Columbea

The 46 Orders

Paleognathae

Galloanserae

Columbea

Otidae

Gruae

Ardeae

Telluraves

Afroaves

Australaves

NEOAVES Sibley, Ahlquist, & Monroe, 2008

Neoaves is the sister clade to Galloanserae. Jarvis et al. (2014) show that it has two parts: Columbea and Passerea.

COLUMBEA Jarvis et al., 2014

Columbea contains about 3½% of the avian tree. The sister relationship between Columbimorphae and Mirandornithes in Jarvis et al. (2014) is not so suprising as they have been included in Metaves since it was proposed by Fain and Houde (2004).

Jarvis et al.'s results make it clear that the original Metaves hypothesis does not hold up in larger datasets. It was in fact an artifact of using the β-fibrogen gene, particularly the 7th intron. Nonetheless, some of the signal was real, and Columbea remains once the Eurypygimorphae, Strisores, and Hoatzin have been moved elsewhere.

Mirandornithes Sangster, 2005

Van Tuinen et al. (2001) seem to have been the first to recognize that flamingos and grebes were sister taxa. Their reanalysis of Sibley and Ahlquist (1990) shows that other data to support this association were available earlier, but that the association had been missed by Sibley and Ahlquist, who did not include both of them in any of their individual trees. They did not suspect flamingos and grebes could be closely related.

Later analysis — Chubb (2004), Cracraft et al. (2004), Ericson et al. (2006a), Brown et al. (2008), Hackett (2008), Gibb et al. (2013), McCormack et al. (2013), and Jarvis et al. (2014) — have also supported Mirandornithes. Further, Mayr (2004) identified morphological and oological evidence in favor of Mirandornithes. See also Mayr (2007, 2008). As with the genetic data, previous morphological investigations also did not consider the possibility of a relationship here. It was just too inconceivable.

It's optional whether to lump the grebes and flamingos into one order, or treat them as two. I've switched to the conventional treatment with two orders. The large morphological differences and ancient split from the flamingos (e.g., Jarvis et al., 2014; Brown et al., 2008) support this.

PHOENICOPTERIFORMES Fürbringer, 1888

Phoenicopteridae: Flamingos Bonaparte, 1831

Phoenicopteridae tree

The genetics of flamingos were studied by Torres et al. (2014). They found that all 6 flamingos shared a recent common ancestor, perhaps 4-6 million years ago. They also recommended sinking Phoeniconaias into Phoenicoparrus, as they considered the distinction between them “arbitrary and uninformative”. Based on the fossil record, Torres et al. (2014) also suggested that the grebe-flamingo split occurred around the beginning of the Oligocene, i.e., about 33 million years ago. Mayr (2014a), using morphological evidence, found that the Eocene Juncitarus is a stem representative of Mirandornithes. As such, it gives little information about the timing of the flamingo-grebe, although Mayr suggested middle Eocene. Jarvis et al. (2014) date the split at about 56 mya, roughly at the beginning of the Eocene. I have to wonder whether the crown group may also not be as recent as Torres et al. think, especially since they suggest that their divergence dates among flamingos fit better with an even more recent split between grebes and flamingos. This suggests to me that the older mitochondrial dates are more likely to be correct, and that the true date is probably toward (or past!) the long end of the range. The SACC has decided to keep the current genera, and I follow them here.

3 genera, 6 species HBW-1

PODICIPEDIFORMES Fürbringer, 1888

Podicipedidae: Grebes Bonaparte, 1831

6 genera, 24 species HBW-1

There's no complete molecular phylogeny of the grebes. However, Fjeldså (2004) provides a morphological phylogeny. Christidis and Boles (2008) have also adopted Fjeldså's phylogeny. For the present, I also use it with some modification based on Ogawa et al. (2015).

Ogawa et al. (2015) found that Rollandia is embedded in Podiceps, so it has been merged into Podiceps. The arrangement within the expanded Podiceps is based on Ogawa et al. (2015).

Bochenski's (1994) osteological study of the grebes found that the Great Grebe is significantly different from the other Podiceps grebes. He created the genus Podicephorus for it. Fjeldså endorsed this view by putting Podicephorus in an unresolved trichotomy with Aechmophorus and the remaining Podiceps. The SACC has yet to consider the issue.

Tricolored Grebe, Tachybaptus tricolor, has been split from Little Grebe, Tachybaptus ruficollis. The subspecies vulcanorum is included in Tricolored Grebe. See Mlíkovský (2010).

Based on Ogawa et al. (2015), the Eared Grebe, Podiceps californicus, has been split from the Black-necked Grebe, Podiceps nigricollis. Ogawa et al. did not find a clear separation between Silvery Grebe, Podiceps occipitalis, and Junin Grebe, Podiceps taczanowskii. Further study is needed to clarify the situation.

