A Review of the Australian Fossil Storks of the Genus Ciconia (Aves: Ciconiidae), With the Description of a New Species

Only a single species of stork, the Black-necked Stork Ephippiorhynchus (= Xenorhynchus) asiaticus, occurs in Australia today, and is known from several fossil localities from the Early Pliocene. Two species of smaller fossil storks are also known, one previously named and one described here. The former, found in the Darling Downs, southeastern Queensland, was named Xenorhynchus nanus De Vis, 1888. Some later authors suggested that this species should be transferred to the living genus Ciconia; this decision is confirmed here, the name for this species becoming Ciconia nana. The second species of small stork comes from several Late Oligocene and Early Miocene sites at Riversleigh, northwestern Queensland. This taxon is referred to the genus Ciconia and distinguished as a new species, C. louisebolesae. It constitutes the earliest record of the Ciconiidae from Australia. BOLES, WALTER E., 2005. A review of the Australian fossil storks of the genus Ciconia (Aves: Ciconiidae), with the description of a new species. Records of the Australian Museum 57(2): 165–178. Records of the Australian Museum (2005) Vol. 57: 165–178. ISSN 0067-1975 www.amonline.net.au/pdf/publications/1440_complete.pdf The classification of living storks (Ciconiidae) by Kahl (1979) admitted 17 species in six genera in three tribes, whereas that of Hancock et al. (1992) recognized 19 species in six genera in two tribes. The family is represented in Australia by a single living species, the Black-necked Stork, or Jabiru, Ephippiorhynchus (= Xenorhynchus auct.) asiaticus (Latham, 1790). Storks are rather well represented in the world fossil record, although no comprehensive review of them has been attempted. The earliest records come from the Late Eocene of Egypt (Ciconiidae gen. and sp. indet. and Leptoptilos sp. indet.) (Miller et al., 1997). After taxa incorrectly referred to this family were removed (Olson, 1985), the earliest named species became Palaeoephippiorhynchus dietrichi Lambrecht, 1930 (Late Oligocene; Egypt). The identity of the older Eociconia sangequanensis Hou, 1989 (Middle Eocene; China) as a stork needs to be confirmed (Unwin, 1993). Other Tertiary-aged storks are known from North America, Europe and Asia (references in Olson, 1985; Bickart, 1990). Quaternary-aged palaeospecies are known for several extant genera. The fossil record of this family in Australia has not been studied in detail. Much of the Australian fossil stork material is comparable in size and morphology to E. asiaticus. Specimens assigned to this species are known from Pliocene and Pleistocene localities in northeastern and southeastern Queensland and northeastern South Australia (Archer, 1976; Baird, 1991a; Boles & Mackness, 1994; Molnar & Kurz, 1997; Vickers-Rich, 1991). 166 Records of the Australian Museum (2005) Vol. 57 The first stork reported from Australia was described by C.W. De Vis, who named several species (De Vis, 1888, 1892, 1905); however, all but Xenorhynchus nanus are now known to have been misidentified to family (amended family identifications summarized by van Tets & Rich, 1990). Material of a new species of stork from Oligo-Miocene deposits at Riversleigh, northwestern Queensland, was mentioned briefly in the literature with little elaboration (Boles, 1991, 1997; Vickers-Rich, 1991). It is the purpose of this study to review X. nanus and the undescribed Riversleigh stork. Both are here considered to belong to the extant genus Ciconia. This genus has an extensive fossil record. Three of the living species of Ciconia have been recorded from Quaternary deposits (Brodkorb, 1963). Several fossil taxa have been assigned to Ciconia, but many are based on single specimens. A large, but unnamed species of Ciconia from the Late Miocene-Early Pliocene of Arizona is known from numerous skeletal elements (Bickart, 1990), as is another large form, C. maltha L. Miller, 1910, from the Quaternary of North America and Cuba (Miller, 1910; Howard, 1942; Feduccia, 1967). Other palaeospecies include C. stehlini Jánossy, 1992 (Early Pleistocene, Hungary, tarsometatarsi, tibiotarsi, ulna, phalanges), C. gaudryi Lambrecht, 1933 (Late Pliocene of Greece, humerus), C. minor Harrison, 1980 (Late Miocene, Kenya, distal femur) and C. sarmatica Grigorescu & Kessler, 1977 (Late Miocene, Romania, proximal carpometacarpus). Lambrecht (1933) cited records of indeterminate species of Ciconia from the Pleistocene of California and Late Pliocene of France, and Olson & Rasmussen (2001) recorded two indeterminate species from North Carolina, one Middle Miocene in age, the other Early Pliocene. Late Pleistocene or Quaternary reports of this genus include those by Ono (1984; Honshu, Japan; Ciconia sp.), Steadman et al. (1994; northeast Mexico; Ciconia sp. or Mycteria sp.) and Suarez & Olson (2003; Cuba; Ciconia sp.). Species of Ciconia and Mycteria are rather generalized