ADVERTISEMENT
  About the SA Blog Network













Tetrapod Zoology

Tetrapod Zoology


Amphibians, reptiles, birds and mammals - living and extinct
Tetrapod Zoology Home

Ratites in trees: the evolution of ostriches and kin, and the repeated evolution of flightlessness (ratite evolution part II)

The views expressed are those of the author and are not necessarily those of Scientific American.


Email   PrintPrint



There really aren't enough good reconstructions of Aepyornis, the recently extinct elephant bird of Madagascar, out there. So here's another one. Image by Brian Choo. Read on for more about elephant birds.

Regular Tet Zoo readers will recall the article from March on ratite and tinamou evolution. Ratites, just in case you don’t know, are the flightless kiwi, ostriches, rheas, emus and so on, while tinamous are their diminutive, flight-capable, superficially pheasant-like close relatives. Together, ratites and tinamous are grouped together as the palaeognaths (properly: Palaeognathae). Some archaic fossil groups from the Cenozoic – most notably the long-billed, flight-capable lithornithids of the European Paleogene – are regarded as palaeognaths outside of the tinamou + ratite clade.

As discussed or hinted at in the previous palaeognath-themed article, several major areas of controversy and uncertainty make our evolving understanding of palaeognaths an area of special interest. Thoughts about the history of these birds have involved major, long-running debates on biogeography, the evolution of flightlessness and large body size, the alleged important of paedomorphosis, the issue of whether the ratite body shape evolved once or several times, and the matter of whether palaeognaths are ‘primitive’ with respect to other birds or not. My initial plan was to publish all of my ratite/tinamou-themed thoughts in one go. But I clearly have precognitive abilities of some sort, since no sooner had I published the part I article than I learnt about a couple of new and exciting papers that were set to appear in the imminent future. It seemed wisest to wait for those papers to come out… so here we are. Read on.

Lest we forget, ratites can be weird, and also weirdly cute. Here's a Darwin's or Lesser rhea (Rhea/Pterocnemia pennata), lying on its belly on the grass. Weird. Photo by Darren Naish.

Ratite polyphyly: a neat idea, shame it’s not supported by evidence

In the previous article I discussed the sometimes popular (but likely erroneous) idea that ratites might owe their unusual anatomical features to paedomorphosis (the process in which developmental timing changes such that adults retain features more usually typical of their juvenile stage). I don’t think that the data supports the view that paedomorphy might have been important in ratite evolution, but the idea that it is ties in with another evolutionary hypothesis sometimes entertained about these birds: are they all really close relatives, or could they, actually, have arisen independently from distantly related ancestors? This is known as the polyphyly hypothesis (‘polyphyly’ being the term used for cases where members of a group do not share a single ancestor, but have some or many disparate evolutionary origins).

Nobody has ever formally proposed a model of palaeognath polyphyly like the one you see whimsically depicted here. However, some of the relationships hinted at in this diagram have been proposed by some supporters of palaeognath non-monophyly. Image by Darren Naish.

There was a time when the idea of ratite polyphyly was fairly popular. I was surprised to read, in Hans Hvass’s 1963 Birds of the World that “It was at one time thought that [ratites] were all related; but there is no doubt that they originate from different groups of early birds” (Hvass 1963, p. 199). As often with views of this sort, the notion of ratite polyphyly was never explained in detail, and usually only alluded to in cursory fashion. To be fair, Victorian arch-anatomist Richard Owen made some specific statements about the concept of ratite polyphly, suggesting that ostriches were allied to bustards, and that kiwi and moa were close to megapodes. And the idea that ostriches might be allied to (or descended from) a weird group of didactyl crane-like birds from the Eocene and Oligocene called eogruids was promoted for a while during the 1970s and 80s.

Diversity in ratite pelvic anatomy. They're different, but are they really _that_ different? Redrawn after a diagram in Feduccia (1996), image by Darren Naish.

But if you read the books and articles of Alan Feduccia and Storrs Olson, you get the impression that various ratites could have evolved independently from such groups as ducks, geese, ibises or the crane-like eogruids… it being, alas, about impossible to pin these authors down on specific statements, or on specific bits of evidence that might really link the different ratite lineages to any non-ratite ones. In fact, this view of ratite polyphyly seemingly arose because the several ratite lineages look different in detail.

Think about wings, for example. All ratites have a humerus that’s somewhat longer than the lower arm, but while ostriches and rheas have long wing feathers that attach to long, tridactyl, two-clawed hands, emus and cassowaries have very short wings with a substantially reduced, single-clawed carpometacarpus and poorly ossified (or unossified) splint for the alula. Kiwi have stick-thin wings where the manus is monodactyl, and moa have no wings at all. An interesting and perhaps surprising amount of variation is also seen in ratite scapulocoracoid and pelvic anatomy. Take this statement: “The diversity of ratite pelvic structure would be astounding if all derived directly from a single flightless common ancestor and is perhaps best explained by ratites having evolved through a combination of increasing body size and neoteny” (Feduccia 1996, p. 273).

Diversity in the hands of living ratites. At top left, a de-feathered ostrich wing (photo by John Hutchinson); at lower left, de-feathered emu wing (photo by Mathew Wedel); at right, manual skeletons from Parker (1888).

Personally, I think that things have been over-stated, since you can see profound disparity in the skeletons of numerous animal groups where lineages have been separate for some tens of millions of years. Nevertheless, the point remains that ratites are different enough in anatomy for us to at least suspect that something weird has gone on as goes the evolution of their bauplan. In other words, the very different wing and pelvic anatomies seen in the different ratite lineages might indicate that those different lineages evolved their ratite-type morphology independently. [Image below by Rei.]

Ostrich wings are proportionally longer than those of many other ratites, and their long, showy remiges make them look much larger in the live animal. Image by Rei, licensed under Creative Commons Attribution-Share Alike 3.0 Unported license.

But does this mean that the different ratite groups emerged independently from different neognaths? Could ducks or ibises or cranes or whatever evolve into ‘ratites’ if only they were given enough time? (this being the possibility that Feduccia and Olson appear to favour). The short answer is no: ratites are united by a list of anatomical characters that aren’t seen in other birds (e.g., Bledsoe 1988, Lee et al. 1997, Livezey & Zusi 2007, Mayr & Clarke 2003, Bourdon et al. 2009), and a large number of molecular studies consistently find ratites to group together, and to group together with tinamous (e.g., Haddrath & Baker 2001, Hackett et al. 2008, Harshman et al. 2008, Suh et al. 2011, Smith et al. 2013, Yuri et al. 2013, Mitchell et al. 2014). There have, in other words, never been good indications from quantitative analyses that palaeognaths are not monophyletic.

A tree for ratites

Palaeognaths: an impressive amount of diversity in size and shape. Or is that disparity? Image by Darren Naish.

Within the palaeognath clade, how might the different lineages be related to one another? Over the decades, people have of course proposed several good, sensible and perfectly logical ideas about palaeognath phylogeny. Tinamous (endemics to the Americas) are small, capable of flight and without the anatomical specialisations of ratites, so they’re presumably the sister-group to ratites. Moa and kiwi are both endemic to New Zealand, so it seems logical to assume a close affinity between both groups. Cassowaries and emus both inhabit Australia and look fairly alike anyway, and ostriches and rheas look about enough alike to indicate that they’re presumably close relatives, that pesky Atlantic Ocean being a bit of a pain.

The history of ideas about palaeognath phylogeny are complex (err, just as they always are) and I’m not about to provide a thorough summary here (if you want to see such a summary, consult Sibley & Ahlquist (1990)). Three things about ‘historical’ views on palaeognath evolution are especially intereting. (1) Tinamous were often grouped with galliforms, mostly because authors were confused either by convergence, or by the shared presence of primitive characters; (2) several authors favoured ideas about ratite polyphyly like those discussed above; and (3) at least some authors who interpreted ratites as monophyletic arranged the lineages in a fairly ‘modern’ phylogeny (e.g., Mivart 1877).

The ‘modern era’ of palaeognath phylogeny originated with Joel Cracraft’s 1974 study (Cracraft 1974). Based mostly on the distribution of select skeletal characters, he proposed that kiwi and moa were close kin, and that they were outside a clade that contained all other ratites, the topology of which was (elephant birds + ((emus + cassowaries) + (ostriches + rheas))). Note, however, that he only used 25 characters (Cracraft 1974), and subsequent workers pointed to problems with certain aspects of his analysis.

Cracraft's (1974) hypothesis of palaeognath phylogeny. I used the same taxonomic names that he did (some of which are a bit odd compared to those used by others). Image by Darren Naish.

Modern palaeognath phylogenies - this one is based on that recovered by van Tuinen et al. (1998) - have tended to find ostriches to be the sister-group to remaining ratites. Image by Darren Naish.

The idea that ostriches and rheas are close relatives (a hypothesis supported by Cracraft) has often been popular – they look superficially similar and are fairly alike in ecology and behaviour – but it isn’t supported by molecular analysis, nor by more recent examinations of the distribution of anatomical details. Ostriches have most frequently been recovered as the sister-group to the remaining ratites (Prager et al. 1976, van Tuinen et al. 1998, Harshman et al. 2008, Phillips et al. 2010, Smith et al. 2013, Baker et al. 2014, Mitchell et al. 2014), though note that rheas have occupied this position in some other studies (Lee et al. 1997). Cassowaries and emus are uncontroversially recovered as sister-taxa in just about all analyses. What about kiwi? The fact that moa are endemic to New Zealand has of course contributed to the idea that they’re most closely related to kiwi. Cracraft (1974) found two tarsometatarsal characters that seemed to support this possibility. For a while, this meant that people thought of the proportionally enormous kiwi egg as a sort of evolutionary holdover from far bigger-bodied ancestors. This is a somewhat weird idea, given how labile traits like egg size are in other bird groups. [In image below, moa photo by Ghegoghedo.]

Extinct ratites: where do they fit in the phylogeny? At left, the moa Emeus and Pachyornis (image by Ghedoghedo, licensed under Creative Commons Attribution-Share Alike 3.0 Unported license.); at right, the elephant bird Aepyornis (image in public domain). Note the interesting difference in toe length and sternum size.

What about the extinct ratites: the elephant birds (or aepyornithids) and the moa (or dinornithiforms)? The news about moa isn’t news anymore – it was a huge surprise when first announced but is now familiar stuff. Moa, it seems, are the sister-group to tinamous. This relationship was first reported by Phillips et al. (2010) and later reported or discovered by Smith et al. (2013), Baker et al. (2014) and Mitchell et al. (2014). Note that a few other studies had also found tinamous to be nested somewhere within ratites (Harshman et al. 2008, Hackett et al. 2008, Faircloth et al. 2012). If correct*, this all means that ratites as conventionally conceived aren’t monophyletic, since tinamous – which have never been regarded as ratites under any understanding of the word – are deeply nested within the clade that includes all ratite lineages.

* I’ve recently learnt about an unpublished thesis that points to problems with published phylogenies nesting tinamous within ratites (Scherz 2013). I’m not convinced that it overturns the signal that seems to be emerging, but it does point to ongoing problems concerning lack of data.

Tinamous are really ok once you get to know them. This is a Patagonian or Ingouf's tinamou (Tinamotis ingoufi); image by CHUCAO, licensed under Creative Commons Attribution-Share Alike 3.0 Unported license.

Finding tinamous to be nested within ratites is a big deal [adjacent tinamou image by CHUCAO]. Given that ratites are flightless, does this mean that flight re-evolved in the lineage leading to tinamous? That would be pretty radical and surprising, not least because this has (so far as we know) never happened elsewhere in any other bird lineage. It would also be a big deal because all the ratite lineages – that is, all the outgroups to tinamous in these new phylogenetic topologies – possess a set of features which seemingly prevent them from ever evolving the ability to fly again (like those significantly simplified, reduced wings, modified sternum and scapulocoracoid, absence of the furcula and heavy-boned, massive hindlimbs).

What’s the alternative? Obviously, that flightlessness evolved independently on several occasions within palaeognaths (Harshman et al. 2008, Phillips et al. 2010, Smith et al. 2013). Exactly how many times depends on the preferred topology. Three or four times looks most likely, but it was perhaps as many as five times… in ostriches, in the cassowary-emu clade, in kiwi, in rheas, and in moa (what about elephant birds? Hold on) (Harshman et al. 2008, Phillips et al. 2010, Smith et al. 2013). So, here we have an explanation as to why the members of the different ratite lineages look so distinct as goes their wings and hips and so on: they evolved their big, flightless forms independently from smaller, flight-capable ancestors. If this is true, ratites like ostriches and rheas and elephant birds are a good example of parallelism – the phenomenon whereby close relatives convergently evolve similar appearances.

The newest palaeognath phylogeny: according to Mitchell et al. (2014), tinamous are close to moa (again), but now kiwi have a sister-group relationship with elephant birds. Say what? Image by Darren Naish.

The newest news on palaeognath phylogeny concerns Mitchell et al.’s (2014) paper in Science. They successfully retrieved mitochondrial DNA from the elephant bird taxa Aepyornis hildebrandti and Mullerornis agilis, which is a big deal in itself. But while elephant birds look something like moa or ostriches, the DNA shows “unequivocally” that they’re closest to kiwis, a surprising result that not only seems discordant with anatomy and ecology but also with distribution. In fact, it seems to absolutely contradict the idea that ratites were ancestrally flightless and owe their distribution to continental breakup, and can only sensibly be explained by over-water dispersal (Mitchell et al. 2014).

The point has already been made elsewhere online that an entire kiwi could just about fit inside the egg of Aepyornis; both taxa now seem to be close relatives. Image by Kyle Davis and Paul Scofield, Canterbury Museum.

The shape of palaeognath phylogeny now suggests that flight capability was widespread, even ubiquitous, across the clade, with large size and flightlessness evolving independently, apparently early on in the Cenozoic and in the wake of the KPg extinction event. Mitchell et al. (2014) in fact even suggest that big-bodied palaeognath groups evolved large size because they were among the first animals with this evolutionary potential within their respective ecosystems: those groups that arrived later on in the same places were then ‘denied’ the opportunity to also evolve large size in the same way. So, kiwi and tinamous are small because moa and rheas, respectively, ‘got their first’. Intriguing stuff.

This is far from all that there is to say about palaeognath evolution and history, of course, but (for now) we have to stop there. We’ll no doubt be coming back to this fascinating group again. For previous Tet Zoo articles on ratites and neornithine bird evolution in general, see…

Refs – -

Baker, A. J., Haddrath, O., McPherson, J. D. & Cloutier, A. 2014. Genomic support for a moa-tinamou clade and adaptive morphological convergence in flightless ratites. Molecular Biology and Evolution doi: 10.1093/molbev/msu153

Bledsoe, A. H. 1988. A phylogenetic analysis of postcranial skeletal characters of the ratite birds. Annals of Carnegie Museum 57, 73-90.

Bourdon, E., de Ricqles, A. & Cubo, J. 2009. A new transantarctic relationship: morphological evidence for a Rheidae-Dromaiidae-Casuariidae clade (Aves, Palaeognathae, Ratitae). Zoological Journal of the Linnean Society of London 156, 641-663.

Cracraft, J. 1974. Phylogeny and evolution of the ratite birds. Ibis 116, 494-521.

Faircloth, B. C., McCormack, J. E., Crawford, N. G., Harvey, M. G. & Brumfield, R. T. & Glenn, T. C. 2012. Ultraconserved elements anchor thousands of genetic markers spanning multiple evolutionary timescales. Systematic Biology 61, 717-726.

Feduccia, A. 1996. The Origin and Evolution of Birds. Yale University Press, New Haven & London.

Hackett, S. J., Kimball, R. T., Reddy, S., Bowie, R. C. K., Braun, E. L., Braun, M. J., Cjojnowski, J. L., Cox, W. A., Han, K.-L., Harshman, J., Huddleston, C. J., Marks, B., Miglia, K. J., Moore, W. S., Sheldon, F. H., Steadman, D. W., Witt, C. C. & Yuri, T. 2008. A phylogenomic study of birds reveals their evolutionary history. Science 320, 1763-1768.

Haddrath, O. & Baker, A. J. 2001. Complete mitochondrial DNA genome sequences of extinct birds: ratite phylogenetics and the vicariance biogeography hypothesis. Proceedings of the Royal Society of London B 268, 939-945.

Harshman, J., Braun, E., Braun, M., Huddleston, C., Bowie, R., Chojnowski, J., Hackett, S., Han, K., Kimball, R., Marks, B., Miglia, K., Moore, W., Reddy, S., Sheldon, F., Steadman, D., Steppan, S., Witt, C., and Yuri, T. 2008. Phylogenomic evidence for multiple losses of flight in ratite birds. Proceedings of the National Academy of Sciences 105: 13462-13467.

Hvass, H. 1963. Birds of the World. Eyre Methuen, London.

Lee, K., Feinstein, J. & Cracraft, J. 1997. The phylogeny of ratite birds: resolving conflicts between molecular and morphological data sets. In Mindell, D. P. (ed) Avian Molecular Evolution and Systematics. Academic Press, pp. 173-211.

Livezey, B. C. & Zusi, R. L. 2007. Higher-order phylogeny of modern birds (Theropoda, Aves: Neornithes) based on comparative anatomy. II. Analysis and discussion. Zoological Journal of the Linnean Society 149, 1-95.

Mayr, G. & Clarke, J. 2003. The deep divergences of neornithine birds: a phylogenetic analysis of morphological characters. Cladistics 19, 527-553.

Mitchell, K. J., Llamas, B., Soubrier, J., Rawlence, N. J., Worthy, T. H., Wood, J., Lee, M. S. Y. & Cooper, A. 2014. Ancient DNA reveals elephant birds and kiwi are sister taxa and clarifies ratite bird evolution. Science 344, 898-900.

Mivart, S. G. 1877. On the axial skeleton of the Struthionidae. Transactions of the Zoological Society 10 (1), 1-52.

Parker, W. K. 1888. On the presence of claws in the wings of the Ratitae. Ibis 6, 124-128.

Phillips, M. J., Gibb, G. C., Crimp, E. A. & Penny, D. 2010. Tinamous and moa flock together: mitochondrial genome sequence analysis reveals independent losses of flight among ratites. Systematic Biology 59, 90-107.

Prager, E. M., Wilson, A. C., Osuga, D. T. & Feeney, R. E. 1976. Evolution of flightless land birds on southern continents: transferrin comparison shows monophyletic origin of ratites. Journal of Molecular Evolution 8, 283-294.

Scherz, M. 2013. The Paraphyly of Ratites Just Doesn’t Fly. Unpublished Zoology Honours thesis, University of Edinburgh.

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

Smith, J. V., Braun, E. L. & Kimball, R. T. 2013. Ratite nonmonophyly: independent evidence from 40 novel loci. Systematic Biology 62, 35-49.

Suh, A., Paus, M., Kiefmann, M., Churakov, G., Franke, F. A., Brosius, J., Kriegs, J. O. & Schmitz, J. 2011. Mesozoic retroposons reveal parrots as the closest living relatives of passerine birds. Nature Communications Aug 23;2:443. doi: 10.1038/ncomms1448

Yuri, T., Kimball, R. T., Harshman, J., Bowie, R. C. K., Braun, M. J., Chojnowski, J. L., Han, K.-L., Hackett, S. J., Huddleston, C. J., Moore, W. S., Reddy, S., Sheldon, F. H., Witt, C. C. & Braun, E. L. 2013. Parsimony and model-based analyses of indels in avian nuclear genes reveal congruent and incongruent phylogenetic signals. Biology 2013, 2, 419-444.

Darren Naish About the Author: Darren Naish is a science writer, technical editor and palaeozoologist (affiliated with the University of Southampton, UK). He mostly works on Cretaceous dinosaurs and pterosaurs but has an avid interest in all things tetrapod. His publications can be downloaded at darrennaish.wordpress.com. He has been blogging at Tetrapod Zoology since 2006. Check out the Tet Zoo podcast at tetzoo.com! Follow on Twitter @TetZoo.

The views expressed are those of the author and are not necessarily those of Scientific American.





Rights & Permissions

Comments 661 Comments

Add Comment
  1. 1. DavidMarjanovic 6:34 am 05/24/2014

    As often with views of this sort, the notion of ratite polyphyly was never explained in detail, and usually only alluded to in cursory fashion.

    I think it wasn’t so much a hypothesis of its own as part of phylopessimism (Tassy 1998) = polyphyly of everything. In the middle 20th century, it dawned on people that there are no “reliable characters” that can “prove” the monophyly of a taxon; many concluded that, basically, we can’t know anything – everything might be polyphyletic: mammals could be descended from both “ictidosaurs” and “bauriamorphs”, “reptiles” could be descended from seymouriamorphs and “microsaurs” and who knows what else, lissamphibians from whole bunches of temno- and lepospondyls, tetrapods from “osteolepiforms” and porolepiforms, Hesperornis and Ichthyornis were happily assigned deeply nested positions within Neornithes, and so on. (Intriguingly, it was widely maintained that Archaeopteryx was a direct ancestor of all later birds; but imagine if dromaeosaurs, oviraptorosaurs etc. etc. had been as well known then as they are now.)

    Pascal Tassy (1998): L’Arbre à remonter le temps. Diderot.

    Link to this
  2. 2. DavidMarjanovic 6:37 am 05/24/2014

    we can’t know anything

    Except of course when, by some subjective metric, the fossil record is dense enough that you can just “see” the phylogeny in it. Hence stratophenetics. Often, the resulting phylogenetic hypotheses have turned out to be less parsimonious than possible.

    Link to this
  3. 3. darkgabi 6:37 am 05/24/2014

    these recent papers on palaeognaths are just so cool. it’s actually very interesting to see how some groups regarded as “primitive” are just superfically so… that they’re actually pretty derived in comparison to a supposed plesiomorphic archetype. so many reversals one must be very careful when making comparisons and using them as a model for anything. like it happaned to amphibians at some point. i’m a doorknob when it comes to mammals but after reading de beer, i have the feeling monotremes follow the same line in many aspects..

    Link to this
  4. 4. Zoovolunteer 6:53 am 05/24/2014

    Aepyornis and Mullerornis? I thought there was only one species of elephant bird on Madagascar. Were they both contemporary with the early human colonists or just Aepyornis?

    Link to this
  5. 5. naishd 6:57 am 05/24/2014

    Thanks for great comments :) Good point, David, as goes ‘phylopessimism’… but there’s also the point that people during the late 1800s/early 1900s had very broad views on what was possible – it being thought that convergence was more rampant within Tetrapoda that we think it is today, for example.

    Zoovolunteer (comment # 4): there are numerous named aepyornithids – it’s a big taxonomic mess that still needs resolving. Work is underway.

    Link to this
  6. 6. Pristichampsus 7:18 am 05/24/2014

    My comment from facebook:

    The fact that Brain Brian Choo restored the Aepyornis as vaguely emu-colored seemed odd to me, until I saw the tinamou photo in the blog post that has a vaguely similar facial color to an emu, and to the aforementioned drawing. Phylogenetic bracketing FTW. In truth I really like the new drawing, Brian. Only bad part really is that the National Geographic drawing from years ago, with Aepyornis drawn with ostrich colors, is now completely outdated. I always liked that drawing, too.

    And Darren’s reply from facebook:

    There’s always the temptation to reconstruct aepyornithids with something other than dull grey or brownish, shaggy plumage. But, fact is, dull greys and browns are ubiquitous across palaeognaths. We don’t know enough about aepyornithid habitat preference to say anything intelligent, though their foot morphology suggests to me that they were birds of dry grasslands and other open habitats. We should be having this discussing in the Tet Zoo comments field, you know.

    He said this discussion would fit better here, so, by all means discuss, everyone. Phylogenetic bracketing re extinct ratite reconstruction.

    Incidentally, we know a bit more about moa coloration now thanks to some rather interesting work on various preserved feathers (genetic testing and some interesting photography of the specimens). What I would like to know is, apparently there are bluish and purplish moa feathers in some museums, does this mean whole animals hued in dull pastel purples and blues, or simply some accents around the face or ankles? Also, of course, whitish feathers could be involved around the face, as well as the belly.

    Link to this
  7. 7. Gigantala 9:08 am 05/24/2014

    So now there’s basically what appears to be a mid-Cenozoic australasian ratite radiation, with the appearence of mainland cassowaries and emus and birds that dispersed to New Zealand to become kiwis and to Madagascar to become elephant birds. Thanks to Proapteryx, we can tell palaeognaths were still flying around as recently as the Miocene, so kiwis and elephant birds being relatively “young” clades is exciting, to say the least.

    On Mullerornis: some of the most recent elephant bird remains belong to it, if I’m not mistaken. Makes sense, that the smaller elephant birds would survive longer than the hulking Aepyornis.

    Link to this
  8. 8. Tayo Bethel 9:13 am 05/24/2014

    Could phylogeneticbracketing be applied to vocalizations, as well? How would the display calls of tinamous sound if scaled up to moa size and tracheal morphology? *Music and science an combine to come up with interesting and, it seems, unanswerable questions*

    Link to this
  9. 9. John Harshman 9:49 am 05/24/2014

    Moa, it seems, are the sister-group to tinamous. This relationship was first reported by Harshman et al. (2008)…

    If only that were true. We did the analysis — pairing the published mitochondrial genome data with our own nuclear data that lacked moas — and it did come out with moas and tinamous as sister taxa, but we decided not to publish that result for reasons that now seem very silly.

    Was there perhaps a garbled paragraph in there? Because you start out with moas and tinamous and end up with tinamous inside ratites, the latter being what was actually the big message of that paper.

    Now *that* relationship was actually first reported by Elzanowski (1995). But I like to think that our paper made it conclusive.

    Elzanowski, A. 1995. Cretaceous birds and avian phylogeny. Courier Forschungsinstitut Senckenberg 181:37-53.

    Link to this
  10. 10. John Harshman 10:15 am 05/24/2014

    I haven’t been able to analyze Scherz 2013 very closely, but I do notice that he excluded almost all the nuclear data from analyses, retaining only a few protein-coding bits. This seems like a problem to me.

    Link to this
  11. 11. Christopher Taylor 10:26 am 05/24/2014

    Some archaic fossil groups from the Cenozoic – most notably the long-billed, flight-capable lithornithids of the European Paleogene – are regarded as palaeognaths outside of the tinamou + ratite clade.

    Though I noted that in the supplementary info for Mitchell et al. (2014), Lithornis is placed in a combined phylogenetic analysis as sister to the tinamous, and hence also within the palaeognath crown group. I can think of a couple of reasons for adding a pinch of salt to this: if tinamous are relatively plesiomorphic compared to other palaeognaths, they could be dragging Lithornis over. But I note that Mlíkovský (2002) synonymised Lithornithidae and Tinamidae in his much-maligned Cenozoic Birds of the World, so Mitchell et al. have raised the possibility that this time, Mlíkovský was right.

    I’m also reminding myself that it has been explicitly suggested that lithornithids are a paraphyletic grade rather than a single clade. I’m not sure if anyone has ever run a formal phylogenetic analysis including the various lithornithids as separate taxa; maybe it’s time someone did.

    Link to this
  12. 12. naishd 10:33 am 05/24/2014

    With reference to comment # 9.. yes, there’s a mangled paragraph, I’ll rearrange it asap (can’t do so right now).

    Moa life appearance (comment # 6): I’m (slowly) putting together a manuscript on moa life appearance, and it includes a discussion of what we think we know about colours.

    Link to this
  13. 13. Gigantala 10:34 am 05/24/2014

    “Paleognathous Birds from the Early Tertiary of the Northern Hemisphere” suggested that Lithornis was at the base of Palaeognathae with Pseudocrypturus and Paracathartes formed a clade closer to “ratites”, if I recall correctly.

    Link to this
  14. 14. BrianL 11:02 am 05/24/2014

    That description of ratite wings made me wonder: what do aepyornithid wings look like? I read they only consist of small humeri, is that true? If so, is there a possibility they might have been entirely encased inside the body in life?

    Of course, with both lithornids and *Proapteryx* mentioned, it’s time to bring out other fossil ratites.
    I recall reading (though I unfortunately can’t remember where!) that Paleocene *Diogenornis*, normally allied to rheas, has also been considered closest to emu and cassowaries. If so, it too may have been part of this Australasian-Indian Ocean lineage of ratites, though its old age is somewhat inconvenient then, perhaps. The same applies to Seymour Island’s unnamed Eocene ratite. Of course, stem-moa or stem-tinamou would possibly have traveled across Antarctica too and there’s little to preclude the theoretical possibility of yet another flightless lineage, so that’s not very conclusive.
    Then, of course, there’s European *Remiornis* and *Paleotis*. The latter’s ‘reports of ‘ostrichiness’ are greatly exaggarated’, if we follow Mayr. In fact, Mayr suggest that these two might be most closely allied to each other so we might be looking at yet another extinct ratite clade here.

    @Darren:
    I recall you one time mentioning a giant flightless bird from the Paleogene of Europe (From the Paris basin or Messel, I believe) that was neither a ratite nor a gastornithid. It is some years ago by now, do you recall this? I can’t seem to find it anywhere. And if so, has anything new come up about this bird?

    Link to this
  15. 15. Andreas Johansson 11:52 am 05/24/2014

    Feduccia:
    “The diversity of ratite pelvic structure would be astounding if all derived directly from a single flightless common ancestor and is perhaps best explained by ratites having evolved through a combination of increasing body size and neoteny”

    The implicit assumption here, and in some of Darren’s later comments, appears to be that flightless forms should be expected to be more pelvically conservative than flighted ones. Is there a particular reason to think so?

    Elephant birds inside an otherwise Australasian clade inevitably makes me think of the Malagasy language, which is spoken on Madagascar with all its relatives way away in Southeast Asia and Oceania. Do the monsoon winds or something make it particularly easy to colonize the place from far away in the east compared to nearby Africa? The “Lemuria” landbridge / sunken continent was originally proposed to explain Madagascan affinities with Asia, I believe, so I guess there are more biotic examples?

    Link to this
  16. 16. Jenny Islander 12:14 pm 05/24/2014

    Complete amateur here; please pardon my howlers.

    Does the new data suggest that the ancestral paleognath group (genus, family?) was the Rallidae of its time? I’m thinking of the tendency of rallids, given a habitat that does not require migration, to give rise to flightless species. Or maybe, considering the awesome butt-kicking power of e.g. the ostrich, a better parallel would be Bucorvidae. Don’t-mess-with-me-peckers-about-in-the-open rather than sneaky-scramblers-in-the-reeds?

    Link to this
  17. 17. Gigantala 12:42 pm 05/24/2014

    @Andreas: Exactly! The Dodo + Solitaire clade has already been long used as an example of australasian birds that were dragged by the monsoon winds long in the mid-Cenozoic (Miocene, Oligocene or even earlier depending on who you ask), and mesites may also represent such (since they’re supposedly sister taxa to pigeons), so it’s not unconceivable that elephant birds simply follow this trend.

    @Jenny: Seems to be the case, given how many times flightless was lost and how widespread they were. Probably even ecologically analogous too, if you consider lithornithid appearent shorebird like habits.

    Link to this
  18. 18. Sebastian Marquez 1:22 pm 05/24/2014

    @ BrianL, you thinking of Palaeotis?

    Link to this
  19. 19. Sebastian Marquez 1:24 pm 05/24/2014

    urrg, nevermind, how could i miss it in the comment itself :P

    Link to this
  20. 20. Yodelling Cyclist 1:57 pm 05/24/2014

    Awesome article/series. What’s with the teenage mutant ninja cassowaries leaping into battle in the illustrations? ;-)

    Jesting aside, could any of the pros give a quick idea to us in the cheap seats as to when and where the palaeognaths and neognaths diverged? It would be helpful to have a slightly broader idea as to what is happening here.

    As to Moa colouration, this would be another problem solved by the John Conway clone ‘em and taste ‘em project.

    Link to this
  21. 21. vdinets 3:34 pm 05/24/2014

    Andreas Johansson: yes, there is a strong current between Africa and Madagascar that makes the crossing very difficult (although Nile crocs have apparently made it relatively recently). Malagasy language has some words from languages of South Asia, so it appears that the settlement of Madagascar by people of Asian origin was a result of a long migration with many stops. They were perhaps somewhat similar to modern “sea gypsies” culturally.

    Link to this
  22. 22. Jenny Islander 3:59 pm 05/24/2014

    @Gigantala: Perhaps a lithornithid, or something like it, as a sort of magpie of the shoreline, pecking about for this and that and expanding its territory inland if there happens to be food available? And if there’s lots of food inland, but there aren’t any mustelids or other species that make life on the ground a very bad idea, then there’s no pressure to keep the wings, rather pressure to improve the legs.

    Makes me wonder if somewhere in an unexcavated fossil bed there’s a paleognath the size of a pitta.

    Link to this
  23. 23. leecris 4:32 pm 05/24/2014

    Do you plan to follow up the monitors t-shirt with one with the above image of ratites? If so, I’d be interested in having one…

    Link to this
  24. 24. josimo70 4:36 pm 05/24/2014

    Taking Madagascar as focus, three different landbridges were proposed: from East Africa, India and Antarctica (Gunnerus Ridge).
    Am Nat. 2006 Dec;168(6):730-41. Epub 2006 Oct 31.
    Vicariant origin of malagasy reptiles supports late cretaceous antarctic land bridge.
    Noonan BP1, Chippindale PT.

    Link to this
  25. 25. Gigantala 5:02 pm 05/24/2014

    @Jenny Islander: Indeed. There’s even something of a precedent for that, albeit with northern hemisphere Palaeogene lithornithids: the more basal Lithornis is rather ibis like, bearing a long bill and wings akin to those of soaring birds like vultures, while Paracathartes and Pseudocrypturus seem to have become more terrestrial, with shorter and thicker bills and wings more akin to those of tinamous and Galliformes.

    Link to this
  26. 26. LeeB 1 7:02 pm 05/24/2014

    The reference to Baker et al (2014) just below the picture of the Moa and Elephent bird skeletons isn’t shown in the reference list at the bottom; it would be interesting to know what this is.

    Hippopotamus got to Madagascar as well as crocodiles, but given the proximity to Africa the number of animals that made it over the Mozambique channel is rather small.
    And Psittacula parakeets are common on Indian Ocean islands suggesting that colonisation from Asia or further east is at least as easy as from Africa; in the early tertiary if islands on the ninety east ridge were above water this might have provided a stepping stone between Australia and the western Indian Ocean islands for flying birds.

    That the ancestral ratites were volant and became flightless numerous times is interesting; but perhaps at least as interesting is that flying ratites other than Tinamou are now non-existent.

    This might suggest that they are unable to compete successfully with other birds.

    Also given the number of ratite lineages colonising available continents in the early tertiary you could wonder if there were further ratites to be found on India, Antarctica or even on the Kerguelan microcontinent.

    And there are currently about three recognised species in Aepyornis and three or four in Mullerornis; given the changes in Moa taxonomy when recent studies were done I would expect the number of Elephant bird species to change when they are properly studied.

    LeeB.

    Link to this
  27. 27. Gigantala 7:13 pm 05/24/2014

    “This might suggest that they are unable to compete successfully with other birds.”

    For obvious reasons, we have little information about whereas neognaths outcompeted flying palaeognaths. There have been some suggestions in literature that “lithornithids” were replaced by scolopacids, but A) there’s a very large gap between the first unambiguous scolopacid and the last “lithornithids” and B) if things like Cimolornis are Charadriiformes (which probably aren’t, but let’s pretend for the sake of the argument that they are) then that means that “lithornithids” and conventional shorebirds have co-existed for a very long period of time.

    Plus there’s the fact that Proapteryx shows that flying, probing palaeognaths were alive for considerably longer than previously thought; Saint Bathans also has several fossils of Charadriiforme shorebirds, though no scolopacids if I’m not mistaken.

    Link to this
  28. 28. naishd 7:37 pm 05/24/2014

    Thanks to all for continuing comments, great stuff! Did you all know that Tet Zoo is (so far as I can tell) far and away the biggest bringer-in of comments on the whole SciAm network? Hmm. Anyway…

    leecris (comment # 23): sure, I’ll do a palaeognath t-shirt if people will buy it. Which images are you thinking of?

    LeeB1 (comment # 26): apologies, the missing Baker et al. reference is…

    Baker, A. J., Haddrath, O., McPherson, J. D. & Cloutier, A. 2014. Genomic support for a moa-tinamou clade and adaptive morphological convergence in flightless ratites. Molecular Biology and Evolution doi: 10.1093/molbev/msu153

    There is so much new palaeognath stuff I initially planned to cover here: the new papers on moa species-level diversity and phylogeny, the news on Miocene kiwi, the new work on the phylogenetic position of Emuarius

    Link to this
  29. 29. Allen Hazen 8:26 pm 05/24/2014

    Umm… Are Palaeognaths (including Lithornithids) united by non-palatal characters?
    Because I know more Greek than ornithology, and palaeo means old, suggesting that palaeognathy (the osteological condition the critters get their name from) is (in the broad perspective of supra-ordinal avian phylogeny) a plesiomorphic character…
    Does this thought … make the placement of Lithornithids seem even MORE problematic?

    Link to this
  30. 30. Allen Hazen 8:31 pm 05/24/2014

    Umm, again. Maybe I posted too soon. The Wikipedia article “Neognathae” says
    “Neognathae” means “new jaws”, but ironically it seems that the supposedly “more ancient” paleognath jaws are among the few apomorphic (more derived) features of this group as compared to the neognaths.”

    (Mind you, as the Wikipedia bureaucracy helpfully points out (citation needed), there is no source given for this particular remark in the article.

    Link to this
  31. 31. Heteromeles 8:38 pm 05/24/2014

    Dumb idea, but if we’re talking about dispersal and subsequent extinction of flying lithornithids, are we perhaps talking about a clade that dispersed along the ancient Tethys? Perhaps the reason that they disappeared is that the Tethys disappeared too? Or are the fossils all wrong for that?

    My thought is that the Tethys may have been a dispersal pathway, rather than a barrier, if we’re talking about a group of island-hopping birds. Most of the islands the clade was on have disappeared (into Europe, for example), but a few survived here and there.

    Rails and pigeons do the same thing now, but the insular rails are on much more modern islands. Perhaps they are latecomers to island-hopping?

    Link to this
  32. 32. Gigantala 8:46 pm 05/24/2014

    ^”Lithornithids” also occur in North America, though the most recent forms do indeed come from Europe (specifically, Messel and similar sites).

    @Allen: The palaeognath condition was so named because of a misconception of ancestrality in comparision with the neognath palate. Some Cretaceous ornithurine groups do have a similar palatal anatomy, which helped with that misconception.

    Link to this
  33. 33. Anthea Fleming 9:17 pm 05/24/2014

    A small comment on Rhea – I once saw a displaying male in Melbourne Zoo. It was dancing, with its wings extended sideways, showing that they are quite long and well-feathered – as it waved them, they bent and flexed to a remarkable extent, showing that the outer bones (fore-arm etc.) are reduced to a cartilaginous framework. Obviously, they would be unlikely to fossilize. The display was very beautiful. This display was quite Ostrich-like, but much more graceful.

    Link to this
  34. 34. vdinets 11:30 pm 05/24/2014

    What I find interesting is that there are other lineages (i. e. hummingbirds) that are now mostly confined to the Neotropics, having gone extinct elsewhere, but are clearly capable of competing with groups of non-Neotropic origin.

    Link to this
  35. 35. JoseD 11:31 pm 05/24/2014

    Yodelling Cyclist: “Awesome article/series. What’s with the teenage mutant ninja cassowaries leaping into battle in the illustrations? ;-)

    I was wondering the same thing, given that it looks just like 1 of the cassowaries in this link.

    Link to this
  36. 36. leecris 12:24 am 05/25/2014

    @Darren When I click on it, this is the name shown for the image I’d like on a shirt: ratite-montage-2-Mar-17-2014-600-px-tiny-Mar-2014-Darren-Naish-Tetrapod-Zoology.jpg

    Link to this
  37. 37. SRPlant 3:43 am 05/25/2014

    There was a very natty ratite discovered recently in a Parisian apartment that had been boarded up for 68 years;

    http://moco-choco.com/2014/01/16/a-parisian-apartment-frozen-in-time-for-70-years/

    Ostremu?

    Link to this
  38. 38. MattPhillips 5:41 am 05/25/2014

    Nice article, and interesting that the Harshman paper had Moa/Tinamou, but then left it out. We did the same in a 2004 paper in MBE (Harrison et al). It was outside the scope of the paper and we didn’t want to court controversy.

    Somewhat similar, the Phillips et al 2010 Syst Biol (Moa/Tinamou) paper was the first to publish the elephant bird/kiwi relationship. “elephant bird …. best
    placement was as sister to kiwi (−lnL = 6007.35)”.

    I decided not to make a bigger deal out of it, knowing more data was needed – which Mitchell et al now provide. Hindsight is unhelpful.

    Link to this
  39. 39. DavidMarjanovic 5:53 am 05/25/2014

    The closest relatives of the Malagasy language are spoken in southern Borneo.

    people during the late 1800s/early 1900s had very broad views on what was possible – it being thought that convergence was more rampant within Tetrapoda that we think it is today, for example

    I’m not sure if that’s a separate phenomenon. From a few decades later, take the widespread idea that owls and “caprimulgiforms” must be closely related because a relationship between owls and “falconiforms” would just be too obvious, or the widespread idea that rodents and lagomorphs likewise can’t be close relatives for the same reason.

    the Kerguelan microcontinent

    AFAIK, it’s purely oceanic crust, like Iceland.

    Did you all know that Tet Zoo is (so far as I can tell) far and away the biggest bringer-in of comments on the whole SciAm network?

    This is my complete lack of surprise.

    Link to this
  40. 40. Tayo Bethel 7:58 am 05/25/2014

    This is just a pet theory of mine, but perhaps modern ratites evolved from the Galliform analogs rather than the snipe analogs.

    BtW, tinamous are weird. A really fascinatinggroup of birds.

    Link to this
  41. 41. Tayo Bethel 8:50 am 05/25/2014

    @DavidMarjanovic
    A bit off-topic, but where can I find the latest news on owl taxonomy, do you know? Have the relationships of Strigiformes to other orders resolved to anyone’s satisfaction?

    Link to this
  42. 42. Gigantala 9:11 am 05/25/2014

    @Tayo: Well, currently we can’t know for sure, for obvious reasons. Based on the “lithornithid grades”, you are probably right, since Paracathartes and Pseudocrypturus DID resemble more tinamous than the ibis like Lithornis and kin.

    However, given Proapteryx and such, it is equally possible that snipe-like body types were retained across the ratite phylogenetic tree, and that tinamou-like forms evolved independently multiple times (and subsequently became the classical ratites).

    Link to this
  43. 43. John Harshman 9:48 am 05/25/2014

    #41: No, they haven’t. Owls are “land birds” (sometimes called Telluraves), but other than that there’s nothing.

    Now if you want to know about relationships within owls, that’s better known. This, for example.

    Link to this
  44. 44. Yodelling Cyclist 10:46 am 05/25/2014

    Phylopessimism is a great term. I have a strong fear that mesozoic dinosaur phylogenies are as wrong as mammal phylogenies were pre-DNA work. Thing is, we’ll never really know.

    Link to this
  45. 45. Zoovolunteer 2:28 pm 05/25/2014

    One other odd feature of many ratites is that the male often has more involvement in incubation than in other birds. In many cases, even the polygynous species of tinamous, most of the incubation is by the male, and females compete for access to a male and his nest site rather than males competing for females. Compared with almost all other birds, where either a pair shares nest care equally or the female does most or all of the incubation and brood care, this seems rather odd. Any ideas why?

    Link to this
  46. 46. Gigantala 4:06 pm 05/25/2014

    May or may not be an ancestral trait to Maniraptora, since troodonts and oviraptors found to incubate the eggs are also male*.

    * Yes, there may be a troodontid pair, but as far as we know it could be a male + male couple. Homosexuality does exist in birds, after all.

    Link to this
  47. 47. LeeB 1 6:12 pm 05/25/2014

    David, the kerguelan microcontinant is formed of oceanic crust like iceland; but like iceland parts of it were subaerial for a long time; from the Cretaceous until well into the Tertiary.

    There are the remains of wood from large trees found on Kerguelan island itself and from drillholes on the offshore plateau.

    So if there were forests there in the early Tertiary there is the potential for large flightless birds.

    LeeB.

    Link to this
  48. 48. adzebill 7:43 pm 05/25/2014

    I had a look at the Paris elephant bird material, and knowing what we do now about the amount of sexual dimorphism in similarly-sized moa, I wouldn’t be surprised if there turned out to be only one species of Aepyornis and one of Mullerornis.

    Link to this
  49. 49. Gigantala 7:56 pm 05/25/2014

    In fact, such has been proposed several times, though more like “smaller species represent juveniles” rather than “extreme sexual dimorphism”.

    Link to this
  50. 50. Jerzy v. 3.0. 5:23 am 05/26/2014

    Re: male parental care
    In ratities and non-avian dinosaurs it may be the effect of parental brooding plus large clutch size plus precociality.

    Precocial birds have mechanism which slows and speeds egg development in a clutch, so all the chicks hatch at once. Don’t know details but it seems fascinating. In ostriches there is equally interesting phenomenon, that male and females must invest in incubating eggs which they cannot be sure are their own. Evolutionary biology goes dodgy.

    Link to this
  51. 51. Tayo Bethel 9:01 am 05/26/2014

    Have any hypotheses on the evolution of male parental care in ratites and (possibly) other maniraptorans been put forward? Since data on parental care in in more basal archosaurs is lacking, how early archosaurs got down tothe parenting business is probably the most mysterious part of their biology. Inliving crocodilians, females take care of these matters, with occasional helping being reportedby territorial males. So how mighta completely reversed system have arisen?

    Link to this
  52. 52. Jerzy v. 3.0. 11:05 am 05/26/2014

    @Tayo Bethel
    I would be also interested, for I have one.

    Anyway, I would propose checking whether nests of some larger dinosaurs contained eggs with different size, shape and/or texture which were laid by more than one female. Citipati is smaller than the Ostrich, but still. :)

    Link to this
  53. 53. John Scanlon FCD 11:11 am 05/26/2014

    Re phylopessimism: it is definitely a vice. Phylo-optimism (that’s a word now) never stopped anyone from testing a hypothesis, or all hypotheses.
    When I was presenting lab tours of Riversleigh fossils including Emuarius and dromornithids, I often talked about how the new view of drom relationships (anseriform not ratite) led to a huge difference in their reconstructed appearance, particularly in showing them with pennaceous rather than filamentous feathers. I’d explain that the similarly stringy feathers of emu, cassowary and kiwi are derived from more normal plumes as seen in ostrich, rhea, tinamou and neognaths. It was an excellent point (the EPB), though a bit of tangent (my demos often went for an hour at a time and were never the same twice).
    But what now? We have moa feathers that look intermediate in morphology (Rawlence et al. 2009, doi: 10.1098/rspb.2009.0755), but what do we know of aepyornithid feathers? (Nothing?) And as flightlessness is hardly consistent with oceanic dispersal, the kiwi/moa/casuariid conditions must all be independently derived anyway… apparently?

    Link to this
  54. 54. irenedelse 11:50 am 05/26/2014

    @ BrianL, Sebastian Marquez:

    In addition to Palaeotis, there is another possible ancient ratite from the Paris basin, Remiornis. But it’s only known from partial remains and not much is certain about this taxon AFAIU.

    Link to this
  55. 55. Yodelling Cyclist 12:07 pm 05/26/2014

    The point about phylopessimism is that tests conducted purely on the basis of morphology may very well be wrong. Darren discusses a suit of alternative phylogenies for rattites above, all suggested prior to DNA work, all/most of which were at least plausible, and all of which were dead wrong in the light of the molecular evidence.

    For species/groups which are sufficiently recent for molecular work to be conducted, there may well be reason for phylo-optimism. But non-avian dinos, to choose a popular example, are not and the more (apparent, but well founded) madness that emerges from the phylogenies of extant mammals and birds the worse it gets. Phylo-optimists may state that with sufficient characters these relationships can be resolved morphologically, but a great many extinct species are known from fragmentary remains, and many critical points along a hypothetical evolutionary trajectory may never be observed for taphonomic reasons (think proto-pterosaurs or bats – too small, too light and inhabiting (probably) forest habitats).

    In fact rattites maybe another example of this – their hypothetical highly volant ancestors may have hopped their way around the southern hemisphere at high latitudes. The fossils required to make sense of all this lost beneath the ice cap, sunk islands and simply islands where the taphonomy was poor. Heck, even Australia and New Zealand have huge chunks of the Cenozoic for which the fauna are very poorly known.

    Break out the pessimism, and pass the bottle.

    Link to this
  56. 56. Yodelling Cyclist 12:33 pm 05/26/2014

    Oh, and why do proto-rattites have to fly? Penguins dispersed by swimming. Maybe rattites are super oceanic loons. ( I jest – a bit.)

    Link to this
  57. 57. Gigantala 12:46 pm 05/26/2014

    To be fair, flying palaeognaths are represented as much as any clade of small birds is: only spotted in fossil sites with a high degree of preservation. Small flying birds aren’t a feature in several Cenozoic fossil sites, for instance.

    It is entirely reasonable that more flying palaeognaths will show up in Eocene to early Miocene fossil sites eventually, as Proapteryx and the Messel lithornithid have.

    Link to this
  58. 58. ectodysplasin 7:39 pm 05/26/2014

    Worth pointing out that “phylopessimism” and “phylo-optimism” are realy two sides of the same coin. In both cases, you’re optimizing one set of characteristics at the expense of another set of characteristics.

    In a lot of cases, phylopessimism may be the result of taxonomic inertia. Pre-phylogenetic workers came up with hypotheses for the origins of major groups, and many had pet hypotheses about the origins of certain characteristics. Phylogenetic work simply codified those interpretations. In some cases, the original hypotheses have never been rigorously tested.

    Link to this
  59. 59. DavidMarjanovic 4:53 am 05/27/2014

    #41: No, they haven’t. Owls are “land birds” (sometimes called Telluraves), but other than that there’s nothing.

    Oh. I was going by Gerald Mayr’s morphological phylogenies; on the molecular side there’s nothing.

    I have a strong fear that mesozoic dinosaur phylogenies are as wrong as mammal phylogenies were pre-DNA work.

    I don’t. That’s because most pre-molecular mammal phylogenies weren’t made by phylogenetic analysis – they were based on evolutionary scenarios and “reliable characters” (or “the most reliable characters”). That’s also (among the reasons) why they were so poorly resolved.

    It’s a coincidence that DNA and phylogenetic analysis have come to mammal phylogeny at the same time.

    troodonts and oviraptors found to incubate the eggs are also male

    Reference, please; do they lack medullary bone?

    the kerguelan microcontinant is formed of oceanic crust like iceland

    That makes it not a continent, is my pedantic point.

    I agree of course that it may well have had its own flightless birds.

    Worth pointing out that “phylopessimism” and “phylo-optimism” are realy two sides of the same coin. In both cases, you’re optimizing one set of characteristics at the expense of another set of characteristics.

    In phylopessimism, you don’t optimize anything: you notice that the congruence between sets of characteristics is too low for your liking, so you despair, exaggerate all uncertainties and assert nothing can be known to any reasonable degree.

    Romerograms where each bubble turns (well, backwards in time) into a stippled line that bends but ends in a question mark before reaching any other line or bubble: that’s phylopessimism.

    Link to this
  60. 60. Gigantala 10:14 am 05/27/2014

    Varricchio et al. (2008) specifically pointed the absence of medullary bone in incubating Troodon specimens, for starters. Likewise, the Citipati specimens have also been established as male.

    Geoffrey et. al (2013) more recently found a couple of Troodon in association with a nest, but in true heteronormative fashion they did not test for medullary bones and assumed it was a male/female pair willy nilly.

    Link to this
  61. 61. Yodelling Cyclist 10:47 am 05/27/2014

    “…despair, exaggerate all uncertainties and assert nothing can be known to any reasonable degree.”

    True, but in my own discipline (physical chemistry) I’ve far too often seen the absolute opposite – model for a reaction proposed, discrepancies waved aside as mere details only to have it come right back at you that the mere trifle you dismissed absolutely and massively disproves the model.

