Descriptions and phylogenetic relationships of two new genera and four new species of Oligo-Miocene waterfowl (Aves: Anatidae) from Australia

Descriptions and phylogenetic relationships of two new genera and four new species of Oligo-Miocene waterfowl (Aves: Anatidae) from Australia TREVOR H. WORTHY* Department of Earth and Environmental Sciences, The University of Adelaide, Adelaide, SA 5005, Australia Received 17 September 2007; accepted for publication 30 May 2008 The Tertiary anatid fossils (Aves: Anatidae) from Oligocene and Miocene deposits in Australia are described. Most fossils derive from the Late Oligocene – Early Miocene (26–24 Mya) Etadunna and Namba Formations, respectively, in the Lake Eyre and Lake Frome Basins of SouthAustralia. The local faunas from these two formations contain the same suite of anatid species. Two new genera, the oxyurine Pinpanetta, with three new species (Pi. tedfordi, 18 specimens; Pi. vickersrichae, 15 specimens; Pi. fromensis, 20 specimens), and the tadornine Australotadorna, for a large new species known from eight specimens, are established. Three anatid bones from the Waite Formation (c. 8 Mya) at Alcoota, Northern Territory reveal the presence of a tadornine that is neither Australotadorna nor an extant Tadorna species, and an indeterminate duck about the size of Malacorhynchus. Phylogenetic analyses establish Pinpanetta as a basal member of an oxyurine (stiff-tailed duck) radiation. Oxyurines are found to include the Recent Stictonetta and Malacorhynchus as basal members, along with the fossil taxa Mionetta,Manuherikia, and Dunstanetta, and the traditionally included Recent Oxyura,Biziura,Thalassornis, and Nomonyx. © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454. ADDITIONAL KEYWORDS:Australotadorna – Oxyurinae – phylogeny – Pinpanetta.zoj_483 411..454 INTRODUCTION The Tertiary fossil record for Australian avifaunas is relatively rich with faunas derived from Lower Oligocene and younger deposits (Rich et al., 1991; Vickers-Rich, 1991). The most significant sources of Tertiary fossil birds include the Oligocene to Pliocene sequences of inland South Australia (Stirton, Tedford & Miller, 1961; Woodburne et al., 1994) particularly those near lakes Palankarinna, Pinpa, Ngapakaldi, and Yanda (Rich & van Tets, 1982; Pledge & Tedford, 1990; Rich et al. (1991; Vickers-Rich, 1991). The richest avifaunas derive from the Oligo-Miocene (26–24 Mya) fluvial–lacustrine sequences in the Lake Eyre Basin from the Etadunna and Namba Formations (Woodburne et al., 1994; Alley, 1998). Faunal correlations indicate that the abundant vertebrate faunas in the Carl Creek Limestone (Archer et al., 1997, 1999, 2006) from deposits at Riversleigh in north-western Queensland are in part of similar Oligo-Miocene age but extend through to the Middle Miocene. In the numerous sites around Riversleigh, mammals dominate, and birds although relatively few, have high diversity (Boles, 1993a–c, 1995a, b, 1997a–d, 1998, 1999, 2001, 2005a–c). After these Late Oligocene – Early Miocene faunas, a substantial gap exists in the Australian record. Only two significant faunas for birds are known from the Middle–Late Miocene. The Middle Miocene 10–12 Mya Camfield Beds at Bullock Creek in the Northern Territory, although containing an abundance of aquatic and stream bank species (Murray & Megirian, 1992; Archer et al., 1999; Megirian, Murray & Schwartz, 2004) has, relative to the Etadunna and Namba formations, a much smaller and less diverse avifauna with no waterfowl. Slightly younger, and also containing an important avifauna, is the Waite Formation (c. 8 Mya) at Alcoota (Murray & Megirian, *E-mail: trevor[email protected] Zoological Journal of the Linnean Society, 2009, 156, 411–454. With 7 figures © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 411 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 1992). In both Bullock Creek and Alcoota, mihirung birds (Aves: Dromornithidae) dominate the avifauna and other birds are rare (Rich, 1979; Rich & van Tets, 1982; Rich et al., 1991; Vickers-Rich, 1991; Murray & Vickers-Rich, 2004). Other prominent components of all these Tertiary avifaunas are flamingos and palaelodids together in Phoenicopteriformes (Miller, 1963; Rich et al., 1987; Baird & Vickers-Rich, 1998) and pelicans, Pelecaniformes (Rich & van Tets, 1981). Since Vickers-Rich’s (1991) review, there have been notable reports of ratites, Casuariidae (Boles, 1992) and megapodes, Megapodiidae (Boles & Ivison, 1999) from the South Australian deposits. Waterfowl (Aves: Anseriformes) fossils are a significant component of various Australian Tertiary fossil faunas (e.g. Tedford et al., 1977; Rich & van Tets, 1982; Pledge, 1984; Pledge & Tedford, 1990; Tedford & Wells, 1990; Rich et al., 1991; VickersRich, 1991; Tedford, Wells & Barghoorn, 1992; Boles & Mackness, 1994; Boles, 1997c). To date, the only waterfowl faunas to have been studied are some of Plio-Pleistocene age. All nine presumed extinct PlioPleistocene anatid species named by De Vis from Australian-wide deposits (De Vis, 1888, 1889, 1905) were referred to living species by Olson (1977). The extensive faunas of Pliocene age that mainly derive from the Tirari Formation exposed in sediments at Lake Palankarinna (Mampuwordu Member, Palankarinna Fauna), and Lake Kanunka (Kanunka Fauna), and along the Warburton River (Toolapinna Fauna) in South Australia contain a few anseriforms (Tedford, Williams & Wells, 1986; Tedford & Wells, 1990; Tedford et al., 1992), but most are extant species (pers. observ.) and are not covered here. Fossil anseriforms are common in the Late Oligocene – Miocene lacustrine deposits of the Lake Eyre Basin in central Australia (Rich et al., 1991; Vickers-Rich, 1991), but their identity and phylogenetic relationships have been neglected to date. Such data will be especially relevant to understanding the evolutionary origins of the modern fauna. Taxa such as the endemic and monotypic Anseranas (Anseranatidae), and among anatids, Dendrocygna, the anserines Cereopsis and Cnemiornis,Biziura, Oxyura, and Stictonetta, are primitive members of the Recent fauna (Frith, 1964; Madsen, McHugh & de Kloet, 1988; Livezey, 1986, 1989, 1996a, 1997a, b; Sibley & Ahlquist, 1990; Christidis & Boles, 1994; Sraml et al., 1996; Worthy et al., 1997; Sorenson et al., 1999; Donne-Goussé, Laudet & Hänni, 2002; Dickinson, 2003; Callaghan & Harshman, 2005; McCracken & Sorenson, 2005). Similarly, Malacorhynchus is also considered relatively basal (Brush, 1976; Frith, 1977; Olson & Feduccia, 1980) with Fullagar, in Kear (2005: 442) suggesting this taxon is a part of the old endemic component of Australia’s avifauna with no close relatives. These taxa are all monotypic or of low diversity and combined with their basal position within Anseriformes could be presumed to have a long history in the region. However, to date there is no Australian fossil record for any of them. Recently several fossils from the northern hemisphere have been referred to Anseranatidae. Olson (1999) described Anatalavis oxfordi from the Eocene London Clay of Europe, and referred it and Anatalavis rex (Shufeldt, 1915) from the Palaeocene Hornerstone Formation in New Jersey, USA, to Anseranatidae, although this attracted some debate (Dyke, 2001; Mayr, 2005). More recently, Mourer-Chauviré, Berthet & Hugueney (2004) described Anserpica kiliani from the Late Oligocene Créchy Quarry in France and referred it to Anseranatidae. None of the other Australian endemic taxa listed above have a fossil record elsewhere in the world. There is considerable overlap of Recent taxa between New Zealand and Australia with congeneric species in Biziura,Malacorhynchus, Oxyura,Aythya,Anas, and Tadorna (Turbott, 1990; Worthy, 2005), and with Cnemiornis considered the sister taxon to Cereopsis (Worthy et al., 1997). From the extensive St Bathans Fauna of Early Miocene age, 19–16 Mya, in New Zealand (Worthy et al., 2007), six anatids are known, but none are referrable to Recent genera. A phylogenetic analysis of the New Zealand fossils Manuherikia and Dunstanetta (Worthy & Lee, 2008) suggests that these taxa are basal oxyurines, which together with Biziura,Thalassornis,Oxyura,Stictonetta, Malacorhynchus, and the European Early Miocene Mionetta, form either a clade or a grade of ‘oxyurines’ relatively basal within Anatidae. It is therefore the aim of the present work to describe the species represented by waterfowl fossils from the Australian Oligocene – Miocene and to determine their phylogenetic relationships. This information is fundamental to an understanding of the evolution of the modern waterfowl fauna of Australasia and is potentially significant in the context of the global evolution of waterfowl. The occurrence of waterfowl in both the Etadunna and the Namba Formations will allow assessment of taxonomic similarity of the contained local faunas and so test previous faunal correlations and hypotheses of relative ages of these formations. Similarly, an understanding of the Oligo-Miocene Australian waterfowl faunas will reveal their relationships to those from the extensive Early Miocene New Zealand St Bathans Fauna. Lastly, this knowledge can contribute to ecological reconstructions of the biomes from which the fossil faunas derived. 412 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 METHODS ABBREVIATIONS Institutions: AM, Australian Museum, Sydney, Australia; AMNH, Fossil Amphibian, Reptile, and Bird Collections, Division of Paleontology, American Museum of Natural History, New York, United States of America; ANWC, Australian National Wildlife Collection, CSIRO, Canberra, Australia; BMNH, The Natural History Museum, London, United Kingdom; CM, Canterbury Museum, Christchurch, New Zealand; NMNZ, Museum of New Zealand Te Papa Tongarewa, Wellington (formerly National Museum of New Zealand, Dominion Museum, and Colonial Museum), New Zealand; MV, Museum Victoria, Melbourne, Victoria, Australia; NTM, Museums & Art Galleries of the Northern Territory, Darwin, Australia; QM, Queensland Museum, Brisbane, Queensland, Australia; SAM, South Australian Museum, Adelaide, South Australia, Australia; UCMP, University of California, Museum of Paleontology, Berkeley, California, USA; USNM, Division of Birds, Smithsonian Institution, Washington D.C., USA. Anatomical nomenclature: Names for specific bone landmarks follow Baumel & Witmer (1993) with English translations, or follow Howard (1929), thereafter. Some common terms are abbreviated as follows: L is left and R is right elements. L or R elements are sometimes prefixed with either ‘p’, ‘s’, or ‘d’, to indicate that the either proximal, shaft, or distal parts, respectively, of the element is represented; tuber, tuberculum; LF, local fauna; Mya, million years ago, indet, indeterminate. Measurements: TL, total length; PW, proximal width; SW, shaft width; DW, distal width. Anatomical abbreviations: a, acrocoracoid; ap, alular process; bc, bicipital crest; bf, brachial fossa; cdf, crus dorsale fossa (median crest); cf, cranial fossa; cg, capital groove; cr, capital ridge; csc, scapular cotyla of coracoid; dc, dorsal condyle; dcf, dorsal lobe clavicle facet; dcr, deltoid crest; dls, distal ligamental scar; dpf, dorsal pneumotricipital fossa; ds, distal synostosis; dsf, dorsal sternal facet; dt, dorsal tubercle; ec, ectocnemial crest; ecp, ectepicondylar prominence; ect, dorsal ridge carpal trochlea; ep, extensor process; fa, flexor attachment; fal, facet for anterior articular ligament; fcc, fovea carpalis caudalis (cuneiform fossa); fp, flexor process; hh, humeral head; hf, humeral facet of coracoid; icf, ventral carpal (infratrochlear) fossa; ict, ventral ridge carpal trochlea; ls, scars for attachment external ligament; M., musculus; ma, medial angle; mc, medial condyle; MII, os metacarpale majus (major metacarpal); MIII, os metacarpale minus (minor metacarpal); ol, osseus lamina; p, procoracoid; pc, procnemial crest; pls, proximal ligamental scar; pp, pisiform process; ps, proximal synostosis; sa, supraspinatus attachment scar; spm, superficial pronator muscle attachment; ss, supracoracoidal sulcus; tb, tendinal bridge; tg, scapulotricipital groove; TII, TIII, and TIV, trochlea for metatarsi II, III and IV respectively; vc, ventral condyle; vcf, ventral lobe clavicle facet; vpf, ventral pneumotricipital fossa; vt, ventral tubercle. COMPARATIVE MATERIAL The comparative material examined is listed in the Supporting Information. IDENTIFICATION OF FOSSIL MATERIAL Fossil collections in the following institutions were searched for anseriforms of Oligocene and Miocene age from Australia: Australian Museum (AM), Museum Victoria (MV), Queensland Museum (QM), and South Australian Museum (SAM). A large collection of fossil avian material housed at AM, including material from the University of California Museum of Paleontology (UCMP) and the American Museum of Natural History (AMNH), which had been assembled for other purposes, was also examined. All specimens at these institutions provisionally identified as anseriforms were borrowed for this study. Those whose identities were established as anseriforms form the basis of this report. PHYLOGENETIC ANALYSES The phylogenetic analyses were aimed principally at determining the relationships of the three extinct Australian anatids sufficiently well known for phylogenetic evaluation. These analyses built on those described in Worthy & Lee (2008) and used similar methodologies. In the present analyses 61 terminal taxa were included. The New Zealand Miocene fossil taxa Manuherikia lacustrina,Dunstanetta johnstoneorum and the European Oligo-Miocene Mionetta blanchardi were retained in these analyses as they are near contemporaneous with the Australian fossils. Gallus gallus and Anhima cornuta were defined as outgroups, as Galliformes is the sister group to Anseriformes (e.g. Sibley & Ahlquist, 1990; Ericson, 1997) and Anhimidae is the basal anseriform family (Livezey, 1986, 1997a; Clarke et al., 2005; Livezey & Zusi, 2007). An additional set of analyses were made including Presbyornis, the Palaeogene sister taxon to Anatidae (Ericson, 1997; Livezey, 1997c), to determine whether this dataset obtained a similar sistergroup relationship, and to see whether its addition affected the relationships within Anatidae. AUSTRALIAN TERTIARY ANATIDS 413 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 As found by Worthy & Lee (2008), preliminary analyses of this dataset grouped all divers in a single clade, whereas molecular and immunological evidence indicates this result is an artefact of homoplasy (see Worthy & Lee, 2008: fig. 8). This is the result of morphological convergence towards specialized diving that affects multiple parts of the skeleton (e.g. McCracken et al., 1999), for example, loss of pneumaticity has evolved independently in oxyurines (sensu Worthy & Lee, 2008), and in mergines and aythyines (O’Connor, 2004). Hence backbone constraints that reflected well-supported genetic clades were employed. A major effect of these constraints is to enforce the separation of certain diving taxa in line with genetic evidence: Oxyura is a basal form and not closely related to mergines and aythyines. Other divers, such as Biziura,Thalassornis, and the fossil taxa will then associate with one of these separated diving groups, based on the phylogenetic data. Essentially three clades, each with basal polytomies, were enforced above Anseranas. The most inclusive clade comprised Stictonetta, Cygnus +Branta +Anser brachyrhynchus,Dendrocygna,Oxyura, and the ‘middle clade’. The middle clade had a basal a polytomy of Tadorna ferriginea, T. tadornoides,T. tadorna,Alopochen and the ‘upper clade’. The upper clade