in their morphology compared to the large, long-legged Ephippiorhynchus and Jabiru, the heavy-bodied Leptoptilos and somewhat aberrant Anastomus. Any fossil stork remains not exhibiting characters of these more distinctive genera were frequently allocated to one of the more “typical” ones. The problem of deciding whether a fossil form based on single or fragmentary elements has been correctly assigned to genus is compounded by the heavy reliance by the current taxonomy on behavioural (Kahl, 1972, 1979; Slikas, 1998) or molecular characters (Slikas, 1997). Materials and methods Taxonomic nomenclature follows Kahl (1979). Osteological terminology follows Baumel & Witmer (1993), except that as terms of position and direction anterior is used rather than rostral or cranial and posterior rather than caudal. Most of the measurements follow the methods of Steadman (1980) or van den Driesch (1976), and were made with digital calipers and rounded to the nearest 0.1 mm. Several factors hamper the ease of using the fossil record of storks from elsewhere for evaluating that of Australia. Generic-level taxonomy of the Ciconiidae has changed substantially, with several formerly monotypic genera now merged with others. New palaeogenera were often based on material that exhibited some morphological intermediacy between two nominal genera that have since been synonymised; this is particularly so in the expanded concept of Ciconia. The more inclusive generic concepts result in a broader morphological range across the constituent species, into which the palaeospecies may fit comfortably. Published diagnoses of such fossil forms must be assessed with caution because some of the characters may no longer apply to the genus sensu lato. Another difficulty is that many of the species of fossil storks have been based on isolated fragments, confounding comparison between nominal taxa for which common osteological elements are not known. Moreover, many extant taxa are poorly represented in skeletal collections and of those specimens that do exist, individuals from zoos form a high proportion. In addition to any developmental abnormalities the latter may have, most likewise lack provenance and are frequently unsexed. Osteological diagnosis of Ciconiidae The skeletal elements can be recognized as belonging to this family on the basis of the following suites of characters. Diagnoses are restricted to those portions of the elements represented by the fossils, both here for the family and subsequently for generic level taxa in the respective species accounts. Cranium. The lateral indentations at the orbits are shallow (in dorsal view); fossae glandulae nasalis are absent. The processus postorbitalis is long, and the temporal fossae well defined and rather extensive posteriorly. There is a single small circular fontanelle orbitocranialis situated at the posterior border of septum interorbitalis where it joins the braincase. Quadrate. The anterior and posterior borders of the bladelike processus orbitalis are straight or slightly tapering through most of its length. The process is more or less straight (in posterior view) but not strongly flattened, with the distal end somewhat inflated; it is not incised posteriorly, twisted nor inflected medially or ventrally. The processus oticus is broad and not compressed laterally; the processus mandibularis is deep mediolaterally (in ventral view). The condylus medialis and combined condyli lateralis and caudalis are long and thin, and converge laterally at an acute angle; the sulcus intercondylaris is moderately large, particularly on its medial half. The short, broad projection of the condylus lateralis extends anteriorly along the lateral side, at its anterior end supporting the cotyla quadratojugalis, which is located just above the posteroventral border of the element; the part of the projection between the cotyla and the posterior end of the quadrate comprises about half of its length. Humerus. The element has a pronounced sigmoid curvature, with a particularly marked anterior bend in the distal end (in dorsal view). Proximal end. In anterior view, the long axis and distal border of the caput humeri are oriented dorsodistally-ventroproximally; the caput humeri is moderately short. The sulcus ligamentum transversus and incisura capitis are deep. The tuberculum dorsale is distinct and triangular. The fossa pneumotricipitalis is large. The distal margin of the crista bicipitalis forms a nearly right Boles: fossil birds of the stork genus Ciconia 167 angle with the shaft. The intumescentia humeri is inflated, particularly distally. The crista deltopectoralis is prominent, with its apex more or less level with the distal end of the crista bicipitalis. Distal end. The fossa musculus brachialis is large and deep, particularly ventrodistally, and is angled sharply dorsoproximally-ventrodistally relative to the shaft. The tuberculum supracondylare ventrale is elongate and situated along a prominent ridge. The epicondylus