Podicipedidae tree

Columbimorphae Latham, 1790

MESITORNITHIFORMES Wetmore 1960

Sharpe had earlier tried to establish such a suborder (Mesitides), but based it on the preoccupied genus name Mesites. I haven't been able to find uses of Mesitornis as a type genus earlier than Wetmore.

Mesitornithidae: Mesites Wetmore, 1960 (1850)

2 genera, 3 species HBW-3

PTEROCLIFORMES Huxley, 1868

Huxley (1868) first separated the sandgrouse as an order, using the name Pteroclomorphae, which modernizes to Pterocliformes.

I follow Jarvis et al. (2014) where the mesites are sister to the sandgrouse, with the combination sister to the doves and pigeons. Hackett et al. (2008) also considered the three groups a clade, but with a different arrangement. Jarvis et al. also estimated that the divisions between the three clades date back to the early Paleocene.

Pteroclidae: Sandgrouse Bonaparte, 1831

2 genera, 16 species HBW-4

Click for genus-level Columbidae tree
Click for species-level
Pteroclidae tree

There has been some controversy about how to spell the family name. Although TiF has used Pteroclidae for a long time, back in 2010 both Pteroclidae (Clements, HBW, Sibley-Monroe) and Pteroclididae (AOU, BLI, Howard-Moore, IOC) were in general use, and even Pterocleidae had also been used. Since then, Pteroclidae has won the day. The name indicates it is known for its wing, i.e., “-cles” takes the same meaning as in names such as Heracles. By analogy with Heraclidae/Heracleidae, it would then appear that either Pteroclidae or Pterocleidae would be correct. Most importantly, the first is the form used by Bonaparte when he established the family-group name in 1831 (as the subfamily Pteroclinae), and is used here.

The arrangement here is based on Cohen (2011). There's a bit of ambiguity in Cohen's results. According to the nuclear DNA, the Pin-tailed Sandgrouse, Pterocles alchata, is the basal species. However, mitochondrial DNA puts it sister to Burchell's Sandgrouse, Calopterocles burchelli, albeit with mediocre support. Because of that ambiguity, I've put Burchell's Sandgrouse in its own genus, Calopterocles (Roberts, 1922). Further, Pterocles is now restricted to the Pin-tailed Sandgrouse. Most of the wedge-tailed sandgrouse, with a long 10th primary have been of the former Pterocles species have been moved to Syrrhaptes, which is buried deep inside that clade. The other former Pterocles have been separated as Nyctiperdix (Roberts, 1922, type bicinctus).

COLUMBIFORMES Latham, 1790

Columbidae: Doves, Pigeons Leach, 1820

47 genera, 341 species HBW-4

Click for genus-level Columbidae tree
Click for genus-level
Columbidae tree

I have based the organization of the Columbidae on the paper by Pereira et al. (2007), which provides a comprehensive DNA-based phylogentic tree. Although the dodos and Rodriguez Solitaire (genera Raphus and Pezophaps) have been traditionally considered a separate family in the Columbiformes, the DNA says otherwise. Shapiro et al. (2002) and Pereira et al. (2007) found that these two genera are buried deeply within the Columbidae, in the Raphinae. The Reunion Solitaire seems to have actually been an ibis! See Mourer-Chauviré et al. (1995).

The DNA testing shows three major clades. A basal clade includes a subclade consisting of the New World genera Geotrygon, Leptotila, Zenaida, together with a subclade of New World pigeons (including the Passenger Pigeon), as well as typical pigeons, cuckoo-doves, and turtle-doves. It is sister to the other two clades together. One of them consists of the New World Ground-Doves. The other contains all other doves. I rank the three major clades as subfamilies: Columbinae, Claravinae, and Raphinae.

The name Peristerinae is sometimes used for Claravinae. However, this is incorrect as Peristerinae is based on the genus Peristera (Swainson 1827), which is a junior homonym of the mollusc genus Peristera (Rafinesque 1815). Thus Peristerinae is not available. The genus Peristera (Swainson 1827) was replaced by Claravis (Oberholser 1899). Todd (not Richmond) established the subfamily Claravinae in 1917, which he misspelled as Claraviinae (Dickinson and Raty, 2015).

The arrangement of Claravinae is based on Sweet and Johnson (2015). As a result, the Purple-winged Ground-Dove, Claravis geoffroyi and Maroon-chested Ground-Dove, Claravis mondetoura have been moved to Metriopelia.

The name Raphidae has often been used for a family containing dodos and solitaires. However, the dodos and solitaires turn out to be nested well within our subfamily Raphinae, and even deeper in the family Columbidae.