included aythyines (Aythya affinis,A. novaeseelandiae,A. australis), mergines (Lophodytes,Somateria), Anas superciliosa,andAix. The polytomies in this constraint means that only very robust genetic clades were enforced: the branching order of most taxa within each level was still largely free to vary depending on the phylogenetic signal; and the remaining 36 ingroup taxa (not included in these backbone constraints) were completely free to associate as dictated by the phylogenetic character data. The dataset employed by Worthy & Lee (2008) was extended by seven humeral and ten pelvic characters for a total dataset of 150 characters (145 osteological, five integumental). Characters were based primarily on those in Livezey (1986, 1996a) with some from Howard (1929), Raikow (1971), Worthy et al. (1997); see Worthy & Lee (2008: supporting information, appendix 1) and new characters in Appendix 1 herein. All characters were scored following original examination of specimens (Appendix 2), except data for Presbyornis, which was scored from data in Howard (1955), Ericson (1997, 1999, 2000), and Livezey (1997c). Missing data were identified as either: (1) inapplicable characters (coded as ‘-’) which could not be objectively scored in a particular taxon because of extensive divergence obscuring homology, or (2) unknown characters (coded as ?) that were not preserved in the (often incomplete) specimens examined, but which potentially could be determined. Although distinguished in our matrix (Appendix 2), PAUP and MrBayes treat both types of missing data in the same fashion. A total of 35 multistate characters varied as morphoclines and could potentially be coded as ordered (numbers 2 11 12 19 24 29 31 37 48 52 53 55 59 60 66 72 77 80 83 85 89 93 97 105 107 108 116 118 120 121 124 129 138 142 146). Preliminary analyses were performed with these characters as unordered or as ordered. The results were very similar (see Worthy & Lee, 2008), and so those reported here focus on the ‘ordered’ analyses. The phylogenetic analyses used PAUP* 4.0b10 (Swofford, 2000). Parsimony analyses used heuristic searches with tree-bisection-reconnection branch swapping, and 1000 random addition replicates per search. Trees were rooted with outgroups forming a polytomy at the base of the tree. When calculating tree lengths, multistate taxa were treated as polymorphisms rather than as ambiguous. Bootstrapping used heuristic searches and the same options. Analyses were performed without and with molecular backbone constraints, but only those using constraints are reported here for the reasons outlined in Worthy & Lee (2008). Bayesian analyses The program MrBayes 3.1.2 (Ronquist & Huelsenbeck, 2003) was used to determine posterior probabilities for clades in the tree. The analyses were performed with the same characters and ordering assumptions as above; however, the outgroup was restricted to Gallus gallus as multiple outgroups were not allowed. MrBayes does not have an explicit command to effect a backbone constraint, so this was achieved by the addition of a second character block. In this character block, taxa in the molecular backbone constraint were coded with dummy characters that enforced the six clades assumed in the backbone constraints. Ten binary characters supporting each of the assumed clades were sufficient to generate posterior probabilities of 1.0 for assumed backbone relationships. The remaining taxa were coded as missing data for all characters in the ‘backbone’ character block. The analysis was then performed with the morphological (1–150) and ‘backbone’ (151–210) characters partitioned into two datasets, with topology linked but all other parameters unlinked across the two character blocks. The signal in the ‘backbone’ character block was sufficiently strong to enforce the assumed backbone constraints, whereas the positions of all the other taxa (and resolution of the polytomies in the backbone constraint tree) were generated by the morphological data. The MrBayes file is in the online Supporting Information. 414 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 The following priors were used. Characters were assumed to have rate variability distributed according to gamma parameter (rates =gamma) with flat prior distribution (0–200). In the morphological dataset, only variable characters were assumed to be have been included (coding =variable), whereas in the backbone constraint data partition, only parsimony-informative characters were assumed to be included (coding =informative); these assumptions were consistent with the structure of the data matrix. Topology was linked across the two character blocks, but the rate variability parameter, rate matrix, and branch lengths were unlinked across the two partitions [unlink shape =(all) Statefreq =(all) unlink brlens =(all)]. Unlinking parameters between character blocks means the analysis (and retrieved signal) of the morphological data is not affected by the structure of the backbone matrix. After preliminary analyses with varying generation times, burnin, and sampling frequency, the following Markov chain Monte Carlo parameters were used. Two independent analyses were run simultaneously to check for adequacy of convergence, each for 5 000 000 generations, sampled every 1000 generations. To improve exploration of tree topology space, the heating parameter was set to 0.22, six chains (one cold and five incrementally-heated) per analysis were used, and branch swapping was set at three times the default (nswaps =3). The time to convergence for topology, stationarity, and all parameters was checked using TRACER v. 1.3 (Rambaut & Drummond, 2004); the first 1000 sampled trees were discarded as burnin. A standard ‘all-compat’ consensus tree, where all clades are shown regardless of posterior probabilities values, was produced by combining the post-burnin samples from both runs. Posterior probability values for a node are the percentage of sampled generations that have that node. FOSSIL SITES The specimens described below mainly derive from localities in two main depositional basins of Lake Eyre Basin. The western sub-basin was formerly identified as the Lake Eyre Sub-basin and the eastern one Tarkarooloo Sub-basin by Vickers-Rich (1991) and Woodburne et al. (1994), but they now are named the Tirari Sub-basin and Callabonna Sub-basin, respectively (Tedford et al., 1986; Krieg et al., 1990; Callen, Alley & Greenwood, 1995; Alley, 1998). Sites in the Tirari Sub-basin that have revealed anseriform fossils are at Lake Palankarinna 28°46–47′S, 138° 24′E and Ngapakaldi 28° 17′S, 138° 17′E; those from the Callabonna Sub-basin are at Lake Pinpa (=Pine Lake) 31° 8′S, 140° 13′E, Lake Namba, 31° 12′S, 140° 14′E, Lake Yanda, 31° 0.05′S, 140° 18.5′E, and those associated with Billeroo Creek, 31° 6′S, 140° 14′E (Rich et al., 1991; Vickers-Rich, 1991). These sites often have identification codes or name equivalents, for example, the Lake Palankarinna site ‘Tedford Locality Site 2’, has the University of California Museum of Paleontology site code V-5375. These codes are included whenever known so that locality data is as fully cross referenced to specimen data as possible. From Lake Pinpa, many specimens derive from Sites A and C. Site A was located at the south-west end of Lake Pinpa, CURNAMONA (prov. ed.) sheet grid coordinates 318146 and Site C is an area on the western shore of the lake that ‘extends north from the E-W cross lake track to about the location of the base of my measured section of 1971, e.g. grid coord. 317148, CURNAMONA....sheet.’ (R. Tedford, pers. comm., 30 August 2006). Several of the following specimens derive from expeditions organized jointly between various institutions identified by the prefix in collection codes as follows: QMAM, joint Queensland Museum – American Museum expeditions of 1971 (QMAM 47, 65, 66 and 74) and 1973 (numbers >100) led by Dick Tedford; VSQ, a joint Victoria, South Australia and Queensland museums expedition. The main vertebrate-bearing beds in the Lake Eyre Basin are in the Etadunna Formation in the Tirari Sub-basin, (Woodburne et al., 1994), and in the Namba Formation of the Callabonna Sub-basin (Callen & Tedford, 1976; Tedford et al., 1977). We follow Woodburne et al. (1994) in accepting a Late Oligocene 26–24 Mya age for the Etadunna Formation, and for the nomenclature of local faunas and fossil mammal zones. They correlated Zone A, the oldest mammal zone in the Etadunna Formation, or the Minkina LF, with the Pinpa LF of the Namba Formation, thereby suggesting a Late Oligocene age for this fauna. Secondly, Woodburne et al. (1994) correlated the superjacent Zone B, containing the Ditjimanka LF at Lake Palankarinna, with the Ericmas Fauna in the upper part of the Namba Formation. The youngest fauna in the Etadunna sequence is the Ngama LF from Mammalon Hill, Lake Palankarinna (Pledge, 1984). Woodburne et al.’s (1994) revised chronology based on magnetostratigraphy and biochronology has led to a considerably older age being attributed to these units than the previous Middle Miocene age that was based on palynological evidence (Callen & Tedford, 1976; Woodburne et al., 1985), and is now widely followed (e.g. Archer et al., 1997; Alley, 1998). Reassessment of pollen samples from the Namba Formation indicates a Late Oligocene – Early Miocene age for a rainforest flora from the base of the unit and a dry sclerophyll forest similar to Late Miocene – Pliocene floras elsewhere in its upper section (Martin, 1990). AUSTRALIAN TERTIARY ANATIDS 415 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Fossil anatids are known from only two sites in the Riversleigh deposits (19° 00′S, 138° 39′E): details of site localities are available from the University of New South Wales or Queensland Museum on request. A single bone derives from Sticky Beak Site, part of the System A sequence, considered to be of Late Oligocene – Early Miocene age as contained faunas have taxa in common with the Late Oligocene faunas from Lake Palankarinna (Archer et al., 1997). The other is from Ringtail Site, in the Ray’s Amphitheatre Sequence on Gag Plateau, attributed to the System C assemblage. System C overlies System A deposits on Gag Plateau and so are younger, and were considered Middle Miocene in age, possibly 16–14 Mya and slightly older than the Bullock Creek LF (Archer et al., 1997). The Riversleigh ‘System’ nomenclature has been replaced by Faunal Zones, with System A equating to Faunal Zone A, etc, (Travouillon et al., 2006). Map references are given as recorded by the collectors from the following sheets: CURNAMONA (prov. ed.) sheet is CURNAMONA, SH 54-14, provisional ed. 1964, Series R502, 1 : 250,000, 1973 reprint. RESULTS SYSTEMATIC PALAEONTOLOGY ORDER ANSERIFORMES WAGLER, 1831 FAMILY ANATIDAE LEACH, 1820: SWANS,GEESE,DUCKS SUBFAMILY OXYURINAE PHILLIPS, 1926: STIFF-TAILED DUCKS The following fossil taxa are referred to Anatidae rather than Anseranatidae and Anhimidae by the following unique combination of humeral characters: 1, a wide caput humeri (head) with its distal margin caudally roughly at right angles to the shaft; 2, a relatively broad fossa pneumotricipitalis ventralis housed in an inflated crista bicipitalis; 3, a distinct dorsal pneumotricipital fossa. The following unique combination of humeral characters indicates that they are related to a clade termed Oxyurinae (Oxyura, Nomonyx,Thalassornis,Biziura,Malacorhynchus,Stictonetta, Mionetta,Dendrochen,Manuherikia, and Dunstanetta) by Worthy & Lee (2008): (1) a distinct capital shaft ridge that is directed towards the tuber. dorsale; (2) elevated dorsal tubercle; (3) an elongate crista deltopectoralis that is concave dorsally; (4) and a closed or nonpneumatic ventral pneumotricipital fossa. These characters are consistent with other traits. The fossils are much smaller than any anserine and derived relative to Anseranas, Anhima, and anserines with an elongate epicondylus ventralis (entepicondyle), extending distally nearly to the same extent as the condylus ventralis (not markedly shorter). The new taxa are more derived than both anserines and dendrocygnines as (1) the capital shaft ridge is directed towards the dorsal tubercle rather than more ventrally, and has the same caudal elevation (is level with) the crus dorsale fossae (median crest), rather than being significantly more elevated or caudad, of the median crest; and (2) the humeri have a closed or nonpneumatic ventral pneumotricipital fossa. They are excluded from Tadorninae by having a closed ventral pneumatic fossa. They are excluded from Anatinae, including all anatines, aythyines, and mergines, by retention of the plesiomorphic characters of a distinct capital shaft ridge, and an elevated dorsal tubercle. GENUS PINPANETTA GEN.NOV. Type species: Pinpanetta tedfordi sp. nov. Diagnosis: Oxyurines in which humeri have the following unique combination of characters: incisura capitis (capital groove) forming either a very shallow or no notch in proximal profile; ventral pneumotricipital fossa closed or nonpneumatic, not extending under median crest; dorsal pneumotricipital fossa narrower than ventral one; dorsal tubercle about as wide as long, not elongate; attachment of M. scapulohumeralis cranialis (supraspinatus) an elongate ridge, extending distally to point level with junction of bicipital crest and shaft; tuber. supracondylare ventrale (facet for anterior ligament) buttressed cranially; ectepicondylar prominence distinct; and attachment of pronator brevis is an isolated pit on ventral facies. Etymology: After Lake Pinpa and the Pinpa Local Fauna from which many specimens derive, and for ‘netta’, duck in Greek. Description and comparison: Within Anatidae, lack of a well-developed notch at the ventral end of the capital groove, as shown by Pinpanetta, is found only in dendrocygnines, anserines, and tadornines. Thalassornis,Biziura,Oxyura,Nomonyx,Stictonetta, Malacorhynchus,Mionetta, and Manuherikia are all more derived with their proximal profile interrupted by a distinct notch at the ventral end of the capital groove. The relatively narrow dorsal pneumotricipital fossa in Pinpanetta is shared with dendrocygnines, anserines, and tadornines. Within oxyurines, Mionetta,Manuherikia,Thalassornis,Biziura,Stictonetta, and Malacorhynchus also retain the narrow dorsal pneumotricipital fossa, but in Oxyura and Nomonyx the fossa is relatively wider. Pinpanetta shares the derived condition of a closed or nonpneu416 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 matic ventral pneumotricipital fossa with all oxyurines except Stictonetta and Nomonyx, in which taxa the fossa is pneumatic. In Pinpanetta, the attachment scar for m. latissimus dorsi posterioris commences proximal to and links to the end of the deltoid crest before extending down the shaft, as in Anseranas, some Dendrocygna species, and Cereopsis. This character state is therefore probably the primitive condition, and is shared with Mionetta,Thalassornis, Oxyura,andNomonyx. In the assumed more derived states there is no connection of the scar with the deltoid crest, whether the scar begins anterior of the end of the crest, e.g. Biziura and Manuherikia, or commences level with its end e.g. Stictonetta, Malacorhynchus, tadornines, and most anatines. The tuber. ventrale slightly overhangs the ventral pneumotricipital fossa in Pinpanetta, rather than being directed proximally. This is a derived condition accentuated in most diving anatids, e.g. Biziura and Oxyura, and so differs from Stictonetta and Mionetta where the ventral tubercle is directed proximally. The dorsal tubercle is about as wide as long, as seen in Anseranas,Dendrocygna, anserines, and some oxyurines. The