PSITTACIFORMES (PARROTS)

Ksepka, DT & Clarke, JA, 2015, Phylogenetically vetted and stratigraphically constrained fossil calibrations within Aves, Palaeontologia Electronica (Miami, Fla.) 18, pp. 1-25 : 14-15

PSITTACIFORMES (PARROTS)

Node Calibrated. Divergence between Psittaciformes and Passeriformes (see below).

Fossil Taxon. Pulchrapollia gracilis ( Dyke and Cooper, 2000) .

Specimen. NHMUKA6207: partial skeleton including rostral tip of beak, quadrate, two vertebrae, partial scapula, partial humerus, complete ulna, partial carpometacarpus, phalanx I-1 and II-1, complete femur, partial tibiotarsus, complete tarsometatarsus, and nine pedal phalanges. A coracoid associated with the holotype is now considered to belong to another taxon ( Mayr, 2001b).

Phylogenetic Justification. Phylogenetic analysis of morphological data ( Dyke and Cooper, 2000; Mayr, 2005c), phylogenetic analysis of combined morphological and molecular data ( Ksepka et al., 2011, Ksepka and Clarke, 2012).

Minimum Age. 53.5 Ma.

Age Justification. NMH A6207 was collected from the Walton Member (Division A2 of King, 1981) of the London Clay Formation at Walton-on-the-Naze, England. The Walton Member correlates to the upper part of Chron C24r and to calcareous nannoplankton zone NP10-11 ( Rhodes et al., 1999). Precise dating of Chron C24r is difficult and estimates vary depending on alternate cyclostratigraphic methods applied to extrapolate the length of the stratigraphic gap between the top of C24r and horizons dated using radiometric means. Because a hard minimum calibration should be conservative, we selected the youngest estimate (inclusive of error) for the top of C24r (53.54+/-0.04) presented among the alternate estimates of Westerhold et al. (2007).

Phylogenetic Position of Psittaciformes . Consensus is building for a sister group relationship between Psittaciformes and Passeriformes (Psittacopasserae of Suh et al., 2011) ( Figure 11 ). The phylogenomic analysis of Hackett et al. (2008) supported this pairing, as did most analyses of a 30 nuclear loci dataset (Wang et al., 2012) and analysis of retroposon insertions ( Suh et al., 2011). A trichotomy including Psittaciformes , Passeriformes , and Falconidae was recovered by the nuclear gene study of Ericson et al. (2006). However, mitochondrial studies have placed Psittaciformes near the base of Neoaves ( Pratt et al., 2009) or as sister taxon to Strigiformes ( Pacheco et al., 2011). Morphological analyses have supported a sister group relationship between Coliiformes and Psittaciformes ( Mayr and Clarke, 2003) , near the base of a "higher land bird" assemblage ( Livezey and Zusi, 2006, 2007), or, with a more limited outgroup sampling, as sister group to Passeriformes ( Ksepka et al., 2011) . Despite any uncertainty over the higher level relationship of Psittaciformes , Pulchrapollia gracilis shares unique characters with extant parrots (Dyke and Cooper, 2001; Mayr, 2002) and remains in the same position relative to crown Psittaciformes even when disparate groups such as Musophagidae , Coliiformes , and Passeriformes are recovered as the extant sister taxon of Psittaciformes ( Dyke and Cooper, 2000, Ksepka et al., 2011; Mayr, 2011). Thus, we consider the placement of Pulchrapollia gracilis as a stem parrot to be secure.