    While I’m sure absolute, rampant phylopessimism maybe going too far, there has to be fear, floating around that dismissing uncertainties is in fact picking and choosing what is important and what is not, and this is perhaps most true for extinct taxa known from fragmentary remains for which there are no molecular data.

    Link to this
  62. 62. Yodelling Cyclist 11:00 am 05/27/2014

    Actually, that’s a point that would be very interesting to have answered: how different would in terms of suggested evolutionary relationships would romerograms circa 1985, say, and a modern cladograms be for the broad group of non-avian dinosaurs and a similar pair for mammals (and a further pair for cenozoic birds)?

    My understanding may well be wrong, and if so I sincerely apologise, but if one is revolutionised by the inclusion of molecular data while the other is not (due to the lack of such data), then doesn’t that at least hint that the other may well be just as wrong? Or were mammalogists just really …wrong.. while previous dinosaur palaeontologists found just the right features to assign evolutionary relationships?

    Link to this
  63. 63. vdinets 1:10 pm 05/27/2014

    Jerzy: I don’t think you can get the number of females from egg variability. You’d have to know the extent of individual variability per female to do that. Also, it is entirely possible that there were brood parasites among dinosaurs.

    Link to this
  64. 64. ectodysplasin 1:42 pm 05/27/2014

    @David,

    In phylopessimism, you don’t optimize anything: you notice that the congruence between sets of characteristics is too low for your liking, so you despair, exaggerate all uncertainties and assert nothing can be known to any reasonable degree.

    Romerograms where each bubble turns (well, backwards in time) into a stippled line that bends but ends in a question mark before reaching any other line or bubble: that’s phylopessimism.

    Sure, but that’s still based on optimization of certain characters, even if it’s all eyeballing it.

    An example.

    The tetrapod diphyly hypothesis of Holmgren is based on optimization of a few characteristics of the nasal capsule that Holmgren considered important. Holmgren discarded a bunch of obvious tetrapod synapomorphies because he thought the differences in nasal capsule morphology were so important.

    Mammalian or cynodont diphyly is a result of apparent rampant convergence between some therocephalians and some cynodonts (secondary palate, heterodonty, etc).

    Lissamphibian polyphyly is a function of all sorts of issues with lissamphibian origins, including a lack of good intermediate forms, convergence in dermal skull reduction, and some issues of convergence in the loss of the cavum internasale in frogs, caecilians, and amniotes. There are some additional issues at play, as well, but much of the historic debate on the subject exists in reference to comparative credibility of evolutionary hypotheses concerning one set of characters vs another.

    Your comments about Romer diagrams are well-taken, but it’s worth noting that in these situations, the dashed lines ending nowhere are a result of there being no character suite that the worker was willing to hang their hat on.

    Link to this
  65. 65. ectodysplasin 2:00 pm 05/27/2014

    @YC:

    My understanding may well be wrong, and if so I sincerely apologise, but if one is revolutionised by the inclusion of molecular data while the other is not (due to the lack of such data), then doesn’t that at least hint that the other may well be just as wrong? Or were mammalogists just really …wrong.. while previous dinosaur palaeontologists found just the right features to assign evolutionary relationships?

    Mammal phylogeny suffers from the fact that mammals all started out pretty small and a few lineages got really large and specialized, whereas a few other lineages remained small and plesiomorphic. So there’s some long-branch issues (between scandentians, macroscelids and ‘insectivores’ especially) as well as some convergence issues (between elephants and ungulates). Some of this may be the result of major diversification in the post-KPg. The morphology itself was intractable for these reasons. The molecules were able to provide an answer to questions where the morphology was insuffient.

    With non-avian dinosaurs, we have a bit better idea of the morphology in most cases. There’s no massive rapid diversification of all major clades. I have my suspicions that ornithiscians might have some issues, but everything else seems to be relatively robust.

    Link to this
  66. 66. Jenny Islander 4:40 pm 05/27/2014

    @LeeB1 & DavidM.: Oh, awesome. I looked up the Kerguelen Microcontinent at Basement Geographer and learned that it is “three times the size of Japan” and had coniferous forests. This means that it’s easily three times the size of New Zealand, with a comparable climate before its most recent submergence 20 mya. Wow. That is a heck of a lot of room. And as far as I can tell from Miocene world maps, it was too far from hospitable shores for rafting to allow for many survivors, although some reptiles may have made it.

    Birds, spiders, insects, bats, possibly amphibians, maybe some lizards . . . I can’t find data on the length of time between the most recent and second most recent submergences, but there might have been time for some spectacular radiations.

    Link to this
  67. 67. LeeB 1 6:37 pm 05/27/2014

    David, sorry I misinterpreted your comment, it’s just that people commonly call it a microcontinent when they discuss such things.

    Jenny, it’s bigger than New Zealand now, but New Zealand (and New Caledonia)have been bigger in the past.
    Look up the size of Zealandia in the early tertiary, and then consider that most of it went underwater by the earliest Miocene.
    Exactly how much of Zealandia was above water at any particular time seems unclear but certainly more than the current extent of New Zealand.

    LeeB.

    Link to this
  68. 68. Jenny Islander 8:19 pm 05/27/2014

    @LeeB1 no. 67: Yes, I’ve seen Zealandia as well. Is the consensus that the recent fauna of New Zealand are remnants of a, how would you say it, Zealandia assemblage? Or that they aren’t, because at one point New Zealand was some shoals and a rock or two? Or is this also in dispute?

    Link to this
  69. 69. LeeB 1 12:03 am 05/28/2014

    This is still disputed.
    A lot of the flora and fauna have links to that of Australia in the Tertiary including such things as short-tailed bats and Podocarp and Kauri trees; this suggests they could have dispersed from Australia as new Zealand re-emerged from the sea; but other fauna like the fossil mammal and the Tuatara are sufficiently distinct that it looks like there must have been enough islands above water that they could survive even during the period of maximum submersion of Zealandia.

    So I suspect the modern fauna is a mix of original old flora and fauna and more recent immigrants, both from Australia and potentially also from New Caledonia.

    LeeB.

    Link to this
  70. 70. Jerzy v. 3.0. 6:21 am 05/28/2014

    @vdinets
    If there is enough variability in the sample, you can check if egg size (or whatever) distribution is unimodal or multimodal.

    Nest parasitism in dinosaurs – it was proposed for Citipati and Byronosaurus, and occurs in most todays egg-laying groups. Sounds like cool idea which not made it to All Yesterdays.

    Link to this
  71. 71. DavidMarjanovic 7:05 am 05/28/2014

    Mammal phylogeny suffers from the fact that mammals all started out pretty small and a few lineages got really large and specialized, whereas a few other lineages remained small and plesiomorphic. So there’s some long-branch issues (between scandentians, macroscelids and ‘insectivores’ especially) as well as some convergence issues (between elephants and ungulates). Some of this may be the result of major diversification in the post-KPg. The morphology itself was intractable for these reasons. The molecules were able to provide an answer to questions where the morphology was insuffient.

    I disagree. The most recent phylogenetic analyses of morphological data (not all of them published yet) have been getting results that are more and more similar to the molecular trees – O’Leary et al. (2013) being the big exception, as is Livezey & Zusi (2007) for birds. It really looks like the difference isn’t molecules vs. morphology, but phylogenetic analysis vs. its absence. The trick here is that making a morphological data matrix for phylogenetic analysis of Placentalia is easily a PhD thesis, while making a molecular one is a matter of weeks.

    Hypotheses of dinosaur phylogeny have changed a lot since 1985. Saurischian monophyly is now taken for granted, while back then it was rather a minority opinion; the “Carnosauria”/”Coelurosauria” dichotomy of Theropoda has been trounced (stepwise); birds are dinosaurs beyond reasonable doubt, while back then the skeptics still had a few intriguing arguments; the “Prosauropoda” mess has been largely sorted out (not completely, though); sauropod phylogeny is now quite well understood, while back then it was all very wishy-washy and differed in important features like the position of the titanosaurs; the understanding of ornithischian phylogeny was in an even worse state, while now it has greatly advanced (even though I agree with ectodysplasin that the root of the tree and the former “hypsilophodontids” still haven’t been sorted out).

    Sure, but that’s still based on optimization of certain characters, even if it’s all eyeballing it.

    You’re right. Phylogenetics with one or two characters used to be common; phylopessimism is more a reaction to the fact that this doesn’t work.

    So I suspect the modern fauna is a mix of original old flora and fauna and more recent immigrants, both from Australia and potentially also from New Caledonia.

    This seems to be the very recent consensus.

    Link to this
  72. 72. DavidMarjanovic 7:06 am 05/28/2014

    …The monophyly of Dinosauria itself was a rarely held opinion in 1985!

    Link to this
  73. 73. ectodysplasin 11:27 am 05/28/2014

    Chatterjee’s “Tyrannosaurids evolved from Postosuchus” hypothesis notwithstanding, which other authors supported dinosaur polyphyly?

    Link to this
  74. 74. ectodysplasin 11:49 am 05/28/2014

    @David,

    I disagree. The most recent phylogenetic analyses of morphological data (not all of them published yet) have been getting results that are more and more similar to the molecular trees – O’Leary et al. (2013) being the big exception, as is Livezey & Zusi (2007) for birds. It really looks like the difference isn’t molecules vs. morphology, but phylogenetic analysis vs. its absence. The trick here is that making a morphological data matrix for phylogenetic analysis of Placentalia is easily a PhD thesis, while making a molecular one is a matter of weeks.

    Well, with the O’Leary analysis, it’s a big and comprehensive dataset. The issue is that there’s a great deal of character atomization, and a lot of the characters involve implicit hypotheses about how certain characteristics evolve. Other morphological analyses make use of different character diagnoses and may not atomize the characters to the same degree. However, this is a good example of the research inertia problem I mentioned previously; a lot of the character diagnoses are based on pre-phylogenetic hypotheses of evolutionary series made by people like Malcolm McKenna, and have simply been implemented in a phylogenetic matrix. The hypotheses themselves are not really being tested. This goes for a bunch of early tetrapod stuff as well, where the character selection as it is implemented in a lot of analyses reflects existing biases of pre-phylogenetic workers and thus assumes certain results.

    Link to this
  75. 75. Andreas Johansson 1:50 pm 05/28/2014

    ectodysplasin wrote:
    Chatterjee’s “Tyrannosaurids evolved from Postosuchus” hypothesis notwithstanding, which other authors supported dinosaur polyphyly?

    I dunno if I’ve ever heard of an actual polyphyletic origin hypothesis before, but I recall popular books saying things like “dinosaurs make up two extinct orders of the Archosauria”, which in retrospect sounds like deliberately avoiding implying anything either way about whether they’re more closely related to one another than to other archosaurs.

    Link to this
  76. 76. John Harshman 6:53 pm 05/28/2014

    Every BANDit or MANIAC supports dinosaur polyphyly in the same way Chatterjee does.

    And Romer clearly was into polyphyly. For example, from the 1966 edition of Vertebrate Palontology: “Further, the dinosaurs were not a single group but were already divided at their first appearance into two distinct stocks, related only in that both were archosaurs descended from the primitive ruling reptiles, the Thecodontia. The term “dinosaur” is thus one which can be used only in a popular sense; scientifically the dinosaurs are arrayed in two separate orders, defined below.” And “The dinosaurs are divided into two orders, the Saurischia and the Ornithischia, both included in the subclass Archosauria but no more closely related to one another than to the other members or descendants of the ruling reptile group — the crocodiles, pterosaurs, and birds.” As was the custom in those days, he gave no justification for his conclusion.

    Link to this
  77. 77. ectodysplasin 7:13 pm 05/28/2014

    @John Harshman

    Every BANDit or MANIAC supports dinosaur polyphyly in the same way Chatterjee does.

    Not really. Chatterjee was exploring diversity of crurotarsans and seeing some very dinosaur-like morphologies in croc-line archosaurs. It took a while for folks to figure out what weirdos like Shuvosaurus actually were because they do show some serious convergence with derived theropod groups. Same goes with isolated fragments, such as teeth of Revueltosaurus. Until the diversity of crurotarsan morphology became more widely understood, it made reasonable sense to take some of that morphological similarity as a given.

    AFAICT, the dinosaur-bird denialists are a different breed, draw different lines (or refuse to draw lines) and generally point to hypothetical transitional series as “proof” that the phylogenies are all wrong. That’s not what Chatterjee was doing. Chatterjee was trying to understand some very dinosaur-like crurotarsans before we had the context necessary to make sense of these animals.

    Link to this
  78. 78. John Harshman 8:00 pm 05/28/2014

    Not really.

    By “in the same way” I meant that Chatterjee thought some animals traditionally considered to be dinosaurs were not really dinosaurs. That’s all.

    Link to this
  79. 79. ectodysplasin 8:27 pm 05/28/2014

    Fair enough. I just feel like Chatterjee deserves better than to be lumped in with dino-bird denialists, when his mistakes were understandable given the state of Triassic archosaur paleontology in the 70s and 80s.

    Link to this
  80. 80. naishd 9:13 pm 05/28/2014

    Honorary mention of Alan Charig, who was promoting/supporting dinosaur polyphyly during the 1980s and, I think, the 1990s too. I can recall a time when the concept of dinosaur monophyly – promoted by Bakker, Benton and others – was seen as new and ‘contra mainstream’.

    Link to this
  81. 81. JoseD 11:15 pm 05/28/2014

    DavidMarjanovic: “Reference, please; do they lack medullary bone?”

    Varricchio et al. 2008 & yes. However, Birchard et al. have since shown that “the study in question failed to account for factors known to affect egg and clutch size in living bird species. A new scaling analysis of avian clutch mass demonstrates that type of parental care cannot be distinguished by conventional allometry because of the confounding effects of phylogeny and hatchling maturity. Precociality of young but not paternal care in the theropod ancestors of birds is consistent with the available data” ( http://rsbl.royalsocietypublishing.org/content/9/4/20130036.full ).

    Link to this
  82. 82. Andreas Johansson 3:39 am 05/29/2014

    Romer wrote:
    “The dinosaurs are divided into two orders, the Saurischia and the Ornithischia, both included in the subclass Archosauria but no more closely related to one another than to the other members or descendants of the ruling reptile group — the crocodiles, pterosaurs, and birds.”

    That strikes me as a rather curious statement. Presumably he didn’t think crocs, pterosaurs, birds, saurischians, and ornithischians formed a hard polytomy, but that there was a true branching order, even if we didn’t or couldn’t know it?

    Link to this
  83. 83. John Harshman 9:08 am 05/29/2014

    What Romer meant, I imagine, is that they all emerged from the amorphous lump on the bubble diagram named “thecodonts”, and that’s all you need to know. But we can’t ask him.

    Link to this
  84. 84. DavidMarjanovic 9:39 am 05/29/2014

    However, this is a good example of the research inertia problem I mentioned previously; a lot of the character diagnoses are based on pre-phylogenetic hypotheses of evolutionary series made by people like Malcolm McKenna, and have simply been implemented in a phylogenetic matrix. The hypotheses themselves are not really being tested. This goes for a bunch of early tetrapod stuff as well, where the character selection as it is implemented in a lot of analyses reflects existing biases of pre-phylogenetic workers and thus assumes certain results.

    Absolutely. Ruta & Coates (2007) is chock full of characters that must have been taken from diagnoses of established taxa.

    Honorary mention of Alan Charig, who was promoting/supporting dinosaur polyphyly during the 1980s and, I think, the 1990s too. I can recall a time when the concept of dinosaur monophyly – promoted by Bakker, Benton and others – was seen as new and ‘contra mainstream’.

    All seconded. I grew up with lots of popular books saying, basically, “there’s no evidence that Saurischia and Ornithischia are sister-groups [of course they didn't use that term!], so there’s no reason to think they were; therefore we should suppose that they were not sister-groups, while also not supposing that anything else we know is the sister-group of either”. Spot the logical fallacy of phylopessimism.

    Charig, though, wondered very loudly if the quadrupedal sauropods had descended from the quadrupedal rauisuchians (and kept saying so till he died). That’s not phylopessimism, it’s the “phylogenetics with one character” that phylopessimism was the reaction to.

    What Romer meant, I imagine, is that they all emerged from the amorphous lump on the bubble diagram named “thecodonts”, and that’s all you need to know.

    Phylopessimism.

    Link to this
  85. 85. John Harshman 11:33 am 05/29/2014

    Let’s not forget that G. G. Simpson defined “monophyletic” as “descending from a group of equal or lesser taxonomic rank”. So birds are monophyletic because they descend from Class Reptilia. They weren’t all that concerned with rigorous branching trees in those days.

    Link to this
  86. 86. SJCrum 5:18 pm 05/29/2014

    deleted

    Link to this
  87. 87. naishd 5:21 pm 05/29/2014

    Ok… now I realise why your comments were going immediately into my spam file. Back they go. Goodbye.

    Link to this
  88. 88. DavidMarjanovic 8:25 am 05/30/2014

    I dimly remember the name SJCrum, but not what they did. Creationism?

    G. G. Simpson defined “monophyletic” as “descending from a group of equal or lesser taxonomic rank”.

    o_O

    They weren’t all that concerned with rigorous branching trees in those days.

    That’s my point: they had been concerned with them earlier, in the times of Haeckel and Fürbringer.

    Link to this
  89. 89. DavidMarjanovic 12:56 pm 05/30/2014

    Oh. I just read that Proapteryx is only known from “an incomplete femur and a tentatively referred quadrate – not much to go on in terms of whether it was volant or not, except to say that the inferred size of this bird is consistent with it being volant.” Given the tiny size of the flightless Messelornis, Proapteryx may well have been completely flightless as well, there’s just no way to tell for sure yet!

    Link to this
  90. 90. Gigantala 5:11 pm 05/30/2014

    According to Göhlich et al 2013, though, the proportions of the femur are akin to those of flying birds (they used the Banded Rail as an example), so either it was volant or it was recently flightless.

    Link to this
  91. 91. DavidMarjanovic 8:34 am 05/31/2014

    The proportions of an incomplete femur in comparison to what? ~:-|

    Link to this
  92. 92. Gigantala 11:24 am 05/31/2014

    To quote Göhlich et al 2013:

    “The proportion of femur size to quadrate size seen in Proapteryx lies intermediate between those those observed in similar-sized but distantly related birds such as Banded Rail, Gallirallus philippensis, e.g., SAM B36299, and the Australian Little Bittern, Ixobrychus dubius[...]“

    Link to this
  93. 93. irenedelse 3:47 pm 06/1/2014

    Talking of ratites, some enterprising quacks are selling oil from emu fat as a cure-all:
    http://drjengunter.wordpress.com/2014/06/01/emu-oil-schmaltz-is-the-new-cure-for-everything/

    I knew of ostrich farms, but emu?

    Link to this
  94. 94. irenedelse 3:52 pm 06/1/2014

    Oh, wait. A little googling tells me that emu ranches are indeed a thing.

    Link to this
  95. 95. DavidMarjanovic 6:43 am 06/3/2014

    The proportion of femur size to quadrate size

    That’s assuming the “incomplete femur” and the “tentatively referred quadrate” belong not just to the same species, which is already “tentative”, but also to individuals of similar size.

    Ornithologists seem to routinely make such extrapolations. I don’t think they should.

    Link to this
  96. 96. vdinets 3:08 pm 06/5/2014

    Jerzy (#70) Thanks for the Byronosaurus tip, I’ve never heard of this!

    Link to this
  97. 97. Grehan1896 5:23 pm 06/23/2014

    A close relationship between the elephant birds and kiwis is only discordant with distribution only if one assumes that they have to be next door to each other. There is no biogeographic imperative for such an assertion. A Madagascar-New Zealand distribution no more negates a pre drift origin than for obligate subterranean freshwater fish relatives in Madagascar and Australia. The Madagascar-New Zealand relationship only shows that kiwis did not necessarily differentiate with respect to a known Australian relative.

    John Grehan

    Link to this
  98. 98. naishd 5:34 pm 06/23/2014

    John Grehan (comment # 97): a Madagascar-New Zealand distribution would require use of terrestrial links that disappeared during the Jurassic, a timing discordant with what we understand of the bird fossil record. I’m not sure that I understand the point you’re trying to make, but the main point of interest in the new study is that a Madagascar-New Zealand link is not consistent with overland dispersal or vicariance, but, rather, with over-water dispersal. My apologies if I’ve misunderstood what you’re saying.

    Link to this
  99. 99. Grehan1896 10:42 am 06/24/2014

    The assertion that the Jurassic timing is not discordant with what “we understand” of the bird fossil record is not correct. There is nothing discordant with the fossil record. A Jurassic timing does precede the oldest recognized ratite fossil, but the oldest fossil is just the oldest fossil, not necessarily the actual age of origin for a taxon. There are plenty of species that have no fossil record, but we do not say that they came into being yesterday.

    John Grehan

    Link to this
  100. 100. naishd 10:49 am 06/24/2014

    “The assertion that the Jurassic timing is not discordant with what “we understand” of the bird fossil record is not correct” (comment # 99, John Grehan)

    No, it is correct. Every piece of evidence we have shows that crown-bird evolution post-dated the end of the Jurassic by some reasonable margin. I think you might have a biased agenda.

    Link to this
  101. 101. Grehan1896 1:41 pm 06/24/2014

    What would you cite as evidence?

    Link to this
  102. 102. naishd 4:03 pm 06/24/2014

    The bird fossil record and divergence dates based upon it.

    Link to this
  103. 103. Grehan1896 4:35 pm 06/24/2014

    There’s the weak point in your contention. The fossil record for birds (or any other group for that matter) only records the oldest fossil. Fossils therefore only provide the minimal estimate of the age of a group. The same also applies to divergence dates which are also usually calibrated using fossils to molecular divergence dates are minimal only. The application of priors does not change that since there is no objective way to ascertain how much older a group is compared with its oldest fossil.

    In the case of ratites there is a biogeographic pattern that shows phylogenetic breaks matching the Atlantic Ocean, the Mozambique Channel, and the Pacific/Tasman Sea boundary. The principal crown ratite lineages are also allopatric with respect to each other except for the overlap occuring in New Zealand between the kiwi and moa linages. This pattern is concordant with a Mesozoic origin for ratites (certainly not a problem given recent discovery of duck related fossils and Cretaceous fossil parrots (any maybe other groups?).

    So I contend that my view that the fossil record does not preclude Mid-Mesozoic origins for the ratites is not falsified by your interpretation of the fossil record and molecular divergence estimates which misinterpret fossil dates as the actual or maximal age of origin. Conversely, the biogeographic correlation with tectonics does support a Mid-Mesozoic origin.

    John Grehan

    Link to this
  104. 104. naishd 4:51 pm 06/24/2014

    I will agree with you that “fossils … only provide the minimal estimate of the age of a group”.

    However, the application of Occam’s Razor means that we have to consider the numerous relevant data points we have (early members of the palaeognath, galloanserine and neoavian lineages, early members of the various stem-bird lineages, and so on) as relevant, not irrelevant, to the issue. After all, we’re supposed to construct hypotheses by compiling all available data points as best as possible. Furthermore, given data indicating that flight-capable behaviour was present not only in very early palaeognaths, but also throughout taxa across the palaeognath tree, the burden of evidence outweighs the historic view – the one you seem to be promoting – that extant ratite distribution is linked to vicariance.

    Link to this
  105. 105. irenedelse 5:30 pm 06/24/2014

    @naishd
    “Every piece of evidence we have shows that crown-bird evolution post-dated the end of the Jurassic by some reasonable margin.”

    @Grehan1896
    “This pattern is concordant with a Mesozoic origin for ratites”

    Even a non specialist like me can see at least one of the problems in your objection to the reconstruction of ratite origins detailed in the article: Mesozoic does not equal Jurassic. By a good margin.

    Link to this
  106. 106. Yodelling Cyclist 6:05 pm 06/24/2014

    I’d just like to thank previous commenters on taking the time to correct me previously in the comment thread.

    Btw, since various bird groups seem prone to hybridisation, is there any possibility that such events maybe complicating our understanding of ratites?

    Link to this
  107. 107. Grehan1896 6:10 pm 06/24/2014

    To naishd
    OK so what about these “data points” is evidence against a vicariance origin for ratites?
    How does inferred flight-capable behavior provide evidence that outweighs biogeographic evidence for vicariance.
    To irenedelse
    Agreed, Mesozoic does not equal Jurassic (Mid Jurassic). As in my earlier postings, the Mozambique break suggests a correlated Mid Jurassic origin for the break between ostriches and other ratites.

    John Grehan

    Link to this
  108. 108. Christopher Taylor 11:17 pm 06/24/2014

    The application of priors does not change that since there is no objective way to ascertain how much older a group is compared with its oldest fossil.

    Nevertheless, the recognition that a given fossil may not represent the true earliest representative of its lineage does not give carte blanche to project that lineage as indefinitely far as may be convenient.

    Fossil occurrences do not exist in a vaccuum. If ostriches diverged from other palaeognaths in the mid-Jurassic, then that implies that palaeognaths of some kind must have been present since at least since that time. They must have presumably also been reasonably widespread, if they were present on multiple continents during that time. In that case, why are there no ratite fossils from this period? Why, indeed, are there no fossils of crown-group birds of any kind from this period?

    It is theoretically possible that simply none were preserved, that ‘absence of evidence is not evidence of absence’. But this ignores the fact that other bird lineages are known from the Jurassic and Cretaceous fossil record, and sometimes in reasonable numbers. Why should enantiornithines be preserved in the fossil record, but not neornithines? Why confuciosornithids? Why apsaravids? Why hesperornithids? Why ichthyornithids? As I said, it is possible that neornithines just happened to fly under the radar. But it is also possible that I will be struck by a meteor before I finish this comment.

    There have been suggested neornithines identified from late Cretaceous deposits, but most of them are pretty rudimentary. In fact, I think that there may be only one well-supported neornithine known from before the K/Pg boundary: Vegavis (unless someone wants to defend Polarornis to me). And even then, I think David Marjanović may have mentioned something in the works questioning Vegavis‘ status as a neornithine. If that pans out (emphasis on ‘if’), then we’re back to the fossil record being not incompatible with crown-group birds not even appearing until after the end of the Mesozoic.

    Absence of evidence may not be evidence of absence, but it is evidence of the absence of evidence. And that’s not something we can a priori dismiss.

    Link to this
  109. 109. Grehan1896 9:59 am 06/25/2014

    Response to Christopher Taylor

    “Nevertheless, the recognition that a given fossil may not represent the true earliest representative of its lineage does not give carte blanche to project that lineage as indefinitely far as may be convenient.”

    True, but I was not making that assertion.

    “Fossil occurrences do not exist in a vacuum. If ostriches diverged from other palaeognaths in the mid-Jurassic, then that implies that palaeognaths of some kind must have been present since at least since that time.”

    Yes

    “They must have presumably also been reasonably widespread, if they were present on multiple continents during that time.”

    Yes.

    “In that case, why are there no ratite fossils from this period? Why, indeed, are there no fossils of crown-group birds of any kind from this period?”

    This is a similar question to why do most species not have a fossil record? If one believes that there must be a fossil record if something exists then certainly the fossil record ‘proves’ there were no ratites before their earliest fossils.

    “It is theoretically possible that simply none were preserved, that ‘absence of evidence is not evidence of absence’. But this ignores the fact that other bird lineages are known from the Jurassic and Cretaceous fossil record, and sometimes in reasonable numbers. Why should enantiornithines be preserved in the fossil record, but not neornithines? Why confuciosornithids? Why apsaravids? Why hesperornithids? Why ichthyornithids?”

    Sure, that is an argument one may make – that if certain groups were present in the fossil record others must be. But is only a theoretical argument. The oldest kiwi fossil is only one million years old. Why should the kiwi not be preserved as older fossils when more ‘advanced’ birds are? If one takes the position that fossils must exist as the absolute criterion of existence then one is forced to ignore tectonic evidence. Why, for example, should the two mail clades of brown kiwi in New Zealand not only show an allopatric distribution, but one that is not only correlated with the Alpine fault (not the Alpine divide), but also dislocated in a manner that matches the 480 km of dextral strike-slip displacement that began 20 million years ago (and the kiwi pattern is known to be matched by at least 100 other taxa).

    “There have been suggested neornithines identified from late Cretaceous deposits, but most of them are pretty rudimentary. In fact, I think that there may be only one well-supported neornithine known from before the K/Pg boundary: Vegavis (unless someone wants to defend Polarornis to me). And even then, I think David Marjanović may have mentioned something in the works questioningVegavis‘ status as a neornithine. If that pans out (emphasis on ‘if’), then we’re back to the fossil record being not incompatible with crown-group birds not even appearing until after the end of the Mesozoic.”

    We are only back to an absence of fossils from that time.

    “Absence of evidence may not be evidence of absence, but it is evidence of the absence of evidence. And that’s not something we can a priori dismiss.”

    It is only absence of FOSSIL evidence, not necessarily that absence of evidence from other sources – such as biogeography.

    John Grehan

    Link to this
  110. 110. naishd 10:15 am 06/25/2014

    In response to comment # 109 (and in agreement with what Chris Taylor already said above), I would only repeat the point that we should construct hypotheses on the basis of as many data points as possible. There are numerous fossils that are relevant to this issue – the oldest kiwi fossil is not “only one million years old” as per comment # 109, there is also a kiwi at c. 20 Ma, and of course there are numerous other fossil palaeognaths going all the way back to the Paleogene (possibly the latest Cretaceous).

    These fossils are all consistent with a diversification that is predominantly or wholly post-Mesozoic, as is phylogenetic and anatomical data showing that flight-capable behaviour was the primitive condition for most palaeognath lineages, as is the pattern of the bird fossil record in general (the rarity of Late Cretaceous neornithines should be accepted as a genuine pattern, not simply arm-waved away*). So, these lines of evidence are inconsistent with an idea that is based on modern distribution alone.

    * With reference to Vegavis, the doubt about its position as a neornithine came from early results later included in Lee et al. (2014).

    Link to this
  111. 111. Grehan1896 10:36 am 06/25/2014

    “In response to comment # 109 (and in agreement with what Chris Taylor already said above), I would only repeat the point that we should construct hypotheses on the basis of as many data points as possible.”

    And so I asked for what are the data points you consider as evidence against a Jurassic origin for ratites.

    “There are numerous fossils that are relevant to this issue – the oldest kiwi fossil is not “only one million years old” as per comment # 109, there is also a kiwi at c. 20 Ma,

    Great to know that there is a fossil kiwi at 20 Ma (I did not know that). So does that mean kiwis originated at 20 Ma? No it does not. And the biogeographic evidence shows that the extant brown kiwi itself has at least a 20 Ma origin.

    “and of course there are numerous other fossil palaeognaths going all the way back to the Paleogene (possiblythe latest Cretaceous).”

    Sure.

    “These fossils are all consistent with a diversification that is predominantly or wholly post-Mesozoic,”
    Only if you read the fossil record as literal – that groups appear when the fossils appear, and that there can be no contradictory evidence from other sources.

    “as is phylogenetic and anatomical data showing that flight-capable behavior was the primitive condition for most palaeognath lineages,”
    Not sure I understand your argument about flight-capable behavior with respect to timing the origin of ratites. Please clarify.

    “as is the pattern of the bird fossil record in general (the rarity of Late Cretaceous neornithines should be accepted as a genuine pattern, not simply arm-waved away*).

    I’m not arm waving the rarity or absence away. I am arguing that biogeographic evidence suggests that ratites existed much earlier than their oldest recognized fossil.

    “So, these lines of evidence are inconsistent with an idea that is based on modern distribution alone.”

    Again, they are only inconsistent if one expects fossils to be present when a taxon is present. Just because one might have that expectation does not make it so.

    “* With reference to Vegavis, the doubt about its position as a neornithine came from early results later included in Lee et al. (2014).”

    Interesting that even when fossils do appear, there can be disputes over its identity.

    In summary, the position of naishd appears to be that fossils constitute evidence of absence of ratites in the Mid Jurassic because it is the expectation of naishd that if ratites were present so too would their fossils. In contrast I take the view that fossils only represent minimal dates and that biogeographic analysis involving spatial distributional and tectonic correlations (as shown by the kiwi example) provide an independent source of historical evidence of origins older than the oldest fossil.

    John Grehan

    Link to this
  112. 112. naishd 10:48 am 06/25/2014

    This is bad time for me to be involved in a discussion where I’m asked to provide specific data points — I’m a bit short on time, apologies. For dates on fossil palaeognaths that are all substantially younger than those we might regard as consistent with the sort of relatively ancient origin you’re positing, see the many sources cited above, plus the Lee et al. (2014) paper I keep mentioning, plus my review of the bird fossil record (Naish 2012).

    If the modern distribution of these birds is taken to be pre-eminent with regard to the timing of their evolution, and if fossils cannot be taken as a reliable guide to when divergences might have occurred, where does it end? Did primates, rodents, crocodiles and Trichechus manatees originate in the Cretaceous, since their distribution can ‘only’ be explained by the splitting of South America from Africa c. 100 Ma ago. Or… might it be more likely that dispersal was involved? Again, the argument has to involve the fact that, yes, dispersal does seem to have played a major role in the distribution of these birds.

    Lee, M. S. Y, Cau, A., Naish, D., Dyke, G. J. 2014. Morphological clocks in palaeontology, and a Mid-Cretaceous origin of crown Aves. Systematic Biology 63, 442-449.

    Naish, D. 2012. Birds. In Brett-Surman, M. K., Holtz, T. R. & Farlow, J. O. (eds) The Complete Dinosaur (Second Edition). Indiana University Press (Bloomington & Indianapolis), pp. 379-423.

    Link to this
  113. 113. Grehan1896 11:27 am 06/25/2014

    “This is bad time for me to be involved in a discussion where I’m asked to provide specific data points — I’m a bit short on time, apologies. “

    No worries. You referred to these data points so naturally I was interested to know the nature of your evidence.

    “For dates on fossil palaeognaths that are all substantially younger than those we might regard as consistent with the sort of relatively ancient origin you’re positing, see the many sources cited above, plus the Lee et al. (2014) paper I keep mentioning, plus my review of the bird fossil record (Naish 2012).”

    You keep referring to the fossil record as not being ‘consistent’ with the relatively ancient origin I (and others) propose. But it’s only inconsistent if you believe the fossils should have been found and recognized if the groups existed at an earlier time.

    “If the modern distribution of these birds is taken to be pre-eminent with regard to the timing of their evolution”

    Modern distribution is not necessarily ‘pre-eminent’, but it is a source of evidence independent of the fossil record and provides another potential measure of divergence age.

    “ and if fossils cannot be taken as a reliable guide to when divergences might have occurred,”

    They are reliable as the minimum age at which divergences might have occurred.

    “where does it end? Did primates, rodents, crocodiles and Trichechus manatees originate in the Cretaceous, since their distribution can ‘only’ be explained by the splitting of South America from Africa c. 100 Ma ago. Or… might it be more likely that dispersal was involved?”

    Ah, so now we come to the general issues. Biogeographically that answer regarding the origin of primates, rodents, crocodiles and manatees in the Cretaceous is yes. Primates in particular have been subject to a very detailed analysis showing how modern primate distributions within the continents are correlated with tectonic features that support Mesozoic and early Cenozoic origins. At various phylogenetic levels one sees a predominantly vicariant (allopatric) pattern of distribution which confounds the idea that they arose by chance dispersal.

    “Again, the argument has to involve the fact that, yes, dispersal does seem to have played a major role in the distribution of these birds.”

    Well here is your bias. But my bias, if you will, is that there is no evidence for dispersal to have played a major role in the distribution of each ratite lineage. I would go further to assert that it is a fiction – a created scenario based on a literal reading of the fossil record where the oldest fossils are treated as actual or maximal evidence of age of divergence.

    John Grehan

    Link to this
  114. 114. irenedelse 11:33 am 06/25/2014

    @John Grehan:

    “Interesting that even when fossils do appear, there can be disputes over its identity.”

    And?
    I’m sorry if this sounds flippant, but fossils don’t come with labels. And some times, the remains are ambiguous. It doesn’t mean that they are useless, or that one can interpret the fossil record any which way. Sometimes ambiguous or lacking data is just that. Using this paucity to suggest there might have existed Neornithes as early as the mid-Jurassic is, at the very least, premature.

    “In summary, the position of naishd appears to be that fossils constitute evidence of absence of ratites in the Mid Jurassic because it is the expectation of naishd that if ratites were present so too would their fossils.”

    Now, I wonder if we read the same comment. Especially with the article cited, Lee et al. (2014), on molecular clock data placing the origin of crown birds at mid-Cretaceous. Which doesn’t contradict the fossil record, since as you yourself say, the earliest fossil known isn’t the same as first member of a taxon… But mid-Cretaceous is not unreasonably far from what the fossil record shows either: one would expect that the earliest crown bird taxa would not have had a large population at the start, and hence left scarce fossils. So it’s not surprising or inconsistent with current theories that their first possible fossils come from late Cretaceous, and that they only “exploded” in the Cenozoic, after an extinction event that left a lot of ecological space for the few surviving taxa.

    In short, you don’t need extra hypotheses to explain modern birds, including Paleognaths.

    Link to this
  115. 115. naishd 12:03 pm 06/25/2014

    The fun continues. With reference to John Grehan’s comment # 113… “But my bias, if you will, is that there is no evidence for dispersal to have played a major role in the distribution of each ratite lineage.”

    Regarding my own position, I don’t think a case of pro-dispersal bias exists, since I actually think that each case has to be evaluated on its own merits – indeed, I have supported ancient vicariance events in the past (there is a whole chapter about this – it pertained to Michael Heads’s work on birds of paradise and New Guinea – in my 2010 Tetrapod Zoology Book One). But, as goes the idea that there is no reason to favour dispersal, absence of evidence has to be accepted as meaning something in historical scenarios.

    Furthermore, rather than being simple and neat, the notion that palaeognath distribution is explained by vicariance does not match well the shape of the cladogram – why would we see a pattern where (mostly African) ostriches and (South American) rheas diverge before an Australasian-Madagascan-South American clade, and why would Australasian taxa (cassowaries, emus, kiwi) group with Madagascan ones (elephant birds)? This is discordant with the actual pattern of continental breakup, and this smacks of a classic case of favouring a hypothesis no matter what the data points say.

    Link to this
  116. 116. Grehan1896 12:12 pm 06/25/2014

    “I’m sorry if this sounds flippant, but fossils don’t come with labels. And some times, the remains are ambiguous. It doesn’t mean that they are useless, or that one can interpret the fossil record any which way. Sometimes ambiguous or lacking data is just that. Using this paucity to suggest there might have existed Neornithes as early as the mid-Jurassic is, at the very least, premature.”

    I think we agree about the ambiguity of identifying fossils. My point is that this ambiguity or uncertainty can affect how the fossil record is read. But to be clear, I am not using the paucity (or apparent absence) or ratites in the mid-Jurassic. All I am saying is that biogeographic evidence suggests that the absence of fossils is an artifact of preservation or discovery rather than actual absence of the taxon.

    “Now, I wonder if we read the same comment. Especially with the article cited, Lee et al. (2014), on molecular clock data placing the origin of crown birds at mid-Cretaceous. Which doesn’t contradict the fossil record, since as you yourself say, the earliest fossil known isn’t the same as first member of a taxon… “

    Yes, any origin proposed before the oldest fossil does not contradict the fossil reocord.

    “But mid-Cretaceous is not unreasonably far from what the fossil record shows either:”

    What is unreasonable? Early Cretacous? Late Jurassic? Mid Jurassic. What is ‘reasonable’ is a personal statement of belief, not a scientific criterion.
    “one would expect that the earliest crown bird taxa would not have had a large population at the start, and hence left scarce fossils.”

    I have no prejudgment about that.

    “ So it’s not surprising or inconsistent with current theories that their first possible fossils come from late Cretaceous, and that they only “exploded” in the Cenozoic, after an extinction event that left a lot of ecological space for the few surviving taxa.”

    Or it would not be surprising or consistent with the biogeographic evidence that ratites (and its sister group) originated in the Jurassic and only diversified substantially in Cenozoic.

    In short, you don’t need extra hypotheses to explain modern birds, including Paleognaths.”

    I do not know what you mean by “extra hypothesis”. There appear to be two hypotheses for ratites – either they evolved at or about the age of their oldest known fossil, or some time earlier. In the latter case, biogeographic evidence indicates mid-Jurassic for the initial divergence and other divergence events associated with Mesozoic for the ratites.

    John Grehan

    Link to this
  117. 117. ectodysplasin 12:19 pm 06/25/2014

    Sure, a vicariance-only model can be tweaked enough to marginally explain what data we have. However, it does not explain the current data as well as a vicariance and dispersal model.

    Link to this
  118. 118. Yodelling Cyclist 12:23 pm 06/25/2014

    Mr. Grehan, the vicariance hypothesis is not supported by the DNA or morphological evidence discussed in the main body of the article, and for which Darren has provided the external links.

    Moving away from the fossil debate, which now seems to be chasing its tail a little, surely these are the results which must be attacked and overcome for vicariance to be regarded as the most plausible scenario once more.

    This is not just a matter of the presence/absence of fossil data.

    Link to this
  119. 119. Grehan1896 12:24 pm 06/25/2014

    ectodysplasin,

    Please explain how a “vicariance-only” model only marginally explains what “data we have” (and describe that data) and why it does not explain the current data as well as a vicariance and dispersal model (and please describe the vicariance and dispersal model you have in mind).

    John Grehan

    Link to this
  120. 120. irenedelse 12:27 pm 06/25/2014

    @John Grehan:

    “I am not using the paucity (or apparent absence) or ratites in the mid-Jurassic. All I am saying is that biogeographic evidence suggests that the absence of fossils is an artifact of preservation or discovery rather than actual absence of the taxon.”

    What you are saying, in effect, is that “biogeographic evidence” trumps up both fossil record and molecular data. Which sounds to me like an extraordinary claim.

    (And yes, indeed, I wrote “extra hypotheses” when I meant extraordinary. My bad, blame writing on the phone.)

    Link to this
  121. 121. naishd 12:29 pm 06/25/2014

    One more note (reference to comment # 119): I don’t think it’s fair to keep asking people to “describe the data” (comments # 107 and # 119) – this is what we go to the primary literature for! Again, see the refs cited above… The papers by Mindell et al. (2014) and Lee et al. (2014), among others, provide the data we’re talking about: the locations and ages of palaeognath and other bird fossils.

    Link to this
  122. 122. Grehan1896 12:39 pm 06/25/2014

    “What you are saying, in effect, is that “biogeographic evidence” trumps up both fossil record and molecular data. Which sounds to me like an extraordinary claim.”

    No, that is not a correct interpretation of my position. Darren has agreed with me that the oldest fossil does not preclude earlier origins. The question then becomes one of what other evidence might there be about the age of a taxon before its earliest fossil?

    One approach might be to use a sister group fossil. For example, recognized angiosperm fossils only go back into the early Cretaceous (at least the last time I looked, but I have not kept up with this question), but their gymnosperm relatives go back to the Triassic, suggesting that angiosperms diverged in the Triassic also and were present at that time even though they are not recorded or recognized in the fossil record.

    Another approach is to use biogeographic and tectonic correlations. If an extant group shows a phylogenetic break coinciding with a Mesozoic tectonic feature, for example, then the inference from biogeography is that the divergence dates from the time of that tectonic feature even if that date is older than the fossil record.

    So my view is that the biogeographic prediction does not trump the fossil record (or molecular data which amounts to the same thing since molecular divergence estimates calibrated using fossils only represent minimal divergence estimates) since the fossil record does not preclude an origin older than the older fossils.

    There is nothing inherently extraordinary about a mid-Jurassic origin, other than being extraordinary to popular belief.

    John Grehan

    Link to this
  123. 123. THoltz 12:44 pm 06/25/2014

    One wonders what panbiogeoraphers due to explain the existence of life on volcanic islands… Or, how taxa can know to disperse to volcanic islands, but know to avoid colonizing continental islands?

    Link to this
  124. 124. naishd 12:47 pm 06/25/2014

    John Grehan (comment # 122) says “There is nothing inherently extraordinary about a mid-Jurassic origin”. Assuming you’re referring to palaeognaths (as I think you are), the claim sure is extraordinary given that all other data-points indicate a mid-Jurassic origin for birds as a whole!

    Link to this
  125. 125. Yodelling Cyclist 12:48 pm 06/25/2014

    Ah, another good ol’ academic debate which leaves you exhausted and wanting a bear attack.

    For the record, bears may have evolved once, in the Permian, and their global distribution might be due to vicariance. We just haven’t found their fossils in Australia – yet.

    Link to this
  126. 126. Yodelling Cyclist 12:49 pm 06/25/2014

    Dammit, I should always re-read posts for phrasing. Please mentally substitute “craving” for “wanting”. It’s funnier.

    Link to this
  127. 127. Grehan1896 1:05 pm 06/25/2014

    “One wonders what panbiogeoraphers due to explain the existence of life on volcanic islands… Or, how taxa can know to disperse to volcanic islands, but know to avoid colonizing continental islands?”

    Assuming this is not just a rhetorical question, I would suggest reading the chapters on the central Pacific and the Hawaiian islands. There’s more than I expect readers would want me to go over to answer your question in a meaningful way here, but I would be happy to address any questions that arise if you wanted to contact me directly at Calabar.John@gmail.com

    Also, there is a paper I wrote on the Galapagos in 2001 you can obtain here.

    Link to this
  128. 128. Grehan1896 1:07 pm 06/25/2014

    Assuming you’re referring to palaeognaths (as I think you are), the claim sure is extraordinary given that all other data-points indicate a mid-Jurassic origin for birds as a whole!

    If “all other data points” (presumably you mean other than biogeographic data points) indicate a mid-Jurassic origin for birds as a while then there would appear to be nothing extraordinary about the basal lineages of extant birds also diverging about that time.

    John Grehan

    Link to this
  129. 129. Grehan1896 1:10 pm 06/25/2014

    “Ah, another good ol’ academic debate which leaves you exhausted and wanting a bear attack.
    For the record, bears may have evolved once, in the Permian, and their global distribution might be due to vicariance. We just haven’t found their fossils in Australia – yet.”

    I can sympathize with how such debates may seem to those not directly involved. There is no biogeographic evidence that I am aware of for bears evolving in the Permian.

    Link to this
  130. 130. Grehan1896 1:15 pm 06/25/2014

    Oops, forgot the citation. The chapters are in the book “Molecular panbiogeography of the Tropics. 2012. Michael Heads. University of California Press.

    Link to this
  131. 131. irenedelse 1:28 pm 06/25/2014

    “Assuming this is not just a rhetorical question”

    Note to self: don’t drink while reading Tet Zoo comments. I might want to use that keyboard again.

    “There is no biogeographic evidence that I am aware of for bears evolving in the Permian.”

    At this point, bear attacks sound rather attractive. ><

    Link to this
  132. 132. Yodelling Cyclist 1:29 pm 06/25/2014

    There is no biogeographic evidence that I am aware of for bears evolving in the Permian.

    Good to know, thanks. Then again,they are widely distributed and flightless. A Pangea origin should be considered.

    Link to this
  133. 133. irenedelse 1:33 pm 06/25/2014

    @YC:

    Because there are no flighted bears today, or at any point in the fossil record, doesn’t mean we couldn’t posit hypothetical flying bears. The past is wild with possibilities.

    Link to this
  134. 134. Grehan1896 1:54 pm 06/25/2014

    Of course as a matter of speculation, anything is theoretically possible. But in science there are the constraints of evidence. But of course the nature of evidence is not always that straightforward. Evidence in the form of the fossil record being treated as more or less the actual timing of divergence is a case in point. There are molecular divergence papers out there that acknowledge in the introduction that fossils only provide minimal dates, but then in their analysis they go on to treat fossils as providing actual or maximal dates.

    With respect to ratites it is not just an open ended speculation for a mid_Jurassic origin of ratites, but one based on evidence in the form of a biogeographic and tectonic correlation. This date is older than the oldest fossil and older than dates (such at 135 Ma) proposed for the origin of ratites (dates that are also older than the oldest fossil), but the date is not “inconsistent” with the fossil record as such. It is only inconsistent with the view that fossils necessarily provide an accurate date of origin for ratites.

    John Grehan

    Link to this
  135. 135. ectodysplasin 2:34 pm 06/25/2014

    @John,

    Of course as a matter of speculation, anything is theoretically possible.

    The issue is not the theoretical possibility of a hypothesis, but the credibility of the hypothesis.

    But in science there are the constraints of evidence. But of course the nature of evidence is not always that straightforward.

    This is correct, but it does not allow you to disregard entire swaths of data simply because they disagree with your interpretation.

    Evidence in the form of the fossil record being treated as more or less the actual timing of divergence is a case in point.

    Sure. But if you’re going to argue for a massive gap in the record of a number of lineages, it is up to you to justify that. “It is required by my model” is not sufficient justification.

    There are molecular divergence papers out there that acknowledge in the introduction that fossils only provide minimal dates, but then in their analysis they go on to treat fossils as providing actual or maximal dates.

    Two things.

    1) You seem to be under the impression that molecular methods have no addressed these issues with new methods. The reality is, the current molecular models (e.g. BEAST) do not treat these as hard numbers.

    2) In addition, the order of divergences is equally an issue, and one which you have not addressed. All molecular analyses suggest an order of divergence that is not in line with the order of continental breakup. So, you’re not only suggesting that the fossil record is anomalously gappy, but you’re also suggesting that the molecular phylogenies are wrong…..with no good reason for doing so.

    With respect to ratites it is not just an open ended speculation for a mid_Jurassic origin of ratites, but one based on evidence in the form of a biogeographic and tectonic correlation.

    What correlation? The only correlation you have is that all modern ratites are Gondwandan. That’s not a correlation. That’s one data point. If you try to correlate actual ratite clades with tectonics, that correlation collapses. Which is why panbiogeographers then argue that the phylogeny itself is wrong.

    This date is older than the oldest fossil and older than dates (such at 135 Ma) proposed for the origin of ratites (dates that are also older than the oldest fossil), but the date is not “inconsistent” with the fossil record as such. It is only inconsistent with the view that fossils necessarily provide an accurate date of origin for ratites.

    It’s also substantially older than the first neoavian fossil, the oldest ornithuran, and most avian diversity in general.

    When we talk about science and hypotheses, we talk about setting up direct tests of those hypotheses i.e. if we assume that a hypothesis is true, do additional forms of data or additional observations run counter to it or do they align with it? This is the fundamental test of any scientific hypothesis. If we assume panbiogeography to be true, we would expect various observations of molecular phylogeny and similarity that we do not observe. We would expect an antiquity of ornithurans and neoavians that we do not observe. And so on. Thus, we have to remain skeptical of the panbiogeography hypothesis, because so many lines of evidence run counter to it.

    Link to this
  136. 136. ectodysplasin 2:49 pm 06/25/2014

    The Permian bear discussion is cute, but there are quite a number of other taxa worthy of discussion for which vicariance-only hypotheses have been forwarded and found wanting. Panbiogeographers would have us believe that advanced cichlids date back to the Jurassic, that anthropoids date back to the Jurassic, and so on. These are all patently ridiculous hypotheses.

    It’s also worth pointing out that dispersal appears to be easier than previously thought. There’s substantial evidence of recent dispersal by boid snakes, by rhacophorid frogs, and a bunch of other animals. Transoceanic dispersal itself is no longer an ad hoc hypothesis that must be doubted with severity at all turns. It is well-chronicled and reasonably well-understood by mainstream biogeographers.

    Link to this
  137. 137. Grehan1896 3:16 pm 06/25/2014

    Two things.
    “1) You seem to be under the impression that molecular methods have no addressed these issues with new methods. The reality is, the current molecular models (e.g. BEAST) do not treat these as hard numbers.”

    The reality is that molecular divergence dates cannot falsify an earlier origin.

    “2) In addition, the order of divergences is equally an issue, and one which you have not addressed. All molecular analyses suggest an order of divergence that is not in line with the order of continental breakup. “

    The order of divergence does not have to match the order of continental drift.

    “So, you’re not only suggesting that the fossil record is anomalously gappy, but you’re also suggesting that the molecular phylogenies are wrong…..with no good reason for doing so.”

    Not I am not (at least not because may not match the sequence of drifting).

