ribs are distinct and randomly scattered on the block anterior to the fused lumbar vertebrae and lateral to the thoracic vertebrae. They are extremely slender and not as long as the sternal ribs.
Pectoral girdle
The pectoral girdle is extremely well preserved but the only element that may be completely described is the scapula. Only the right proximal end of the coracoid is visible, as the rest is overlain by the humerus. The left coracoid is completely obscured. The furcula is represented only by the two termini of the clavicle branches exposed at the lateral sides of the scapulae. The scapula is exposed dorsally as being completely fused with the coracoid, its relatively deep glenoid is situated at its proximoventral end, its proximodorsal margin is crescentic and relatively broadened, and slightly distal to the proximal end the scapular shaft is its narrowest. The dorsal margin of the scapula (the side approaching the vertebrae) is relatively rounded and from this point posteriorly the blade gradually decreases in thickness. The posterior blade of the scapula is thin and slightly expanded with a reduced terminus.
On the dorsal surface of the proximal shaft there is a shallow groove that runs posteriorly past the midsection where it then attenuates. The ventral margin of the scapula is also relatively rounded from its proximal end to the scapular blade but at the distal end it becomes thin. Consequently, the entire element is elongated with a thick proximal end and thin distal end. The coracoid basically resembles the morphology of C. sanctus, having an extremely expanded and thick proximal end and an extremely thick dorsal margin where it contacts the scapula and composes the anterior margin of the glenoid fossa. There is an expansive surface for articulation with the clavicle, the lateral margin is projected, and from the proximal point of its fusion with the scapula it forms a 90° curvature extending posteriorly and becoming thin. Its distal end is overlain by the humerus, obscuring its morphology. The clavicle branches of the furcula are extremely robust and its termini articulate with the large clavicle articular facets. The clavicle is the most robust element on the pectoral girdle.
Sternum
Morphology of this element is vague due to being obscured by the scapula, thoracic vertebrae, and ribs. But its right side appears to be planar, heart-shaped, relatively elongated with a rather acute terminus, and a lateral process is absent.
Forelimb
Compared to C. sanctus, the proximal pneumatic foramen on the humerus is smaller, indicating that it is gradually becoming lost, and therefore the depressed region surrounding the foramen has become more shallow and the proximolateral wall has become thin. In addition, the lateral tuberosity is extremely inconspicuous, and the medial tuberosity is not as well developed as on C. sanctus, although a medial crest is still present which extends from the proximal end to the humeral shaft. The shaft is straight, short, slightly laterally compressed, expands laterally toward the distal end, and there is a distinct lateral ectepicondylar process in the shape of a longitudinal prominence distal to which is a low and concave ectepicondyle. These latter features are absent in C. sanctus.
At the distal end of the humerus the medial margin of the entepicondyar process curves toward the midsection of the shaft but an actual entepicondyle is indistinct. Although there is no true olecranon fossa, there is a small depressed region. Distolaterally, there is a distinct autapomorphic ulnar condylar groove which is undocumented on any other fossil bird, although extant taxa possess a similar groove which is extremely shallow. The radius and ulna are relatively robust, and the straight-shafted radius being slightly thinner than the ulna. The proximal ulna is slightly laterally oblique and is expanded with a large humeral articular facet that is slightly concave at its midsection.
A large spaceous cavity is present proximodorsally that extends as a distal groove and gradually narrows, but upon reaching the distal end it again broadens and finally, at the terminus of the ulna, again forms into a relatively deep cavity. At the distoventral ulna there is a projected and laterally expanded angle in addition to an irregularly shaped ulnare in articulation. The radius contacts the ulna with a proximal process, its proximal end is small, has an extremely narrow articular facet for the humerus, and a shallow groove is present proximally which extends and disappears by the midportion of the shaft. The cross-section of the shaft distal to the midportion is nearly circular but at its distal end it expands into a boot where it articulates with what appears to be a radiale.