derived state (an elongate tubercle) is seen in other oxyurines (some Oxyura species, Nomonyx, Stictonetta,Manuherikia, and Dunstanetta), most tadornines and all anatines. An elongated attachment of the supraspinatus is derived relative to a short attachment characteristic of Anseranas,Dendrocygna species, and anserines. In Pinpanetta, the attachment is as elongate as in Manuherikia, Thalassornis, Oxyura, Biziura and Malacorhynchus, but it is much shorter in Stictonetta,Nomonyx, and Mionetta. Distally, in Pinpanetta, the facet for the anterior ligament is cranially buttressed, as in Mionetta, Manuherikia,Dunstanetta,Malacorhynchus,Stictonetta, and Nomonyx, and so is derived relative to the unbuttressed state where the facet is parallel to the shaft in Anseranas,Thalassornis,Biziura, and Oxyura. The space between the facet for the anterior ligament and the dorsal condyle is wider than this facet, versus narrower in Thalassornis and Dendrocygna bicolor, which latter state is regarded as more primitive (Woolfenden, 1961). In Pinpanetta,the sulcus scapulotricipitalis (scapulotricipital groove) extends from the caudal surface around the distal margin, as in most oxyurines, tadornines and anatines, and is therefore derived compared to Anseranas,Cnemiornis and Biziura, which either lack or have a barely defined scapulotricipital groove caudally, and to Dendrocygna,Thalassornis and some anserines, where the groove exists only on the caudal face. Although the epicondylus dorsalis is well developed and dorsally prominent level with the proximal margin of the dorsal condyle in Pinpanetta, a distinct ectepicondylar prominence of similar or relatively larger size to that in Malacorhynchus is also present. A distinct ectepicondylar prominence is present in Anseranas, anserines and Dendrocygna, and so its presence in Pinpanetta is a retained plesiomorphy shared with Thalassornis and Stictonetta, but the derived condition (lack of ectepicondylar prominence) is found in Oxyura,Nomonyx, Biziura, most tadornines, and all Anatinae. Pinpanetta has the attachment of the pronator brevis (sensu Howard, 1929) in an isolated pit on the ventral facies of the ventral epicondyle, thus distinguishing it from Thalassornis, Oxyura,Nomonyx, and Biziura which have a derived state with the attachment area fused with the ventral margin of the facet for the anterior ligament. The brachial fossa is elongate, with well-defined margins, is separated from the ventral margin by a narrow rounded ridge, and lacks secondary deepening distoventrally, as in Manuherikia.InStictonetta, the fossa is poorly defined and flat. SPECIES PINPANETTA TEDFORDI SP.NOV.(FIG.1) Holotype: SAM P.41257, complete R humerus (Fig. 1A, F), reassembled from two pieces; shaft with some wear on distal margin of bicipital crest and on caudal shaft surface; light brown in colour. Diagnosis: A species of Pinpanetta about the size of Oxyura australis, characterized by the following features: dorsal pneumotricipital fossa excavated below head; elongate deltoid crest with about 50% length extending distad of bicipital crest; ventral pneumotricipital fossa, from median crest, wider than its length measured from the ventral tubercle to the distal end of the bicipital crest – shaft junction; shaft narrows distally; facet for the attachment of anterior ligament directed distoventrally; and on ventral facies, attachment of pronator brevis positioned towards cranial facies, not centrally. Etymology: For R. H. (Dick) Tedford whose efforts over many years have revealed much about the Namba and Etadunna Formations and their contained faunas and whose expeditions collected many of the following specimens. Type locality: Young Bucks Quarry, site code RV-9002, Lake Palankarinna, 28° 47′S, 138° 24′E, Tirari Subbasin, Lake Eyre Basin, SA, collected N.S. Pledge et al., xii.1992. Horizon: Stratigraphy/Age/Fauna: Etadunna Formation, Late Oligocene 24–26 Mya, Minkana LF, Zone A. Distribution: Late Oligocene (24–26 Mya): Lake Palankarinna, Etadunna Formation, Minkana LF, AUSTRALIAN TERTIARY ANATIDS 417 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Figure 1. Right humeri of Pinpanetta species, A–E, cranial view, and F–J, caudal view. A, F, Pinpanetta tedfordi SAM P.41257, holotype; B, G, Pi. tedfordi UCMP 56998, paratype; C, D, H, I, Pi. vickersrichae SAM P.42703, holotype two nonarticulating fragments of one bone; and E, J, Pi. fromensis SAM P.43128, holotype. Scale bars =10 mm. See main text for abbreviations. 418 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Zone A and Ditjimanka LF, Zone B, Member 7; Lake Pinpa, Namba Formation, Pinpa LF. Measurements of holotype: TL =65.6 mm, PW = (maximum width from dorsal tubercle) 14.5 mm, SW =4.0 mm, DW =9.0 mm, depth dorsal condyle = 5.3 mm. Paratypes: Lake Palankarinna, Etadunna Formation: SAM P.42699, s+dR humerus, Neville’s Nirvana, Minkina LF, Zone A, collection code VSQ 1978-40P. UCMP 46173, s+dL humerus, Tedford Locality Site 2, Ditjimanka LF, Zone B, Member 7, site code UCMP V-5375, collected by R. A. Stirton 1954, collection code RAS #4803. UCMP 56998, well preserved R humerus with the shaft broken and joined with some plaster infill caudally, Stirton Site 2, Ditjimanka LF, Zone B, Member 7, site code UCMP V-5375, collected by R. H. Tedford et al., 1957, collection code RHT#450 (Fig. 1B, G). Lake Pinpa; Namba Formation, Pinpa LF: SAM P.43133 (formerly AMNH 10957), dR humerus, collected by R. H. Tedford et al., 1971 at Site C, collection code QMAM 151; SAM P.43130 (formerly AMNH 10835), d+sL humerus, collected by R. H. Tedford et al., 1971 at Site C, collection code QMAM 252. Referred material: Humeri – SAM P.23480, part pR humerus, Lake Pinpa, Namba Formation, Pinpa LF. SAM P.27846, dR humerus, west of site SIAM, Lake Palankarinna, Etadunna Formation, Ditjimanka LF, Zone B. SAM P.41262, dL humerus, White Sands Basin, Lake Palankarinna, Etadunna Formation, Ditjimanka LF, Zone B. QM F52743 (=AR17105), dL humerus, Ringtail Site, Gag Plateau, System C, Riversleigh. Measurements: See Table 1. Description and comparison: Humeri of Pi. tedfordi have the following additional features: dorsal pneumotricipital fossa variably excavated under the head, slightly in holotype, marked in UCMP 56998 as in Malacorhynchus; pit for attachment of ligamentum collaterale dorsale on dorsal face of ectepicondyle, deep, divided by median ridge as in all anatids, although this median ridge is unusually weak in SAM P.41257; olecranal fossa well marked; brachial fossa elongate, aligned up shaft, in holotype c. 2.1 mm wide by 5 mm long with proximal dorsal margin barely extending past midshaft width, relatively deeper and extends closer to dorsal margin in other specimens, e.g. UCMP 46173, UCMP 56998, SAM P.43133. Although distinguished from all oxyurines as indicated in the generic description, humeri of Pinpanetta tedfordi are most similar to those of Oxyura and Malacorhynchus. Both are distinguished from Pi. tedfordi by a marked notch at the ventral end of the capital groove and a shorter deltoid crest. Oxyura further differs as follows: relatively wider ventral pneumotricipital fossa and bicipital crest; distinct groove dorsad of median crest; attachment of pronator brevis fused with ventral facies of ventral epicondyle; and attachment of anterior ligament not buttressed anteriorly. Malacorhynchus humeri further differ by: lack of distal narrowing of the shaft; pocket of ventral pneumotricipital fossa much deeper; distal margin of bicipital crest more convex; facet for attachment of anterior ligament more distally directed; attachment of pronator brevis on ventral facies more central. The New Zealand Early Miocene fossils Manuherikia and Dunstanetta share with Pi. tedfordi a distally narrowing shaft, but differ in the features listed above, notably with a marked notch in their proximal profile, a shorter deltoid crest, an elongate dorsal tubercle, a deeper ventral pneumotricipital fossa that extends under the median crest, and a relatively wider dorsal pneumotricipital fossa. QM F52743, from Ringtail Site, System C, is one of only two confirmed anatid bones from the Carl Creek limestone at Riversleigh. It is slightly larger but otherwise indistinguishable from SAM P.43130 from the Namba Formation, which is one of the larger specimens attributed to Pinpanetta tedfordi (Table 1). Table 1. Measurements (mm) of humeri of Pinpanetta tedfordi Specimen TL PW from DC SW min DW Depth DC SAM P.41257 65.6 14.5 4.0 9.0 5.3 SAM P.42699 – – 4.1 8.7 5.0 UCMP 46173 – – 4.0 8.6 4.8 UCMP 56998 66.8 13.9 4.5 9.2 5.5 SAM P.43133 – – – 8.9 – SAM P.43130 – – 4.7 9.5 5.5 QM F52743 – – – 9.7 5.4 Abbreviations as in Methods; and DC, dorsal condyle; min, minimum. AUSTRALIAN TERTIARY ANATIDS 419 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Measurements of holotype: TL =50.85 mm, PW from dorsal tubercle =11.0 mm, length deltoid crest = 14.4 mm, SW =3.8 mm, DW =8.3 mm, depth dorsal condyle =4.8 mm. Paratypes: SAM P.42675, pL humerus with worn ventral margin to bicipital crest, and SAM P.42676, dL humerus (DW =7.8 mm, SW =3.8 mm), both Site 2, Billeroo Creek, 31° 06.205′S; 140° 13.912′E; Namba Formation, Pinpa LF, collected by THW and A. Camens, June 2007. Referred material: The following specimens are referred to Pi. fromensis: those other than humeri on the basis of expected size given the size of the humerus. Humerus: SAM P.42674, dL humerus, SAM P.42677, worn dR humerus, both Site 2, Billeroo Creek, 31° 06.205′S 140° 13.912′E, Namba Formation, Pinpa LF. AMNH 10954, pL humerus; Lake Pinpa, Site C, collection code QMAM 151, Namba Formation, Pinpa LF, measurements – PW =11.1 mm. Description: The reconstructed holotype humerus of Pi. fromensis retains damage to the ventral side of the head resulting in loss of the floor of the capital groove and so preventing the form of the end of the capital incision from being determined. However, AMNH 10954 reveals the proximal profile to have a shallow notch, one deeper than the other Pinpanetta species. The deltoid crest is missing the proximal 5 mm but the remaining section is relatively high, extending 3.2 mm above the adjacent cranial surface. It is concave dorsally, and the attachment scar for m. latissimus dorsi posterioris links to its distal end before extending farther distally. Pinpanetta fromensis differs from Pi. tedfordi and Pi. vickersrichae as follows: dorsal pneumotricipital fossa more deeply excavated, so capital groove opens to fossa from shelf, more excavated under head, and fossa forms a groove adjacent to median crest; deltoid crest shorter. It differs from Pi. tedfordi by the ventral pneumotricipital fossa being as wide as long and that the shaft diameter does not narrow distally. Humeri of Mal. membranaceus are only slightly larger than Pi. fromensis and have a similar development of the dorsal pneumotricipital fossa, e.g. SAM B.39385; however, they differ as follows: ventral end of the capital groove with more distinct notch in proximal profile; ventral pneumotricipital fossa relatively wider; deltoid crest shorter, beside whose distal end the attachment scar for m. latissimus dorsi posterioris commences and passes distad of without connection; brachial fossa smaller; facet for attachment of anterior ligament craniodistally directed (rather than cranioventrally). Humeri of the similar-sized Man. minuta of the St Bathans Fauna in New Zealand differ from those of Pi. fromensis as follows: dorsal tubercle distinctly elongate rather than about as wide as long; proximal profile at ventral end of capital groove distinctly notched; ridge for attachment of supraspinatus, although extending to level with end of bicipital crest in both taxa, is much less prominent, such that there is no distinct groove dorsad of median crest; intumescentia humeri less inflated cranially; deltoid crest less elevated from cranial surface; brachial fossa markedly deepened distoventrally rather than relatively flat and even depth. The ventral tubercle extends at right angles to the ventral pneumotricipital fossa in Pi. fromensis. In all diving taxa, it is directed slightly distally to overhang the fossa, sometimes markedly, e.g. Oxyura. The deeper dorsal pneumotricipital fossa may be associated with a diving habit as it is deepened and broadened to an extreme in Oxyura; however, it is deep in Malacorhynchus which is not a specialized diving taxon. Most specialized divers have a distally narrow humeral shaft, e.g. Oxyura, Aythya,Mergus, a feature not seen in Pi. fromensis. Similarly, the facet for the attachment of the anterior ligament is directed cranially, rather than distoventrally as seen in specialist divers across diverse clades of anatids e.g. Oxyura,Aythya, and Mergus. These observations suggest that P. fromensis was not a specialized diver, and in this was similar to Pi. vickersrichae. Ulna: SAM P.22837, L ulna, Mammalon Hill, Lake Palankarinna, Etadunna Formation, Ngama LF, Zone D, Member 8, measurements – preserved length = 44.7 mm, estimated TL =46 mm. The single available specimen is complete, but worn proximally so that the structure of the dorsal cotylar process and of the tuber. bicipitale ulnae in the incisura radialis is undeterminable. It has the following features: brachial fossa shallow; tuberculum for ventral collateral ligament not separated from ventral cotyla by deep groove (as in Man. minuta). It is smaller than ulnae of Nettapus with which it shares a shallow brachial fossa. If correctly associated with the humerus, this taxon has a comparatively short ulna. Carpometacarpus (Fig. 4): SAM P.42700, L carpometacarpus (Fig. 4A–C), Neville’s Nirvana, Lake Palankarinna, collection code VSQ 1978-40P, Etadunna Formation, Minkina LF, Zone A, measurements – TL =31.8 mm, PW =7.6 mm. AMNH 10938, L carpometacarpus lacking only the minor metacarpal, Lake Pinpa, Site C, Namba Formation, Pinpa LF, measurements – TL =29.0 mm, PW =7.1 mm. SAM 426 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 P.41261, worn L carpometacarpus, White Sands Basin, Lake Palankarinna, Etadunna Formation, Ditjimanka LF, Zone B, measurements – estimated length =29.5 mm, PW =6.5 mm. AMNH 10725, worn pR carpometacarpus, Lake Pinpa, Site C, collection code QMAM 264, Namba Formation, Pinpa LF. AMNH 10745, pL carpometacarpus, Lake Pinpa, Site C, Namba Formation, Pinpa LF, measurements – PW =7.0 mm. AMNH 10748, worn L carpometacarpus, Lake Pinpa, Site C, collection code QMAM 243, Namba Formation, Pinpa LF, measurements – PW =6.9 mm. AMNH 10951, pR carpometacarpus, Lake Pinpa, Site C, collection code QMAM 151, Namba Formation, Pinpa LF, measurements – PW =6.7 mm. The well-preserved specimens SAM P.42700 and AMNH 10938 enable the following character states to be determined: external rim of trochlea carpalis with a shallow carpal notch; both the anterior carpal fossa and cuneiform fossa are present, deep; internal carpal fossa deep, distinct foramen in base, separated from extensor process by rounded ridge; one ligamental facet for ligamentum ulnocarpo-metacarpale dorsale below the carpal rim (not two); short proximal synostosis or region from intermetacarpal space to proc. alularis (pollical facet); minor metacarpal not grooved at the proximal synostosis; flexor attachment marked by single scar distal of proximal synostosis of minor and major metacarpals; distal synostosis shorter than distal width; facets for digits II and III of approximate equal distal length. These specimens are slightly smaller than that of Nettapus, and bigger than those of Man. minuta. Coracoid (Fig. 4): SAM P.41301, worn R coracoid (Fig. 4D), with the acrocoracoid, the tip of the procoracoid, the tip of the medial angle, and the lateral process missing, SAM