    “What correlation? The only correlation you have is that all modern ratites are Gondwandan. That’s not a correlation. That’s one data point. If you try to correlate actual ratite clades with tectonics, that correlation collapses. Which is why panbiogeographers then argue that the phylogeny itself is wrong.”

    I am curious. Where have panbiogeographer has argued that the (any any particular) ratite phylogeny is wrong?

    The correlation is spatial. In reference to molecular phylogenetic breaks proposed by Smith ete al (2013) there are three main allopatric clades – (1. Africa: Ostriches), (2. Pacific basins moas, tinamous, and rheas), and (Indian Ocean (elephant birds, emus, cassowaries and kiwis). This gives two biogeographic breaks. The first biogeographic break is between Africa and Madagascar (Middle Jurassic) and the Atlantic Ocean (Early Cretaceous). The second is somewhere in the new Zealand region where two clades overlap (the only region of overlap between any of the main clades). The moa-tinamou-rhea clade has a distribution similar to other organisms with different means of dispersal such as Coriaria (angiosmpersm) and the Penthesilenula incase group of ostrocods (see Heads, 2014).

    “It’s also substantially older than the first neoavian fossil, the oldest ornithuran, and most avian diversity in general.”

    True, but that is not a falsification.

    “When we talk about science and hypotheses, we talk about setting up direct tests of those hypotheses i.e. if we assume that a hypothesis is true, do additional forms of data or additional observations run counter to it or do they align with it? This is the fundamental test of any scientific hypothesis. If we assume panbiogeography to be true, we would expect various observations of molecular phylogeny and similarity that we do not observe. We would expect an antiquity of ornithurans and neoavians that we do not observe. And so on. Thus, we have to remain skeptical of the panbiogeography hypothesis, because so many lines of evidence run counter to it.”

    But there has not been any indication that the panbiogeographic hypotheses runs counter to any other lines of evidence since all those other lines of evidence seem (as far I as I can tell) one or other representation of fossil dates as the actual or maximal divergence dates for a taxon.

    John Grehan

    Link to this
  138. 138. Grehan1896 3:20 pm 06/25/2014

    “The Permian bear discussion is cute, but there are quite a number of other taxa worthy of discussion for which vicariance-only hypotheses have been forwarded and found wanting. Panbiogeographers would have us believe that advanced cichlids date back to the Jurassic, that anthropoids date back to the Jurassic, and so on. These are all patently ridiculous hypotheses.”

    You may think they are ridiculous, but that does not mean that they are scientifically so.

    “It’s also worth pointing out that dispersal appears to be easier than previously thought. There’s substantial evidence of recent dispersal by boid snakes, by rhacophorid frogs, and a bunch of other animals. Transoceanic dispersal itself is no longer an ad hoc hypothesis that must be doubted with severity at all turns. It is well-chronicled and reasonably well-understood by mainstream biogeographers.”

    Sure, things are seen to move about. But interestingly when one looks at global biogeography one does not find everything everywhere, but more often differentiated taxa are allopatric and their phylogenetic breaks match various tectonic features suggesting that they there ancestor was widely distributed before those tectonic events. Maybe it is biogeography that helps us understand the evolutionary significance of how and when organisms move about rather than the other way around.

    Biogeographic patterns have even been used to successfully predict the existence of Mesozoic tectonic features that were then unknown to geologists.

    John Grehan

    Link to this
  139. 139. Yodelling Cyclist 3:30 pm 06/25/2014

    Damn, another good pint.

    Link to this
  140. 140. Grehan1896 3:50 pm 06/25/2014

    A’m afraid it will take more than a few pints.

    Link to this
  141. 141. Grehan1896 4:03 pm 06/25/2014

    Darren Naish,

    Exactly how much older than the oldest known fossil are ratites allowed to be? And why.

    John Grehan

    Link to this
  142. 142. naishd 4:25 pm 06/25/2014

    I think it’s clear by now why the majority of interested parties are not satisfied with panbiogeographical hypotheses; thank you to everyone who has contributed here. As regards my own dissatisfaction with the idea (a follow-on from comment # 115), I would strongly agree with ectodysplasin’s point in comment # 136: we have excellent reasons from modern animals for thinking that dispersal is, often, easy. Giant tortoises have literally been discovered in the act of swimming between islands, birds like cattle egrets have colonised new continents (including the fringes of Antarctica), iguanas have been observed rafting across a seaway, and so on.

    If we stick with palaeognaths as a case-study, a model of palaeognath phylogeny is said (comment # 137) to be consistent with continental breakup. But — what about the new phylogenetic hypotheses? As discussed in the article way above, rheas don’t group with moa and tinamous: rather, the tinamou + moa clade is closer to the Australasian ratite + Madagascan elephant bird clade that it is to rheas. I mean – this topology demands a totally different pattern of continental breakup. What then? Do you just pick and choose the phylogenies you favour, or do you devise a new model of continental fragmentation? I would say this is a case where the panbiogeographic hypothesis is, again, discordant with other data.

    As for “how much older than the oldest known fossil are ratites allowed to be?” (comment # 141), they should be younger than the numerous early neornithine, ornithurine and ornithuromorph fossils we have. That’s a lot of data-points (there are numerous taxa involved). Even with a generous bit of slop, this puts palaeognath origins no later than the Early Cretaceous – something like 120 Ma ago at most, and this generously takes the total lineage into account, not the crown-group only. This is addressed in the paper I did with Mike Lee et al., cited above…

    We’re going round in circles here.

    Link to this
  143. 143. irenedelse 4:40 pm 06/25/2014

    Wait, I have the solution! Let’s go back to pre-Wegener biogeography! Temporary land bridges between continents are a logical hypothesis too, and consistent with vicariance as the cause to the observed Paleognaths distribution! Right?

    Link to this
  144. 144. Grehan1896 5:47 pm 06/25/2014

    “I think it’s clear by now why the majority of interested parties are not satisfied with panbiogeographical hypotheses; thank you to everyone who has contributed here.”

    It is certainly clear that the majority of those who have commented do not like the panbiogeographic hypothesis (or evidence which is more to the point).

    “As regards my own dissatisfaction with the idea (a follow-on from comment # 115), I would strongly agree with ectodysplasin’s point in comment # 136: we have excellent reasons from modern animals for thinking that dispersal is, often, easy.”

    We do not. We only have some observations of animals and plants moving about. That does not mean that the origin of vicariant groups was not geologically mediated.

    “Giant tortoises have literally been discovered in the act of swimming between islands, birds like cattle egrets have colonised new continents (including the fringes of Antarctica), iguanas have been observed rafting across a seaway, and so on.”

    True, but again, that does not prove anything about the origin of vicariant groups.

    “If we stick with palaeognaths as a case-study, a model of palaeognath phylogeny is said (comment # 137) to be consistent with continental breakup. But — what about the new phylogenetic hypotheses? As discussed in the article way above, rheas don’t group with moa and tinamous: rather, the tinamou + moa clade is closer to the Australasian ratite + Madagascan elephant bird clade that it is to rheas. I mean – this topology demands a totally different pattern of continental breakup.”
    No it does not. I do not understand this insistence that a phylogenetic pattern of relationship must match a particular geological breakup sequence. There is no necessary relationship unless the geological sequence is correct and is responsible for the isolation and differentiation sequence.

    “What then? Do you just pick and choose the phylogenies you favour, or do you devise a new model of continental fragmentation? I would say this is a case where the panbiogeographic hypothesis is, again, discordant with other data.”

    Again no. It seems to me that you have very little knowledge of panbiogeographic methodology or of examples of analysis (it might be a good thing to read the two recent books by Heads).

    So with respect to the latest phylogeny (and who knows how long that will stay together given the various changes over the years in both molecular and morphological reconstruction), one has a very similar overall pattern. There is still the primary phylogenetic break with the Mozambique Channel and the Atlantic Ocean. The rhea is basal with respect to a trans-Pacific group (moa and tinamous) that overlaps in New Zealand with an Indian Ocean group (kiwi, elephant bird, emu/cassowary). The main difference here is biogeographic overlap between the rhea and the tinamous member of its sister group. The overlap between the Indian Ocean and Pacific Ocean groups remains (and this overlap is itself evidence of dispersal subsequent to their original differentiation.

    “As for ‘how much older than the oldest known fossil are ratites allowed to be?’ (comment # 141), they should be younger than the numerous early neornithine, ornithurine and ornithuromorphs fossils we have.”

    If the formation of fossils is contingent upon conditions allowing for fossils, why should the order of appearance of fossils necessarily follow phylogenetic sequence?

    “That’s a lot of data-points (there are numerous taxa involved). Even with a generous bit of slop, this puts palaeognath origins no later than the Early Cretaceous – something like 120 Ma ago at most,”

    But other molecular theorists have come up with older predictions. Of course you may disagree with them, but the point is that there is more than one predicted oldest divergence date.

    “and this generously takes the total lineage into account, not the crown-group only. This is addressed in the paper I did with Mike Leeet al., cited above…”

    I have the paper and I will take a look in due course and see what might be added.

    “We’re going round in circles here.”

    Circles or not, I think the nature of the exchange reflects our different views. We are not going to change our respective views, but it is interesting to work out the precise basis for our different perspectives.

    John Grehan

    Link to this
  145. 145. naishd 6:05 pm 06/25/2014

    It’s true that I haven’t read the books by Heads, but I have read several of his papers* and was under the impression that I knew what the general methodology/philosophy was. Anyway…

    * see previous comments: I was really fair on him in what I said about birds-of-paradise (Naish 2010).. too fair, I think, since I was not aware at the time of his broader hypotheses.

    John: I may not agree with you at all, but I want to thank you for leaving these interesting comments here, and – above all – for being extremely cordial and reasonable in your comments.

    Link to this
  146. 146. Grehan1896 6:39 pm 06/25/2014

    “John: I may not agree with you at all, but I want to thank you for leaving these interesting comments here, and – above all – for being extremely cordial and reasonable in your comments.”

    Darren, very kind of you to say so. I have had many, many years of being in the minority so it does not worry me that most do not agree with me and I am comfortable with that. I do find quite a few people who are in a majority position are not comfortable with minority views and who will make disparaging comments. But you have been very respectful throughout. This is very different to some who have published in Systematic Biology who decry panbiogeography as akin to creationism and use words to the effect that panbiogeography should be banned from publication.

    John Grehan

    Link to this
  147. 147. ectodysplasin 6:40 pm 06/25/2014

    @John,

    The reality is that molecular divergence dates cannot falsify an earlier origin.

    This is essentially special pleading. Selective nonpreservation of some lineages in some intervals and not in others is an ad hoc hypothesis that can be modeled and tested, and we can put numbers on how credible such a hypothesis is. You’re not doing that. You’re just saying “well, it’s a non-zero possibility, therefore my idea is right” and that’s bad science.

    The order of divergence does not have to match the order of continental drift.

    Obviously not, but you need to be able to explain situations where biogeographical outliers are deeply nested and the timing does not line up.

    The correlation is spatial. In reference to molecular phylogenetic breaks proposed by Smith ete al (2013) there are three main allopatric clades – (1. Africa: Ostriches), (2. Pacific basins moas, tinamous, and rheas), and (Indian Ocean (elephant birds, emus, cassowaries and kiwis). This gives two biogeographic breaks. The first biogeographic break is between Africa and Madagascar (Middle Jurassic) and the Atlantic Ocean (Early Cretaceous). The second is somewhere in the new Zealand region where two clades overlap (the only region of overlap between any of the main clades). The moa-tinamou-rhea clade has a distribution similar to other organisms with different means of dispersal such as Coriaria (angiosmpersm) and the Penthesilenula incase group of ostrocods (see Heads, 2014).

    1. What are you correlating? You’re claiming there are strong correlations, but you’re not actually presenting two variables that you’re testing the correlation of. So you’re really just saying “these things are correlated” while only investigating a single set of data (occurrence).

    2. Flightless birds do not live in water. They do not “occur” in basins. They occur within ranges on land. The fact that you are placing animals that have overlapping occurrences within different theoretical “basins” is intellectually dishonest, because you have no reason to place moas in the pacific basin and kiwis in the Indian basin besides your own personal biases.

    True, but that is not a falsification.

    I could draft up a rough preservational model that could estimate the credibility of your hypothesis, though. So, maybe not strictly Popperian, but definitely Bayesian.

    You seem to think you can propose rampant nonpreservation and not introduce additional parameters into your inferences, but that is not the case.

    But there has not been any indication that the panbiogeographic hypotheses runs counter to any other lines of evidence since all those other lines of evidence seem (as far I as I can tell) one or other representation of fossil dates as the actual or maximal divergence dates for a taxon.

    I think you critically misunderstand how inferences regarding preservation in the fossil record actually work.

    Link to this
  148. 148. Stripeycat 6:53 pm 06/25/2014

    Comment 144: “If the formation of fossils is contingent upon conditions allowing for fossils, why should the order of appearance of fossils necessarily follow phylogenetic sequence?”

    It needn’t. However, when we have reasonably good fossil assemblages from many times and places (including some amazing Lagerstaetten) and they only show the hypothesised ancenstral/basal forms and none of the derived lineages your hypothesis requires, you need to explain why. Why are none of the Chinese sites turning up any neoavians, even though they are preserving similarly-sized species like /Confusciornis/ and /Protopteryx/? Why does Solnhoffen have /Archaeopteryx/ and small pterosaurs, but no neoavians, or even enantiornithines? There may be possible explanations for why some sites and lineages aren’t preserved (like extreme rarity, or unusual behaviour), but if you are proposing something of the order of 70 million years of ghost lineage, the onus is on you to explain why; and happenstance becomes increasingly unlikely as the time extends, and the known assemblages become larger and probably more representative.

    Also, right back in comment 123, you mentioned the need for an angiosperm ghost lineage back to the mid-Triassic, to the first appearance of gymnosperms. Why?

    Link to this
  149. 149. ectodysplasin 7:13 pm 06/25/2014

    and more…

    You may think they are ridiculous, but that does not mean that they are scientifically so.

    In some cases, yes, it does. Once again, hypotheses of non-preservation are ad hoc hypotheses that can and should be modeled and tested.

    Sure, things are seen to move about. But interestingly when one looks at global biogeography one does not find everything everywhere, but more often differentiated taxa are allopatric and their phylogenetic breaks match various tectonic features suggesting that they there ancestor was widely distributed before those tectonic events. Maybe it is biogeography that helps us understand the evolutionary significance of how and when organisms move about rather than the other way around.

    Biogeographic patterns have even been used to successfully predict the existence of Mesozoic tectonic features that were then unknown to geologists.

    Biogeography certainly has its place in understanding the history of life, but it should not be given primacy a priori.

    Secondly, statements such as:

    differentiated taxa are allopatric and their phylogenetic breaks match various tectonic features suggesting that they there ancestor was widely distributed before those tectonic events

    must be given proper scrutiny and their underlying assumptions must first pass the sniff test, and one must be careful to distinguish between a priori and a posteriori statements. Assuming a priori that a vicariance-only model is appropriate is pretty dangerous and doesn’t actually permit you to test the questions you claim to want to test.

    Link to this
  150. 150. Grehan1896 7:20 pm 06/25/2014

    Last two postings cover quite a lot. I will comment soon.

    John Grehan

    Link to this
  151. 151. Christopher Taylor 9:38 pm 06/25/2014

    If “all other data points” (presumably you mean other than biogeographic data points) indicate a mid-Jurassic origin for birds as a while then there would appear to be nothing extraordinary about the basal lineages of extant birds also diverging about that time.

    What Darren was referring to was the basalmost divergence of birds sensu lato, i.e. between Archaeopteryx and other members of the bird lineage. From that point on the tree, the divergence of extent lineages is not very basal at all.

    For example, recognized angiosperm fossils only go back into the early Cretaceous (at least the last time I looked, but I have not kept up with this question), but their gymnosperm relatives go back to the Triassic, suggesting that angiosperms diverged in the Triassic also and were present at that time even though they are not recorded or recognized in the fossil record.

    Not quite: it means that the angiosperm stem-lineage must have been present in that time. And while their position as such has not been entirely settled, there are indeed candidate stem-angiosperms going back far enough: Caytonia, Glossopteridales, etc.

    If the formation of fossils is contingent upon conditions allowing for fossils, why should the order of appearance of fossils necessarily follow phylogenetic sequence?

    It may not follow religiously, but in this case it may be expected to at least approximate. Because there is such a thing as probability.

    Let me offer a counter-example. The fossil record of nematodes is hopelessly fragmentary: I am aware of only one well-characterised fossil nematode (the Devonian Palaeonema, which has been assigned to the Enoplia) from earlier than the Cretaceous amber. Yet despite this paucity, we would be surprised if many of the higher divergences in nematode phylogeny hadn’t happened a long time ago. Molecular dating has suggested that members of the nematode crown group diverged in the Cambrian.

    So in the case of nematodes, the fossil record is not regarded as reliable evidence for divergence dates. But there are reasons for this: the potential for good preservation of nematodes is hopelessly low, as evidenced not only by their unexpectedly late appearance in the fossil record but also by their extremely sporadic appearance afterwards.

    As both I and Ectodysplasin have pointed out above, this argument doesn’t work out for birds. The preservation potential of birds is lower than that of some taxa—birds lose out to bivalves—but it is far higher than that of nematodes. Once birds do appear in the fossil record, they maintain a fairly consistent appearance in it. And for the most part, the order of appearance of bird lineages in the fossil record does approximate their order of phylogenetic divergence.

    Link to this
  152. 152. Grehan1896 11:29 pm 06/25/2014

    “What Darren was referring to was the basalmost divergence of birds sensu lato, i.e. between Archaeopteryx and other members of the bird lineage. From that point on the tree, the divergence of extent lineages is not very basal at all.”

    What I think you mean is that there is no fossil record of divergence in extant lineages at that time.

    “Not quite: it means that the angiosperm stem-lineage must have been present in that time.”

    By stem lineage I presume you refer to angiosperms that did not diversify into their modern groups until they appeared in the Cretaceous(?) Either way my point was that a sister group fossil record is sometimes used to argue for the presence of the other sister group for extended periods of time where there is no fossil record.

    “And while their position as such has not been entirely settled, there are indeed candidate stem-angiosperms going back far enough: Caytonia, Glossopteridales, etc.”

    Agreed.

    “It may not follow religiously, but in this case it may be expected to at least approximate. Because there is such a thing as probability.”

    Fair enough as an opinion, but it does not mean that it is necessarily so.

    “So in the case of nematodes, the fossil record is not regarded as reliable evidence for divergence dates”.
    So fossil calibrated molecular divergence estimates are not used for nematodes?

    “Once birds do appear in the fossil record, they maintain a fairly consistent appearance in it. And for the most part, the order of appearance of bird lineages in the fossil record does approximate their order of phylogenetic divergence.”

    Maybe so, but that does not falsify the possibility of such lineages having an origin much earlier than their oldest fossil.

    John Grehan

    Link to this
  153. 153. Grehan1896 11:37 pm 06/25/2014

    I guess the question is how such hypotheses are “tested”. Certainly not my molecular divergence estimates that give minimal ages only.

    Biogeography certainly has its place in understanding the history of life, but it should not be given primacy a priori.

    Yes, its place is in providing an independent source of evidence to methods such as molecular divergence estimation. My argument is, as I have said, only that biogeographic correlation with tectonics offers an alternative divergence estimate to that of fossil (or stratigraphcially) calibrated molecular minimal divergence estimates.

    It is not given primacy. In the case of ratites it gives an alternative model for the origin of ratite distributions to that of molecular divergence which can only give minimal estimates.

    “must be given proper scrutiny and their underlying assumptions must first pass the sniff test, and one must be careful to distinguish between a priori and a posteriori statements. Assuming a priori that a vicariance-only model is appropriate is pretty dangerous and doesn’t actually permit you to test the questions you claim to want to test.”

    In actuality, a vicariance-only model (for those cases analyzed by that method) is not assumed a priori. The model is a predication of the historical process based on how vicariant distributions are spatially related to each other. Heads’ books give numerous examples of the approach.

    John Grehan

    Link to this
  154. 154. Grehan1896 11:44 pm 06/25/2014

    “There may be possible explanations for why some sites and lineages aren’t preserved (like extreme rarity, or unusual behaviour), but if you are proposing something of the order of 70 million years of ghost lineage, the onus is on you to explain why; and happenstance becomes increasingly unlikely as the time extends, and the known assemblages become larger and probably more representative.”

    I know it seem daunting that such a possibility exists. But there is still no scientific imperative as to what should or should not happen in the fossil record just because it does not meet personal expectations. If it were not for the biogeographic evidence I would certainly consider the absence to be real, but for now I have to consider that there are modern birds with Mesozoic correlations and yet no fossils. Either the correlation is an artifact or the fossil record is. My personal inclination is in favor of the reality of the biogeography and I also accept that others such as yourself are inclined in favor of the fossil record as it stands.

    “Also, right back in comment 123, you mentioned the need for an angiosperm ghost lineage back to the mid-Triassic, to the first appearance of gymnosperms. Why?”

    Sister taxa diverge at the same time.

    Link to this
  155. 155. Christopher Taylor 12:21 am 06/26/2014

    Either way my point was that a sister group fossil record is sometimes used to argue for the presence of the other sister group for extended periods of time where there is no fossil record.

    But in the case of palaeognaths, there is no sister group fossil record for the time you claim either.

    I’m a sloth. Your argument is invalid.

    Link to this
  156. 156. naishd 5:32 am 06/26/2014

    There’s another point I want to make here: it’s been alluded to obliquely, but needs to be stated more clearly, I think.

    Those of us interested in animals and other organisms are never experts on everything – we tend to specialise on one, two, or a few groups. Ergo, for specialist knowledge it is advisable to talk to people who know well the specific group of interest. When thinking about birds, say, from the point of view of biogeography, it may be easy to gain a broadbrush view, but not so easy to appreciate the many details and obscure data points that need to be considered.

    What I’m getting at is that it’s knowledge of the finer details of deep-time history that has allowed palaeornithologists (and others) to construct the detailed hypotheses they have. As discussed in comments above (comments 108, 110, 135), the pattern of the bird fossil record is regarded by most of us as being discordant with the idea that palaeognaths, even crown-birds in general, could have been present in the Middle Jurassic (as they’d need to be for palaeognath vicariance to hold up). So, it isn’t just that we have a sensible and logical phylogenetic backbone on which to hang our hypotheses (example: in the bird branch of Maniraptora, we see Archaeopteryx-like animals branching first, then jeholornithids, confuciusornithids, enantiornithines, then numerous early ornithuromorphs, then early ornithurines like hesperornithines – all of this happening before we get to crown-birds), it’s also that the pattern of anatomical, behavioural, ecological and biogeographical details we see is consistent with this ‘backbone’. In the early ornithuromorphs of the Cretaceous, for example, we see increasingly more crown-bird-like characters being assembled in step-wise fashion, and we also see the ecology and biogeography of these birds changing as a logical precedent to the settings in which ornithurines appeared (Ornithurae being the clade that includes crown-birds).

    In other words, the fine details of the tree – the huge mass of data we have, it representing a body of evidence much greater than mere occurrences in geological time – paint a consistent and robust model of the timing of bird evolution. The panbiogeographic model simply ignores all that, and pulls all lineages way, way back into the Jurassic, requires extensive ghost lineages right across the tree, and would mean that groups are not associated with the ecological, biogeographical, climatic or environmental settings or events we currently think they are, quite simply because those settings or events were very different, or did not exist, so further back in time. It proposes that we have EVERYTHING wrong, not just that we’re missing a few early occurrences of certain groups.

    In short, the model is not only grossly unparsimonious but also looks clumsy and superficial, basically requiring that we throw out all the details we see across the tree – not just the datings of fossils. I’ve used birds as the example here, but the same goes for many other groups that might be mentioned (primates, crocodiles and other terrestrial groups present on far-flung continents). So, to favour panbiogeography, you seemingly have to have the broadest, most superficial understanding of the groups you’re interested in. Palaeognaths are birds and they evolved in the Jurassic. Err, no; it’s a bit more complicated than that.

    Link to this
  157. 157. irenedelse 7:04 am 06/26/2014

    “In other words, the fine details of the tree – the huge mass of data we have, it representing a body of evidence much greater than mere occurrences in geological time – paint a consistent and robust model of the timing of bird evolution. The panbiogeographic model simply ignores all that, and pulls all lineages way, way back into the Jurassic, requires extensive ghost lineages right across the tree, and would mean that groups are not associated with the ecological, biogeographical, climatic or environmental settings or events we currently think they are, quite simply because those settings or events were very different, or did not exist, so further back in time.”

    Very good point. And that’s where the proverbial devil is in the details! When advocating an alternative hypothesis, one needs to show that it better explains the data than the status quo. For paleognath origins, once you look at it closely, panbiogeography introduces more problems, not fewer.

    Link to this
  158. 158. Grehan1896 7:39 am 06/26/2014

    “But in the case of palaeognaths, there is no sister group fossil record for the time you claim either.

    I’m a sloth. Your argument is invalid.”

    Sorry, I did not mean to imply that there was. I was only pointing out that this was a method sometimes employed to infer a long history for a group when there was no fossil record. I was not saying that this was the case for palaeognaths.

    John Grehan

    Link to this
  159. 159. Grehan1896 7:53 am 06/26/2014

    “), the pattern of the bird fossil record is regarded by most of us as being discordant with the idea that palaeognaths, even crown-birds in general, could have been present in the Middle Jurassic (as they’d need to be for palaeognath vicariance to hold up).”

    Only because there is an expectation that there should be fossils to support that model.

    “paint a consistent and robust model of the timing of bird evolution.”

    Its consistent with a particular view and it is robust in that it appears not to be challenged in any way.

    “The panbiogeographic model simply ignores all that, and pulls all lineages way, way back into the Jurassic”

    Not all lineages.

    “requires extensive ghost lineages right across the tree”

    Most living species are ghost lineages (never liked that term myself) since they lack a fossil record.

    “and would mean that groups are not associated with the ecological, biogeographical, climatic or environmental settings or events we currently think they are, quite simply because those settings or events were very different, or did not exist, so further back in time.”

    That’s an interesting point. And yes, the original ecology, biogeographical, climatic or environmental settings may have been very different to the present, but this does not mean that they are evolutionarily disconnected. For example, one may find organisms that are currently subalpine, but were originally coastal or mangrove, but were transported into their present condition by tectonic uplift.

    “It proposes that we have EVERYTHING wrong, not just that we’re missing a few early occurrences of certain groups.”

    I don’t know that EVERYTHING is wrong.

    “In short, the model is not only grossly unparsimonious but also looks clumsy and superficial, basically requiring that we throw out all the details we see across the tree – not just the datings of fossils. I’ve used birds as the example here, but the same goes for many other groups that might be mentioned (primates, crocodiles and other terrestrial groups present on far-flung continents).”

    The converse of this argument is that the extensive details for so many taxa that generate a coherent tectonic pattern have to be thrown out.

    “So, to favour panbiogeography, you seemingly have to have the broadest, most superficial understanding of the groups you’re interested in. Palaeognaths are birds and they evolved in the Jurassic. Err,no; it’s a bit more complicated than that.”

    So is the biogeography.

    Link to this
  160. 160. Grehan1896 7:59 am 06/26/2014

    “Very good point. And that’s where the proverbial devil is in the details! When advocating an alternative hypothesis, one needs to show that it better explains the data than the status quo.”

    It all depends on what one means by “better explains”. Unfortunately there is no universal criterion. How does one know that an explanation is better? Often the argument is that if an explanation is better then it will be accepted by the majority. The history of science shows that this is not always the case. Barbara McClintok had a better genetic explanation with her jumping genes theory only to be ridiculed for years by her ‘better informed’ peers.

    “For paleognath origins, once you look at it closely, panbiogeography introduces more problems, not fewer.”
    Only more problems for the conventional model.

    Link to this
  161. 161. irenedelse 8:37 am 06/26/2014

    “It all depends on what one means by “better explains”. Unfortunately there is no universal criterion. How does one know that an explanation is better.”

    Now, that is just playing with words.

    Link to this
  162. 162. irenedelse 8:45 am 06/26/2014

    “Only because there is an expectation that there should be fossils to support that model.”

    The trouble is when there *is* a fossil record (as detailed by naishd in comment #156 for the evolution of birds) and someone insist on building a theory that doesn’t take it into account…

    Link to this
  163. 163. naishd 8:47 am 06/26/2014

    John says (comment # 160)…

    “It all depends on what one means by “better explains”. Unfortunately there is no universal criterion. How does one know that an explanation is better?”

    Definition of one explanation being better than another = one explanation has better explanatory power, better matches data points (see comment # 156 and others), makes predictions that are supported by further discoveries (see comment # 135 and others). I’m still not sure whether science is a democracy or not (a complex meta-issue), but the number of qualified people you can convince also indicates which hypothesis is ‘better’. Hence the poor standing of vicariance as having explanatory power behind palaeognath distribution.

    Link to this
  164. 164. Grehan1896 9:27 am 06/26/2014

    “Definition of one explanation being better than another = one explanation has better explanatory power, better matches data points (see comment # 156 and others)”

    The definition is fine, but it’s the application to particulars that can be problematic. As in the McClintock example, she had an explanation with better explanatory power and better matched data points. That was not enough.

    “makes predictions that are supported by further discoveries (see comment # 135 and others).”
    Even this criterion is tricky. Panbiogeography has, for example, made predications about geology that have been supported by further discoveries. That did not persuade many that the panbiogeographic explanation was better. Most seem to just ignore these achievements.
    The panbiogeographic model predicts ratites in the Jurassic. In this case, the future discovery of a fossil would provide support.

    “I’m still not sure whether science is a democracy or not (a complex meta-issue), but the number of qualified people you can convince also indicates which hypothesis is ‘better’. Hence the poor standing of vicariance as having explanatory power behind palaeognath distribution.”

    This is an interesting perspective on the philosophy of science. Choices as to what is better often depend on one’s perspective (and this is not just playing with words). For example, the morphogenetic evidence for a sister group relationship between humans/fossil hominids and orangutans is not accepted as the better explanation by the majority of ‘qualified people’ (not sure who is qualified or not) because most in the field believe that sequence similarity is the absolute proof of a closer relationship between humans and chimpanzees.

    John Grehan

    Link to this
  165. 165. Grehan1896 9:29 am 06/26/2014

    “The trouble is when there *is* a fossil record (as detailed by naishd in comment #156 for the evolution of birds) and someone insist on building a theory that doesn’t take it into account…”

    I do take the fossil record into account. But given the biogeographic evidence I am inclined to interpret the fossil record differently to most.

    John Grehan

    Link to this
  166. 166. Grehan1896 10:04 am 06/26/2014

    Darren,

    I do not have ready access to your 2010 book. Is there any way you can send me a scan of the section dealing with Michael Heads’s work on birds of paradise and New Guinea?

    John Grehan

    Link to this
  167. 167. Halbred 12:31 pm 06/26/2014

    But this Jurassic palaeognath thing can never be disproven, so it’s not really scientific, is it? If we continue to NOT find palaeognaths in Jurassic rocks, you can continue to say “you just haven’t found them yet,” or “they must have lived in environments with low preservation potential.”

    Do you see how this might come across as intellectually dishonest? That’s how it’s sounding to me.

    Link to this
  168. 168. naishd 12:45 pm 06/26/2014

    Re: comment # 166, I don’t have a pdf of the book chapter to hand, but you can read a prototype of this now-very-dated take on the subject here at Tet Zoo ver 1. It dates to March 2006.

    Link to this
  169. 169. ectodysplasin 2:07 pm 06/26/2014

    John,

    I think you fundamentally misunderstand how absence data in the fossil record are and can be handled in formal research. The absence of fossils in an interval does not give researchers license to freely speculate on which organisms were around at the time but have not been preserved. This especially applies to very large gaps where fossils are to be expected. You have to demonstrate using multiple lines of evidence that the organisms had to be there.

    All you’re doing is saying that your theoretical framework implies that these animals must be there. When the gap is as long as you’re claiming, then this suggests you are wrong to apply your model to this specific problem.

    Link to this
  170. 170. Grehan1896 4:22 pm 06/26/2014

    “I think you fundamentally misunderstand how absence data in the fossil record are and can be handled in formal research. The absence of fossils in an interval does not give researchers license to freely speculate on which organisms were around at the time but have not been preserved.”

    Agreed. But I am not freely speculating. I have evidence in the form of a tectonic correlation. Now I would also agree that others such as you may not consider that evidence sufficiently compelling to accept a much earlier date for the origin of the group in question. I also accept that the geological correlation could be an artifact, but then so too might the ‘gap’.

    “This especially applies to very large gaps where fossils are to be expected. You have to demonstrate using multiple lines of evidence that the organisms had to be there.”

    Well certainly each of us has our own particular criteria as to what one ‘must’ do. I guess it depends on what one requires in the way of multiple lines of evidence.

    “All you’re doing is saying that your theoretical framework implies that these animals must be there.”
    Well all approaches involve a theoretical framework and pretty much say the same thing – that whatever theoretical framework is applied leads to whatever conclusion that is made.

    “When the gap is as long as you’re claiming, then this suggests you are wrong to apply your model to this specific problem.”

    Maybe, maybe not.

    John Grehan

    Link to this
  171. 171. Grehan1896 4:29 pm 06/26/2014

    “But this Jurassic palaeognath thing can never be disproven, so it’s not really scientific, is it? If we continue to NOT find palaeognaths in Jurassic rocks, you can continue to say “you just haven’t found them yet,” or “they must have lived in environments with low preservation potential.”
    Do you see how this might come across as intellectually dishonest? That’s how it’s sounding to me.”

    Yes I can see how. This starts getting into broader questions of philosophy about scientific method. Lakatos, for example, argued that research programs all have a hard core that is protected from falsification, and yet such programs still work when they generate new knowledge. He identified theoretically progressive research programs that generated novel predications that were not yet corroborated (but they may be already ‘falsified’). He also identified empirically progressive research programs where corroboration was produced by an independent research program. I guess a case in point would be the predication made in the 1950’s that Galapagos was associated with a major tectonic feature that was then unknown to science because seafloor mapping had not yet generated such results. In the interim I suppose the predication could not be disproven in the absence of finding the tectonic feature. Maybe that’s not a perfect comparison. Perhaps I should not even try as I am not a philosopher of science. And I used Lakatos only as an illustration.

    John Grehan

    Link to this
  172. 172. irenedelse 5:07 pm 06/26/2014

    “But I am not freely speculating. I have evidence in the form of a tectonic correlation.”

    Others already asked but again: correlation between what data? New-Zealand kiwis are genetically closer to Madagascar elephant birds than to New-zealand moa. That’s more of a tectonic paradox than a correlation.

    Link to this
  173. 173. ectodysplasin 6:19 pm 06/26/2014

    No, no, he already explained that. Kiwis aren’t from New Zealand, they’re from the Indian Ocean basin. Moas aren’t from New Zealand, they’re from the southern Pacific Ocean basin. The criteria for making this distinction aren’t arbitrary, they’re based on tectonic correlations.

    Link to this
  174. 174. Grehan1896 6:49 pm 06/26/2014

    “No, no, he already explained that. Kiwis aren’t from New Zealand, they’re from the Indian Ocean basin. Moas aren’t from New Zealand, they’re from the southern Pacific Ocean basin. The criteria for making this distinction aren’t arbitrary, they’re based on tectonic correlations.”

    That’s quite an accurate understanding. Yes, each of the NZ genera belong (according to the ‘current’ molecular reconstruction)to different groups with their respective nearest relatives occuring on the other side of different ocean basins. That suggests that age of each group is correlated with the age of formation of those basins. But there is also the phylogenetic break between groups such as between Ostriches and the rest with the Atlantic basin and Mozambique basin.

    John Grehan

    Link to this
  175. 175. irenedelse 7:14 pm 06/26/2014

    @grehan1896

    “That suggests that age of each group is correlated with the age of formation of those basins.”

    And here we are! Your hypothesis is that there is a correlation between age of those groups with tectonic events. However, you have earlier said (in your comment #170): “I have evidence in the form of a tectonic correlation.” That’s what I was commenting on at #172.

    So, which is it, please?

    Oh, and, talking of basins, I’d be curious to see your reply to ectodysplasin’s comment #147.

    Link to this
  176. 176. ectodysplasin 7:22 pm 06/26/2014

    Note to self: when calling out circularity, be explicit. People miss sarcasm, apparently.

    Link to this
  177. 177. Grehan1896 7:51 pm 06/26/2014

    “That suggests that age of each group is correlated with the age of formation of those basins.”

    And here we are! Your hypothesis is that there is a correlation between age of those groups with tectonic events.”

    Yes

    However, you have earlier said (in your comment #170): “I have evidence in the form of a tectonic correlation.”

    Yes.

    “So, which is it, please?”

    I think I see what you find problematic, but I am guessing here. There is a spatial correlation that is identified in terms of phylogenetic breaks and particular tectonic features, and there are clades that span particular tectonic features. This is a biogeographic correlation. It leads to the prediction that the origin of the tectonic feature and the origin of the distribution range or phylogenetic breaks share a common history (that is what I meant by correlation here).

    “Oh, and, talking of basins, I’d be curious to see your reply to ectodysplasin’s comment #147.”

    This comment?

    “Others already asked but again: correlation between what data? New-Zealand kiwis are genetically closer to Madagascar elephant birds than to New-zealand moa. That’s more of a tectonic paradox than a correlation.”
    It’s not tectonic in the sense of a tectonic feature that marks a zone of tectonic activity responsible for a particular geomorphology. Madagascar and New Zealand just refer to two geographic (really geopolitical) areas. There is nothing inherently paradoxical about the distribution that I see.

    John Grehan

    Link to this
  178. 178. Heteromeles 10:15 pm 06/26/2014

    Um, how have we managed to have a huge discussion about mesozoic tectonic basins without mentioning the Tethys? That was the important one for a lot of tropical dispersal. If you’re going to start spinning stories in different eras, it’s kind of reasonable to talk about the tectonic basins that actually existed during the eras you’re trying to frame.

    Link to this
  179. 179. irenedelse 2:50 am 06/27/2014

    @John Grehan:

    “I think I see what you find problematic, but I am guessing here. There is a spatial correlation that is identified in terms of phylogenetic breaks and particular tectonic features, and there are clades that span particular tectonic features. This is a biogeographic correlation.”

    What I find problematic is that this use of the term “correlation” could cover any description of the range of any group of species. There is a phylogenetic break here and there are some tectonic features there. What correlation is there other than “at a certain point in the past, the two land masses where those species occur were part of the same continent”?

    ““Oh, and, talking of basins, I’d be curious to see your reply to ectodysplasin’s comment #147.”

    This comment?

    “Others already asked but again: correlation between what data? [...]“”

    Sorry, no, you are quoting me at comment #172. I meant what ectodysplasin wrote at #147:

    “1. What are you correlating? You’re claiming there are strong correlations, but you’re not actually presenting two variables that you’re testing the correlation of. So you’re really just saying “these things are correlated” while only investigating a single set of data (occurrence).

    “2. Flightless birds do not live in water. They do not “occur” in basins. They occur within ranges on land. [...]“

    Link to this
  180. 180. ectodysplasin 9:54 am 06/27/2014

    Honestly I’m more curious about when John thinks the Pacific Basin formed.

    Link to this
  181. 181. Grehan1896 9:56 am 06/27/2014

    Heteromeles said

    Um, how have we managed to have a huge discussion about Mesozoic tectonic basins without mentioning the Tethys? That was the important one for a lot of tropical dispersal. If you’re going to start spinning stories in different eras, it’s kind of reasonable to talk about the tectonic basins that actually existed during the eras you’re trying to frame.”

    I agree, it’s definitely reasonable to talk about the tectonic basins that actually existed during the eras I was referring to. However, there does not appear to be any evidence that ratites were tethyan – that the Tethys tectonics molded ratite distribution. The current evidence indicates that the Indian Ocean, Atlantic Ocean, and Pacific basins were involved for ratites rather than Tethys. However, along the Tethyan boundary one does get overlap between the Pacific and Indian Ocean ratites in New Zealand.

    A more northern tethyan extension is found in the plant Coriaria mostly around the Pacific margin and then embedded within the Himalaya and Europe. Another Tethyan example is the genus Austrobryonia with the tribe Bryonieae in Australia, central and southwestern Asia, the Mediterranean region and the Canary Islands (see Heads, 2014).

    So for ratites one need not involve Tethys any more than for the subterranean freshwater fish group Milyeringa and Tuphleotris in Australia and Madagascar respectively which is disjunct relative to an intervening landmass (India) as are elephant birds and kiwis (Australia).

    John Grehan

    Link to this
  182. 182. Grehan1896 10:38 am 06/27/2014

    irenedelse

    “What I find problematic is that this use of the term “correlation” could cover any description of the range of any group of species. “

    It could, I suppose, but in panbiogeography the correlation is either where a distribution spans or parallels a tectonic or geomorphological feature (ocean basins being the largest tectonic feature, the active tectonic center would be presented by the spreading ridges) or have a distributional or phylogenetic boundary or break at a tectonic/geomorphological feature.

    A couple of illustrative examples: In Australia there are distributions that show an interesting match with the former extent of land and seas in the Cretaceous. Birds of paradise in New Guinea where disjunctions match transform fault movement, freshwater fish disjunctions that match the tectonics of Baja California. The differentiation of anthropoids and prosimians at the Lebombo monocline.

    “There is a phylogenetic break here and there are some tectonic features there. What correlation is there other than “at a certain point in the past, the two land masses where those species occur were part of the same continent”?”

    This historical context or interpretation is not the correlation. The correlation is between patterns, patterns of distributions and patterns of tectonics/geomorphology (spreading ridges, synclines, anteclines, marine transgressions, transform faults, subduction/obdunction zones, basaltic plateaus, propagating fissures etc. It is the historical interpretation of these features by geologists that provides a historical insight into the evolution of organisms or groups with correlated distributions.

    “1. What are you correlating? You’re claiming there are strong correlations, but you’re not actually presenting two variables that you’re testing the correlation of. So you’re really just saying “these things are correlated” while only investigating a single set of data (occurrence).

    The correlation may be identified individually for a particular group. In that sense it is not a statistical correlation. There are correlations that may be identified for multiple taxa. Presumably that is amenable to statistical analysis. For example, there are over 100 taxa showing disjunctions either side of the Alpine fault in New Zealand. Some statistical work has been done in this field, but it has not yet been well explored. Monte Carlo methods have indicated that globally correlated distributional connections and boundaries lie well outside the 95% probability limits of random distribution.

    “2. Flightless birds do not live in water. They do not “occur” in basins. They occur within ranges on land. [...]“

    True, but the ocean basin correlation provides a biogeographic hypothesis of which geosector is most likely involved in the origin of a group. If a group has an Indian Ocean baseline then one my hypothesize the origin of the distribution is linked more to the history of that basin that events in the Pacific or Atlantic basins. This approach may be used even if one is positing overwater dispersal (e.g. for elephant birds and kiwis one might argue that the birds flew between across the Indian ocean rather than going the other way across the Pacific and then Atlantic).

    John Grehan

    Link to this
  183. 183. Grehan1896 10:49 am 06/27/2014

    Ectodysplasin: “Honestly I’m more curious about when John thinks the Pacific Basin formed.”

    If you mean the modern Pacific Basin then my thinking would be no different from what geologists think, and if I recall correctly, that is based on the oldest known seafloor sediments in the Jurassic. However, that is not to say that distributions correlated with the basin are immediately Jurassic. Pacific paleogeography involves the formation of large igneous plateaus and island arcs that by the early Cretaceous that have moved across the Pacific, some of which have become imbedded within the Americas and Asia.

    John Grehan

    Link to this
  184. 184. ectodysplasin 11:33 am 06/27/2014

    Thanks for establishing that you don’t know the first thing about tectonics.

    Link to this
  185. 185. Grehan1896 11:54 am 06/27/2014

    “Thanks for establishing that you don’t know the first thing about tectonics.”

    You might be so kind as to back up your assertion?

    John Grehan

    Link to this
  186. 186. ectodysplasin 12:09 pm 06/27/2014

    The Pacific Ocean wasn’t formed in the Jurassic. It is the remnant of Panthalassia, which goes back quite a bit further. The fact that the oldest Pacific crust is Jurassic is due to active subduction of oceanic crust along both margins of the Pacific Ocean, a process which, judging from the age of back-arc accretion along both East Asia and western North America, has been going on for the entire history of the Phanerozoic and then some.

    I’m also not sure where you got the idea that large igneous plateaus have wandered across the Pacific Ocean and accreted to the Americas. You seem to be conflating igneous provinces (such as Hawaii) with back-arc archipelagos (e.g. the San Juans). These two types of island chains originate from different processes and, to date our understanding of how these systems contribute to continent formation differs substantially.

    Link to this
  187. 187. Grehan1896 12:33 pm 06/27/2014

    “The Pacific Ocean wasn’t formed in the Jurassic. It is the remnant of Panthalassia, which goes back quite a bit further. The fact that the oldest Pacific crust is Jurassic is due to active subduction of oceanic crust along both margins of the Pacific Ocean, a process which, judging from the age of back-arc accretion along both East Asia and western North America, has been going on for the entire history of the Phanerozoic and then some.”

    Sure, but I was not focusing on ultimate origins of the Pacific. I have no problem with your description. What interests me is what Pacific paleogeographies contributed to the existence of trans-Pacific distribution.

    “I’m also not sure where you got the idea that large igneous plateaus have wandered across the Pacific Ocean and accreted to the Americas. You seem to be conflating igneous provinces (such as Hawaii) with back-arc archipelagos (e.g. the San Juans). These two types of island chains originate from different processes and, to date our understanding of how these systems contribute to continent formation differs substantially”

    Well I did not state where I got the “idea”. I was referring to formations such as the East Sulawesi ophiolite, the Shatsky Rise, Ontong Java plateau, Gorgona plateau, Caribbean plateau, Hikurangi plateau.

    John Grehan

    Link to this
  188. 188. ectodysplasin 1:39 pm 06/27/2014

    You said that the Pacific Ocean basin in Jurassic in age, and that this is the geological consensus. You also said that this is what necessitates the existence of Jurassic ratites, despite the lack of any fossil or molecular evidence for this. You have developed that idea in these comments as an explanation for phylogenetic relatedness of some South American paleognaths and some New Zealand paleognaths.

    Proposing Cretaceous dispersal along igneous provinces is also a smokescreen, because these igneous provinces don’t float across the ocean and become “imbedded” (sic) in other continents.

    In addition, you have cited the Alpine Fault in NZ as a biogeographic barrier, but have not addressed the substantial rain shadow effect that the southern alps produce. I could easily show disjunction on either side of any mountain range, due to the same effect. You’re asserting the primacy of tectonism as a cause without any acknowledgement or treatment of the more obvious biotic factors (e.g. water availability, temperature, etc) at play. Asserting that a bird adapted to rainforest underbrush has an “Indian Ocean Basin” distribution, whereas a dryland/upland bird has a “Pacific Ocean Basin” distribution is circular at best and ignores key biotic factors (distribution of habitat) which should be considered first.

    I am generally amenable to reasonable discussion about tectonism and how it contributes to dispersal or endemism of taxonomic groups. However, such reasonable discussion requires that all parties involved have a reasonable understanding and appreciation of the biotic factors that govern distribution of the taxa under discussion as well as the actual tectonic history of a region. It also requires that parties be willing to state and adhere to conditions which, if met, limit the applicability of their model. I’m not seeing that in this conversation.

    Link to this
  189. 189. Heteromeles 1:53 pm 06/27/2014

    Should we be discussing the outcomes of providing supplemental feeding to habituated individuals of the obscure species Stultus saxum? Nah, why bother? Pardon the interruption.

    Link to this
  190. 190. Grehan1896 1:57 pm 06/27/2014

    “You said that the Pacific Ocean basin in Jurassic in age, and that this is the geological consensus.”
    Sorry – I meant to indicate that this is a minimal age.

    “You also said that this is what necessitates the existence of Jurassic ratites, despite the lack of any fossil or molecular evidence for this.”

    Actually it was the opening of the Mozambique Channel that was cited. Sure, there are no fossils from this time, but even molecular divergence estimates extend ratites well before the oldest fossils.

    “You have developed that idea in these comments as an explanation for phylogenetic relatedness of some South American paleognaths and some New Zealand paleognaths.”

    At this time I do not have any particularly strong opinion about the Pacific range for the ratites. The origin may be Gondwanic with respect to Antarctica or lie further within the Pacific basin. The Pacific ratites form a pattern found in many other groups with greater or lesser ranges in details (Nothofagus for example).

    “Proposing Cretaceous dispersal along igneous provinces is also a smokescreen, because these igneous provinces don’t float across the ocean and become “imbedded” (sic) in other continents.”

    There are geologist who take a different view about that.

    “In addition, you have cited the Alpine Fault in NZ as a biogeographic barrier”

    No I did not. It is a biogeographic boundary for some taxa, and many show correlated disjunctions.

    “ but have not addressed the substantial rain shadow effect that the southern alps produce. I could easily show disjunction on either side of any mountain range, due to the same effect.”

    The mountain range is not the same as the Alpine Fault.

    “You’re asserting the primacy of tectonism as a cause without any acknowledgement or treatment of the more obvious biotic factors (e.g. water availability, temperature, etc) at play.”

    If they are that obvious they can be incorporated into the analysis.

    “Asserting that a bird adapted to rainforest underbrush has an “Indian Ocean Basin” distribution, whereas a dryland/upland bird has a “Pacific Ocean Basin” distribution is circular at best and ignores key biotic factors (distribution of habitat) which should be considered first.”

    Not sure I understand the circularity. Kiwi and elephant birds are disconnected across the Indian Ocean basin (using the minimum distance criterion) and the moa-tinamous are disconnected by the Pacific basin. I assume these are the rainforest and dryland/upland birds you are referring to.

    “I am generally amenable to reasonable discussion about tectonism and how it contributes to dispersal or endemism of taxonomic groups. However, such reasonable discussion requires that all parties involved have a reasonable understanding and appreciation of the biotic factors that govern distribution of the taxa under discussion as well as the actual tectonic history of a region. It also requires that parties be willing to state and adhere to conditions”

    as you define them I presume

    “which, if met, limit the applicability of their model. I’m not seeing that in this conversation.”

    If you find the conversation unproductive then certainly don’t let me trouble you any further.

    John Grehan

    Link to this
  191. 191. irenedelse 2:00 pm 06/27/2014

    “What interests me is what Pacific paleogeographies contributed to the existence of trans-Pacific distribution.”

    Pacific and Indian ocean paleogeography very likely involved flighted stem-Paleognaths dispersing over water. After all, let’s not forget tinamous in all the kiwi and elephant bird controversy: they are still flighted, and nested within “Ratites”.

    Link to this
  192. 192. irenedelse 2:12 pm 06/27/2014

    @Heteromeles #189:

    Depriving a living, sentient creature of its accustomed source of nourishment would be cruelty.

    Link to this
  193. 193. Grehan1896 2:17 pm 06/27/2014

    “Pacific and Indian ocean paleogeography very likely involved flighted stem-Paleognaths dispersing over water. “

    And what makes that “very likely”.

    “After all, let’s not forget tinamous in all the kiwi and elephant bird controversy: they are still flighted, and nested within “Ratites”.”

    True, but that goes back to the first point above.

    John Grehan

    Link to this
  194. 194. naishd 2:45 pm 06/27/2014

    A sometimes useful test of someone’s position with respect to a given theory or hypothesis concerns their approach to falsification. As stated in several comments upthread (by myself and several other commenters), I don’t think that any of the evidence supports a pre-Cretaceous origin for neornithine birds, let alone for palaeognaths or crown-palaeognaths in particular. But… I would change my mind if a pre-Cretaceous neornithine was discovered.

    So, John — what sort of evidence would disprove your hypothesis that palaeognath distribution is contingent on continental fragmentation? Because, at the moment, it’s sounding like a conviction or belief, the mandate of a philosophy, more than a hypothesis. Sorry if this sounds accusational or confrontational but, well, there it is.

    Link to this
  195. 195. naishd 3:23 pm 06/27/2014

    Oh, and… first one to 200 comments wins the internet.

    Link to this
  196. 196. keesey@gmail.com 3:28 pm 06/27/2014

    I know others have gone into it, but I still feel like John Grehan isn’t quite grasping how unlikely his scenario is from given the accepted phylogeny. It requires a remarkable number of coincidences.

