A coraciiform-like bird quadrate from the Early Eocene ...

3 downloads 0 Views 403KB Size Report
Apr 7, 2015 - in the Alcediniformes on morphological data (Mayr 2009; Clarke ..... PAN, Krakow), Judy White and Robert Prys-Jones (Natural History Museum ...
CSIRO PUBLISHING

Emu http://dx.doi.org/10.1071/MU14084

A coraciiform-like bird quadrate from the Early Eocene Tingamarra local fauna of Queensland, Australia Andrzej Elzanowski A,C and Walter E. Boles B A

Museum and Institute of Zoology, Polish Academy of Sciences, 64 Wilcza Street, 00-679 Warszawa, Poland. Australian Museum Research Institute, Australian Museum, Sydney, 6 College Street, NSW 2010, Australia. C Corresponding author. Email: [email protected] B

Abstract. A fragmentary quadrate from the earliest Eocene Tingamarra local fauna of Murgon, Queensland, combines characters of the coraciiform birds and their closest relatives, a separate pterygoid facet on the orbital process (a feature unknown in this clade), and similarities to the Upupiformes (Upupidae and Phoeniculidae). The new fossil provides the first Paleogene, and thus the oldest, record of the coraciiform-like birds from the Southern Hemisphere. The Tingamarra bird may well represent one of the named families of Paleogene birds whose quadrates remain largely unknown. A great variation of pneumaticity-related features among the coraciiform birds reveals morphogenetic instability in the development of pneumatic diverticula from the tympanic cavity.

Received 9 September 2014, accepted 11 January 2015, published online 7 April 2015

Introduction Recent advances in avian phylogenetics provide strong evidence for the monophyly of ‘higher land birds’, that is, most arboreal birds exclusive of the hoatzin, cuckoos and turacos (Mayr et al. 2003). Nested among ‘higher land birds’ is a clade that includes Picocoraciae (Mayr 2011), Leptosomus, Trogonidae (Hackett et al. 2008: node D) and, with weaker support, Coliidae (Ericson et al. 2006: node 3). The Picocoraciae include the Piciformes and coraciiform birds (an informal term for non-piciform Picocoraciae). The latter comprises rollers (Coraciidae and Brachypteraciidae, jointly classed as Coracii); kingfishers, motmots and todies (jointly classed as Alcediniformes); and hoopoes and woodhoopoes (jointly Upupiformes), which together with the hornbills are jointly classed as Bucerotes (Clarke et al. 2009; Mayr 2009).The relationships among the families of coraciiform birds, however, remain far from settled. Meropidae have been retrieved as equidistant from Alcediniformes and Coracii on sequence data (Ericson et al. 2006; Hackett et al. 2008), included in the Alcediniformes on morphological data (Mayr 2009; Clarke et al. 2009), and retrieved as a sister group to the Bucerotes on mixed morphological and sequence data (Clarke et al. 2009). The Bucerotes have usually been retrieved as the basal group of the crown Picocoraciae (Ericson et al. 2006; Hackett et al. 2008; Mayr 2009), although Clarke et al. (2009) recovered this clade as a sister group to all Piciformes. Such uncertainties make it rather difficult to determine relationships of a fragmentary fossil that combines characters of several extant taxa. It is now well established that the present-day diversity of Passeriformes (perching birds) was preceded by a considerable Eocene diversity of arboreal non-passerines (Feduccia 1996), including six extinct families of the Picocoraciae, a fossil family of mousebirds (Sandcoleidae) and the Zygodactylidae, which Journal compilation Ó BirdLife Australia 2015