The subfamily Columbinae has been studied in more detail by Johnson and Clayton (2000), Johnson et al. (2001), Gonzalez et al. (2009a), and Johnson and Weckstein (2011). Zenaidini and the species in Streptopelia and Columba, but not Patagioenas, have been rearranged accordingly. Although Johnson et al. (2001) argued that Nesoenas should be merged into Streptopelia, Cheke (2005) makes the case for instead moving the Malagasy Turtle-Dove to Nesoenas and distinguishing the Laughing and Spotted Doves in a separate genus. The results of Gonalez et al. (2009a) support this move.

In the case of Nesoenas, both Nesoenas and Homopelia (type picturatus) have equal priority, being both named in the same work of Salvadori 1893. Cheke acts as first reviser, choosing Nesoenas. The Laughing Dove is the type of Stigmatopelia (Sundevall 1873) while the Spotted Dove is the type of Spilopelia (also Sundevall 1873). Cheke attempts to designate Stigmatopelia as the genus, but as John Penhallurick points out Schodde and Mason had previously (1997) chosen Spilopelia, so Spilopelia they become.

The work by Johnson and Weckstein (2011) showed that Geotrygon as usually constituted is paraphyletic with its members belonging to 3 separate clades. The Olive-backed Quail-Dove is sister to Leptotila. Banks et al. (2013) establish the genus Leptotrygon for it. The remaining quail-doves fall into at least two groups. Although Johnson and Weckstein did not include the type of Geotrygon (versicolor), a further analysis by Banks et al. (2013) included it and chrysia. The other “Geotrygon” form a clade sister to Zenaida. Banks et al. (2013) have now established the name Zentrygon for it (type costaricensis).

The English names of White-faced Dove, Turacoena manadensis, and Black Dove, Turacoena modesta, have been changed to White-faced Cuckoo-Dove and Black Cuckoo-Dove to match recent IOC changes. There are also several splits.

The remaining splits are based on Ng et al. (2016).

Jønsson et al. (2011a) and Moyle et al. (2013) studied the Alopecoenas ground-doves and Gallicolumba bleeding-hearts. Jønsson et al. found that Gallicolumba was paraphyletic, and recommended splitting it into Gallicolumba (bleeding-hearts) and Alopecoenas (Australasian ground-doves). Their phylogenetic trees seemed to indicate there was still paraphyly even with the restricted Gallicolumba. The further analysis by Moyle et al. revealed issues with pseudogenes that were causing some confusion. Their results support placing the Crested Pigeon in the genus Ocyphaps. Finally, the three extinct Alopecoenas are most likely closely related to A. sanctaecrucis and A. stairi, and their placement reflects this.

The Purple Quail-Dove / Indigo-crowned Quail-Dove Geotrygon purpurata, has been split from Sapphire Quail-Dove, Geotrygon saphirina based on SACC proposal #566.

Gibb and Penny (2010) investigated the fruit-doves and close relatives. They confirmed Shapiro et al.'s (2002) result that Alectroenas and Drepanoptila are embedded in fruit-dove genus Ptilinopus. Cibois et al. (2014) concured in their more complete analysis of the enlarged Ptilinopus. Moreover, they were able to include most Ptilinopus species. The current arrangement of Ptilinopus is based on their results. They suggested a six genus treatment that retained Alectroenas and Drepanoptila. However, this may mean that Ptilinopus is not monophyletic, and it does not materially solve the problem of Ptilinopus heterogeneity. I do not recommend it at this time.

The Black-banded Fruit-Dove, Ptilinopus alligator, has been split from Banded Fruit-Dove, Ptilinopus cinctus. Given present knowledge, it's fairly arbitrary whether to split or lump these allopatric forms. However, one is in Australia, the other in the Lesser Sundas, and the current tendency is for such forms to be split.

Based on Rheindt et al. (2011a), the Maroon-chinned Fruit-Dove, Ptilinopus subgularis is split into 3 species: Oberholser's Fruit-Dove, Ptilinopus epius, Banggai Fruit-Dove, Ptilinopus subgularis, and Sula Fruit-Dove, Ptilinopus mangoliensis.

Columbinae Leach, 1820

Zenaidini Bonaparte, 1853

Columbini Leach, 1820

Claravinae: American Ground-Doves Todd, 1917

Raphinae: Old World Doves and Pigeons Wetmore, 1930 (1835)

Phabini: Australasian Pigeons and Doves Bonaparte, 1853

Raphini Wetmore, 1930 (1835)

Treronini: Green-Pigeons G.R. Gray, 1840

Turturini G.R. Gray, 1840

Ptilinopini: Fruit-Doves and Imperial-Pigeons Selby, 1835

Previous Page Next Page