    Here are the major clades of winged theropods that crown-group avians belong to:

    - Aves sensu stricto (crown group, split into Pan-Palaeognathae and Pan-Neognathae)
    - Carinatae sensu stricto (also includes Ambiortiformes)
    - an unnamed clade that also includes Ichthyornithes and Hesperornithes
    - Euornithes (also includes Patagopterygidae, Hongshanornithidae, Songlingornithidae, Gansus, etc.)
    - Ornithothoraces (also includes Enantiornithes)
    - Pygostylia (also includes Confuciusornithidae)
    - Avebrevicauda (also includes Omnivoropterygidae)
    - Euavialae (also includes Dalianraptor, Zhongjianornis, etc.)
    - Ornithurae sensu lato (also includes Shenzhouraptor)
    - Eumaniraptora (also include dromaeosaurids, unenlagiines, troodontids, Archaeopteryx, Anchiornis, Aurornis, Xiaotingia, etc.)
    - Paraves (also includes Pedopenna)
    - Chuniaoae (also includes Oviraptorosauria)
    - Maniraptora (also includes Therizinosauria)
    (Wings might go back a bit further, to Maniraptoriformes, but let’s just stop here.)

    Illustrated here: http://phylopic.org/name/667b8664-5dd7-487e-ba56-c6c5c95055a2/lineage/

    Here are the earliest known occurrences of sister groups:
    - Ambiortiformes: Barremian (EK)
    - Hesperornithes: Cenomanian (early LK)
    - basal euornithians: Barremian (EK)
    - Enantiornithes: Barremian (EK — possibly earlier in EK)
    - Confuciusornithidae: Hauterivian (EK)
    - Omnivoropterygidae: Aptian (EK)
    - basal euavialians: Aptian (EK)
    - Shenzhouraptor: Aptian (EK)
    - basal eumaniraptors: Oxfordian (LJ)
    - Pedopenna: MJ or LJ
    - Oviraptorosauria: EK
    - Therizinosauria: EJ? (if Eshanosaurus belongs — otherwise EK)

    Now, clearly there are ghost lineages. Notably, Oviraptorosauria may have one longer than the MJ and the LJ combined (although that hinges on the identification of a particularly controversial specimen). But there’s still a general order in this — there’s some signal. In most cases, the earliest occurrence of a “side branch” is coeval with or earlier than the earliest occurrence of the “side branch” that is closer to crown-group avians.

    John Grehan’s scenario pushes the palaeognathian/neognathian split down to the Jurassic. This must have occurred after all these other splits, giving all of these taxa ghost lineages that stretch tens of millions of years. For example, Archaeopteryx must have existed alongside not just stem-palaeognathians and stem-neognathians, but early ambiortiforms, hesperornithians, ichthyornithians, various basal euornithians, enantiornithians, confuciusornithids, etc.

    But look what this means: NEARLY EVERY FOSSIL WE FIND MUST BE A RELIC IN ITS OWN TIME. We only find non-ornithurians in the Jurassic. And these are so morphologically “primitive” that we still can’t sort out which later lineages they’re allied to.

    Why? Why do we only find the relics, when there must have been dozens of more derived lineages living alongside them? What would make basal taxa so prominent in the fossil record, and derived taxa so scarce?

    Link to this
  197. 197. irenedelse 3:33 pm 06/27/2014

    @John Grehan #192:

    Parsimony makes dispersion likely. Why the need to explain the distribution of Paleognaths as if the whole group had always been flightless, when there are extant flighted species?

    Link to this
  198. 198. ectodysplasin 3:41 pm 06/27/2014

    John,

    “I am generally amenable to reasonable discussion about tectonism and how it contributes to dispersal or endemism of taxonomic groups. However, such reasonable discussion requires that all parties involved have a reasonable understanding and appreciation of the biotic factors that govern distribution of the taxa under discussion as well as the actual tectonic history of a region. It also requires that parties be willing to state and adhere to conditions”

    as you define them I presume

    Incorrect, but nice of you to play the victim card. I’m sure there are some situations where some aspect of your theoretical framework can be used fruitfully. What I want is for you to define the assumptions of your method and the sorts of data that violate the assumptions of your method.

    Link to this
  199. 199. Grehan1896 4:03 pm 06/27/2014

    “Parsimony makes dispersion likely. Why the need to explain the distribution of Paleognaths as if the whole group had always been flightless, when there are extant flighted species?”

    What kind of “parsimony”. What makes it “likely” that flight was involved just because the birds (assuming for the sake of argument) could fly? What evidence do you propose?

    John Grehan

    Link to this
  200. 200. irenedelse 4:18 pm 06/27/2014

    There’s something ironic here. If we take a step back from the nitty gritty of the debate, isn’t the distribution of ratites an interesting example of paleobiogeography at work? If I understand correctly, the current reconstitution of Paleognath evolution involves several groups dispersing and colonizing a fragmented Gondwana in the late Cretaceous or early Paleogene, and taking advantage of the niches opened by the K/Pg extinction for large(ish) land animals on the different continents. Convergence shaped similarly moa and elephant birds (as well as emus, ostriches, etc.) while in different niches some their close relatives stayed small and in the case of tinamous, kept the ability to fly. Further tectonic activity cemented the geographic separation of otherwise related groups.

    Link to this
  201. 201. irenedelse 4:26 pm 06/27/2014

    @John Grehan:

    “What makes it “likely” that flight was involved just because the birds (assuming for the sake of argument) could fly?”

    Ha. Of course, they *could* have rafted across the ocean. South American monkeys did it, didn’t they?

    Link to this
  202. 202. Grehan1896 4:27 pm 06/27/2014

    “Incorrect, but nice of you to play the victim card.”
    Nope – just pointing out that different people have different ideas about these matters.
    “I’m sure there are some situations where some aspect of your theoretical framework can be used fruitfully.”

    Thank you. Others may disagree of course.

    “ What I want is for you to define the assumptions of your method and the sorts of data that violate the assumptions of your method.”

    1. Distribution patterns constitutes an empirical database for biogeographical analysis.
    2. Distribution patterns provide information about where, when, and how animals and plants evolve.
    3. The spatial and temporal component of these distribution patterns can be graphically represented.
    4. Testable hypotheses about historical relationships between the evolution of distributionsand earth history can be derived from geographic correlations between distribution graphs and geological/geomorphic features.

    I’m not sure the method can be falsified as such since it can be applied to any data. I guess it if were proven in some way that the multitude of vicariant correlations with tectonics represented artifacts (i.e. were not historically connected) I suppose the method might be invalidated.

    Also, please clarify your contention about circularity with respect to my comment:

    Not sure I understand the circularity. Kiwi and elephant birds are disconnected across the Indian Ocean basin (using the minimum distance criterion) and the moa-tinamous are disconnected by the Pacific basin. I assume these are the rainforest and dryland/upland birds you are referring to.

    Also, what would represent a falsification of your hypothesis that kiwis and elephant birds flew between their current locations?

    John Grehan

    Link to this
  203. 203. Grehan1896 4:28 pm 06/27/2014

    “What makes it “likely” that flight was involved just because the birds (assuming for the sake of argument) could fly?”

    Ha. Of course, they *could* have rafted across the ocean. South American monkeys did it, didn’t they?”

    They did? Evidence? Still not answered the question

    John Grehan

    Link to this
  204. 204. Grehan1896 4:37 pm 06/27/2014

    “There’s something ironic here. If we take a step back from the nitty gritty of the debate, isn’t the distribution of ratites an interesting example of paleobiogeography at work? If I understand correctly, the current reconstitution of Paleognath evolution involves several groups dispersing and colonizing a fragmented Gondwana in the late Cretaceous or early Paleogene, and taking advantage of the niches opened by the K/Pg extinction for large(ish) land animals on the different continents. Convergence shaped similarly moa and elephant birds (as well as emus, ostriches, etc.) while in different niches some their close relatives stayed small and in the case of tinamous, kept the ability to fly. Further tectonic activity cemented the geographic separation of otherwise related groups.”

    Its a nice story.

    John Grehan

    Link to this
  205. 205. Grehan1896 4:57 pm 06/27/2014

    “I know others have gone into it, but I still feel like John Grehan isn’t quite grasping how unlikely his scenario is from given the accepted phylogeny. It requires a remarkable number of coincidences.”

    It’s not the phylogeny that is the problem. It what is or is not recorded in the fossil record

    “John Grehan’s scenario pushes the palaeognathian/neognathian split down to the Jurassic. This must have occurred after all these other splits, giving all of these taxa ghost lineages that stretch tens of millions of years. “
    Yes that is the implication. But then some molecular divergence theorists proposed 135 Ma for ratites, if that’s ok, what’s another 30 odd million years?

    “For example, Archaeopteryx must have existed alongside not just stem-palaeognathians and stem-neognathians, but early ambiortiforms, hesperornithians, ichthyornithians, various basal euornithians, enantiornithians, confuciusornithids, etc.”

    Yes

    “Why? Why do we only find the relics, when there must have been dozens of more derived lineages living alongside them? What would make basal taxa so prominent in the fossil record, and derived taxa so scarce?”
    Yes, that’s the core question. Either they were not there (as most do believe and I can quite understand that), or they were and there was something about the derived lineages that made them both non-apparent in the early fossil record and also survived the K/T event. I’m the first to recognize that I am sticking my neck out, and I have no final ‘proof’ that is going to make anyone change their minds. But the active dispersal model sets up its own problems. It’s easy to say that a kiwi or elephant bird flew one way or the other, but less easy to explain why it did not jump across to Africa, did not drop off in Australia (including Tasmania) or bothered to take a trip to New Caledonia or even New Guinea.

    John Grehan

    Link to this
  206. 206. ectodysplasin 5:08 pm 06/27/2014

    @John,

    Not sure I understand the circularity. Kiwi and elephant birds are disconnected across the Indian Ocean basin (using the minimum distance criterion) and the moa-tinamous are disconnected by the Pacific basin. I assume these are the rainforest and dryland/upland birds you are referring to.

    These animals do not live in oceanic basins. They have fixed ranges on land. To claim that kiwis and elephant birds as a clade have an Indian Ocean distribution is an assumption based on the idea that these oceanic basins are the important factor. You cannot then use this interpretation as evidence of correlation. To do that is circular.

    Link to this
  207. 207. Grehan1896 5:09 pm 06/27/2014

    “A sometimes useful test of someone’s position with respect to a given theory or hypothesis concerns their approach to falsification. As stated in several comments upthread (by myself and several other commenters), I don’t think that any of the evidence supports a pre-Cretaceous origin for neornithine birds, let alone for palaeognaths or crown-palaeognaths in particular. But… I would change my mind if a pre-Cretaceous neornithine was discovered.”

    It’s not quite true that none of the evidence supports a pre-Cretacouos origin. I gree that the fossil record does not, but the biogeographic evidence does. On the other hand the fossil record does not provide evidence against a pre-Cretaceous origin other than the interpretation that the absence of fossils is real.

    “So, John — what sort of evidence would disprove your hypothesis that palaeognath distribution is contingent on continental fragmentation? Because, at the moment, it’s sounding like a conviction or belief, the mandate of a philosophy, more than a hypothesis. “

    Sorry that it comes across that way, but the geological model is indeed a hypothesis based on the way the patterns and boundaries of ratite taxa are spatially related to various Mesozoic tectonic features. I am not sure I can think of a disproof. It would be like asking what would disprove that a flighted ancestor of kiwis and elephant birds flew between the two localities.

    “Sorry if this sounds accusational or confrontational but, well, there it is.”

    No it does not. One has to be direct as possible to make the point.

    John Grehan

    Link to this
  208. 208. Grehan1896 5:14 pm 06/27/2014

    On a general note, my approach to falsification is probably a little different from most (all) reading this list. I am more interested in corroboration than falsification. I read somewhere that relativity theory was falsified the day after it was published (or something like that). People engaged in witchcraft make falsifiable hypotheses all the time. Barbara McClintock was ridiculed for years over her theory of mobile genes. Perhaps she was asked how she would falsify her theory too.

    John Grehan

    Link to this
  209. 209. ectodysplasin 5:32 pm 06/27/2014

    also…

    1. Distribution patterns constitutes an empirical database for biogeographical analysis.
    2. Distribution patterns provide information about where, when, and how animals and plants evolve.
    3. The spatial and temporal component of these distribution patterns can be graphically represented.
    4. Testable hypotheses about historical relationships between the evolution of distributionsand earth history can be derived from geographic correlations between distribution graphs and geological/geomorphic features.

    I’m not sure the method can be falsified as such since it can be applied to any data. I guess it if were proven in some way that the multitude of vicariant correlations with tectonics represented artifacts (i.e. were not historically connected) I suppose the method might be invalidated.

    Okay, so maybe you do not understand what “assumption” means in this context. I’m talking about the conditions that a sample of data must meet for the method to operate as expected. For example, the assumptions of a Student’s T test will be violated by datasets exhibiting high skedasticity. The assumptions of a phylogenetic analysis will be violated by a dataset with nonindependent character evolution. And so on. In other words, any method has implicit or explicit assumptions that a dataset must meet in order for that method to be appropriate to use.

    You’ve done an adequate job of explaining your theory, but you have not done an adequate job of explaining the limitations of the methods you’re proposing, and the minimum conditions any dataset must meet for application of your method to provide robust results.

    This isn’t about trying to falsify the method writ large. This is about trying to understand conditions where your method will fail due to the condition of the data.

    Link to this
  210. 210. irenedelse 5:36 pm 06/27/2014

    “They did? Evidence? Still not answered the question”

    That’s no more absurd than a mid-Jurassic origin of Paleognaths. Less, even, since rafts of vegetation drifting long distance with small animals clinging to them is something observed today, but fossil evidence is far from even pointing at such an early origin for crown birds. (See Naish’s and Keesey’s detailed remarks about the fossil record.)

    Similarly, flighted birds flying across a body of water, especially if there are a few island stops along the route, is not an extraordinary hypothesis. (Need this really be pointed out?) Which is why I answered earlier: parsimony. You may not agree about what constitutes parsimony, but then we are back in philosophical territory.

    Also, you seem to be missing something: I’m not proposing a novel interpretation here, either for Platyrhines or for Paleognaths, but only citing current mainstream hypotheses for the evolution of these groups.

    And for the record, to the question “what would make me change my mind”: a Jurassic neornithe, definitely.

    Link to this
  211. 211. Grehan1896 5:48 pm 06/27/2014

    “These animals do not live in oceanic basins. They have fixed ranges on land. To claim that kiwis and elephant birds as a clade have an Indian Ocean distribution is an assumption based on the idea that these oceanic basins are the important factor. You cannot then use this interpretation as evidence of correlation. To do that is circular.”
    I’m trying to think through your perspective. To rephrase your argument in phylogeny – one links taxa using shared derived characters based on the idea that these characters are the important factor. You cannot then use this interpretation as evidence of correlation (of biological relationship cf spatial relationship).
    You probably outclass me on the theory of methodology as I admit I am no theorist. I cannot even design quantitative algorithms for spatial analysis. From my perspective the question starts which whether or how the distribution of any group is spatially related with geomorphological features. The statement that kiwis and elephant birds have an Indian Ocean baseline is nothing more than an observable fact based on the minimal spanning criterion. The method does provide evidence that the spatial features of this group overlap or span the Indian Ocean rather than the Pacific or some other Ocean basin or other geomorphological feature.

    Still interested in what makes it “likely” that elephant bird or kiwi ancestors flew from one or other locality.

    John Grehan

    Link to this
  212. 212. ectodysplasin 5:53 pm 06/27/2014

    @John,

    On a general note, my approach to falsification is probably a little different from most (all) reading this list. I am more interested in corroboration than falsification. I read somewhere that relativity theory was falsified the day after it was published (or something like that). People engaged in witchcraft make falsifiable hypotheses all the time. Barbara McClintock was ridiculed for years over her theory of mobile genes. Perhaps she was asked how she would falsify her theory too.

    You’re conflating falsification of a specific scenario with falsification of hypotheses of mechanisms. Most of the philosophy of science debate has focused on the study of universal mechanisms; this is a repeated point of Lakatos, Popper, Quine, and lots of other workers. In such a case, it is important to distinguish between universal principles (e.g. general relativity) and localized or situationally-dependent principles (e.g. lateral gene transfer) and to design falsification/corroboration tests that are appropriate.

    This does not apply to specific historical descriptions of what has happened. You repeatedly have responded to incredulity towards Jurassic paleognaths by saying “well, it’s not impossible” which suggests you misunderstand how we assess credibility specific historical scenarios. In such situations, we tend towards the principle of parsimony inasmuch as we don’t want to propose all sorts of weird processes occurring that we do not have independent reason to invoke, and if a result requires us to invoke complicated and singular processes, then we have to remain skeptical of this result until independent lines of evidence are presented which require those singular and complicated processes.

    This is why everyone has made a point of enumerating all the multiple lines of evidence that suggest overseas ratite dispersal long after the Jurassic.

    No one here is claiming that vicariance does not have a place in understanding distribution of organisms. Vicariance very clearly does have a place in understanding distribution of organisms. The only thing people are questioning is your claim that dispersal does not, and that dispersal definitely has nothing to do with paleognath distribution.

    Link to this
  213. 213. keesey@gmail.com 6:07 pm 06/27/2014

    “It’s easy to say that a kiwi or elephant bird flew one way or the other, but less easy to explain why it did not jump across to Africa, did not drop off in Australia (including Tasmania) or bothered to take a trip to New Caledonia or even New Guinea.”

    Who says they didn’t? I don’t understand how you can posit a ghost lineage of stem-kiwis stretching back *150 million years* but then turn around and say that there were categorically never any stem-kiwis in Australinea. Is the fossil record less complete than most researchers think or more complete than most researchers think? You can’t have it both ways.

    Link to this
  214. 214. ectodysplasin 6:09 pm 06/27/2014

    @John,

    I’m trying to think through your perspective. To rephrase your argument in phylogeny – one links taxa using shared derived characters based on the idea that these characters are the important factor. You cannot then use this interpretation as evidence of correlation (of biological relationship cf spatial relationship).

    Phylogeny is based in analysis of molecular sequence data. We know that molecular sequences are vertically transmitted within organisms and we have generally good reason to trust the pattern of ancestry gleaned from comparison of these sequences. In this case, we’re looking at an intrinsic feature of these organisms (DNA sequence data) and using it to generate an inference of relatedness (phylogeny).

    Taxonomic distribution is an extrinsic feature of species, and one which we can directly observe by going out in the field and recording observations. We can use multiple field observations to infer “actual” distribution based on various biogeographical modeling tools.

    So basically we have two independent things:

    *Phylogeny, which is inferred from intrinsic characteristics (sequence, morphology, etc)

    *Biogeography, which is inferred fro extrinsic characteristics (field records)

    We can then make comparisons of these data in such a manner that we can test hypotheses about the relationship between phylogeny and biogeography, because these two sets of inferences come from separate sources of raw data.

    The problem with your approach is that you do not have this separation of inference. You assign biogeographical provinces based on minimum spanning trees that are generated on the basis of the phylogeny. In other words, you are generating your biogeographical provinces in a non-independent manner from generation of phylogeny, because you are using phylogeny to infer these provinces. Thus, the fact that you have correlation between the provinces and the clades means absolutely nothing.

    In the specific case of the moa and kiwi, you have defined animals from overlapping territory as belonging to two separate biogeographical provinces. The justification for this appears to be because of their phyogenetic relationships, and not anything intrinsic or extrinsic about the actual geographical ranges of these animals. The summary results you generate with such an analysis are meaningless because you are simply replicating the phylogenetic result. It’s circular.

    Link to this
  215. 215. MichaelHeads 6:34 pm 06/27/2014

    Christopher Taylor (Comment 151) wrote: ‘…for the most part, the order of appearance of bird lineages in the fossil record does approximate their order of phylogenetic divergence’.

    I’d be interested to know if this has ever been argued in print.

    Link to this
  216. 216. Grehan1896 7:16 pm 06/27/2014

    “Phylogeny is based in analysis of molecular sequence data.”
    Only molecular phylogeny is based on analysis of molecular sequence data.

    “ We know that molecular sequences are vertically transmitted within organisms and we have generally good reason to trust the pattern of ancestry gleaned from comparison of these sequences.”

    Well, they’re not necessarily infallible.

    “In this case, we’re looking at an intrinsic feature of these organisms (DNA sequence data) and using it to generate an inference of relatedness (phylogeny).”
    Distribution is intrinsic in the sense that no organism has no location

    “So basically we have two independent things:
    *Phylogeny, which is inferred from intrinsic characteristics (sequence, morphology, etc)
    *Biogeography, which is inferred fro extrinsic characteristics (field records)
    We can then make comparisons of these data in such a manner that we can test hypotheses about the relationship between phylogeny and biogeography, because these two sets of inferences come from separate sources of raw data.
    The problem with your approach is that you do not have this separation of inference. You assign biogeographical provinces based on minimum spanning trees that are generated on the basis of the phylogeny. ”

    But it is the biogeographic method the generates the assignment, not the phylogeny.

    In other words, you are generating your biogeographical provinces in a non-independent manner from generation of phylogeny, because you are using phylogeny to infer these provinces. Thus, the fact that you have correlation between the provinces and the clades means absolutely nothing.”

    But you miss the point – it is a correlation with a particular province.

    In the specific case of the moa and kiwi, you have defined animals from overlapping territory as belonging to two separate biogeographical provinces. The justification for this appears to be because of their phyogenetic relationships, and not anything intrinsic or extrinsic about the actual geographical ranges of these animals.

    Not correct. The phylogeny does not make their distribution Indian Ocean, only the spatial criterion.

    “The summary results you generate with such an analysis are meaningless because you are simply replicating the phylogenetic result. It’s circular.”

    It may appear that way to you, but it is not.

    John Grehan

    Link to this
  217. 217. ectodysplasin 7:34 pm 06/27/2014

    So, once again, why is the moa “pacific” and the kiwi “indian ocean”?

    What criterion do you use to make that distinction when these animals have/had overlaping ranges? Obviously not a spatial criterion, so then what?

    Link to this
  218. 218. ectodysplasin 7:39 pm 06/27/2014

    In addition, no one is claiming that phylogeneticmethods are infallible. We are only claiming that the methods and their limitations are well-understood, and that the results are independently generated from biogeographical results. As a result, the chain of inference from raw data to phylogeny to phylogeography is reasonably transparent. This is not the case with your minimum spanning trees, which obfuscate a lot of the inferential process.

    Link to this
  219. 219. Grehan1896 8:05 pm 06/27/2014

    “Who says they didn’t? I don’t understand how you can posit a ghost lineage of stem-kiwis stretching back *150 million years* but then turn around and say that there were categorically never any stem-kiwis in Australinea. Is the fossil record less complete than most researchers think or more complete than most researchers think? You can’t have it both ways.”

    Yep, it could be that they did and just became extinct everywhere else. But it’s not just these birds, but everything else besides. If it were not for the match with tectonics that one sees with so many intercontinental groups I would agree that extinction would be just as much an explanation as the absence of fossils.

    John Grehan

    Link to this
  220. 220. Grehan1896 8:08 pm 06/27/2014

    In addition, no one is claiming that phylogenetic methods are infallible.

    OK

    We are only claiming that the methods and their limitations are well-understood,

    Perhaps

    and that the results are independently generated from biogeographical results.

    sure

    As a result, the chain of inference from raw data to phylogeny to phylogeography is reasonably transparent.

    What do you mean by phylogeography?

    This is not the case with your minimum spanning trees, which obfuscate a lot of the inferential process.

    I disagree.

    John Grehan

    Link to this
  221. 221. keesey@gmail.com 8:25 pm 06/27/2014

    “If it were not for the match with tectonics that one sees with so many intercontinental groups I would agree that extinction would be just as much an explanation as the absence of fossils.”

    But is it a such a good match that tectonics completely explains *everything*? Remember that you didn’t have an explanation for why relict forms would be preferentially preserved in the fossil record. That same problem exists for other taxa as well: primates, rodents, etc. You’re invoking a lot of ghost lineages and, worse, a lot of unexplained preservational bias, just to avoid a phenomenon (dispersal) that we already know exists.

    Link to this
  222. 222. Grehan1896 9:42 pm 06/27/2014

    “But is it such a good match that tectonics completely explains *everything*? Remember that you didn’t have an explanation for why relict forms would be preferentially preserved in the fossil record. That same problem exists for other taxa as well: primates, rodents, etc. You’re invoking a lot of ghost lineages and, worse, a lot of unexplained preservational bias, just to avoid a phenomenon (dispersal) that we already know exists.”
    I agree that there is a fossil record anomaly if one accepts the biogeographic evidence. On the other hand, of one took the fossil record at face value then one is left with a biogeographic anomaly. It is not a case of avoiding dispersal. Yes we know it exists, but how it works in relation to the origin of vicariant (allopatric) differentiation is not self evident.
    I refer again to the acceptance of some molecular theorists for the origin of ratites at 135 Ma. Evidently they had no problem with a ghost lineage of at least 75 Ma. If this is acceptable, what criterion would one draw for what is not acceptable – 80 Ma, 90 Ma, 100 Ma?

    John Grehan

    Link to this
  223. 223. Grehan1896 10:10 pm 06/27/2014

    “They did? Evidence? Still not answered the question”
    That’s no more absurd than a mid-Jurassic origin of Paleognaths.

    No I was not trying to say that it was absurd. I was only interested in what you would present as evidence.

    “Less, even, since rafts of vegetation drifting long distance with small animals clinging to them is something observed today, “

    Sure, rafts have been found with animals and plants. But how does one know to what extent such events had much to do with vicarious distributions, and kiwis in particular?

    but fossil evidence is far from even pointing at such an early origin for crown birds. (See Naish’s and Keesey’s detailed remarks about the fossil record.)

    True, it does not point to it, but it does not say anything other than an absence of fossils.

    “Similarly, flighted birds flying across a body of water, especially if there are a few island stops along the route, is not an extraordinary hypothesis. (Need this really be pointed out?)”
    It’s not a matter of being extraordinary, but what evidence is there that such abilities operated in the evolution of vicarious distributions.

    “Which is why I answered earlier: parsimony. You may not agree about what constitutes parsimony, but then we are back in philosophical territory.”
    True, with parsimony one has to specify the criteria for a choice of parsimony.

    “Also, you seem to be missing something: I’m not proposing a novel interpretation here, either for Platyrhines or for Paleognaths, but only citing current mainstream hypotheses for the evolution of these groups.”

    Understood. And I presume you also support such hypotheses.

    “And for the record, to the question “what would make me change my mind”: a Jurassic neornithe, definitely.”
    I do not see how that would change your mind about kiwi and elephant bird distribution evolving through flight. One could still invoke that mechanism, only in an earlier setting. I was interested in what would change your mind about them having flown between the localities.

    John Grehan

    Link to this
  224. 224. keesey@gmail.com 12:09 am 06/28/2014

    “I refer again to the acceptance of some molecular theorists for the origin of ratites at 135 Ma. Evidently they had no problem with a ghost lineage of at least 75 Ma. If this is acceptable, what criterion would one draw for what is not acceptable – 80 Ma, 90 Ma, 100 Ma?”

    You’re right that there is no good way to determine a hard date. Molecular clocks extrapolated from observed intergenerational mutation rates in extant taxa can give a good ballpark estimate, but it gets fuzzy back in time.

    But 135 Mya is a very early molecular estimate (and probably using some kind of calibration other than mutation rates in extant taxa), and it’s still tens of millions of years off from what your hypothesis requires. Other molecular estimates indicate as little as 78 Mya. The average estimate in the literature is about 106 Mya: http://timetree.org/index.php?taxon_a=struthio&taxon_b=corvus&submit=Search

    So, in fact, you also have to explain why they’re so very wrong — even the extreme ones.

    And I refer again to the paradox. If there were dozens of eumaniraptoran lineages, including stem-kiwis, stem-ostriches, stem-moas, etc., in the Late Jurassic, then why do the only Late Jurassic eumaniraptoran fossils ALL have such basal morphology that we can’t even tell if they are troodontids, deinonychosaurs, or on the lineage to ornithurians? None of them are even in the neighborhood of the crown group. Nothing in the Jurassic has pygostyles, alulae, carpometacarpi, tarsometatarsi, retroverted halluces, ungual-free manual digits, scapulocoracoids, keeled sterna, toothlessness. They may well not even have powered flight.

    At this point I find that kind of preservational bias a lot harder to believe than the idea of a handful of early, flying palaeognathians could cross between continents in the Late Cretaceous or Cenozoic.

    Link to this
  225. 225. naishd 5:04 am 06/28/2014

    I absolutely agree with Keesey (comment # 224), and with others who have made the same argument. The idea that palaeognaths have to have a lineage stretching back to the time of continental fragmentation does not fit with any of the copious data we have – it is an incredibly unparsimonious hypothesis that creates a swarm of problems and questions. And – I’m sorry – but it’s obvious that it’s based on the underlying belief of panbiogeography proponents that the fragmentation of landmasses trumps everything else, our knowledge of the fossil record, of phylogenetic patterns, of the ecological and environmental settings associated with the origins of certain groups and so on all being ignored or deemed unworthy of attention.

    Conclusion: it is not difficult to see why panbiogeographers have a bad name and why the panbiogeography movement is generally ignored.

    Link to this
  226. 226. MichaelHeads 5:21 am 06/28/2014

    Panbiogeography has a bad name with those who read the fossil record literally, but not with others – e.g. Oxford University Press (Craw et al., 1999. Panbiogeography: Tracking the history of life), University of California Press (Heads 2012, Molecular panbiogeography of the tropics) and Cambridge University Press (Heads 2014, Biogeography of Australasia: A molecular approach). It’s also well-respected in megadiverse countries such as Mexico and Brazil.

    Christopher Taylor (Comment 151) wrote: ‘…for the most part, the order of appearance of bird lineages in the fossil record does approximate their order of phylogenetic divergence’.

    I’d be interested to know if this has ever been argued in print.

    Link to this
  227. 227. naishd 5:41 am 06/28/2014

    I can’t speak for whatever’s been published in the non-panbiogeography literature on non-tetrapods, but in the modern tetrapod literature – on biogeographic discussions of amphibian, reptile, bird and mammal distribution – panbiogeographical hypotheses are either not mentioned at all, or decried as discordant with everything else we know (I’m thinking here of recent published discussions about frog, caecilian, manatee, chameleon, crocodile, palaeognath and passerine distribution). Clearly, there is a specialised panbiogeography literature that cites and favourably discusses other panbiogeography literature, but the movement is certainly not “well respected” in general. Indeed, I’ve been surprised to see how much hostility and negatively is associated with the mention of the subject (you’d have to see what happened in my twitter feed to appreciate what I’m getting at).

    As for “literal reading of the fossil record” — see the comments above (many of them, but especially comments # 156, 196 and 224). It isn’t as if the adherance to non-panbiogeographic hypotheses relies on rigid literal reading of a select few data-points — rather, it’s that all the fossil evidence we have – ALL OF IT – is inconsistent with what panbiogeography predicts (I am referring here to the palaeognath debate, not organisms in general). In other words, panbiogeography mandates that the fossil record is irrelevant and we can ignore it.

    As for Chris Taylor’s statement (comment # 151) that “the order of appearance of bird lineages in the fossil record does approximate their order of phylogenetic divergence”, see comments 196 and 224, and there are numerous published topologies that match his contention. For Cenozoic birds, see Mayr’s Paleogene Fossil Birds and my own review of the avian fossil record in The Complete Dinosaur 2.

    Link to this
  228. 228. irenedelse 6:24 am 06/28/2014

    @John Grehan:

    “True, it does not point to it, but it does not say anything other than an absence of fossils.”

    And round and round we go…

    ““And for the record, to the question “what would make me change my mind”: a Jurassic neornithe, definitely.”
    I do not see how that would change your mind about kiwi and elephant bird distribution evolving through flight. One could still invoke that mechanism, only in an earlier setting.”

    The point is that *if* we had evidence for modern birds occurring around the time of the fragmentation of Gondwana, it wouldn’t be *necessary* to use the dispersion hypothesis to explain modern Paleognaths distribution. It wouldn’t prove that they didn’t fly, but it would make vicariance a more parsimonious hypothesis than it is now.

    Link to this
  229. 229. MichaelHeads 7:17 am 06/28/2014

    Darren, you wrote: ‘In other words, panbiogeography mandates that the fossil record is irrelevant and we can ignore it’.

    We certainly don’t ignore the fossil record; our publications are full of references to fossils. Fossils are tremendously useful – e.g. if you accept the oldest fossil passerine is Eocene, that rules out younger events being involved in their origin.

    But placing groups accurately on a phylogeny using morphology is often very difficult, and placing fragmentary fossils (with no soft parts etc. etc.) on a phylogeny is much more difficult. Placing older fossils (e.g. Mesozoic birds) on a phylogeny is even more risky.

    For example, tinamous do not have a typical ratite morphology, so they were not placed in Ratitae, where are they are now (thanks to sequencing studies). If tinamous had gone extinct in, say, the Pleistocene ice ages, tinamou fossils (without usable DNA) would not be recognised as Ratitae.

    The molecular studies are indicating that there has been a very large amount of parallelism, even more than was recognised in the pre-molecular era. For example, earlier birds do not have a pygostyle (Keesey, comment 224), but that may not be very informative about which lineage they belonged to. The condensation and telescoping of the body (e.g. the transition from a pinnate tail to a palmate one, giving a pygostyle etc. in birds) is one of the most important parallel trends in vertebrates and is seen throughout the group.

    Thanks for the references, I’ll look at them closely. In the areas I’m more familiar with (the tropics, Australasia) the vast majority of lineages do not have any pre-Quaternary fossil record. For example, the New Zealand Mohouidae are now thought to be sister to a large global complex (Corvida), while the NZ Callaeidae are thought to be sister to the other large global complex of passerines (Passerida), but neither NZ family has pre-Quaternary fossils.

    Link to this
  230. 230. naishd 7:39 am 06/28/2014

    Thank you for the response, Michael.

    As an aside on the position of tinamous: since the new molecular studies were published, a few research teams have noted (so far only in pers. comms. – not in print) that they did, actually, recover tinamous nested within ratites when running anatomy-based datasets alone. They assumed that this was incorrect on the basis of the then consensus. Indications are that the position of tinamous within ratites (viz, their forming a clade with moa) will turn out to be similar to the afrotherian situation: once people look hard enough, targeted searches will allow them to find morphological characters that support the new topology.

    On the usefulness of fossils: the discussion above with John Grehan makes it clear that panbiogeographic hypotheses simply involve special pleading at every corner — the argument amounting to “well, sure, you have fossils, but they’re not fossils of the right age to be useful”. Response: “ok, what about the fossils of this group?”. Response: “well, sure, you have fossils, but they’re not fossils of the right age to be useful”, ad nauseum. As stated many times above, in the case of palaeognaths and a panbiogeographic hypothesis, you have to ignore fossils pertaining to all nodes on the bird cladogram extending back to bird origins! I’m sorry, but that’s ridiculous. We aren’t talking about nematodes here.

    Link to this
  231. 231. Grehan1896 9:42 am 06/28/2014

    “I can’t speak for whatever’s been published in the non-panbiogeography literature on non-tetrapods, but in the modern tetrapod literature – on biogeographic discussions of amphibian, reptile, bird and mammal distribution – panbiogeographical hypotheses are either not mentioned at all, or decried as discordant with everything else we know (I’m thinking here of recent published discussions about frog, caecilian, manatee, chameleon, crocodile, palaeognath and passerine distribution).”

    Most of what I have seen of the current specialised vertebrate literature using molecular divergence estimates is that the generation of divergence dates are considered falsifiers of any earlier dates. So it is not surprising that the panbiogeographic evidence would not be considered (and often enough I doubt that the practitioners are even aware of the literature.

    “Clearly, there is a specialised panbiogeography literature that cites and favourably discusses other panbiogeography literature, but the movement is certainly not “well respected” in general.”

    It is not just well respected in panbiogeographic literature, but in broader biogeographic circles as well. Just not in traditional biogeography.

    “Indeed, I’ve been surprised to see how much hostility and negatively is associated with the mention of the subject (you’d have to see what happened in my twitter feed to appreciate what I’m getting at).”

    I’m not surprised as most people consider their views to be ‘right’ according to common sense, logic etc., and usually have little tolerance for alternatives that appear to fly in the face of that common sense. The hostility directed towards Barbara McClintock was apparently sufficiently intimidating for her to discontinue publishing.

    John Grehan

    Link to this
  232. 232. Grehan1896 9:48 am 06/28/2014

    “You’re right that there is no good way to determine a hard date.”

    We are agreed on that – at least for fossils.

    “Molecular clocks extrapolated from observed intergenerational mutation rates in extant taxa can give a good ballpark estimate, but it gets fuzzy back in time.”

    But how does one know a “good ballpark estimate” that is any different from the fossil record? (This is more of a rhetorical question).

    “But 135 Mya is a very early molecular estimate (and probably using some kind of calibration other than mutation rates in extant taxa), and it’s still tens of millions of years off from what your hypothesis requires. “

    The point is that it was tens of millions of years off the fossil record as well.

    “Other molecular estimates indicate as little as 78 Mya.”

    This estimate is still about 20 Ma before the oldest fossil (I had to do a quick check so maybe I am wrong about that – if so please correct me)

    “The average estimate in the literature is about 106”

    Is the average estimate (average of different studies I presume) acceptable (that would give 40 Ma of ghost lineages in the Cretaceous).

    “Mya: http://timetree.org/index.php?taxon_a=struthio&taxon_b=corvus&submit=Search
    So, in fact, you also have to explain why they’re so very wrong — even the extreme ones.”
    Which of the molecular dates above is considered wrong?

    John Grehan

    Link to this
  233. 233. keesey@gmail.com 10:17 am 06/28/2014

    “Which of the molecular dates above is considered wrong?”

    I would say that, in the aggregate, as ballpark estimates, they’re on to something. They all indicate the Cretaceous. That’s what all the fossil/phylogenetic evidence points to as well.

    But under your hypothesis, they’re all horribly wrong. Every single one of them. Why?

    Link to this
  234. 234. keesey@gmail.com 10:37 am 06/28/2014

    “For example, earlier birds do not have a pygostyle (Keesey, comment 224), but that may not be very informative about which lineage they belonged to.”

    Of course it isn’t — it’s a plesiomorphy.

    “The condensation and telescoping of the body (e.g. the transition from a pinnate tail to a palmate one, giving a pygostyle etc. in birds) is one of the most important parallel trends in vertebrates and is seen throughout the group.”

    And we do see it in maniraptors — the oviraptorosaur Nomingia has a mass of fused vertebrae at the end of its tail as well. But we are capable of detecting that kind of parallelism by looking at the entire morphology.

    But are you seriously suggesting that pygostyles might have evolved convergently *between various crown group lineages*? That the concestor of extant birds had a long, bony tail with unfused vertebrae? Please tell me that’s not what you’re suggesting.

    Link to this
  235. 235. Grehan1896 10:39 am 06/28/2014

    “But under your hypothesis, they’re all horribly wrong. Every single one of them. Why?”

    I consider them all be be minimal estimates. Of course for ratites as a whole they are about 100 Ma off the mark. As there are no ratite fossils in the Cretaceous (if I am correct on that) I would be interested to know from those contributors who have strong views about the age of taxa to tell me which of the molecular divergence ages that precedes the oldest fossil they consider to be ‘reasonable’ and on what criterion.

    John Grehan

    Link to this
  236. 236. Grehan1896 10:45 am 06/28/2014

    They all indicate the Cretaceous. That’s what all the fossil/phylogenetic evidence points to as well.

    But they all only indicate the Cretaceous as a minimal estimate. And when you say Cretaceous, they do not all give the same dates. If Cretaceous is OK, does that mean the 135 Ma estiamte is OK (and is it supported by your understanding of the fossil/phylogenetic record), and if so would another 15 Ma be OK to place ratite origins at the end of the Jurassic?

    John Grehan

    Link to this
  237. 237. Grehan1896 12:57 pm 06/28/2014

    There is such diversity and range of comments posted that I regret that I do not absorb all of them and some I miss in responding. So it’s by accident when items are overlooked. Checking back there was the following comment by Darren (#142):
    “But — what about the new phylogenetic hypotheses? As discussed in the article way above, rheas don’t group with moa and tinamous: rather, the tinamou + moa clade is closer to the Australasian ratite + Madagascan elephant bird clade that it is to rheas. I mean – this topology demands a totally different pattern of continental breakup.”

    With the introduction of the vicariance biogeography method of matching phylogenetic cladograms with geological breakup sequence cladograms back in the 1970’s and 80’s I can understand how this relationship became so widely cemented. However, it does not apply to panbiogeography where the method is not to match phylogenetic sequence with theorized historical sequences, but between biological distribution (in a phylogenetic context) and geomorphological distribution.
    There is no panbiogeographic requirement, therefore, that a relationship and its geographic location has to match any particular breakup sequence.

    “ I would say this is a case where the panbiogeographic hypothesis is, again, discordant with other data.”
    But the data (sequence of breakup) are not necessarily applicable to the differentiation history.

    John Grehan

    Link to this
  238. 238. keesey@gmail.com 4:28 pm 06/28/2014

    “I consider them all be be minimal estimates.”

    Why? Do you think mutation rates have sped up in the Cenozoic?

    “If Cretaceous is OK, does that mean the 135 Ma estimate is OK (and is it supported by your understanding of the fossil/phylogenetic record), and if so would another 15 Ma be OK to place ratite origins at the end of the Jurassic?”

    First of all, that estimate is not for “ratite origins”, but the split between the lineage leading to palaeognathians and the lineage leading to neognathians. “Ratite” has been shown to be an unnatural group, but assuming you mean “palaeognathian”, they would not have appeared until later. (It’s a crown group.)

    Secondly, that is an extreme estimate, so I’d consider it less likely (if still somewhere in the realm of possibility).

    Thirdly, as I’m sure you’re well aware, the breakup of Pangaea didn’t occur right at the J/K boundary — it started over 180 million years ago. Under your hypothesis, if I understand it correctly, there would have to be multiple palaeognathian lineages in place already by that point. The origin of palaeognathians, and the split between stem-palaeognathians and stem-neognathians would have to have been even earlier. So you really need to add more than 55 million years onto what is already an extreme estimate. You’re asking for palaeognathians to have already diversified at a time when we don’t even have any eumaniraptoran fossils!

    I still don’t understand why you would weight the geographic evidence so heavily. We know birds can pass between islands and continents. We know they can settle new, distant land (volcanic islands) and even become flightless on them. And we don’t even need to invoke this happening very often — we can still have most birds living in the same area they hatched. A few dispersals explain it without invoking bizarre anomalies in the fossil record and the rate of mutation.

    And with that, I think I’m pretty much done with this discussion, unless someone can address those anomalies.

    Link to this
  239. 239. MichaelHeads 4:32 pm 06/28/2014

    Keesey (234) suggested: ‘we are capable of detecting that kind of parallelism by looking at the entire morphology’.

    Molecular work has rejected many long-established groups that were based on traditional morphology. For example, dicotyledonous plants are now thought to include monocotyledons, ratite birds now include tinamous, and the mammalian order Artiodactyla now includes Cetacea. Many other cases are known at lower taxonomic levels.Yet fossil-calibrated molecular biogeographic studies depend on the correct identification of fossils and their assignment to a position in a phylogeny, and this is done using traditional morphological homologies. Morphological analyses of living taxa have been wrong so often that it seems strange to base the molecular clock calibration entirely on morphological analysis of fossil material, a much more difficult task.

    Link to this
  240. 240. keesey@gmail.com 5:01 pm 06/28/2014

    “As there are no ratite fossils in the Cretaceous (if I am correct on that)”

    Again, “ratites” are an unnatural group, but there is a possible stem-palaeognathian from around the Campanian-Maastrichtian boundary (~70 Mya): Limenavis.

    Link to this
  241. 241. keesey@gmail.com 5:16 pm 06/28/2014

    @MichaelHeads, please see comment #71 above.

    And as a further point to that: morphological analyses do sometimes have a problem sorting out the branching order when a clade undergoes rapid diversification, such as at the root of Placentalia or Neoaves. Another good example of this, which I explicitly mentioned above, is Eumaniraptora, where we have trouble telling the exact branching order of Dromaeosauridae, Microraptoria, Troodontidae, Unenlagiinae, Ornithurae (sensu lato), etc.

    But this does not mean that all morphology-based analysis is worthless!!! Elsewhere it has been pretty robust. There’s been no upset to the general structure of Aves (Palaeognathae-Neognathae split, with a Galloanserae-Neoaves split within Neognathae) or to Mammalia (Monotreme-Therian split, with a Marsupialia-Placentalia split within Theria). Cetacea may have moved closer to Hippopotamidae, but its internal structure remained the same (Odontoceti-Mysticeti split).

    The deep nesting of crown group Aves within Eumaniraptora is something we can be pretty sure of.

    Link to this
  242. 242. Grehan1896 6:03 pm 06/28/2014

    “I consider them all be be minimal estimates.”
    Why? Do you think mutation rates have sped up in the Cenozoic?

    No, but since they are calibrated using fossils they are minimal estimates.

    “If Cretaceous is OK, does that mean the 135 Ma estimate is OK (and is it supported by your understanding of the fossil/phylogenetic record), and if so would another 15 Ma be OK to place ratite origins at the end of the Jurassic?”

    First of all, that estimate is not for “ratite origins”, but the split between the lineage leading to palaeognathians and the lineage leading to neognathians. “Ratite” has been shown to be an unnatural group, but assuming you mean “palaeognathian”, they would not have appeared until later. (It’s a crown group.)

    Sorry, I get fixed on old terms – palaeognathan is fine.

    “Secondly, that is an extreme estimate, so I’d consider it less likely (if still somewhere in the realm of possibility).”

    Ok, so if the end of the Jurassic is extreme but within the realm of possibility. How far back is also within the realm of possibility in your opinion?

    “Thirdly, as I’m sure you’re well aware, the breakup of Pangaea didn’t occur right at the J/K boundary — it started over 180 million years ago. Under your hypothesis, if I understand it correctly, there would have to be multiple palaeognathian lineages in place already by that point. The origin of palaeognathians, and the split between stem-palaeognathians and stem-neognathians would have to have been even earlier. So you really need to add more than 55 million years onto what is already an extreme estimate. You’re asking for palaeognathians to have already diversified at a time when we don’t even have any eumaniraptoran fossils!”

    True. But then what fossil records would not fit into the picture requiring differentiation at 135Ma?

    “I still don’t understand why you would weight the geographic evidence so heavily.”
    I sympathize with your conundrum.

    “ We know birds can pass between islands and continents.”

    Yes, but how much ability among modern birds is responsible for patterns of differentiation is another matter.

    “We know they can settle new, distant land (volcanic islands) and even become flightless on them.”

    Now you are making a leap to assert that we “know” that flightless species on distant volcanic islands resulted from their having flown such distances. What evidence would you cite?

    “And we don’t even need to invoke this happening very often — we can still have most birds living in the same area they hatched. A few dispersals explain it without invoking bizarre anomalies in the fossil record and the rate of mutation.”

    Trouble is, these dispersals are entirely imaginary. In many cases there are no anomalies in the fossil record – the kiwi is a good case where the claimed kiwi fossil at 16-20 Ma is not out of place with the expectation of modern species existing at that time as indicated by transform fault disjunction.

    And with that, I think I’m pretty much done with this discussion, unless someone can address those anomalies.

    OK, but a pity to bow out when you end with a number of assertions that raise questions as above.

    John Grehan

    Link to this
  243. 243. DavidMarjanovic 8:56 pm 06/28/2014

    there is no objective way to ascertain how much older a group is compared with its oldest fossil

    That’s not quite true (me & Laurin 2007, 2008, 2013a, b; and references therein).

    your interpretation of the fossil record and molecular divergence estimates which misinterpret fossil dates as the actual or maximal age of origin

    I’m afraid that’s a strawman.

    How does inferred flight-capable behavior provide evidence that outweighs biogeographic evidence for vicariance.

    What biogeographic evidence exactly question mark. Joel Cracraft seems to have blithely assumed that ostriches are fundamentally African, but there’s at best no evidence for this; they show up in Eurasia at the same time (early Miocene), if not earlier. There were flightless birds (well, at least one: Eremopezus) in Africa before it collided with Eurasia, but there’s no evidence of ostriches in Africa before that happened.

    I think David Marjanović may have mentioned something in the works questioning Vegavis‘ status as a neornithine

    I did; this issue is now briefly discussed in a paper in the latest issue of JVP (I’ll provide the ref on Monday and maybe quote the whole thing, brief as it is). The most important factor seems to be that the analysis which found it to be an anseriform had too few outgroups.

    The oldest kiwi fossil is only one million years old. Why should the kiwi not be preserved as older fossils when more ‘advanced’ birds are?

    So you didn’t even know that New Zealand has no known terrestrial fossil record from the entire Cenozoic except for the last million years and the single locality of St. Bathans (which is about 20 Ma old and has yielded Proapteryx, which has been discussed in this very thread!)!?!

    If you didn’t know that – frankly, why do you even talk about this topic?

    I have to go to bed, more tomorrow.

    ░░░░░░░░░░░░░░░░░░░░░░░

    David Marjanović & Michel Laurin (2007): Fossils, molecules, divergence times, and the origin of lissamphibians. Systematic Biology 56: 369–388. (open access)
    David Marjanović & Michel Laurin (2008): Assessing confidence intervals for stratigraphic ranges of higher taxa: The case of Lissamphibia. Acta Palaeontologica Polonica 53: 413–432. (open access)
    David Marjanović & Michel Laurin (2013a): The origin(s) of extant amphibians: a review with emphasis on the “lepospondyl hypothesis”. Geodiversitas 35: 207–272. (open access if the new website is finally up now)
    David Marjanović & Michel Laurin (2013b, printed 2014): An updated paleontological timetree of lissamphibians, with comments on the anatomy of Jurassic crown-group salamanders (Urodela). Historical Biology 26: 535–550. (find me in Google Scholar and drop me an e-mail)

    Link to this
  244. 244. DavidMarjanovic 9:00 pm 06/28/2014

    I just scrolled up to the latest comment before mine and found this gem:

    Now you are making a leap to assert that we “know” that flightless species on distant volcanic islands resulted from their having flown such distances. What evidence would you cite?

    Uh… they’re volcanic. They consist of oceanic crust. They didn’t exist before a few million years ago (the maximum by far, some 30 Ma, is the case of Fiji) and were never connected to any continent. They only ways to reach them are to fly there, or to swim for weeks.

    A few cases of birds accidentally crossing an ocean and establishing a population on the other side have happened in historical times and have been observed.

    Link to this
  245. 245. LeeB 1 10:16 pm 06/28/2014

    Birds crossing oceans and founding new populations are not that uncommon; the cattle egret crossing the Atlantic is well known, but here are some from Australia to New Zealand: Silvereye (Zosterops lateralis) in the 1850′s; Spur-winged Plover (Vanellus miles) 1932; White Faced Heron (Ardea novaehollandiae) 1940′s; Welcome Swallow (Hirundo neoxena) 1958; Australian Coot (Fulica atra) 1958.
    Other creatures have made the same trip but are less noticeable.
    A good example is that the Australian Moreton Bay Fig (Ficus macrophylla) and Port Jackson Fig (F. rubiginosa)were widely planted in Northern New Zealand but were unable to reproduce without their fig wasp pollinators. However in the 1960′s the pollinating wasp for F. rubiginosa was blown over from Australia and in the 1990′s the pollinator of F. macrophylla arrived.
    Then some other wasps that can only reproduce in already pollinated figs arrived; I think they are up to about five species by now.
    The juvenile fig trees are now noticeable around Auckland growing as epiphytes on trees such as oaks and Canary Island Date Palms.
    Also some Australian butterflies get blown to New Zealand with some regularity; some of which have then bred here.

    LeeB.

    Link to this
  246. 246. Heteromeles 10:53 pm 06/28/2014

    @239: Molecular phylogenetics blew up the old group of dicotyledones, it did not embed monocots (which are a good group) within dicots. There are now multiple ancient lineages (all of which were pre-cladistics considered more or less dicots due to their leaf and wood morphology), monocots, and eudicots (which contain the vast majority of what have always been considered dicots).

