Putting Hypsipetes in the passerine tree

ResearchBlogging.org

One of my long-running plans on Tet Zoo has been to review passerine phylogeny. After decades of people saying that oscine passerines are (except larks and corvids) far too alike for anyone to construct a sensible phylogeny, a flurry of (mostly molecular) studies have meant that - as I like to say - the passerine tree is finally coming together. Alas, I'm nowhere near able to review current hypotheses of passerine evolution in depth (and didn't I say something recently about not starting any new major Tet Zoo projects?).

i-56760579fb18310af126924da37660b8-Hypsipetes-white-vs-black-headed-forms-Feb-2011.jpg

Well done to those of you who correctly identified the 'mystery' bird as an Asian black bulbul or Himalayan black bulbul Hypsipetes leucocephalus; more specifically, as a white-headed example of the nominate form H. l. leucocephalus. This individual was photographed in captivity here in England: wild ones occur from eastern Afghanistan through northern Pakistan and India to southern China, and also in parts of Myanmar, Laos and Thailand. It's somewhat ironic that a bird known as the 'Black' bulbul can have a strikingly white head and breast, and indeed most populations and species within the species complex that includes H. l. leucocephalus are mostly greyish, dark-headed birds [photo below - by J. M. Garg, from wikipedia - shows a properly dark-headed H. l. psaroides, photographed in Himachal Pradesh, India].

i-81f530558bf3f2183e6592accbb26e3d-Black-Bulbul_JM-Garg-wikipedia-Feb-2011.jpg

I said that there was, potentially, a lot to say about this species. Some of you suggested that the bird might instead be H. madagascariensis. If we're being as specific as possible then that's wrong, but it's not entirely wrong, since the two have been regarded as part of the same species by some authors (see the historical review in Dickinson & Dekker 2002). There is in fact a whole complex of white-headed and dark-headed bulbuls of this sort (occurring in Madagascar, the Mascarenes, southern India and Sri Lanka); authors have differed whether they regard these as many distinct species, or as subspecies and/or intermediates of just one or a few species.

Warren et al. (2005) examined the phylogeny and biogeography of these confusing birds, using mtDNA from 32 individuals. Because H. philippinus from the Philippines was the sister-taxon to all the other taxa and because Asian birds long referred to as 'H. madagascariensis' were also recovered as 'basal' in the tree, it seems that these bulbuls originated in Asia and then spread across the western Indian Ocean, apparently some time between the Late Pliocene and Middle Pleistocene (Warren et al. 2005). A more recent expansion of H. madagascariensis from out of Madagascar apparently led to the colonisation of Aldabra and the Comoros [this discussion treats H. philippinus as a species of Hypsipetes, but it probably warrants placement in another clade... read on].

As is so often the case, the molecular lineages identified by Warren et al. (2005) didn't match with the species conventionally identified on the basis of morphology. The old, Asian 'H. madagascariensis' population was substantially removed from 'true' H. madagascariensis of Madagascar, for example, while populations of both the Comoro black bulbul H. parvirostris and Philippine bulbul H. philippinus were non-monophyletic (Warren et al. 2005). Also interesting is that the relatively large, greenish Hypsipetes species found on various Indian Ocean islands were not close relatives (as usually thought), but represented convergently similar, independently evolved descendants of small, grey species. [A little bit of Asiatic/Indian Ocean bulbul variation is depicted in the image below... ignore the Hypocolius].

i-bc399171b6abad02fd19ef44acf5be6d-bulbuls-in-books-Feb-2011.jpg

The fact that these Indian Ocean bulbuls came from continental Asia mirrors the colonisation history of at least a few other Indian Ocean birds, squamates and other animals, including owls, magpie-robins, and the group of pigeons that includes dodos and solitaires. Once upon a time, it was thought that just about all of the animals of the Mascarenes came from India or south-east Asia, but we now know that many groups also invaded from Africa, including sunbirds, kestrels and mabuya skinks. While the phylogeny of Indian Ocean bulbuls indicates that they colonised the big islands first, note that this isn't necessarily what happened to all the groups that invaded the region: fruit bats, for example, seem to have colonised the small, remote islands first.

And also worth noting - in part because, again, it's kind of contrary to 'tradition' - is that weird little island endemics can sometimes invade continents and produce continental radiations (Filardi & Moyle 2005). This notion of 'upstream' colonisation is contrary to the better known idea whereby continental colonists move 'downstream' across ocean archipelagos.