There has been no displacement of the carpal elements, which is a rare condition. The carpal sequence is extremely well preserved and their morphology is distinct. The ulnare is irregular in shape and rather spherical with a large articular facet for the ulna but smaller articular facets for the other carpals and metacarpals. It predominantly articulates with MtIII. The radiale is large, boot-shaped, maintains a flat and elongated articular facet for the radius, and a large, deep, and concave facet for MtII. Between the ulnare and radiale are an additional two small spherical carpals.
The metacarpals are slightly distinct from those on C. sanctus. This is predominantly displayed in the more robust nature of MtII and the nature of its articulation with the carpals, as its dominant proximal articular facet is large and crescentic. The first metacarpal is slightly smaller and has been displaced distally. MtIII is laterally compressed to become thin and weak in addition to being slightly reduced. Its shaft is slightly curved approaching MtII, proximal articular facet is extremely small, and although the distal articular facet for the phalanges is slightly inflated, it is not expanded. MtI is ladle-shaped with a relatively thin distal end that maintains a relatively large and concave articular facet for the first phalanx.
The thin lateral edges of MtI are nearly as thin as those on MtII and it is in tight contact with the latter with the medial margin becoming thicker. This metacarpal’s proximal articular surface is extremely narrow, is not on the same plane as the proximal surface of MtII, and has actually lost an articular facet for the carpals. MtII is the most robust of these elements with a proximal surface as a large crescentically shaped trochlea, rather resembling the laterally surfaced morphology of extant taxa in which carpal articulation occurs proximolaterally. The lateral margins of the trochlea are expanded with a relatively deep medial depression that appears to represent a medial ligamental fossa. The shaft is robust with a short groove at its proximal end. Distally, the articular facet for the first phalanx of digit two is large with a process that extends to articulate with the expansive facet on the phalanx.
The articular surface at the proximal end of the first phalanx of digit I is convex and maintains a process that extends to facilitate a better articulation with the metacarpal trochlea. The ungual of digit I is the largest and most recurved in the series, resembling the condition of C. sanctus. Digit II has three phalanges the first of which is short and broad with a large proximal articular facet. The second phalanx is larger with a more robust proximal end and a distal end that gradually thins to articulate with an extremely small ungual. Digit III has five phalanges which are all short, with the exception of the ungual. The first and second phalanges of digit III are particularly short. Digit III is the most slender and weakest in addition to being slightly shorter than digit II, and ts ungual is relatively large and hooked.
Pelvic girdle
In dorsal perspective, the pelvic girdle is extremely well preserved with all three elements in natural articulation. The iliac body is elongated as is the preacetabular region which is also expanded with a relatively thick dorsal margin, an extremely thin ventral margin, and rather rounded anterior margin. The acetabular region is thickened, its anterior margin is projected to form a process, although it is unlike that on extant taxa, where there is a dorsal antitrochanter, and as such the position of the C. suniae process differs. Worthy of note is that dorsal to the acetabulum on the ilium there is a posterior, obliquely directed, long, and deep groove that nearly reaches the end of the ilium. The ilium posterior to the acetabulum is much shorter than the anterior portion and is narrow and rounded, with a slightly acute terminus. The ischium is extremely autapomorphic as there are three processes extending off the main body: The first is relatively short, broad, and forms the posterior wall of the acetabulum. The second process is plate-shaped, extends, and expands obliquely dorsally from the distomedial side of the ilium to the vertebral column, appearing as though it encircles the most posterior portion of the synsacral vertebrae. The third and largest process is one which extends posteriorly to the distal end. As previously stated, this morphology it is extremely autapomorphic with both ischia projecting to form broad and spacious arms that laterally encircle the unfused synsacral caudal vertebrae. They become broad and expansive at the terminus of the ilium and have a reflected angle at their distal ends that are medially constricted and a dorsal margin that is relatively thick. The ventral margin is linear with a terminus that extends directly to the ventral side of the synsacral caudal vertebrae. Whether the distal ends are fused or not is unclear as they are overlain completely by the caudals.
Pubis
The pubis is still the longest element of the three sacral elements but is more conservative and simple in morphology compared to the well developed anterior ilium and autapomorphic ischium posterior to it. It is present as a narrow, elongated, and thickened plate with its anterodorsal margin composing the ventral wall of the acetabulum. The dorsal section extends anteroventrally to surpass the anterior margin of the acetabulum, which represents the origin of the “pectineal process” of extant avian forms. The pubes extend posteriorly and at the lateral side of the ischium extend directly to the most posterior fused caudal centra but they do not expand. There is a very slightly projected, low dorsal ridge and the terminus is rounded.