    So far as anyone arguing about insular flightlessness, I’d suggest hunting down a copy of Sherwin Carlquist’s old Island Biology. He’s whole got chapters discussing dispersal mechanisms (e.g. wings) and secondary loss of these mechanisms, with fairly exhaustive lists of the same. It’s long out of print, but that’s the point. None of this is new, and the evidence has been around for decades if not centuries.

    Link to this
  247. 247. Grehan1896 11:54 pm 06/28/2014

    “I just scrolled up to the latest comment before mine and found this gem:

    Now you are making a leap to assert that we “know” that flightless species on distant volcanic islands resulted from their having flown such distances. What evidence would you cite?

    Uh… they’re volcanic. They consist of oceanic crust. They didn’t exist before a few million years ago (the maximum by far, some 30 Ma, is the case of Fiji) and were never connected to any continent. They only ways to reach them are to fly there, or to swim for weeks.”

    David, you did sort of walk into that one. There is more to the age of taxa than the age of the islands. Many taxa are dated (by molecular divergence) as being older than the islands they occupy. Even though individual islands may not be very old, they are often parts of clusters of islands that have formed in the vicinity over a very long time. The ability of organisms to move about between these islands certainly contributes to their survival, but is it necessarily the key to their original presence. It is often assumed that volcanic islands have had no contact with any other geological formation, but geological and tectonic research has provided evidence to the contrary.

    “A few cases of birds accidentally crossing an ocean and establishing a population on the other side have happened in historical times and have been observed.”
    As noted on other posts, the question is how much this kind of movement is responsible for the patterns of endemism and relationship that do occur.

    Since you feel so certain about Fiji in particular I would be interested to know what you think of the paper Heads, M. 2006. Seed plants of Fiji: an ecological analysis. Biological Journal of the Linnaean Society 89, 407-431. Here is an excerpt from the abstract:
    The complex geology of Fiji is determined by its position between two subduction zones of opposite polarity, the Vanuatu and Tonga Trenches, in what is currently a region of transform faulting. The large islands comprise fragments of island arcs that have amalgamated and welded together. There has been considerable uplift as well as subsidence in the islands and it is suggested that both these processes have had drastic effects on the altitudinal range of the taxa. Limestone and mangrove floras could have provided a widespread, diverse ancestral species pool from which freshwater swamp forest, lowland rainforest, dry forest, secondary forest, thickets, and montane forest have been derived during phases of uplift.

    Another ‘classic’ volcanic island system is the Galapagos where almost everyone assumes that all the endemics originated by overwater dispersal from other places, and yet geological reconstructions for the late Cretaceous and early Cenozoic identify a large island arc system that passed over the Galapagos hotspot. This intersection may be responsible for much of the endemic life that managed to survive by colonizing new Galapagos islands as the old ones sank (and is concordant with the location of some mainland relatives that are confined to terranes of Pacific origin.

    John Grehan

    Link to this
  248. 248. Grehan1896 12:21 am 06/29/2014

    LeeB notes that “Birds crossing oceans and founding new populations are not that uncommon etc. etc.”
    Yes, but the question remains, to what extent does this represent the evolution of vicariant differentiation. The correlation of distributions with tectonics for many groups that have been studies represents evidence that in these cases the ability of individual organisms to fly, walk, raft etc may have more to do with survival than the origin of the distributions.

    “Also some Australian butterflies get blown to New Zealand with some regularity; some of which have then bred here.”

    Some, being very few indeed. It’s a while since I have looked at this, but I recollect that the level of endemism is very high. Same for most NZ insects. There is one group of tiger moths where the females are flightless. No relatives in Australia, but related species in S America (all with flightless females) all the way up the Andes to the Caribbean.

    John Grehan

    Link to this
  249. 249. Grehan1896 12:22 am 06/29/2014

    So far as anyone arguing about insular flightlessness, I’d suggest hunting down a copy of Sherwin Carlquist’s old Island Biology. He’s whole got chapters discussing dispersal mechanisms (e.g. wings) and secondary loss of these mechanisms, with fairly exhaustive lists of the same. It’s long out of print, but that’s the point. None of this is new, and the evidence has been around for decades if not centuries.”

    I actually have this book. What he regards as dispersal mechanisms are really more about survival mechanisms.

    John Grehan

    Link to this
  250. 250. LeeB 1 12:57 am 06/29/2014

    Interesting that you can accept that organisms can survive on volcanic island groups presumably dispersing from one to the other as the older ones erode away and new ones form yet have trouble with organisms dispersing to the island groups even though the last century or so of New Zealand’s biological history demonstrates that this long distance dispersal occurs with some regularity; especially with flighted organisms such as birds.

    LeeB.

    Link to this
  251. 251. MichaelHeads 12:57 am 06/29/2014

    Comment 246: ‘Molecular phylogenetics blew up the old group of dicotyledones, it did not embed monocots (which are a good group) within dicots’.

    If you look up a pre-molecular text, angiosperms comprised two groups, dicots (including Amborella, magnoliids etc. as well as what are now termed eudicots) and monocots. Monocots are now embedded in the old ‘dicots’. In the same way, Ratitae Merrem 1813 now includes tinamous. (There’s no need to change the name to Palaeognathae).

    Link to this
  252. 252. irenedelse 6:18 am 06/29/2014

    Panbiogeography: let’s rewrite vertebrate evolution, plant evolution, ecology, geology, taxonomy, epistemology… Wheee!

    Link to this
  253. 253. Gigantala 9:48 am 06/29/2014

    The sad thing is that arguing for a vicariant distribution of ratites makes even less sense when you consider the MANY northern hemisphere palaeognaths like Palaeotis and the “lithornithids”. So ratites evolved in the Triassic? Because if we’re equating their diversity with continental drift, we might as well go all the way.

    Modern Palaeognathae distribution is not vicariant, it is relictual. Palaeognaths were a globally distributed clade until the late Eocene, where northern hemisphere species with the possible exception of early ostriches became extinct; we know for a fact that the distribution of modern clades isn’t restricted to specific gondwannan landmasses.

    Link to this
  254. 254. DavidMarjanovic 10:03 am 06/29/2014

    There’s no need to change the name to Palaeognathae

    Why do you say “change”? Palaeognathae Pycraft, 1900, has been the name for “ratites” + tinamous ever since it was coined.

    David, you did sort of walk into that one. There is more to the age of taxa than the age of the islands. Many taxa are dated (by molecular divergence) as being older than the islands they occupy. Even though individual islands may not be very old, they are often parts of clusters of islands that have formed in the vicinity over a very long time.

    Still, the cluster as a whole is volcanic in the vast majority of cases, and has never been in contact with a continent. Hawaii continues through Midway and the Emperor seamounts, so anything endemic to Hawaii today may be older than any individual currently existing island if you allow for dispersal between neighboring islands – but what continent has the whole thing ever been in contact with?

    It is often assumed that volcanic islands have had no contact with any other geological formation, but geological and tectonic research has provided evidence to the contrary.

    Examples, please. And you’re misusing the term formation.

    Since you feel so certain about Fiji in particular I would be interested to know what you think of the paper Heads, M. 2006. Seed plants of Fiji: an ecological analysis. Biological Journal of the Linnaean[*] Society 89, 407-431. Here is an excerpt from the abstract:

    Thanks for the ref, I’ll read the paper tomorrow. What you quote doesn’t mention any connection to any continent; the iguanas may well have been there for 30 Ma, but they must have gotten there by overwater dispersal (probably from Asia).

    * No, Linnean, because it’s named for Carl von Linné as he was called after 1761.

    Another ‘classic’ volcanic island system is the Galapagos where almost everyone assumes that all the endemics originated by overwater dispersal from other places, and yet geological reconstructions for the late Cretaceous and early Cenozoic identify a large island arc system that passed over the Galapagos hotspot.

    And what continent were those island arcs ever connected to?

    BTW, is there something in the water in Galápagos that increases the fidelity of DNA replication by a factor of 10 or 20? You’re postulating extremely slow evolution here.

    “Also some Australian butterflies get blown to New Zealand with some regularity; some of which have then bred here.”

    Some, being very few indeed.

    “Few” doesn’t mean “not enough”. Madagascar was colonized by placental mammals five times in 66 million years (not counting bats or humans), and it’s full of placentals now.

    Link to this
  255. 255. Heteromeles 10:50 am 06/29/2014

    @251: Yes, the Stultus saxum are getting quite a bit of food now. Well done, all. I hope this is what Darren wanted.

    Anyway, I’ll add to the feeding frenzy, despite internet lore warning that this is a really bad idea, by pointing out that MichaelHeads would have failed a botany exam with the idea that monocots are now embedded in dicots. The term dicots was discarded over a decade ago, because it doesn’t adequately describe the reality the botanists see in the data at hand.

    Resurrecting dicots and claiming that monocots are crammed into them happens to be wrong, because we have another group name for the group that includes both monocots and dicots: flowering plants, aka angiosperms. Dicots are not all flowering plants, and that’s why this answer is wrong, and why the term “dicots” is no longer used. Most of them are now in a group called “eudicots,” but eudicots and dicots are not synonymous.

    Link to this
  256. 256. Grehan1896 11:29 am 06/29/2014

    “Interesting that you can accept that organisms can survive on volcanic island groups presumably dispersing from one to the other as the older ones erode away and new ones form yet have trouble with organisms dispersing to the island groups even though the last century or so of New Zealand’s biological history demonstrates that this long distance dispersal occurs with some regularity; especially with flighted organisms such as birds.”

    LeeB.

    I would be interested to know what you cite, other than molecular divergence postdating any geological connections, as evidence that this process occurred with some regularity for the origin of New Zealand’s endemic biota.

    John Grehan

    Link to this
  257. 257. Grehan1896 11:33 am 06/29/2014

    “Panbiogeography: let’s rewrite vertebrate evolution, plant evolution, ecology, geology, taxonomy, epistemology… Wheee!”

    My understanding that epistemology is the theory and nature of knowledge so that would not have to be rewritten. Darwin’s work required a re-writing of evolution and McClintock’s work required a re-writing of genetics. Wheee!

    John Grehan

    Link to this
  258. 258. Grehan1896 12:10 pm 06/29/2014

    David, you did sort of walk into that one. There is more to the age of taxa than the age of the islands. Many taxa are dated (by molecular divergence) as being older than the islands they occupy. Even though individual islands may not be very old, they are often parts of clusters of islands that have formed in the vicinity over a very long time.

    “Still, the cluster as a whole is volcanic in the vast majority of cases, and has never been in contact with a continent.”

    A cluster of volcanic islands does not have to be in direct contact with a continent for a geological connection to have existed for the biota.

    “ Hawaii continues through Midway and the Emperor seamounts, so anything endemic to Hawaii today may be older than any individual currently existing island if you allow for dispersalbetween neighboring islands – but what continent has the whole thing ever been in contact with?”

    Again, no direct contact. But there are a number of geological models requiring major basaltic plateau in the central Pacific. You can read about these in more detail in Heads’ books.

    “It is often assumed that volcanic islands have had no contact with any other geological formation, but geological and tectonic research has provided evidence to the contrary.

    Examples, please. And you’re misusing the term formation.”

    Island arc moving across the Galapagos.

    “Since you feel so certain about Fiji in particular I would be interested to know what you think of the paper Heads, M. 2006. Seed plants of Fiji: an ecological analysis. Biological Journal of the Linnaean[*] Society 89, 407-431. Here is an excerpt from the abstract:
    Thanks for the ref, I’ll read the paper tomorrow. What you quote doesn’t mention any connection to any continent;”

    Fiji and other volcanic islands rest upon various ridges that were formerly subarial and connected with continental areas in the region. More detail on that is in Heads, M. 2008. Panbiogeography of New Caledonia, south-west Pacific: basal angiosperms on basement terranes, ultramafic endemics inherited from volcanic island arcs and old taxa endemic to young islands. Journal of Biogeography 35, 2153-2175.”

    “the iguanas may well have been there for 30 Ma, but they must have gotten there by overwater dispersal (probably from Asia).”

    Why probably from Asia when there are no close relatives in Asia? According to the molecular phylogeny of Noonan and Sites (2010) the iguanas of Fiji and Tonga are sistergroup to Dipsosaurus of Baja California and SW USA and Mexico.

    “And what continent were those island arcs ever connected to?”

    North, Central, and South America.
    “BTW, is there something in the water in Galápagos that increases the fidelity of DNA replication by a factor of 10 or 20? You’re postulating extremely slow evolution here.”

    Need to be explicit as to what you think is the problem.

    “Few” doesn’t mean “not enough”.

    True, but there is nothing to indicate that any of the endemic species arose this way.

    “Madagascar was colonized by placental mammals five times in 66 million years (not counting bats or humans), and it’s full of placentals now.”

    You assert this as fact, but it is not.

    John Grehan

    Link to this
  259. 259. Grehan1896 12:16 pm 06/29/2014

    “Modern Palaeognathae distribution is not vicariant, it is relictual.”

    Modern ratites (including Tinamous) have a distribution that is largely vicarianct. That is not precluded by being relictual.

    “Palaeognaths were a globally distributed clade until the late Eocene, where northern hemisphere species with the possible exception of early ostriches became extinct;”

    Sure, that is what I have been saying – a globally distributed CLADE – not just some parts of it.
    “we know for a fact that the distribution of modern clades isn’t restricted to specific gondwannan landmasses.”

    Agreed, this has been pointed out in earlier postings. In particular there is an Indian Ocean subclade and a Pacifici subclade.

    John Grehan

    Link to this
  260. 260. irenedelse 12:40 pm 06/29/2014

    “The sad thing is that arguing for a vicariant distribution of ratites makes even less sense when you consider the MANY northern hemisphere palaeognaths like Palaeotis and the “lithornithids”.”

    Especially since several of these ancient Palaeognaths from Europe seem to have been good flyers (cf. Peter W. Houde, 1988, Paleognathous birds from the early Tertiary of the northern hemisphere, through Google Books.) As an aside, the presence of several palaeognath lineages which are either flightless or with reduced flight capacity is interesting. Adaptation to a terrestrial lifestyle seemed to have been an obvious “choice” for this group of birds. Wonder if they were preadapted for this or if it’s only the consequence of the niches they occupy.

    “Still, the cluster as a whole is volcanic in the vast majority of cases, and has never been in contact with a continent. Hawaii continues through Midway and the Emperor seamounts, so anything endemic to Hawaii today may be older than any individual currently existing island if you allow for dispersal between neighboring islands – but what continent has the whole thing ever been in contact with?”

    In this discussion of volcanic archipelagos, land colonisation and dispersion from neighbouring islands, may I bring up the case of Surtsey?

    Historical observations of dispersion over water and colonisation of a few newly erected islands by plants, fungi, invertebrates, birds and pinnipeds. Means of dispersion include insects and other invertebrates carried by drifting logs and on birds’ feet or feathers. Adult individuals can be carried this way but also larvae and eggs. Birds fly, marine mammals swim, seeds and spores are carried by the wind, by marine currents, on the animals or in their guts…

    “Anyway, I’ll add to the feeding frenzy, despite internet lore warning that this is a really bad idea”

    Let’s say that observation of the resultant micro-ecosystem is often quite educative.

    Link to this
  261. 261. irenedelse 1:06 pm 06/29/2014

    “Darwin’s work required a re-writing of evolution and McClintock’s work required a re-writing of genetics.”

    What, no mention of Galileo? Tsk.

    “Sure, that is what I have been saying – a globally distributed CLADE – not just some parts of it.”

    The point, as Gigantala said, is that if you want to explain a global distribution for “Ratites” by simple vicariance, you need to push the origin of Aves back to the Triassic, when Gondwana parted ways with Laurasia. Don’t you notice the problem?

    (By the way, why the insistence on the name Ratites anyway? As David mentioned, Palaeognaths were defined as Ratites + Tinamous, after all. Is it a thing about older names? Or is it the need to distinguish your field with a particular vocabulary? I’m not trying to get snarky here but to understand.)

    Link to this
  262. 262. irenedelse 1:08 pm 06/29/2014

    Darn, must have put too many links in my post, it looks like it’s in moderation.

    Link to this
  263. 263. BrianL 1:14 pm 06/29/2014

    Having been reading all these comments and the endless back-and-forths here, I think the whole discussion, as far as I can remember everything that has been said, would benefit from John Grehan answering the following questions briefly:
    - Why does overwater dispersal explain Silvereyes showing up in New Zealand in recent times, but can it a priori not explain kiwi colonising New Zealand in far more distant times?
    - Do you think molecular datings are inherently wrong?
    - Do you think any fossil that contradicts what you would expect from panbiogeography *has to be* incorrectly identified?
    - Do you think fossils are at all relevant to the conclusions you derive from panbiogeographic principles?
    - Would you agree that you consider panbiogeography the answer to all current biogeographic phenomena, but not historic ones?
    - Do you disagree that overseas dispersal events have been demonstrated?
    - Why do you think species move between islands but not from continents to islands?
    - Why are kiwi neccessarily ‘Indian Ocean’ in origin and moa neccessarily ‘Pacific Ocean’ in origin?
    - Were Madagascar’s hippos the result of Mesozoic vicariance?
    - Do you think that all oceanic islands were originally connected to continents?
    - Do you explain Paleogene European ratites as having been the result of vicariance too? If not, why do you consider them able to disperse but should Gondwanan (and yes, your Pacific and Indian Ocean subclades would be Gondwanan) paleognaths not have been able to?
    - Do you understand why most commenters on here consider the existence of Jurassic or even early Cretaceous ratites and primates very unlikely?
    - I get the impression that you consider everything argued by other people here to be part of an anti-panbiogeographic agenda and that alone means any counter-evidence they provide has to be wrong. Is this impression correct?
    - Do you think *all* biogeographical research should consider vicariance to be null hypothesis for any pattern observed? If not for all recent divergences, why for all ancient ones?

    I think your answers to these questions would be enlightening for everyone following this thread.

    Link to this
  264. 264. Grehan1896 1:15 pm 06/29/2014

    “What, no mention of Galileo? Tsk.”

    There are probably endless examples, I chose a couple from more recent times.

    “Sure, that is what I have been saying – a globally distributed CLADE – not just some parts of it.”
    The point, as Gigantala said, is that if you want to explain a global distribution for “Ratites” by simple vicariance, you need to push the origin of Aves back to the Triassic, when Gondwana parted ways with Laurasia. Don’t you notice the problem?”

    And yet there was agreement that the group was globally distributed

    “(By the way, why the insistence on the name Ratites anyway? As David mentioned, Palaeognaths were defined as Ratites + Tinamous, after all. Is it a thing about older names? Or is it the need to distinguish your field with a particular vocabulary? I’m not trying to get snarky here but to understand.)”

    Names are just labels. Ratite works for me, palaeognath works for others.

    John Grehan

    Link to this
  265. 265. Grehan1896 1:30 pm 06/29/2014

    Brian L. Nice questions!

    - Why does overwater dispersal explain Silvereyes showing up in New Zealand in recent times, but can it a priori not explain kiwi colonising New Zealand in far more distant times?

    Because it only shows that silvereyes did this. It does not show that kiwis necessarily did.

    - Do you think molecular datings are inherently wrong?
    They are “wrong” when they are interpreted as actual or maximal.

    - Do you think any fossil that contradicts what you would expect from panbiogeography *has to be* incorrectly identified?
    No.

    - Do you think fossils are at all relevant to the conclusions you derive from panbiogeographic principles?
    Of course. The panbiogeographic literature is full of examples using the fossil record.

    - Would you agree that you consider panbiogeography the answer to all current biogeographic phenomena, but not historic ones?
    This is a bit of a loaded question (not that you intended it to be that way – perhaps I should say ambiguous to my understanding). Panbiogeography is an approach that can be used for current and historical (fossil) phenomena.

    - Do you disagree that overseas dispersal events have been demonstrated?
    Yes

    - Why do you think species move between islands but not from continents to islands?
    Between islands was referring to relatively local movements (such as within the Hawaiian or Galapagos clusters). That is not to say that some organisms have very wide ranging dispersal capabilities. The biogeographic question is how much do such abilities have to do with the actual course of spatial evolution for groups with vicariant distributions. Individuals may have their different opinions on this, but the matter is not self evident. Panbiogeographic analysis suggests that the patterns are more concordant overall with vicariance than the amount of chance dispersal proposed by molecular dispersalists.

    - Why are kiwi neccessarily ‘Indian Ocean’ in origin and moa neccessarily ‘Pacific Ocean’ in origin?
    Because the nearest kiwi relative (if one accepts the most recent phylogeny as true) is in Madagascar. A minimum spanning link (this is a spatial parsimony criterion) would go across the Indian ocean. That of the moa goes across the Pacific.

    - Were Madagascar’s hippos the result of Mesozoic vicariance?
    Perhaps. I have not analyzed that group.

    - Do you think that all oceanic islands were originally connected to continents?
    No. But the answer involves more complexities that can be best addressed by reading the section on the Pacific in Molecular Panbiogeography of the Tropics.

    - Do you explain Paleogene European ratites as having been the result of vicariance too? If not, why do you consider them able to disperse but should Gondwanan (and yes, your Pacific and Indian Ocean subclades would be Gondwanan) paleognaths not have been able to?
    Which ratites are you referring to?

    - Do you understand why most commenters on here consider the existence of Jurassic or even early Cretaceous ratites and primates very unlikely?
    Yes. Although interestingly some molecular theorists think otherwise for ratites at least.

    - I get the impression that you consider everything argued by other people here to be part of an anti-panbiogeographic agenda and that alone means any counter-evidence they provide has to be wrong. Is this impression correct?

    No. What gives you that impression? The counter evidence does not have to be wrong, it’s just that we are in disagreement over that evidence.

    - Do you think *all* biogeographical research should consider vicariance to be null hypothesis for any pattern observed? If not for all recent divergences, why for all ancient ones?

    No, there should be multiple hypotheses.
    I think your answers to these questions would be enlightening for everyone following this thread.
    Well, enlightenment is something we all search for. Thank you for taking the time to think these questions through. No one contributing to this discussion is necessarily going to change their mind about anything, but that’s OK.

    John Grehan

    Link to this
  266. 266. Gigantala 2:15 pm 06/29/2014

    “Modern ratites (including Tinamous) have a distribution that is largely vicarianct. That is not precluded by being relictual.”

    Except for the fact that several clades demonstrate this to be the case, most notably ostriches, which until very recently were widspread across the Old World (in fact, european fossils far predate african ones).

    “Sure, that is what I have been saying – a globally distributed CLADE – not just some parts of it.”

    Except you’re arguing that modern palaeognath diversity specifically tie in to gondwanna palaeogeographical features.

    “Agreed, this has been pointed out in earlier postings. In particular there is an Indian Ocean subclade and a Pacifici subclade.”

    Both the tinamou+moa clade and the emu+cassowary+elephant bird+kiwi clade would have been of antarctic origin, especially if Diogenornis is more closely related to australasian ratites than to rheas.

    “Which ratites are you referring to?”

    “Lithornithids”, Remiornis, Palaeotis, et cetera. Really, it is very telling that virtually all “ratites are vincariant” literature has to ignore these taxa.

    Link to this
  267. 267. irenedelse 2:28 pm 06/29/2014

    John Grehan at comment #265 (about my asking him what he thought about Palaeognaths in the northern hemisphere during the Cenozoic):

    “And yet there was agreement that the group was globally distributed”

    What is in question is not *that* Palaeognaths were globally distributed. It’s about *how* they came to occupy various land masses. You argue for vicariance as an explanation of the distribution of today’s Palaeognaths in the southern hemisphere. But since related groups (Lithornis, Palaeotis, etc.) used to live in Europe too, do you think we should say that their distribution too is explained by vicariance and the breaking up of Pangaea?

    Link to this
  268. 268. DavidMarjanovic 3:59 pm 06/29/2014

    …OK, I have illustrated the consequences of trying to deal with a thread on two separate days: I’ve repeated part of the discussion.

    From comment 127:

    Also, there is a paper I wrote on the Galapagos in 2001 you can obtain here.

    Naturally, this paper answers some of my questions! However, I find it curious that the paper doesn’t try to evaluate dispersal hypotheses on their own; sea currents are nowhere mentioned in the text or the maps, and it is tacitly assumed that all of the discussed clades completely lack a fossil record outside their current distribution – which is screamingly wrong for iguanas, to pick the example I know.

    The paper tries to show that vicariance explanations aren’t completely impossible (according to geology and exclusively extant biogeography). What it should have tried to show is that vicariance hypotheses are more parsimonious than dispersal hypotheses. And frankly, the reviewers ought to have pointed this out.

    From later comments:

    If “all other data points” (presumably you mean other than biogeographic data points) indicate a mid-Jurassic origin for birds as a while then there would appear to be nothing extraordinary about the basal lineages of extant birds also diverging about that time.

    “Birds as a whole” includes Archaeopteryx and goes way, way, way beyond the crown-group. Cretaceous birds are plentiful in the fossil record as we have come to know it in the last few years, and most or all of them go on various positions in the stem, not the crown. The stratigraphic fit of bird phylogeny is pretty good (see comment 196); as long as evidence to the contrary remains absent, we have to expect the crown-group (Neornithes) to be much younger than birds as a whole.

    The reality is that molecular divergence dates cannot falsify an earlier origin.

    It’s not like falsification is all science can do. There’s also parsimony. (In fact, there’s a lot of parsimony hidden in falsification.)

    The order of divergence does not have to match the order of continental drift.

    What criterion do you have left, then, to test your own biogeographic hypotheses???

    We are not going to change our respective views

    What the fuck.

    How can a scientist say such a thing?!?

    I never know in advance when I’m going to change my views. I never know in advance what new evidence might come to my attention or what new hypotheses are going to be tested. And neither do you, or anybody else.

    So fossil calibrated molecular divergence estimates are not used for nematodes?

    Can’t be, because there’s nothing to calibrate them with. Read Raaum et al. (2005) on how to choose calibration dates.

    It all depends on what one means by “better explains”. Unfortunately there is no universal criterion.

    You fail at basic science theory. Start anew, do not pass Go, do not collect 200 dolaritos.

    For example, the morphogenetic evidence for a sister group relationship between humans/fossil hominids and orangutans is not accepted as the better explanation by the majority of ‘qualified people’ (not sure who is qualified or not) because most in the field believe that sequence similarity is the absolute proof of a closer relationship between humans and chimpanzees.

    1) “Absolute proof” is a disgusting strawman. Proof is for math, logic and American alcohol!
    2) It’s not just sequence similarity. It’s a long list of morphological features as well, plus the fact that many similarities between orang-laut and orang-utan are symplesiomorphies – as shown by the fact that they’re shared with other apes.

    I do take the fossil record into account. But given the biogeographic evidence I am inclined to interpret the fossil record differently to most.

    I’m sorry, this claim is no different to when Ken Ham or Kent Hovind insist that scientists and creationists look at the same evidence and just interpret it differently. Much like what they keep doing, you have demonstrated in this very thread that you don’t know the fossil record of birds or their phylogeny; remember when you implied that Archaeopteryx was just outside the crown-group?

    You also haven’t explained why Neornithes should have a so much worse preservation potential than Enantiornithes throughout the Cretaceous. Did you know Enantiornithes exists?

    The differentiation of anthropoids and prosimians at the Lebombo monocline.

    What? Both strepsi- and haplo(r)rhines are from Eurasia, showing up there in the Eocene.

    “Prosimii” is paraphyletic: the tarsiers are closer to us than to the lemurs & lorises.

    The Pacific Ocean wasn’t formed in the Jurassic. It is the remnant of Panthalassia, which goes back quite a bit further. The fact that the oldest Pacific crust is Jurassic is due to active subduction of oceanic crust along both margins of the Pacific Ocean, a process which, judging from the age of back-arc accretion along both East Asia and western North America, has been going on for the entire history of the Phanerozoic and then some.

    Specifically, North America on the one side and East Antarctica + Australia on the other broke apart some 750 Ma ago. This has been uncontroversial for easily 15 years now.

    At this time I do not have any particularly strong opinion about the Pacific range for the ratites. The origin may be Gondwanic with respect to Antarctica or lie further within the Pacific basin. The Pacific ratites form a pattern found in many other groups with greater or lesser ranges in details (Nothofagus for example).

    All of these are obvious ancient inhabitants of Outer Gondwana ( = Gondwana after Africa had broken away).

    Its a nice story.

    It’s a parsimonious story. Have you got a more parsimonious one?

    some molecular divergence theorists proposed 135 Ma for ratites, if that’s ok, what’s another 30 odd million years?

    I don’t find 135 Ma for a fictitional subset of paleognaths OK at all, no. But even apart from that, the longer a ghost lineage is, the less parsimonious is the hypothesis that this ghost lineage is real; that should be obvious! It’s quantitative, not black-or-white as you appear to believe.

    Panbiogeography has a bad name with those who read the fossil record literally, but not with others – e.g. Oxford University Press (Craw et al., 1999. Panbiogeography: Tracking the history of life), University of California Press (Heads 2012, Molecular panbiogeography of the tropics) and Cambridge University Press (Heads 2014, Biogeography of Australasia: A molecular approach).

    …Did you just seriously try to cite publishers instead of publications as evidence???

    And what’s with this strawman of “reading the fossil record literally”? It sounds like you don’t know what you’re talking about.

    It’s also well-respected in megadiverse countries such as Mexico and Brazil.

    When people’s support for a hypothesis can be predicted from where they work, something is massively, massively wrong either with that hypothesis or with its alternative(s): one of them is – I can’t see a way around this conclusion – propagated by authority and/or by lack of access to publications, not because it explains the data better than any alternative. National science is bad science.

    you’d have to see what happened in my twitter feed to appreciate what I’m getting at

    Speaking of your twitter feed, Darren, the main page of this blog used to show it to us. Then one day it was gone; at the time I assumed that was a temporary glitch, but it has never come back. What happened? Did the benevolent SciAm overlords fear it was distracting people from SciAm?

    The condensation and telescoping of the body (e.g. the transition from a pinnate tail to a palmate one, giving a pygostyle etc. in birds) is one of the most important parallel trends in vertebrates and is seen throughout the group.

    Really? I don’t know about teleosts, but it’s really rare in tetrapods – much less common than elongation, in fact.

    Molecular work has rejected many long-established groups that were based on traditional morphology.

    “Traditional morphology” did not, however, use phylogenetic analysis. It used scenario-based arguments that were founded on picking one or a few characters which were subjectively deemed “important” or “reliable” – and on ignoring all others as “probably convergent”. That wasn’t science, it was art!

    Molecular phylogenetics has used phylogenetic analysis pretty much from the start; it’s really not surprising that the results of phylogenetic analyses of molecular data hold up better than the results of no phylogenetic analysis at all.

    Phylogenetic analysis of morphological data continues to lag behind that of molecular data, because making a dataset for phylogenetic analysis from morphological data is a lot more work than doing that from molecular data. However, where both kinds of data have been analyzed, the results tend to agree pretty well; I can only think of one exception that I can’t begin to explain (squamate phylogeny).

    morphological analyses do sometimes have a problem sorting out the branching order when a clade undergoes rapid diversification, such as at the root of Placentalia or Neoaves.

    Some of these cases don’t have a branching order in the first place; retroposon insertions show incomplete lineage sorting in Neoaves, Scrotifera, and IIRC Placentalia.

    Mammalia (Monotreme-Therian split, with a Marsupialia-Placentalia split within Theria)

    Incidentally, it took molecular biologists a long time to get this one right. From morphology it’s plain obvious, but with molecules you run headfirst into long-branch attraction.

    No, but since they are calibrated using fossils they are minimal estimates.

    This just shows you don’t know how fossils are used to provide calibration dates. :-|

    Usually, their ages are used as hard minimum ages, and no maximum ages are applied (except sometimes a very relaxed one to the root of the tree). This has long been known to generate overestimates.

    Australian Coot (Fulica atra)

    The Australian coot is F. atra australis. F. a. atra spans northern Africa and Eurasia from sea to shining sea.

    Anyway, I’ll add to the feeding frenzy, despite internet lore warning that this is a really bad idea

    First, not everyone who says something wrong is a troll; a troll is someone who tries to make people upset and then laughs when that worked. Second, what’s bad about feeding trolls till they explode?

    A cluster of volcanic islands does not have to be in direct contact with a continent for a geological connection to have existed for the biota.

    So you suddenly accept dispersal from a continent to volcanic islands?

    But there are a number of geological models requiring major basaltic plateau in the central Pacific.

    A basaltic plateau isn’t a continent; and no continent has been anywhere near the central Pacific in a long time.

    Island arc moving across the Galapagos.

    The Galápagos islands aren’t part of an island arc… there’s no subduction going on anywhere nearer to them than the rim of South America, and that subduction zone lacks an island arc.

    “the iguanas may well have been there for 30 Ma, but they must have gotten there by overwater dispersal (probably from Asia).”

    Why probably from Asia when there are no close relatives in Asia?

    Keyword “are”. There used to be iguanas all over Eurasia all the way to the Oligocene.

    According to the molecular phylogeny of Noonan and Sites (2010) the iguanas of Fiji and Tonga are sistergroup to Dipsosaurus of Baja California and SW USA and Mexico.

    The closest living relative isn’t necessarily the sister-group.

    Oh, speaking of Tonga (which has not just iguanas, but even frogs): was there ever a way to walk from Fiji to Tonga?

    “BTW, is there something in the water in Galápagos that increases the fidelity of DNA replication by a factor of 10 or 20? You’re postulating extremely slow evolution here.”

    Need to be explicit as to what you think is the problem.

    Visible mutation rates across generations depend on the fidelity of DNA replication & repair and on natural selection. The latter acts with different strength on different genes, and doesn’t act at all on much of the genome (even within genes, the 3rd position in a codon is often redundant); the former doesn’t vary much. As far as I know, the genomes of the species on Galápagos differ from their closest relatives elsewhere by amounts that are compatible with recent divergences, about as recent as the age of the archipelago. You, however, postulate that these divergences happened in the Cretaceous instead. That means you propose extremely slow rates of evolution. As I just said, proposing mind-blowingly strong stabilizing selection doesn’t work, because much of the genome is invisible to selection in the first place. You are, therefore, proposing mind-blowingly high fidelity of DNA replication & repair! You just haven’t noticed.

    I think you should try the reductio ad absurdum approach found in two papers from early 2004 (at least one of them is in the Journal of Paleontology): if you assume that the abovementioned Polarornis from the end-Cretaceous of Antarctica was a crown-group loon, it follows that the crown-group of birds is as old as the Earth itself.

    “Madagascar was colonized by placental mammals five times in 66 million years (not counting bats or humans), and it’s full of placentals now.”

    You assert this as fact, but it is not.

    I assert it as a well-supported hypothesis. Haven’t you read the papers about it? Not even Yoder et al. (2007, Systematic Biology)?

    Fact: Madagascar has no known Cenozoic terrestrial fossil record except for the last million years or so.
    Fact: Madagascar has a pretty rich end-Cretaceous terrestrial fossil record.
    Fact: all placentals known from Madagascar are either bats, or humans, or have been brought along by humans, or belong to one of these clades: 1) lemurs; 2) euplerid carnivorans; 3) nesomyid rodents; 4) tenrecs (including the extinct Plesiorycteropus); 5) crown-group hippos.
    Fact: these clades are not each other’s closest relatives; instead, each of them has a sister-group in Africa, so 5 is the minimum number of times placentals must have reached Madagascar (not counting bats flying all over the place and humans cheating with ships). You’ll probably agree Madagascar is not where the first placentals lived.
    Fact: there are a few mammals in the Cretaceous (and Jurassic) fossil record of Madagascar, but (so far) not one of them is a placental.
    Fact: Madagascar broke away from Africa in the Jurassic and hasn’t been connected to Antarctica and/or the Seychelles + India since at least the end of the Cretaceous, probably the middle of the Late Cretaceous. Thus, anything that has reached it in the Cenozoic must have dispersed over water.

    As mentioned, the placentals of Madagascar all have their closest relatives in Africa – not in India, the mammal-free Seychelles (again excepting bats and meddling humans), or South America/Antarctica/Australia. Because Madagascar separated from Africa first, you need to postulate an elaborate geographic pattern of extinction, in addition to the usual ghost lineages, if you want to claim that all five of these clades walked to Madagascar when that was still possible.

    Several molecular dating studies have concluded that the placentals of Madagascar all diverged from their closest extant relatives in the Cenozoic (Eocene to Miocene, IIRC). All of these studies used calibration dates from the fossil record (obviously from places outside of Madagascar) as minimum ages, meaning their results are more likely over- than underestimates.

    Where is your more parsimonious alternative?

    “What, no mention of Galileo? Tsk.”

    There are probably endless examples, I chose a couple from more recent times.

    Dude, do we really need to explain to you that the Galileo Gambit is a logical fallacy? Yes, people laughed at Galileo, but (as has famously been pointed out) they also laughed at Bozo the clown; the trick is that Galileo’s hypothesis, and Clintock’s, explained the known data more parsimoniously than the alternatives.

    You need to demonstrate that panbiogeography explains the known data more parsimoniously than occasional dispersal. You haven’t even tried to do that.

    - Why are kiwi neccessarily ‘Indian Ocean’ in origin and moa neccessarily ‘Pacific Ocean’ in origin?
    Because the nearest kiwi relative (if one accepts the most recent phylogeny as true) is in Madagascar. A minimum spanning link (this is a spatial parsimony criterion) would go across the Indian ocean. That of the moa goes across the Pacific.

    Don’t you think that both go across Antarctica instead? Remember, the Earth isn’t flat…

    - Were Madagascar’s hippos the result of Mesozoic vicariance?
    Perhaps. I have not analyzed that group.

    Wow.

    You know what? Next time, learn about a topic before you shoot your mouth off about it. You just said “perhaps” to the equivalent of Precambrian bunnies.

    - Do you understand why most commenters on here consider the existence of Jurassic or even early Cretaceous ratites and primates very unlikely?
    Yes. Although interestingly some molecular theorists think otherwise for ratites at least.

    They don’t so much think otherwise as just not understand the matter much better than you do. If you want to read really cringeworthy things in the peer-reviewed scientific literature, I recommend molecular papers from the late 90s and early 00s.

    Link to this
  269. 269. naishd 4:55 pm 06/29/2014

    I am sorely tempted to close the comments here. Thoughts? Thank you, David, thank you.

    Link to this
  270. 270. irenedelse 5:08 pm 06/29/2014

    Nothing to add. Maybe a mention of Alfred Wegener and the fact that revolutionnary ideas often get extensively revised themselves before they get accepted as the new “orthodoxy”. (Tectonic plates don’t imply continents ploughing the seafloor, but the seafloor itself expanding.)

    Link to this
  271. 271. Gigantala 5:09 pm 06/29/2014

    Please do. I want a rational conversation about palaeognaths, not more trolling.

    Link to this
  272. 272. MichaelHeads 5:10 pm 06/29/2014

    Comment 255: …MichaelHeads would have failed a botany exam with the idea that monocots are now embedded in dicots. The term dicots was discarded over a decade ago, because it doesn’t adequately describe the reality the botanists see in the data at hand.

    Resurrecting dicots and claiming that monocots are crammed into them happens to be wrong, because we have another group name for the group that includes both monocots and dicots: flowering plants, aka angiosperms. Dicots are not all flowering plants, and that’s why this answer is wrong, and why the term “dicots” is no longer used. Most of them are now in a group called “eudicots,” but eudicots and dicots are not synonymous.

    My reply: Molecular studies changed the ideas on angiosperm classification, even at the basal node. Instead of having two groups, dicots and monocots, as in the old view,the monocots are now embedded in the old ‘dicots’, which is now rejected as a group.

    Link to this
  273. 273. MichaelHeads 5:27 pm 06/29/2014

    Comment 212: ‘your claim that dispersal does not, and that dispersal definitely has nothing to do with paleognath distribution’.

    The ratite distributions are very interesting because they show such a high level of allopatry. (John seems to be the only one who has mentioned allopatry in this thread, but it’s a very striking phenomenon). Lithornithids in the far north; ostriches south of them, in Africa and Asia (fossil); elephant birds etc. in Madagascar and Australasia (with the families all allopatric); moas etc.in New Zealand and South America. The whole world is divided up very neatly. Why? How? Furthermore, the division is not random – the patterns of each of the clades are also repeated in many other groups (e.g. the trans Indian Ocean basin and trans-Pacific basin disjunctions). The simplest explanation for the global pattern is in-situ vicariance. But there is some overlap (moas and kiwis overlap in New Zealand, rheas and tinamous ovrlap in South America south of the Amamzon). This can be explained by dispersal. Panbiogeography explains allopatry by vicariance, overlap by dispersal.

    Link to this
  274. 274. MichaelHeads 5:30 pm 06/29/2014

    Comment 268: You know what? Next time, learn about a topic before you shoot your mouth off about it. You just said “perhaps” to the equivalent of Precambrian bunnies.

    Mesozoic vicariance for hippos on Madagascar would only imply Precambrian bunnies if molecular evolution took place at a constant rate.

    Link to this
  275. 275. MichaelHeads 5:36 pm 06/29/2014

    Comment 254: Why do you say “change”? Palaeognathae Pycraft, 1900, has been the name for “ratites” + tinamous ever since it was coined.

    You don’t need to change a name for a group whenever the delimitation of the group changes, . Ratitae 1813 is a much older, much better-known name than Paleognathae 1900. The fact that one of the members (tinamous) doesn’t have a ratite carina is irrelevant. Just because ‘Planta longifolia’ is discovered to have a subspecies with short leaves, you don’t need to change the name.

    Link to this
  276. 276. irenedelse 5:39 pm 06/29/2014

    This is heading back to a Permian radiation of bears, I tell you…

    Link to this
  277. 277. MichaelHeads 5:48 pm 06/29/2014

    Comment 268: Panbiogeography has a bad name with those who read the fossil record literally, but not with others – e.g. Oxford University Press (Craw et al., 1999. Panbiogeography: Tracking the history of life), University of California Press (Heads 2012, Molecular panbiogeography of the tropics) and Cambridge University Press (Heads 2014, Biogeography of Australasia: A molecular approach).

    …Did you just seriously try to cite publishers instead of publications as evidence???

    MH: No, both are cited.

    DM: And what’s with this strawman of “reading the fossil record literally”? It sounds like you don’t know what you’re talking about.

    MH: So just how much older than its oldest fossil can a clade be? Pick any group and give me an example. It’s tricky. This is now acknowledged as ‘the problem of the priors’ in bayesian dating (the standard current technique).

    MH: The condensation and telescoping of the body (e.g. the transition from a pinnate tail to a palmate one, giving a pygostyle etc. in birds) is one of the most important parallel trends in vertebrates and is seen throughout the group.

    DM: Really? I don’t know about teleosts, but it’s really rare in tetrapods – much less common than elongation, in fact.

    MH: So you still have a tail, do you?

    Link to this
  278. 278. MichaelHeads 5:57 pm 06/29/2014

    Did volcanic island taxa all arrive by long-distance dispersal from the continents?

    Comment 244: Uh… they’re volcanic. They consist of oceanic crust. They didn’t exist before a few million years ago (the maximum by far, some 30 Ma, is the case of Fiji) and were never connected to any continent.

    The age of the individual island is not important. The key information is the age and history of the structure that produced them. In the case of Fiji, this is a subduction zone that separated from Gondwana (Australian part) in the Cretaceous and migrated into the Pacific. The taxa on the islands survive as metapopulations, constantly colonizing new, individually ephemeral islands from prior islands in the region. (See my paper in Syst Biol 2011).

    Link to this
  279. 279. MichaelHeads 6:08 pm 06/29/2014

    I wrote that ‘Molecular work has rejected many long-established groups that were based on traditional morphology’.

    In comment 268, David replied ‘“Traditional morphology” did not, however, use phylogenetic analysis’.

    Phylogenetic analysis, as used in non-molecular work (e.g. by paleontologists), relies on traditional morphology. In a vast number of cases, molecular studies have suggested that this can be very misleading. (I’m not ‘against’ morphology, I’ve used it a lot myself and obviously the morphology has to be accounted for. But the traditional approaches seem flawed).

    Link to this
  280. 280. irenedelse 6:35 pm 06/29/2014

    “MH: So you still have a tail, do you?”

    This may sound like a trifle compared to Mesozoic hippos and crown birds, but by “condensation and telescoping of the body”, do you mean just reduction of the tail?

    “The age of the individual island is not important. The key information is the age and history of the structure that produced them.”

    This is getting tricky. Ephemeral islands? With Mesozoic and early Cenozoic sea levels, I would expect a lot of amphibious lineages in these hypothetical metapopulations.

    Besides, once again, how does panbiogeography square with European relatives of Ostriches from the Paleocene and Eocene?

    And, whether vicariance and/or dispersion are involved, why not take into account Antarctica as a possible stepping stone between Africa, Australia and South America?

    Link to this
  281. 281. Grehan1896 7:09 pm 06/29/2014

    Darren,

    The conversation has been going back and forth well enough for those interested to respond so I am not sure why you are tempted to close the comments after David decides to rant a bit and use what would be considered in some circles to be bad and impolite language.

    In all fairness, I hope you will allow the opportunity for response – which I continue to do in a polite and respectful way.

    John Grehan

    Link to this
  282. 282. naishd 7:21 pm 06/29/2014

    I appreciate that the responses and discussion here have been sensible and cordial — well done, John and everyone else. I admit to being highly frustrated by a complete reluctance and reticence on your part, however, to appreciate how what you’re proposing is, I say again, so contrary to the evidence we have that it can be disregarded as a sensible explanation. So, hippos in the Jurassic?

    Furthermore, I absolutely see David’s point (comment # 268) that you’re saying things that make you seem like an anti-scientist (e.g., you said that you can’t change your mind on the opinion you have). Hmm, so, once you’re a member of the church of panbiogeography, you can never leave?

    Link to this
  283. 283. LeeB 1 7:40 pm 06/29/2014

    The arrival of biota in New Zealand before historic times is attested by numerous examples; the Australian Black swan existed in New Zealand when the first Maori arrived, was exterminated and then reintroduced from Australia.
    As the species was the same in New Zealand and Australia dispersal seems the most parsimonious hypothesis for it’s occurrence.
    Other examples would be the presence of species in the Genera Porphyrio, Gallirallus, Petroica and Rhipidura in both Australia and/or Polynesia and Melanesia and New Zealand.
    The New Zealand species of Rhipidura is sufficiently close to a species in Australia and New Caledonia that they have in the past been considered conspecific.

    Also the New Zealand Eagle Harpagornis moorei was close to the Australian Little Eagle according to the genetic evidence, sufficiently so that it has been considered to have arrived in the last million years.
    And the duck Chenonetta finschi from New Zealand was closely related to the Australian C. jubata and so again it is considered a recent dispersal.

    LeeB.

    Link to this
  284. 284. Heteromeles 7:57 pm 06/29/2014

    @268: bravo, David, you’ve outdone yourself.

    I too am happy to see the respectful tone.

    However: In my book (which is far from the only book) trolls thrive not on misery, but on sucking attention in, and there’s no upper limit to how much attention an ego can absorb when it’s convinced it is right and suffering for righteousness’ sake. Note that being right and being righteous aren’t the same thing.

    IMHO that goes on all sides, incidentally, although I think the tendency is more pronounced on one side than the other. To me, the best solution is the same one one would hypothetically use with a black hole: stop feeding it.

    Link to this
  285. 285. Grehan1896 8:06 pm 06/29/2014

    “I admit to being highly frustrated by a complete reluctance and reticence on your part, however, to appreciate how what you’re proposing is, I say again, so contrary to the evidence we have that it can be disregarded as a sensible explanation. So, hippos in the Jurassic?”

    I did not comment on hippos. I have not analysed that group.

    “Furthermore, I absolutely see David’s point (comment # 268) that you’re saying things that make you seem like an anti-scientist (e.g., you said that you can’t change your mind on the opinion you have). Hmm, so, once you’re a member of the church of panbiogeography, you can never leave?”

    Its the reality I have observed over a number of decades, that in matters such as these, scientists rarely change their mind. It seems that a particular perspective is developed when younger and that tends to prevail when older. So of course anyone can become or unbecome a panbiogeographer (and some have). Never implied otherwise. But in matters of biogeography I know from experience that scientists (amateur or professional) tend to stay with what they have committed to. Nothing wrong with that. Its just life. George Gaylord Simpson opposed continental drift and plate tectonics to the end of his days, contrary evidence notwithstanding. Again, nothing wrong with that. Scientists are people.

    I will comment further about the age of the Mozambique channel later. No time right now.

    John Grehan

    Link to this
  286. 286. Grehan1896 8:42 pm 06/29/2014

    “IMHO that goes on all sides, incidentally, although I think the tendency is more pronounced on one side than the other. To me, the best solution is the same one one would hypothetically use with a black hole: stop feeding it.”

    I agree fully. If one finds that a discussion becomes uncomfortable or unproductive from one’s point of view, then one need not contribute any further.

    John Grehan

    Link to this
  287. 287. Christopher Taylor 9:35 pm 06/29/2014

    286 comments, people. Two. Hundred. And. Eighty. Six. Comments.

    And Irene wins the Internet. All Internets to Irene.

    Link to this
  288. 288. MichaelHeads 1:04 am 06/30/2014

    Christopher Taylor (Comment 151) wrote: ‘…for the most part, the order of appearance of bird lineages in the fossil record does approximate their order of phylogenetic divergence’.

    This is a fascinating, bold prediction and would be of great interest if it could be substantiated. I asked whether anyone had suggested it in print, and Darren kindly pointed me to Mayr’s 2009 book and his own ‘Tetrapod Zoology 2′. Our library doesn’t have TZ2 (Darren, you’ll be pleased to know I’ve ordered it), but I had a look through Mayr’s concise, beautifully-illustrated volume. Unless I’ve missed it, he doesn’t make the bold claim, or even consider it as a possibility. The few groups I’ve checked so far contradict it. I mentioned the ‘basal’ passerids and corvids (Callaeidae and Mohouidae) earlier. In another example, the ‘basal’ passerine clade, the New Zealand Acanthisittidae, are known as fossils back to the Miocene, but there are Eocene (ambiguous) and Oligocene passerines, including both oscines and suboscines. These ‘closely resemble their modern counterparts’. Presumably the Acanthisittidae did exist before the Miocene, but are not known from then because of gaps in the fossil record.

    Darren, do you actually suggest something like ‘the order of appearance of bird lineages in the fossil record does approximate their order of phylogenetic divergence’ in your book? If so, could you supply the quote? Has anyone ever said or implied this anywhere else in print?

    Link to this
  289. 289. MichaelHeads 2:02 am 06/30/2014

    Comment 280: ‘This may sound like a trifle compared to Mesozoic hippos and crown birds, but by “condensation and telescoping of the body”, do you mean just reduction of the tail?’

    Not at all, that’s just one, fairly minor example, but one that’s close to home. The point was, with huge amounts of morphological parallelism occurring right across vertebrates, it becomes very difficult to place lineages (especially old, ‘bizarre’ ones) on a phylogeny.

    “The age of the individual island is not important. The key information is the age and history of the structure that produced them.”

    Comment 280: This is getting tricky. Ephemeral islands? With Mesozoic and early Cenozoic sea levels, I would expect a lot of amphibious lineages in these hypothetical metapopulations.

    There are frogs (is that what you meant?) in Fiji and New Zealand, and the debate has always been how did they disperse there. With migrating subduction zones (slab rollback) and their associated island arcs, the frogs could have dispersed with the arc.

    Comment 280: Besides, once again, how does panbiogeography square with European relatives of Ostriches from the Paleocene and Eocene?

    The really fascinating thing about the ratites for biogeographers is the allopatry. All nine families (including lithornithids of North America, London Clay, and Denmark, but I’m not considering the other, more ambiguous fossil groups) are almost perfectly allopatric, with only two areas of substantial overlap (moas and kiwis in New Zealand, and rheas and tinamous in southern South America). It’s a very dramatic example of global allopatry at a high (family) level. The comparatively minor areas of overlap can be explained by secondary dispersal blurring the edges, while a simple explanation for the allopatry is vicariance.

