New records of perching birds from the latest Paleocene/earliest Eocene Fur Formation of Denmark

Anette Vedding Kristoffersen, Geological Museum, Øster Voldgade 5-7, 1350 Copenhagen

Introduction

Three partly articulated postcranial skeletons (Da.k.7, Da.k.22 and FU100) from the early Tertiary Fur Formation of northern Jutland, Denmark, have been studied. They are preserved as part and counterpart in “cementstone”, which has split along the fossil plane. Splitting and partly imbedding of the bones allows in many cases only a two-dimensional observation, often concealing details. The birds represented by the fossils are approximately the size of the common European wren, Troglodytes troglodytes. The morphology of the hind limb in Da.k.7 and FU100 show that these birds were perching and as a basis for systematic interpretations they have been compared morphologically with recent perching birds belonging to the “higher” landbird assemblage as defined by Olson (1985). The “higher” landbird assemblage constitutes the avian orders; Strigiformes (owls), Caprimulgiformes (nightjars, oilbirds, frogmouths and allies), Apodiformes (swifts and hummingbirds), Coraciiformes (bee-eaters, kingfishers, rollers, todies, motmots and allies), Piciformes ( jacamars, honeyguides, toucans, woodpeckers and allies), Coliiformes (mousebirds), Bucerotiformes (hoopoes, woodhoopoes and hornbills) and Passeriformes (suboscines and oscines).

Description and comparison

Da.k.7 (Fig.1): The fossil specimen consists of humerus, ulna and radius of both wings (the distal ends of left ulna and radius are not preserved), the left coracoid, the furcula, and the sternum is represented by the sternal carina pointing out of the sediment. A complete, articulated left hind limb intersects the radius and ulna of the left wing. Remains of the ungual hornsheaths of the foot and imprints of feathers are preserved.

Characters of the coracoid and humerus are distinctive and indicate affinity with the Apodiformes. The prominent and ventrally directed ventral tubercle, the rounded outline of the pectoral crest and broad proximal and distal articulating ends in the humerus, are reminiscent of Apus. The skeletal elements of Da.k.7 are, however much less specialized than in the true swift genus Apus, and more similar to the tree swifts, Hemiprocnidae, and some fossil genera of Apodiformes. Hemiprocnidae, differs from the Apodidae, particular in the wing elements, e.g. by a less specialized humerus, with a longer and more slender shaft and the M. extensor metacarpi radialis and M. tensor propatagialis brevis processes positioned on the lower half of the humerus. Also the elongation of the hand elements is less pronounced.

Characters such as the sternal projection of the coracoid, the size and shape of the ventral tubercle, the outline of the bicipital and pectoral crest, the size, depth, position and shape of the brachial fossa, the position and size of the condyles of the humerus together with the size and proportions strongly suggests that Da.k.7 belong to the species Eocypselus vincentiHarrison 1984, known from the Lower Eocene London Clay. The M. tensor propatagialis brevis and the M. extensor metacarpi radialis tubercles cannot be demonstrated in Da.k.7. This might be due to breakage or partly covering by sediment of the dorsal face of the humerus in the distal end, especially if the tubercles are as weakly developed as inEocypselus.

The hind limb of Da.k.7 differ from Apus, in having a very prominent cnemial crest which serves as the attachment for the flexor muscles and a strong anisodactyle foot, with a long backwards pointing hallux, positioned at the same level as the other three digits and the strongly curved unguals with large flexor tubercles are typical of perching birds. The true swifts posses a pamprodactyle foot (all four digits pointing forward), however the tree swifts posses a more typical perching foot in concordance with the fossil specimen.

Da.k.22 (Fig.2): This fossil specimen consists of a complete, articulated right wing skeleton, the left humerus, the sternum and the proximal parts of both coracoids seen in dorsal view.

It exhibits a mixture of characters recognizable in the avian families Alcedinidae, Meropidae, Trogonidae and the Archaeotrogonidae, strongly indicating affinity with the Coraciiformes. The strong medial position of the sternal facet in the coracoid is similar to both the Alcedinidae and the Archaeotrogonidae. In the sternum, the characteristic external spine is similar to the Trogonidae, whereas the thick cranial pilae are more similar to the Alcedinidae and Meropidae. The characters of the humerus, with the rather long and medially positioned humeral head and the almost pointed appearance of the dorsal condyle, show the closest similarity to the Trogonidae and Archaeotrogonidae.

The projection in the proximal part of the radius resembles that of the Meropidae and to some extent Trogonidae. Separate characters of the carpometacarpus can be recognized in all of the Coraciiform families. In digit II, the elaboration of the cranial pila and the biconcavity and the shape of phalanx II are reminiscent of the Alcedinidae.

