Journal of Ornithology

, Volume 149, Issue 2, pp 223–228

Earliest fossil record of the Certhioidea (treecreepers and allies) from the early Miocene of Germany


    • Forschungsinstitut SenckenbergSection of Ornithology
Original Article

DOI: 10.1007/s10336-007-0263-9

Cite this article as:
Manegold, A. J Ornithol (2008) 149: 223. doi:10.1007/s10336-007-0263-9


A complete tarsometatarsus of a passerine bird from the early Miocene (MN 3) of Petersbuch (Bavaria, Germany) is identified as an extinct representative of the climbing Certhioidea, i.e., a clade comprising treecreepers (Certhiidae), nuthatches and wallcreepers (Sittidae). The fossil specimen represents the so far earliest evidence of a representative of the Certhioidea and is described as †Certhiops rummeli gen. et sp. nov. Similarities to other climbing passerines are discussed.




Several current DNA-sequence based phylogenies support a clade Certhioidea, comprising wrens (Troglodytidae) and gnatcatchers (Polioptilidae), as well as nuthatches (Sittidae) and treecreepers (Certhiidae) (Barker et al. 2002; 2004; Cracraft et al. 2004; Beresford et al. 2005; Alström et al. 2006). In fact, Certhioidea is one of few well-defined clades within Passerida (Cracraft et al. 2004). Several studies indicate that Sittidae and Certhiidae are sister groups (e.g. Beresford et al. 2005, but see Harshman 2007), descending from a last common ancestor capable of climbing on vertical surfaces such as tree trunks. The Wallcreeper (Tichodroma muraria), which clings on vertical rock faces, was not included in these studies, but its affiliation to the Sittidae is corroborated by similarities in behaviour and a shared derived tail pattern (Heinroth and Heinroth 1924–26; Vaurie 1957; Löhrl 1964, 1988; Harrap and Quinn 1996) as well as by DNA-DNA-hybridization studies (Sibley and Ahlquist 1990; Sheldon and Gill 1996). The relationships of the Spotted Creeper (Salpornis spilonotus) of Africa and India are uncertain, but it is currently included within Certhiidae (Sibley and Ahlquist 1990; Dickinson 2003). In the following, the informal term “climbing Certhioidea” is used for the Sittidae/Certhiidae clade.

The fossil record of climbing Certhioidea is both sparse and as yet almost restricted to European and North American localities of late Pliocene and Pleistocene age (3–1.5 mya; Jánossy 1974; Brodkorb 1978; Parmalee and Klippel 1982; Mlíkovský 2002). Ballmann (1973: 58) parenthetically mentioned the presence of Sittidae in deposits of La Grive-Saint-Alban (France) which date from the middle Miocene (MN 7 + 8; 13.5−11.1 mya; Steininger 1999), but a description of these fossils has never been published (see also Olson 1985: p. 140).

New collections from early Miocene (MN 3) karstic fissure fillings near Petersbuch (Bavaria, Germany) yielded an isolated but well-preserved right tarsometatarsus which is assignable to the climbing Certhioidea by means of several shared derived characters. It is the earliest record of a representative of this group of birds. Because the characters in question seem to be correlated to locomotory habits, comparisons were also made with corresponding bones of climbing and non-climbing passerines.


Terminology of anatomical structures follows Baumel and Witmer (1993) and Vanden Berge and Zweers (1993) if not stated otherwise. Generally, English versions of anatomical terms are used when it makes the text more readable. For the same reason abbreviations of certain flexor muscles (Table 1) are used in the text.
Table 1

Abbreviations for the deep flexor muscles of the hind limb and their tendons


Deep flexor muscles


flexor digitorum longus


flexor hallucis longus


flexor perforatus digiti II


flexor perforans et perforatus digiti II


flexor perforatus digiti III


flexor perforans et perforatus digiti III


flexor perforatus digiti IV

Nomenclature of modern species follows Dickinson (2003), but “ Furnariidae” is set in quotation marks to indicate that it does not denote a monophyletic group, according to recent DNA-sequence analyses (Irestedt et al. 2002; Chesser 2004; Fjeldså et al. 2005).