ventralis is strongly produced as a triangular projection. The epicondylus dorsalis and processus flexorius are rudimentary. The ventral side of the distal end is flat (in anterior view) with the processus supracondylare dorsalis prominent, angling moderately to very abruptly to shaft. The fossa olecrani is broad and shallow, and extends proximally from, and dorsoventrally across, the condylus ventralis humeri. Ulna. Proximal end. The proximal end is straight in relation to the shaft, i.e. there is no inflection from the midline of the shaft. The margins of the impressio m. brachialis are pronounced, with the anterior margin the more extensive distally. The tuberculum lig. collateralis ventralis is slightly bulbous but does not overhang the impressio m. brachialis and has a relatively short distal extension along its border. The incisura radialis is more proximodistally oval (narrower, longer) than circular and the impressio m. scapulotricipitalis is small with little distal extension. Tibiotarsus. Proximal end. The proximal end is deeper than wide because the region level with the incisura tibialis between the cristae cnemialis and the facies articularis is elongated. The surface is mostly level, with a small to at most moderate rise towards the cristae cnemialis. The cristae cnemialis are not strongly developed proximally, but are rather broad (in proximal view). They form a more or less 90° angle, and from this junction, the crista cnemialis lateralis is about twice the length of the crista cnemialis cranialis. The crista cnemialis cranialis is situated towards, but not at, the medial edge, with only a slight indentation separating them; it is long distally, angling smoothly into the shaft (in medial view). There is an expanded articular surface at the end of the crista cnemialis lateralis with a flattened anterolateral face, which projects both anteromedially and posterolaterally (in proximal view). Distal end. The shaft is long, thin and straight, with the posterior surface rounded and the anterior surface flattened for most of its length, taken up by a very broad and shallow sulcus extensorius, which deepens for a short extent just proximal of the pons supratendineus. There is a large, prominent papilla for M. tibialis cranialis centred directly proximal to the area intercondylaris and level with the distal border of the pons supratendineus. The pons supratendineus is restricted to the medial half of shaft, with its distal border strongly developed into a ridge. The scar on the lateral face of the shaft is large and proximodistally elongated. The sulcus m. fibularis is moderately deep. The distal end of the element has little mediolateral expansion, and the medial border of the shaft does not flare strongly outwards proximal to the condylus medialis. The condyli lateralis and medialis are more or less parallel and directly distal to the respective borders of the shaft, are longer anteroposteriorly than proximodistally, and have about the same distal extension; the condylus lateralis extends further proximally. The condylus medialis is notched distally. The area intercondylaris is a deep circular pit centred on the midline of shaft, extending between the pons supratendineus and the condylus medialis. The sulcus intercondylaris is deep (in distal view) and the trochlea cartilaginis tibialis is shallow with prominent borders. Tarsometatarsus. Proximal end. The eminentia intercotylaris is narrow, with the lateral border abrupt and the medial one sloping (in dorsal view). In proximal view, the rims of the cotylae are rounded and (in dorsal view) the medial rim of cotyla medialis is blunt or rounded. The hypotarsi comprises two parallel cristae hypotarsi separated by a single large sulcus hypotarsi, which is deep throughout its length; it is centred mediolaterally on the plantar face. There is no small secondary groove within the sulcus hypotarsi. Distal end. The sulcus extensorius occupies the greater part of the length of the anterior surface, making the distal third of the shaft relatively flat, and then angling from the midline of the shaft to the lateral side at the distal end, extending into the foramen vasculare distale but not beyond that into the incisura intertrochlearis lateralis. The fossa metatarsi I is a long proximodistally elongated oval, terminating distally on a ridge extending towards the trochlea metatarsi II. The fossa supratrochlearis plantaris is markedly excavated lateral to this ridge. The trochleae are not inflated proximally nor do they join the shaft abruptly; the shaft bulges laterally just proximal to the trochlea metatarsi IV, meeting it with relatively little demarcation. The trochleae form a shallow but obvious curve (in distal view). The trochleae metatarsi II and IV are more or less equal in length and shorter than trochlea metatarsi III. Genus Ciconia Brisson, 1760 Ciconia Brisson (1760). Ornithologia sive Synopsis Methodica, 1: 48, 361—type species: Ciconia = Ardea ciconia Linnaeus,