North (a white sand locality), Lake Palankarinna, 28° 46.503′S; 138° 24.164′E, Etadunna Formation, Ditjimanka LF, Zone B. MV 222424, cranial part L coracoid preserved from the mid shaft (Fig. 4E), Lake Pinpa, (location code MV No. 2348), Namba Formation, Pinpa LF. Together these specimens show the following: acrocoracoid not pneumatic, dorsoventral plane of acrocoracoid aligned nearly at right angles to plane of sternal end; both dorsal and ventral lobes of clavicle facet extensively overhang supracoracoidal sulcus; ventral clavicle facet does not overhang ventral shaft facies; supracoracoidal sulcus not excavated below humeral facet; scapular cotyla large and oval; procoracoid lacks a foramen; dorsal surface of the blade without pneumatic fossa; ventral facies flat; and ventral sternal facet present but not prominent of the ventral facies. These coracoids are very similar to those referred to Pi. tedfordi and Pi. vickersrichae, but are smaller than both. Among extant taxa, SAM P.41301 is most similar to coracoids of Nettapus pulchellus and Mal. membranaceus, although it is smaller. It is more similar to Malacorhynchus and differs from Nettapus in the shape of the humeral facet, which is widest at about mid-length (not widest and rounded towards acrocoracoidal end as in Nettapus). Similarly, MV 222424 is very similar to both Pi. tedfordi and Malacorhynchus in the form of the clavicle facet that extensively overlaps the supracoracoidal sulcus. Measurements: SAM P.41301 – estimated total medial length =27 mm, SW =3.2 mm; MV 222424 – length scapular cotyla-head =8.3 mm; SW =2.9 mm; length humeral facet =6.0 mm. Scapula: SAM P.42673, L scapula, Billeroo Creek, Site 2, 31° 06.205′S; 140° 13.912′E, Namba Formation, Pinpa LF, measurements – width acromion to humeral facet =6.7 mm, shaft height =2.2 mm. Small anatid scapula, shaft with parallel dorsal and ventral margins, acromion not directed dorsally from shaft, coracoidal articulation globose and prominent, humeral facet laterally orientated. Tibiotarsus: AMNH 10953, pL tibiotarsus, Lake Pinpa, Site C, Namba Formation, Pinpa LF, measurements – preserved length =9.0 mm, PW =5.8 mm. SAM P.42678, dL tibiotarsus with broken lateral condyle, Billeroo Creek, Site 2, 31° 06.205′S 140° 13.912′E, Namba Formation, Pinpa LF, measurements – SW =2.5 mm. These specimens are smaller than those referred to Pi. tedfordi, but have a similar morphology. They reveal that the medial and lateral articular facets of the proximal end have a similar depth (rather than the medial one extending markedly caudad of the lateral one), and that they are separated caudally by a distinct notch. The structure of the cnemial crests is not determinable. Tarsometatarsus: AMNH 10838, worn dR tarsometatarsus, Lake Pinpa, Site C, Namba Formation, Pinpa LF, measurements – DW =c. 6.0 mm. This specimen is referred to Anatidae as it lacks a fossa metatarsi I (metatarsal fossa) and the plantar exit for the distal foramen opens in a groove into the lateral intertrochlear notch, and is referred on size to Pi. fromensis. Trochlea metatarsal II does not extend distad of the lateral intertrochlear notch. SUBFAMILY TADORNINAE REICHENBACH 1849–1850: SHELDUCKS The following taxon is referred to the tadornines as the humerus has the following characters: (1) AUSTRALIAN TERTIARY ANATIDS 427 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 proportions as in Tadorna; (2) elevated and elongate dorsal tubercle; (3) capital shaft ridge directed at area between dorsal tubercle and head; (4) dorsal pneumotricipital fossa narrow, not excavated under head; (5) ventral pneumotricipital fossa large, more than half proximal width, wider than long, pneumatic; (6) the osseous lamella extending from the caudal shaft surface around the distal margin of the ventral pneumotricipital fossa merges with the bicipital crest/ base of fossa in ventral half of fossa; (7) deltoid crest dorsally concave; (8) facet for anterior ligament elevated, directed distally; (9) scapulotricipital sulcus a distinct groove caudally, extends around distal end; (10) distinct ectepicondylar prominence; (11) flexor process equal distal extent to dorsal condyle; Character 1 is derived in tadornines with Anseranas, dendrocygnines, and anserines plesiomorphic with an elongate humerus. Character 3 is derived in tadornines relative to Anseranas and anserines where the ridge is directed towards the head. Character 5 is derived in tadornines with Anseranas, dendrocygnines, and anserines relatively plesiomorphic with a narrow fossa. Character 6 is considered an apomorphy of tadornines with the plesiomorphic state of the lamella remaining elevated from the floor of the fossa and extending up under the ventral tubercle in Anseranas, dendrocygnines, and anserines: anatines are more derived with the lamella merging with the floor of the fossa in the dorsal part of the fossa. Character 7 and the elevated dorsal tubercle are retained plesiomorphic features that distinguish tadornines from anatines that have derived states (deltoid crest convex or flat dorsally; dorsal tubercle not elevated off adjacent facies). Character 8 is derived relative to the unelevated state in Anseranas or the low elevation of the facet in anserines. Character 10 is a retained plesiomorphy with the prominence most developed in the tadornines (Miotadorna,Alopochen) that distinguishes tadornines from the anatines in which it is apomorphically lost. Humeri of Anseranas differ greatly, for example, more elongate; capital shaft ridge directed towards the head; a relatively small ventral pneumotricipital fossa largely occluded by a broad osseous lamella extending from the caudal shaft surface; short flexor process; no scapulotricipital groove caudally or distally; and facet for anterior ligament not elevated. Those of anserines, geese and swans, are more elongate; the osseous lamella extending from the caudal shaft surface into the ventral pneumotricipital fossa remains elevated off the base of the fossa and extends up under the ventral tubercle; and the capital shaft ridge is directed towards the head. Dendrocygna species are all smaller; the dorsal tubercle is near circular; the osseous lamella extending from the caudal shaft surface into the ventral pneumotricipital fossa is as in anserines; and the ventral pneumotricipital fossa is small, much less than half proximal width. Humeri of the oxyurines, Biziura,Oxyura,and Malacorhynchus differ in several ways but all have a closed or nonpneumatic ventral pneumotricipital fossa. Stictonetta, apart from the smaller size, differs with the capital shaft ridge directed towards the dorsal tubercle and so has a wider dorsal pneumotricipital fossa. All anatines differ by the derived loss of the capital shaft ridge and loss of elevation of the dorsal tubercle, and by the dorsal surface of the deltoid crest being flat or convex. GENUS AUSTRALOTADORNA GEN.NOV. Type species: Australotadorna alecwilsoni sp. nov. Diagnosis: Humerus more robust than other tadornines, ventral pneumotricipital fossa larger, and caudal margin of bicipital crest in ventral view forming a single plane from ventral tubercle to junction with shaft rather than forming a marked angle with the proximal section at near right angles to the shaft. Etymology: For its inferred ancestral relationship to the shelduck Tadorna, and that it derives from Australia. Description and comparison: Australotadorna gen. nov. shares with tadornines the above features and that the capital groove forms only a shallow notch in the proximal profile, not a deep one as in most oxyurines and all anatines. The ventral pneumotricipital fossa is large and, with its internal diameter at about 46% of proximal width, is broader than that of all other tadornines. It is highly pneumatic and thus differs from Miotadorna where bone struts fill the fossa inside of the median crest. The capital shaft ridge is better developed and farther separated from the dorsal tubercle than it is in Tadorna species, but is in this similar to Miotadorna from New Zealand. As in Miotadorna, it has the dorsal tubercle separated from the capital ridge by a flaring groove extending onto the dorsal surface. Alopochen differs with a dorsal tubercle about as wide as long, rather than elongate, and by the capital shaft ridge being more directed to the dorsal tubercle. The latter results from a deeper and broader dorsal pneumotricipital fossa that extends under the head. Chenonetta and Hymenolaimus, two genera often associated with tadornines, have a much wider dorsal pneumotricipital fossa with the capital shaft ridge more weakly developed and directed towards the dorsal tubercle. The well developed ectepicondylar prominence is better developed than in Tadorna but similar to the development in Miotadorna and Alopochen. The brachial fossa was 428 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 secondly deepened distoventrally and did not extend closer than 3 mm of the dorsal shaft margin. Such a secondary deepened brachial fossa is seen in Miotadorna and Alopochen. SPECIES AUSTRALOTADORNA ALECWILSONI SP.NOV.(FIGS 5, 6) Holotype: SAM P.43141 (formerly AMNH 11499), L humerus (Fig. 5B, D, F), reassembled from several fragments with plaster infill in areas of shaft so that distal end is distorted ventrally; dorsal margin of deltoid crest lost,a7by5mmhole in ventral part of intumescentia humeri, distal margin of bicipital crest worn, and craniodistally part of the shaft proximal to the attachment of the anterior ligament extending through the brachial fossa is lost; light brown in colour. Diagnosis: As for genus. Etymology: For Alec Wilson, owner of Frome Downs Station who has freely allowed access to the fossil sites and helped palaeontologists over many years. Type locality: Lake Pinpa (=Pine Lake), 31° 8′S, 140° 13′E, Lake Eyre Basin, Callabonna (=Tarkarooloo) Sub-basin, SA, north-west side Pine Lake, near grid coordinates 317148 on the CURNAMONA (prov. ed.) sheet (R. H. Tedford, pers. comm. 30 Aug 2006), collected by R. H. Tedford et al., 1971, collection code QMAM 74. Horizon: Stratigraphy/Age/Fauna: Namba Formation, Late Oligocene 24–26 Mya, Pinpa LF. Distribution: Late Oligocene (24–26 Mya): Lake Pinpa, Namba Formation, Pinpa LF; possibly Lake Palankarinna, Etadunna Formation (see below). Figure 5. Left humeri of tadornines in caudal (A, B), cranial (C, D) and ventral (E, F) views: A, C, E, modern T. tadornoides SAM B.39591; and B, D, F, Australotadorna alecwilsoni (SAM P.43141). The arrow points to the planar caudoventral margin of the bicipital crest compared to the angled margin in Tadorna. Scale bar =10 mm. See main text for abbreviations. AUSTRALIAN TERTIARY ANATIDS 429 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Measurements of holotype: TL =132.4 mm (as preserved bent), estimated restored TL =134 mm, PW (maximum from dorsal tubercle) =28.7 mm, SW = 10.2 mm, DW =22.4 mm, depth dorsal condyle = 12.6 mm. Paratype: Coracoid (Fig. 6): SAM P.43137 (formerly AMNH 11497), cranial part R coracoid (Fig. 6I), Lake Pinpa, north-west side near CURNAMONA (prov. ed.) sheet grid co-ordinates 317148, R. H. Tedford, pers. comm. 30 Aug 2006, collection code QMAM 65, R. H. Tedford et al., July 1971, Namba Formation, Pinpa LF. Measurements – preserved length =38 mm, length humeral facet =15 mm, SW =c. 7.5 mm. Description and comparison Humerus: As for genus. Coracoid: SAM P.43137 preserves only the dorsal half of the cranial end so lacks the tip of the acrocoracoid, the ventral clavicle facet and the ventral half of the supracoracoidal sulcus. It is of appropriate size for Australotadorna, given the size of the humerus SAM P.43141. It has the following features: deep circular scapular cotyla, c. 7 mm in diameter; short supracoracoidal sulcus not undercutting the humeral facet; ridge separating supracoracoidal sulcus from humeral facet relatively short, broad and rounded; shallow fossa with small pneumatic foramina adjaFigure 6. Fossil tadornine bones compared with modern Tadorna tadornoides SAM B.39591. Tadorna tadornoides: A,C. proximal right carpometacarpus; and H, dorsal view cranial half coracoid. Fossils referred to Australotadorna alecwilsoni: B, D, E, proximal right carpometacarpus; F, distal right tibiotarsus SAM P.36762 in anterior view; G, cranial part right coracoid (SAM P.24531) in dorsal aspect; I, cranial part right coracoid (SAM P.43137) in dorsal aspect. Fossils referred to an undetermined tadornine from Alcoota: J, left radius UCMP 65985 in dorsal aspect; and right carpometacarpus NT P.2913 in K, ventral; L, dorsal; and M, caudal views. Scale bars =10 mm. See main text for abbreviations. 430 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 cent to the scapular cotyla on sulcus side of this ridge; dorsal clavicle facet markedly overhangs sulcus as sharp crest with pneumatic foramina under it; procoracoid short and robust, although tip lost; procoracoid lacking a foramen in it, although insufficient of the shaft, remains to determine if it had a notch as seen in Cereopsis; supracoracoidal sulcus extended as a groove along ventral side of procoracoid; ventral facies apparently flattened adjacent procoracoid, not very deep as in Cereopsis. Although SAM P.43137 is of similar size to coracoids of Anseranas, it differs as follows: it lacks a procoracoidal foramen; it has a sharp dorsal lobe of the clavicle facet and pneumatic foramina in the supracoracoidal sulcus penetrating the acrocoracoid (both lacking in Anseranas). The pneumatic acrocoracoid suggests an affinity with anserines, but only Cereopsis has a similarly sharp overhanging dorsal clavicle lobe. Tadornines, e.g. Alopochen and Miotadorna also have pneumatic foramina under the clavicle facet, so their presence is a plesiomorphic feature. Whereas the fossil shares with Cereopsis a relatively broad, short, rounded ridge between the supracoracoidal sulcus and the humeral facet, a groove along the ventral side of the procoracoid is not seen in anserines but is in tadornines. The broad, rounded ridge craniad of the scapular cotyla differs from tadornines where it is narrower, but this may relate to the larger size of the fossil. Larger specimens of Miotadorna, e.g. NMNZ S.42315 have a short relatively inflated ridge approaching the condition seen in SAM P.43137. A shallow pneumatic fossa dorsally immediately craniad of the scapular cotyla in the fossil has its homologue in a small elongate pneumatic fossa on the supracoracoidal sulcus side of the ridge in tadornines, e.g. in Tadorna tadornoides and Miotadorna. The insertion on the shaft of the procoracoid is relatively shorter than in Tadorna and Miotadorna but breakage means its medial, ventral, and cranial extents are not determinable. In summary, SAM P.43137 is of appropriate size for Australotadorna and shares more features with tadornines than other groups of waterfowl, but it is distinguished by several apomorphies, such as a more circular scapular cotyla and shorter procoracoidal insertion. Their significance is obscure at present, but given that Australotadorna is several million years older than the next known tadornine, Miotadorna, such differences are likely to be plesiomorphic features. Referred material: The following specimens are referred to Australotadorna alecwilsoni on the basis of their appropriate size and tadornine-like morphology. Coracoid (Fig. 6): SAM P.24531, cranial part R coracoid (Fig. 6G), Tedford Locality, Lake Palankarinna, Etadunna Formation, Ditjimanka LF, Zone B. Measurements – SW =5.8 mm, length humeral facet c.