Hindlimb
On the right hindlimb elements distal to the right tarsometatarsus are missing but the left side is exceptionally complete. This limb is extremely robust with a very slightly curved, relatively long, and robust femur. Although its head is rather enlarged, it nevertheless lacks a neck, resembling the morphology of its reptilian ancestors. Distinctive morphology of the femur includes the particularly well developed fourth trochanteric ridge proximally, which is larger than on many archaic reptiles. The fourth trochanter on the extant Alligator sinensis is relatively well developed and projects laterally but it is positioned relatively low on the shaft. The shaft of C. suniae is very slightly anteroposteriorly curved and the posterodistal end has an elongated and broadened trough initiating at the lower half of the shaft, and the lateral margins are thinned. The distal end maintains a deep trough, the position of which corresponds to the popliteal region on extant birds and thus it probably represents the attachment for Popliteus musculature. Distomedial and distolateral condyles are well developed, the intercondylar groove is narrow, a short process extends posteriorly off the lateral condyle, while another process on the fibular side facilitates the Flexor musculature. Between these processes there is a deep depression. The medial condyle is relatively broad and large with a particularly noticeable depression dorsal to it.
There is a small triangular bone which has been modified from tendon situated lateral to the distomedial condyle that should represent a patella. The fibula is extremely slender and exceeds the length of the tibiotarsus by three-quarters. Its proximal end is relatively broad with a convex articular surface. On the medial side of the shaft there is a ridge to facilitate the tibiotarsus. The tibiotarsus is robust and is slightly longer than the femur. Both termini are not very expanded, the proximal articular facet is extremely planar and articulates principally with the medial condyle of the femur. From an anterior perspective, the proximal end appears to maintain a lateral cnemial crest which resembles extant Aves by being a relatively large semi-circular structure. On the posterolateral shaft there is a relatively long fibular crest. Extant Aves possess a proximomedial cnemial crest but on this specimen it is represented only by a small process. The shaft is slightly compressed anteroposteriorly. Distally, genuine medial and lateral condyles are absent and instead there are two differentially sized articular facets, the medial of which is relatively long and broad while the lateral facet is relatively short. There is a depression dorsal to these articular facets which represents the location for the subsequent development of the supratendinal bridge. On the lateral side of the distolateral articular surface, there is a relatively pronounced bulge, suggesting the presence of an ectepicondylar depression. Whether or not a depression is present on the entepicondylar region is still unknown due to lack of specimen preparation.
The tarsometatarsus is the shortest element on the hindlimb, being relatively broad. and exposed on its posterior side. Its proximal end is the only portion of it that is fused, which is distinct from other specimens, but resembles the Early Cretaceous birds from Western Liaoning Province. MtIII and MtIV are nearly equivalent in length, MtV is only visible in proximal cross section as the majority of it, including the proximal articular surface, is missing, but in lateral perspective it appears to be relatively flat. In cross-section, the shaft of MtIII appears to be hollow. MtI is located posterodistal to MtII with the two being in tight association. The distal trochleas on all the metatarsals are relatively broad with that of MtIII being the largest. There is an extremely narrow groove on the posterior MTIV which becomes deeper and broader approaching its distal end. In posterior perspective the proximal ends of MtII and MtIII are spaciously broad, MtIV is only visible at the lateral margin, and only toward the distal end is its shaft noted to also expand and constitute a major portion of the tarsometatarsus. Distally MtIII and MtIV are equivalent in length and MtII is the shortest with its trochlea situated proximal to that of MtIII.
Confuciusornis suniae measurements (mm).