All-inclusive Hypsipetes vs smaller Hypsipetes, and separate Ixos, Microscelis, Iole, Hemixos and Tricholestes

As hinted at from this little discussion, the species included within Hypsipetes are variable in size and looks. There are mostly light species, mostly dark species, crested ones, crestless ones, and small ones and large ones. Depending on how you have your genericometer is calibrated, these could easily be subsumed within the one 'genus' ooor they could best be split up into at least five different 'genera'. And, surprise surprise, over the years a large number of different generic names have been created for the different species: the result being that, if Hypsipetes sensu lato should be shot to pieces, there are already names associated with the different fragments.

i-961fb5c1e0843e5af8b7f21837acb6c3-Brown-eared-bulbul-wikipedia-Feb-2011.jpg

Indeed, noting that "The issue of the correct generic names for the species within this complex has, apparently, been long avoided by those who have had to deal with it", Gregory (2000, p. 164) brought attention to the generic names that are available for the different subdivisions of Hypsipetes sensu lato. Among these, the name Hypsipetes Vigors, 1831 itself has to go with H. psaroides, a junior synonym of H. leucocephalus. Then there's Ixos Temminck, 1825, which has to go with the Sumatran bulbul Ix. virescens and similar species (and which, you'll note, pre-dates Hypsipetes!*), Microscelis Gray, 1840, which has to go with the Brown-eared bulbul M. amaurotis [shown here, from wikipedia] and similar species, Iole Blyth, 1844, which has to go with the Buff-vented bulbul I. olivacea and similar species, Hemixos Blyth, 1845, which has to go with the Ashy bulbul H. flavala and similar species, and Tricholestes Salvadori, 1874, the type species of which is the Hairy-backed bulbul T. criniger. There are several other available names too. Some ornithologists have been using at least some of these names for a while as they feel that the constituent taxa look 'different enough' to warrant separation from Hypsipetes sensu stricto. As we'll see, phylogenetic studies might lend support to the idea that Hypsipetes sensu lato should be split up, and hence that these names should come back into general use. The composite image below shows (left to right) Buff-vented bulbul Iole olivacea [photo by Lip Kee Yap, from wikipedia], Chestnut bulbul Hemixos castanonotus [photo by Charles Lam, from wikipedia], and Mountain bulbul Ixos mcclellandii [photo from wikipedia].

* See Dickinson & Gregory (2002) for a discussion of the nomenclatural history of Ixos. It was long ignored as a potential senior synonym of Hypsipetes because its type species was (wrongly) thought to be Turdus phoenicopterus Temminck, 1821, the bird now known as Campephaga phoenicea (the Red-shouldered cuckoo-shrike). Now that Ixos is definitely associated with bulbuls, Pycnonotidae Gray, 1840 is a junior synonym of Ixodidae Bonaparte, 1838. Because this name is already widely used for a group of ticks, there were plans at one stage to get the new name Ixosidae used for bulbuls (Bock 1994). Dickinson & Gregory (2002) said of this matter that it seemed most sensible that "the status quo should be maintained and the family name Pycnonotidae retained" (p. 88), and this is far as things have gone. Thanks to Lars Dietz for bringing Dickinson & Gregory (2002) to my attention.

i-038deecf3c150e137d0ff1e7bbd14104-Hypsipetes-group-bulbul-motley-Feb-2011.jpg

Bulbuls within the passerine tree

It is universally agreed that Hypsipetes is a bulbul - a member of the passerine group Pycnonotidae - though this, of course, doesn't necessarily mean that it's true. In the 'traditional' passerine classifications that tended to make it into both textbooks and the popular literature (that is, those of Gadow, Stresemann, Wetmore and Mayr), bulbuls were typically placed within an assemblage that also includes other robust-billed, mostly frugivorous passerines, like waxwings, leafbirds, orioles, cuckooshrikes and bowerbirds.

Sibley & Ahlquist (1990) were the first authors to propose that oscine passerines consisted of two major radiations: one including all crow-like forms (Corvida) and one including all sparrow/finch/warbler-like forms (Passerida). Passerida, in turn, consisted of three clades: Muscicapoidea, Sylvioidea and Passeroidea. I'll be here all day if I start to elaborate on the content, membership and historical nomenclature of these groups, so I'll stop there. We're only interested for now in Sylvioidea: a group thought by Sibley & Ahlquist (1990) to include nuthatches, tits, tree-creepers, wrens, kinglets, bulbuls, leaf warblers, white-eyes and babblers. Nuthatches, treecreepers, wrens and gnatcatchers - grouped together as Certhioidea - may represent a distinct lineage closer to Muscicapoidea and Passeroidea (most likely Muscicapoidea) than to the remaining birds originally included in Sylvioidea, and tits (Paridae and Remizidae or Remizinae) seem to be a distinct lineage as well. The 'remainders' have indeed been recovered as a clade for which the name Sylvioidea has been co-opted (Alström et al. 2006).

Incidentally, one problem with the names that have been used for the major passerine clades is that explicit definitions have (to my knowledge) never been published. Ergo, when a certain group of workers recovers a particular clade, they then find themselves being uncertain as to which name it should have. Somebody really should work on this, hint hint. I'll make things a little clearer here (perhaps) by saying that Sylvioidea should be all oscine passerines closer to the respective type species of Sylvia than to the respective type species of Muscicapa (representing the Muscicapoidea) and Passer (representing the Passeroidea). A substantially simplified phylogeny of Passerida is shown below [all images from wikipedia] (sorry it's so small: I can only post images at a maximum width of 490 pixels).

i-846cb4b164c6e45a1bc756d497f19bb7-Phylogeny-of-Passerida-Feb-2011-update-490-px.jpg

Bulbuls among the sylvioids

In recent years, Sylvioidea has become famous for being fabulously confusing and poorly resolved, incorporating as it does the babblers, laughingthrushes, bulbuls, greenbuls, African warblers, white-eyes and so many others. Luckily, I don't need to worry about the majority of those groups now. Within this massive clade, Alström et al. (2006) found bulbuls to be the sister-group to a sylvioid clade that included Hirundinidae (swallows and martins), Cettidae (bush warblers and kin), Aegithalidae (long-tailed tits) and Phylloscopidae (leaf warblers and kin). Larks (Alaudidae) were the sister-taxon to this 'bulbuls + other sylvioids' clade.