=12.3 mm. The specimen lacks the entire sternal blade, and the tip of the acrocoracoid including the clavicle facets. It has a standard anatid form: the scapular cotyla is circular and deep; procoracoid lacks a foramen; and the supracoracoidal sulcus does not have a secondary fossa below the humeral facet. Coracoids of Phoenicopteriforms, present in the Etadunna Formation, differ with a procoracoidal foramen. It is of similar size to coracoids of tadornines. The dorsal facies of the ridge leading from the humeral facet to the acrocoracoid is like that in SAM P.43137, short and broadly inflated. It is smaller than SAM P.43137; however, the size difference is within that seen in sexual dimorphism in modern anatids. SAM P.43137 is pneumatic below the dorsal clavicle facet but this cannot be determined in SAM P.24531. In summary, this specimen may be referred to Australotadorna alecwilsoni or a very similar taxon. Carpometacarpus (Fig. 6): All from Lake Pinpa, Site C, Namba Formation, Pinpa LF, collected by R. H. Tedford et al., 1971. AMNH 10728: pL carpometacarpus, PW =16 mm. AMNH 10818, pL carpometacarpus (Fig. 6B, D, E), collection code QMAM 175, PW =15.7 mm. AMNH 10864, worn pR carpometacarpus, collection code QMAM 175, PW =14.6 mm. These specimens preserve only the proximal half of the element. They are about the size of Tadorna tadornoides but have a shorter extensor process. They are tentatively referred to Australotadorna on the basis of size and that there is no evidence of another large anatid in the source fauna. They have the following features: shallow infratrochlear fossa, a deep cranial fossa, moderately deep cuneiform fossa, an elongate proximal synostosis, and the minor metacarpal is not grooved at the fornix. Only AMNH 10818 allows the morphology of the ligamental scars proximally on the dorsal surface. The proximal one is prominent, but a second, shallow and elongate distal scar usually found in tadornines is present only as an obscure elongate hollow, but polishing of the bone by windblown sand may have obscured its real form. Scapula: UCMP 57156, R scapula, Lake Palankarinna, Tedford Locality Site 3, UCMP V-5762, R. H. Tedford 1957, collection code RHT#465, Etadunna Formation, Minkina LF, Zone A, Member 4, measurements – acromion to ventral margin of humeral facet =14.7 mm, depth of collum scapulae (blade) = 5.9 mm. UCMP unreg., R scapula, Leaf Locality, Lake AUSTRALIAN TERTIARY ANATIDS 431 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Ngapakaldi, UCMP V-6313, collection code RAS #5294, Wipajiri Formation, Early Miocene, Kutjamarpu LF, measurements – blade depth =5.2 mm. UCMP 57156 is missing most of the blade. It has the general shape of anatid scapulae, particularly a compressed dorsal margin to the blade, acromion extending craniad of tuberculum coracoideum, and the coracoid tubercle prominent of a line joining the humeral facet and acromion. It is distinguished from Anseranas and anserines by the lack of pneumatic foramina either laterally or medially. In size, it is similar to that of Tadorna tadornoides and thus larger than scapulae of Biziura,Oxyura,Dendrocygna,Malacorhynchus,Stictonetta, and Aythya and other anatines. UCMP unreg., worn and lacking most of the blade and the acromion, is similar to UCMP 57156, but preserves little of diagnostic value. Tibiotarsus (Fig. 6): SAM P.36762, dR tibiotarsus (Fig. 6F), White Sands/SAM North, RV-7247, Lake Palankarinna, Etadunna Formation, Ditjimanka LF, Zone B, Member 6. Of typical anatid shape, SAM P.36762 is smaller than tibiotarsi of Anseranas, but is slightly larger and has a stouter shaft than Tadorna tadornoides (Table 3). It is thus smaller than anserines, e.g. Cereopsis, and larger than Biziura,Stictonetta, oxyurines, and all anatines. SAM P.36762 is also similar to tibiotarsi of Tadorna in qualitative features and differs from those of Anseranas as follows: short tendinal bridge; distal end with distal margin near right angles to the shaft (not sloped distally); medial condyle forming notch with shaft at proximo-anterior junction (rather than a wide angle); and medial prominence small and occluded by medial condyle in anterior view (large, protuberant). It differs from those of Dendrocygna by larger size and greater medial inflection of the medial condyle. As it has tadornine features it is tentatively referred to Australotadorna alecwilsoni. Measurements: See Table 3. GENUS INDETERMINATE,MAGN.TADORNA SP. Humerus: AMNH 10966, fragment pR, Lake Pinpa, Site A, R.H. Tedford 1971, Namba Formation, Pinpa LF, measurements – width of head =16.0 mm, depth of head =8.0 mm. Description: This fragment preserves only the head and part of the dorsal tubercle. It is tentatively referred to Anseriformes as the caudal margin of the head is roughly at right angles to the presumed line of the shaft. The dorsal pneumotricipital fossa did not extend under the head. It is smaller than SAM P.43141 and of similar size to Tadorna, but nothing further can be determined. ALCOOTA TADORNINE,GENUS INDETERMINATE, MAGN.TADORNA TADORNOIDES (FIG.6) The following two specimens are referred to Tadorninae but in the absence of more diagnostic elements they are not named. Table 3. Measurements (mm) and selected ratios (as percentages) of the fossil tibiotarsus SAM P.36762 compared to summary statistics [mean (range), standard deviation] for those for Tadorna tadornoides (five M, five F), and Anseranas semipalmata (one F, two M), all from SAM (catalogue numbers T. tadornoides: SAM B 39568, 39575, 39584-5, 39592, 39866, 39874, 39877-9; Anseranas, as below) SW Caudal width distal condyles Height medial condyle Depth between distal condyles Tadorna tadornoides n=10 5.3 (5.1–5.8) 0.21 9.1 (8.7–9.6) 0.31 12.8 (12.1–13.5) 0.42 8.2 (7.5–8.5) 0.33 Anseranas, F, 39824 6.8 10.5 15.5 10.8 Anseranas, M, 36790 7.6 11.7 17.7 11.9 Anseranas, M, 48035 8 11.2 18.5 12.6 SAM P.36762 6.4 10 14.6 c. 8.4 SW/CWD SW/HMC CWD/HMC SW/DD Tadorna tadornoides n=10 57.77 (56.0–60.4) 1.35 41.22 (39.8–43.0) 0.32 71.36 (69.5–73.2) 1.21 64.83 (60.7–68.3) 2.48 Anseranas, F, 39824 64.96 42.94 66.10 63.87 Anseranas, M, 36790 64.76 43.87 67.74 62.96 Anseranas, M, 48035 71.43 43.24 60.54 63.49 SAM P.36762 64.00 43.84 68.49 76.19 Abbreviations: SW, shaft width; CWD, caudal width distal condyle; HMC, height of medial condyle; DD, depth between distal condyles; F, female; M, male. 432 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 NTM P.2913: R carpometacarpus (Fig. 6K–M), Alcoota, quarry unknown, Alcoota Homestead, NT, 22° 52′S, 134° 27′E, Waite Formation, Miocene, c. 8 Mya, Alcoota LF. Measurements – PW =16.7 mm, length to distal end intermetacarpal space =64 mm, proximal synostosis wider (7.5 mm) than long (c. 6.5 mm). NTM P.2913 lacks the synostosis metacarpalis distalis (distal synostosis) and the minor metacarpal. It has the following features: dorsal rim of the carpal trochlea distinctly notched; infratrochlear fossa shallow with base slightly above ventral facies of extensor process and lacking a large foramen in its base; proximal margin of extensor process at right angles to shaft; extensor process less than half proximal width; alular process distinctly distad of distal end of ventral rim of carpal trochlea; small distinct fossa dorsally adjacent to alular process; dorsal facies of carpal trochlea has two distinct ligamental attachments for the external ligament, one proximal, slightly elevated, caudally directed, and larger of pair, one distal and more elongate, level with the cuneiform fossa; cuneiform fossa extends deeper than the adjacent shaft facies; anterior carpal fossa present; the minor metacarpal at the proximal synostosis not grooved, has a prominent tuberosity on ventral margin centred on the synostosis; tuberosity for flexor attachment distad of the proximal synostosis; and cranial facies of the major metacarpal is convex. This fossil is referred to Anatidae because of the apomorphic presence of a distinct notch in the dorsal rim of the carpal trochlea. The conformation of the ligament scars on the dorsal facies (distal one more elongate, larger than proximal one of pair, and level with the cuneiform fossa, proximal one slightly elevated and caudally directed) is an apomorphy for Tadorninae (including Hymenolaimus) and Anserinae (including Cereopsis): in all anatines the distal scar is smaller (second scar lacking in Anseranas). Both can be excluded from Cygnini by the lack of the apomorphic state for cygnines of a distally sloping extensor process, and from all anserines by markedly smaller size. The carpometacarpus of Anseranas, which in females may be of similar size to the fossil, differs markedly as follows: the infratrochlear fossa has a large foramen in its base and is markedly elevated above the ventral facies of the extensor process; it lacks an anterior carpal fossa; it has a paired flexor attachment; the dorsal rim of the carpal trochlea has a very shallow or incipient notch; the cranial facies of the major metacarpal is markedly flattened; it has only a single scar for the external ligament attachment. Cereopsis, apart from being larger has a markedly more robust extensor process, more robust and ventrally directed pisiform process, and flattened cranial facies to the major metacarpal. Most members of Anserinae (including Cereopsis) lack an anterior carpal fossa (exception, Anser brachyrhynchus)and most have a rounded or very weakly grooved minor metacarpal at the proximal synostosis (exceptions Anser brachyrhynchus and some individual Branta canadensis). However, the small distinct fossa adjacent to the alular process is seen in Cereopsis. This fossil is similar in size to carpometacarpi of Tadorna and so larger than Biziura,Oxyura,Stictonetta,Malacorhynchus,Anas,Aythya, and other anatines. Although it is of similar size to the Australian endemic Tadorna tadornoides, the latter has a more drawn-out tip to the extensor process, where proximodistal width increases evenly from the cranial tip of the process. Secondly, the minor metacarpal is distinctly grooved in T. tadornoides as in most tadornines including the Miocene Miotadorna from New Zealand (flattened in Hymenolaimus), but is not grooved in the fossil. The fossil differ further from Tadorna in that the alular process is distad of the carpal trochlea and by the presence of a small distinct fossa adjacent to alular process. It lacks the apomorphic state for Hymenolaimus of a very deep cuneiform fossa bound dorsally by an expanded dorsal rim of the carpal trochlea. NTM P.2913 differs from the older fossils tentatively referred to Australotadorna as follows: extensor process longer (44 vs. 37% proximal width); alular process distally offset from distal end of carpal trochlea; depth at proximal synostosis less despite greater proximal width. UCMP 65985: L radius (Fig. 6J), Alcoota, Paine Quarry, V6346, Alcoota Homestead, NT, 22° 52′S, 134° 27′E, collected by Woodburne et al., 1963, Waite Formation, Miocene, c. 8 Mya, Alcoota LF. Measurements – TL =128.6 mm, PW =7.6 mm, DW =8.7 mm. The fossil is reconstructed with damage rebuilt in plaster in its proximal quarter. It differs markedly from Anseranas radii in having a more robust shaft and that the distal end lacks the abrupt widening, both dorsally and ventrally, from the shaft. Also, the depressio ligamentosa (ulnar depression) is centrally located on the ventral surface and bound by a large ligamental prominence ventrally, whereas in Anseranas the ulnar depression is in two parts divided by a median ligamental tuberosity. The fossil is smaller and straighter that anserine radii, but differs little from radii of Tadorna and is the size of a large T. tadornoides. It differs from T. tadornoides by the ulnar depression being deeper and more centrally located on the ventral facies and the sulcus tendinosus (tendinal groove) on the dorsal surface is deeper. Proximally, the capital tuberosity is relatively larger. The carpometacarpus (NTM P.2913) and the radius AUSTRALIAN TERTIARY ANATIDS 433 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 (UCMP 65985) derive from the same local fauna, and are of appropriate size to come from a single species. Pending discovery of more diagnostic elements this taxon is not named. OLIGO-MIOCENE INDETERMINATE ANATID FOSSILS UCMP unreg.: L scapula, indet. small anatid; Leaf Locality, Lake Ngapakaldi, UCMP V-6313, collected by R. A. Stirton, collection code RAS#5274, Wipajiri Formation, Early Miocene, Kutjamarpu LF, measurements – length acromion to ventral side humeral facet =7.7 mm, depth blade =3.5 mm. The specimen has typical anatid form with a relatively short acromion. MV 222431: R manus phalanx II.1, anatid indet., Neville’s Nirvana, Lake Palankarinna, MV site number 2396, I-V5367, Etadunna Formation, Minkana LF, Zone A, measurements – TL =14.9 mm. SAM P.42671: R manus phalanx II.1 lacking distal end, anatid indet., Site 2, Billeroo Creek, 31° 06.205′S; 140° 13.912′E, THW and A. Camens June 2007, Frome Downs Station, South Australia, Namba Formation, Pinpa LF. SAM P. 42680: pL and dL radius, anatid indet., Lake Pinpa, Site 6, 31° 08.289′S; 140° 12.679′E, THW and A. Camens June 2007, Namba Formation, Pinpa LF. AMNH 10935: pL radius, anatid indet., Lake Pinpa, Site C, Namba Formation, Pinpa LF. UCMP 65978: pL carpometacarpus, anatid indeterminate cf. Malacorhynchus membranaceus SAM B.39384, Alcoota, Paine Quarry, UCMP V-6345, collection code RAS 5478, collected by Stirton 1962. QM F52742 (=AR11004): dR femur with lateral condyle and fibular trochlea both eroded caudally and medial condyle eroded cranially, anatid indet., Sticky Beak Site (System A), Riversleigh, Queensland, Late Oligocene – early Miocene. Measurements: DW > 21.2 mm, SW =8.7 mm. This specimen has a typical anatid shape and differs from Anseranas in two main ways: (1) the linea intermuscularis caudalis from the middle of the bone is aligned on the medial side, mesad of the nutrient foramen, and was directed towards the now missing femoral head (aligned laterally and directed towards lateral side in Anseranas); (2) the tuber. M. gastrocnemialis Lateralis is elongate (8 by 3.5 mm), not near circular. That this impression is relatively short and does not have a distinct bend in the proximal half indicates that it is plesiomorphic relative to anserines, exclusive of Cereopsis and Cnemiornis, and all other anatids, where the scar is more elongate to varying degrees and distinctly bent medially. In so far as preserved the specimen differs from Cereopsis, by a shallower popliteal fossa and a broader ridge separating that fossa from the tuber. M. gastrocnemialis lateralis. Little more can be said from this fossil other than that an anatid of similar size and evolutionary grade to Cereopsis existed in the System A faunas at Riversleigh. RESULTS – PHYLOGENETIC ANALYSES PARSIMONY ANALYSES Parsimony analyses were made on the dataset with all characters unordered or with 35 characters ordered, in both instances employing the backbone constraint outlined in the Methods. In the ordered analysis, a strict consensus of the four shortest trees found, length 1269, consistency index (CI) =0.2593, homoplasy index (HI) =0.8125, retention index (RI) =0.5950, is shown in Figure 7. The monophyly of Anatidae was strongly supported (bootstrap 82%, Table 4 clade A). The basal Table 4. Significant clades shown in Fig. 7, and the unambiguous apomorphies (UA) defining them. Only characters with a consistency index >0.5 are listed Clade Taxon Number UA Character, State change (x to y), CI A Anatidae 21 4(0–1), 0.750; 26(0–1), 1.00; 30(0–1), 0.571; 87(0–1), 1.000; 89(0–1), 0.500; 91(0–1), 1.000; 107(0–1), 0.667; 109(0–1), 0.667; 113(0–1), 1.000; 115(0–1), 1.000; 146(1–2), 1.000. B Anatidae exclusive of anserines 9 None with CI >0.5, but one significant ambiguous apomorphy 119(0–2), 0.500 C ‘oxyurines’ 6 119(2–0), 0.500 The characters are as in Worthy & Lee (2008) and Appendix 1. Abbreviations as in Methods. 