Premaxilla length 39.0
Premaxilla to anterior margin of nares 12.0
Preserved maxilla length 26.0
Maximum breadth of preserved parietal 26.0
Pterygoid length 24.0
Breadth of mandible articulation 7.0
Breadth of cervical 6 9.0
Length of cervical 6 8.0
Length of cervical 6 neural spine 2.5
Scapula length 42.5
Proximal breadth of coracoid 7.5
Distance between clavicle termini 21.0
Sternum length 43.0
Approximate breadth of sternum 24.0
Approximate length of sternal rib 49.0
Approximate breadth of sternal rib 1.7
Humerus length 51.0
Approximate breadth of proximal humerus 28.0
Length of proximal humeral pneumatic foramen 6.0
Humerus breadth at midshaft 10.5
Breadth of distal humerus 1.05
Ulna length 46.0
Ulna breadth at midshaft 4.5
Radius length 44.0
Radius breadth at midshaft 3.0
Radiale length 6.5
Ulnare length 5.5
Metacarpal I length 8.0
Metacarpal I breadth 5.5
Metacarpal II length 25.5
Metacarpal II proximal breadth 7.5
Metacarpal II midshaft breadth 4.0
Metacarpal III length 18.0
Phalanx of manus digit I length 19.0
Ungual of manus digit I length (including sheath) 20.0
First phalanx of manus digit II length 18.5
First phalanx of manus digit II breadth 6.0
Second phalanx of manus digit II length 20.0
Ungual of manus digit II length 8.0
First phalanx of manus digit III length 5.0
Third phalanx of manus digit III approximate length 14.0
Fourth phalanx of manus digit III length 15.0
Ungual of manus digit III length 16.0
Length of thoracic 8 5.0
Breadth of thoracic 8 anterior articular surface 4.0
Synsacral lumbar vertebrae approximate length 13.0
Synsacral lumbar vertebrae diapophyseal breadth 11.0
Synsacral sacral vertebrae length 18.0
Synsacral sacral diapophyseal breadth 14.0
Synsacral caudal vertebrae length 32.0
Synsacral caudal vertebrae breadth 8.0
Ilium length 33.5
Ilium preacetabular length 23.0
Anterior ilium approximate breadth 8.0
Preserved length of ischium 22.5
Transverse process of ischium length 10.0
Pubis length 47.0
Pubis terminal fusion breadth 2.0
Femur length 45.0
Femur proximal breadth 10.0
Femur distal breadth 7.0
Tibiotarsus length 55.0
Tibiotarsus proximal breadth 8.0
Tibiotarsus distal breadth 7.0
Fibula approximate breadth 41.0
Fibula proximal breadth 4.5
Tarsometatarsus length 26.0
Tarsometatarsus proximal breadth 6.0
Tarsometatarsus distal breadth 7.0
Pes digit I length including ungual 15.0
Pes digit II length including ungual 22.5
Pes digit III approximate length without ungual 18.0
Pes digit III ungual length 10.0
Pes digit IV length without ungual 19.0
Pes digit IV ungual length 12.0
Phalanges and unguals
The manus and pes digits are all relatively short but the unguals are all relatively large and recurved. Pedal digit I is extremely short but the articular surface for its ungual is extremely broad and thus from a ventral perspective, the distal end is broader than the proximal end and the ventral surface of the shaft is slightly concave. The ungual on pes digit I is the smallest in the series. Digit II has two phalanges, each of which are longer than the pes digit I phalanx, the articular surfaces at each end are expanded, the ventral surface is relatively concave and there is a longitudinal groove present. Lateral grooves on the ungual are particularly well developed. Digit III has three phalanges, the first two of which are relatively long and robust and the third phalanx is relatively short although its proximal end is expanded. The ungual of digit III is equivalent in size to the ungual on digit II. Digit IV has four relatively short and thick phalanges although its length is nearly equivalent to digit III and its ungual morphology is also nearly equivalent to that of digit III. The last phalanx is the longest in the sequence.
Comparison and discussion
This new Late Jurassic primitive avian specimen should be assigned to the genus Confuciusornis based upon following characters: the jaws are edentulous; and have grooved surficial ornamentation and a keratinous beak, the proximal humerus maintains a pneumatocoel, the ungual on digit I is particularly enlarged, the ungual on digit 2 is particularly small, and the sternum is extremely well developed. Justification for the erection of the new species C. suniae is based upon characters distinct from the other known species of Confuciusornis including the autapomorphic “V” shaped vacuity on the premaxilla, elongated nasal process of the premaxilla, transversely expanded cervical vertebrae, extremely small neural spines, sternum heart-shaped and lacking lateral processes, fusion of the lumbar and synsacral vertebrae, and fusion of the posterior caudals. C. suniae corresponds to the size of C. sanctus and as described above, differs from the latter in its extremely distinct “V” shaped surface of the premaxilla, the function of which is still not determined.