A few other positions for bulbuls have been recovered by other studies - Sefc et al. (2003), for example, found them to be the sister-group to swallows - but most workers have recovered a similar topology to Alström et al. (2006), though sometimes with less resolution (e.g., Johansson e al. 2008). Some taxa conventionally included within Pycnonotidae - namely the nicators, greenbuls and Madagascan endemics Oxylabes and Crossleyia - belong elsewhere within Sylvioidea (Cibois et al. 2001, Beresford et al. 2005, Johansson et al. 2008).

Sorting out the bulbuls

Where does Hypsipetes fit among bulbuls as a whole? A few phylogenetic studies devoted to the relationships and diversification of bulbuls have now been published. In a study mostly designed to test the phylogeny of the Criniger bulbuls of Asia and Africa (better known as the Asian bearded bulbuls), Pasquet et al. (2001) found Hypsipetes to be part of an 'Asian clade' that also included Pycnonotus and the Asian Criniger taxa (Pycnonotus isn't exclusive to Asia, and invaded Africa at some point in its history). The African Criniger species belonged in an African clade together with Bleda, Phyllastrephus and other bulbuls, so Criniger was recovered as non-monophyletic.

i-c061fb291d65ec092d43771f449454dc-Alophoixus-phaeocephalus-Lip-Kee-Yap-wikipedia-Feb-2011.jpg

And so was Hypsipetes (here I'm talking about Hypsipetes sensu lato), since three of its species formed a clade that was closer to the Asian bearded bulbuls than was a H. propinquus + H. criniger clade (Pasquet et al. 2001). Because the name Criniger Temmink, 1820 should remain associated with its type species (African C. barbatus), the Asian species previously included in the genus should be renamed: some authors have already done this, and have used Alophoixus Oates, 1889 (originally coined for 'C.' phaeocephalus, the Grey-headed bearded bulbul or Yellow-bellied bulbul [shown here: photo by Lip Kee Yap, from wikipedia]) for the Asian bearded bulbuls. This is what Pasquet et al. (2001) chose to do, though they did allude to the idea that Alophoixus should be used for phaeocephalus alone and that a new name should be created for the remaining Asian bearded bulbuls. There's another idea that they didn't mention: this being that all 'Asian crinigers' be sunk into the otherwise paraphyletic Hypsipetes. To my knowledge, this latter idea has never been seriously proposed.

Incidentally, note that two of the three or so polytypic genera included in that small study were found to be non-monophyletic: a common theme in phylogenetic studies of passerines. I don't think it's as widely realised as it should be that all too many polytypic 'genera' of tradition are artificial, non-monophyletic assemblages.

Results essential identical to those of Pasquet et al. (2001) were reported by Jønsson & Fjeldså (2006) but, then, their's was a supertree study. Moyle & Marks (2006) produced a larger analysis incorporating a lot more species. They also found Hypsipetes sensu lato to form a clade with the Asian bearded bulbuls (the Hook-billed bulbul Setornis criniger was in there as well), and again recommended that the name Alophoixus be used for the Asian bearded bulbuls.

Most recently, Oliveros & Moyle (2010) included numerous Asian bulbuls within a much larger phylogeny and recovered a similar topology [a section of their phylogeny is shown below]. It has many implications for the species included Hypsipetes sensu lato. Here's where we come back to all those generic synonyms mentioned earlier on. Oliveros & Moyle (2010) recovered a distinct structure within 'Hypsipetes-group bulbuls', and found many of the divergences within this clade to be equally as old as, or older than, the divergences present between many other of the Asian bulbul 'genera'.

i-7f7b88fefc67fae053b3740327447696-Oliveros-&-Moyle-2010-bulbul-phylogeny-section-Feb-2011.jpg

In order of divergence from older to youngest, they recovered clades that corresponded to Iole, Hemixos, Hypsipetes sensu stricto, Microscelis, and Ixos. Even this treatment doesn't result in a totally tidy nomenclature, since the species conventionally included in Ixos didn't form a clade. The Hairy-backed bulbul T. criniger - often included in Hypsipetes - wasn't a part of the 'Hypsipetes group' at all, but was recovered as the sister-taxon to the clade that includes the Asian bearded bulbuls (those 'Asian crinigers') and all the 'Hypsipetes-group bulbuls' (Oliveros & Moyle 2010).

As usual, you could make the argument that we should stick to tradition and just lump all of the species concerned into one super-inclusive Hypsipetes genus and be done with it... but, to my mind, it seems more informative (and more interesting) to have the taxonomy reflect the phylogeny. None of these changes affect Hypsipetes leucocephalus - since it will be forever associated with the name Hypsipetes - but they do very much affect our view of Hypsipetes in the old, inclusive sense of the name. At last, Hypsipetes goes only with those confusing white- and dark-headed bulbuls like the white-headed Black bulbul that started this whole discussion.