    Chance dispersal can explain anything, but it wouldn’t give a satisfying account for the beautiful overall allopatry or the standard distribution of the individual clades, e.g. the Madagascar – New Zealand distribution of elephant birds + kiwis, which is shared with other vertebrates, as well as invertebrates and plants. The Africa – Asia (fossil) distribution of ostriches (not in Madagascar!) is also a very common pattern. Also see cassowaries in New Guinea and NE Queensland, and emus of the rest of Australia (some very local overlap in NE Queensland). Again, this is a very common pattern (e.g. in marsupials, Fig. 4.10 in my 2014 book).

    The allopatry of the ratite families could, in theory, be explained by a ‘founder takes all’ model, but this would not account for the overlap in New Zealand and South America.

    The Paleotididae of Germany are controversial – they’ve been treated as sister to ostriches or to rheas, but ‘evidence for neither hypothesis is convincing’ (Mayr,2005). In any case, they do not overlap with Lithornithidae. They may well be ostrich relatives, and the Africa – Europe – Asia (not Madagascar or India/Sri Lanka!) distribution this would give is one of the most common distributions on Earth. I’m not sure why you think this might be a problem for a vicariance model. (Note that vicariance does not mean ’caused by Gondwana breakup’ – this is only one example of the process).

    Link to this
  290. 290. irenedelse 2:24 am 06/30/2014

    I can see why Darren wanted to close the comments. If we are going to re-hash things already discussed ad nauseam with John Grehan, what’s the use in continuing with the back and forth, Mr. Heads?

    I can’t comment at length right now as I have to go to work, but I must say that I’m not exactly favorably impressed by the way the two panbiogeography proponents have offered a lot of scepticism of other people’s theories but little in the way of evidence to support their own. It’s not enough to point at the things currently accepted theories *don’t* explain. If you think your alternative hypothesis has a *better explanatory power*, you should be able to substantiate it with positive data.

    Link to this
  291. 291. naishd 4:13 am 06/30/2014

    Michael (comment # 288): it will take me a while before I have time to compose a reply, though note that a detailed response has already been given (comment # 196, by Mike Keesey). Thank you for ordering my Tetrapod Zoology book, but I think I was referring to my bird chapter in The Complete Dinosaur 2, only because this provides an overview of the avian fossil record and the timing of origin of the various groups.

    Link to this
  292. 292. Dartian 4:20 am 06/30/2014

    I’ve stayed out of this discussion thus far. There is, as others have already said, very little point in discussing with an individual (or individuals) who openly states that he won’t change his opinion on the subject under discussion – and I, too, agree that that is pretty the most un-scientific (if not downright anti-scientific) attitude anyone could have.

    However, I can’t let this factually incorrect statement pass unrefuted:

    John Grehan:
    George Gaylord Simpson opposed continental drift and plate tectonics to the end of his days, contrary evidence notwithstanding.

    Simpson was an outspoken opponent of the continental drift and plate tectonics theories as late as in the 1960′ies, but after that he accepted the evidence. Grudgingly, perhaps, but he accepted it. In fact, he explicitly wrote in his autobiography about this issue that:
    [S]ince reaching my seventies I have myself changed my mind” (on this subject).

    Emphasis added for the benefit of Mr. Grehan. Thus, kindly refrain from trying to recruit Simpson as your intellectual ally in the future.

    As for David’s comment #268; bravo! Darren, please bestow a Tet Zoo knighthood upon him already. ;)

    Link to this
  293. 293. MichaelHeads 6:10 am 06/30/2014

    Darren has just tweeted: Darren Naish @TetZoo · Jun 28
    #Panbiogeography debate at #TetZoo: Heads says PBG “well respected”, taken seriously in Brazil & Mexico. Can anyone confirm if this is true?

    There are now many panbiogeographic studies coming out of Latin America. To cite just one nice example: Echeverry, A., & Morrone, J. J. (2013). Generalized tracks, area cladograms and tectonics in the Caribbean. Journal of Biogeography, 40(9), 1619-1637.
    Readers may also be interested in two large volumes:

    Lorente Bousquets, J., and J.J. Morrone (eds.). 2001. Introducción a la biogeografía en Latinoamérica: Teorías, conceptos, methodos y aplicaciones’, ed. Universidad Nacional Autónoma de México: Mexico City. [In Spanish]. The book is dedicated: ‘A la memoria de Léon Croizat, precursor de la biogeografía en latinoamérica, por atreverse a mirar más allá de su espacio y tiempo’.

    Juan Morrone is the most cited biogeographer in Latin America.

    de Carvalho, C.J.B., and E.A.B. Almeida (eds.) 2011. Biogeografia da América do Sul: Padrões e Processos. Roca: São Paulo. [In Portuguese]. The first paragraph in this volume, in the ‘Apresentação’ by the editors, begins:
    ‘A little more than 30 years ago, Leon Croizat began his ‘Analytic and synthetic biogeography (Panbiogeography’ of the Americas’ [a 1975 book] with the following declaration ‘It is not all bold to state that there is to-day no biogeography known in Latin America worth the title of scientific’. He continued, arguing that we do not have a philosophy, a general idea or a nexus of basic notions that would permit us to confront problems of geographic distribution…’. The editors went on to discuss this claim positively.
    Claudio (Curitiba) and Eduardo (São Paulo) are currently preparing a second edition of this book. Both editions have long chapters on panbiogeography.

    For a bibliography on the subject, see the website by Mauro Cavalcanti in Rio (note that the list of species and genera named after Croizat is very incomplete): http://panbiog.eu5.org/?Home

    Link to this
  294. 294. MichaelHeads 6:33 am 06/30/2014

    Irene (comment 290) wrote: ‘I can see why Darren wanted to close the comments. If we are going to re-hash things already discussed ad nauseam with John Grehan, what’s the use in continuing with the back and forth, Mr. Heads?’

    Well as they say in Jamaica, iron sharpeneth iron…

    ‘If you think your alternative hypothesis has a *better explanatory power*, you should be able to substantiate it with positive data’.

    I cited quite a bit of data in my last post. Here’s another example of data that is consistent with differentiation in ratites being caused by vicariance. The kiwi, Apteryx, comprises two main clades, brown and spotted. Sequencing studies of extant and sub-fossil populations show that the brown kiwi clade is divided into two allopatric groups of species. Their mutual boundary lies along the Alpine fault (a plate boundary), not the main geographic divide, which runs parallel with it but 10 km to the east.

    A molecular phylogeny of the moa, Megalapteryx, shows sister clades that are disjunct along the Alpine fault, between Fiordland and Nelson. The gap, hundreds of kilometres, is filled by other clades. The gap is consistent with strike slip (horizontal) displacement along the fault.

    Both patterns, in the kiwi and the moa, could be explained by chance dispersal. But it is always good to have more than one hypothesis for a given phenomenon – this is the principle of ‘multiple working hypotheses’ advocated by geologists (see Chamberlin in Science 1965). One alternative to chance would be vicariance caused by tectonic activity at the plate margin, prior to the uplift of the main mountain range.

    Link to this
  295. 295. Christopher Taylor 6:39 am 06/30/2014

    as they say in Jamaica, iron sharpeneth iron…

    Only if applied correctly. Otherwise everything just ends up blunter than before.

    Link to this
  296. 296. Grehan1896 11:29 am 06/30/2014

    Dartian said:

    “I’ve stayed out of this discussion thus far. There is, as others have already said, very little point in discussing with an individual (or individuals) who openly states that he won’t change his opinion on the subject under discussion – and I, too, agree that that is pretty the most un-scientific (if not downright anti-scientific) attitude anyone could have.”

    It is my expectation that I would not change my mind. However, yes it is theoretically possible and I did not meant to suggest otherwise.

    “However, I can’t let this factually incorrect statement pass unrefuted:

    “[S]ince reaching my seventies I have myself changed my mind” (on this subject).
    Emphasis added for the benefit of Mr. Grehan. Thus, kindly refrain from trying to recruit Simpson as your intellectual ally in the future.”

    Thank you for the correction. Whatever I read a long time in the past was either mistaken or I misread. However, your request does not follow.

    John Grehan

    Link to this
  297. 297. Grehan1896 12:09 pm 06/30/2014

    I’ve made an effort here to respond to points from David #268 while not being distracted by some of the emotive content (and with that removed the David’s comments do stand out as being worthwhile).

    …OK, I have illustrated the consequences of trying to deal with a thread on two separate days: I’ve repeated part of the discussion.
    From comment 127:
    “Naturally, this paper answers some of my questions! However, I find it curious that the paper doesn’t try to evaluate dispersal hypotheses on their own; sea currents are nowhere mentioned in the text or the maps, and it is tacitly assumed that all of the discussed clades completely lack a fossil record outside their current distribution – which is screamingly wrong for iguanas, to pick the example I know.”
    There was not need to discuss the dispersal hypotheses as others had done that and my focus was on biogeographic patterns.
    There was no tacit assumption about the fossil record outside the current distribution. Do you know anything about the fossil record of the groups in question that would change their biogeography in any particular way?

    “The paper tries to show that vicariance explanations aren’t completely impossible (according to geology and exclusively extant biogeography). What it should have tried to show is that vicariance hypotheses are more parsimonious than dispersal hypotheses. And frankly, the reviewers ought to have pointed this out.”

    No paper is perfect.

    From later comments:
    If “all other data points” (presumably you mean other than biogeographic data points) indicate a mid-Jurassic origin for birds as a while then there would appear to be nothing extraordinary about the basal lineages of extant birds also diverging about that time.

    “Birds as a whole” includes Archaeopteryx and goes way, way, way beyond the crown-group. Cretaceous birds are plentiful in the fossil record as we have come to know it in the last few years, and most or all of them go on various positions in the stem, not the crown. The stratigraphic fit of bird phylogeny is pretty good (see comment 196); as long as evidence to the contrary remains absent, we have to expectthe crown-group (Neornithes) to be much younger than birds as a whole.”

    I do understand that you and others on this list have that view.

    “The reality is that molecular divergence dates cannot falsify an earlier origin.
    It’s not like falsification is all science can do. There’s also parsimony. (In fact, there’s a lot of parsimony hidden in falsification.)”
    Ok, but fact remains that molecular divergence dates cannot falsify an earlier origin.

    “The order of divergence does not have to match the order of continental drift.
    What criterion do you have left, then, to test your own biogeographic hypotheses???”

    Tectonic correlation (as covered in several postings)

    “We are not going to change our respective views
    What the fuck.
    How can a scientist say such a thing?!?”
    No need for abusive language. This is a prediction. Perhaps I should have said it is very unlikely.

    “I never know in advance when I’m going to change my views. I never know in advance what new evidence might come to my attention or what new hypotheses are going to be tested. And neither do you, or anybody else.”
    True, but I would be very surprised if you changed your mind with the current evidence. If something new crops up in the future that might change everything or not. One might think that scientists would respond positively to panbiogeography successfully predicting geological structures then unknown to geologists. But oh no.

    “So fossil calibrated molecular divergence estimates are not used for nematodes?
    Can’t be, because there’s nothing to calibrate them with. ReadRaaum et al. (2005) on how to choose calibration dates.”

    There is no other group related to nematodes at some level that can be used?

    “It all depends on what one means by “better explains”. Unfortunately there is no universal criterion.
    You fail at basic science theory. Start anew, do not pass Go, do not collect 200 dolaritos.”

    Maybe, maybe not.

    1) “Absolute proof” is a disgusting strawman. Proof is for math, logic and American alcohol!

    Perhaps, but according to some leading luminaries in the field the relationship between humans and chimpanzees is a fact. That is what I meant by absolute proof (but I like the allusion to EOH).

    “2) I do take the fossil record into account. But given the biogeographic evidence I am inclined to interpret the fossil record differently to most. I’m sorry, this claim is no different to when Ken Ham or Kent Hovind insist that scientists and creationists look at the same evidence and just interpret it differently. “

    Well, there are different interpretations about the fossil record, even among evolutionists.

    “You also haven’t explained why Neornithes should have a so much worse preservation potential than Enantiornithes throughout the Cretaceous.”

    Agreed, I don’t have an ‘explanation”. Interestingly, when discussing a parrot fossil from the Cretaceous the author did speculate on the lack of fossils for this and other extent groups.

    “The differentiation of anthropoids and prosimians at the Lebombo monocline.
    What? Both strepsi- and haplo(r)rhines are from Eurasia, showing up there in the Eocene.”
    Yes, but the Lebombo monocline is not in Africa.
    “Prosimii” is paraphyletic: the tarsiers are closer to us than to the lemurs & lorises.

    Only in some molecular reconstructions. Tarsiers are largely allpatric to both anyway.

    “Specifically, North America on the one side and East Antarctica + Australia on the other broke apart some 750 Ma ago. This has been uncontroversial for easily 15 years now.”

    Was not referring to that.

    “Its a nice story.
    It’s a parsimonious story. Have you got a more parsimonious one?”
    Yes, abut according to biogeographic criteria over which we disagree.

    “some molecular divergence theorists proposed 135 Ma for ratites, if that’s ok, what’s another 30 odd million years?
    I don’t find 135 Ma for a fictitional subset of paleognaths OK at all, no. But even apart from that, the longer a ghost lineage is, the less parsimonious is the hypothesis that this ghost lineage is real; that should be obvious! It’s quantitative, not black-or-white as you appear to believe.”
    It’s also well-respected in megadiverse countries such as Mexico and Brazil.
    “When people’s support for a hypothesis can be predicted from where they work, something is massively, massively wrong either with that hypothesis or with its alternative(s): one of them is – I can’t see a way around this conclusion – propagated by authority and/or by lack of access to publications, not because it explains the data better than any alternative. National science is bad science.”

    But it is transnational. In this case there is more resistance to panbiogeography from American and UK than some other major regions (encompassing more than one country and culture.

    “Molecular work has rejected many long-established groups that were based on traditional morphology.
    “Traditional morphology” did not, however, use phylogenetic analysis.”

    Not correct. The analysis I published on humans and orangutans was based on morphology and phylogenetic analysis.
    “ a troll is someone who tries to make people upset and then laughs when that worked.”

    Evidence of this here?

    A cluster of volcanic islands does not have to be in direct contact with a continent for a geological connection to have existed for the biota.
    So you suddenly accept dispersal from a continent to volcanic islands?

    It can and does happen. No sudden change. Question remains, how does one know whether this applies to the origin of various biotas on volcanic islands.

    “But there are a number of geological models requiring major basaltic plateau in the central Pacific.
    A basaltic plateau isn’t a continent; and no continent has been anywhere near the central Pacific in a long time.”

    Maybe true, but I was not invoking continents.

    “Island arc moving across the Galapagos.
    The Galápagos islands aren’t part of an island arc… there’s no subduction going on anywhere nearer to them than the rim of South America, and that subduction zone lacks an island arc.”
    Of course not. I did not say that they were. I said that an island arc passed over the hotspot (and now parts are imbedded with in the Americas.
    “Keyword “are”. There used to be iguanas all over Eurasia all the way to the Oligocene.”

    OK, but there is currently no evidence that the Fiji iguanas have their nearest relatives in Asia (that I know of).

    “According to the molecular phylogeny of Noonan and Sites (2010) the iguanas of Fiji and Tonga are sistergroup to Dipsosaurus of Baja California and SW USA and Mexico.
    The closest living relative isn’t necessarily the sister-group.”

    It’s the known sister group. There may be other closer sistergroups that have become extinct.

    Oh, speaking of Tonga (which has not just iguanas, but even frogs): was there ever a way to walk from Fiji to Tonga?

    Is this a rhetorical question?

    “. As far as I know, the genomes of the species on Galápagos differ from their closest relatives elsewhere by amounts that are compatible with recent divergences, about as recent as the age of the archipelago. You, however, postulate that these divergences happened in the Cretaceous instead. That means you propose extremely slow rates of evolution. As I just said, proposing mind-blowingly strong stabilizing selection doesn’t work, because much of the genome is invisible to selection in the first place.You are, therefore, proposing mind-blowingly high fidelity of DNA replication & repair! You just haven’t noticed.”

    Perhaps this means there is a problem with these molecular theories (I know you won’t agree).

    “Madagascar was colonized by placental mammals five times in 66 million years (not counting bats or humans), and it’s full of placentals now.” You assert this as fact, but it is not.
    I assert it as a well-supported hypothesis. Haven’t you read the papers about it? Not even Yoder et al. (2007, Systematic Biology)?
    I’ve read Yoder. Nothing there to make it a well supported hypothesis.
    “Fact: these clades are not each other’s closest relatives; instead, each of them has a sister-group in Africa, so 5 is the minimum number of times placentals must have reached Madagascar (not counting bats flying all over the place and humans cheating with ships). You’ll probably agree Madagascar is not where the first placentals lived.”

    To be honest, I have no data to say whether the ancestors of such groups were first in what is now ‘Madagascar’ or in ‘Africa’

    Fact: there are a few mammals in the Cretaceous (and Jurassic) fossil record of Madagascar, but (so far) not one of them is a placental.

    “anything that has reached it in the Cenozoic must have dispersed over water.”

    That would be the case given that geological isolation.

    “ Because Madagascar separated from Africa first, you need to postulate an elaborate geographic pattern of extinction, in addition to the usual ghost lineages, if you want to claim that all five of these clades walked to Madagascar when that was still possible.”

    Or these groups were not present in those areas that became those other continents.

    “Dude, do we really need to explain to you that the Galileo Gambit is a logical fallacy? Yes, people laughed at Galileo, but (as has famously been pointed out) they also laughed at Bozo the clown; the trick is that Galileo’s hypothesis, and Clintock’s, explained the known data more parsimoniously than the alternatives.”

    Point was that an explanation may be better, but it does not meant that it will be accepted.

    “You need to demonstrate that panbiogeography explains the known data more parsimoniously than occasional dispersal. You haven’t even tried to do that.”

    Heads has responded so I will not repeat. You may disagree with the demonstration, but that is not to say that the demonstration has not been made.

    “Don’t you think that both go across Antarctica instead? Remember, the Earth isn’t flat…”

    No, it’s a tectonic correlation (but Antarctica could be included if fossil relatives in each group were found there).

    “- Were Madagascar’s hippos the result of Mesozoic vicariance? Perhaps. I have not analyzed that group. Wow.”

    So?

    You know what? Next time, learn about a topic before you shoot your mouth off about it.

    The topic being hippos? If so, then I did not, and I pointed out that I have not studied that group.

    Yes. Although interestingly some molecular theorists think otherwise for ratites at least.
    They don’t so much think otherwise as just not understand the matter much better than you do. If you want to read really cringeworthy things in the peer-reviewed scientific literature, I recommend molecular papers from the late 90s and early 00s.

    OK, so what you are saying is that your point of view on molecular age is superior to those molecular theorists who have proposed origins with which you disagree.

    John Grehan
    Link to this

    Link to this
  298. 298. ectodysplasin 1:29 pm 06/30/2014

    I can think of a number of taxa with good to decent fossil records for which panbiogeography would offer a positively misleading result. A perfect example would be lungfishes, which a panbiogeographer would assign a Gondwandan distribution and an origin in the Jurassic or early Cretaceous. Another example would be cryptobranchid salamanders, which would be assigned a Pacific distribution, despite extensive fossil evidence for a holarctic distribution until quite recently. And so on and so forth. Panbiogeographers would also miss the presence of fossil marsupials in Eurasia, among other regions, and would infer some bizarre tectonic scenarios. When you’re looking at taxa with very, very few living representatives, you have a much greater chance of getting trivial artifacts, and drawing vast interpretations on the basis of those artifacts is really, really questionable.

    There are a bunch of methodological issues that I’ve brought up above, and additional methodological issues worth discussing (e.g. “what does a minimum spanning tree actually report?” and “how do different clustering algorithms affect the topology of a MST?” among other things). I’m not against the idea that PBG might have a use in certain circumstances, but this idea that PBG evidence should trump phylogenetic data and the fossil record so extensively seems ludicrous. Tools should be measured by their usefulness. PBG seems to be useful only in very specific circumstances.

    Link to this
  299. 299. Grehan1896 2:28 pm 06/30/2014

    “298. Ectodysplasin
    “I can think of a number of taxa with good to decent fossil records for which panbiogeography would offer a positively misleading result.

    “A perfect example would be lungfishes, which a panbiogeographer would assign a Gondwandan distribution and an origin in the Jurassic or early Cretaceous.”
    I have not analyzed lungfishes. Need to explain please.

    “Another example would be cryptobranchid salamanders, which would be assigned a Pacific distribution, despite extensive fossil evidence for a holarctic distribution until quite recently.

    I have not studied this group. A quick look shows China, Japan, and eastern N. America. Those groups alone could suggest a Pacific basin, but not necessarily as the N American species are border the Atlantic. You say “despite extensive fossil evidence for a holarctic distribution until quite recently”. This seems to say that quite recently there is now no extensive fossil evidence for a holarctic distribution. If there are fossils through N America and Eurasia it would be interesting to see if affinities are closer between those either side of the Pacific or Atlantic.

    “And so on and so forth. Panbiogeographers would also miss the presence of fossil marsupials in Eurasia, among other regions, and would infer some bizarre tectonic scenarios.”

    I cannot figure this out. As already stated, fossils are included in analyses and have been extensively in the panbiogeographic literature.

    “ When you’re looking at taxa with very, very few living representatives, you have a much greater chance of getting trivial artifacts, and drawing vast interpretations on the basis of those artifacts is really, really questionable.”

    But would apply to any analysis, biogeographic or otherwise. In science one works with information that one has. To date I am not aware of any new fossil evidence arising that has fundamentally changed the panbiogeographic picture elucidated from living representatives (this is over the last few decades and I may have forgotten something).

    “There are a bunch of methodological issues that I’ve brought up above, and additional methodological issues worth discussing (e.g. “what does a minimum spanning tree actually report?”

    It is a spatial parsimony criterion – that in the absence of other evidence (e.g. phylogeny, diversity concentration) the shortest distance between different localities is treated as generating the least number of adhoc spatial explanations required for that distribution. Now one may disagree with this, but I am only providing the explanation here.

    “and “how do different clustering algorithms affect the topology of a MST?” among other things). “

    Please give an example

    I’m not against the idea that PBG might have a use in certain circumstances, but this idea that PBG evidence should trump phylogenetic data and the fossil record so extensively seems ludicrous.

    As said before, it does not automatically trump anything. It may or may not corroborate other evidence or interpretations. One may make a choice about what to support.

    “Tools should be measured by their usefulness. PBG seems to be useful only in very specific circumstances.”
    It may be a good idea to read Mike’s books to make that kind of judgment (and Craw et al as an introductory text).

    John Grehan

    Link to this
  300. 300. Halbred 2:42 pm 06/30/2014

    At the risk of running this topic further into the ground, John, your most recent comment intrigues me:

    “One may make a choice about what to support.”

    Why do you support panbiogeography at all? What is the advantage? As has been demonstrated, it requires a good deal of special pleading, it difficult to square with other lines of evidence, and isn’t putting you on any Christmas card lists.

    What is your personal motivation for advocating PBG? What are you getting out of this?

    Link to this
  301. 301. Grehan1896 3:25 pm 06/30/2014

    “At the risk of running this topic further into the ground, John, your most recent comment intrigues me:
    “One may make a choice about what to support.”
    Why do you support panbiogeography at all? What is the advantage?”

    I find that it provides a coherent and integrated understanding of biogeographic patterns and history.

    “As has been demonstrated, it requires a good deal of special pleading, it difficult to square with other lines of evidence, and isn’t putting you on any Christmas card lists.”

    The reference to “special pleading” has some subjective connotations. One might say that those who accept the fossil record at face value are making a good deal of special pleading as well – and just because it may be a majority view does not change that.”

    “What is your personal motivation for advocating PBG? What are you getting out of this?”

    Probably nothing much different from any other scientist that advocates a position or a view. I’ve always had an interest in sharing views and communicating with those who have different views. If I were to take a selfish goal it would be that I benefit from such exchanges.

    Halbred – do you advocate anything in science? If so, why?

    John Grehan

    Link to this
  302. 302. DavidMarjanovic 3:25 pm 06/30/2014

    First of all, apologies to John Grehan: I had somehow managed to overlook the “necessarily” in “No one contributing to this discussion is necessarily going to change their mind about anything” (comment 265), so the version in comment 144 that lacks “necessarily” was all I had to go with. That’s where about half of my anger came from.

    Second, people who know me from a blog with more traffic (where threads this length are common) won’t be surprised at the length of comment 268. A lot was wrong or questionable or otherwise required a comment, so I replied to it all – that comes to me naturally; it’s not some kind of special effort. I have SIWOTI syndrome.

    Mesozoic vicariance for hippos on Madagascar would only imply Precambrian bunnies if molecular evolution took place at a constant rate.

    No; all this requires is that the rates of molecular evolution can’t vary by several orders of magnitude. You’re basically proposing no limit at all.

    Apart from this, you would really need to explain why abundant, large, semiaquatic, osteosclerotic* animals like crown-group hippos, which have, as expected, a very good fossil record in the Pleistocene (preceded by a similarly good fossil record of stem-hippos in the Eocene through Miocene), have no fossil record at all before that time. Keep into account that the less abundant, equally semiaquatic crocodiles have a quite good fossil record all the way through to the Early Jurassic.

    * Adult hippos can’t even swim. They walk on the bottom.

    You don’t need to change a name for a group whenever the delimitation of the group changes, . Ratitae 1813 is a much older, much better-known name than Paleognathae 1900. The fact that one of the members (tinamous) doesn’t have a ratite carina is irrelevant.

    Both names have been widely used, with different meanings, by ornithologists ever since 1900. They don’t designate the same thing. Ratitae designates a group that has turned out to be para- or polyphyletic depending on how you define it, while Palaeognathae designates a clade as it did in 1900; there’s no reason to suddenly ditch Palaeognathae now.

    Comment 268: Panbiogeography has a bad name with those who read the fossil record literally, but not with others – e.g. Oxford University Press (Craw et al., 1999. Panbiogeography: Tracking the history of life), University of California Press (Heads 2012, Molecular panbiogeography of the tropics) and Cambridge University Press (Heads 2014, Biogeography of Australasia: A molecular approach).

    …Did you just seriously try to cite publishers instead of publications as evidence???

    MH: No, both are cited.

    What you wrote says that Oxford U Press, U of Cal Press and Cambridge U Press are the “others” who don’t “read the fossil record literally”, illustrated with the examples in the parentheses. If that’s not what you meant, express yourself more clearly next time.

    So just how much older than its oldest fossil can a clade be? Pick any group and give me an example. It’s tricky. This is now acknowledged as ‘the problem of the priors’ in bayesian dating (the standard current technique).

    I’ve already said it’s not a black-and-white thing. How about you follow my suggestion to read Raaum et al. (2005; linked to above) and my papers (cited above)? Two of mine, as I said, are open-access, and if you want the other two (including the open-access one that is unavailable right now because the website has been under alleged construction for half a year) or for that matter that by Raaum et al. (2005), find me in Google Scholar and drop me an e-mail.

    It’s also a bit odd to simply call Bayesian inference “standard”. Maximum likelihood isn’t exactly dead.

    MH: The condensation and telescoping of the body (e.g. the transition from a pinnate tail to a palmate one, giving a pygostyle etc. in birds) is one of the most important parallel trends in vertebrates and is seen throughout the group.

    DM: Really? I don’t know about teleosts, but it’s really rare in tetrapods – much less common than elongation, in fact.

    MH: So you still have a tail, do you?

    My tail is condensed & telescoped, but the rest of me isn’t; I have about three fewer presacral vertebrae than my Pennsylvanian ancestors did. Frogs have less than half as many presacrals as their Pennsylvanian ancestors, possibly only a third – and they have about a third of the number I have.

    You wrote “condensation and telescoping of the body”, not just of the tail.

    Tail reduction is common among mammals. Among archosaurs or indeed limbed diapsids, however, the short-tailed birds (Avebrevicauda) are the only case of serious tail reduction. That’s because the tail is the site of the biggest, most important leg retractor, the M. caudifemoralis longus; tail reduction was only possible in maniraptorans after large-scale reorganization of locomotion and posture.

    The age of the individual island is not important. The key information is the age and history of the structure that produced them. In the case of Fiji, this is a subduction zone that separated from Gondwana (Australian part) in the Cretaceous and migrated into the Pacific. The taxa on the islands survive as metapopulations, constantly colonizing new, individually ephemeral islands from prior islands in the region. (See my paper in Syst Biol 2011).

    Where’s the evidence for specifically Cretaceous fauna on them, then? And what makes you so certain that the subduction zone always had, throughout its entire existence, emergent islands in dispersal distance of each other, and that this began early enough that it was possible to settle the first such islands from Australia?

    Google Scholar can’t find your paper; please tell me which volume and issue it’s in.

    Phylogenetic analysis, as used in non-molecular work (e.g. by paleontologists), relies on traditional morphology. In a vast number of cases, molecular studies have suggested that this can be very misleading.

    Frankly, this sounds like you’re just remixing your treknobabble without understanding it.

    I repeat: traditionally, phylogenetics using morphological data was an art that built evolutionary scenarios from subjectively “reliable”/”important” characters; the science of phylogenetics, whether using molecular or morphological data (or both), finds different results; there remain many large groups for which no phylogenetic analysis that used a sufficiently large amount of morphological data has ever been done, because it’s just too much work, so analyses of molecular data are all we have at the moment.

    Phylogenetic analysis of morphological data does not rely on evolutionary scenarios (unless you invest a lot of work in twisting your data matrix’s arm behind its back – a case is taken apart in my other 2008 paper, open access again). But I’m not sure if that’s even what you were trying to say!

    after David decides to rant a bit and use what would be considered in some circles to be bad and impolite language

    I’m a scientist. I have trained long and hard to call a spade a spade.

    In case you’re wondering, I have no respect for anybody who commits logical fallacies and pontificates about topics he doesn’t know well enough to form an opinion. I am, in fact, angry; I’m not pretending.

    I fail to see how that’s relevant to this discussion, though. :-|

    In all fairness, I hope you will allow the opportunity for response – which I continue to do in a polite and respectful way.

    Don’t feel constrained. Whether there were crown-group paleognaths or crown-group hippos in the Middle Jurassic doesn’t depend on how much we love each other.

    To me, the best solution is the same one one would hypothetically use with a black hole: stop feeding it.

    Trolls, like bullies, don’t just fade away in a puff of Hawking radiation when you don’t give them your attention. They just come over and take it. I’m speaking from rich experience.

    That said, I don’t think anybody in this thread is a troll. A troll bullshits in order to manipulate people’s emotions as part of a power trip (“I’m making you cry, and you know it, and you can’t do anything about it”); a study has found they’re usually sociopaths (people without empathy).

    George Gaylord Simpson opposed continental drift and plate tectonics to the end of his days, contrary evidence notwithstanding. Again, nothing wrong with that.

    Absolutely, there’s something very wrong with that: it’s intellectual dishonesty. It’s nothing less than a preference for ideological thinking over science!

    For Simpson, not exactly a stranger to the scientific method, to keep that up all the way to 1984 would have required immense cognitive dissonance, even denial of measured movements! I’m not at all surprised to learn (from comment 292) that he didn’t in fact do that, but changed his mind around the time the geologists did.

    The point was, with huge amounts of morphological parallelism occurring right across vertebrates, it becomes very difficult to place lineages (especially old, ‘bizarre’ ones) on a phylogeny.

    That point is wrong, though. Cladograms with a consistency index of 0.3 or 0.2 can still be robust.

    As you add data to a matrix for phylogenetic analysis, the signal adds up, and the noise cancels itself out.

    There are frogs (is that what you meant?) in Fiji and New Zealand, and the debate has always been how did they disperse there. With migrating subduction zones (slab rollback) and their associated island arcs, the frogs could have dispersed with the arc.

    That’s not working out: the frogs in NZ and Fiji + Tonga are about as distantly related as extant frogs can be! The NZ frogs (Leiopelma) have their sister-group (Ascaphus) in the Pacific Northwest of North America, and together these two clades form the extant sister-group to all other extant frogs put together. The Fiji frogs, 3 species of Platymantis, are ceratobatrachid ranoids, so their closest relatives are “found in the Malay Peninsula, Borneo, the Philippines, Palau, [...] New Guinea, and the Admiralty, Bismarck, and Solomon Islands” (Wikipedia: Ceratobatrachidae – I’m not linking for fear of triggering moderation by too many links); this distribution (Palau!!!) suggests good transoceanic dispersal abilities. I can’t find anything about Tonga right now, but I’m pretty sure that’s Platymantis again.

    The Africa – Asia (fossil) distribution of ostriches (not in Madagascar!) is also a very common pattern.

    Yeah, because it’s been possible to walk back and forth between Asia and Africa for about 23 million years now.

    Also see cassowaries in New Guinea and NE Queensland, and emus of the rest of Australia (some very local overlap in NE Queensland). Again, this is a very common pattern (e.g. in marsupials, Fig. 4.10 in my 2014 book).

    o_O That’s ecological: cassowaries live in tropical rainforests, emus in drier habitats. As it happens, tropical rainforest on the Australian continent is limited to New Guinea and a few patches in NE Queensland…

    The Paleotididae of Germany are

    Palaeotidids are like Highlanders: there can be only one. :-) Palaeotis weigelti is the only putative paleognath from Germany; Remiornis heberti is from France, and that’s it for Europe before the Pliocene ostriches of Ukraine, now that Eleutherornis is considered some kind of phorusrhacid.

    list of species and genera named after Croizat

    …eh… naming something after him doesn’t imply agreement with every last one of his ideas. Croizat was a very important figure in the history of biogeography; that’s reason enough to immortalize him. People keep naming species and genera after Darwin long after his Lamarckian theory of heredity has turned out to be completely wrong.

    I’ve made an effort here to respond to points from David #268 while not being distracted by some of the emotive content

    You’re not getting an extra cookie for behaving like a scientist :-)

    There was no tacit assumption about the fossil record outside the current distribution. Do you know anything about the fossil record of the groups in question that would change their biogeography in any particular way?

    I mentioned the iguanas. On the one hand, there’s this:

    “Many complete skeletons of Geiseltaliellus longicaudus [sic – should be longicauda, a noun in apposition] KUHN are described from the Middle [sic!] Eocene of the fossillagerstätten Geiseltal near Halle an der Saale and Grube Messel near Darmstadt, Germany. The investigation supports the close similarity to the Corytophanidae with its subfamily Basiliscinae. But for practical reasons it was necessary to erect the subfamily Messelosaurinae nov. based upon the following characters: position of the Foramen parietale in the frontoparietal suture; lacking of an occipitally enlarged parietal crest; coronoid with a large processus labialis. The Messelosaurinae includes: Geiseltaliellus longicaudus KUHN, 1944 from the Lower and Middle Eocene of France, Belgium and Germany, Cadurciguana hoffstetteri AUGE,1987 from the Upper Eocene of France, Aciprion formosum COPE, 1873 from the Lower Middle Oligocene of the USA, Cypressaurus hypsodontus HOLMAN, 1972 and Holmanisaurus oligocenicus (HOLMAN, 1972) both from the Lower Oligocene of Canada. Capitolacerta dubia KUHN, 1944 and Geiseltaliellus louisi (AUGE, 1990a) are junior synonyms of G. longicaudus.”

    The stem-corytophanid Geiseltaliellus has also been found in Sossís (Spain, late Eocene).

    On the other hand, there’s the stem-polychrotid Afairiguana from the Green River Formation (early Eocene, Wyoming).

    Both of these require a long ghost lineage at the base of Iguanidae sensu stricto, to which the iguanas of Fiji and Tonga belong.

    The oldest known crown-iguanoids come from the middle of the Late Cretaceous of Mongolia. This is compatible with Oligocene crown-iguanids, and it is further evidence that Iguanoidea has a Laurasian, not a Gondwanan origin.

    “The paper tries to show that vicariance explanations aren’t completely impossible (according to geology and exclusively extant biogeography). What it should have tried to show is that vicariance hypotheses are more parsimonious than dispersal hypotheses. And frankly, the reviewers ought to have pointed this out.”

    No paper is perfect.

    *eyeroll*
    Who’s talking about “perfect”! Not all imperfections are equal!

    Ok, but fact remains that molecular divergence dates cannot falsify an earlier origin.

    Nothing in phylogenetics or divergence dating can strictly speaking be falsified; parsimony is all we have. You continue to imply that everything that isn’t falsified is equally likely; this continues to be a logical fallacy!

    “The order of divergence does not have to match the order of continental drift.”

    “What criterion do you have left, then, to test your own biogeographic hypotheses???”

    Tectonic correlation (as covered in several postings)

    Won’t you always get tectonic correlation if you just project it far enough in time?

    “So fossil calibrated molecular divergence estimates are not used for nematodes?”

    “Can’t be, because there’s nothing to calibrate them with. Read Raaum et al. (2005) on how to choose calibration dates.”

    There is no other group related to nematodes at some level that can be used?

    It’s been known for years that if all your calibration points come from outside the clade of interest, the ages of cladogeneses inside the clade of interest will be systematically overestimated. My 2007 paper shows this for one method and cites earlier papers that show it for another.

    I like the allusion to EOH

    What’s EOH?

    “You also haven’t explained why Neornithes should have a so much worse preservation potential than Enantiornithes throughout the Cretaceous.”

    Agreed, I don’t have an ‘explanation”.

    Then work on that, or I’ll continue to hold that the fossil record makes your scenario extremely unlikely.

    Interestingly, when discussing a parrot fossil from the Cretaceous the author did speculate on the lack of fossils for this and other extent [sic] groups.

    Yeah… that jaw fragment has not convinced many people. You see, there are no parrot features in it; either it’s specifically a crown-group lory, or it’s not a parrot at all. Considering all the Eocene stem-parrots which don’t have parrot beaks, the latter option is more parsimonious.

    Yes, but the Lebombo monocline is not in Africa.

    It is specifically in South Africa, Lesotho, Swaziland and Mozambique; and the Google Books result says it’s “the eastern side of the giant South African dome”. So… what are you talking about?

    “‘Prosimii’ is paraphyletic: the tarsiers are closer to us than to the lemurs & lorises.

    Only in some molecular reconstructions.

    No, in all of them, and also in all phylogenetic analyses of morphological data. The hypothesis that Prosimii is a clade is limited to the time when phylogenetics was an art.

    The names Strepsi- and Haplo(r)rhini both date back to that time as well, BTW; they’re not new!

    “It’s a parsimonious story. Have you got a more parsimonious one?”

    Yes, abut according to biogeographic criteria over which we disagree.

    Please explain how, given your criteria, it is more parsimonious. I want to count the assumptions.

    “Molecular work has rejected many long-established groups that were based on traditional morphology.
    “Traditional morphology” did not, however, use phylogenetic analysis.”

    Not correct. The analysis I published on humans and orangutans was based on morphology and phylogenetic analysis.

    …Not sure what you’re replying to here.

    “But there are a number of geological models requiring major basaltic plateau in the central Pacific.”

    “A basaltic plateau isn’t a continent; and no continent has been anywhere near the central Pacific in a long time.”

    Maybe true, but I was not invoking continents.

    But island biota has to come from a continent at some point. Why did you mention those basaltic plateaus, then?

    It’s the known sister group. There may be other closer sistergroups that have become extinct.

    That’s not what “sister-group” means. Except in a polytomy, a clade can only have one sister-group by definition.

    “Oh, speaking of Tonga (which has not just iguanas, but even frogs): was there ever a way to walk from Fiji to Tonga?”

    Is this a rhetorical question?

    No; but when I wrote that, I wasn’t yet sure that you accept transoceanic dispersal within unspecified limits, so the answer to my question may not be relevant anyway.

    Perhaps this means there is a problem with these molecular theories

    Perhaps it does. Have you got any other evidence that it does? Have you come up with another theory that explains the currently known data?

    I’ve read Yoder. Nothing there to make it a well supported hypothesis.

    Details, please.

    To be honest, I have no data to say whether the ancestors of such groups were first in what is now ‘Madagascar’ or in ‘Africa’

    Even if you arbitrarily declare the fossil record to contain “no data”, you need to postulate astounding numbers of dispersals out of Madagascar followed by an astounding rate and selectivity of extinction in Madagascar if you want to claim that Madagascar is where Placentalia comes from.

    Or these groups were not present in those areas that became those other continents.

    By “were not present”, you can only mean that some miracle prevented them from walking a few steps farther.

    Point was that an explanation may be better, but it does not meant that it will be accepted.

    Irrelevant as long as you don’t show that your explanation is better. You’ve hardly tried! Do that first, and then we can talk about sociology of scientists.

    OK, so what you are saying is that your point of view on molecular age is superior to those molecular theorists who have proposed origins with which you disagree.

    In some cases, yes. It’s not like I have no experience with molecular dating (me & Laurin 2007). Also, I’m a paleontologist; this allows me to tell that molecular biologists tend not to know a lot of paleontology, often leading to misinterpretations of the fossil record, on which calibration dates are often built (me & Laurin 2007, 2013a, and references therein).

    Panbiogeographers would also miss the presence of fossil marsupials in Eurasia

    To be fair, it’s not quite clear if any of them are crown-group marsupials; probably they’re all on the stem.

    I have not studied this group. A quick look shows China, Japan, and eastern N. America. Those groups alone could suggest a Pacific basin, but not necessarily as the N American species are border the Atlantic. You say “despite extensive fossil evidence for a holarctic distribution until quite recently”. This seems to say that quite recently there is now no extensive fossil evidence for a holarctic distribution. If there are fossils through N America and Eurasia it would be interesting to see if affinities are closer between those either side of the Pacific or Atlantic.

    Andrias, today restricted to China and Japan, used to cover Europe and western Asia. Its split from the North American Cryptobranchus may involve the Bering landbridge (north of the Pacific) or the Thule landbridge (across the northern Atlantic, involving the hotspot of Iceland) or perhaps the de Geer landbridge (the contact of Greenland with Spitzbergen). This is further complicated by the unsorted mess of fossil C. and/or A. species from North America and by additional cryptobranchid genera from the Paleogene of Mongolia and surroundings (like Aviturus). The closest known relative of the cryptobranchid crown-group is called Ukrainurus after the country it was found in.

    I cannot figure this out. As already stated, fossils are included in analyses and have been extensively in the panbiogeographic literature.

    …so you helpfully offered us a panbiogeographic perspective on paleognaths that completely ignored the fossils from the northern hemisphere. *slow clap*

    Link to this
  303. 303. DavidMarjanovic 3:28 pm 06/30/2014

    Note to self: at least 5 links are okay.

    Link to this
  304. 304. Halbred 3:49 pm 06/30/2014

    I don’t know, man. I advocate a view of science that is elegant and parsimonious. Basically, the KISS principle. Whatever theory squares the most evidence while leaving the fewest loose threads.

    For me, panbiogeography is not that. It forces me to make WAY too many assumptions or, at least, make claims like “there are Jurassic ratites but we haven’t found them yet.” Something about that statement makes me deeply uncomfortable.

    Link to this
  305. 305. Grehan1896 4:34 pm 06/30/2014

    I don’t know, man. I advocate a view of science that is elegant and parsimonious. Basically, the KISS principle. Whatever theory squares the most evidence while leaving the fewest loose threads.

    For me, panbiogeography is not that. It forces me to make WAY too many assumptions or, at least, make claims like “there are Jurassic ratites but we haven’t found them yet.” Something about that statement makes me deeply uncomfortable.

    That’s OK with me. We each make our choices in science.

    John Grehan

    Link to this
  306. 306. keesey@gmail.com 5:02 pm 06/30/2014

    “Among archosaurs or indeed limbed diapsids, however, the short-tailed birds (Avebrevicauda) are the only case of serious tail reduction.”

    Pterodactyloids!

    “tail reduction was only possible in maniraptorans after large-scale reorganization of locomotion and posture.”

    Of course, that still applies.

    Link to this
  307. 307. DavidMarjanovic 5:16 pm 06/30/2014

    That’s OK with me. We each make our choices in science.

    No. The scientific method consists of falsification and parsimony. Choice is not part of it.

    Pterodactyloids!

    Oopsie. :-] And anurognathids, too – though whether that’s independent is still a bit controversial, IIRC.

    Link to this
  308. 308. MichaelHeads 5:37 pm 06/30/2014

    David (comment 302), you wrote that: ‘Ratitae designates a group that has turned out to be para- or polyphyletic depending on how you define it, while Palaeognathae designates a clade as it did in 1900; there’s no reason to suddenly ditch Palaeognathae now’.

    Ratitae is a name – the delimitation of the group it denotes can change. This is a fundamental nomenclatural principle – you don’t need a new name every time the delimitation of a group changes, that would just lead to chaos. I know this doesn’t apply as a legal obligation to higher rank-clades, but the principle is the same. The name Ratitae (whether or not the group includes tinamous) is 87 years older than the name Palaeognathae and is very well known, even to many in the general public. Why do you think Darren used it for the title of this thread? Let’s face it, no-one outside our little group has heard of Palaeognathae.

    It was suggested that panbiogeography has a ‘bad name’, and I pointed out that in some circles it doesn’t. I mentioned that it doesn’t have a bad name in Latin America, and that Oxford, Cambridge, Berkeley etc. publish out work. You responded:

    ‘What you wrote says that Oxford U Press, U of Cal Press and Cambridge U Press are the “others” who don’t “read the fossil record literally” … If that’s not what you meant, express yourself more clearly next time’.

    Just because Oxford etc. publish our work doesn’t mean that they agree with it, or have any particular view on fossils (although presumably the reviewers did), but it does mean that we don’t have a ‘bad name’ with them. This is clear enough.

    I wrote: ‘So just how much older than its oldest fossil can a clade be? Pick any group and give me an example. It’s tricky. This is now acknowledged as ‘the problem of the priors’ in bayesian dating (the standard current technique)’.

    You wrote: ‘I’ve already said it’s not a black-and-white thing’.

    So you can’t pick any group and give an example?

    You wrote: ‘How about you follow my suggestion to read Raaum et al. (2005; linked to above)’

    I’ve read it. I cited and quoted it in my 2010 paper on primates (Zool. Scripta 39: 107).

    Link to this
  309. 309. Grehan1896 5:52 pm 06/30/2014

    302. DavidMarjanovic
    “First of all, apologies to John Grehan: I had somehow managed to overlook the “necessarily” in “No one contributing to this discussion is necessarily going to change their mind about anything” (comment 265), so the version in comment 144 that lacks “necessarily” was all I had to go with. That’s where about half of my anger came from.”

    No apologies necessary. I also forgot that I used that qualifier.

    “Google Scholar can’t find your paper; please tell me which volume and issue it’s in.”
    Copies of Mike’s papers care available at http://johngrehan.net/index.php/panbiogeography/panpublications/michael-heads-publications

    I’ve made an effort here to respond to points from David #268 while not being distracted by some of the emotive content

    You’re not getting an extra cookie for behaving like a scientist

    There was no tacit assumption about the fossil record outside the current distribution. Do you know anything about the fossil record of the groups in question that would change their biogeography in any particular way?

    “I mentioned the iguanas. On the one hand, there’s this:………. require a long ghost lineage at the base of Iguanidae sensu stricto, to which the iguanas of Fiji and Tonga belong.”

    How do the localities listed change the biogeography?

    “No paper is perfect.
    *eyeroll*
    Who’s talking about “perfect”! Not all imperfections are equal!
    True. No arguments there.

    Ok, but fact remains that molecular divergence dates cannot falsify an earlier origin.
    Nothing in phylogenetics or divergence dating can strictly speaking be falsified; parsimony is all we have.

    Ok, but papers have rejected earlier origins that predate the molecular estimate. So it seems we are in agreement here.

    “You continue to imply that everything that isn’t falsified is equally likely; this continues to be a logical fallacy!”

    I did not realize I was making that implication.

    “Tectonic correlation (as covered in several postings)
    Won’t you always get tectonic correlation if you just project it far enough in time?”
    Interesting question. I do not have an answer for that either.

    “It’s been known for years that if all your calibration points come from outside the clade of interest, the ages of cladogeneses inside the clade of interest will be systematically overestimated. My 2007 paper shows this for one method and cites earlier papers that show it for another.”

    Ok

    “I like the allusion to EOH
    What’s EOH?”
    Sorry –ethyl alcohol.

    “You also haven’t explained why Neornithes should have a so much worse preservation potential than Enantiornithes throughout the Cretaceous.”
    Agreed, I don’t have an ‘explanation”.
    Then work on that, or I’ll continue to hold that the fossil record makes your scenario extremely unlikely.

    That’s OK with me.

    Interestingly, when discussing a parrot fossil from the Cretaceous the author did speculate on the lack of fossils for this and other extent [sic] groups.
    Yeah… that jaw fragment has not convinced many people.

    OK.

    “It is specifically in South Africa, Lesotho, Swaziland and Mozambique; and the Google Books result says it’s “the eastern side of the giant South African dome”. So… what are you talking about?”

    Sorry, typo, I should have said not in Asia.

    “‘Prosimii’ is paraphyletic: the tarsiers are closer to us than to the lemurs & lorises.
    Only in some molecular reconstructions.
    No, in all of them, and also in all phylogenetic analyses of morphological data.

    I’ll have to go back and check with molecular literature, but you are incorrect that all phylogenetic analyses of morphological data support tarsiers being closer to anthropoids than other prosimians (unless you are referring to just quantitative analyses).

    Yes, abut according to biogeographic criteria over which we disagree.
    Please explain how, given your criteria, it is more parsimonious. I want to count the assumptions.

    The preponderance of correlated tectonic and biological distributions.

    “Traditional morphology” did not, however, use phylogenetic analysis.”
    Not correct. The analysis I published on humans and orangutans was based on morphology and phylogenetic analysis.
    …Not sure what you’re replying to here.”
    You said that traditional morphology did not use phylogenetic analysis. I was just giving an example where traditional morphology did use phylogenetic analysis.
    “But island biota has to come from a continent at some point. Why did you mention those basaltic plateaus, then?”

    Because they represent a possible source for the volcanic island biotas and even some of the continental biotas.

    It’s the known sister group. There may be other closer sistergroups that have become extinct.
    That’s not what “sister-group” means. Except in a polytomy, a clade can only have one sister-group by definition.
    Sure, then I’m not sure what you were disagreeing with.

    Perhaps this means there is a problem with these molecular theories
    Perhaps it does. Have you got any other evidence that it does? Have you come up with another theory that explains the currently known data?
    Well of course, but then we already disagree about that.

    I’ve read Yoder. Nothing there to make it a well supported hypothesis.
    Details, please.
    If I get time I will do that.

    “Even if you arbitrarily declare the fossil record to contain “no data”, you need to postulate astounding numbers of dispersals out of Madagascar followed by an astounding rate and selectivity of extinction in Madagascar if you want to claim that Madagascar is where Placentalia comes from.”

    I do not know where (geographically) the Placentalia originally come from.

    “By “were not present”, you can only mean that some miracle prevented them from walking a few steps farther.”

    No miracle. There are plenty of examples of organisms that do not span the entire geographic range available to them.

    “Irrelevant as long as you don’t show that your explanation isbetter. You’ve hardly tried! Do that first, and then we can talk about sociology of scientists.”

    Whether I show that my explanation is better or not is not determined by whether you agree that it is or not.

    Panbiogeographers would also miss the presence of fossil marsupials in Eurasia
    To be fair, it’s not quite clear if any of them are crown-group marsupials; probably they’re all on the stem.
    I have not studied this group. A quick look shows China, Japan, and eastern N. America. Those groups alone could suggest a Pacific basin, but not necessarily as the N American species are border the Atlantic. You say “despite extensive fossil evidence for a holarctic distribution until quite recently”. This seems to say that quite recently there is now no extensive fossil evidence for a holarctic distribution. If there are fossils through N America and Eurasia it would be interesting to see if affinities are closer between those either side of the Pacific or Atlantic.

    Andrias, today restricted to China and Japan, used to cover Europe and western Asia. Its split from the North American Cryptobranchus may involve the Bering landbridge (north of the Pacific) or the Thule landbridge (across the northern Atlantic, involving the hotspot of Iceland) or perhaps the de Geer landbridge (the contact of Greenland with Spitzbergen).

    Or not.

    This is further complicated by the unsorted mess of fossil C.and/or A. species from North America and by additional cryptobranchid genera from the Paleogene of Mongolia and surroundings (like Aviturus). The closest known relative of the cryptobranchid crown-group is called Ukrainurus after the country it was found in.