434 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Figure 7. The strict consensus tree obtained from the parsimony analysis with 35 characters ordered. Support values above lines at each node show bootstrap >50% and Bayesian credibility values >70% (100% =*). Values below lines are numbers of unambiguous synapomorphies for each node. Clades A, B, and C are referred to in text and Table 4. AUSTRALIAN TERTIARY ANATIDS 435 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 ephaga, unconstrained in the analyses, is more basal than the true shelducks (tadornines) in this grade, and may even be separated from them by the oxyurine clade, although there is only weak support for this. The mid-Oligocene fossil Telornis impressus, if correctly interpreted as a tadornine by Agnolin (2004), may give support for this basal position for Chloephaga. Within the clade of shelducks, Tadorna is paraphyletic, with T. radjah pairing with Alopochen. In parsimony analyses, Miotadorna lay within a clade of shelducks basal to Tadorna, but in the Bayesian analysis it separated from them, but without significant support. The significant amount of missing data for this taxon undoubtedly reduces support for the tadornine clades. Above the shelducks, the unconstrained Chenonetta jubata paired with Che. finschi supporting the generic synonymy of these taxa (Worthy & Olson, 2002), contra (Livezey, 1989, 1997a). The unconstrained position of Hymenolaimus as the most derived taxon in this grade of tadornines does not support Livezey’s (1997a, b) conclusion, that this aberrant form is a tadornine. These data in contrast suggest that both Chenonetta and Hymenolaimus are outside of Tadorninae and would be better classed as basal anatines. DIVING AND THE OXYURINE LINEAGE Several features of the skeleton of P tedfordi indicate that it was a specialized diver, e.g. nonpneumatic and distally narrowing humerus and a flattened tibiotarsus. That the basal members of the oxyurine clade, Stictonetta,Mionetta, Malacorhynchus, are/were not specialist divers, suggests diving evolved just once within the clade. The fossil taxa indicate that this had happened by the latest Oligocene – Early Miocene. The slightly younger Man. lacustrina and Dunstanetta from New Zealand were specialist divers (Worthy & Lee, 2008). In Europe, the smaller waterfowl from the latest Oligocene to the Middle Miocene were dominated by Mionetta blanchardi (Cheneval, 1983; Bochenski, 1997; Mlíkovský, 2002, 2003; Mourer-Chauviré et al., 2004). This species lacked specialist diving apomorphies in its legs and is a basal oxyurine in these analyses. Other waterfowl taxa co-existing with Mionetta blanchardi are not nearly so well characterized; however, a size-equivalent of Pinpanetta fromensis and Manuherikia minuta is present, i.e. Mionetta natator (Milne-Edwards). A larger form, Mionetta consobrina (Milne-Edwards), is doubtfully distinct and may refer to larger specimens of Mi. blanchardi (Livezey & Martin, 1988), but neither have specialized diving features. Therefore diving, characteristic of the most deeply nested taxa in the oxyurines, evolved independently of other diving anatids, e.g. the aythyines and mergines, necessarily just once within the lineage. A decrease in the number of skeletal elements that are pneumatic within more specialized divers is a well-documented feature within anseriforms (O’Connor, 2004). However, the present results indicate that the nonpneumatic ventral tricipital fossa of the humerus evolved long before diving adaptations in oxyurines, and thus question whether it is functionally correlated with diving. Stictonetta, which is a specialized filter feeder in shallow water (Marchant & Higgins, 1990), not a specialized diver, is consistently found here to be the most basal oxyurine. It has a pneumatic humerus and also differs from other oxyurines in having reticulated tarsi like anserines and dendrocygnines. The fossil taxon Mi. blanchardi is the next most derived oxyurine above Stictonetta and its humerus is nonpneumatic, but its leg bones lack strong adaptations towards diving, such as dorsoventrally curved femora, flattened tibiotarsus, or shortened tarsometatarsus with trochlea II markedly proximally located. Similarly, Malacorhynchus has a nonpneumatic humerus, and is not a diver but a specialized surface dabbler. It is only in oxyurines more derived than these taxa that diving becomes prevalent and so a nonpneumatic ventral pneumotricipital fossa which characterizes Oligo-Miocene oxyurines may have predisposed this group towards a diving habit. In support of this suggestion are the results of a recent analysis of Romainvillia by Mayr (2008) that reveals this taxon to be the most basal anatid, and its tricipital fossa was not pneumatic. In a study of pneumaticity of the post-cranial skeleton of anseriforms, O’Connor (2004) determined that specialized diving had to have evolved multiple times when he assumed Livezey’s (1997a) phylogeny. Similarly, evolution of diving at least five times is inferred from the sequence Callaghan & Harshman (2005) adopted. They placed Thalassornis as a discrete tribe within Dendrocygninae, Biziura as a discrete lineage, Oxyura as a lineage basal to tadornines, and mergines and aythyines as separate deeply nested anatine groups. In contrast to the sequence advocated by Livezey (1997a) or Callaghan & Harshman (2005), the phylogenetic relationships found here (Fig. 7) reduces the number of times specialized diving necessarily evolved to just two: one in oxyurines and one in a clade of anatines that included both aythyines and mergines. COMMENT ON CORRELATION OF LOCAL FAUNAS The distribution of all three Pinpanetta species supports the previous broad biocorrelation of the Pinpa LF from the Namba Formation with the Minkina LF (Zone A), Ditjimanka LF (Zone B) and 442 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Ngama LF (Zone D) of the Etadunna Formation (Woodburne et al., 1994). The occurrence of the phoenicopteriform Palaelodus wilsoni throughout the Lake Palankarinna sequence and at Lake Pinpa (Baird & Vickers-Rich, 1998) further supports these correlations. No other bird groups are sufficiently well known to add further data to these faunal comparisons. The recovery of Pi. tedfordi in Ringtail Site, from System C, at Riversleigh, if borne out by further specimens, presents a marked extension in geographic range of this taxon and also perhaps in time. System C sites, which equate to Faunal Zone C sites (e.g. Travouillon et al., 2006), are considered to be Middle Miocene in age (Archer et al., 1997) and slightly older than the Bullock Creek LF. The fossil Pinpanetta bone has several implications: Firstly, if its larger size is paralleled in further specimens by other morphological differences, then the Ringtail form would be a distinct species. Alternatively, if the presence of Pi. tedfordi is confirmed by further material in Ringtail Site, it would corroborate other faunal data that suggest an affinity with Faunal Zone B sites (Travouillon et al., 2006). Ringtail Site is characterized by a number of aquatic taxa, especially numerous lungfish and turtles, and the platypus Obdurodon, which latter taxon is also present in System B sites (Archer et al., 2006; Travouillon et al., 2006). This aquatic component may mean the Ringtail fauna was deposited coeval with the limestone formation, whereas most other ‘System C’ sites on Gag Plateau are cave infill deposits which would necessitate their being somewhat younger. If so, the age of the Ringtail fauna could be closer to those of Faunal Zone B sites, which are intermediate between those of Faunal Zones A and C (Archer et al., 1997). Sites from Faunal Zone A have a taxon found in the Ngama LF, uppermost in the Etadunna sequence and about 23 Mya, and six taxa otherwise known in the Kutjamarpu LF from sediments overlying the Etadunna at Lake Ngapakaldi (Archer et al., 1997; Travouillon et al., 2006). Some Faunal Zone C sites share taxa, e.g. Neohelos sp., with the Middle Miocene Bullock Creek LF, suggesting similar ages for their faunas. The Bullock Creek LF has no anatids to shed any light on this issue. Elsewhere in Australasia the fossil record does not support longevity of anatid species. The anatid fauna of the Early Miocene (19– 16 Mya) of New Zealand contains at least five genera and six species, none of which persist into the Recent and, globally, no modern genera are unequivocally known earlier than the Late Miocene (Worthy et al., 2007). This suggests that turnover in anatids is relatively rapid compared to other birds, as genera such as Aegotheles and Collocalia,Menura,Cacatua, and Orthonyx did persist from the Early–Middle Miocene to the present (Boles, 1993a, b, 1995b, 2001; Worthy et al., 2007). It therefore seems unlikely that a species of anatid would have persisted from 23–15 Mya. Together these observations suggest the age of Ringtail Site is likely to be closer to those of Faunal Zone B sites, perhaps about 20–18 Mya, and thus much closer in age to the Etadunna local faunas. If so the presence of an anatid species from, or very similar to one from, the Etadunna formation would be less surprising. ACKNOWLEDGEMENTS I thank the following curators and collection managers for enabling access to fossil specimens in their care: Walter Boles, AM; Carl Mehling, Fossil Amphibian, Reptile, and Bird Collections, Division of Paleontology, AMNH; Alan Tennyson and Gillian Stone, NMNZ; David Pickering and Tom Rich, Palaeontology Dept., MV; Dennis Rice, SAM; Dirk Megirian, Museums & Art Galleries of the Northern Territory, Darwin; Rod Wells, Flinders University of South Australia; Kristen Spring and Heather Janetzki, QM; Cécile Mourer-Chauviré, Université Claude Bernard, Lyon 1, Centre des Sciences de la Terre, Villeurbanne, Cedex, France. The following collection managers and curators helped greatly by loaning reference material for this study: notably Philippa Horton, SAM who not only provided ready access to the collections in her care but facilitated the loans; Rory O’Brien, Vertebrates, MV; Robert Palmer, ANWC; Alan Tennyson, Gillian Stone, and Sandy J.A. Bartle, NMNZ; Storrs Olson, James Dean, and Christopher Milensky, USNM; Joanne Cooper, BMNH; Paul Scofield, CM. Joanne Cooper and Paul Scofield each went out of their way to help by scoring data for a rare specimen in their care. I thank both Neville Pledge (SAM) and Tom Rich (MV) for providing data on fossil sites some of the specimens were derived from. The study was significantly enriched by the fieldwork to Lake Pinpa and Billeroo Creek, which a grant from the Sir Mark Mitchell Foundation to Aaron Camens and THW. I am pleased to acknowledge the role Alec Wilson has had, who as owner of Frome Downs Station has provided access to the sites to palaeontologists over many years. I thank the University of Adelaide for supporting this fieldwork with a suitable expedition vehicle, and Aaron Camens and Rebecca Candy for willing help searching out bird fossils. I particularly thank Rod Wells for introducing me to the Frome Basin sites and sharing data from the earlier expeditions, and Richard Tedford for answering questions about site data from previous expeditions. Michael S. Y. Lee was particularly helpful with his guidance in the phylogenetic analyses and comments on drafts of the text. Similarly, I thank both Neville Pledge and Walter AUSTRALIAN TERTIARY ANATIDS 443 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Boles for comments that improved drafts of the text, and Mike Archer for discussions re. Riversleigh deposits. REFERENCES Agnolin FL. 2004. Revisión sistemática de algunas aves deseadenses (Oligoceno Medio) descriptas por Ameghino en 1899. Revista del Museo Argentino de Ciencias Naturales, n.s 6: 239–244. Alfaro MES, Zoller S, Lutzoni F. 2003. Bayes or bootstrap? A simulation study comparing the performance of Bayesian Markov chain Monte Carlo sampling and bootstrapping in assessing phylogenetic confidence. Molecular Biology and Evolution 20: 255–266. Alley NF. 1998. Cainozoic stratigraphy, palaeoenvironments and geological evolution of the Lake Eyre Basin. Palaeogeography, Palaeoclimatology, Palaeoecology 144: 239–263. Archer M, Arena DA, Bassarova M, Beck RMD, Black K, Boles WE, Brewer P, Cooke BN, Crosby K, Gillespie A, Godthelp H, Hand SJ, Kear BP, Louys J, Morrell A, Muirhead J, Roberts K, Scanlon JD, Travouillon KJ, Wroe S. 2006. Current status of species-level representation in faunas from selected fossil localities in the Riversleigh World Heritage Area, northwestern Queensland. Alcheringa Special Issue 1: 1–17. Archer M, Arena R, Bassarova M, Black K, Bramwell J, Cooke B, Creaser P, Crosby K, Gillespie A, Godthelp H, Gott M, Hand SJ, Kear B, Krikmann A, Mackness B, Muirhead J, Musser A, Myers T, Pledge N, Wang Y, Wroe S. 1999. The evolutionary history and diversity of Australian mammals. Australian Mammalogy 21: 1– 45. Archer M, Hand SJ, Godthelp H, Creaser P. 1997. Correlation of the Cainozoic sediments of the Riversleigh World Heritage Fossil Property, Queensland, Australia. In: Aguilar J-P, Legendre S, Michaux J, eds. Actes du Congrès BiochroM’97, Mémoires et Travaux de l’Institut de Montpellier de l’Ecole Pratique des Hautes Etudes 21: 131–152. Aslanova SM, Burczak-Abramowicz NI. 1968. A fossil swan from the Maykopian of Azerbaydzhan. Acta Zoologica Cracoviensia 13: 325–338, pls XI–XIV. Baird RF, Vickers-Rich P. 1998. Palaelodus (Aves: Palaelodidae from the Middle to Late Cainozoic of Australia). Alcheringa 22: 135–151. Baumel JJ, Witmer LM. 1993. Osteologia. In: Baumel JJ, King AS, Breazile JE, Evans HE, Vanden Berge JC, eds. Handbook of avian anatomy: Nomina anatomica avium, 2nd edn. Publications of the Nuttall ornithological club, Vol. 23. Cambridge, Massachusetts, USA, 45–132, figs 4.1–4.18. Bochenski Z. 1997. List of European fossil bird species. Acta Zoologica Cracoviensia 40: 293–333. Boles WE. 1992. Revision of Dromaius gidju Patterson and Rich, 1987, with a reassessment of its generic position. In: Campbell KE, Jr, ed. Papers in avian paleontology honouring Pierce Brodkorb. Natural History Museum Los Angeles County Sciences Series 36: 195–208. Boles WE. 1993a. A new cockatoo (Psittaciformes: Cacatuidae) from the Tertiary of Riversleigh, northwestern Queensland, and an evaluation of rostral characters in the systematics of parrots. Ibis 135: 8–18. Boles WE. 1993b. A logrunner Orthonyx (Passeriformes: Orthonychidae) from the Miocene of Riversleigh, northwestern Queensland. Emu 93: 44–49. Boles WE. 1993c. Pengana robertbolesi, a peculiar bird of prey from the Tertiary of Riversleigh, northwestern Queensland, Australia. Alcheringa 17: 19–25. Boles WE. 1995a. The world’s oldest songbird (Aves: Passeriformes). Nature 374: 21–22. Boles WE. 1995b. A preliminary analysis of the Passeriformes from Riversleigh, Northern Queensland, Australia, with the description of a new species of Lyrebird. Courier Forschungsinstitut Senckenberg 181: 163–170. Boles WE. 1997a. Fossil songbirds (Passeriformes) from the Early Eocene of Australia. Emu 97: 43–50. Boles WE. 1997b. A kingfisher (Halcyonidae) from the Miocene of Riversleigh, Northwestern Queensland, with comments on the evolution of kingfishers in Australo-Papua. Memoirs of the Queensland Museum 41: 229–234. Boles WE. 1997c. Riversleigh birds as palaeoenvironmental indicators. Memoirs of the Queensland Museum 41: 241– 246. Boles WE. 1997d. Hindlimb proportions and locomotion of Emuarius gigju (Patterson & Rich, 1987) (Aves: Casuariidae). Memoirs of the Queensland Museum 41: 235–240. Boles WE. 1998. A budgerigar Melopsittacus undulatus from the Pliocene of Riversleigh, northwestern Queensland. Emu 98: 32–35. Boles WE. 1999. A new songbird (Aves: Passeriformes: Oriolidae) from the Miocene of Riversleigh, northwestern Queensland. Alcheringa 23: 51–56. Boles WE. 2001. A swiftlet (Apodidae: Collocaliini) from the Oligo-Miocene of Riversleigh, northwestern Queensland. Memoir of the Association of Australasian Palaeontologists 25: 45–52. Boles WE. 2005a. A review of the Australian fossil storks of the genus Ciconia (Aves: Ciconiidae), with a description of a new species. Records of the Australian Museum 57: 165– 178. Boles WE. 2005b. A new flightless gallinule (Aves: Rallidae: Gallinula) from the Oligo-Miocene of Riversleigh, northwestern Queensland, Australia. Records of the Australian Museum 57: 179–190. Boles WE. 2005c. Fossil honeyeaters (Meliphagidae) from the Late Tertiary of Riversleigh, north-western Queensland. Emu 105: 21–26. Boles WE, Ivison TJ. 1999. A new genus of the dwarf megapode (Galliformes: Megapodiidae) from the Late Oligocene of Central Australia. Proceedings of the 4th International meeting of the Society of Avian Paleontology and Evolution. Washington, USA, June 1996. Smithsonian Contributions to Paleobiology 89: 199–206. Boles WE, Mackness B. 1994. Birds from the bluff downs local fauna, allingham formation, Queensland. Records of the South Australian Museum 27: 139–149. Brodkorb P. 1964. Catalogue of fossil birds. Part 2 (Anseri444 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 formes through Galliformes). Bulletin of the Florida State Museum Biological Sciences 8: 195–335. Brush AH. 1976. Waterfowl feather proteins: analysis of use in taxonomic studies. Journal of Zoology, London 179: 467–498. Callaghan D, Harshman J. 2005. Taxonomy and systematics. In: Kear J, ed. Ducks, geese and swans. 