The fossil specimen described here is deposited in Naturmuseum Augsburg, Germany (NMA). Comparisons were made with the following specimens of extant taxa in the collections of the Forschungsinstitut Senckenberg, Frankfurt am Main, Germany (SMF) and the Natural History Museum, London/Tring, Great Britain (NHM).

suboscines: “Furnariidae”: Pseudocolaptes boissoneautii NHM 1891,7.20.341. Dendrocolaptidae: Dendrocincla fuliginosa NHM S/1974.11.36, Sittasomus griseicapillus SMF 235, Xiphorhynchusfuscus SMF 4145. oscines: Climacteridae: Cormobates leucophaea NHM S/1964.1.97. Paridae: Parus major SMF 4475, SMF 5583. Troglodytidae: Thryothorus ludovicianus SMF 6487, Troglodytes troglodytes SMF 5558, SMF 6590. Sittidae: Sitta europaea SMF 5319, SMF 5577, Sitta frontalis SMF 3717, SMF 5607, Tichodroma muraria SMF 6633. Certhiidae: Certhiafamiliaris, Certhia brachydactyla SMF 4363. Parulidae: Mniotilta varia SMF 225, Seiurus aurocapillus NHM S/1979.17.8.

A specimen of the Spotted Creeper (Salpornis spilonotus) was not available and comparisons are therefore limited to the few descriptions of Orenstein (1977).

The hypotarsal pattern of representatives of all but 5 of the 46 extant “families” of Passeriformes recognized by Sibley and Monroe (1990) were investigated previously (Manegold et al. 2004). These data were considered for the present comparisons.


Passeriformes (Linnaeus, 1758)

Eupasseres Ericson et al., 2002

Oscines sensu Ericson et al. (2002)

Certhioidea Cracraft et al., 2004

Certhiops gen. nov.

Type species:Certhiops rummeli sp. nov.

Diagnosis: Certhiops differs from all other passerines studied in having an indistinct, blurred tuberositas m. tibialis cranialis which is shifted towards the medial side of the bone (Fig. 1a). The presence of a large, stout eminentia intercotylaris is likewise remarkable (Fig. 1a).
Fig. 1

Right tarsometatarsus in dorsal view of: aCerthiops rummeli gen. et sp. nov.; b European Nuthatch (Sitta europaea); c Wallcreeper (Tichodroma muraria); d Short-toed Treecreeper (Certhia brachydactyla); e Olivaceous Woodcreeper (Sittasomus griseicapillus). Abbreviations: arc Arcus extensorius; clh crista lateralis hypotarsi; dme distal medial extension; fvd foramen vasculare distale; fvp foramina vascularia proximalia; iitr incisura intertrochlearis; pme proximal medial extension; tf trochlear furrow; ttc Tuberositas m. tibialis cranialis; II–IV trochlea metatarsi II–IV. Scale bar = 1 mm

Etymology: From Certhia and the Latinized Greek suffix-ops meaning appearance, used here to indicate the resemblance to the extant taxa of climbing Certhioidea. The genus name is male.

Certhiops rummeli sp. nov.

Holotype: NMA 2007/51/2021; right tarsometatarsus, with partly broken trochlea metatarsi IV and slightly damaged rim of cotyla lateralis (Figs. 1a, 2a, d, 3a).
Fig. 2