    So, nothing to preclude a Pacific or Atlantic relationship at this time.

    I cannot figure this out. As already stated, fossils are included in analyses and have been extensively in the panbiogeographic literature.
    …so you helpfully offered us a panbiogeographic perspective on paleognaths that completely ignored the fossils from the northern hemisphere. *slow clap*
    Mike commented on the northern hemisphere fossils I think.

    John Grehan

    Link to this
  310. 310. irenedelse 6:09 pm 06/30/2014

    @MichaelHeads:

    “I cited quite a bit of data in my last post. Here’s another example of data that is consistent with differentiation in ratites being caused by vicariance. The kiwi, Apteryx, comprises two main clades, brown and spotted. Sequencing studies of extant and sub-fossil populations show that the brown kiwi clade is divided into two allopatric groups of species. Their mutual boundary lies along the Alpine fault (a plate boundary), not the main geographic divide, which runs parallel with it but 10 km to the east.

    A molecular phylogeny of the moa, Megalapteryx, shows sister clades that are disjunct along the Alpine fault, between Fiordland and Nelson. The gap, hundreds of kilometres, is filled by other clades. The gap is consistent with strike slip (horizontal) displacement along the fault.”

    Very interesting indeed, but you’re talking here about the distribution of Kiwi on the one hand and of Moa on the other… Not about a the Moa-Tinamou clade or the Kiwi-Elephant bird clade. So even if panbiogeographic concepts can be useful if we consider the two groups of lower taxonomic rank (Kiwi, Moa) within Zealandia, we can’t just extrapolate to a larger clade like the one made by Kiwi and Elephant birds and spanning the Indian ocean. The dispersion hypothesis, proposed by Mitchell and al. (2014) does look here like the more parsimonious one.

    Link to this
  311. 311. ectodysplasin 6:11 pm 06/30/2014

    @John,

    I have not analyzed lungfishes. Need to explain please.

    Modern lungfishes occur in South America, Africa, and Australia. However, there’s an extensive fossil record of crown and stem lungfishes on all continents. Where you draw the line between stem and crown lungfishes is important. I’ve seen quite a lot of people try to assign a Jurassic age to the crown on the basis of biogeography, but the fossil evidence suggests a lowermost Triassic origin at the very latest.

    I have not studied this group. A quick look shows China, Japan, and eastern N. America. Those groups alone could suggest a Pacific basin, but not necessarily as the N American species are border the Atlantic. You say “despite extensive fossil evidence for a holarctic distribution until quite recently”. This seems to say that quite recently there is now no extensive fossil evidence for a holarctic distribution. If there are fossils through N America and Eurasia it would be interesting to see if affinities are closer between those either side of the Pacific or Atlantic.

    The fossil evidence suggests that until the Plio-Pleistocene, there were cryptobranchids all over the place. Sorry if my wording confused you. As for Atlantic/Pacific, the North American cryptobranchids are not especially close to the Atlantic Ocean; there are two separate and disjunct populations in different portions of the Mississippi River drainage, but all are separated from the Atlantic by at least one mountain range. In addition, there are fossil cryptobranchids from Western Nebraska (center of the continent) and from Saskatchewan (an area that drains to the Arctic Ocean), as well as the various Eurasian taxa that David noted above.

    I cannot figure this out. As already stated, fossils are included in analyses and have been extensively in the panbiogeographic literature.

    How do you know if the fossils are crown group or stem group? In other words, how do you know which fossils to include and which to exclude? What are your criteria?

    But would apply to any analysis, biogeographic or otherwise. In science one works with information that one has. To date I am not aware of any new fossil evidence arising that has fundamentally changed the panbiogeographic picture elucidated from living representatives (this is over the last few decades and I may have forgotten something).

    Well, apparently that’s because there are a lot of taxa which you have not studied.

    However, this does not change the fact that your position on ratite evolution does not apparently follow your theoretical understanding of the limitations of the methods.

    It is a spatial parsimony criterion – that in the absence of other evidence (e.g. phylogeny, diversity concentration) the shortest distance between different localities is treated as generating the least number of adhoc spatial explanations required for that distribution. Now one may disagree with this, but I am only providing the explanation here.

    Whoa, whoa, whoa there. That’s a massive, massive misuse of the term “parsimony” and the term “ad hoc hypothesis.” Moreover, the treatment of range data (which is not necessarily point-localized in real life) dramatically affects how one might assume “ad hoc hypotheses of dispersal.” Additionally, you have to address why absolute map distance is a reasonable proxy for these “ad hoc hypotheses of geographic dispersal” when overland dispersal and overseas dispersal might be very different.

    “and “how do different clustering algorithms affect the topology of a MST?” among other things). “

    Please give an example

    UPGMA vs centroid clustering will give you different dendrograms due to different assumptions behind the clustering methods.

    As said before, it does not automatically trump anything. It may or may not corroborate other evidence or interpretations. One may make a choice about what to support.

    The fact you see fit to repeatedly argue for Jurassic ratites suggests that you do indeed think it trumps other sources of evidence. Which is fine, but requires justification that you have not offered in your comments here.

    It may be a good idea to read Mike’s books to make that kind of judgment (and Craw et al as an introductory text).

    It should not be terribly difficult for you to briefly explain the utility of your methods. You should not have to refer me elsewhere.

    Link to this
  312. 312. irenedelse 6:17 pm 06/30/2014

    As an aside, a suggestion for all but especially John and Michael: when quoting someone, especially if there are multiple quotes in one post, using the italics tag makes it more readable. It can be hard in such a long thread to keep track of what part is answering to what! Thanks.

    Link to this
  313. 313. Grehan1896 6:18 pm 06/30/2014

    David,

    I was wrong about my recollection of having read Yoder et al 2007 (it was Yoder & Nowak 2006). Please give me the full citation as I did not see it in the publication list for Yoder.

    Thanks,

    John Grehan

    Link to this
  314. 314. Grehan1896 8:00 pm 06/30/2014

    As an aside, a suggestion for all but especially John and Michael: when quoting someone, especially if there are multiple quotes in one post, using the italics tag makes it more readable. It can be hard in such a long thread to keep track of what part is answering to what! Thanks.

    I hate to appear dimmer than some think I am already, but where do I find the italics tag?

    John Grehan

    Link to this
  315. 315. Halbred 8:18 pm 06/30/2014

    There isn’t one–you have to do it manually: (remove spacing in the coming example)
    Torosaurus means “fenestrated lizard.”
    Becomes
    Torosaurus means “fenestrated lizard.”

    (not “bull lizard” as is often mistranslated. If Marsh had wanted to use that, he would have used “Taurosaurus.”)

    Link to this
  316. 316. Halbred 8:19 pm 06/30/2014

    Well, dammit. The spacing did nothing.

    something something

    There.

    Link to this
  317. 317. Halbred 8:19 pm 06/30/2014

    OH WHAT THE HELL

    Link to this
  318. 318. Christopher Taylor 8:51 pm 06/30/2014

    <i>italics</i> = italics

    Link to this
  319. 319. Grehan1896 9:09 pm 06/30/2014

    italics

    Link to this
  320. 320. Grehan1896 9:10 pm 06/30/2014

    It works! Decades since I ever wrote html. Thanks!

    John Grehan

    Link to this
  321. 321. Stripeycat 9:30 pm 06/30/2014

    Hmmm, this is a fascinating thread that’s really taken off the last few days I’ve been enfootballed. Going right back to my question about angiosperm origins (148, 151 from Ectodysplasin, 154 from John G) John and I were apparently talking about stem and crown groups respectively; fine, but not really very good evidence for ghost lineages (as he initially implied). Stem-angiosperms almost certainly exist in known fossils (as 151 said), but we can’t positively distinguish them from stem-gymnosperms (and given how divergent the gymnosperm crown is, maybe from weird crown-gymnosperms). Anyway, not relevant to ratites/paleognaths, but thanks for answering.

    On the Galapagos/island arc thing, do you say an arc that became part of the Cordillera crossed over the hotspot at some point? I suppose that’d be pretty hard to prove or disprove, although we’d still want to know where the arc’s fauna came from. Since it’d be moving W-E, how would it carry the predominantly American fauna without overwater dispersal?

    There seems to be an attitude from Mike and John that if vicariance is good in some circumstances, then it’s good in all: the NZ Alpine Fault is totally irrelevant to discussions of (say) early paleognath evolution. No-one denies the use of biogeography in some cases, but one must consider each application on its own merits, without irrelevant appeals to authority.

    Further to the above, I’d suggest a general brush-up in introductory logic: there’s a whole bundle of circular arguments (assigning clades to tectonic basins on the basis of the phylogeny, then claiming that the phylogeny supports the geographical inferences) that they both repeatedly decline to address.

    There’s also the willingness to dismiss as SEP any counterarguments relating to fossil preservation – Occam’s Razor isn’t the only tool in the box, but when you need many freak occurences for your hypothesis to work, the ball is in your court. While not a strict logic failure, probabilities must be considered (after all, it’s not /impossible/ or even logically fallacious that everything happens because fairies; it’s just damned unlikely). While early/mid-Jurassic crown birds aren’t that unlikely, they’re on a par with mid-Silurian terrestrial vertebrates.

    Link to this
  322. 322. Grehan1896 11:55 pm 06/30/2014

    321. Stripeycat
    On the Galapagos/island arc thing, do you say an arc that became part of the Cordillera crossed over the hotspot at some point?

    That’s possible, but I doubt anyone will be able to accurately geologically reconstruct the extent and connections of various arcs in the eastern Pacific. Two arcs that most pertain to the Galapagos are what are now the Greater Antilles and the various chunks embedded in Central America and NE S America (I am generalizing about this as there were probably more, but these have been the two principal components in geological reconstructions.

    I suppose that’d be pretty hard to prove or disprove

    Geologically yes, although one geological study looked at terranes in Ecuador and noted that some of the geochemistry was very much the same as the magma produced at the Galapagos hotspot

    although we’d still want to know where the arc’s fauna came from.
    One might, but it will depend on what one is looking for in the way of an answer. One might might say we still want to know where a continent’s fauna came from. At a certain point this probably becomes an insoluble question since we do not have the detail in the fossil record that we might see in the living taxa today. Admittedly for some groups one might be able to trace some detail. For analysis of living taxa the principal concern for evolutionary biogeography has been the origin of vicariant distributions. Beyond those patterns one is left with the origin of the ancestor and that would be a far more problematic proposition.

    Since it’d be moving W-E, how would it carry the predominantly American fauna without overwater dispersal?

    But what makes the fauna “American”. Just because some of the fauna of the Galapagos has American relatives does not make the group “American” biogeographically speaking. The group might just as well be considered “Galapagan”. In panbiogeography at least, biota are not classified by what continent they occupy. One alternative is to classify the fauna and flora with respect to the tectonic basins involved. If this is new to you I suggest reading Mike’s books for working examples.

    There seems to be an attitude from Mike and John that if vicariance is good in some circumstances, then it’s good in all

    That has never been our attitude.

    the NZ Alpine Fault is totally irrelevant to discussions of (say) early paleognath evolution.

    But was a good illustration of tectonic correlation and included a nice ratite example for two species groups suggesting each of these species differentiated at least 20 Ma. I know that means a ghost lineage for about that time given that modern kiwis have no pre-Quaternary fossil record and the 16 Ma fossil is sufficiently different to be placed in an different genus. Hmm, I wonder if extant kiwi species can be distinguished on skeletal features alone.

    No-one denies the use of biogeography in some cases, but one must consider each application on its own merits, without irrelevant appeals to authority.
    Sure. Don’t know what you mean about the appeals to authority

    Further to the above, I’d suggest a general brush-up in introductory logic: there’s a whole bundle of circular arguments (assigning clades to tectonic basins on the basis of the phylogeny, then claiming that the phylogeny supports the geographical inferences) that they both repeatedly decline to address.

    I realize that panbiogeography is probably quite unfamiliar to you (at least that is what I infer from your assertions), but the assignment of clades to tectonic basins is not made on the basis of the phylogeny since phylogeny is biological and has no spatial characteristics.

    There’s also the willingness to dismiss as SEP any counterarguments relating to fossil preservation –
    Well one might argue the converse regarding dismissal of biogeography.

    Occam’s Razor isn’t the only tool in the box, but when you need many freak occurences for your hypothesis to work, the ball is in your court. While not a strict logic failure, probabilities must be considered (after all, it’s not /impossible/ or even logically fallacious that everything happens because fairies; it’s just damned unlikely).

    Thank you for the opinion.

    While early/mid-Jurassic crown birds aren’t that unlikely, they’re on a par with mid-Silurian terrestrial vertebrates.
    Interesting opinion

    John Grehan

    Link to this
  323. 323. LeeB 1 12:04 am 07/1/2014

    Actually regarding kiwi distributions you might want to see Shepherd et. al’s paper in PlosOne of August 2 2012.
    http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0042384#pone-0042384-g002

    Figure 1 shows the distributions of all the Kiwi species derived from recent and subfossil DNA and this doesn’t seem to show the matches with the Alpine Fault that you mention.
    And lesser spotted kiwi bones can be differentiated fron the other species on size whereas the bones of the others are not differentiable morphologically; hence the use of DNA to identify them.

    LeeB.

    Link to this
  324. 324. LeeB 1 12:21 am 07/1/2014

    And the Encyclopedia of New Zealand moa distributions here: http://www.teara.govt.nz/en/interactive/11365/distribution-of-moa

    are not a match with the alpine fault either; they seem to reflect better the disjunction between the wetter west coast of the South Island and the drier east side; and a major disjunction between the North and South Islands.
    LeeB.

    Link to this
  325. 325. MichaelHeads 1:00 am 07/1/2014

    Comment 323 Shepherd 2012: ‘this doesn’t seem to show the matches with the Alpine Fault that you mention’.

    Thanks for the query. Have a close look at their Fig. 1 (bottom figure, brown kiwis). The basal phylogenetic break is between the red + blue symbols, and the black symbols. The break lies between Haast and Okarito – both west of the main divide, but separated by the Alpine fault.

    Link to this
  326. 326. MichaelHeads 1:10 am 07/1/2014

    Comment 324: the Encyclopedia of New Zealand moa distributions here: http://www.teara.govt.nz/en/interactive/11365/distribution-of-moa

    ‘are not a match with the alpine fault either’

    These maps don’t show the molecular clades though. have a look at Bunce et al 2009 PNAS 106: 20646. (Free access). The molecular clades of Megalapteryx are shown in Fig. 4.

    Link to this
  327. 327. LeeB 1 1:31 am 07/1/2014

    Yes but the authors give a divergence date for the Megalapteryx clade from the other Moa of 5.8my, and the Megalapteryx clades match the wet forests to the west of the Southern Alps and the drier forests to the east, with a total lack of Megalapteryx in the central part of the eastern South Island where it was too dry for the forests they inhabited.

    LeeB.

    Link to this
  328. 328. MichaelHeads 1:35 am 07/1/2014

    Comment 302. I wrote: ‘Also see cassowaries in New Guinea and NE Queensland, and emus of the rest of Australia (some very local overlap in NE Queensland). Again, this is a very common pattern (e.g. in marsupials, Fig. 4.10 in my 2014 book)’.

    David, you replied: ‘That’s ecological: cassowaries live in tropical rainforests, emus in drier habitats. As it happens, tropical rainforest on the Australian continent is limited to New Guinea and a few patches in NE Queensland…’.

    Cassowaries live in cold rainforest,as well as hot rainforest. There is plenty of cold rainforest in SE Australia/Tasmania, very similar to that in montane New Guinea, but there are no cassowaries anywhere in the region. There are also large areas of dry, open grassland and savanna in New Guinea (e.g. Fly area, Markham area) where there are no emus, only cassowaries. Cassowaries are well known from savanna and grassland (e.g.Davies, 2002, Ratites ans tinamous, Oxford U.P.).

    In any case, aridification in Australia began in the Miocene (e.g. Davis et al., 2013, Global Change Biology 19: 1970), while emu (Dromaiidae) fossils are known from the Oligocene (Mayr, 2009) and are already distinct from cassowaries (tip of premaxilla rounded as in emus, not pointed as in cassowaries).

    So simple ecological sorting doesn’t work. An alternative idea is that emus occupy open habitat because they are in Australia – they are not in Australia because they like open habitat. Cassowaries are mainly in rainforest because they are a New Guinea group, and New Guinea (at the moment) is largely covered in rainforest.

    Link to this
  329. 329. MichaelHeads 1:53 am 07/1/2014

    Comment 298: ‘I can think of a number of taxa with good to decent fossil records for which panbiogeography would offer a positively misleading result. A perfect example would be lungfishes, which a panbiogeographer would assign a Gondwandan distribution and an origin in the Jurassic or early Cretaceous’.

    No, we try and incorporate fossil dates (and localities) and fossil calibrated clock dates whenever possible; these are often very useful as minimum ages. So lungfish distributions can’t be the result of Gondwana distribution, because of the older fossils. Another example would be the New Zealand frog – this can’t be the result of Tasman Sea rifting (starting at 84 Ma)because fossil-calibrated clock dates (minimum ages!) are too old. Gondwana breakup is not the only means of vicariance, although it is often assumed to be. For example, pre-drift rifting often lasts for tens of millions of years and is probably just as important.

    Link to this
  330. 330. naishd 4:12 am 07/1/2014

    Cassowaries: myself and Richard Perron have a paper on cassowary evolution in press and I’ll be discussing it here in time. The fossil, molecular and anatomical data indicates that cassowaries are ancestrally Australian, migrating via terrestrial connections to New Guinea in several waves during times of low sea-level. Fossil cassowaries from Australia are equally as old as those from New Guinea, plus cassowaries still occur on suitable habitat on Australia. In other words, I agree with suggestions that cassowary absence from much of Australia is due to lack of suitable habitat.

    Link to this
  331. 331. MichaelHeads 4:35 am 07/1/2014

    Comment 327: Yes but the authors give a divergence date for the Megalapteryx clade from the other Moa of 5.8my…

    I’m sorry, I just don’t take the date very seriously as anything other than a minimum clade age. For example, the calibrations (Bunce et al., Supp. Info) include: ‘Casuariidae (Emu vs. cassowary). Normal distribution, median at 25.5 Ma [the oldest emu fossil date], 95% range from 16 to 35 Ma.’ In other words, the maximum age of emu and cassowary is stipulated as 35 Ma (with 95% credibility!) by the authors before any analysis (as a ‘prior’). Why not 36 or 37, or 47, or 67…? It’s completely arbitrary. (For more details, see my paper on bayesian dating in J. Biogeogr. 39: 1749. 2012).

    On the other hand, the Bunce et al. phylogeny is very interesting.

    … and the Megalapteryx clades match the wet forests to the west of the Southern Alps and the drier forests to the east.

    It’s not that simple. For example, the eastern Clade A inhabits areas (e.g. south of the Nelson lakes) that are wetter than substantial areas occupied by the western clade B + C (e.g. north of the Nelson lakes). The sampling isn’t great, but it looks as though it’s just another case of Alpine fault disjunction. This is now documented in more than 200 plants and animals, including kiwis. (This is from a book on New Zealand that I’ve just submitted. If you’re interested in more details, let me know).

    Link to this
  332. 332. Grehan1896 12:37 pm 07/1/2014

    And lesser spotted kiwi bones can be differentiated from the other species on size whereas the bones of the others are not differentiable morphologically

    This would seem to make things difficult if one were looking to fossil corroboration of their phylogenetic age. If one cannot make distinctions how would one know that one has found a fossil of a particular species without having DNA sequences? If the brown kiwis are at least 20 Ma old (and as pointed out by Mike, the distributions are disjunct with respect to the fault) then it would seem that one might find their fossils but not be able to identify them.
    John Grehan

    Link to this
  333. 333. Grehan1896 12:49 pm 07/1/2014

    330. naishd
    Cassowaries: myself and Richard Perron have a paper on cassowary evolution in press and I’ll be discussing it here in time. The fossil, molecular and anatomical data indicates that cassowaries are ancestrally Australian

    This will be interesting to see – that the data actually indicate that cassowaries are “ancestrally” Australin, or will it be a particular theoretical interpretation that leads to that hypothesis.

    migrating via terrestrial connections to New Guinea in several waves during times of low sea-level. Fossil cassowaries from Australia are equally as old as those from New Guinea, plus cassowaries still occur on suitable habitat on Australia. In other words, I agree with suggestions that cassowary absence from much of Australia is due to lack of suitable habitat.

    But the point Mike made is that there are other large areas of suitable habitat in Australia so their absence from much of Australia could be due to lack of suitable habitat, but that does not explain their absence from those areas where there is suitable habitat. Biogeographic analysis does, however, provide a coherent explanation that does not lead to the habitat incongruence.

    Since the paper is already in press, is there anything preventing you circulating a pre-print at this time so it will be possible to assess the basis of your biogeographic claims?

    John Grehan

    Link to this
  334. 334. keesey@gmail.com 12:53 pm 07/1/2014

    “the 16 Ma [kiwi] fossil is sufficiently different to be placed in an different genus”

    Genera are bookkeeping bins erected artificially for our convenience — they don’t *mean* much of anything.

    A more interesting question is whether it’s a stem-kiwi or part of the crown. (I’d assume the former.)

    Link to this
  335. 335. Grehan1896 2:09 pm 07/1/2014

    Genera are bookkeeping bins erected artificially for our convenience — they don’t *mean* much of anything.
    If it were referring to something phylogenetically informative based on some analysis then they can refer to monophyletic units that do have some phylogenetic meaning. In this case I suspect this is not the case.

    A more interesting question is whether it’s a stem-kiwi or part of the crown. (I’d assume the former.)
    I take it by “stem-kiwi” that you mean it will have a sister group relationship with all the living kiwis.

    John Grehan

    Link to this
  336. 336. keesey@gmail.com 2:41 pm 07/1/2014

    A stem group is the total group minus the corresponding crown group. So a stem-kiwi would be something sharing more ancestry with living kiwis than with anything else alive, but not descended from the last common ancestor of living kiwis.

    There is a movement in some fields to convert generic names to clade names. This tends to lead to a lot of splitting, and ignores that some type species might be ancestral to others. (E.g., it’s awfully difficult to use “Australopithecus” for a clade.) But that still doesn’t measure difference, since a clade can be any size, from a single taxonomic unit to the clade of all life (Panbiota).

    Link to this
  337. 337. keesey@gmail.com 5:30 pm 07/1/2014

    Come to think of it, under the new phylogeny gleaned from elephant bird DNA, elephant birds would also be stem-kiwis. Although that starts to get into an odd area where you could quibble about whether something that went extinct in the 1600s should be considered part of a stem group. (Most would probably say yes, but….)

    I discussed this sort of issue here: http://3lbmonkeybrain.blogspot.com/2009/02/extinct-or-extant.html

    Link to this
  338. 338. Halbred 6:28 pm 07/1/2014

    Mike, something I’ve always wondered about stems and crowns: for extinct groups (like, say, ceratopsids), what would the “crown” be? Does that term even apply?

    Would the “crown” be the Triceratopsini, because members of that group were the last ones alive, and they are clearly on a different branch from their closest relatives (Arrhinoceratops, Anchiceratops)? That is, even though Eotriceratops was extinct before the end, it was still on the Triceratopsini branch, so it’s within the crown?

    Am I even making sense?

    Link to this
  339. 339. keesey@gmail.com 8:16 pm 07/1/2014

    Hallbred,

    No, the terms do not apply within a completely extinct clade. I’ve seen a few publications try to use them, but it’s not the proper usage. There’s no such thing as a crown ceratopsid. (Although it’d be awesome if there were!)

    Link to this
  340. 340. Halbred 8:28 pm 07/1/2014

    Man, would it ever be. Thanks for answering!

    Link to this
  341. 341. MichaelHeads 3:06 am 07/2/2014

    330 Darren, it’ll be very interesting to see your molecular evidence for cassowary dispersal from Australia. Are you really suggesting that the New Guinea species are all nested in the Queensland population?!

    You say: ‘I agree with suggestions that cassowary absence from much of Australia is due to lack of suitable habitat’. But there is extensive suitable habitat (rain forest) in eastern mainland Australia south of Cairns and also in Tasmania, where cassowaries are not present.

    And as I mentioned, cassowaries are well-known from savanna and grassland in New Guinea, which indicates that there are vast areas of suitable habitat in drier Australia.

    Also, Mayr (2009) concluded that the Oligocene fossils from Riversleigh (far NW Queesnland) are emus, not cassowaries, and this was before aridification began.

    I have a question: at high altitude in the New Guinea mountains (~2500 m)I’ve often heard cassowaries roaming in the forest very late at night, at 10 or 11, well after dark (which is always near 6 pm). I assumed this was well-known, but I can’t find any reference to it in the ornithological literature. Can anyone help? Aren’t kiwis supposed to the only nocturnal ratites?

    Link to this
  342. 342. MichaelHeads 3:12 am 07/2/2014

    337 under the new phylogeny gleaned from elephant bird DNA, elephant birds would also be stem-kiwis.

    No, in the Mitchell et al. 2014 phylogeny (Fig. 1), kiwis and elephant birds are reciprocally monophyletic – they’re sister-groups.

    Link to this
  343. 343. keesey@gmail.com 3:32 am 07/2/2014

    MichaelHeads,

    That is not inconsistent with elephant birds being part of the kiwi stem group. (See definition of “stem group” above.)

    Link to this
  344. 344. BrianL 3:43 am 07/2/2014

    Call this tooting my own horn, but I’m rather fond of the fact that my question about Madagascar’s hippos (and especially the answer given to that by John Grehan) has apparently reached memetic levels.

    Link to this
  345. 345. naishd 5:53 am 07/2/2014

    Cassowaries… In comment # 341, Michael says…

    Darren, it’ll be very interesting to see your molecular evidence for cassowary dispersal from Australia. Are you really suggesting that the New Guinea species are all nested in the Queensland population?!

    I repeat what I said earlier: fossil cassowaries are not limited to New Guinea – there are specimens of similar age from both New Guinea and Australia. Furthermore, the specimens concerned appear to be outside of the crown-cassowary clade (I am not talking about Emuarius, which now seems to be a stem-emu). Indeed we are not suggesting that the cassowaries of New Guinea are nested within the Queensland population of C. casuarius. However, our molecular analysis shows C. casuarius to be the siser-taxon to the others. The overall picture that emerges is of an origin on Australia followed by dispersal to, and endemic radiation on, New Guinea. The fact that cassowaries are lacking from what seems like suitable habitat from parts of Australia and Tasmania may not be informative with respect to their historical distribution – animals do not occur in all the places we might predict for lots of reasons. In any case, one of the Australian fossil cassowaries is from down in the south-east: the Wellington Valley in New South Wales… somewhat south of Cairns!

    Nocturnal habits are reasonably well known (as is: often mentioned), but not yet the subject of detailed study, in cassowaries. The more we learn, the more it seems that ‘diurnal’ birds do a lot more during darkness than just sleep (cf. swans, plovers, peregine falcons, bald eagles).

    Link to this
  346. 346. Grehan1896 11:27 am 07/2/2014

    142. naishd4:25 pm 06/25/2014

    As for “how much older than the oldest known fossil are ratites allowed to be?” (comment # 141), they should be younger than the numerous early neornithine, ornithurine and ornithuromorph fossils we have. That’s a lot of data-points (there are numerous taxa involved). Even with a generous bit of slop, this puts palaeognath origins no later than the Early Cretaceous – something like 120 Ma ago at most, and this generously takes the total lineage into account, not the crown-group only.

    In previous discussions there has been an understandable objection to the idea of ratites (let alone extant birds in general) having arisen before the separation of Madagascar. Although the date for this event is often referred to as about 160 Ma, the complete separation was not immediate because the movement of Madagascar was oblique relative to Africa along the Davies Fracture Zone. This means that Madagascar is proximate or in contact with mainland Africa for a period of time after the initial opening of the Somali and Mozambique basins. It took me a while to dig up something specific as to when this process was completed. It seemed that there was more precision and reference to the inception than to the completion. A recent paper by Colin Reeves, (2014 The position of Madagascar within Gondwana and its movements during Gondwana dispersal. Journal of African Earth Sciences 94 , 45–57) provides some clarification. His models shows Madagascar still in contact with Africa at 145 Ma and the position shown indicates that this would have been maintained for some time after, although there is nothing specified. By 120 Ma the process is well past complete.

    The Reeves model does allow for final separation by about 140 Ma – into the early Cretaceous. If one were to suppose that ratites diverged from that time, this would give a 20 Ma difference between what Darren would find acceptable and what the tectonic reconstruction might predict. So that might be still too much to accept (ghost lineages of about 50 Ma being OK, but 70 Ma not being OK), but I thought the tectonic details interesting, and may be interesting to others.

    John Grehan

    Link to this
  347. 347. Grehan1896 11:57 am 07/2/2014

    Darren,

    If you do not want to release your in press article at this time (although I would be interested to know what the problem would be), could you be a little more explicit as to exactly what evidence you see as pointing to an origin on Australia followed by dispersal to, and endemic radiation on, New Guinea

    Thanks,
    John Grehan

    Link to this
  348. 348. DavidMarjanovic 12:02 pm 07/2/2014

    Unfortunately, I don’t currently have time to catch up with replying, though I’ve read the entire thread so far. Just this much for now:

    Since the paper is already in press, is there anything preventing you circulating a pre-print at this time so it will be possible to assess the basis of your biogeographic claims?

    Preprints are, culturally, extremely rare outside of physics; and while I doubt – because of its subject – that the paper is in Nature or Science or anything like them, the most prestigious journals will cancel publication even at this point if you “break the embargo” on your own paper or if anybody else does.

    Unless, I guess, if it’s literally in the press. :-)

    Link to this
  349. 349. naishd 12:18 pm 07/2/2014

    Yeah, sorry, not ready to release information on the paper yet. Might be a while…. but watch this space!

    Link to this
  350. 350. Grehan1896 1:12 pm 07/2/2014

    OK, if it has a journal embargo that’s understandable. But I would still be interested in an explicit statement as to the nature of the evidence for Darren’s posted assertions.

    John Grehan

    Link to this
  351. 351. naishd 1:25 pm 07/2/2014

    Well, there’s only so many times I can repeat the same thing. See comment # 345.

    Link to this
  352. 352. Grehan1896 2:42 pm 07/2/2014

    351. naishd
    Well, there’s only so many times I can repeat the same thing. See comment # 345.

    I had seen the comment as a series of assertions rather than evidence, but I will go over the comment below:
    fossil cassowaries are not limited to New Guinea – there are specimens of similar age from both New Guinea and Australia. Furthermore, the specimens concerned appear to be outside of the crown-cassowary clade (I am not talking about Emuarius, which now seems to be a stem-emu).

    Would I be correct to understand from this that Emuarius is the sistergroup of Dromaius?

    However, our molecular analysis shows C. casuarius to be the siser-taxon to the others.

    Meaning sister group of Emuarius + Dromaius?

    The overall picture that emerges is of an origin on Australia followed by dispersal to, and endemic radiation on, New Guinea

    This is an assertion about historical events. It does not present any evidence and it was the evidence that I was asking about – what constitutes evidence for an origin on Australia (I presume you mean the common ancestor of Casuarius + Dromaius + Emuarius or are you referring to something else?). So what is the evidence?

    The fact that cassowaries are lacking from what seems like suitable habitat from parts of Australia and Tasmania may not be informative with respect to their historical distribution – animals do not occur in all the places we might predict for lots of reasons.

    Agreed, and that is the point – habitat alone does not ‘explain’ the distribution of Casuarius.
    In any case, one of the Australian fossil cassowaries is from down in the south-east: the Wellington Valley in New South Wales… somewhat south of Cairns!

    Which would show that Casuarius had a wider range in the past than at the present – in which case I presume you are inferring that the cassowary could have been in other suitable habitat in the past, but became extinct in those localities.

    Earlier literature that I have seen seemed to indicate that the locality record was problematic, that there was insufficient evidence for corroboration. Has this since changed?

    John Grehan

    Link to this
  353. 353. Stripeycat 3:53 pm 07/2/2014

    John #352
    /Emuarius/ is indeed a probable sister-taxon to modern emus. Its relationship to cassowaries is more distant.

    As far as I understand what Darren said in 345, within /Casuarius/, /C. casuarius/ is less closely related to the exclusively New Guinean species than they are to each other (ie it diverged first, then they did). It isn’t clear from what he says whether New Guinea /C. casuarius/ groups with the mainland strain of the species, or with the others. If the latter, then it is very likely that all the New Guinean populations evolved from a single incursion from Australia. If the /C. casuarius/ populations nest one inside each other, then they probably came from the broader group; and if they’re separate groups then either they came from a third source population, or a large chunk of the parent population died out.

    Link to this
  354. 354. MichaelHeads 4:37 pm 07/2/2014

    Hi Darren, you wrote (345): ‘our molecular analysis shows C. casuarius to be the siser-taxon to the others. The overall picture that emerges is of an origin on Australia followed by dispersal to, and endemic radiation on, New Guinea’.

    If C. casuarius is sister to the others, another possibility would be simple vicariance between emus and cassowaries, and then vicariance within the cassowaries, with some blurring (overlap) around the edges caused by subsequent dispersal. In this model there would be no need for any long-distance dispersal, just breakup of a global ancestor.

    The fact that cassowaries are lacking from what seems like suitable habitat from parts of Australia and Tasmania may not be informative with respect to their historical distribution – animals do not occur in all the places we might predict for lots of reasons.

    Yes, but we have to explain the distribution data (fossil and extant) that we do have. As I mentioned, cassowaries are well-known from savanna and grassland, but are absent from this type of vegetation in Australia; emus are in savanna and grassland, but are absent from this type of vegetation in New Guinea. The conjunction of the two absences looks significant.

    Emus and cassowaries are for the very great majority of their range (living and fossil) allopatric, with minor overlap along the east coast of Australia. Perhaps there are more fossils indicating more overlap, i.e. more dispersal, but the overall allopatry is impressive and can be explained very simply by vicariance. Likewise, the allopatry (living and fossil) with the related kiwi in New Zealand and elephant bird in Madagascar is striking and can be explained by vicariance (or, of course, by chance!).

    In the same way, the very high levels of allopatry in the other ratite families can be explained simply by vicariance, while the overlap of kiwis and moas, and of rheas and tinamous, can be explained by dispersal.

    Nocturnal habits are reasonably well known (as is: often mentioned), but not yet the subject of detailed study, in cassowaries.

    Great – thanks! I’d be very grateful indeed for any reference. I’ve mentioned this in a submitted manuscript so I’m very keen to know more. I’m arguing that nocturnality in kiwis is not the result of adaptation to special New Zealand conditions, but is just a local byproduct of a prior, general trend to nocturnality in ratites and other birds.

    Link to this
  355. 355. irenedelse 5:12 pm 07/2/2014

    @John Grehan #332:

    “But what makes the fauna “American”. Just because some of the fauna of the Galapagos has American relatives does not make the group “American” biogeographically speaking. The group might just as well be considered “Galapagan”.”

    The past is full of adventures. Iguanas, finches, giant turtles… So, did all Sauropsids originating from one small volcanic archipelago?

    Link to this
  356. 356. Grehan1896 5:18 pm 07/2/2014

    The past is full of adventures. Iguanas, finches, giant turtles… So, did all Sauropsids originating from one small volcanic archipelago?

    Very funny, but the point is that just because a group is in two places it does not automatically mean that the distribution is defined by one or other place. Various taxa in the Galapagos biota have been called ‘American’ because they have relatives on the American mainland. But this does not make the group ‘American’ (unless you want to include the Galapagos under the term ‘American’.

    John Grehan

    Link to this
  357. 357. Stripeycat 6:11 pm 07/2/2014

    John #322
    Thanks for the information re: the Galapagos hotspot. Can you remember which of the Galapagos magmas the Ecuador similarity is to, because that might tell us something about the overall history of the hotspot. More generally, can you recommend open-access sources on the subject (I’m mostly housebound, and don’t have either paid journal access or a University library in easy reach)?

    Two arcs that most pertain to the Galapagos are what are now the Greater Antilles and the various chunks embedded in Central America and NE S America

    Hmm. Isn’t that to do with the earlier phase of activity at the hotspot (before 80mya)? I’d got the impression it was dormant between then and the Miocene phase that formed the Carnegie Ridge, which is too long a time without islands in the vicinity of the hotspot for the fauna to have survived.

    One might might say we still want to know where a continent’s fauna came from.
    We do, and to a good extent we can. Plenty of groups lived in one place, crossed into a different continent when they met up, and when they separated again, iverged. Something like Eurasia-North America in the last 40my did it again and again, and we can track different groups.
    We can also have a stab at larger, older groups too: the fossil record suggests Theria developed in Laurasia, but only ancestral Metatherians and some placentals made it into S. America before the separation.
    Placentalia itself divides early on quite neatly into Boreoeutherians stuck in Laurasia, and Xenoarthrans in S. America and Afrotherians in Africa, probably dividing in a hard tritomy (certainly too close together for current molecular work to separate cleanly), which would also explain the lack of fossil data. (Sorry those are all mammal-related, but they’re off the top of my head.)

    But what makes the fauna “American”. Just because some of the fauna of the Galapagos has American relatives does not make the group “American” biogeographically speaking. The group might just as well be considered “Galapagan”. In panbiogeography at least, biota are not classified by what continent they occupy.

    Um. A lot of the American faunas evolved /in situ/, or came in from other continents, and this is fairly well recorded. Unless you’re seriously suggesting that, say, /Buteo/ evolved by convergent evolution in several places at once! Or, less outrageous but still unlikely, multiple colonisations of N. America. (If you accept that the Messel bird of prey is not a falcon, then we have accipteriforms from c.40mya in Europe; if not, there’s
    still a decent record for /Buteo/ itself, including basal forms, back to the Oligocene in N. America.)

    One alternative is to classify the fauna and flora with respect to the tectonic basins involved.

    I don’t see why that’s any more or less helpful than classifying them into areas between barriers.

    I know that means a ghost lineage for about that time given that modern kiwis have no pre-Quaternary fossil record

    New Zealand is notoriously poor in Tertiary and Quaternary fossils of anything, and kiwis live in woodland areas with poor fossilisation potential at the best of times, so the lack of fossils isn’t surprising. What’s significant in the examples I gave in my first post is that they clearly do preserve bird-like animals from a variety of niches and ecosystems, but no
    birds. That demands an explanation.

    Don’t know what you mean about the appeals to authority
    Mentioning irrelevant successes (and the kiwi example is irrelevant to the evolution of early palaeognaths) is an example of an informal fallacy that’s called either the appeal to or the argument from authority – the comparison doesn’t strengthen the argument, but is used to create a general aura of reliability around the source.

    I realize that panbiogeography is probably quite unfamiliar to you (at least that is what I infer from your assertions), but the assignment of clades to tectonic basins is not made on the basis of the phylogeny since phylogeny is biological and has no spatial characteristics.

    I understood that, at its most basic, you drew lines between related species on a map and looked for patterns (I’m aware this is a gross oversimplification of a potentially very useful analytical tool, but understood that to be the underlying concept). So you have to have some idea of what constitute related species.

    No-where (despite being repeatedly asked) have you explained why it’s reasonable to assign sympatric moas and kiwis to the Pacific and Indian Oceans respectively, unless you have prior assumptions about their relationships to other groups. I’m afraid that frustration over this did colour the rest of my previous post a little.

    Well one might argue the converse regarding dismissal of biogeography
    No. It’s a possibility that’s considered, but given a probability (at least in the paleognath case) lower than dispersal, because of other evidence (fossil and molecular, from multiple sources).

    On Occam’s razor, I should have been less sarcastic, so you’d be less inclined to dismiss the argument. Likelihood matters. The chance of crown-birds (and their less basal ancestors) existing but never being fossilised anywhere for 60MY or so is tiny. If you had a reason (other than they didn’t exist), even an unevidenced hypothesis, for the necessary preservation bias, you would strengthen your case dramatically

    Link to this
  358. 358. ectodysplasin 6:46 pm 07/2/2014

    Wait, is it just me, or is Grehan completely rejecting decades of island biogeography?

    Link to this
  359. 359. MichaelHeads 6:47 pm 07/2/2014

    Mitchell et al 2014 show elephant birds and kiwis as reciprocally monophyletic sister-groups.

    Comment 343 That is not inconsistent with elephant birds being part of the kiwi stem group. (See definition of “stem group” above.)

    The definition you gave above is:

    A stem group is the total group minus the corresponding crown group.

    There are different definitions of stem group. The usual one, based on Hennig’s ‘stammgruppe’, is that a stem group is a *paraphyletic* group composed of all the members of a group that are not in the crown group (i.e. the smallest clade with extant members). Hennig thought the stem group is ancestral to the crown group. I don’t think this is valid (a paraphyletic group isn’t necessarily the ancestor of a group included in it), but it does have a certain logic.

    You’re using ‘stem group’ in a very different sense, to refer to a *monophyletic* sister group (elephant bird) that went extinct just 300 years ago. But based on the phylogeny there is no reason to assume that the elephant bird is a ‘stem group’ ancestral to the kiwi, unless you assume that if a group is extinct it must be an ancestor of its living relatives.

    Take the example of two allopatric sister groups that evolved by normal vicariance of a widespread common ancestor. If one goes extinct, does that mean it suddenly becomes the ancestor? No, its just an extinct sister group.

    Link to this
  360. 360. Grehan1896 6:52 pm 07/2/2014

    357. Stripeycat

    Couple of immediate comments. Response to other points later.

    No-where (despite being repeatedly asked) have you explained why it’s reasonable to assign sympatric moas and kiwis to the Pacific and Indian Oceans respectively, unless you have prior assumptions about their relationships to other groups. I’m afraid that frustration over this did colour the rest of my previous post a little..

    I sometimes lose track (no pun) of every question. Sorry. With respect to moas and kiwis I did not intentionally indicate that one would assign ocean basins without having some prior assumption about their relationships to other groups. But it’s not just their relationships to other groups, but where those other groups are located with respect to major tectonic features that result in the different baselines.

    re: the Galapagos hotspot. Can you remember which of the Galapagos magmas the Ecuador similarity is to, because that might tell us something about the overall history of the hotspot. More generally, can you recommend open-access sources on the subject (I’m mostly housebound, and don’t have either paid journal access or a University library in easy reach)?

    There is a rather brief web page at http://johngrehan.net/index.php/panbiogeography/galapagos
    But on the first line you will see a link to my 2001 paper that will give you more detail, and that refers to the geochemical link.

    John Grehan

    Link to this
  361. 361. Grehan1896 10:18 pm 07/2/2014

    357. Stripeycat

    Two arcs that most pertain to the Galapagos are what are now the Greater Antilles and the various chunks embedded in Central America and NE S America
    Hmm. Isn’t that to do with the earlier phase of activity at the hotspot (before 80mya)? I’d got the impression it was dormant between then and the Miocene phase that formed the Carnegie Ridge, which is too long a time without islands in the vicinity of the hotspot for the fauna to have survived.

    I don’t know about that, but Werner et al 2003 say that islands existed continuously above the Galapagos hot spot over at least the last 17 m.y. Before that there really is not much of a record as the older islands are subducted at the plate boundary. The oldest parts of the Carnegie Ridge off the coast of Ecuador are estimated to be 20 Ma. The earlier geological record is lost. Their paper is available free at http://onlinelibrary.wiley.com/doi/10.1029/2003GC000576/pdf

    The Galapagos provides a nice example of how biogeographic patterns can preserve aspects of history that are no longer accessible to geology.

    John Grehan

    Link to this
  362. 362. DavidMarjanovic 9:36 am 07/3/2014

    Haven’t had time to catch up, won’t have time soon.

    Wait, is it just me, or is Grehan completely rejecting decades of island biogeography?

    It’s not just you.

    There are different definitions of stem group. The usual one, based on Hennig’s ‘stammgruppe’, is that a stem group is a *paraphyletic* group composed of all the members of a group that are not in the crown group (i.e. the smallest clade with extant members).

    The elephantbird clade is part of the kiwi stem group. It alone is not the entire stem group…

    a paraphyletic group isn’t necessarily the ancestor of a group included in it

    …what… …by definition, a paraphyletic group is (or contains) an ancestor of every group that descends from it; every stem group is ancestral to the corresponding crown group. That’s what “paraphyletic” means.

    The Galapagos provides a nice example of how biogeographic patterns can preserve aspects of history that are no longer accessible to geology.

    Be careful to ensure this doesn’t make them untestable.

    Link to this
  363. 363. Grehan1896 11:57 am 07/3/2014

    The Galapagos provides a nice example of how biogeographic patterns can preserve aspects of history that are no longer accessible to geology.
    Be careful to ensure this doesn’t make them untestable.

    The ‘testability’ in this case came through in the corroborated novel geological predictions. Not that that made any difference to most Galapagos people.

    John Grehan

    Link to this
  364. 364. Grehan1896 12:20 pm 07/3/2014

    358. ectodysplasin

    Wait, is it just me, or is Grehan completely rejecting decades of island biogeography?

    Yes got it right. But I would say rejecting decades of traditional island biogeography. The origin of biota on volcanic islands is not assumed to be concordant with a simple model of dispersal from the nearest mainland when it comes to understanding patterns of endemism involving volcanic islands. Biogeographic analysis suggests that the biota of an individual volcanic island needs to be considered in broader context of origin and survival. The means of dispersal of animals and plants that have supposed to demonstrate long distance dispersal from continents may be more important as means of survival that allows persistence of a group within an oceanic region when stranded on volcanic islands following the former presence of other geological landscapes (such as moving continents, basaltic plateaus, island arcs). Mike has termed this biogeographic process as a metapopulation model. An example is the Galapagos tortoise that cannot swim very well, but nevertheless its ability to float for at least some time has been regarded as evidence that it could have drifted from the American mainland. Another possibility is that the tortoise originally colonized the hotspot from a former arc (such as the Panama-Costa Rica arc [that one geologists described to me as being a huge area) and its floating ability allowed it to colonize newer Galapagos islands as the older ones eroded and subsided beneath the sea.

    John Grehan

    Link to this
  365. 365. Grehan1896 2:16 pm 07/3/2014

    357. Stripeycat Unless you’re seriously suggesting that, say, /Buteo/ evolved by convergent evolution in several places at once! Or, less outrageous but still unlikely, multiple colonisations of N. America. (If you accept that the Messel bird of prey is not a falcon, then we have accipteriforms from c.40mya in Europe; if not, there’s
    still a decent record for /Buteo/ itself, including basal forms, back to the Oligocene in N. America.)

    I cannot comment in any depth on the biogeography of Buteo that you give as an example as I have not studied this group. According to Reising et al 2003 the more basal clades are New World and the more terminal clades are Old World. Interesting (for me) is that there is an endemic Galapagos species that is sistergroup to Buteo swainsoni which has a breeding range in N America and a non breeding range in Argentina, and vagarent records over much of S America, but apparently not in the Galapagos.

    John Grehan

    Link to this
  366. 366. Yodelling Cyclist 2:20 pm 07/3/2014

    Possibly a very dumb question here, but what is the panbiogeography take on human distribution?

    Link to this
  367. 367. Grehan1896 3:06 pm 07/3/2014

    366. Yodelling Cyclist Possibly a very dumb question here, but what is the panbiogeography take on human distribution?

    I should emphasize that there is no ‘the’ panbiogeography take on anything. Panbiogeography is a research program that based on certain assumptions leads to analyses and conclusions that may are may not be agreed upon by all those who participate.

    With respect to ‘human’ do you mean extant humans? Fossil human? Origins? And what ‘take’ do you mean with respect to what? Please elaborate.

    John Grehan

    Link to this
  368. 368. Yodelling Cyclist 4:15 pm 07/3/2014

    Extant humans. How would a panbiogeographer conclude that anatomically modern humans reached their current distribution.

    Link to this
  369. 369. keesey@gmail.com 5:02 pm 07/3/2014

    Re Michael Heads’ confusion over my usage of “stem group”, David Marjanovic has it. I am using Hennig’s definition. Elephant birds are stem-kiwis but not all stem-kiwis are elephant birds. Kiwi ancestors (up to their last common ancestor with any other extant organisms) are also stem-kiwis. Anything within the kiwi total group (=the first kiwi ancestor that is not ancestral to anything else extant, and all of its descendants) but not within the kiwi crown group (=the last common ancestor of extant kiwis, and all of its descendants) is a stem-kiwi.

    I find “sister group” to be a problematic term unless you are discussing maximal clades (a.k.a., branch-based or stem-based clades, including total groups) or are very clear about it being with a certain context (i.e., a phylogenetic hypothesis). Saying elephant birds are the sister group of kiwis within the context of that molecular analysis is correct — saying they are the sister group of kiwis without specifying a context is going to prove incorrect when something else closer to kiwis is found.

    Link to this
  370. 370. keesey@gmail.com 5:05 pm 07/3/2014

    Whoops.

    “Kiwi ancestors (up to their last common ancestor with any other extant organisms)”

    …should be:

    “Kiwi ancestors (up to, but not including, their last common ancestor with any other extant organisms) “

    Link to this
  371. 371. Grehan1896 6:02 pm 07/3/2014

    368. Yodelling Cyclist
    Extant humans. How would a panbiogeographer conclude that anatomically modern humans reached their current distribution.

    This question is more about how an individual taxon moves about rather than how does differentiation take place whereby related taxa occupy different localities. I have not studied human movements so I for one would not presume any authority on the matter. In the 1950’s Croizat did theorize (based on some comparative anthropology) overwater migrations of people from central coastal Asia to America and that model has since found support among paleoanthropologists. Croizat also suggested that the first Americans did not have precisely the same phenotypic appearance as modern native Americans and that view has also found recent support. Croizat’s timing of the first arrivals was 40,000-80,000 years. At present I don’t think paleoanthropologists have proposals within that range, but some time ago there were figures of up to 30,000 years.

    Michael Heads may or may not have anything to add on how humans migrated.

    John Grehan

    Link to this
  372. 372. MichaelHeads 6:21 pm 07/3/2014

    I wrote that: a paraphyletic group isn’t necessarily the ancestor of a group included in it

    David (362) replied: …what… …by definition, a paraphyletic group is (or contains) an ancestor of every group that descends from it; every stem group is ancestral to the corresponding crown group. That’s what “paraphyletic” means.

    That was Hennig’s (1966) interpretation of paraphyly, but he was writing in the pre-vicariance era. Take a phylogeny: (a (b (c (d))), in which a, b, c and d are all allopatric and derived by simple vicariance of a widespread common ancestor. a, b, and c form a paraphyletic group. d is not derived from a-c.

    For more detail, see Fig. 1.6 in my 2014 book ‘Biogeography of Australasia’.

    The elephantbird clade is part of the kiwi stem group. It alone is not the entire stem group…

    The Mitchell et al. phylogeny of the trans-Indian Ocean clade (is there another term for it?) is: ((elephant bird + kiwi) (emu + cassowary)). There is no evidence here that the elephant bird is part of the stem group of kiwi. The only thing ‘special’ about the elephant bird is that it went extinct 300 years ago. If kiwis had gone extinct 300 years ago and elephant birds were still extant, would you say that kiwis were part of the stem group of elephant birds?

    Darren, I’m still very keen to see the references to nocturnal cassowaries that you mentioned. I’d also be interested in any other references to nocturnality in non-kiwi ratites.

    Link to this
  373. 373. Grehan1896 6:25 pm 07/3/2014

    369. keesey@gmail.com
    I find “sister group” to be a problematic term unless you are discussing maximal clades (a.k.a., branch-based or stem-based clades, including total groups) or are very clear about it being with a certain context (i.e., a phylogenetic hypothesis). Saying elephant birds are the sister group of kiwis within the context of that molecular analysis is correct — saying they are the sister group of kiwis without specifying a context is going to prove incorrect when something else closer to kiwis is found.

    I wonder if problematic depends on what one is used to. I find the term ‘stem group’ confusing, but then it’s not a term I have used. When analyzing hominid (anything more closely related to humans than any other primate) evolution I only referred to sister taxa.

    I would agree that context is important for assigning sister taxa – one need to refer to the supporting analysis. But finding something else closer to kiwis does not necessarily make the sister grout relationship incorrect. True, the new taxon would be the sistergroup to ‘ kiwis’, but the elephant birds would reamin the sistergroup of ‘new taxon + kiwis’, or if the term kiwis were applied to the ‘new taxon’ then nothing would need to change in the sistergroup statement.