If you made it though the whole of this article, you might now have a rough idea of bulbul phylogeny and diversity, but we've hardly touched on them at all really. At least I've started - at long last - to make one minor inroad into the terrifying, gargantuan nexus that is the world of passerine phylogeny. Much more one day.

For previous Tet Zoo articles on passerines see...

Refs - -

Alström, P., Ericson, P. G. P., Olsson, U. & Sundberg, P. 2006. Phylogeny and classification of the avian superfamily Sylvioidea. Molecular Phylogenetics and Evolution 38, 381-397.

Beresford, P., Barker, F. K., Ryan, P. G. & Crowe, T. M. 2005. African endemics span the tree of songbirds (Passeri): molecular systematics of several evolutionary 'enigmas'. Proceedings of the Royal Society of London B 272, 849-858.

Bock, W. J. 1994. History and nomenclature of avian family-group names. Bulletin of the American Museum of Natural History 222, 1-281.

Cibois, A., Slikas, B., Schulenberg, T. S. & Pasquet, E. 2001. An endemic radiation of Malagasy songbirds is revealed by mitochondrial DNA sequence data. Evolution 55, 1198-1206.

Dickinson, E. C. & Dekker, R. W. R. J. 2002. Systematic notes on Asian birds. 25. A preliminary review of the Pycnonotidae. Zoologische Verhandelingen Uitgegeven door het Rijksmuseum van Natuurlijke Historie te Leiden 340, 93-114.

- . & Gregory, S. M. S. 2002. Systematic notes on Asian birds. 24. On the priority of the name Hypsipetes Vigors, 1831, and the division of the broad genus of that name. Zoologische Verhandelingen Uitgegeven door het Rijksmuseum van Natuurlijke Historie te Leiden 340, 75-91.

Filardi, C. E. & Moyle, R. G. 2005. Single origin of a pan-Pacific bird group and upstream colonization of Australasia. Nature 438, 216-219.

Gregory, S. M. S. 2000. Nomenclature of the 'Hypsipetes' bulbuls (Pycnonotidae). Forktail 16, 164-166.

Johansson, U. S, Fjeldså, J. & Bowie, R. C. K. 2008. Phylogenetic relationships within Passerida (Aves: Passeriformes): a review and a new molecular phylogeny based on three nuclear intron markers. Molecular Phylogenetics and Evolution 48, 858-876.

Jonsson, K., & Fjeldsa, J. (2006). A phylogenetic supertree of oscine passerine birds (Aves: Passeri) Zoologica Scripta, 35 (2), 149-186 DOI: 10.1111/j.1463-6409.2006.00221.x

Moyle, R. G. & Marks, B. D. 2006. Phylogenetic relationships of the bulbuls (Aves: Pycnonotidae) based on mitochondrial and nuclear DNA sequence data. Molecular Phylogenetics and Evolution 40, 687-695.

Oliveros, C. H. & Moyle, R. G. 2010. Origin and diversification of Philippine bulbuls. Molecular Phylogenetics and Evolution 54, 822-832.

Pasquet, Ã., Han, L.-X., Khobkhet, O. & Cibois, A. 2001. Towards a molecular systematics of the genus Criniger, and a preliminary phylogeny of the bulbuls (Aves, Passeriformes, Pycnonotidae). Zoosystema 23, 857-863.

Sefc, K. M., Payne, R. B. & Sorenson, M. D. 2003. Phylogenetic relationships of African sunbird-like warblers: Moho (Hypergerus atriceps), Green hylia (Hylia prasina) and Tit-hylia (Pholidornis rushiae). Ostrich 74, 8-17.

Sibley, C. G. & Ahlquist, J. A. 1990. Phylogeny and Classification of Birds. New Haven: Yale University Press.

Warren, B. H., Bermingham, E., Prys-Jones, R. P. & Thebaud, C. 2005. Tracking island colonization history and phenotypic shifts in Indian Ocean bulbuls (Hypsipetes: Pycnonotidae). Biological Journal of the Linnean Society 85, 271-287.

More like this

Making Hypsipetes stand for the entire Ixos+Microscelis+Hypsipetes clade reflects phylogeny just as well as renaming one of the Ixos clades - both will make genera monophyletic.

The data aren't all that good, but I think they still point to Paridae (and by implication Stenostiridae) being within Sylvioidea. The big question is what Regulus is. And of course Passerida includes some groups you haven't shown that either clearly are or may be outside any of the superfamilies, like Petroicidae. And they're all just passerines anyway.

By John Harshman (not verified) on 02 Feb 2011 #permalink

Yeah, I knowingly and deliberately left out kinglets, bombycillids, African hyliotas and petroicids. Johansson et al. (2008: cited above) found parids to be closest to sylvioids, but have you seen Nabholz et al. (2010)? They applied shot-gun transcriptome sequencing to obtain protein-sequencing genes from mtDNA and generated a topology (Muscicapoidea + (Paridae + (Sylvioidea + Passeroidea))). Hmm.