If, as in the extant family Anseriformes, it is an adaptation for aquatic habitats, then the rostrum would be spaciously expanded and not have a bifurcated orifice. Therefore, the autapomorphic character of the beak causes one to also consider the corresponding morphology of the hindlimb, in which the fourth digit is as long as the third digit and maintains the largest ungual among the series. Metatarsals III and IV are also similar in length indicating that at very least the species is adapted to riparian habitats and possibly to fully aquatic habitats. Other significant autapomorphies include the arcuate orbital region with an expanded and thickened margin which reinforces the frontal and provides it with an angled edge, which is a rarely documented feature on primitive birds. Whether it represents a conservative function, facilitates a cranial crest, or perhaps represents a combination of the two is unknown. Extant migratory birds have more substantial orbital regions compared to predators, song birds, or ground dwelling birds. And although wading birds have particularly thickened and broadened orbital regions their morphology is unlike that of C. suniae. On the new specimens of C. sanctus, cervical vertebrae are poorly preserved but it is possible the centra are relatively broad.
The cervicals of C. suniae, however are unexpectedly distinct. Firstly, the diapophyses are broadened and there is a lateral projection. Secondly, the centra are thin and expanded. Thirdly, the neural spine is low and small. Fourthly, the anterior diapophyseal articular facets are positioned anterolateral to the centrum articulation and project anteriorly, which implies increased mobility between the cervicals, unlike the simplified general condition of the postzygapophyes being posteriorly compressed upon the prezygapophyses of its counterpart. Moreover, the anterior and posterior articulations are multiple, complex, and not in tight articulation. This may possibly be a function of supplementing articulation because the centra are not sufficiently modified (saddle shaped) and would thus provide more flexibility in the cervical region as opposed to the inflexible and rigid mode of its reptilian ancestors. This specimen’s dorsal vertebrae still retain primitive characters such that individual dorsal vertebrae recovered in the field would be diagnosed as belonging to a small member of the Reptilia and not Aves. The fusion of the lumbar and sacral vertebrae is another apomorphic character for this species and to date is the first documentation of this character in a specimen from the Early Cretaceous of Western Liaoning.
Its occurrence here indicates that not only is this character contemporaneous with the other members of the fauna, but also that the degree of evolutionary development differed between taxa. That is to say that the environment was selecting for this phenomenon. The pneumatocoel in the proximal humerus is conspicuously more reduced than in C. sanctus, which is regarded as an apomorphic character. It was previously predicted that Confuciusornis would possess pneumatocoels independent from the hollow limb shaft, representing a character for flight adaptation. C. suniae slightly postdates C. sanctus and as such the documentation of a conspicuously smaller humeral pneumatocoel contradicts the concept of reducing skeletal weight to adapt to flight, unless the organism is capable of making the pneumatocoels confluent in order to attain the same objective.
A comparison of these two species indicates that both have apomorphic forelimb morphology in which the third metacarpal is fused to the second metacarpal, particularly at their proximal ends, the shafts have become more slender and slightly curved, and there is a tendency to fuse the carpals. Some of the carpals have become fused with the proximal second metacarpal while the proximal end of the second metacarpal has developed a relatively large trochlea, which distinctly resembles later avian forms. Obviously, the forelimb of Confuciusornis is still not a well developed wing and its flight stroke is determined to be very weak because the basic characters determining a wing morphology lie in the reduction and incipient fusion of the carpals while digits become reduced or approach fusion in addition to the extreme reduction in ungual phalanges by the retention of only one or two. Additionally, the length of the humerus still slightly exceeds that of the radius and ulna, further indicating a weak power stroke. An autapomorphic character of C. suniae lies in the relatively robust long bones of the forelimb and hindlimb. This is particularly noticeable in the hindlimb where there is hardly any distinction of robusticity between the femur, tibiotarsus, and the tarsometatarsus. The tibiotarsus is only slightly longer than the femur and the tarsometatarsus does not attain one-half the length of the tibiotarsus.
