PELECANIDAE, Rafinesque, 1815

Smith, ND & Ksepka, DT, 2015, Five well-supported fossil calibrations within the " Waterbird " assemblage (Tetrapoda, Aves), Palaeontologia Electronica 18 (1), pp. 1-21 : 7-9

STEM PELECANIDAE

Pelicans

Node Calibrated. This node represents the split between Pelecanidae (monotypic family comprised of eight extant species of Pelecanus ) and their extant sister taxon, which has been variable among molecular ( Ericson et al., 2006; Hackett et al., 2008) and morphological ( Livezey and Zusi, 2007; Smith, 2010; Mayr, 2011b) estimates of phylogeny. Fossil Taxon. Pelecanus sp.

Specimen. NT-LBR-039 (holotype; FSL-367.087 cast). A nearly complete skull and mandible, and several articulated cervical vertebrae.

Phylogenetic Justification. Phylogenetic justification is based on three cranial apomorphic characters of Pelecanidae that are preserved in the specimen ( Louchart et al., 2011). These include a long, spatulate rostrum with two ridges on the ventral surface subparallel to the edges, long and thin mandibular rami, and a rostrally located syndesmotic intraramal hinge in the mandible. None of these characters are present in any other bird species, thus they will optimize as synapomorphies uniting NT-LBR-039 and extant pelicans regardless of which avian group constitutes the extant sister taxon of Pelecanidae .

Minimum Age. 28.1 Ma

Soft Maximum Age. None specified.

Age Justification. The specimen was collected from Pichovet, 3 km northeast of Vachéres, Luberon, in southeastern France (43º55'N, 5º40'E) ( Louchart et al., 2011). The specimen was found in fine-grained limestone, indicative of a coastal freshwater lagoon depositional environment ( Louchart et al., 2011). The deposits from Pichovet have been biostratigraphically correlated with the Mammal Paleogene biostratigraphic zone MP 24, which is within the Rupelian ( Mourer-Chauviré, 1985; Legendre and Lévêque, 1997; Mayr, 2006). The temporal age range for the Rupelian has been inferred to be 28.1–33.9 Ma (BiochroM, 1997; Gradstein et al., 2012; Cohen et al., 2013; Walker et al., 2013). Following best practices for justifying minimum age constraints ( Parham et al., 2012), we apply the youngest possible age for the Rupelian, inclusive of error: 28.1 Ma.

Phylogenetic position of Pelecanidae and the Oligocene Pelecanus sp. Traditional taxonomy and phylogenetic studies based on morphological data ( Cracraft, 1985; Livezey and Zusi, 2007 and references therein) have suggested the Pelecanidae are sister-taxon to Suloidea ( Sulidae , Anhingidae , Phalacrocoracidae ), with this larger group constituting the sister taxon to Fregatidae . This group of "pelecaniforms" exclusive of tropicbirds ( Phaethontidae ) is known as Steganopodes ( Chandler, 1916). The study of Smith (2010) also recovered a monophyletic Steganopodes, but with Fregatidae as the sister taxon to Suloidea and Pelecanidae as the successive sister taxon to this clade. However, Mayr (2011b) recovered a Pelecanidae + Balaenicipitidae clade that is sister taxon to Scopidae , with this larger group in a basal polytomy with other clades of Ciconiiformes and Pelecaniformes . Mayr's (2011b) result is more similar to the topologies that have long been recovered by molecular studies. Van Tuinen et al. (2001) provided some of the first molecular support for a Pelecanus + Balaeniceps + Scopus clade, and subsequent studies with more taxon-sampling and diverse genetic data have largely confirmed this grouping ( Fain and Houde, 2004; Ericson et al., 2006; Hackett et al., 2008). Thus, the relative utility of the Oligocene Pelecanus sp. (NT-LBR-039) as a fossil calibration is highly dependent on the topology of the phylogeny being temporally calibrated (e.g., it would represent the oldest record of a Pelecanus + Balaeniceps + Scopus clade, but not the oldest record of a monophyletic Steganopodes).