The concordance in preserved elements in Da.k.22 and the Archaeotrogons are limited, however, in the few similarly preserved elements a very good agreement in characters can be seen (Fig. 4). The most distinct difference between Da.k.22 and the Archaeotrogons is the actual size, where the Archaeotrogons are two to two and a half time larger than Da.k.22. A new species and possibly even a new genus should be erected for a position of Da.k.22 in this family. Any elements associated with the hind limb has in this fossil specimen not been preserved. The proportions of the preserved skeletal elements and the possible allocation with the Coraciiformes, however, suggests that this was a perching bird.

FU100 (Fig.3): This fossil specimen consists of a complete, articulated left wing skeleton, the proximal end of the right humerus, a large part of the shoulder girdle, including the sternum, the right coracoid, the right scapula and fragments of the furcula. The right hind limb is represented by a tibiotarsus and a broken tarsometatarsus, upon which the radius and ulna of the right wing are superimposed. Imprints of feathers, including a flight feathers, are visible.

The combination of a proximal, rectangular pectoral crest, a caudal position of the humeral head and the high and prominent external condyle of the humerus and prominent craniolateral processes, slender and relatively long lateral processes and a high external spine in the sternum indicate a close relationship of FU100 with Bucerotiformes and the Passeriformes.

A ventral torsion of metacarpal I, craniocaudally flattened metacarpal III, a deep carpal fovea and a narrow carpal trochlea, further support the hypothesis of a close relationship between FU100, Passeriformes and Bucerotiformes. However, the character mosaic and in many bones less specialized features of FU100 prevent it from being classified with either of these two orders.

Differences in the elaboration of the sternal rostrum is noticeable between FU100 and species from the latter two orders. The internal spine present in the hoopoe is not demonstrable in FU100, nor is the external spine in FU100 dorsally bifurcated as seen in the majority of passerine birds. However, an ancestral position of FU100 to the Bucerotiformes and Passerifomes would be possible.

The elaboration of the procoracoid process and the external spine in FU100 is less specialized than in both the hoopoe and the passerines. Phylogenetically, a continuing growth in height of the procoracoid process, from the state seen in FU100, could result in a fusion with the acrocoracoid process, as seen in the Bucerotiformes.

In the passerine lineage the coracoid could have developed, from the state seen in FU100, towards a more reversed condition, with a less pronounced procoracoid process, as seen e.g. in the tyrants, to the slender and delicate coracoid of e.g. the robin, which has a nearly completely reduced procoracoid process. Similar series of transformation could be imagined for characters of the sternum, with a deepening of the sternal incisions, a cranial direction of the craniolateral processes and a bifurcation of the external spine in the passerine lineage and in the Bucerotiform lineage the development of an internal spine.

The hind limb of FU100 show, as Da.k.7, features typical of perching birds, with a large cnemial crest and the metatarsal I at the same level as the distal end of the remaining metatarsals. Also a single disarticulated, strongly curved ungual is observable.

Conclusion

The “higher” landbirds have since 1898 (Beddard) been grouped together on the basis of the absence of the ambiens muscle. The interrelationship and evolution of the higher landbirds is, however a controversial subject. Feduccia (1977, 1979) on the basis of the morphology of the stapes in Recent perching birds (defined by Feduccia as birds closely related to and including the passerines) concluded that the passerines were a monophyletic group, that the Coraciiformes should be divided into the Alcediniformes and Coraciiformes, that the Trogonidae (formerly regarded as a separate order) should be included in the Alcediniformes and that the hoopoes, woodhoopoes and hornbills should be separated from but closely allied with the Coraciiformes.

Based on the three specimens from the Fur Formation, litterature on other early Tertiary bird fossils and the model by Feduccia (1979), I have attempted a new model (Fig.4) for the evolution of the “higher” landbirds. The three fossil birds from the Fur Formation, and most other early Tertiary birds, show that in the latest Paleocene/earliest Eocene, the higher landbirds had differentiated, but were still in the beginning of their diversification, being much less divergent, as evidenced by their character mosaic. In many features they combine characters from several of the avian orders within the higher landbird assemblage.

This character mosaic can provide information on the evolution and interrelationship of the taxa in the higher landbird assemblage.