The short tarsometatarsus is a symplesiomorphic character among primitive birds and is perhaps due to the organisms being not fully adapted to their habitats. However the degree of habitat adaptation may be more fully reflected in the measurement index of the pedal elements. The hindlimb of C. suniae has one additional prominent plesiomorphic character: the fourth trochanter is extremely well developed for facilitating the attachment of the Caudofemoralis and laterally the Caudofemoralis brevis. But there is an extremely well developed fibular groove on the distolateral femoral condyle which is not documented on any other Late Jurassic bird. Another apomorphic character of the hindlimb lies in the well developed fibular crest on the lateral side of the tibiotarsus, which is not documented on C. sanctus or Archaeopteryx. The remaining characters of the hindlimb, however, are all relatively primitive. In general, this new specimen of Confuciusornis is in many respects more derived than other specimens. Undoubtedly, this is due to its occurrence in a higher stratigraphic level and also because it has a greater adaptation for flight, which evolved relatively rapidly and involved the morphological modification of several organs. The degree of modification on this specimen exceeds that for intraspecific variation. Within the evolutionary process, after the appearance of Confuciusornis in western Liaoning, there occurs regional speciation, particularly between different habitats, such that appropriate conditions in isolated habitats would lead to the genesis of species while within a single geographic region, and stimulated by a number of factors (environmental fluctuation for instance), the inheritance of modified fundamental genetic codes would also provide speciation.
C. chuonzhous has a particularly small ungual on the first digit of the pes and the remaining pedal unguals are not very curved, differing distinctly from C. suniae. In 1995, global interest was stimulated after the publications in Nature and Chinese Science Bulletin regarding the discovery of C. sanctus within the Late Jurassic basal mudstones of the Yixian Fm. in the Shangyuan region of Beipiaoshi, western Liaoning. This was followed by subsequent discoveries of numerous specimens, some of which were in pristine condition. The attention provided to Confuciusornis was primarily due to three factors: First, this was the first well preserved specimen of a primitive bird discovered in over a hundred years, since the excavations of Archaeopteryx in Germany, and it was contemporaneous with the latter. Second, despite the numerous characters shared with Archaeopteryx, the Chinese genus is edentulous, representing the oldest member of Aves with a keratinous beak replacing the dentition. Moreover, the humerus possesses the autapomorphic character of a pneumatocoel and the sternum is relatively well developed, differing widely from that of Archaeopteryx. Third, the autapomorphic nature of abundant new Confuciusornis specimens provokes workers into renewed debate regarding the origin of Aves.
Both genera possess plesiomorphic characters for the class Aves, including the presence of feathers, absence of a postorbital, large orbit, a large antorbital fenestra, and a well developed sternum. Furthermore, both maintain cranial elements that are relatively thick and not fully pneumaticized, the nasal process of premaxilla is extremely short, carpals are unfused, digits are separated, there are three large and acute unguals, vertebral centra are amphicoelous, and the hallux is in opposition to the other metacarpals. Both genera share fundamental plesiomorphic characters in addition to distinctly derived characters. That is to say, the two are noticeably distinct by the presence of full dentition on one and the earliest occurrence of a keratinous beak replacing the dentition on the other. In addition to the dentition, the crania of the two genera are also distinct in many characters: The margins of the premaxilla, maxilla, and mandible of Confuciusornis are ornamented with long longitudinal striations between which are troughs, a condition which is absent on Archaeopteryx, and may be interpreted as the genesis of the keratinous beak. Also, after the divergence from its reptilian ancestors, Confuciusornis inhabited ecological niches that differed from Archaeopteryx and its ancestors. However, there is still an extremely primitive character retained on both taxa, represented by the presence of the extremely short process of the premaxilla, which is a character inherited from reptilian ancestors. In addition, the two genera share apomorphic characters, indicating that they lay within the same evolutionary phase (or level). The nasal process of the premaxilla only attains the anterior margin of the external nares, a condition also documented in numerous archosaurs such as the Pseudosuchia, Crocodylia, and several members of the Dinosauria. An additional shared character includes the presence of an exceptionally large splenial, the only difference between the two lies in its slightly more elongated and acutely triangular morphology on Archaeopteryx whereas on Confuciusornis it is long, thin, and plate-shaped with a rounded terminus.