Fossil record of total group Pelecanidae . Protopelicanus cuvieri , represented by an isolated femur from the late Eocene Paris Gypsum of France, has previously been considered to share affinities with pelicans, sulids, and pelagornithids, but as Mayr (2009a, p. 80) notes, the affinities of this material should be considered indeterminate until more complete remains and/or apomorphy-based diagnoses can be advanced. Beyond NT- LBR-039, the next oldest fossil stem Pelecanidae is Miopelecanus gracilis from the early Miocene of France, which is represented by a cranium and caudal portion of the rostrum ( Olson, 1985). Louchart et al. (2011) suggested that this taxon might be referable to Pelecanus . However, regardless of this taxonomic uncertainty, there is currently no evidence to support the phylogenetic placement of either NT-LBR-039 or Miopelecanus gracilis within crown pelicans (e.g., crown Pelecanidae /crown Pelecanus ). Other Neogene and Quaternary records of total group Pelecanidae are limited, but members of the clade are known from all continents, with the exception of Antarctica (Stidham et al., 2014).

Based on the topologies of most higher-level avian phylogenies based on molecular data (e.g., Ericson et al., 2006; Hackett et al., 2008), NT-LBR-039 would calibrate splits within a Pelecanus + Balaeniceps + Scopus clade. The oldest member of total group Balaenicipitidae is Goliathia andrewsi from the early Oligocene lower sequence of the Jebel Qatrani Formation of Egypt ( Brodkorb, 1980; Rasmussen et al., 1987). The holotype of Goliathia andrewsi is a complete ulna (NHMUK A883), and a distal end of a right tarsometatarsus lacking the trochleae from the upper sequence of the Jebal Qatrani Formation has also been referred to this species ( Rasmussen et al., 1987; Smith, 2013, figure 3). Scopus xenopus , represented by a distal tarsometatarsus and partial coracoid from the early Pliocene of South Africa, constitutes the only known fossil record of total group Scopidae ( Olson, 1984) . Neither Goliathia andrewsi nor Scopus xenopus have been evaluated in a phylogenetic analysis.

Based on the topologies recovered from most morphological studies, NT-LBR-039 would either calibrate the split between Pelecanidae and a Fregatidae + Suloidea clade (Smith, 2010); or between Pelecanidae and Suloidea ( Cracraft, 1985; Livezey and Zusi, 2007). In the former case, NT-LBR-039 would be superseded as a fossil calibration by the older stem-frigatebird Limnofregata (Smith, 2010; see below). In the latter cases, there are several potentially older members of total group Suloidea that might supersede NT-LBR-039. These include (but are not limited to): 1) Masillastega rectirostris from the middle Eocene of Messel, an isolated skull that may be a stem member of Sulidae ( Mayr, 2002b) ; 2) Eostega lebedinskyi from the middle Eocene of Romania, an incomplete mandible that may also be a stem sulid and/or a senior synonym of Masillastega ( Mlíkovsky, 2002, 2007; Mayr, 2009a); 3) Phocavis maritimus , represented by a tarsometatarsus from the late Eocene to early Oligocene Keasey Formation of northwest Oregon ( Goedert, 1988) and a member of the extinct Plotopteridae , which may constitute the sister taxon to a Phalacrocoracidae + Anhingidae clade (Smith, 2010); and 4) the enigmatic Protoplotus beauforti , a nearly complete skeleton from lacustrine sediments of Sumatra (likely Paleocene in age; Whateley and Jordon, 1989), that has been allied with various members of the Pelecaniformes (van Tets et al., 1989; Smith, 2010). With the exception of Protoplotus , most of these specimens are represented by fragmentary or incomplete material; and in all cases, their relationships among extant groups of waterbirds have remained controversial, and have not been tested in phylogenetic analyses.