ad Proximal right tarsometatarsus in lateral view of aCerthiops rummeli gen. et sp. nov.; b Wallcreeper (Tichodroma muraria); c Olivaceous Woodcreeper (Sittasomus griseicapillus); d Great Tit (Parus major). ej Hypotarsal canal pattern of eCerthiops rummeli gen. et sp. nov.; f Velvet-fronted Nuthatch (Sitta frontalis); g Wallcreeper (Tichodroma muraria); h Short-toed Treecreeper (Certhia brachydactyla); i White-throated Treecreeper (Cormobates leucophaea); j Great Tit (Parus major). Arrow indicates proximal protruding crista hypotarsi lateralis. Abbreviations: clh crista lateralis hypotarsi; cpl crista plantaris lateralis (lateral crista); emi eminentia intercotylaris; tb tendinal bridge; for foramen; fp 2–4, fpp 2–3 hypotarsal canals for the tendons of the deep flexor muscles explained in Table 1. Not to scale
Fig. 3

Right tarsometatarsus in distal view of aCerthiops rummeli gen. et sp. nov.; b European Nuthatch (Sitta europaea); c Wallcreeper (Tichodroma muraria); d Short-toed Treecreeper (Certhia brachydactyla); e Olivaceous Woodcreeper (Sittasomus griseicapillus). Abbreviations: tf trochlear furrow; II–IV trochlea metatarsi II–IV. Not to scale

Diagnosis: Same as for genus.

Type locality and horizon: “Petersbuch 62”, a karstic fissure filling close to the village Petersbuch north of Eichstätt (Bavaria, Germany). According to the collector, M. Rummel (personal communication), this site dates from the early Miocene and is assigned to the Mammal Neogene Zone 3 (MN 3). It is contemporaneous with the fossil site Wintershof West (Bavaria, Germany) and 20.5–18 my old (Mein 1999; Steininger 1999).

Dimensions of the holotype: Total length is 19.8 mm. The tarsometatarsus of †Certhiops rummeli is slightly smaller, stockier, and more robust than the corresponding bone of an extant European Nuthatch (Sitta europaea) (Fig. 1a–b). In the Wallcreeper (Tichodroma muraria) as well as the Short-toed Treecreeper (Certhia brachydactyla), the tarsometatarsus is more slender and elongated (Fig. 1c–d).

Etymology: The new species is named in honour to Dr. Michael Rummel who collected the specimen and made it generously available for description.

Description and comparison:Certhiops rummeli can be assigned to the Passeriformes and Eupasseres, respectively, because it shares with them a prominent crista plantaris lateralis (Boles 2005; Fig. 2a) and the characteristic hypotarsal canal pattern (Manegold et al. 2004). It shares the following derived characters with extant representatives of the climbing Certhioidea. These characters are regarded as apomorphies for a clade comprising †Certhiops rummeli, Certhiidae, and Sittidae (only characters 5–7 could be checked for Salpornis based on illustrations and descriptions of Orenstein 1977).
  1. 1.

    The crista lateralis hypotarsi reaches farther proximally than the crista medialis hypotarsi and it almost protrudes the eminentia intercotylaris (Figs. 1, 2a–b). In other passerines, the crista lateralis hypotarsi is not protruding (Fig. 2d).

  2. 2.

    The hypotarsal canals for the deep flexor tendons are characteristically arranged (see Table 1 for abbreviations): The diameter of the canal for the fpp3-tendon is markedly smaller than the diameter of the single canal for the fp3- and fp4-tendons. This size difference is obvious in the fossil even though the fpp3-canal is confluent with the canals for the fp2- and fpp2-tendons. The fp3,fp4-canal lays also more plantarly with respect to the fpp3-canal (Fig. 2e–f, h–i). In other passerines, the fpp3-canal is on the same level with the fp3,fp4-canal and both are of similar diameter (Fig. 2g).

  3. 3.

    In the fossil specimen as well as in Tichodroma and Certhia, the fp3,fp4-canal is incompletely ossified and forms a shallow, plantarly open furrow (Fig. 2e, h–i). In Sitta, however, the canal is completely ossified (Fig. 2f) as is the case in the majority of Eupasseres (Manegold et al. 2004).

  4. 4.