Despite the relatively short premaxilla on Confuciusornis, the nasals and anterior frontals are narrow and long, indicating a more narrow and elongate rostral region than on Archaeopteryx, which has a more broadened anterior frontal. The quadrate morphology on Confuciusornis is basically similar to that of Archaeopteryx although on the latter the quadrate contact with the braincase is a matter of significance as its contact is posterodorsal and because of the absence of the squamosal there is controversy regarding its veracity. The quadrate-squamosal contact in Confuciusornis is distinct. On the ancestral reptilian forms there is an intimate relationship between the quadrate and quadratojugal and on the majority of archosaurs the mandible articulates with both these elements, but it is only in the ornithischian dinosaurs and other related diapsids that the quadratojugal has become reduced and migrated anteriorly to become a portion of the jugal and isolated from contact with the mandible. On extant Aves the quadratojugal and jugal are completely fused and it is only on juvenile specimens that the quadratojugal is independent and in contact with the anteroventral portion of the quadrate. On both Confuciusornis and Archaeopteryx the quadratojugal is still independent and is basically present as a triangular element that is in contact with the posterolateral side of the jugal and fused with the ventrolateral side of the quadrate, but still does not compose part of the jaw articulation.
The posterior portion of the Confuciusornis skull differs from Archaeopteryx as the former is not constricted or narrow and instead is relatively broad. It also differs from the vast majority of coelurosaurian dinosaurs and is more similar to the Crocodiliformes. Postcranially, Confuciusornis is extremely distinct from Archaeopteryx, primarily in its vertebral morphology. To date, there has been no specialized study of Archaeopteryx vertebrae, but current data indicates that the cervical centra are elongated and a ventral keel is absent, whereas in Confuciusornis the centra are relatively short, the neural arch is particularly enlarged, and there is a relatively well developed ventral keel which are characters shared with several thecodonts and Crocodylians. The sternum of Confuciusornis also differs by the presence of a long and relatively deep elliptical groove on its lateral side (it is reported that several small theropods also possess this character) which may represent a character for weight reduction, and also slightly resembles the morphology on Crocodiliformes. Confuciusornis caudal vertebrae are also distinct from Archaeopteryx by having reduced centra with rather elongated diapophyses and displaying a general developmental direction of the gradual and complete loss of the tail, whereas on Archaeopteryx the caudal centra are extremely elongated and nearly all the centra lack diapophyses, with the exception of several centra proximal to the sacrum.
Not only is there an absence of the trend toward tail reduction, but conversely, the tail appears to be functional. The forelimb of Confuciusornis is most characteristic: its proximal humerus is extremely inflated, particularly transversely expanded, and maintainins a pneumatocoel. This morphology is absolutely autapomorphic, although it does resemble the humerus on some Crocodiliformes even to the extent of sharing several characters with extant crocodiles such as Alligator sinensis. The humerus of Sphenosuchus from the Late Triassic-Early Jurassic of South Africa is also particularly broadened with an extremely thickened proximal margin and extremely well developed deltoid process. Obviously, the pneumatocoel on the Confuciusornis humerus is for the reduction of weight and thereby the facilitation of flight. On Archaeopteryx, the proximal humerus is only slightly expanded, resembles a thin rectangular plate, and its humeral head is extremely small and inconspicuous.
On the anconal aspect of the Confuciusornis proximal humerus lateral to the pneumatocoel, there is a relatively long conspicuous crest for facilitating the Posterior latissimus dorsi, which is nearly absent in Archaeopteryx and the coelurosaurs, but is relatively well developed on the Crocodiliformes, where it also facilitates the attachment of the Teres major. The expansion of the proximal humerus on Confuciusornis facilitates the further development of the Biceps which in turn is intimately related to a more flexible pectoral girdle. Additional forelimb distinction lies in the large unguals on its first and third digits, particularly on digit I which is extremely large and recurved, while the ungual on the second digit is reduced. These are functional adaptations and extremely significant as the second digit phalanges are extremely robust, representing the primary element for facilitating flight as opposed to the grasping of branches or tree trunks.