Fossil Record of Pan-Psittaciformes. A substantial record of Eocene stem parrots is now known ( Figure 12 ). Several partial skeletons of unnamed fossil taxa have been reported from the London Clay ( Mayr and Daniels, 1998). These taxa are interpreted as more closely related to crown clade Psittaciformes than Pulchrapollia gracilis ( Mayr, 2002) . Slightly younger but more complete records of stem Psittaciformes are provided by Cyrilavisolsoni and Cyrilavis colburnorum , known from several complete and partial skeletons from the Fossil Butte Member of the Green River Formation (51.66 ± 0.09 Ma: see above discussion concerning the age of the Fossil Butte Member). An even larger collection of skeletons representing multiple species is known from the Messel Formation (~47 Ma) ( Mayr and Peters, 1998). Additional Middle Eocene records have been reported from Geiseltal ( Mayr, 2000a). Late Eocene records of Psittaciformes include isolated elements assigned to the Quercypsittidae ( Mourer-Chauviré, 1992) . Vastanavidae, a group of that may also represent stem parrots, are known from isolated elements from the Early Eocene of India ( Mayr et al., 2010).

Two putative stem parrots from the Early Eocene Fur Formation are contemporary with, or slightly older than, Pulchrapollia gracilis , but do not meet all of the criteria for inclusion as calibration points laid out above. These fossils were described as Mopsitta tanta and Pseudasturidae incertae sedis ( Kristofferson, 2002; Waterhouse et al., 2008). Both taxa are represented solely by a humerus, and the psittaciform affinities of Mopsitta tanta are now considered doubtful given a lack of unambiguous synapomorphies supporting a relationship with parrots and recognition of similarities to the stem ibis Rhynchaeites ( Mayr and Bertelli, 2011) . Regardless of phylogenetic status, it is uncertain whether the age of these fossils is greater than that of Pulchrapollia gracilis . Numerous volcanic ash layers occur in the Fur Formation, and these are designated by a series of numbers, ranging from +140 for the highest ash in the formation to -39 for the lowest. Two radioisotopic dates have been reported from these ashes: a date of 54.04+/-0.14 Ma from layer +19 and a date of 54.52 +/- 0.05 Ma from layer -17 ( Chambers et al., 2003). Unfortunately, the horizon from which these putative fossil parrots were recovered was not reported, so whether these dates provide a minimum age, maximum age, or together bound the age of the fossils cannot be determined. Given their incomplete nature and controversial phylogenetic status, we recommend against utilizing the Fur fossils from divergence studies pending further evidence. However, we note the older of the two radiometrically dated layers is only ~1 Ma older than the hard minimum age of Pulchrapollia gracilis and so the potential increase in calibration age is slight.

A partial mandible from the Late Cretaceous Lance Formation was attributed to a parrot by Stidham (1998), though this record has remained controversial ( Dyke and Mayr, 1999; Stidham, 1999). Following the guidelines above, we do not consider this fossil as a calibration point due to its incomplete nature and the lack of phylogenetic analysis supporting its referral.

Fossil Record of Related Clades. As with Coliiformes , it is unlikely that further work on the higher-level affinities of Psittaciformes will significantly impact the calibration age. The oldest reported record of Passeriformes , the probable sister taxon of Psittaciformes , is also from the Early Eocene ( Boles, 1995). Paleocene fossil records of Passeriformes are otherwise restricted to the Oligocene. However, if the extinct Zygodactylidae represent the closest fossil relatives of Passeriformes as proposed by Mayr (2004b, 2008b), the record of the stem lineage leading to Passeriformes would also include many additional Eocene and Oligocene records from Europe and North America. The oldest records of Coliiformes and Strigiformes are both from the Late Paleocene ( Houde and Olson, 1992; Rich and Bohaska, 1976).

Figures

FIGURE 11. Phylogenetic tree showing position of Pulchrapollia gracilis, illustrating the divergence calibrated by NMH A6207. Placement of Pulchrapollia gracilis as a stem parrot is supported by multiple analyses (see text).

FIGURE 12. Stratigraphic distribution of fossil Pan-Psittaciformes in the Paleogene. The minimum possible age of the calibrating specimen NMH A6207 is presented, and midpoints of age ranges are shown for the ages of other Paleogene parrot specimens. Note that the London Clay Formation also contains multiple associated skeletons and isolated bones in addition to the holotype of Pulchrapollia gracilis that serves as a calibration point.