    Shaft with triangular medial extension just distal to the fossa metatarsi I which is comparatively smooth in Tichodroma muraria (Fig. 1a–c). [A similar structure is also present in wrens (Troglodytidae), tits (Paridae), and honeyeaters (Meliphagidae; Boles 2005)].

  5. 5.

    The trochleae metatarsorum II–IV are proximodistally short, and separated from each other by broad incisurae intertrochleares (Figs. 1a–d, 3a–d).

  6. 6.

    The trochlea metatarsi III bears a deep trochlear furrow (Figs. 1a–d, 3a–d). This furrow is especially deep in Certhia (Figs. 1d, 3d; Moreno 1986).

  7. 7.

    The trochlea metatarsi III is large and protrudes dorsally as well as plantarly. In both †Certhiops and Tichodroma, the lateral rim of this trochlea is smaller than the medial one, while both rims are of almost equal size in extant nuthatches and treecreepers (Figs. 1a–d, 3a–d).

Certhiops rummeli might be a stem lineage representative of either Certhiidae or Sittidae or both, but no synapomorphies were found supporting any of these hypotheses. On the other hand, the extant representatives of the climbing Certhioidea are distinguished from the fossil specimen by several characters. In Certhia and Salpornis, the trochlea metatarsi II is strongly tilted in dorso-medial direction (Fig. 3d; Orenstein 1977: Fig. 21), while it is more or less parallel to the trochlea metatarsi III in other passeriforms (this character might be a new synapomorphy for Certhia and Salpornis). The foramina vascularia proximalia of Certhia are staggered and shifted towards the lateral margin of the arcus extensorius. Certhia is also peculiar in having an additional medial extension of the shaft which is situated approximately on the level of the proximal margin of the fossa metatarsi I (Fig. 1d).

The European and the Velvet-fronted Nuthatches (Sitta europaea and S. frontalis) differ from †Certhiops rummeli as well as from Tichodroma, Certhia and Salpornis in having a more slender trochlea metatarsi II. This trochlea is especially large and bulky in Tichodroma. Finally, the foramen vasculare distale is situated further proximally in Tichodroma muraria than in any other representative of Certhioidea.

Characters 1, 3, 5 and 6 are also present in Neotropical “Furnariidae” and Dendrocolaptidae, e.g. Streaked Tuftedcheek (Pseudocolaptes boissonneautii), Lesser (Xiphorhynchusfuscus), Olivaceous (Sittasomus griseicapillus), and Plain-brown Woodcreepers (Dendrocincla fuliginosa), as well as White-throated Treecreeper (Cormobates leucophaea), one of the Australo-Papuan Climacteridae (Figs. 1e, 2c, j; 3e). These similarities are, however, best explained as due to convergence, because several additional characters distinguish these taxa from climbing Certhioidea including †Certhiops. In particular, “Furnariidae”, Dendrocolaptidae, and Climacteridae possess
  • comparatively small trochleae metatarsorum II–IV of about the same size which exceed the dorsal surface of the tarsometatarsus but very slightly (Fig. 1e), and

  • a deeply furrowed trochlea metatarsi II (Fig. 3e);

  • a medial flange at the shaft of the tarsometatarsus is missing.

In addition, “Furnariidae” and Dendrocolaptidae have an ossified tendinal bridge connecting the crista plantaris lateralis with the hypotarsus (Fig. 2c) and bordering a small foramen (“peroneal foramen” of Orenstein 1977) which is passed in Dendrocolaptidae by a tendon of M. fibularis longus (Raikow 1993: Fig. 14). A similar bridge and foramen were described for honeyeaters (Meliphagidae) and Australian babblers (Pomatostomus spp.; Boles 2005), and are also present in several other Eupasseres (Orenstein 1977, and own observations). Dendrocolaptidae are further distinguished by the sudden broadening of the distalmost part of their tarsometatarsus (Fig. 1e).