may or may not be related to the Passeriformes (Mayr 2009). To date, all the Eocene arboreal non-passerines are known only from Europe and North America (Clarke et al. 2009; Mayr 2009): the Early Eocene of the London Clay (England) and Green River Formation (Wyoming), the Middle Eocene of Messel and Geiseltal (Germany) and the Late Eocene of the Paris Gypsum and Quercy (France). In view of the identification of the passerines from Tingamarra (Boles 1995, 1997), this raises the possibility of vicariant northern and southern origins of the Picocoraciae and Passeriformes, respectively. Here we describe a single quadrate of a coraciiform-like bird from the earliest Eocene of Australia. The specimen, QMF22781, in the collection of the Queensland Museum, Brisbane, Australia, was recovered from the non-marine, Early Eocene Tingamarra Local Fauna, which is 54.6  0.5 million years old (Boles et al. 1994; Boles 1999). It is thus slightly older than the Fossil Butte Member of the Green River Formation, which yielded hitherto the oldest Picocoraciae (Ksepka and Clarke 2010). The Tingamarra locality produced several postcranial remains of Presbyornis-like ‘graculavid’ (Boles 1999), an anhimid-like quadrate (Elzanowski and Boles 2012) that may or may not represent the same taxon as the ‘graculavid’, and a fragmentary carpometacarpus and tarsometatarsus that have been identified as the oldest passerines (Boles 1995, 1997). No other avian remains have ever been described from this site, but more than 40 specimens remain unstudied. While most of them are fragmentary and many unlikely to be referable further than Aves, they do demonstrate that there is a greater diversity of taxa than has been thus far reported. The quadrate is one the most frequently and best preserved cranial bones in the fossil record of birds and yet has been largely ignored because of the lack of detailed studies of its morphology. Except for occasional and fragmentary to sketchy representations, www.publish.csiro.au/journals/emu

B

Emu

the quadrate morphology has not been studied in detail in any arboreal bird, which made it necessary to survey its variation among the Picocoraciae and their consecutive outgroups. Materials and methods The Tingamarra material was collected by small, incremental trowel loads from the area a person was working. The Tingamarra sediments are illite/smectite clays that expand and contract considerably as they get wet or dry. As a result nearly all the fossils are broken, but their fragments remain in a close association. In conjunction with low fossil densities in the deposit, this enables a confident assignment of fragments to an individual specimen. The specimen QMF22781, in the collections of the Queensland Museum (QMF),was found by Kirsten Crosby in a small block of sediment. The size, preservation, colour and morphology are consistent with the two pieces representing a single specimen. For comparisons, we first sampled all major avian clades to narrow down the search to the clade comprising the Picocoraciae and their consecutive outgroups, and then compared quadrate morphologies in all families of this clade and selected passerine families. Our detailed survey covered 42 species of 35 genera (see Table S1 in Supplementary Material available online only) in 27 families: Strigidae, Coliidae, Trogonidae, Leptosomidae, Upupidae, Phoeniculidae, Bucerotidae, Brachypteraciidae, Coraciidae, Meropidae, Todidae, Momotidae, Alcedinidae, Bucconidae, Galbulidae, Lybiidae, Indicatoridae, Picidae and nine passerine families (two suboscine and seven oscine). To find the most probable affinities of the fossil, we conducted a standard character analysis including outgroup comparisons and a group-specific character analysis, the latter to assess the reliability of the available characters. The anatomical terminology (formal in Fig. 1 and anglicised in the text) follows Elzanowski and Stidham (2010) with a few additions. Description QMF22781 is a right quadrate in two pieces, the dorsal (proximal) piece with the intact head (Fig. 1, A1–A3) and the ventral (distal) piece with the lateral process damaged (missing the quadratojugal area) and the medial process intact (Fig. 1, A4; Fig. 2, A1–A2). The middle part of the bone (most of its body) is missing and only the ventral base of the orbital process is preserved. The dimensions reveal a bird approximately the size of the Green Woodhoopoe (Phoeniculus purpureus), that is, larger than the hoopoes (Upupa) (Table 1). The intercapitular vallecula is distinct but narrow, narrower than the transverse diameter of the otic capitulum (Fig. 1, A1–A3). The otic capitulum is narrower (transversely) than the squamosal capitulum, but slightly larger rostrocaudally. The otic capitulum is discoid with little rostral and no caudal slope. The squamosal capitulum has a distinct and broad rostral slope and caudal slope. Distal to it is a faint lateral supraorbital crest extending towards the broken base of the orbital process. Adjacent to the squamosal capitulum is an adductor tubercle with a rostral expansion (Fig. 1, A2). Extending from the adductor tubercle is a bulky, blunt tympanic crest, which continues on the ventral piece where it bounds the caudomedial foramen (Fig. 1, A3). The ventral (distal) piece bears the caudal crest that extends