Eocypselus vincenti (including Da.k.7) is the oldest known swift taxon and is regarded as a generalized Apodiform by Harrison (1984), with a basal position in this order, probably as an ancestral form. This early Tertiary swift shows a greater degree of morphological similarity to the Hemiprocnidae (tree swifts) suggesting that the tree swifts might constitute the stemgroup of the true swifts. In some characters (especially of the sterum and coracoid) these early Tertiary swifts show some similarity to FU100 as well, indicating that the Apodiformes is more closely related to the Passeriformes than previously assumed. It has also been suggested that the Trochlii is closer allied with the Passeriformes than the remaining Apodiformes.

The majority of the available characters of Da.k.22 are shared with the Alcedinidae, Trogonidae and Archaeotrogonidae and support the inclusion of the latter two families in the Coraciiformes sensu Olson 1985. Several palaeospecies from the Green River Formation in Wyoming, North America, previously referred to the Piciformes (Feduccia and Martin,1976), have later been referred to the suborder Coracii (Olson, 1985) and were in 1992 by Olson and Houde included in a new order, the Sandcoleiformes. Judging from photographs several characters in the coracoid, ulna and humerus of Sandcoleiform species appear very similar to the trogons and particular the humerus is similar also to the Archaeotrogons and Da.k.22. The Sandcoleiformes also exhibit characters reminiscent of the Coliiformes. Most likely the Sandcoleiformes is very close to the basal Coraciiform stock and have given rise to the Coliiformes and Piciformes. In spite of the inclusion of the Archeotrogons in the Coraciiformes sensu Olson (1985), Mourer-Chauviré noted similarities with the Caprimulgiformes as well, implying that the Apodiformes and Caprimulgiformes are other lineages arising from the Coraciiformes.

The last specimen, FU100, indicates that the hoopoes and passerines are more closely related to each other than has formerly been presumed and that the passerines possibly originated in the northern hemisphere. According to Olson (1985), the earliest record of passerines is from the late Oligocene of France. The Upupidae are recorded from the Eo-Oligocene Phosphorites du Quercy of France (Olson, 1985). Thus, the geological age of FU100 does not exclude the possibility of an ancestral position of FU100.
Little is known of the relationship of the Strigiformes with the remaining higher landbirds. Being the oldest known taxa in the higher landbird assemblage suggests that it also has a basal position in this assemblage, or that it should rather be excluded from the “higher” landbirds. Further investigation and more fossils are needed to elucidate this question.

From the above assumptions it seems that the Coraciiformes is the basal group of “higher” landbirds, from which the remaining “higher” landbird taxa have arisen. Finally the three bird fossils support the hypothesis that the “higher” landbird assemblage is a monophyletic group.

References

Beddard, F.E. 1898. The structure and classification of birds. Longmans, Green, London
Feduccia, A. 1977. A model for the evolution of perching birds. Systematic Zoology26: 19-31.
Feduucia, A.1979. Comments on the phylogeny of perching birds. Proc. Biol. Soc. Wash. 92 (4): 689-696.
Feduccia, A. & Martin, L.D. 1976. The Eocene zygodactyl birds of North America (Aves Piciformes). Smithsonian Contributions to Paleobiology27: 101-110.
Harrison, C.J.O. 1984. A revision of the fossil swifts (Vertebrata, Aves, suborder Apodi), with descriptions of three new genera and two new species. Meded. Werkgr. Tert. Kwart. Geol.21 (4): 157-177.
Houde, P. & S. L. Olson. 1992. A radiation of Coly-like birds from the Eocene of North America (Aves Sandcoleiformes new order). In, K. E. Campbell, Jr. (ed.), Papers in avian palaeontology, honoring Pierce Brodkorb. Natural History Museum of Los Angeles County Sci. Ser. 36:137-160.
Mourer-Chaviré, C. 1980. The archaeotrogonidae of the Eocene and Oligocene Phosphorites du Quercy (France). Nat. Hist. Mus. Los Angeles County Contrib. Sci.330: 17-31.
Olson, S.L. 1985. The fossil record of birds, in (ed.) Farner, D.S., King, Jr., Parker, K.C. Avian biology VIII 80-218.

Fig.3 Specimen FU100, part and counterpart. Abbreviations R.DI-III, right digit 1-3; R.Fe, right femur; R.Sc, right scapula; R.Ti, right tibiotarsus; R.Tm, right tarsometatarsus; U, ulnare. For other abbreviations and signatures see Fig.1 & 2.

Fig.3 Specimen FU100, part and counterpart. Abbreviations R.DI-III, right digit 1-3; R.Fe, right femur; R.Sc, right scapula; R.Ti, right tibiotarsus; R.Tm, right tarsometatarsus; U, ulnare. For other abbreviations and signatures see Fig.1 & 2.