The classification and naming of fossil taxa based on isolated bones has always to be performed with caution. However, the distinctiveness of the specimen described in the present study justifies the introduction of a new species name and its identification as an extinct representative of climbing Certhioidea. Its resemblances to Dendrocolaptidae, “Furnariidae”, and Climacteridae are regarded to be due to convergence. The last mentioned taxa are today restricted to the Neotropical and Australo-Papuan region, respectively, and it seems rather unlikely that they could have a fossil record in the early Miocene of Middle Europe. Nevertheless, Cheneval (2000) identified †Homalopus picoides Milne-Edwards, 1869–1871 from the middle Miocene of France as a fossil dendrocolaptid. The specimen, a distal fragment of a tarsometatarsus, exhibits indeed a deep trochlear furrow of trochlea metatarsi III, but it seems to be too fragmentary to be sure that it belongs to the Passeriformes at all (Mayr and Manegold 2006).

Taking into account that several phylogenetically unrelated groups of passerines capable of climbing on vertical surfaces agree in several derived characters, it seems reasonable to assume that these similarities are in direct relationship to their locomotory properties, even if detailed and sophisticated analyses are missing. Actually, several authors already interpreted the characteristic shape of the tarsometatarsal trochleae of Certhiidae, Climacteridae, Dendrocolaptidae, “Furnariidae”, and Sittidae, but also of woodpeckers (Picidae), and woodhoopoes (Phoeniculidae), as adaptations for climbing on vertical surfaces (Steinbacher 1935; Richardson 1942; Ballmann 1969: p. 52; Feduccia 1973; Orenstein 1977; Moreno 1991; Raikow 1993). Therefore, it is obvious to suppose that †Certhiops rummeli was also able to climb on tree trunks.

The age of the fossil specimen is also remarkable. Not only does it imply that the two major lineages of Certhioidea—one leading to wrens and gnatwrens, the other one to treecreepers, nuthatches, and relatives—already diverged 20 mya. It also represents the so far earliest fossil passerine assignable to an extant subordinated clade within Oscines in the Northern Hemisphere. Known from contemporaneous German fossil deposits are remains of broadbills (Eurylaimidae, Suboscines) (Ballmann 1969). In the Southern Hemisphere, evidence for fossil representatives of lyrebirds (Menura) and logrunners (Orthonyx) from Australian fossil sites is now supposed to date from early Miocene (23.3 mya; Boles 1993, 1995). Remains of several species of honeyeaters (Meliphagidae) stem from the middle to early late Miocene of Australia (Boles 2005) and suggest an already high diversity of passerine birds at that time.


Ältester Fossilnachweis für Certhioidea (Baumläufer und Verwandte) aus dem frühen Miozän Deutschlands

Ein vollständig erhaltener Tarsometatarsus eines Sperlingsvogels aus dem frühen Miozän (MN 3) von Petersbuch bei Eichstätt zeigt hochgradige Übereinstimmungen in abgeleiteten Merkmalen zu heute lebenden Baumläufern (Certhiidae), Kleibern (Sittidae) und ihren nächsten Verwandten, so dass eine Zuordnung des Fossils zu den kletternden Vertretern der Certhioidea möglich ist. Gleichzeitig ist es der älteste Nachweis für diese Vogelgruppe. Das Fossil wird als †Certhiops rummeli gen. et sp. nov. beschrieben, und seine Ähnlichkeiten zu anderen kletternden Sperlingsvögeln werden diskutiert.


For providing access to and loans of fossil and extant specimens I thank Michael Rummel (Naturmuseum Augsburg), Gerald Mayr (Forschungsinstitut Senckenberg) as well as Mark Adams, Joe Cooper, and Robert Prŷs-Jones (Natural History Museum London/Tring). I thank Walter Bock, Antoine Louchart, and Gerald Mayr for their comments on an earlier draft of the manuscript. This study was supported by a German Research Foundation (DFG) grant MA 2328/3-1.

Copyright information

© Dt. Ornithologen-Gesellschaft e.V. 2007