A. Elzanowski and W. E. Boles

to the caudal condyle after branching off from the tympanic crest in the missing middle part of the bone. The medial crest extends from the otic capitulum on the dorsal piece and continues on the ventral piece, where it separates a rather flat basiorbital fossa from a caudomedial foramen and further extends to the pterygoid condyle (Fig. 2, A1). The caudomedial foramen is large, elongate, extends far ventrally and shows two transverse trabeculae inside. Ventral to the foramen and medial crest, the medial aspect of the entire mandibular part is occupied by a roughly rectangular ventromedial depression, which extends to the mandibular condyles (Fig. 2, A1). The preserved ventral base of the orbital process bears on its medial side a distinct pterygoid facet (Fig. 2, A1), which is separated from the pterygoid condyle by a deep and wide embayment with a ligamentous groove at the bottom. The pterygoid condyle is prominent and receives the medial crest. The pterygoid condyle is separated from the medial condyle by a wide intercondylar incisure, which is much wider than the diameter of the pterygoid condyle. The medial condyle bears a medial articular facet, which is slightly convex (Fig. 2, A1), and a large, a saddle-shaped lateral articular facet or lateral trochlea (Fig. 2, A2). The caudal condyle bears a slightly concave articular facet, which is surrounded by a prominent, somewhat irregular and wavy flange,which is concave dorsally (Fig. 2, A1: arrow). Most of the lateral process is gone, but both the breakage site and topography indicate that the caudal facet continued smoothly into the lateral facet and that the preserved caudal condyle is part of a single laterocaudal condyle (Fig. 1, A4). Comparisons The quadrate QMF22781 combines similarities to the Upupiformes and characters that the coraciiforms share with other ‘higher land birds’ (Table 2). Common to all the coraciiforms, Leptosomus, trogons and mousebirds is the laterocaudal condyle with a single, continuous articular facet. In the upupiforms and the trogons, the laterocaudal facet is perfectly undifferentiated, whereas in other taxa the two parts are separated by a depression (with the articular surface continuing across it). Specific to the coraciiforms are (1) a crest that starts from the squamosal capitulum caudally rather than laterally and bifurcates into the tympanic crest,which veers off mediad, and (2) the caudal crest (Table 2: Character 5) that descends to the caudal condyle (strictly speaking, the caudal tip of the laterocaudal condyle). The caudal crest (essentially a ridge supporting the caudal condyle) is most prominent in Upupa (Fig. 1, B3), poorly differentiated in Brachypteracias and Merops, and absent (probably apomorphically) in the kingfishers. The quadrate QMF22781 shows four similarities to the Upupiformes (Table 2: Characters 1–4), two of which are unique. (1) A laterocaudal condyle ends with a sharp-edged flange as in Upupa and Phoeniculidae (Fig. 2, B1–C1). A comparable structure has not been found in any of the surveyed bird taxa. In at least some woodpeckers (Picinae), but not in Jynx and other piciform families, the quadrate bears a peculiar bony collar that projects well dorsal from the articular facet, rather than being continuous with the caudal condyle; it thus seems more comparable with submeatic prominences or processes. In conjunction with a highly

