Anthropoides

Abstract: ABS?RACr.--DNA sequences spanning 1,042 nucleotide bases of the mitochondrial cytochrome-b gene are reported for all 15 species and selected subspecies of cranes and an outgroup, the Limpkin (Aramus guarauna). Levels of sequence divergence coincide approximately with current taxonomic ranks at the subspecies, species, and subfamilial level, but not at the generic level within Gruinae. In particular, the two putative species of Balearica (B. pavonina and B. regulorum) are as distinct as most pairs of gruine species. Phylogenetic analysis of the sequences produced results that are strikingly congruent with previous DNADNA hybridization and behavior studies. Among gruine cranes, five major lineages are identified. Two of these comprise single species (Grus leucogeranus, G. canadensis), while the others are species groups: Anthropoides and Bugeranus; G. antigone, G. rubicunda, and G. vipio; and G. grus, G. monachus, G. nigricollis, G. americana, and G. japonensis. Within the latter group, G. monachus and G. nigricollis are sister species, and G. japonensis appears to be the sister group to the other four species. The data provide no resolution of branching order for major groups, but suggest a rapid evolutionary diversification of these lineages. Received 19 March 1993, accepted 19 August 1993. THE 15 EXTANT SPECIES of cranes comprise the nominate family (Gruidae) of the order Gruiformes, and are currently divided into two subfamilies, Balearicinae and Gruinae (Brodkorb 1967). Balearicine cranes are anatomically unspecialized relative to gruines and are represented by only two extant species in the genus Balearica (the crowned cranes of Africa). Gruines share derived anatomical features such as an anteriorly sculpted sternum (often associated with tracheal coiling inside keel) in which the furcular process is fused to the anteroventral tip of the keel. Three extant gruine genera are recognized: Grus (10 species), Anthropoides (2 species), and Bugeranus (1 species). These genera are defined on the basis of soft anatomical features, although their monophyly has not been addressed by phylogenetic analysis. Fossil balearicines are known from the lower Eocene and later deposits in Eurasia, whereas Gruines date from the late Miocene (Brodkorb 1967). Evolutionary relationships among cranes have been addressed with a variety of different approaches during the past two decades. Archibald (1976) derived the species groups shown in Table 1 on the basis of similarities in unison

Abstract: -I used a DNA-DNA hybridization method to generate more than 1,200 pairwise comparisons among species of the family Gruidae (cranes) and an outgroup Limpkin (Aramidae). A matrix of genetic distances included average delta T,, values for all cells, and reciprocals among cranes, based on 3-10 replicate experiments per cell. I chose delta T,, as the appropriate dissimilarity measure because virtually all homologous single-copy DNA sequences in crane genomes were sufficiently similar to form hybrid duplexes under standard experimental conditions. The normalized percent hybridization (NPH) values approached 100 for all crane species pairs. The delta T, data departed slightly from metricity as a result of experimental variation associated with the measurement of small genetic distances. The DNA data, analyzed with a best-fit tree approach and checked for internal consistency by jackknifing, support the traditional view that crowned cranes (Balearica) are the most ancient lineage of extant gruids. The enigmatic Siberian Crane (Grus leucogeranus) appears as the sister group of the remaining species, which fall into four closely related groups. Bugeranus and Anthropoides are sister groups. The three species of Australasian Grus (antigone, rubicunda, and vipio) form a clade, as do five predominantly Palearctic Grus (grus, monachus, nigricollis, japonensis, and americana). The Sandhill Crane (G. canadensis), while clearly a member of the gruine clade, is an old lineage without close relatives. Received 21 March 1989, accepted 18 May 1989. THE CLASSIFICATIONS of Peters (1934) and Wetmore (1934) established the traditional familylevel distinction of cranes (Gruidae) within the Order Gruiformes. This arrangement, though not universally accepted (Cracraft 1973), has persisted in most recent revisions (e.g. Wetmore 1960, Storer 1971, Sibley et al. 1988). Opinions regarding species affinities within the Gruidae have varied greatly. Following Peters (1934), most workers have accepted the existence of 14 extant species in four genera, though Walkinshaw's (1964) designation of a 15th species has gained many adherents. The crowned cranes (Balearica) are commonly placed in the Balearicinae, apart from the remaining species (Gruinae; Brodkorb 1967) based on their lack of sternal excavation and tracheal convolution. Balearicines, apparently the more ancient lineage, are abundantly represented by Tertiary fossil materials from western Eurasia (Brodkorb 1967). Gruine fossils appear in the Miocene of Europe, but are best represented in North American Pliocene deposits and by a scattering of Pleistocene remains worldwide (Johnsgard 1983). Cracraft (1973) inferred from these ma1 Current address: Museum Support Center, Smithsonian Institution, Washington, DC 20560 USA. terials that cranes diverged from a common ancestor with limpkins (Aramidae) in the late Paleocene. Archibald (1975, 1976) performed the first comprehensive analysis of cranes based on a coherent set of characters. His study of crane unison calls (stereotyped behavior patterns involved in pair-bonding) led him to identify clusters of similar species (Fig. 1). Archibald's work verified the distinctness of crowned cranes (Balearica) and suggested an unexpected relationship between Bugeranus carunculatus (Wattled Crane) and Grus leucogeranus (Siberian Crane). Archibald included leucogeranus as a member of Bugeranus, and recommended "Species Group" status for the remaining Grus and Anthropoides clusters. Wood (1979), like Archibald, found a high level of similarity between Wattled and Siberian cranes, which, aside from complicated multivariate resemblances, lack the exaggerated tracheal convolutions present in other gruines. Ingold et al. (1987) attempted to resolve crane relationships with allele-frequency analysis. Their species arrangement was a major departure from traditional views, although Balearica appears distinct from gruines. Their work involved small sample sizes, and only approxi603 The Auk 106: 603-618. October 1989 This content downloaded from 157.55.39.28 on Thu, 15 Dec 2016 04:53:04 UTC All use subject to http://about.jstor.org/terms 604 CAREY KRAlEwsKI [Auk, Vol. 106