Nabholz, B., Jarvis, E. D. & Ellegren, H. 2010. Obtaining mtDNA genomes from next-generated transcriptome sequencing: a case study on the basal Passerida (Aves: Passeriformes) phylogeny. Molecular Phylogenetics and Evolution 57, 466-470.

Haven't seen it. I'm suspicious of the efficacy of mt data for deep relationships, though this may be shallow enough. Still, the details of analysis can be crucial. Do I understand from the abstract that they're using only four species?

By John Harshman (not verified) on 02 Feb 2011 #permalink

Why was the genus actually called Hypsipetes, if the name Ixos is older?

By Lars Dietz (not verified) on 02 Feb 2011 #permalink

This factor is one that caused Gregory (2000) to state "The issue of the correct generic names for the species within this complex has, apparently, been long avoided by those who have had to deal with it", as quoted above. So, yes, if you regard Hypsipetes and Ixos as congeneric, the latter wins priority, and it seems that most authors ignored this (I suppose I should have emphasised it in the text). Needless to say, the issue goes away if the two are kept separate.

"except larks"

I can see corvids being different from other passerines in morphology, but what makes larks so unique?

You know, seeing as you bring up so many groups that people claim "all look alike except for maybe x" but really aren't once you actuall look at them (passerines, hadrosaurs, ceratopsians, etc.), maybe you need to write a book on the subject. It would be a rather interesting read.

By Anonymous (not verified) on 02 Feb 2011 #permalink

Lars - wow, thanks much... I'm a little frustrated that I missed that paper. Anonymous (grr) - I'll elaborate on the distinctiveness of larks some other time. But, yes, that old chesnut about 'all members of group x all looking alike apart from their heads' is pretty much NEVER true. One might say that it's only endorsed by people who want to make lazy generalisations.

Then again, one must try to find an explanation for the tendency of passerines to pass from one superfamily to another upon molecular study. If you can't tell a sylvioid from a corvoid by looking at their anatomy, what does that say?

By John Harshman (not verified) on 02 Feb 2011 #permalink

I'll make things a little clearer here (perhaps) by saying that Sylvioidea should be all oscine passerines closer to the respective type species of Sylvia than to the respective type species of Muscicapa (representing the Muscicapoidea) and Passer (representing the Passeroidea).

That's always been my preferred usage, with reciprocal definitions for Muscicapoidea and Passeroidea. Certhioidea, on the other hand, is probably best treated as node-based (Certhia + Sitta + Troglodytes would be my preference) so that it can potentially still label the certhioid clade even if that clade should fall within one of the larger clades.

I believe that there is still the niggling little detail that the superfamily name with priority is actually Hirundinoidea, not Sylvioidea, but I won't tell anyone if you don't.

Then again, one must try to find an explanation for the tendency of passerines to pass from one superfamily to another upon molecular study. If you can't tell a sylvioid from a corvoid by looking at their anatomy, what does that say?

But then, I'm not sure that anyone's ever really tried. Earlier passerine classifications were all based on intuitive comparison, until molecular phylogenies became the only game in town. It would be interesting if someone did a proper large-scale morphological analysis of the group (anyone got a spare lifetime or two?) For instance, I believe that Pseudopodoces is not without morphological indicators of its tit ancestry once you're open to the possibility of it having one.

Oh my lord.
What a piece of work that was to read. Nice.

As said above, some clades are said to be "highly similar except the heads" and ceratopsians were the one in particular I had heard about. Not being expert, I took Dougal Dixon's word for it.

I can understand the motivation to simplify like that. It certainly makes things easier. The grand futile ambition of my curiosity about dinosaurs is to postulate a more complete view of their world, because that could be used to extrapolate hypothetical species for which there are no known fossils. So tantalizing! Which would be why I'd simplify dinosaurs, to make them easier to wrap my brain around. No such luck. Nature is a tangled beast.

Which is why the so-called frogamander is almost certainly not a true branching point for the modern species. If the fossils are found, it will prove to be one of hundreds of variations on the combinations of squat-bodied short-tailed lissamphibian, none of which can be definitively pointed to as ancestral to living species. Just like the early birds turned out.

Fascination and frustration are BFF.

By CS Shelton (not verified) on 02 Feb 2011 #permalink

Sorry to through in a COMPLETELY off-topic question but Tet Zoo doesn't have a forum and it seems like the vast majority of the people who post here could easily answer this question.

I'm doing some research for a 6" long sculpture of a Carnotarus that I'm doing just for fun and the wikipedia article states;

"The most distinctive features of Carnotaurus are... and the extremely reduced forelimbs with four basic digits, though only the middle two of these ended in finger bones, while the fourth was splint-like and may have represented an external 'spur.' The fingers themselves were fused and immobile, and lacked claws."

Does this mean that all four fingers lacked claws, or just the two reduced ones? It sounds like NONE of the fingers had claws to me. It cites the journal ref on the page, but I don't have access to any journals so...

If anyone could clarify this for me I'd appreciate it. Again, sorry for the MASSIVELY off-topic post.