Eocene arboreal bird quadrate

Emu

C

Fig. 1. The fossil quadrate QMF22781 compared with the quadrates of extant coraciiforms in dorsal (A1–E1), rostral (A2–E2), caudal (A3–E3) and ventral (A4–E4) views: (A) QMF22781, (B) Upupa epops, (C) Phoeniculus purpureus, (D) Brachypteraciidae gen. sp. NHM S-2010.1.251, (E) Coracias garrulus. Scale bars equal 2 mm. Abbreviations: bo, fossa basiorbitalis; cc, crista caudalis; cl, crista lateralis; cm, crista medialis; cmd, crista medialis dorsal section; cmv, crista medialis ventral section; cs, crista supraorbitalis lateralis; ct, crista tympanica; cv, intumescentia paracondylaris; dc, depressio caudomedialis; dv, depressio ventromedialis; f, adventitious pneumatic foramina; fm, foramen pneumaticum mediale; fp, foramen pneumaticum postcapitulare; lc, condylus mandibularis laterocaudalis; m, condylus mandibularis medialis; mt, lateral trochlea of the medial condyle; o, capitulum oticum; or, processus orbitalis; pt, condylus pterygoideus; po, facies articularis pterygoidea of the orbital process; qj, quadratojugal cotyla; s, capitulum squamosum; t, tuberculum subcapitulare; v, vallecula intercondylaris.

derived position of the woodpeckers (Benz et al. 2006), this structure is most probably not homologous to the flange on the caudal condyle of the upupiform quadrate. (2) A large medial pneumatic foramen extends far ventrally caudal to the medial crest (Fig. 2, A1), whereas in most other Picocoraciae (as well as

most other neovians) the medial foramen is rostral to the medial crest (i.e. in a rostromedial position). In the Phoeniculidae the medial crest appears to be represented by two disconnected parts, dorsal and ventral (Fig. 2, C1) with a depression that may accommodate a pneumatic diverticulum in between. In some

D

Emu

A. Elzanowski and W. E. Boles

Fig. 2. The fossil quadrate QMF22781 compared with the quadrates of extant coraciiformes in medial (A1–E1) and lateral (A2–E2) views: (A) QMF22781, (B) Upupa epops, (C) Phoeniculus purpureus, (D) Brachypteraciidae gesp. nov. NHM S-2010.1.251, (E) Coracias garrulus. The arrows point to the flange-like margins of the laterocaudal condyle. Scale bars equal 2 mm. Abbreviations as for Fig. 1.

specimens the depression is perforated by a small foramen that is caudal to the dorsal leg of the crest. Once the ventral leg of the crest disappears (as may have happened in the Upupidae), the foramen lies caudal to the remaining dorsal part of the crest, which shows the sequence of morphogenetic events leading to an

apparently discontinuous change. (3) A ventromedial depression enclosed between the medial crest (its ventral part that reaches the pterygoid condyle) and the caudal crest (Fig. 2, A1–C1), a character shared at least with Leptosomus. (4) An extensive lateral trochlea of the medial condyle, which also occurs in many

Eocene arboreal bird quadrate

Emu

unrelated non-passerines. Among other coraciiform birds the medial condyle bears a lateral trochlea only in the Alcedinidae, in which the trochlea is large, and Merops, in which the trochlea is diminutive (vestigial or incipient). QMF22781 differs from the extant Upupiformes and all the Bucerotes in having a wide intercondylar incisure (with medial and pterygoid condyles widely separated) and a separate subcapitular tubercle. QMF22781 differs from the Upupiformes alone in lacking any postcapitular foramina and having a strong orbital process in a usual position (approximately mid-height), rather than an extreme ventral one. Except for the postcapitular foramina (which vary between related families such as Coraciidae and Brachypteraciidae), QMF22781 consistently retains plesiomorphic character states (Table 2), which might suggest a stem group representative of the Upupiformes. QMF22781 differs from all extant Picocoraciae and their outgroups, however, in having a separate pterygoid facet at the base of the orbital process (Fig. 2, A1). In the Coliidae, Leptosomus, Trogonidae and most Picocoraciae (including Tricholaema, Indicator and Picidae among Piciformes), the quadrate articulates with the pterygoid via the condyle only, whereas in the Bucconidae, Galbulidae and Ramphastidae the condylar facet continues (to a variable extent) onto the base of the orbital process. Among the ‘higher land birds’, only the Strigidae reveal a tendency to form a disjunct pterygoid facet on the orbital process:

the orbital facet has been found separated from the condylar facet in Strix aluco but continuous in Asio otus. The owls also have the caudomedial pneumatic foramen, but otherwise their quadrate (similar to that of the Coliidae) does not show any specific similarities to QMF22781. Discussion Variation of pneumaticity-related features There is a great variation in the location of pneumatic foramina among families (Fig. 2), as well as some intrafamily variation among genera. The majority of coraciiform birds have the postcapitular foramina, which are lacking in at least some Coraciidae and Bucerotidae. Most of the observed variation is generated by the ossification of the pneumatic canal that descends under the medial crest from the otic part to the basiorbital fossa. The canal may open up at both ends as in Leptosomus, Upupa, Merops and Eurystomus, only dorsally as the medial foramen as in Brachypteraciidae, or only ventrally as the basiorbital foramen as in Coracias. The positioning of the canal under, rather than either in front or behind, the crest accounts for the variable location of the medial foramen: just on the crest in Merops, rostral to the crest in Brachypteraciidae, and caudal in Upupa. In one specimen of Phoeniculus purpureus (SMF5809) the medial foramen (in fact, represented by several small openings) is positioned between two legs of the medial crest, caudal to the dorsal leg and rostral to the ventral leg. No medial foramen is present in another specimen of P. purpureus (Fig. 2, C1) and in Rhinopomastus minor in which a vestigial basiorbital foramen was found. The situation in R. cyanomelas suggests that the pneumatic canal remains open in Phoeniculidae (rather than being divided by a bony bridge as in Upupa) and only small deep medial and basiorbital foramina may mark its ends. Among other ‘higher land birds’, the mousebirds show distinct areas marking a closed medial foramen: in Urocolius macrourus it is rostral, and in Colius striatus caudal, to the medial crest. Ectopic pneumatic foramina in the Phoeniculidae and Coliidae provide additional evidence of morphogenetic instability related to the growth of pneumatic diverticula.

Table 1. Transverse measurements of the quadrates from Tingamarra and extant coraciiform birds Specimen Upupa epops The Tingamarra fossil QMF22781 Phoeniculus purpureus NHM 1904.4.286 Brachypteracias sp. NHM S/2010.1.251 Coracias garrulus

Dorsal widthA Ventral widthB 2.9 3.6 4.0 4.0 4.8

E

3.7 4.3+ (5.0 est.) 5.0 6.0 7.8

A

Maximum, including the subcapitular tubercle. Maximum, between the pterygoid condyle and the rims of the quadratojugal cotyla.

B

Table 2. Variable characters of the quadrate in the Picocoraciae and their closest relatives. Alc, Alcedinidae; Bce, Bucerotidae; Bra, Brachypteraciidae; Buc, Bucconidae; Col, Coliidae; Cor, Coraciidae; FO, QMF22781; Lep, Leptosomus; Mer, Meropidae; Mom, Momotidae; Pho, Phoeniculidae; Tod, Todidae; Tro, Trogonidae; Upu, Upupidae. Parentheses indicate poorly (vestigial or incipient) or indistinctly developed features. Slashes indicate intermediate character states No.

Character

Col

Tro

Lep

FO

Upu

Pho

Bce

Bra

Cor

Mer

Mom

Tod

Alc

Buc

1 2

Caudal condyle: 0 labroid, 1 flange Medial foramen: 0 absent,1 rostromedial, 2 caudomedial Extensive, uniform ventromedial depression: 0 absent, 1 present (Fig. 2 A1, B1, C1) Medial condyle trochlea: 0 absent, 1 present Caudal crest: 0 absent, 1 present Pterygoid and median condyles: 0 widely or 1 narrowly separated, 2 superimposed Orbital process: 0 normal, 1 ventral Postcapitular foramen: 0 absent, 1 present Subcapitular tubercle: 1 separate, 0 merged with squamosal capitulum

0 1/2

0 0

0 (1)A

1 2

1 2

1 (2)

0 1

0 1

0 1

0 1

0 0

0 0

0 0

0 1

0

0

1

1

1

1

0

0

0

0

0

0

0

0

0 0 2

0 0 1

0 0 1

1 1 0

1 1 1/2

1 1 2

0 0 2

0 (1) 0

0 1 0

(1) 0/1 0

0 1 0

0 1 0

1 0 2

0 1 0

0/1 1 1

0 1 0

0 0 0/1

0 0 1

1 1 0

1 1 0

0 0 0

0 1 1

0 0 0

0 1 0

0 1 0

0 1 0

1 1 0

0 0 0

3 4 5 6 7 8 9 A

An irregular meshwork of foramina, mostly rostral to the medial crest (based on a single specimen).