Thus the facility for grasping lay in the first and third digits. Contrary to this, the manus on Archaeopteryx is constructed with the ungual on the third digit being the smallest and the ungual on the second digit retained as an implement for grasping, crawling, and climbing, despite the second digit being more robust as the primary flight mechanism. The caudal regime on both taxa are not modified for flight and it appears as though the feathers functioned only for ornamentation. There are extensive distinctions between sternum morphology of the two genera. Previously, Archaeopteryx specimens either did not preserve a sternum or it was poorly preserved, until the description of A. bavarica (Wellnhofer, 1993), which describes the element as a transversely expanded plate with a slightly crescentic terminus.
The sternum on Confuciusornis is rather well developed with a projected arcuate anterior margin in addition to two well developed anterior processes and a relatively expanded distal end with a slightly acute and small process at its midpoint. Both have obviously yet to develop a carina and are unlike the morphology of later birds; instead the sternum resembles more closely the condition in the Crocodylia and several squamates. Thus, it is extremely significant that the pectoral girdles of these two genera are very similar, such that if one were to consider the morphology of the pectoral girdle as the sole criterion, then both genera appear to be similar enough to represent a single generic rank. This also indicates the two actually share the same basic plesiomorphic condition of primitive members of Aves.
Another significant character shared between the two genera is the fusion of the distal pubes, which on Confuciusornis is not inflated but on Archaeopteryx is inflated to form a pubic boot. But despite their morphological distinction, their derived nature is equivalent, although they have yet to initiate the process of developing the genuine avian “open” pelvic girdle. The fused termini of the Confuciusornis pubes have a suture line represented by a shallow groove. This is quite distinct from the Early Cretaceous avian specimens that have separated pubi. The description of Confuciusornis morphological characters, its resemblance to, and distinction from Archaeopteryx, indicate that these two most primitive birds in the world are extremely distinct from each other, but that they also share numerous plesiomorphic characters which reflect their chronology and also imply that they shared a common reptilian ancestor. The morphology of Confuciusornis indicates that the avian ancestors were not dinosaurs, as is currently fashionable to believe.
This hypothesis is substantiated by the following two points: The avian forelimb (wing) was derived from the Reptilia and transformed into a flight aparatus through 50 adaptive modification and without undergoing reduction. On the contrary, in order to facilitate flight, the radius, ulna, and carpals have become elongated. Thus the reptilian ancestors could not have been those with short forelimbs. The early members of Dinosauria, and particularly the coelurosaurs, all had particularly short forelimbs. Secondly, the avian fusion of the metatarsals differs completely from the dinosaurian mode which is principally represented by the second and fourth metatarsals whereas on birds the principle metatarsal is MtIII. Furthermore, it may be inferred from the morphology of Confuciusornis that more archaic and primitive forms should be present in Middle to Early Jurassic sediments of China and that extensive exposures of Triassic terrestrial sediments should produce genuine avian ancestors.
The Italian genus Megalancosaurus should be subject of more extensive study as its extremely well developed braincase, large orbits, acute rostrum, elongated external nares, and a scapula that resembles the avian condition provoke extreme interest. Perhaps its extremely elongated cervical region was also an adaptation toward an avian mode of life and it is very possible that it represents an avian ancestor. But prior to its formal recognition as such, there is still much more research required. For instance, several workers have suggested that a clavicle and sternum are absent on this genus. This small archosaur has relatively elongated forelimbs; humerus, radius, and ulna are nearly equivalent in length, are gracile, and distinctly rather dexterous. Within the Reptilia this is the only taxon with morphology approaching the avian condition and its age is appropriate for the genesis of the class. If a similar archosaur can be recovered from the Triassic sediments of China then the longstanding controversies regarding the origin of the class Aves will be solved.
Source: Polyglot Paleontologist