Agnolin, F.L. and Chiarelli, P. (2010). "The position of the claws in Noasauridae (Dinosauria: Abelisauroidea) and its implications for abelisauroid manus evolution." Paläontologische Zeitschrift, published online 19 November 2009. doi: 10.1007/s12542-009-0044-2

But then, I'm not sure that anyone's ever really tried. Earlier passerine classifications were all based on intuitive comparison, until molecular phylogenies became the only game in town. It would be interesting if someone did a proper large-scale morphological analysis of the group (anyone got a spare lifetime or two?) For instance, I believe that Pseudopodoces is not without morphological indicators of its tit ancestry once you're open to the possibility of it having one.

You are correct. One reason so many birds were misclassified is that a lot of it was done from study skins, i.e. from surface features only. All museums I know of have many, many more study skins than skeletons or pickles, even today. And it was worse in the past. I wasn't actually thinking of Pseudopodoces specifically, as there are many examples, but yes, Helen James' morphological analysis probably was the first ever detailed look at its skeletal anatomy. I don't actually recall which came first, the skeletal or DNA analysis; they were both published in the same paper.

But sequence is so easy compared to finding morphological characters.

By John Harshman (not verified) on 03 Feb 2011 #permalink

I'm too lazy to check the James et al. paper right now, but I think I recall the authors saying that a few other workers had looked at skeletal morphology and had _also_ found clues as to the non-corvid affinity of this taxon. Actually, I see I covered some of this before. There are a few similar stories. One other: Storrs Olson used skeletal morphology to argue that nicators weren't bulbuls, but rather closer to bush-shrikes (his conclusions haven't been supported by molecular work, but at least he was able to show that they didn't fit within the bulbul clade).

I'm not sure that finding morphological characters is at all difficult, even in passerines: it just needs a lot of time and effort, and everybody now 'knows' that molecular phylogeny is the way to go.

Thanks to others for comments - I don't have time to respond at the moment but might later.

Many of current papers on bird classification are methotodologically poor, and most of proposed clades are artifacts. I would prefer to wait until really good papers are weeded out and consensus is reached within next decade or so.

It would be very strange if birds, of all living things, lacked significant anatomical correlations of their phylogeny, isn't it?

Another Madagascar question. I heard that no Madagascan tetrapod is unquestionably Mesozoic native, and theory that blast wave from K/T impact in Carribean wiped out esentially all life on Madagascar which was directly opposite the globe, and all current vertebrate fauna populated Madagascar in Tertiary. Any comments?

hhhmm...

Seems that their is much more skepticism of the molecular methods than I would have expected. The Passerine relationships as derived from molecular data have largely proven consistent in the last decade, with resolution improving as missing parts of the tree have been filled in. Most molecular studies seem to support the "break up" of Sylviidae into Acrocephalidae, Phylloscopidae, etc. Even the Timaliidae-Sylviidae-Zosteropidae mess is starting to get sorted out.

Sure, there is a lot more work on the alpha taxonomic level, and more work to be done on the generic data, but I really don't see "every molecular studying disagreeing with the one before it" in the lit

By Morgan Churchill (not verified) on 03 Feb 2011 #permalink

I believe that there is still the niggling little detail that the superfamily name with priority is actually Hirundinoidea, not Sylvioidea, but I won't tell anyone if you don't.

Does Hirundinidae have priority over Sylviidae?

Which is why the so-called frogamander is almost certainly not a true branching point for the modern species.

It was never claimed to be. It was found, by the phylogenetic analysis of Anderson et al. (2008b), to be the sister-group of frogs + salamanders + albanerpetontids. According to the same analysis, there is no Lissamphibia -- the mentioned clades are temnospondyls just like Gerobatrachus, while the caecilians are lepospondyls.

I have a few bones to pick with that paper, including the analysis and its results (self & Laurin 2009; Sigurdsen & Bolt 2010; Sigurdsen & Green 2011), but never mind... the take-home message is to never trust a journalist about science.

Many of current papers on bird classification are methotodologically poor, and most of proposed clades are artifacts.

How do you know?

I would prefer to wait until really good papers are weeded out and consensus is reached within next decade or so.

When will consensus equal support by data?

Another Madagascar question. I heard that no Madagascan tetrapod is unquestionably Mesozoic native, and theory that blast wave from K/T impact in Carribean wiped out esentially all life on Madagascar [...], and all current vertebrate fauna populated Madagascar in Tertiary. Any comments?

The only possible exception is Erymnochelys. Most or all others have clearly arrived later.

By David MarjanoviÄ (not verified) on 03 Feb 2011 #permalink

Does Hirundinidae have priority over Sylviidae?

As far as I know, there is no priority rule at the superfamily level. In fact the ICZN has no rules at all at that level.

By John Harshman (not verified) on 03 Feb 2011 #permalink

As far as I know, there is no priority rule at the superfamily level. In fact the ICZN has no rules at all at that level.

No. Superfamilies are family-level taxa, so are linked in priority to their relevant family. It's order-level taxa and above that the ICZN doesn't comment on. And yes, Hirundinidae Rafinesque 1815 does potentially have priority over Sylviidae Leach 1820 (I say 'potentially' because those dates would need checking - there's been some debate over whether Rafinesque's 1815 Analyse de la nature counts as published, for a start). Mind you, family-level taxonomy is absolutely loaded with inconsistencies: priority didn't officially apply at family level until 1961, and there's a great many cases where established names have continued to be used despite not technically having priority.

