Rallidae, Rafinesque, 1815

Worthy, Trevor H., 2011, Australlus, a New Genus for Gallinula disneyi (Aves: Rallidae) and a Description of a New Species from Oligo-Miocene Deposits at Riversleigh, Northwestern Queensland, Australia, Records of the Australian Museum 63 (1), pp. 61-77 : 72

publication ID

https://doi.org/ 10.3853/j.0067-1975.63.2011.1563

persistent identifier

https://treatment.plazi.org/id/039F8789-FFFB-F512-2A33-FB7281C74364

treatment provided by

Carolina

scientific name

Rallidae
status

 

Rallidae View in CoL , gen. et sp. indet.

Material. A single specimen, QM F54524 View Materials (= AR19964), represents a third species of rallid in Riversleigh deposits. It is the left articular of a mandible preserving the part caudal to the fenestra rostralis mandibulae, including the base of the surangular and the cotyla, but it is lacking the medial process and the tip of the processus retroarticularis ( Fig. 2 View Fig ) .

Location. Ringtail Site, considered part of Riversleigh’s Faunal Zone C deposits and hence of Middle Miocene age ( Archer et al., 1997, 2006; Creaser, 1997; Travouillon et al., 2006).

Description

It is considerably smaller than A. disneyi (preserved length 9.3 mm, length of cotyla lateralis 3.1 mm, total anterior width across cotyla 3.8 mm). It has the general form of rallids with the following notable features: 1, the dorsal margin of the crista transversus fossae linking the caudal part of the cotyla lateralis and the processus medialis is near level with only a slight concavity at mid-width, as in A. disneyi , but differing from all rallids examined that have a marked notch in this crista; 2, the cotyla medialis is wider than the cotyla lateralis and extends laterally to the surangular, unlike all rails examined, except Amaurornis and A. disneyi (generally rallids have the base of the surangular aligned with the junction of the two cotylae); 3, the fenestra caudalis mandibulae is small, as in A. disneyi , but the feature varies among rails, being similarly small in Amaurornis , Dryolimnas pectoralis , and Rallina , but relatively larger in Gallinula ventralis , Porzana tabuensis and Gallirallus philippensis ; 4, the fenestra caudalis mandibulae is located relatively far anteriorly such that the distance from it to the cotyla lateralis exceeds the length of the cotyla lateralis laterally, versus this distance being equal to or shorter than the cotyla in all modern rallids examined.

Apart from being much smaller than A. disneyi , this specimen differs as follows: 1, the cotyla lateralis is better emarginated and raised anteriorly; 2, whereas the lateral facies forms an inflated crest on which lies the attachment of ancora ligamenti jugomandibularis externus, the facies is concave above this attachment (convex); and 3, the lateral crest extends anteriad below and anterior to the fenestra caudalis mandibulae (ends caudal to fenestra).

Results of phylogenetic analysis of Australlus disneyi

We were able to score 134 characters of the 381 in Livezey’s (1998) dataset (Appendix 1) for A. disneyi . For Australlus gagensis , the 23 characters of the humerus and 11 of the coracoid able to be assessed do not differ from the states in A. disneyi . These taxa differ mainly by size and the other differences listed above are not captured by Livezey’s (1998) characters, so A. gagensis was not included in the phylogenetic analysis. Initial analyses used all taxa in Livezey’s (1998) dataset and in these the relationship of Australlus disneyi was recovered as sister to the clade of Porphyrio and relatives in a topology of a strict consensus tree otherwise very like that obtained by Livezey. However, the many taxa with considerable missing data contributed to a very large number of shortest trees and many taxa grouped in polytomies. In order to better resolve the tree, taxa were successively trimmed until all those listed in Methods were excluded. Analyses also included ordering characters as used by Livezey (1998) but, as the results did not differ significantly, we ultimately chose to present analyses of the modified database that treated all characters as unordered. In all preliminary analyses, the tree topologies we obtained were similar to, if not the same as, that presented by Livezey (1998). For our preferred analysis with 51 ingroup taxa, all characters were treated as unordered and with equal weight. A heuristic search with 1000 replicates found 1176 shortest trees, length = 805, consistency index = 0.5317, homoplasy index = 0.4683 and retention index = 0.8381.

As found by Livezey (1998), separation of the various gruiform families was well corroborated, but resolution within the Rallidae was poorly supported, although the strict consensus tree ( Fig. 7 View Fig ) was relatively well resolved. The Gallirallus - type rails formed a clade sister to gallinules and crakes (e.g., Porzana ), with all other rallids forming more basal branches in a tree with the same topology as found by Livezey (1998). The analysis strongly supports the inclusion of Australlus disneyi (and thus A. gagensis ) in the Rallidae . Furthermore, within the Rallidae , Australlus is part of a clade of all rallids except Himantornis and Gymnocrex . The analysis suggests a weak affinity of Australlus with swamphens ( Porphyrio and relatives), with which it formed a sister group (bootstrap support 54%), but perhaps equally likely is that Australlus forms an independent lineage from the rallid stem of similar grade to Porphyrio . No support was found for the inclusion of Australlus disneyi with true gallinules, which were united in a clade with bootstrap support of 67% and deeply nested in the tree. These data suggest that Australlus should be considered to comprise relatively basal rallids, perhaps with some affinity to swamphens.

Kingdom

Animalia

Phylum

Chordata

Class

Aves

Order

Gruiformes

Family

Rallidae

Darwin Core Archive (for parent article) View in SIBiLS Plain XML RDF