F

Emu

In the majority of birds, the location of pneumatic foramina is conserved within families or orders, and this holds true even for the position (rostral or caudal) of the medial foramen relative to the medial crest. A striking exception is the Anhimidae, which revealed a prodigious instability in the location of pneumatic foramina (Elzanowski and Stidham 2010). The Picocoraciae and related taxa of ‘higher land birds’ provide another example of morphogenetic instability of pneumatic diverticula that invade the quadrate, which inevitably weakens the strong phylogenetic signal that the position of pneumatic foramina carries in other avian clades. Affinities of the Tingamarra ‘higher land birds’ A unique combination of four detailed similarities (Table 2) provides prima facie evidence for the upupiform relationships of the bird from which the Tingamarra quadrate came. One, the flange of the laterocaudal condyle, is unique and each of the remaining three occurs in one extant family of ‘higher land birds’ (Table 2). The only obvious dissimilarity and presumed autapomorphy is the disjunct pterygoid facet on the orbital process. This feature evolved independently at least five times (and probably more) in other birds (among the Gallifomes, Columbidae, Charadriiformes, Gruiformes, and, to a degree, the Strigiformes), which makes its independent appearance among the Picocoraciae not improbable. The Upupiformes have a very distinctive quadrate, with the orbital process in a ventral position and pointing rostrally (Fig. 2). The ventromedial depression, however, is shared with Leptosomus (Table 2), which seems to support the prevailing placement of the Upupiformes as the earliest branching of the crown Picocoraciae (Ericson et al. 2006; Hackett et al. 2008; Mayr 2009) rather than the sister group to all piciforms (Clarke et al. 2009). A basal position is in agreement with an early Eocene occurrence. Among the Eocene coraciiform birds, the only identified close relatives of the Upupiformes are the Messelirrisoridae, which Mayr (2009) nested within the Bucerotes as the basal Upupiformes (i.e. as a sister group to the Phoeniculidae–Upupidae clade). The Messelirrisoridae share with the extant families a blade-like retroarticular process of the mandible (Mayr 1998, 2009), suggesting a similar conformation of the caudal mandible including the quadratomandibular articulation. Their quadrate remains undescribed, although it seems fairly well preserved in lateral view in at least one specimen identified as Messelirrisoridae indet. (Mayr 1998, plate 6, SMF-ME 600). This shows a deep embayment between the otic part and the orbital process, which is compatible with the upupiforms, but the orbital process is oriented as in the majority of birds, that is, it points much more dorsally than in the hoopoes and woodhoopoes. Thus far known from Europe and North America, the Messelirrisoridae were very small, hummingbird-sized birds, whereas the Tingamarra quadrate indicates the body size of a thrush. While this size difference does not preclude a closer relationship, it makes the membership of the Tingamarra bird in this family improbable. The Upupiformes are almost invariably recovered together, with the Bucerotidae as a sister group, but there is little evidence to claim closer relationship of the Tingamarra bird to the entire clade of Bucerotes. The bucerotid quadrate is highly derived in some respects (especially in its otic part) and otherwise better