2 vols, Chapter 2. Oxford: Oxford University Press, 14–26. Callen RA, Alley NF, Greenwood DR. 1995. Lake eyre basin. In: Drexel JF, Preiss WV, eds. The geology of South Australia. Vol. 2, The Phanerozoic. South Australian Geological Survey. Bulletin 54, 188–194. Callen RA, Tedford RH. 1976. New late Cainozoic rock units and depositional environments, Lake Frome area, South Australia. Transactions of the Royal Society of South Australia 100: 125–167. Cheneval J. 1983. Les Anatidae (Aves, Anseriformes) du gisement Aquitanien de Saint-Gérand-le-Puy (Allier, France). In: Buffetaut E, Mazin JM, Salmon E, eds. Actes du Symposium Paléontologique Georges Cuvier, Montbéliard, France, 1982. Montbéliard: Ville de Montbéliard edit, 85–97. Cheneval J. 1984. Les oiseaux aquatiques (Gaviiformes à Anseriformes) du gisement aquitanien de Saint-Gérand-lePuy (Allier, France). Révision Systematique. Palaeovertebrata 14: 33–115. Cheneval J. 1987. Les Anatidae (Aves, Anseriformes) du Miocene de France. Révision Systématique et évolution. In: Mourer-Chauviré C, ed. L’Évolution des Oiseaux d’Après le Témoignage des Fossils. Documents du Laboratoire Géologiqe de Lyon, Vol. 99. Villeurbanne, Cedex: Departement des Sciences de la Terre, Université Claude-Bernard, 137–157. Christidis L, Boles WE. 1994. The taxonomy and species of birds of Australia and its territories. Hawthorn East, Melbourne: Royal Australasian Ornithologist’s Union, Monograph 2. Clarke JA, Tambussi CP, Noriega JI, Erickson GM, Ketchman RA. 2005. Definitive fossil evidence for the extant avian radiation in the Cretaceous. Nature 433: 305– 308. Crowe TM, Bowie RCK, Bloomer P, Mandiwana TG, Hedderson TAJ, Randi E, Pereira SL, Wakeling J. 2006. Phylogenetics, biogeography and classification of, and character evolution in, gamebirds (Aves: Galliformes): effects of character exclusion, data partitioning and missing data. Cladistics 22: 495–532. Cummings MP, Handley SA, Myers DS, Reed DL, Rokas A, Winka K. 2003. Comparing bootstrap and posterior probability values in a four-taxon case. Systematic Biology 52: 477–487. De Vis CW. 1888. A glimpse of the post-Tertiary avifauna of Queensland. Proceedings of the Linnean Society of New South Wales, Series 2, 3: 1277–1292. De Vis CW. 1889. Additions to the list of fossil birds. Proceedings of the Royal Society of Queensland 6: 55–58. De Vis CW. 1905. A contribution to the knowledge of the extinct avifauna of Australia. Annals of the Queensland Museum 6: 3–25. Delacour J, Mayr E. 1945. The family Anatidae. Wilson Bulletin 57: 3–55. Dickinson EC (ed). 2003. The Howard and Moore complete checklist of the birds of the world, 3rd edn. London: Christophe Helmpp. Donne-Goussé C, Laudet V, Hänni C. 2002. A molecular phylogeny of Anseriformes based on mitochondrial DNA analysis. Molecular phylogenetics and evolution 23: 339– 356. Dyke GJ. 2001. The fossil waterfowl (Aves: Anseriformes) from the Eocene of England. American Museum Novitates 3354: 1–15. Ericson PGP. 1997. Systematic relationships of the Palaeogene family Presbyornithidae (Aves: Anseriformes). Zoological Journal of the Linnean Society 121: 429–483. Ericson PGP. 1999. New material of Juncitarsus (Phoenicopteriformes), with a guide for differentiating that genus from the Presbyornithidae (Anseriformes). In: Olson SL. ed. Avian paleontology at the close of the 20th century. Proceedings of the 4th International Meeting of the Society of Avian Paleontology and Evolution Washington, D. C., 4–7 June 1996. Smithsonian Contributions to Paleobiology 89: 245– 251. Ericson PGP. 2000. Systematic revision, skeletal anatomy, and paleoecology of the New World early tertiary Presbyornithidae (Aves: Anseriformes). PaleoBios 20: 1–23. Eyton TC. 1838. A monograph on the Anatidae-or duck tribe. London: Longman, Orme, Brown, Green and Longman. Frith HJ. 1964. Taxonomic relationships of Stictonetta naevosa (Gould). Nature 202: 1352–1353. Frith HJ. 1977. Waterfowl in Australia. Sydney: A.H. & A.W. Reed. Harrison CJO, Walker CA. 1979. Birds of the British Lower Oligocene. Tertiary Research Special papers 5: 29–39. Howard H. 1929. The avifauna of Emeryville Shellmound. University of California Publications in Zoology 32: 301– 394. Howard H. 1955. A new wading bird from the Eocene of Patagonia. American Museum Novitates 1710: 1–25. Howard H. 1964. Fossil anseriformes. In: Delacour J, ed. The waterfowl of the world. London: Country Life Ltd, 233–326. Johnsgard PA. 1961. The taxonomy of Anatidae – a behavioural analysis. Ibis 103: 71–85. Johnsgard PA. 1968. Waterfowl: Their biology and natural history. Vol. 138. Lincoln, Nebraska: University of Nebraska Press. Xviii +pp. Johnsgard PA. 1978. Ducks, geese and swans of the world. Lincoln and London: University of Nebraska Press. Kear J. (ed). 2005. Ducks, geese and swans. 2 vols. Oxford: Oxford University Press. Krieg GW, Callen RA, Gravestock DI, Gatehouse CG. 1990. Geology. In: Tyler MJ, Twidale CR, Davies M, Wells CB, eds. Natural history of the North East Deserts. Occasional Publications of the Royal Society of South Australia 6, 1–26. Kurochkin EN. 1968. New Oligocene birds from Kazakhstan. Paleontological Journal (a translation of Paleontologicheski Zhurnal) 2: 83–90. AUSTRALIAN TERTIARY ANATIDS 445 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Lambrecht K. 1933. Handbuch der Palaeornithologie. Berlin: Gebrüder, Borntraeger. Livezey BC. 1986. A phylogenetic analysis of Recent anseriform genera using morphological characters. Auk 105: 681– 698. Livezey BC. 1989. Phylogenetic relationships of several subfossil anseriformes of New Zealand. Occasional papers of the Museum of Natural History of the University of Kansas 128: 1–25. Livezey BC. 1995a. Phylogeny and comparative ecology of the stiff-tailed ducks (Anatidae: Oxyurini). Wilson Bulletin 107: 214–234. Livezey BC. 1995b. A phylogenetic analysis of the whistling and white-backed ducks (Anatidae: Dendrocygninae) using morphological characters. Annals of the Carnegie Museum 64: 65–97. Livezey BC. 1996a. A phylogenetic analysis of geese and swans (Anseriformes: Anserinae), including selected fossil species. Systematic Biology 45: 415–450. Livezey BC. 1996b. A phylogenetic reassessment of the tadornine-anatine divergence (Aves: Anseriformes: Anatidae). Annals of the Carnegie Museum 65: 27– 88. Livezey BC. 1997a. A phylogenetic classification of waterfowl (Aves: Anseriformes), including selected fossil species. Annals of the Carnegie Museum 67: 457–496. Livezey BC. 1997b. A phylogenetic analysis of modern shelducks and sheldgeese (Anatidae: Tadornini). Ibis 139: 51–66. Livezey BC. 1997c. A phylogenetic analysis of basal Anseriformes, the fossil Presbyornis, and the interordinal relationships of waterfowl. Zoological Journal of the Linnean Society 121: 361–428. Livezey BC, Martin LD. 1988. The systematic position of the Miocene anatid Anas [?] blanchardi Milne Edwards. Journal of Vertebrate Paleontology 8: 196–211. Livezey BC, Zusi RL. 2007. Higher-order phylogeny of modern birds (Theropoda, Aves: Neornithes) based on comparative anatomy. II. Analysis and discussion. Zoological Journal of the Linnean Society 149: 1–95. Louchart A, Vignaud P, Likius A, Mackaye HT, Brunet M. 2005. A new swan (Aves: Anatidae) in Africa, from the latest Miocene of Chad and Libya. Journal of Vertebrate Paleontology 25: 384–392. Madsen CS, McHugh KP, de Kloet SWR. 1988. A partial classification of the waterfowl (Anatidae) based on singlecopy DNA. The Auk 105: 452–459. Marchant S, Higgins PJ (co-ordinators). 1990. Handbook of Australian, New Zealand & Antarctic birds, Vol. 1. Ratites to Ducks. Melbourne: Oxford University Press. Martin HA. 1990. The palynology of the Namba Formation in the Wooltana-1 bore, Callabonna Basin (Lake Frome), South Australia, and its relevance to Miocene grasslands in central Australia. Alcheringa 14: 247–255. Mayr G. 2005. The Paleogene fossil record of birds in Europe. Biological Review 80: 515–542. Mayr G. 2008. Phylogenetic affinities and morphology of the late Eocene anseriform bird Romainvillia stehlini Lebedinsky, 1927. Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen 248: 365–380. McCracken KG, Harshman J, McClellan DA, Afton AD. 1999. Data set incongruence and correlated character evolution: an example of functional convergence in the hind-limbs of stiff-tail ducks. Systematic Biology 48: 683– 714. McCracken KG, Sorenson MD. 2005. Is homoplasy or lineage sorting the source of incongruent mtDNA and nuclear gene trees in the stiff-tailed ducks (NomonyxOxyura). Systematic Biology 54: 35–55. Megirian D, Murray P, Schwartz L, Von Der Borch C. 2004. Late Oligocene Kangaroo Well Local Fauna from the Ulta Limestone (new name), and climate of the Miocene oscillation across central Australia. Australian Journal of Earth Sciences 51: 701–741. Miller AH. 1963. The fossil flamingos of Australia. The Condor 65: 289–299. Mlíkovský J. 2002. Cenozoic birds of the world, part 1: Europe. Praha: Ninox Presspp. Mlíkovský J. 2003. Middle Miocene birds of Františkovy Lázne˘, Bohemia. Casopsis Národniho muzea, Rada prirodove˘dná (J. Nat. Mus. Nat. Hist Ser.) 172: 109– 113. Mlíkovský J, Švec P. 1986. Review of the Tertiary waterfowl (Aves: Anseridae) of Asia. Vestnik Cestoslovenske Spolecnosti Zoologicke 50: 259–272. Mourer-Chauviré C, Berthet D, Hugueney M. 2004. The Late Oligocene birds of the Créchy Quarry (Allier, France), with a description of two new genera (Aves: Pelecaniformes: Phalacrocoracidae, and Anseriformes: Anseranatidae). Senkenbergiana lethaea 84: 303–315. Murray P, Megirian D. 1992. Continuity and contrast in Middle and Late Miocene vertebrate communities from the Northern Territory. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 9: 195–218. Murray PF, Vickers-Rich P. 2004. Magnificent mihirungs, the colossal flightless birds of the Australian dreamtime. Bloomington and Indianapolis: Indiana University Press. Olson SL. 1977. The identity of the fossil ducks described from Australia by C. W. De Vis. The Emu 77: 127–131. Olson SL. 1985. The fossil record of birds. In: Farner DS, King JR, Parkes KC, eds. Avian biology, Vol. 8. New York: Academic Press, 79–238. Olson SL. 1999. The anseriform relationships of Anatalavis Olson and Parris (Anseranatidae), with a new species from the Lower Eocene London Clay. In: Olson SL. ed. Avian paleontology at the close of the 20th century. Proceedings of the 4th International Meeting of the Society of Avian Paleontology and Evolution Washington, D. C., 4–7 June 1996. Smithsonian Contributions to Paleobiology 89: 231– 243. Olson SL, Feduccia A. 1980. Presbyornis and the origin of the Anseriformes (Aves: Charadriomorphae). Smithsonian Contributions to Zoology 323: 1–24. O’Connor PM. 2004. Pulmonary pneumaticity in the postcranial skeleton of extant Aves: a case study examining Anseriformes. Journal of Morphology 261: 141–161. 446 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Patton JC, Avise JC. 1986. Evolutionary genetics of birds IV. Rates of protein divergence in waterfowl (Anatidae). Genetica 68: 129–143. Peters JL. 1931. Check-list of birds of the world, Vol. 1. Cambridge: Harvard University Press. Phillips JC. 1923. Natural history of the ducks, Vol. 1, Plectropterinae, Dendrocygninae, Anatinae. London: Longmans, Green and Company. Pledge NS. 1984. A new Miocene vertebrate faunal assemblage from the Lake Eyre Basin. A preliminary Report. Australian Zoologist 21: 345–355. Pledge NS, Tedford RH. 1990. Vertebrate fossils. In: Tyler MJ, Twidale CR, Davies M, Wells CB, eds. Natural history of the North East Deserts.Occasional Publications of the Royal Society of South Australia 6: 199–209. Raikow RJ. 1971. The osteology and taxonomic position of the white-backed duck, Thalassornis leuconotus.Wilson Bulletin 83: 270–277. Rambaut A, Drummond AJ. 2004. TRACER version 1.3. Available at: http://beast.bio.ed.ac.uk/Tracer Rich PV. 1979. The Dromornithidae, an extinct family of large ground birds endemic to Australia. Bureau of Mineral Resources, Geology and Geophysics, Australia. Bulletin 184: 1–196. Rich PV, van Tets GF. 1981. The fossil pelicans of Australasia. Records of the South Australian Museum 18: 235– 264. Rich P, van Tets J. [sic] 1982. Fossil birds of Australia and New Guinea: their biogeographic, phylogenetic and biostratigraphic input. In: Rich PV, Thompson EM, eds. The fossil vertebrate record of Australasia. Clayton, Victoria: Monash University Offset printing Unit, 235–384. Rich PV, van Tets GF, Rich THV, McEvey AR. 1987. The Pliocene and Quaternary flamingos of Australia. Memoirs of the Queensland Museum 25: 207–225. Rich TH, Archer M, Hand SJ, Godthelp H, Muirhead J, Pledge NS, Flannery TF, Woodburne MO, Case JA, Tedford RH, Turnbull WD, Lundelius EL, Rich LSV, Whitelaw MJ, Kemp A, Rich PV. 1991. Australian Mesozoic and tertiary terrestrial mammal localities. In: Rich PV, Monaghan JM, Baird RF, Rich TH, eds. Vertebrate palaeontology of Australasia. Melbourne, Pioneer Design Studio and Monash University Publications Committee, 1437, 1005–1070. Ronquist F, Huelsenbeck JP. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574. Sharpe RB. 1899. A hand-list of the genera and species of birds, Vol. 1. London: British Museum (Natural History). Shufeldt RW. 1913. On comparative osteology of Cereopsis novaehollandiae.Emu 12: 209–237. Shufeldt RW. 1914. Contributions to the study of the ‘treeducks’ of the genus Dendrocygna.Zoologische Jahrbücher (Systematik, Geographie und Biologie der Tiere) 38: 1–70. Shufeldt RW. 1915. Fossil birds in the Marsh Collection of Yale University. Transactions of the Connecticut Academy of Arts and Sciences 19: 1–109. Sibley CG, Ahlquist JE. 1990. Phylogeny and classification of birds: a study in molecular evolution. New Haven, CT: Yale University Press. Sibley CG, Ahlquist JE, Monroe BL. 1988. A classification of the living birds of the world based on DNA-DNA hybridization studies. The Auk 105: 409–423. Sorenson MD, Cooper A, Paxinos EE, Quinn TW, James HF, Olson SL, Fleischer RC. 1999. Relationships of the extinct moa-nalos, flightless Hawaiian waterfowl, based on ancient DNA. Proceedings of the Royal Society of London Series B 266: 2187–2193. Sraml M, Christidis L, Easteal S, Horn P, Collet C. 1996. Molecular relationships within Australasian waterfowl (Anseriformes). Australian Journal of Zoology 44: 47–58. Stirton RA, Tedford RH, Miller AH. 1961. Cenozoic stratigraphy and vertebrate paleontology of the Tirari Desert, South Australia. Records South Australia Museum 14: 19–61. Swofford DL. 2000. PAUP*: Phylogenetic analysis Using Parsimony (* and other methods). Version 4.0b10. Computer program and documentation. Sunderland, MA: Sinauer Associates. Tedford RH, Archer M, Bartholomai A, Plane M, Pledge NS, Rich T, Rich P, Wells RT. 1977. The discovery of Miocene vertebrates, Lake Frome area, South Australia. BMR Journal of Australian Geology & Geophysics 2: 53–57. Tedford RH, Wells RT. 1990. Pleistocene deposits and fossil vertebrates from the ‘Dead Heart of Australia’. Memoirs of the Queensland Museum 28: 263–284. Tedford RH, Wells RT, Barghoorn SF. 1992. Tirari Formation and contained faunas, Pliocene of the Lake Eyre Basin, South Australia. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 9: 173–194. Tedford RH, Williams D, Wells RT. 1986. Late Cenozoic sediments and fossil vertebrates. In: Wells RT, Callen RA, eds. The Lake Eyre Basin – Cainozoic sediments, fossil vertebrates and plants, landforms, silcretes and climatic implications. Australasian Sedimentologists Group Field Guide Series No. 4. Geological Society of Australasia, Sydney 177: 42–72. Travouillon K, Archer M, Hand SJ, Godthelp H. 2006. Multivariate analyses of Cenozoic mammalian faunas from Riversleigh, northwestern Queensland. Alcheringa Special Issue 1: 323–349. Turbott EG. (Convener) 1990. Checklist of the birds of New Zealand and the Ross Dependency, Antarctica, 3rd edn. Auckland: Random Century, Ornithological Society of New Zealand, Inc. Vickers-Rich P. 1991. The Mesozoic and Tertiary history of birds on the Australian plate. In: Rich PV, Monaghan JM, Baird RF, Rich TH, eds. Vertebrate palaeontology of Australasia. Melbourne: Pioneer Design Studio and Monash University Publications Committee, 721–808. Weir JT, Schluter D. 2008. Calibrating the avian molecular clock. Molecular Ecology 17: 2321–2328. Wiens JJ, Bonett RM, Chippindale PT. 2005. Ontogeny discombobulates phylogeny: paedomorphosis and higherlevel salamander relationships. Systematic Biology 54: 91– 110. AUSTRALIAN TERTIARY ANATIDS 447 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 Woodburne MO, MacFadden BJ, Case JA, Springer MS, Pledge NS, Power JD, Woodburne JM, Springer KB. 1994. Land mammal biostratigraphy and magnetostratigraphy of the Etadunna Formation (Late Oligocene) of South Australia. Journal of Vertebrate Paleontology 13: 483– 515. Woodburne MO, Tedford RH, Archer M, Turnbull WD, Plane N, Lundelius EL. 1985. Biochronology of the continental mammal record of Australia and New Guinea. South Australia Department of Mines and Energy, Special Publication 5: 347–364. Woolfenden GE. 1961. Postcranial osteology of the waterfowl. Bulletin of the Florida State Museum (Biol. Sci.) 6: 1–129. Worthy TH. 2005. A new species of Oxyura (Aves: Anatidae) from the New Zealand Holocene. Longmann Symposium. Memoirs of the Queensland Museum 51: 255–272. Worthy TH, Holdaway RN, Sorenson MD, Cooper AC. 1997. Description of the first complete skeleton of the extinct New Zealand goose Cnemiornis calcitrans Owen, (Aves: Anatidae), and a reassessment of the relationships of Cnemiornis.Journal of Zoology, London 243: 695–723. Worthy TH, Lee MSY. 2008. Affinities of Miocene (19–16 Ma) waterfowl (Anatidae: Manuherikia, Dunstanetta and Miotadorna) from the St Bathans Fauna, New Zealand. Palaeontology 51: 677–708. Worthy TH, Olson SL. 2002. Relationships, adaptations, and habits of the extinct duck ‘Euryanas’finschi.Notornis 49: 1–17. Worthy TH, Tennyson AJD, Jones C, McNamara JA. Douglas BJ. 2007. Miocene waterfowl and other birds from Central Otago, New Zealand. Journal of Systematic Palaeontology 5: 1–39. doi:1017/S1477201906001957. APPENDIX 1 LIST OF CHARACTERS USED IN THE PHYLOGENETIC ANALYSES Characters 1 to 133 are as given in Worthy & Lee (2008), except that character 63 was revised as follows: 63. Distal end, relative length proc. flexorius (=entepicondyle): 0, short, ends proximad to condylus dorsalis and condylus ventralis (in caudal view, a line across distal extreme of dorsal condyle at right angles to long axis of shaft passes well distad of entepicondyle, e.g. Anseranas); 1, long, distal extent roughly equal to that of the dorsal condyle. Note: in most taxa, the entepicondyle, and the dorsal and ventral condyles have approximately equivalent distal extent. In Anseranas,Branta, and Cygnus, the dorsal and ventral condyles have equal distal extent and the entepicondyle is markedly shorter, so coded ‘0’. In Cereopsis and Anser, the dorsal condyle and entepicondyle have about equal length, and the ventral condyle extends slightly distad of a line drawn across them, but here distal extent relative to the dorsal condyle is prioritized and they are coded ‘1’. New characters 134. Humerus, dorsal pneumotricipital fossa excavated below head: 0, no; 1, yes. 135. Humerus, incisura capitis opens to dorsal pneumotricipital fossa: 0, at equal height; 1, groove elevated above (more caudal) the fossa and often separated from the fossa by a distinct ridge that transverses the groove. 136. Humerus, external tuberosity shape: 0, width roughly equals length; 1, elongate ovate. 137. Humerus, width of space between facet for anterior articular ligament and proximoventral apex of the dorsal condyle: 0, narrow, gap equal to or narrower than width facet; 1, wide, gap wider than facet. Derived from Woolfenden (1961). 138. Humerus, attachment of M. scapulohumeralis cranialis, or supraspinatus (Howard, 1929): 0, poorly defined short and wide; 1, elongate but ending distally well short of junction of bicipital crest and shaft; 2, elongate and robust, often elevated, ending distally level with or distad of junction of bicipital crest and shaft. Note. This attachment scar straddles the median crest, i.e. the ridge between the dorsal and ventral pneumotricipital fossae, so when the dorsal fossa is large and wide as in some divers, e.g. Clangula, the ridge is larger and extends distally further. In this character care needs to be taken to differentiate the scar for the supraspinatus from this ridge. 139. Humerus, fossa olecrani (olecranal fossa): 0, shallow; 1, deep, well defined. All Tadorna species were coded ‘0’. 140. Humerus, fossa pneumotricipitalis ventralis, lamina around the distoventral ventral margin from the caudal facies: 0, more than 1/2 occludes the fossa; 1, partially occludes the fossa and remains elevated off the base of the fossa and extends up under the ventral tubercle e.g. Anser; 2, extends into fossa where it merges with base of the fossa in ventral 1/2 of fossa; 3, extends into fossa and merges with base of fossa in dorsal 1/2 fossa; 4, incomparable as fossa closed/nonpneumatic. Pelvis 141. Pelvis, relative length preacetabular region of synsacrum: 0, long, such that length to distal side of the costal process of the most caudal vertebrae sacrales >40% length; 1, short, equivalent length <40% synsacrum. 448 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 142. Fenestrae intertransversariae: 0, absent; 1, only in distal 1/2 of synsacrum caudad of acetabulum; 2, present in synsacrum over all length caudad of acetabulum. 143. Canalis iliosynsacralis: 0, absent, ilia fused dorsally entirely; 1, paired openings directed caudally. 144. Corpus ischii: 0, with no foramen opening laterally below the antitrochanter at its junction with the acetabulum; 1, with pneumatic foramen. Character 121, Worthy et al. (1997). 145. Antitrochanter: 0, with pneumatic openings medially to fossa renalis, or posteriorly into foramen ilioischiadicum; 1, None. 146. Recessus caudalis fossae: 0, deep, e.g. Gallus;1, shallow and pneumatic; 2, absent. Character 120, Livezey (1986). 147. Foramen ilioischiadicum: 0, very short, much <1/2 length ischium from foramen acetabulum; 1, about 1/2 length ischium from acetabular foramen; 2, long, >1/2 length ischium from acetabular foramen. 148. Tuberculum preacetabulare: 0, very prominent with notch between it and ilium, e.g. Gallus; 1, less prominent, with no notch above it. 149. Pubis: 0, dorsally concave over length; 1, straight or flat. Character 115, Livezey (1986). 150. Pubis, section distad of articulation with distal ischium with flattened caudal expansion, typically rounded in lateral view and with diameter significantly greater than area immediately craniad of it: 0, not so; 1, yes, e.g. Cygnus. Character 117, Livezey (1986). APPENDIX 2 DATA MATRIX USED IN THE PHYLOGENETIC ANALYSES Gallus_gallus0–101000100100 –-0–011–-00(01)00322–- –112111–303–01021–-2 1210022–-100000000(12) 01002000102000–00000 00?001(12)0–1000000000 001000100200010002(01) 01101010001010100100 Anhima_cornuta00100–0200000 0–-0–0010–0100000010 00–10000000000020000 00000020200100001000 220–0–01011000000000 00001211000010001000 10020000123000020001 00010000100010100 Anseranas00000–01–000010 –00100001200000(12)000 0–000000000000000000 00000000000000010000 00100000000100000000 0000(01)01100000000000 00000011004100011001 00010000000010100 Thalassornis00110001?11101 00002100011110321?01 0–?01103031112010110 001122101001100000?1 ?100020?00131–211111 001–1?11112121101011 100?01?0??1??0200000 0002140200121010 Dendrocygna_arborea10110001310 10102012010012100311 1000–111103010112010 11010000002101100000 10111000000011310101 10000101011001121000 01110(01)01100221010?0 001000100??????????? Dendrocygna_autumnalis100100012 10(12)0102012010012110 (01)1(12)0000–101103010 1120101101000000(02) 1 0 11000001011100010001 13101011000012101100 2121(01)0001110(01)0111 0221010?32010001002? ????????? Dendrocygna_bicolor10010001210 10101011010012100211 0000–101103010112010 11020000000101100000 00101000100011310101 0000012(01)1(01)1001121 01001110101110221010 ?0001000000202101211 00 Dendrocygna_eytoni00011001240 10102011010012100112 0000–111(01)0301011201 01101000100210110000 00010100010001131010 1000001011(01)10011210 10011100011101310001 31010001001011012100 0 Dendrocygna_arcuata00010001240 10101012(01)1001211011 20000–11110301011201 011020001000101100(01) 00001110000011113101 AUSTRALIAN TERTIARY ANATIDS 449 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 01100011211110011210 00011100112102310101 00010000002021012100 0 Oxyura_vittata0011100132220 10000201(01)0221113210 0112211103010110111– 21101121101102001011 111000001100231–2111 11111–11110121221110 1110020200310010?041 01111214?????????? Oxyura_jamaicensis00110001322 2010000001(01)02211132 10011220110302011111 02212011221011021011 111110100010(01)0131–2 11111111–11211121221 (01)1011101202003(12) 0 0 10404101001214100012 2000 Oxyura_australis001100013222 010000001(01)012110322 00112201103020111110 22120112210110210101 11110000110(01)0131–21 1111111–11211121221(01) 10111002020031101040 41011112141000121100 Oxyura_maccoa00110101322211 0000101(01)01211032200 11221110301011111022 12011221011021000110 11110010000231–21111 1111–111111212210101 110020200210010?0410 1101214?????????? Nomonyx_dominicus002100012122 01000110100221103220 0011201103010110111– 21201022001112100011 0110(12)0010000131–211 101111–1111012122101 01110020200210010?04 10111110??????????? Stictonetta00210101222111 00000010112110222000 11201103031112110210 20002101101201101101 00000212212311101000 00121121001122110011 10220211231011200000 011003120(01)120101 Biziura_lobata0010000221020 10101100101111032110 00–10110300001001011 0101121021(01)00102012 10110000110(01)121–211 121110–1131112022101 01110(01)2020032(01)210 50010010120411001220 10 Malacorhynchus0121100132221 1000000110111112(12) 1 0 1011201(01)03031011110 21020112201101200111 10100200211202310201 001102211(23)100212210 00111101021021101110 01011012141210122100 Nettapus_pulchellus00211001342 10102001101010111222 02011211103021011011 –1121102201111201201 101001000122(01)231020 11101022112100212210 10111002022011101020 01000112(01)3110012211 1 N._coromandelianus00011001342 10102001101010111222 02011211103031011111 –1121102201111200(12) 0 11010(01)00(01)01221231 12011100022112100212 210(01)011(01)0(01)202201 11010200100011213120 01211?? Cereopsis020121121421010 2100010010110112100(01) 01010122301100100000 000000010110000010(01) 00001(01)0311131010100 01112001100111110001 1102001112(34)10102200 100010010101020001 Cnemiornis_calcitrans0001001204 11010211001001011031 11000–100(01)–20–-–101 –-–0000000–010000000 0–000000000101110101 01000102100110001111 001111022?001–30010? ????0001000120102210 0 Branta_canadensis001121111411 0102022111010110(01) (01) 10101111100221011(01) 0 00110100000020111000 0110000001(12)0311(12) 3 101010001111102100(12) 12210001111200110240 450 T. H. WORTHY © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021 110(34)000000010011200 022001 Anser_brachyrhynchus0001000104 11010202211001011011 10101010100221011000 00101000010211110000 11010000021311231010 10001111103100112210 00111121011024011043 (02)000001001120112200 1 Anser_caerulescens00010(01)0104 11010202211011011021 11101111100321011000 00101000000211110000 01000000001300231010 11001110003100112210 00111120011123111033 2000001001120(01)02200 0 Anser_rossii00012101041101 02011111010110211010 10111003110110000010 10000002111100000101 00000013101310101100 11201121001122100011 11200101231010?32000 0010011200122101 Cygnus_atratus000101120411 010202210101211001 110011101102210010 010010100000020002 000011000001000311 111010100010021011 002122100011102001 011511103020000010 011110121101 Tadorna_ferriginea00010001241 21101011111112110222 00011201103011110110 11020001002111201100 10100000203102310101 00000120131002122100 11111202110210110220 1000111021210121100 Tadorna_tadornoides00010001141 21102010111112110221 00011201103010110110 11010001102111200101 10100000203102310101 00000121131001122100 01111112110221110120 1000111021210122100 Tadorna_tadorna0001201114121 10101111111211022100 01110110301011011011 01000100211120010110 10000021320231010100 00012111100112210001 11101210022111022010 00111021210121110 Tadorna_variegata000100011412 11020101111121102220 00112011030101101101 10100011021112001011 01000012132123101010 00001(12)1121001122100 011112021002(12)111012 01000111021210122100 Tadorna_radjah0021200114121 10201111111211111200 01121110201001011021 01100110211120010110 10000020321231010100 00012111100112210001 11000211021111022110 00112011210120000 Alopochen000121023412110 20121110101102120000 –1011031101100102101 0001(01)02111200101101 00001203211310101000 0010(01)11100112210001 11120211021111023110 10000021210120000 Chloephaga_hybrida001(01)201114 02010202001102?11011 20000–21110301011001 012010000?0211120000 11010000121311231010 10000212111100112110 00111000210021111021 01001100021210021101 Chloephaga_poliocephala001(01)2011 1402010101201102?111 21201010211103211010 111–1010000?02111200 10110100001213112310 10100002110111001122 10001110202100231100 21010001100212101210 01 Somateria_mollisima00113011130 21102011111022110322 10010201003010010110 21121112102111201101 1011100010211231–211 111001–1111112122100 1111102221031011020(03) 1011111141210122010 Lophodytes_cucullatus0011000121 12011101101101211022 AUSTRALIAN TERTIARY ANATIDS 451 © 2009 The Linnean Society of London, Zoological Journal of the Linnean Society, 2009, 156, 411–454 Downloaded from https://academic.oup.com/zoolinnean/article/156/2/411/2732013 by guest on 31 August 2021