Fig.2 Specimen Da.k.22, part and counterpart. Abbreviations Cos, costae; R, radiale; R.Cm, right carpometacarpus; R.Co, right coracoid; St, sternum R.DIII, right digit 3; R.DII, right digit 2; DII-III, digit 2-3; U, ulnare. For other abbreviations and signatures see Fig.1.

Fig.2 Specimen Da.k.22, part and counterpart. Abbreviations Cos, costae; R, radiale; R.Cm, right carpometacarpus; R.Co, right coracoid; St, sternum R.DIII, right digit 3; R.DII, right digit 2; DII-III, digit 2-3; U, ulnare. For other abbreviations and signatures see Fig.1.

 Fig. 1 Fig.1 Specimen Da.k.7, part and counterpart. Abbreviations Fur, furcula; L.Co, left coracoid; L.Fe, left femur; L.Hu, left humerus; L.Ra, left radius; L.Ti, left tibiotarsus; L.Tm, left tarsometatarsus; L.Ul, left ulna; R.Hu, right humerus; R.Ra, right radius; R.Ul, right ulna; S.Ca, sternal carina; I-IV, digit 1-4. Description and comparison Da.k.7 (Fig.1): The fossil specimen consists of humerus, ulna and radius of both wings (the distal ends of left ulna and radius are not preserved), the left coracoid, the furcula, and the sternum is represented by the sternal carina pointing out of the sediment. A complete, articulated left hind limb intersects the radius and ulna of the left wing. Remains of the ungual hornsheaths of the foot and imprints of feathers are preserved. Characters of the coracoid and humerus are distinctive and indicate affinity with the Apodiformes. The prominent and ventrally directed ventral tubercle, the rounded outline of the pectoral crest and broad proximal and distal articulating ends in the humerus, are reminiscent of Apus. The skeletal elements of Da.k.7 are, however much less specialized than in the true swift genus Apus, and more similar to the tree swifts, Hemiprocnidae, and some fossil genera of Apodiformes. Hemiprocnidae, differs from the Apodidae, particular in the wing elements, e.g. by a less specialized humerus, with a longer and more slender shaft and the M. extensor metacarpi radialis and M. tensor propatagialis brevis processes positioned on the lower half of the humerus. Also the elongation of the hand elements is less pronounced. Characters such as the sternal projection of the coracoid, the size and shape of the ventral tubercle, the outline of the bicipital and pectoral crest, the size, depth, position and shape of the brachial fossa, the position and size of the condyles of the humerus together with the size and proportions strongly suggests that Da.k.7 belong to the species Eocypselus vincenti Harrison 1984, known from the Lower Eocene London Clay. The M. tensor propatagialis brevis and the M. extensor metacarpi radialis tubercles cannot be demonstrated in Da.k.7. This might be due to breakage or partly covering by sediment of the dorsal face of the humerus in the distal end, especially if the tubercles are as weakly developed as in Eocypselus. The hind limb of Da.k.7 differ from Apus, in having a very prominent cnemial crest which serves as the attachment for the flexor muscles and a strong anisodactyle foot, with a long backwards pointing hallux, positioned at the same level as the other three digits and the strongly curved unguals with large flexor tubercles are typical of perching birds. The true swifts posses a pamprodactyle foot (all four digits pointing forward), however the tree swifts posses a more typical perching foot in concordance with the fossil specimen. Fig 2 Fig.2 Specimen Da.k.22, part and counterpart. Abbreviations Cos, costae; R, radiale; R.Cm, right carpometacarpus; R.Co, right coracoid; St, sternum R.DIII, right digit 3; R.DII, right digit 2; DII-III, digit 2-3; U, ulnare. For other abbreviations and signatures see Fig.1.Da.k.22 (Fig.2): This fossil specimen consists of a complete, articulated right wing skeleton, the left humerus, the sternum and the proximal parts of both coracoids seen in dorsal view. It exhibits a mixture of characters recognizable in the avian families Alcedinidae, Meropidae, Trogonidae and the Archaeotrogonidae, strongly indicating affinity with the Coraciiformes. The strong medial position of the sternal facet in the coracoid is similar to both the Alcedinidae and the Archaeotrogonidae. In the sternum, the characteristic external spine is similar to the Trogonidae, whereas the thick cranial pilae are more similar to the Alcedinidae and Meropidae. The characters of the humerus, with the rather long and medially positioned humeral head and the almost pointed appearance of the dorsal condyle, show the closest similarity to the Trogonidae and Archaeotrogonidae. The projection in the proximal part of the radius resembles that of the Meropidae and to some extent Trogonidae. Separate characters of the carpometacarpus can be recognized in all of the Coraciiform families. In