A. Elzanowski and W. E. Boles

comparable to the other coraciiform birds. One of its pecularities is a facet for a secondary mandibular articulation on the paracondylar intumescence (caudal to the pterygoid condyle). The intumescence is well developed in Coraciidae and Brachypteraciidae (Fig. 2, D1, E1), but absent in the Upupiformes, which have this area covered by the ventromedial depression (Fig. 2, B1, C1) (Table 2, Character 3). Among the two bone fragments described by Boles (1995, 1997) as the oldest passerine fossils, the carpometacarpus is from a bird the size of a small finch and the distal tibiotarsus (with the distal width of 4.1 mm) is from a bird the size of the Blackbird (Turdus merula). Thus the tibiotarsus and the quadrate QMF22781 come from birds of comparable size, which raises the question of whether they could be conspecific. While the quadrate is clearly non-passeriform, both the quadrate and tibiotarsus could possibly represent the Zygodactylidae, which show striking passerine similarities in the postcranial skeleton and have been proposed to be a sister group to the passerines (Mayr 2009; DeBee 2012). The Zygodactylidae are known in the Eocene of Europe and North America from numerous specimens and some of them, such as Primozygodactylus major, reach the size of a blackbird (Mayr 1998, plate 16). However, the zygodactylid quadrate as exposed in lateral view (DeBee 2012, figs 1.4 and 1.11) is unlike that of extant upupiforms in having the orbital process in a dorsal position with only a shallow dorsal embayment. Yet another possibility is that the passerine-like carpometacarpus and the distal, much less diagnostic fragment of tibiotarsus represent different avian clades, and the latter may possibly belong together with the quadrate irrespective of what group of birds is represented by the small carpometacarpus. The fossil record of the Picocoraciae starts in the Early Eocene of Europe and North America and most of the Eocene fossils represent stem taxa of the named major clades (Mayr 1998, 2009; Clarke et al. 2009). Since stem taxa usually retain more plesiomorphic features (that permit their identification as such) and the Upupiformes may be basal to all the Picocoraciae, including the stem lineages of the Piciformes, Coracii and Alcediniformes (Mayr 2009), then some upupiform characters might have been retained by several stem taxa of Picocoraciae, rather than representing synapomorphies that mark a closer relationships to the Upupiformes. Unfortunately, the most diagnostic details of the quadrate are not exposed in fossil specimens. As partly exposed in caudal view, the quadrate of Primobucco mcgrewi (Ksepka and Clarke 2010, fig. 3) shows the general coraciiform pattern and lacks postcapitular foramina as in the Coraciidae, which seems consistent with the placement of Primobucco among the stem lineages of the Coracii (Clarke et al. 2009). Postcapitular foramina are also lacking in the Bucerotidae, which, in conjunction with the upupiform similarities of QMF22781, could be interpreted as a symplesiomorphy of Bucerotes. The Tingamarra quadrate QMF22781 is the first Southern Hemisphere record of an Eocene coraciiform-like ‘higher land bird’ and supports the view that these birds may have dominated the Eocene arboreal bird communities all over the world. The nearly coeval, Early Eocene record of the Tingamarra bird from Australia and Primobucconidae from the Northern Hemisphere testify to a major radiation of arboreal birds in the Palaeocene or

Eocene arboreal bird quadrate

Emu

earlier, especially in view of the Eocene separation of Australia from the northern continents. Acknowledgements This research was supported by Poland’s National Science Center (NCN) grant 2013/11/B/NZ8/04376 to AE, whereas the collection of QMF22781 was made possible by the Australian Research Council grant to Michael Archer and Suzanne Hand (University of New South Wales, Sydney). We are grateful to Anna Gillespie and Henk Godthelp (Vertebrate Palaeontology Laboratory, University of New South Wales, Sydney) for preparation of the specimen, and to Gerald Mayr (Forschungsinstitut Senckenberg, Frankfurt/M) and two anonymous reviewers for excellent criticisms of the manuscript. AE is indebted to Zbigniew Bochenski (Instytut Systematyki i Ewolucji Zwierz˛a t PAN, Krakow), Judy White and Robert Prys-Jones (Natural History Museum, Tring), and Gerald Mayr for liberal access to skeleton collections.