It would be very strange if birds, of all living things, lacked significant anatomical correlations of their phylogeny, isn't it?

But the entire point of the discussion is that we don't really know whether they do or not. It's quite possible that they will show the appropriate correlations, once someone actually has a proper look.

David M--
Re: frogs + salamanders + albanerpetontids (#20)
Is that, on most recent analysis, left as a trichotomy, or are salamanders + albanerps definitely closer to each other than to anurans?
(Sorry to ask dumb questions: I make some effort to keep up with mammal(-imorph,-iform) but am thoroughly ignorant of the rest of life.)

By Allen Hazen (not verified) on 03 Feb 2011 #permalink

C. Taylor- Amen! I've been thinking about that off and on for a while. By all means, scientists, do the molecular work. But I would be especially keen to see the morphological backup of it. For species are radically different from close relations, we could find out more about the amazing adaptations involved in that. How does one get from basal apodiform to hummingbird? I gotta know.

By CS Shelton (not verified) on 03 Feb 2011 #permalink

No. Superfamilies are family-level taxa, so are linked in priority to their relevant family.

In that case, can it really be the case that there are no sylvioid genera described before 1815? Didn't Linnaeus have a single one?

By John Harshman (not verified) on 03 Feb 2011 #permalink

In that case, can it really be the case that there are no sylvioid genera described before 1815? Didn't Linnaeus have a single one?

Irrelevant. The ICZN recognises three regulated rank-groups: species-group, genus-group and family-group (taxa outside these three are unregulated). Species-group includes species and subspecies, genus-group genera and subgenera, family-group families, subfamilies, tribes and superfamilies. Priority is linked within the groups, but not between them. So the priority of family-group taxa is determined by the oldest family-group name, not by the genera within the families. It doesn't matter when the sylviid genera were published, it matters when the family name Sylviidae was published.

That aside, Hirundo Linnaeus 1758 is an older name than Sylvia Scopoli 1768, which was based on Linnaeus' Motacilla atricapilla.

Morgan (comment 19): I'm not seeing the scepticism you refer to (at least, not in the comment thread here, nor in my article). As Chris Taylor says above, we're noting that nobody has really tried to link morphology with the clades recovered by molecular study, not that the rather consistent molecular trees are suspect. Maybe someone will eventually do what they did with Afrotheria: the molecular support is so good that morphologists went looking for anatomical synapomorphies and eventually found them. In fact, I'm sure that there are morphological characters linking most (though perhaps not all) members of those major passerine clades (anyone want to fund me to demonstrate it?).

CS Shelton (comment 24): on swifts and hummingbirds, here is an excerpt from a long, in-press manuscript by yours truly...

Crown-hummingbirds seem to have diversified during the Miocene when lowland tropical ancestors gave rise to the hummingbird clades of the Andes, West Indies, and temperate regions of Central and North America (Bleiweiss 1998). The jungornithids of Eocene Eurasia appear to be stem-hummingbirds and share limb bone characters with crown-hummingbirds. Unlike crown-hummingbirds, jungornithids appear to have been adapted for gliding, and their short bills suggest that they caught flying insects in swift-like fashion. Feathers preserved in one taxon (Parargornis) suggest that it would have resembled an owlet-nightjar when alive. Stem-hummingbirds more closely related to crown-hummingbirds (namely, Eurotrochilus) inhabited Europe during the Oligocene. Eurotrochilus possesses specializations for hovering flight and its long, slender bill suggests nectarivorous habits.

priority didn't officially apply at family level until 1961

...Oh.

Re: frogs + salamanders + albanerpetontids (#20)
Is that, on most recent analysis, left as a trichotomy, or are salamanders + albanerps definitely closer to each other than to anurans?

IMNSHO that's the least likely of the many options for the phylogenetic position of Albanerpetontidae. The analysis by Anderson et al. (2008b) is the only one that has found albanerpetontids and salamanders together ever since new information on albanerpetontid anatomy was published in 1995.

Anatomically, frogs and salamanders are fairly clearly sister-groups, and the molecules agree more and more.

By David MarjanoviÄ (not verified) on 04 Feb 2011 #permalink

nobody has really tried to link morphology with the clades recovered by molecular study

I wouldn't go that far. Both Julia Clarke and Gerald Mayr have published on that subject. Certainly there's plenty more to be done. Someone, for example, will have to dive into Livezey & Zusi's characters at some point. Though having made a few attempts, I sympathize with anyone who does.

By John Harshman (not verified) on 04 Feb 2011 #permalink

I wouldn't go that far. Both Julia Clarke and Gerald Mayr have published on that subject.

I thought we were only talking about oscine passerines (I know I was!), not bird phylogeny as a whole.

I was more referring to the the comments

"Many of current papers on bird classification are methodologically poor, and most of proposed clades are artifacts"

as well as

"Then again, one must try to find an explanation for the tendency of passerines to pass from one superfamily to another upon molecular study."