digit II, the elaboration of the cranial pila and the biconcavity and the shape of phalanx II are reminiscent of the Alcedinidae. The concordance in preserved elements in Da.k.22 and the Archaeotrogons are limited, however, in the few similarly preserved elements a very good agreement in characters can be seen (Fig. 4). The most distinct difference between Da.k.22 and the Archaeotrogons is the actual size, where the Archaeotrogons are two to two and a half time larger than Da.k.22. A new species and possibly even a new genus should be erected for a position of Da.k.22 in this family. Any elements associated with the hind limb has in this fossil specimen not been preserved. The proportions of the preserved skeletal elements and the possible allocation with the Coraciiformes, however, suggests that this was a perching bird. Fig 3 Fig.3 Specimen FU100, part and counterpart. Abbreviations R.DI-III, right digit 1-3; R.Fe, right femur; R.Sc, right scapula; R.Ti, right tibiotarsus; R.Tm, right tarsometatarsus; U, ulnare. For other abbreviations and signatures see Fig.1 & 2. FU100 (Fig.3): This fossil specimen consists of a complete, articulated left wing skeleton, the proximal end of the right humerus, a large part of the shoulder girdle, including the sternum, the right coracoid, the right scapula and fragments of the furcula. The right hind limb is represented by a tibiotarsus and a broken tarsometatarsus, upon which the radius and ulna of the right wing are superimposed. Imprints of feathers, including a flight feathers, are visible. The combination of a proximal, rectangular pectoral crest, a caudal position of the humeral head and the high and prominent external condyle of the humerus and prominent craniolateral processes, slender and relatively long lateral processes and a high external spine in the sternum indicate a close relationship of FU100 with Bucerotiformes and the Passeriformes. A ventral torsion of metacarpal I, craniocaudally flattened metacarpal III, a deep carpal fovea and a narrow carpal trochlea, further support the hypothesis of a close relationship between FU100, Passeriformes and Bucerotiformes. However, the character mosaic and in many bones less specialized features of FU100 prevent it from being classified with either of these two orders. Differences in the elaboration of the sternal rostrum is noticeable between FU100 and species from the latter two orders. The internal spine present in the hoopoe is not demonstrable in FU100, nor is the external spine in FU100 dorsally bifurcated as seen in the majority of passerine birds. However, an ancestral position of FU100 to the Bucerotiformes and Passerifomes would be possible. The elaboration of the procoracoid process and the external spine in FU100 is less specialized than in both the hoopoe and the passerines. Phylogenetically, a continuing growth in height of the procoracoid process, from the state seen in FU100, could result in a fusion with the acrocoracoid process, as seen in the Bucerotiformes. In the passerine lineage the coracoid could have developed, from the state seen in FU100, towards a more reversed condition, with a less pronounced procoracoid process, as seen e.g. in the tyrants, to the slender and delicate coracoid of e.g. the robin, which has a nearly completely reduced procoracoid process. Similar series of transformation could be imagined for characters of the sternum, with a deepening of the sternal incisions, a cranial direction of the craniolateral processes and a bifurcation of the external spine in the passerine lineage and in the Bucerotiform lineage the development of an internal spine. The hind limb of FU100 show, as Da.k.7, features typical of perching birds, with a large cnemial crest and the metatarsal I at the same level as the distal end of the remaining metatarsals. Also a single disarticulated, strongly curved ungual is observable. Fig. 4 Fig.4 A simplified model showing the supposed evolution and interrelationship of the “higher” landbirds.

Fig.4 A simplified model showing the supposed evolution and interrelationship of the “higher” landbirds.

Fig.1 Specimen Da.k.7, part and counterpart. Abbreviations Fur, furcula; L.Co, left coracoid; L.Fe, left femur; L.Hu, left humerus; L.Ra, left radius; L.Ti, left tibiotarsus; L.Tm, left tarsometatarsus; L.Ul, left ulna; R.Hu, right humerus; R.Ra, right radius; R.Ul, right ulna; S.Ca, sternal carina; I-IV, digit 1-4.

Fig.1 Specimen Da.k.7, part and counterpart. Abbreviations Fur, furcula; L.Co, left coracoid; L.Fe, left femur; L.Hu, left humerus; L.Ra, left radius; L.Ti, left tibiotarsus; L.Tm, left tarsometatarsus; L.Ul, left ulna; R.Hu, right humerus; R.Ra, right radius; R.Ul, right ulna; S.Ca, sternal carina; I-IV, digit 1-4.


Tilmelding til DGF Nyhedsbrev