References Benz, B. W., Robbins, M. B., and Peterson, A. T. (2006). Evolutionary history of woodpeckers and allies (Aves: Picidae): placing key taxa on the phylogenetic tree. Molecular Phylogenetics and Evolution 40, 389–399. doi:10.1016/j.ympev.2006.02.021 Boles, W. E. (1995). The world’s oldest songbird. Nature 374, 21–22. doi:10.1038/374021b0 Boles, W. E. (1997). Fossil songbirds (Passeriformes) from the Early Eocene of Australia. Emu 97, 43–50. doi:10.1071/MU97004 Boles, W. E. (1999). Early Eocene shorebirds (Aves: Charadriiformes) from the Tingamarra Local Fauna, Murgon, Queensland, Australia. Records of the Western Australian Museum 229–238. Boles, W. E., Godthelp, H., Hand, S., and Archer, M. (1994). Earliest Australian non-marine bird assemblage from the Early Eocene Tingamurra Local Fauna, Murgon, southeastern Queensland. Alcheringa 18, 70. doi:10.1080/03115518.1994.9638764 Clarke, J. A., Ksepka, D. T., Smith, N. A., and Norell, M. A. (2009). Combined phylogenetic analysis of a new North American fossil species confirms widespread Eocene distribution for stem rollers. Zoological Journal of the Linnean Society 157, 586–611. doi:10.1111/j.1096-3642.2009.00550.x

G

DeBee, A. M. (2012). A taxonomic and anatomic assessment of the extinct Zygodactylidae (Aves) from the Green River Formation of Wyoming and placement of Zygodactylidae within Aves. M.Sc. Thesis, The University of Texas at Austin. Elzanowski, A., and Boles, W. E. (2012). Australia’s oldest anseriform fossil: a quadrate from the Early Eocene Tingamarra fauna. Palaeontology 55, 903–911. doi:10.1111/j.1475-4983.2012.01166.x Elzanowski, A., and Stidham, T. A. (2010). Morphology of the quadrate in the Eocene anseriform Presbyornis and extant galloanserine birds. Journal of Morphology 271, 305–323. Ericson, P. G. P., Andersson, C. L., Britton, T., Elzanowski, A., Johansson, U. S., Källersjö, M., Ohlson, J. I., Parsons, T. J., Zuccon, D., and Mayr, G. (2006). Diversification of Neoaves: integration of molecular sequence data and fossils. Biology Letters 2, 543–547. Feduccia, A. (1996). ‘The Origin and Evolution of Birds.’ (Yale University Press: New Haven.) Hackett, S. J., Kimball, R. T., Reddy, S., Bowie, R. C. K., Braun, E. L., Braun, M. J., Chojnowski, J. L., Cox, A. W., Han, K., Harshman, J., Huddleston, C. J., Marks, B. D., Miglia, K. J., Moore, W. S., Sheldon, F. H., Steadman, D. W., Witt, C. C., and Yuri, T. (2008). A phylogenomic study of birds reveals their evolutionary history. Science 320, 1763–1768. doi:10.1126/ science.1157704 Ksepka, D. T., and Clarke, J. A. (2010). Primobucco mcgrewi (Aves: Coracii) from the Eocene Green River formation: new anatomical data from the earliest constrained record of stem rollers. Journal of Vertebrate Paleontology 30, 215–225. doi:10.1080/02724630903412414 Mayr, G. (1998). “Coraciiforme” und “piciforme” Kleinvögel aus dem MittelEozän der Grube Messel (Hessen, Deutschland). Courier Forschungsinstitut Senckenberg 205, 1–101. Mayr, G. (2009). ‘Palaeogene Fossil Birds.’ (Springer: Berlin–Heidelberg.) Mayr, G. (2011). Metaves, Mirandornithes, Strisores, and other novelties – a critical review of the higher-level phylogeny of neornithine birds. Journal of Zoological Systematics and Evolutionary Research 49, 58–76. doi:10.1111/j.1439-0469.2010.00586.x Mayr, G., Manegold, A., and Johansson, U. S. (2003). Monophyletic groups within ‘higher land birds’ – comparison of morphological and molecular data. Journal of Zoological Systematics and Evolutionary Research 41, 233–248. doi:10.1046/j.1439-0469.2003.00230.x

www.publish.csiro.au/journals/emu