I just am not seeing this in the molecular data sets. Yes there are still oddball taxa, and it's rough around the edge, but as far as passerine phylogeny goes we have been slowly moving towards stability.

By Morgan Churchill (not verified) on 04 Feb 2011 #permalink

Sibley & Ahlquist (1990) mentioned that the pneumatic fossa of the humerus is single in "Corvida" (now known to be paraphyletic) and double in Passerida as far as known at the time. So this would be one morphological character that supports the molecular phylogeny.
I think behavioral characters may also be useful in phylogeny. I've read an article in the German aviculture magazine "Gefiederte Welt" by someone who had kept various Yuhina species that said that the behavior of Erpornis zantholeuca is very different than those of Yuhinas (it was formerly included in that genus), which fits with the molecular data that say it's a corvoid related to vireos rather than a babbler. The article is:
âWas man nicht unterbringen kann . . .â Timalien. Gefiederte Welt 09-2004.
Unfortunately the author isn't listed on their website, and I don't have the magazine here at the moment.
I think part of the problem with the traditional passerine classificatons, together with a lack of detailed anatomical investigation, is also that most 19th and early 20th century ornithologists based their passerine families on North American and European birds, and especially the Australasian ones were then shoehorned somewhere in those families. (The same thing happened to suboscines before their distinctiveness from oscines was recognized.) Gadow recognized that problem already in 1893, and even recognized the Australian origin of oscines, but he didn't really offer any alternative to the traditional classifications.

By Lars Dietz (not verified) on 04 Feb 2011 #permalink

@Morgan Churchill
Optimism is good thing, definitely.

@Vladimir
Sometime on DINOSAUR mailing list (the best biological mailing list ever ;) )

On the supposed tendency for passerines to change superfamily: it's not all that common these days. Sibly & Ahlquist invented the superfamilies, so I suppose you can't really count anything before that, but I would at least give an honorable mention to their wholesale separation of the Australasian birds from their supposed Eurasian relatives. There have since been many departures and additions. Sure, the overall probability that a species will be found to be in the wrong superfamily is low, but there are also lots of passerines, and there are a fair number of cases. Pseudopodoces and Erpornis have already been mentioned. My favorite is Sapayoa, the New World's Old World suboscine. It isn't as if everything you know is wrong, but there's enough to make it interesting.

By John Harshman (not verified) on 04 Feb 2011 #permalink

I'll also repeat an observation I've made elsewhere: the passerine families that have seen the major re-shufflings are the ones that were never all that well defined in the first place. The really morphologically distinct groups have so far remained inviolate (a swallow is still a swallow, a lark is still a lark) or have only lost a few oddball members that were probably never all that comfortable there in the first place (crows, honeyeaters).

David Marjanovic (#30)
Thanks for reply!
...The Albanerps lasted until the Miocene or Pliocene! So we narrowly missed having another major branch of "crown" Lissamphibia.

By Allen Hazen (not verified) on 04 Feb 2011 #permalink

Families in passerines have generally not fared as well as families in other orders, but are at least much better than genera, which are a mess.

By John Harshman (not verified) on 04 Feb 2011 #permalink

Taylor&Naish- Thanks for the info! I've seen both swifts and hummingbirds up close and it's a pretty amazing change. Cool that we have fossil evidence to some of the process.

Swifts! So cool.

By CS Shelton (not verified) on 05 Feb 2011 #permalink

Sapayoa aenigma is teh awesum.

The Albanerps lasted until the Miocene or Pliocene!

Pliocene (Delfino & Sala 2007).

By David MarjanoviÄ (not verified) on 06 Feb 2011 #permalink

"But, yes, that old chesnut about 'all members of group x all looking alike apart from their heads' is pretty much NEVER true. One might say that it's only endorsed by people who want to make lazy generalisations."

Absolutely agree.

In response to JesseS about Carnotaurus, the manus is incomplete and Bonaparte et al. (1990) probably got identifications wrong based on the articulated manus of Aucasaurus. The latter suggests the supposed carpals are phalanges, and that the large spike-like bone is metacarpal I instead of IV. In any case, Aucasaurus has no phalanges on digits I and IV, and while no manual unguals are preserved for either it or Carnotaurus, we don't know if they were present on digits II and/or III. Note contra your quote, phalanges in digits II and III were not fused.

"Someone, for example, will have to dive into Livezey & Zusi's characters at some point. Though having made a few attempts, I sympathize with anyone who does."

The Mesozoic theropods in that study are terribly miscoded, for what it's worth.

The Mesozoic theropods in that study are terribly miscoded, for what it's worth.

Gerald Mayr didn't much like the Cenozoic theropod codings either. But of course he only mentioned a few characters. Somebody has to check the whole data set. Difficult. I, personally, am very suspicious about the serious lack of homoplasy.

By John Harshman (not verified) on 06 Feb 2011 #permalink

@ Mickey Mortimer;

Thanks that was extremely helpful. I think I'll sculpt it with small claws on II and III. If it turns out to be wrong in the future, well, I'm not as far off as the original Crystal Palace sculpts so...

Sad news: Bradley Livezey has died in a car crash.