    John Grehan

    Link to this
  374. 374. irenedelse 7:41 pm 07/3/2014

    @John Grehan:

    “In the 1950’s Croizat did theorize (based on some comparative anthropology) overwater migrations of people from central coastal Asia to America and that model has since found support among paleoanthropologists. Croizat also suggested that the first Americans did not have precisely the same phenotypic appearance as modern native Americans and that view has also found recent support. Croizat’s timing of the first arrivals was 40,000-80,000 years. At present I don’t think paleoanthropologists have proposals within that range, but some time ago there were figures of up to 30,000 years.”

    Some paleoanthropologists have proposed this kind of scenario, based on a very few, very tentative bits of data. (bits of burnt wood in a cave in Brazil, carbon-dated 40.000 years, but which are probably of natural origin…).

    Most in paleoanthropology go with what most data shows: that the earliest American settlers came from Northeast Asia, either on foot through Beringia or maybe following the coast on canoes, sometimes around 15.000 before present. The genetic data is consistent with the Northeast Asian origin, there’s no uncontroversial human remains or artifacts before that date, and the distribution of human associated species (dogs, chicken, etc.) also follows that pattern. It’s also very dubious that Paleolithic people had the technology to sail through the ocean, as Polynesians did more recently… Humans don’t fly, can’t swim for days, are less buoyant than hippos and tortoises, and they are a bit too big to ride a vegetation raft like lemurians.

    There was until recently a contention that European explorers in the 16th century had seen chickens in Central America in places which had had no previous contact with them, and that it meant Polynesians had arrived there first, but it rested on the interpretation of unclear accounts. Thanks to Thompson et al., 2014, who did a DNA analysis on chickens, we know that it didn’t happen. Still no evidence of trans-Pacific contact with America, either in the Paleolithic or more recently, until the Early Modern period. (With nod to Darren here, who mentioned the chicken DNA paper in the TetZoo podcast!)

    By the way, there are of course a few minority opinions in anthropology, including decidedly fringe ideas, like an incursion of Homo erectus in the Americas in the Middle Pleistocene, and their hybridization with later modern humans. A holdover from the polycentric theory of human evolution, probably, but I know at least one otherwise respected French paleoantheopologist who was still defending this extremely early settlement of the continent during the 1990s. Not that his colleagues were convinced. I remember the bemused smiles…

    Conclusion: it’s always tricky to use works from an area of research you’re not familiar with, even (especially?) when they seem to support one of your hypotheses.

    Link to this
  375. 375. Grehan1896 8:39 pm 07/3/2014

    Some paleoanthropologists have proposed this kind of scenario, based on a very few, very tentative bits of data. (bits of burnt wood in a cave in Brazil, carbon-dated 40.000 years, but which are probably of natural origin…).

    Perhaps so.

    Most in paleoanthropology go with what most data shows: that the earliest American settlers came from Northeast Asia, either on foot through Beringia or maybe following the coast on canoes, sometimes around 15.000 before present.

    I think it’s a bit open ended as to what a ‘coastal’ route really means as I don’t think anyone has any evidence that points to exactly what kind of water travel was involved.

    The genetic data is consistent with the Northeast Asian origin, there’s no uncontroversial human remains or artifacts before that date, and the distribution of human associated species (dogs, chicken, etc.) also follows that pattern.
    Perhaps so.

    It’s also very dubious that Paleolithic people had the technology to sail through the ocean, as Polynesians did more recently… Humans don’t fly, can’t swim for days, are less buoyant than hippos and tortoises, and they are a bit too big to ride a vegetation raft like lemurians.

    Evidently melanesians got around to Australia 40,0000 years ago. As for ‘sailing through the ocean’ – did I say that?

    There was until recently a contention that European explorers in the 16th century had seen chickens in Central America in places which had had no previous contact with them, and that it meant Polynesians had arrived there first, but it rested on the interpretation of unclear accounts. Thanks to Thompson et al., 2014, who did a DNA analysis on chickens, we know that it didn’t happen. Still no evidence of trans-Pacific contact with America, either in the Paleolithic or more recently, until the Early Modern period. (With nod to Darren here, who mentioned the chicken DNA paper in the TetZoo podcast!)

    I made no specific suggestions of this kind.

    Conclusion: it’s always tricky to use works from an area of research you’re not familiar with, even (especially?) when they seem to support one of your hypotheses.

    But I was not doing that.

    John Grehan

    Link to this
  376. 376. irenedelse 10:02 pm 07/3/2014

    @John Grehan:

    “I think it’s a bit open ended as to what a ‘coastal’ route really means”

    Well, going across the Pacific is not it. ;-)

    Of course this is once again a question of parsimony. The less assumptions, etc. If your position is that anything that is not totally impossible is equally valid as an explanation…

    And there are no “perhaps” about the multiple lines of data linking the first settlement of America with Northeast Asia: Genetics, archaeology, geology, history of climate (the presence of the Bering land bridge and the ecological continuity between Siberia and Alaska during the last glaciation): it really would be too long to give all the references! Croizat’s hypothesis may have been reasonable in the 1950s but there’s been 60 years of intense research since.

    “I made no specific suggestions of this kind.”

    See, I was not attacking you, but using your remarks as a starting point to discuss a few points about paleoanthropology and the population of the Americas. You mentioned a hypothesis about dispersion over the water from the “central” part of the Asian coast. This would involve very long distance sailing, or rowing, or drifting. Not impossible, but not as feasible as walking or following the coast in boats. Yes, Melanesians colonized Australia in the timeframe you mention, that’s not controversial, but then, they didn’t have as much water to cross.

    Link to this
  377. 377. irenedelse 10:04 pm 07/3/2014

    Oh, darn. Italics attack. Can we close the tags please?

    Let’s hope this works…

    Link to this
  378. 378. Grehan1896 10:18 pm 07/3/2014

    • “I think it’s a bit open ended as to what a ‘coastal’ route really means”

    Well, going across the Pacific is not it.
    How far from the coast is going across the Pacific?

    Of course this is once again a question of parsimony. The less assumptions, etc. If your position is that anything that is not totally impossible is equally valid as an explanation…

    And if that is not my position?

    And there are no “perhaps” about the multiple lines of data linking the first settlement of America with Northeast Asia: Genetics, archaeology, geology, history of climate (the presence of the Bering land bridge and the ecological continuity between Siberia and Alaska during the last glaciation): it really would be too long to give all the references! Croizat’s hypothesis may have been reasonable in the 1950s but there’s been 60 years of intense research since.

    They once proved that we had 48 chromosomes as well.

    “I made no specific suggestions of this kind.”
    See, I was not attacking you, but using your remarks as a starting point to discuss a few points about paleoanthropology and the population of the Americas. You mentioned a hypothesis about dispersion over the water from the “central” part of the Asian coast. This would involve very long distance sailing, or rowing, or drifting.

    OK. Just to be clear about that.
    Not impossible, but not as feasible as walking or following the coast in boats.

    Perhaps.

    Yes, Melanesians colonized Australia in the timeframe you mention, that’s not controversial, but then, they didn’t have as much water to cross.
    If the first Americans followed the coast then neither did they. Point is that the technology was there.

    John Grehan

    Link to this
  379. 379. LeeB 1 11:06 pm 07/3/2014

    Even if polynesians didn’t take chickens to the Americas in pre-colombian times it doesn’t mean they didn’t visit there.
    They had to get Kumara (sweet potatoes) from somewhere.

    LeeB.

    Link to this
  380. 380. Heteromeles 1:46 am 07/4/2014

    The means of dispersal of animals and plants that have supposed to demonstrate long distance dispersal from continents may be more important as means of survival that allows persistence of a group within an oceanic region when stranded on volcanic islands following the former presence of other geological landscapes (such as moving continents, basaltic plateaus, island arcs). Mike has termed this biogeographic process as a metapopulation model.

    HUNH???

    I don’t get it. I mean, when Sherwin Carlquist early in the 1970s dumped seeds in sea water in a lab and let them sit for months, just to see how prolonged immersion affected viability and hence transport distance, this is supposed to have some relationship with how insular plants survive? And didn’t Darwin do something similar if memory serves?

    When they were doing those aerobiology studies back in the 1960s, and finding fern spores a few miles up in the jet stream, and using this as evidence for how fern spores got to islands (along with spiders, incidentally), this is supposed to have some relationship to how insular plants survive?

    And a metapopulation model…? I don’t think that term means what you think it means. Islands don’t work as true metapopulations, especially when you have different, closely related species (as with the Hawaiian tarplants) on different islands. That’s interisland dispersal, not a metapopulation of one huge species that moves continuously from island to island as populations go extinct on each island.

    So what are you trying to imply with that first statement? Can you be more explicit about what you think is actually going on with experiments such as the ones I just mentioned that has nothing to do with dispersal and everything to do with survival on an island?

    Link to this
  381. 381. DavidMarjanovic 4:09 am 07/4/2014

    “And there are no “perhaps” about the multiple lines of data linking the first settlement of America with Northeast Asia: Genetics, archaeology, geology, history of climate (the presence of the Bering land bridge and the ecological continuity between Siberia and Alaska during the last glaciation): it really would be too long to give all the references! Croizat’s hypothesis may have been reasonable in the 1950s but there’s been 60 years of intense research since.”

    They once proved that we had 48 chromosomes as well.

    That’s the stupidest thing I’ve read since Ken Ham’s latest proclamation of Young-Earth Creationism. In fact, the “argument” from “science has been wrong before, therefore it can all be safely ignored” is one of YEC’s favorites.

    Aren’t you ashamed?

    Link to this
  382. 382. DavidMarjanovic 4:48 am 07/4/2014

    Another possibility is that the tortoise originally colonized the hotspot from a former arc (such as the Panama-Costa Rica arc [that one geologists described to me as being a huge area) and its floating ability allowed it to colonize newer Galapagos islands as the older ones eroded and subsided beneath the sea.

    How did it get to the arc in the first place?

    That was Hennig’s (1966) interpretation of paraphyly, but he was writing in the pre-vicariance era. Take a phylogeny: (a (b (c (d))), in which a, b, c and d are all allopatric and derived by simple vicariance of a widespread common ancestor. a, b, and c form a paraphyletic group. d is not derived from a-c.

    What. Did you manage to confuse para- and polyphyly?

    d is not derived from a, b or c.

    Any group that contains a, b and c and their last common ancestor, but doesn’t contain d, is paraphyletic with respect to d.

    A group that contains a, b and c but not their last common ancestor or the last common ancestors of b and c or c and d is not paraphyletic with respect to d – it’s polyphyletic, having (at least) three independent origins.

    Neither Hennig nor vicariance have anything to do with that. What made you think they do?

    If kiwis had gone extinct 300 years ago and elephant birds were still extant, would you say that kiwis were part of the stem group of elephant birds?

    Yes! That’s how the term is defined!

    Humans don’t fly, can’t swim for days, are less buoyant than hippos and tortoises

    As I mentioned way upthread, we’re a lot more buoyant than adult hippos. They’re so osteosclerotic they can’t swim – they walk on the bottom, and jump up or walk out to breathe.

    “If your position is that anything that is not totally impossible is equally valid as an explanation…”

    And if that is not my position?

    What is your position? See, so far you’ve behaved as if it were your position. Explain what we’ve misinterpreted.

    Link to this
  383. 383. Grehan1896 10:11 am 07/4/2014

    Mike has termed this biogeographic process as a metapopulation model.

    HUNH???

    Yes indeed. Please keep in mind that I am referring to the evolution of vicariant (allopatric) differentiated taxa.

    I don’t get it. I mean, when Sherwin Carlquist early in the 1970s dumped seeds in sea water in a lab and let them sit for months, just to see how prolonged immersion affected viability and hence transport distance, this is supposed to have some relationship with how insular plants survive? And didn’t Darwin do something similar if memory serves?

    Yes to both. The amount of time was supposed to be a predictor of transport distance, but such theorized abilities appear not to be the key to the evolution of vicariant differentiation (allopatric differentiation if you prefer) either on continents or oceans.

    When they were doing those aerobiology studies back in the 1960s, and finding fern spores a few miles up in the jet stream, and using this as evidence for how fern spores got to islands (along with spiders, incidentally), this is supposed to have some relationship to how insular plants survive?

    Yes, when dealing with vicariant differentiated forms. Certainly there are some world wide species that are everywhere and their mode of dispersal allows this. But other forms with the same means are not everywhere, and not only not everywhere, but may be vicariant with respect to related forms in a way that is not predicted by their good means of dispersal.

    And a metapopulation model…? I don’t think that term means what you think it means. Islands don’t work as true metapopulations, especially when you have different, closely related species (as with the Hawaiian tarplants) on different islands. That’s interisland dispersal, not a metapopulation of one huge species that moves continuously from island to island as populations go extinct on each island.

    You are referring to a purely ecological concept of metapopulation. The biogeographic metapopulatioin model involves differentiation as well – that a taxon may persist through the ability of species to disperse locally and colonize new habitats (e.g. island or continental landscapes). As some some species become extinct, others evolve on the new landscapes and so the taxon persists in the area.

    So what are you trying to imply with that first statement? Can you be more explicit about what you think is actually going on with experiments such as the ones I just mentioned that has nothing to do with dispersal and everything to do with survival on an island?
    But it’s NOT survival on an island that is the focus of the metapopulation model, but survival among islands that is made possible through their means of dispersal. Thus, the giant tortoises of Galapagos may have survived at the Galapagos by their ability to move between islands (a necessity if there were no terrestrial contiguity between the present and former islands).
    Since you find this concept new and challenging it might help you to read some of its applications. A couple of examples are:

    Heads 2009. Globally basal centres of endemism: the Tasman-Coral Sea region (south-west Pacific), Latin America and Madagascar/South Africa.

    Heads 2008 Panbiogeography of New Caledonia, south-west Pacific: basal angiosperms on basement terranes, ultramafic endemics inherited from volcanic island arcs and old taxa endemic to young islands.

    They can be accessed at johngrehan.net/index.php/panbiogeography/panpublications/michael-heads-publications

    John Grehan

    Link to this
  384. 384. Grehan1896 10:29 am 07/4/2014

    Another possibility is that the tortoise originally colonized the hotspot from a former arc (such as the Panama-Costa Rica arc [that one geologists described to me as being a huge area) and its floating ability allowed it to colonize newer Galapagos islands as the older ones eroded and subsided beneath the sea.

    How did it get to the arc in the first place?

    How did tortoises in general get to be where they are? The origin of the tortoises on the arc may be the same as the origin of their relatives elsewhere.

    That was Hennig’s (1966) interpretation of paraphyly, but he was writing in the pre-vicariance era. Take a phylogeny: (a (b (c (d))), in which a, b, c and d are all allopatric and derived by simple vicariance of a widespread common ancestor. a, b, and c form a paraphyletic group. d is not derived from a-c.
    What. Did you manage to confuse para- and polyphyly?

    Mike may respond later (from a different time zone), but his statement is correct. Grouping a,b,c together would indeed form a paraphyletic group. In a vicariant origin (and assuming that the phylogenetic sequence represents the differentiation sequence), of all four taxa d is most recently derived from a common ancestor with c.

    d is not derived from a, b or c.
    Any group that contains a, b and c and their last common ancestor, but doesn’t contain d, is paraphyletic with respect to d.

    If one has just these four taxa to hand the statement could be said as Any group that contains a, b and c , but doesn’t contain d, is paraphyletic with respect to d

    A group that contains a, b and c but not their last common ancestor or the last common ancestors of b and c or c and d isnot paraphyletic with respect to d – it’s polyphyletic, having (at least) three independent origins.

    But this was not implied by the original phylogenetic statement with the sequence (a (b (c (d))).

    If kiwis had gone extinct 300 years ago and elephant birds were still extant, would you say that kiwis were part of the stem group of elephant birds?

    Yes! That’s how the term is defined!

    So basically a stem group is the extinct sister group?

    “If your position is that anything that is not totally impossible is equally valid as an explanation…”
    And if that is not my position?
    What is your position? See, so far you’ve behaved as if it were your position. Explain what we’ve misinterpreted.

    My position is that anything that is not totally impossible remains a possibility. Of course I have my preference was to what might constitute an “equally valid” explanation, as does everyone else.

    John Grehan

    Link to this
  385. 385. Grehan1896 10:34 am 07/4/2014

    They once proved that we had 48 chromosomes as well.
    That’s the stupidest thing I’ve read since Ken Ham’s latest proclamation of Young-Earth Creationism.

    Well of course I respect your right to have an opinion, even if I might not respect the opinion!

    “In fact, the “argument” from “science has been wrong before, therefore it can all be safely ignored” is one of YEC’s favorites.”

    Just because creationists might use the same information to support their view does not make that information necessarily wrong. I think orangutans are our closest living relatives. Creationists have (earlier) jumped on the similarities between humans and orangutans to discredit evolution, but just because they have does not affect the scientific validity of the proposed relationship between humans and orangutans.

    Aren’t you ashamed?

    Ashamed of what? Be precise please.

    John Grehan

    Link to this
  386. 386. Heteromeles 10:57 am 07/4/2014

    While we’re talking about human dispersal to islands, I should note that the archeology of human dispersal from Indonesia to Polynesia (and Micronesia) is fairly well documented.

    One critical point is timing.

    Yes, people did make it across Wallace’s gap about 40,000 years ago. That distance is on order 10-20 miles, and it could have been done with a primitive raft without sails. These people then settled every place that could be easily rafted to or walked to, all the way down to Tasmania. They may have explored out to about Vanuatu, but there is no evidence of permanent settlement during this period, although there is evidence that they tried and failed to colonize near islands like New Britain. The flora and fauna of the Solomons, all the way east to New Britain, is insufficient to sustain a human population. Fishermen could make their way along the coast, but they’d have no plants to eat. In that region, Papua New Guinea is the only place where hunter-gatherers could successfully colonize.

    The modern Austronesian speakers showed up about 4,000 years ago (note the difference: 40,000 vs. 4,000 years), and their descendants colonized the deep Pacific as the Polynesians and Micronesians. The Lapita culture which did colonize the remote Pacific islands (and from which the Polynesians descended), showed up around 3700 years ago (~1700 BCE). At that point, Austronesian-speaking people had acquired both the technology they needed for making deep sea sailing vessels and for fishing on remote islands, plus the domestic species (coconut, taro, pigs, chickens, etc–there’s dozens of “canoe people” species) they needed to terraform the atolls and deep islands into places that people could live. We may think of Pacific islands as paradises now, but they are man-made paradises. Absent cultivated plants and domestic animals, they were all unsuitable for settlement by hunter-gatherers.

    Make no mistake: no one in the Pacific had the technology to sail across the deep Pacific before 1700 BCE, and arguably no one sailed across the Pacific until somewhere around 1,000 CE, give or take a few centuries, when the Polynesians made contact on the South American coast and got their hands on sweet potato. Sweet potato then spread back across the Pacific and made its way to Papua New Guinea somewhere around 1400-1500 CE.

    As for rumors (for example, those based on plank-built tomol canoes used by the California Chumash) that the Polynesians made it to California 1,000 years ago and passed on their boat-building technology, the oldest tomol in the archeological record is 2,000 years old, so we can pretty safely say that the chumash or their predecessors developed the technology independently.

    Note that I’m not saying that paleoindians didn’t colonize North America via sea. If they did, they used what some call the “kelp road” by sailing along the coast, eating the abundant marine life that historically characterized the Pacific coast all the way from Asia to Mexico and points south. There’s absolutely no archeological evidence that anyone could have sailed a raft into the wind across the Pacific. No one could do that now, for that matter, because rafts are notoriously bad at sailing into the wind. It took people about 36,000 years to figure out how to make boats that could sail into the wind and thereby colonize the Pacific, and we still don’t give them the credit they deserve.

    Link to this
  387. 387. Grehan1896 11:00 am 07/4/2014

    379. LeeB 1
    Even if polynesians didn’t take chickens to the Americas in pre-colombian times it doesn’t mean they didn’t visit there. They had to get Kumara (sweet potatoes) from somewhere. LeeB

    I had wondered about that also, but I did not mention it as I have not read the literature on that subject. I suppose it might be argued that the kumara just dispersed over the Pacific by itself.

    The model proposed by Croizat for overwater movements of the first Americans did not address whether or not there was any reciprocal movement.

    With respect to some other observations posted about Croizat’s theory I would summarize that some of his predications have been better corroborated that others:
    His idea that the first Americans did not look so much like modern native Americans has received some corroboration.

    His view that the migrations were overwater is compatible with the ‘coastal’ hypothesis, but certainly Croizat’s model is not limited to coastal, but how much more out from the coast these peoples may have moved is at present, unsolvable I think.

    His view about the general locality of Asian origin is probably more complex, and I agree that it may not conform to the currently accepted interpretation of the evidence. Some may consider this a falsification of Croizat’s model and that may be the case.

    On his timescale, as pointed out, it considers an age of origin that predates the currently oldest models generated from palaeoanthropology, and again this may be a falsification or just the absence of corroboration.

    The origin of the first Americans has definitely captured popular imagination and it intrigues many researchers. It will be interesting to see what new insights do or do not emerge in the future.

    John Grehan

    Link to this
  388. 388. naishd 11:11 am 07/4/2014

    Hmm… html tags..

    Link to this
  389. 389. keesey@gmail.com 11:17 am 07/4/2014

    “If kiwis had gone extinct 300 years ago and elephant birds were still extant, would you say that kiwis were part of the stem group of elephant birds?”

    Yes, now you’re getting it.

    See also the link I posted above to a huge post I wrote on the topic.

    Link to this
  390. 390. keesey@gmail.com 11:24 am 07/4/2014

    “But finding something else closer to kiwis does not necessarily make the sister grout relationship incorrect. True, the new taxon would be the sistergroup to ‘ kiwis’, but the elephant birds would reamin the sistergroup of ‘new taxon + kiwis’”

    That’s the problem exactly. It does make calling elephant birds “the sister group to kiwis” incorrect. (Again, unless you’re specifying a context where there is a specific number of taxonomic units.)

    “or if the term kiwis were applied to the ‘new taxon’ then nothing would need to change in the sistergroup statement.”

    This is destabilizing and can lead to unwarranted inferences. I follow the philosophy that terms from neontology are best restricted to crown groups.

    Link to this
  391. 391. naishd 11:27 am 07/4/2014

    I’ll tidy up the html later, sorry it makes a mess of things (trouble starts with comment # 383).

    Link to this
  392. 392. keesey@gmail.com 11:29 am 07/4/2014

    “So basically a stem group is the extinct sister group?”

    No (although that would be part of a stem group). “Stem group” is defined above.

    Link to this
  393. 393. keesey@gmail.com 12:26 pm 07/4/2014

    “When analyzing hominid (anything more closely related to humans than any other primate) evolution I only referred to sister taxa.”

    (I assume you mean *living* primate.) This is increasingly commonly called Hominini now (vernacular form: hominins), with Hominidae shifted to the great ape clade (including humans), but there are still plenty of researchers who use it the “traditional” way as well. (The major impetus, I think, is the finding that traditional “Pongidae” is paraphyletic.)

    I’m writing something on human evolution and I use “stem-humans” all the time. And there seems to be a lot of reticulation within the group (a million-year long divergence with chimpanzees, extensive interbreeding in Middle-Upper Pleistocene Homo species, etc.), so “sister group” has more limited usefulness. (Except, again, when discussing a particular hypothesis generated by a cladistic analysis, where it’s quite useful. Or when discussing extant taxa only.)

    Link to this
  394. 394. Grehan1896 12:35 pm 07/4/2014

    • “When analyzing hominid (anything more closely related to humans than any other primate) evolution I only referred to sister taxa.”

    (I assume you mean *living* primate.)

    Yes.

    This is increasingly commonly called Hominini now (vernacular form: hominins), with Hominidae shifted to the great ape clade (including humans), but there are still plenty of researchers who use it the “traditional” way as well. (The major impetus, I think, is the finding that traditional “Pongidae” is paraphyletic.)

    Yes, this use of taxonomic labels is based on the chimpanzee theory of relationship.

    I’m writing something on human evolution and I use “stem-humans” all the time.

    And I have no objection to that. I happen to have a different preference.

    And there seems to be a lot of reticulation within the group (a million-year long divergence with chimpanzees, extensive interbreeding in Middle-Upper Pleistocene Homo species, etc.), so “sister group” has more limited usefulness. (Except, again, when discussing a particular hypothesis generated by a cladistic analysis, where it’s quite useful. Or when discussing extant taxa only.)

    I have no firm opinions to the contrary.

    John Grehan

    Link to this
  395. 395. irenedelse 12:48 pm 07/4/2014

    @John Grehan:

    “How far from the coast is going across the Pacific?”

    Hum. Are you serious? And this isn’t a rhetorical question.

    “”Of course this is once again a question of parsimony. The less assumptions, etc. If your position is that anything that is not totally impossible is equally valid as an explanation…”

    And if that is not my position?”

    Then please, do clarify your position. Because from what you wrote previously, it looks exactly like what you’re doing. Trying to hold on to Croizat’s hypothesis, no matter what light more recent research has shone on the question, leads you to defend a less parsimonious scenario, in the case of humans as previously for Ratites/Paleognaths.

    Unfortunately for panbiogeography, the shorter path between two landmasses is not necessarily the straight line. You were on sounder ground yesterday, when you wrote that one needed to take into consideration the means of dispersal available to a given species!

    Oh, and please, next time you use italics, remember the / in the last tag…

    Link to this
  396. 396. keesey@gmail.com 1:05 pm 07/4/2014

    Just noticed this:

    “For example, the morphogenetic evidence for a sister group relationship between humans/fossil hominids and orangutans is not accepted as the better explanation by the majority of ‘qualified people’ (not sure who is qualified or not) because most in the field believe that sequence similarity is the absolute proof of a closer relationship between humans and chimpanzees.”

    David Marjanovic already addressed this. I’ll add a reference for a soft-tissue analysis that also supports the usual (Pongo, (Gorilla, (Pan, Homo))) topology:

    http://www.ncbi.nlm.nih.gov/pubmed/21689100

    And I’ll also point at that it’s not just “similarity”, it’s shared derived states. (They happen to be similar as well.)

    Finally I’ll ask: if molecular cladistic analysis isn’t good enough for hominids, why is it good enough for palaeognathes?

    Link to this
  397. 397. keesey@gmail.com 1:10 pm 07/4/2014

    Correction to something I wrote above:

    “Chuniaoae (also includes Oviraptorosauria)”

    Looking back at the papers that used the name “Chuniaoae”, it appears it is not defined and ambiguously used. A paper that came out this week named the minimal oviraptorosaur-paravian clade “Pennaraptora”.

    Link to this
  398. 398. irenedelse 1:12 pm 07/4/2014

    Oh, and by the way…

    @Grehan:

    “”Yes, Melanesians colonized Australia in the timeframe you mention, that’s not controversial, but then, they didn’t have as much water to cross.”
    If the first Americans followed the coast then neither did they. Point is that the technology was there.”

    Don’t forget a big difference with the case of Melanesians: other independent lines of data point toward a more recent date for the settlement of North America… Data that Croizat didn’t have when he proposed his scenario.

    Link to this
  399. 399. irenedelse 1:44 pm 07/4/2014

    @LeeB:

    “Even if polynesians didn’t take chickens to the Americas in pre-colombian times it doesn’t mean they didn’t visit there.
    They had to get Kumara (sweet potatoes) from somewhere.”

    That’s a good point. Occasional contact of Polynesian people with South America probably happened around 1000 CE (date of the earliest remains of cultivated Ipomoea batata, in the Cook islands) or a little earlier. By that time, of course, they had sophisticated navigation technology and a lot of experience of ocean currents. That’s a very different proposition than a Paleolithic cross-Pacific voyage. ;-)

    Link to this
  400. 400. Grehan1896 1:48 pm 07/4/2014

    • 396. keesey@gmail.com
    • Just noticed this:
    I’ll add a reference for a soft-tissue analysis that also supports the usual (Pongo, (Gorilla, (Pan, Homo))) topology:
    http://www.ncbi.nlm.nih.gov/pubmed/21689100

    How does this analysis support the usual topology? What characters do you regard as clastically valid in this respect?

    And I’ll also point at that it’s not just “similarity”, it’s shared derived states
    . (They happen to be similar as well.)

    Of course, how silly of me not to say so.

    Finally I’ll ask: if molecular cladistic analysis isn’t good enough for hominids, why is it good enough for palaeognathes?

    Maybe molecular analysis is good enough for hominids. Problem is that it is not corroborated by the morphogenetic evidence that I have been able to substantiate. Some cladistic theorists take the view that molecular analysis is not really cladistic, but phonetics dressed up in cladistic terminology and techniques. No doubt you and others would disagree and that’s OK. I’m just making an observation here.

    Same for palaeognathes. Maybe the molecular evidence is good enough. Worthy did get a different result with mophogenetics. Maybe it is more accurate, maybe not. But I have not made a study of ratite morphogenetics (as I did for hominoids) and I do not intend to (I do not have the resources to do so). I am happy to look at the biogeographic implications of ratite molecular characters.

    John Grehan

    Link to this
  401. 401. Grehan1896 1:52 pm 07/4/2014

    • 395. Irenedelse • Unfortunately for panbiogeography, the shorter path between two landmasses is not necessarily the straight line. You were on sounder ground yesterday, when you wrote that one needed to take into consideration the means of dispersal available to a given species!

    If I did, I should not have for biogeographic analysis. Please give the quote so I can assess. Thanks.

    John Grehan

    Link to this
  402. 402. Grehan1896 1:54 pm 07/4/2014

    Sorry, the italic off command is not working or I am doing something wrong again .

    John Grehan

    Link to this
  403. 403. irenedelse 2:08 pm 07/4/2014

    @David:

    Thanks for the correction on hippos. I had read your previous comment but had clean forgotten it. To be honest I was only half serious with that comparison.

    (Oh, and we’re at 400+ comments. Great.)

    Link to this
  404. 404. irenedelse 2:41 pm 07/4/2014

    @Grehan #400:

    See comment #364…

    The means of dispersal of animals and plants that have supposed to demonstrate long distance dispersal from continents may be more important as means of survival that allows persistence of a group within an oceanic region when stranded on volcanic islands following the former presence of other geological landscapes (such as moving continents, basaltic plateaus, island arcs).”

    I realize that you were only discussing means of dispersal in the context of redefining what a metapopulation is… But the coincidence was fascinating. You have a tool at your disposal here (the physical and ethological characteristics of the species analyzed, the features of the areas where it occurs, etc.), and you don’t want to use it because…?

    Is panbiogeography just lines on paper then? Where’s the “bio” part? Or even the geography. To discuss over the water dispersal without mention of sea currents, for instance, or vicariance without consideration of ecosystems… Sigh.

    Link to this
  405. 405. Grehan1896 2:50 pm 07/4/2014

    You have a tool at your disposal here (the physical and ethological characteristics of the species analyzed, the features of the areas where it occurs, etc.), and you don’t want to use it because…?

    Because it is not an effective analytical took for the analysis of distribution patterns. Pattern analysis shows that there is no correlation between extent or orientation in the distribution of differentiated taxa. This analysis leads to a different understanding of how dispersal works.

    Is panbiogeography just lines on paper then? Where’s the “bio” part? Or even the geography.

    The bio part is the organisms, the geo part is the location and spatial relationship.

    To discuss over the water dispersal without mention of sea currents, for instance, or vicariance without consideration of ecosystems… Sigh.

    One can do that. And people have done that for decades and made assumptions about such factors being responsible for various plant and animal distributions. If you are interested in the relationship between means of dispersal, ecology, and evolution you might want to take a look at Mike’s monograph on Fiji.

    John Grehan

    Link to this
  406. 406. irenedelse 2:55 pm 07/4/2014

    Tl;dr of my latest comment: basically, what Heteromeles said a lot better at #386!

    Link to this
  407. 407. Heteromeles 3:17 pm 07/4/2014

    Perhaps now would be a bad time to bring up the bipolar asexual disjuncts that a friend of mine once wanted to study. I’ve forgotten which lichen it is, but it’s found in Greenland, Antarctica, and a couple of high points in between. It’s asexual so far as is known, and quite rare.

    I guess the existence of such amphipolar disjuncts demonstrates to panbiogeographers that Greenland and Antarctica were in direct contact with each other in the not too distant past? I mean, the conventional lichen routes of dispersal (bird feathers and high altitude winds) are things they seem to regard as inadequate for explaining how one species can be at both poles (I guess it’s one global metapopulation, using some definition I don’t get?).

    Link to this
  408. 408. irenedelse 3:23 pm 07/4/2014

    @Heteromeles:

    Global metapopulation, or… Global organism?!! We may be looking at Solaris here. ^__^

    Link to this
  409. 409. Grehan1896 3:35 pm 07/4/2014

    • 407. Heteromeles
    • Perhaps now would be a bad time to bring up the bipolar asexual disjuncts that a friend of mine once wanted to study. I’ve forgotten which lichen it is, but it’s found in Greenland, Antarctica, and a couple of high points in between. It’s asexual so far as is known, and quite rare.
    I guess the existence of such amphipolar disjuncts demonstrates to panbiogeographers that Greenland and Antarctica were in direct contact with each other in the not too distant past?

    Instead of guessing, why not read up on the subject?

    I mean, the conventional lichen routes of dispersal (bird feathers and high altitude winds) are things they seem to regard as inadequate for explaining how one species can be at both poles (I guess it’s one global metapopulation, using some definition I don’t get?).

    This assertion ignores the fact that I have said I am referring to the origin of vicariant differentiated entities. Within particular entities such as species, some have a wider and some a narrower range. It would be interesting to know what the high points in between are. And there are other taxa with ‘bipolar’ (northern and southern disjuncts) distributions.

    John Grehan

    Link to this
  410. 410. Heteromeles 4:41 pm 07/4/2014

    Arctic terns, perhaps? Those have an amphipolar distribution.

    So far as anyone knows, the lichen species (singular) I’m talking about occurs on a few mountain tops in between the poles, on several continents. If one believed in long-distance dispersal high in the atmosphere (as studied in the field of aerobiology), then this distribution would make perfect sense. If only vicariance is allowable, then such a distribution makes no sense whatsoever.

    Link to this
  411. 411. MichaelHeads 4:48 pm 07/4/2014

    Getting back to ratites… Keesey (337)suggested that elephant birds are stem-kiwis. David (362) suggested that ‘the elephantbird clade is part of the kiwi stem group. It alone is not the entire stem group…’.

    The phylogeny that Mitchell et al presented provides no evidence that elephant birds are ancestral to kiwis, or part of the ancestor of kiwis. They are an extinct sister group

    I wrote: Take a phylogeny: (a (b (c (d))), in which a, b, c and d are all allopatric and derived by simple vicariance of a widespread common ancestor. a, b, and c form a paraphyletic group. d is not derived from a-c.

    David didn’t accept this, so I’ll try another approach. Why do you think Mitchell et al. titled their paper ‘Ancient DNA reveals elephant birds and kiwi are sister taxa…’ (not ancestors and descendants)? The authors also referred in the text to the two being sister taxa. If there was evidence that elephant birds were indeed ancestral to kiwis or part of the kiwi ancestor (i.e. part of the stem group), that would be very exciting, and the authors would have referred to it (and probably put it in the title). But they didn’t, and they didn’t mention the term ‘stem’. Why not? Were the authors (and the editors and reviewers at Science) mistaken?

    I’m still interested in the nocturnal cassowaries… Darren, could you let us know which journal your in-press cassowary paper will appear in, so I can make sure I don’t miss it when it comes out?

    Link to this
  412. 412. Grehan1896 4:50 pm 07/4/2014

    Heteromeles 410

    So far as anyone knows, the lichen species (singular) I’m talking about occurs on a few mountain tops in between the poles, on several continents. If one believed in long-distance dispersal high in the atmosphere (as studied in the field of aerobiology), then this distribution would make perfect sense. If only vicariance is allowable, then such a distribution makes no sense whatsoever.

    It could make sense either way.

    Link to this
  413. 413. ectodysplasin 5:07 pm 07/4/2014

    Stem-group =/= ancestors

    Stem-group includes all extinct species more closely related to one crown-group than another. Stegosaurus is a stem-bird, but no one would argue that Stegosaurus is ancestral to birds.

    Similarly, elephant birds are part of the kiwi stem group, because elephant birds are an extinct sister-taxon to kiwis. No one is claiming an ancestor-descendant relationship between these taxa.

    Link to this
  414. 414. irenedelse 5:55 pm 07/4/2014

    @Heteromeles:

    “If only vicariance is allowable, then such a distribution makes no sense whatsoever.”

    No, no, it’s more subtle: only vicariance is allowable, unless we’re talking about humans. Then it’s not to be entertained because it wasn’t a panbiogeographer’s hypothesis.

    Link to this
  415. 415. Grehan1896 6:22 pm 07/4/2014

    414. irenedelse
    No, no, it’s more subtle: only vicariance is allowable, unless we’re talking about humans. Then it’s not to be entertained because it wasn’t a panbiogeographer’s hypothesis.

    This is not intelligible.

    John Grehan

    Link to this
  416. 416. ectodysplasin 7:22 pm 07/4/2014

    Sure it is. Delse is simply criticizing your apparent need to vett your interpretations according to how well they align with the opinions of a previous worker.

    Link to this
  417. 417. Grehan1896 7:36 pm 07/4/2014

    Sure it is. Delse is simply criticizing your apparent need to vett your interpretations according to how well they align with the opinions of a previous worker.

    Please explain. Be explicit with respect to the particular example so I can respond in an informed way.

    John Grehan

    Link to this
  418. 418. MichaelHeads 9:26 pm 07/4/2014

    David 362 wrote: every stem group is ancestral to the corresponding crown group.

    Of course; just as the stem (trunk) on a tree (ancestor) produces the crown (descendant). The stem is the ancestor.

    But Ectodysplasin 413 wrote: Stem-group =/= ancestors…
    elephant birds are part of the kiwi stem group, because elephant birds are an extinct sister-taxon to kiwis. No one is claiming an ancestor-descendant relationship between these taxa.

    The first sentence is confusing, but I agree with the second sentence.

    Is a stem group an ancestor of the crown group or not? I think ‘stem group’ must be the ancestor by definition, but I don’t think that a paraphyletic basal complex is necessarily the stem group/ancestor. As I suggested, in a phylogeny: a(b(c, d)), the basal grade formed by a, b and c is not necessarily the stem group/ancestor of a.

    Link to this
  419. 419. Heteromeles 11:28 pm 07/4/2014

    Hard to be sarcastic in here, isn’t it, Irene?

    Let’s get back to that lichen and talk about what an amphipolar distribution might be due to.

    When you have a single species showing up in the same habitats on places of wildly different geology (Andean volcanic mountains, ancient Greenland rocks, the Antarctic peninsula, which could be either, deep ocean basaltic islands, and the Rocky Mountains), then it beggars the imagination that a single vicariance event explains it all. Doubtless, some uninformed person might suggest that it’s range is a remnant of the Cryogenian period 850 million years ago, when it covered the globe, and that it’s been hanging on ever since. Since the species is well embedded in the Ascomycetes (as well as being asexual), and Ascomycetes evolved something like 400 million years after the Cryogenian, such a hypothesis is untenable.

    I’d further suggest that no chain of vicariance events explains the present distribution. However, a simple dispersal hypothesis handily explains it all with admirable parsimony. We know fungal spores get into the upper air, because people find them there.

    If you’re going to say that widespread fungi are due to surface vicariance events, then it’s further absolutely necessary to explain why spores that get into the upper air (where people have found them by directly sampling) all die with absolutely no exceptions, and therefore this is not a viable mode of long-distance dispersal. Considering how hardy most spores are, this is a hard case to prove. For one, sporopollenin, the coating on fungal spores, is sufficiently tough that it’s been known to not fossilize after over 400 million years in one case I know of.

    Link to this
  420. 420. Grehan1896 10:58 am 07/5/2014

    419. Heteromeles
    Hard to be sarcastic in here, isn’t it, Irene?

    Why is it important for you to be sarcastic?

    When you have a single species showing up in the same habitats on places of wildly different geology (Andean volcanic mountains, ancient Greenland rocks, the Antarctic peninsula, which could be either, deep ocean basaltic islands, and the Rocky Mountains), then it beggars the imagination that a single vicariance event explains it all.

    Who is proposing such?

    I’d further suggest that no chain of vicariance events explains the present distribution.

    What constitute vicariance events?

    However, a simple dispersal hypothesis handily explains it all with admirable parsimony.

    Perhaps so. Only one dispesal event? Evidence or speculation?

    We know fungal spores get into the upper air, because people find them there.

    Sure, and birds fly, but that does not meant that each bird species or other higher taxonomic group is everywhere there is suitable habitat.

    If you’re going to say that widespread fungi are due to surface vicariance events, then it’s further absolutely necessary to explain why spores that get into the upper air (where people have found them by directly sampling) all die with absolutely no exceptions, and therefore this is not a viable mode of long-distance dispersal.

    With an undifferentiated species there is no need to invoke vicariance. Obviously over the range of an organisms its means of survival have contributed to its distribution range. It would be interesting to analyze the molecular and morphological genetics of this unidentified species. Are you saying that this fungus (please identify) is only in Greenland, the Antarctic Peninsula, and the Rocky Mountains only? And what is the sister species to this fungus?

    John Grehan

    Link to this
  421. 421. Heteromeles 1:28 pm 07/5/2014

    Sarcasm has a place, especially when sarcasm gets a point across indirectly when people dislike being confronted with a direct statement.

    As for the vicariance argument, as I understood it, you said that it was possible when I first pointed it out, so perhaps you’re being sarcastic when you ask “Who is proposing such? That post is a direct response that, no vicariance is not a legitimate response.

    As for the fungus, I’m ransacking my notes for the name. A friend of mine in grad school proposed studying it for his master’s thesis back in 1999, and I haven’t gotten in touch with him to find the name again. It’s one of the 100,000 holdings at the UW-Madison lichen herbarium, and I’ll post the name when I find it.

    As for that last paragraph, it’s gobbledygook, sorry. There is no such thing as an undifferentiated species, and the statemtn “Obviously over the range of an organisms [sic] its means of survival have contributed to its distribution range” is a meaningless tautology. Organisms only survive where they survive, and they only live where they live. The point of it growing on different, widely separated mountains and coasts that grew at different times is that it did not evolve from a widespread species into a cryo-specialist, because there is no geologic way that could have happened. Geology precludes vicariance in this species.

    The present geographic range is as I remember it when I first learned about it: Greenland, Antarctica, the Rockies, the Andes, and one of the subantarctic islands. My friend was trying to find a way to get someone to fund him to go to Greenland and Antarctica to study this species in its native habitat, and he dropped the project when he couldn’t get funding. I still liked his working title though: “bipolar asexual disjuncts.” Too bad the funding agencies thought it was insufficiently catchy.

    Link to this
  422. 422. Grehan1896 2:12 pm 07/5/2014

    Sarcasm has a place, especially when sarcasm gets a point across indirectly when people dislike being confronted with a direct statement.

    Ok. I guess I don’t see the need myself.

    As for the vicariance argument, as I understood it, you said that it was possible when I first pointed it out, so perhaps you’re being sarcastic when you ask “Who is proposing such? That post is a direct response that, no vicariance is not a legitimate response.

    Admittedly I lose track of some of the discourse due to its disjoined nature, and sometimes the context is not readily apparent. I was not intending sarcasm, but I was interested to know who has proposed a vicariance origin for the example in question.

    As for the fungus, I’m ransacking my notes for the name. A friend of mine in grad school proposed studying it for his master’s thesis back in 1999, and I haven’t gotten in touch with him to find the name again. It’s one of the 100,000 holdings at the UW-Madison lichen herbarium, and I’ll post the name when I find it.

    OK

    As for that last paragraph, it’s gobbledygook, sorry. There is no such thing as an undifferentiated species, and the statemtn

    Of course its differentiated with respect to other species, but I was wondering if the species you were referring to was homogenous throughout or geographically differentiated in any way.

    “Obviously over the range of an organisms [sic] its means of survival have contributed to its distribution range” is a meaningless tautology. Organisms only survive where they survive, and they only live where they live.

    But how did they get to be where they live was what I was addressing. I have no problem with the means of survival also contributing to its current distribution range. But what made the range disjunct seems to be the open ended question.

    The point of it growing on different, widely separated mountains and coasts that grew at different times is that it did not evolve from a widespread species into a cryo-specialist, because there is no geologic way that could have happened. Geology precludes vicariance in this species.

    You have evidence that this species did not formerly have a widespread contiguous distribution?

    The present geographic range is as I remember it when I first learned about it: Greenland, Antarctica, the Rockies, the Andes, and one of the subantarctic islands.

    Interesting that the pattern is Greenland-Rockies (whatever part or whole?)-Andes (some or whole) and a subantarctic island, rather than, for example, Greenland-Himalayas-New Zealand-Subantarctic island.

    My friend was trying to find a way to get someone to fund him to go to Greenland and Antarctica to study this species in its native habitat, and he dropped the project when he couldn’t get funding. I still liked his working title though: “bipolar asexual disjuncts.” Too bad the funding agencies thought it was insufficiently catchy.

    My sympathies (and I am not being sarcastic). Funding is such a fickle endeavor.

    John Grehan

    Link to this
  423. 423. DavidMarjanovic 2:18 pm 07/5/2014

    How did tortoises in general get to be where they are? The origin of the tortoises on the arc may be the same as the origin of their relatives elsewhere.

    Superficially, that would just turn the arrow of the question around: if tortoises evolved on the arc*, how did they spread to continents?

    Less superficially, it doubles the question: how did the ancestors of tortoises reach the arc? Unless you’re saying that the arc is where they had lived all the way back to the origin of terrestrial tetrapods, you can’t evade that question.

    * …even though their two closest relatives, Geoemydidae and Emydidae, are native to Asia and North America, respectively.

    Well of course I respect your right to have an opinion, even if I might not respect the opinion!

    …Is that all you can do to support your hypothesis? Is “before people had looked at enough cells, they thought humans had 48 chromosomes, so there” the best you can do?

    Ashamed of what? Be precise please.

    You were pointed towards “the multiple lines of data linking the first settlement of America with Northeast Asia”. Instead of starting to read up on them or coming up with any counterargument, you made that quip about 48 chromosomes – which only makes “sense” as the fallacy of “science has been wrong before, so it can all be safely ignored; nothing shall endanger my pet hypothesis”.

    Do I seriously need to explain what’s bad about that?

    Seriously?

    Maybe molecular analysis is good enough for hominids. Problem is that it is not corroborated by the morphogenetic evidence that I have been able to substantiate. Some cladistic theorists take the view that molecular analysis is not really cladistic, but phonetics dressed up in cladistic terminology and techniques. No doubt you and others would disagree and that’s OK. I’m just making an observation here.

    1) Judging from the paragraph you wrote next, you mean “morphological”. “Morphogenetic” applies to the early development of the embryo.
    2) Oh dear, you’ve taken a half-digested tidbit of overheard information and are now regurgitating it 20 to 30 years later. Some methods that were used in early analyses – till the early ’00s or so in some cases – were arguably phenetic in that they didn’t try to distinguish plesio- from apomorphic character states or, rather, were distance methods that didn’t look at individual characters at all: DNA/DNA hybridization, WPGMA, UPGMA, and neighbor-joining.
    3) Phonetics is of course something different. Did your spellchecker strike? :-)
    4) For the last 15 years or so, phylogenetic analyses of molecular data have almost exclusively used maximum likelihood and Bayesian inference, which cannot be called phenetic by any stretch of the imagination.

    You’re grasping for straws. Stop embarrassing yourself.

    Oh, BTW, you’ll find molecular phylogeneticists today who restrict the term “cladistics” to parsimony, which they don’t use (outside of special situations) because they don’t need to. Those people will happily tell you that what they’re doing is not cladistics. It is, however, phylogenetics and not phenetics.

    Of course; just as the stem (trunk) on a tree (ancestor) produces the crown (descendant). The stem is the ancestor.

    The stem group, however, contains this stem and all branches except those that belong to the crown group. I agree that the term is somewhat misleading.

    Why is it important for you to be sarcastic?

    Heteromeles inferred that irenedelse was trying to be sarcastic.

    What constitute vicariance events?

    Events that fragment a habitat that was continuous until then.

    Link to this
  424. 424. DavidMarjanovic 2:43 pm 07/5/2014

    Argh, runaway italics again. Let’s see if I can simply close the tag.

    You have evidence that this species did not formerly have a widespread contiguous distribution?

    I do: this contiguous distribution would have needed to go from the Arctic Ocean to the Antarctic Ocean, so it couldn’t be any later than the Jurassic (or the Cretaceous/Paleogene boundary if there really was a short-lived landbridge between North and South America at that time, which is hard to test), and it would have had to span a lot of different habitats, most of them warm and forested, while nowadays this one species is only found in very cold and dry places. You’d need a very unparsimonious scenario to harmonize all of that without overseas dispersal.

    Interesting that the pattern is Greenland-Rockies (whatever part or whole?)-Andes (some or whole) and a subantarctic island, rather than, for example, Greenland-Himalayas-New Zealand-Subantarctic island.

    Even so, Greenland and the Rockies have never been in contact; there has been a more or less continuous mountain range between the Rockies and the Andes for 3 million years now, but in Central America the mountains are low enough to be covered by rainforest.

    South America and Antarctica were in contact till some 37 Ma ago, but the Andes really weren’t high yet. And back then, of course, the Antarctic peninsula was covered in paratropical rainforest, while deciduous trees blanketed the polar interior.

    Either you assume a lot of precisely – ridiculously – parallel evolution in unconnected locations over a really, really long stretch of time, or you have to accept overseas dispersal as an explanation. The only alternative I can see is magic.

    Link to this
  425. 425. Grehan1896 4:05 pm 07/5/2014

    You have evidence that this species did not formerly have a widespread contiguous distribution?
    I do: this contiguous distribution would have needed to go from the Arctic Ocean to the Antarctic Ocean, so it couldn’t be any later than the Jurassic (or the Cretaceous/Paleogene boundary if there really was a short-lived landbridge between North and South America at that time, which is hard to test), and it would have had to span a lot of different habitats, most of them warm and forested, while nowadays this one species is only found in very cold and dry places. You’d need a very unparsimonious scenario to harmonize all of that without overseas dispersal.

    Thanks for the clarification. If I understand you correctly, the current habitat restriction precludes a different ancestral habitat and so that precludes any older origin for the extant distribution than for (presumably youngest) the habitats it currently occupies.

    Either you assume a lot of precisely – ridiculously – parallel evolution in unconnected locations over a really, really long stretch of time, or you have to accept overseas dispersal as an explanation. The only alternative I can see is magic.

    I understand your argument and maybe you are correct. I have no opinion on this particular case (I would need more information than is currently available). I do think it is interesting that there is a restricted pattern of distribution that does not include other suitable habitats, and it would be interesting to know about the sister group distribution. Well, perhaps someone will be able to do a detailed analysis of this distribution some time.

    John Grehan

    Link to this
  426. 426. DavidMarjanovic 5:00 pm 07/5/2014

    If I understand you correctly, the current habitat restriction precludes a different ancestral habitat and so that precludes any older origin for the extant distribution than for (presumably youngest) the habitats it currently occupies.

    *eyeroll* Snap out of it. Only the Sith think in absolutes.

    Parsimony. How probable is it that the same habitat preference evolves so many times independently in the same species – and that quickly enough that it’s still the same species? Remember, it’s asexual – the population can’t be panmictic, in fact, there isn’t a population at all as far as genetics is concerned.

    If you want to be a scientist, you need to recognize that “it’s possible” isn’t enough to make a hypothesis robust. A hypothesis needs to be more probable than all alternatives; if it’s not, it can and should be dismissed as rubbish.

    Link to this
  427. 427. naishd 5:03 pm 07/5/2014

    Aww, dammit… html screwup again?? Sigh…

    Link to this
  428. 428. Grehan1896 5:23 pm 07/5/2014

    • 423. DavidMarjanovic

    • How did tortoises in general get to be where they are? The origin of the tortoises on the arc may be the same as the origin of their relatives elsewhere.
    Superficially, that would just turn the arrow of the question around: if tortoises evolved on the arc*, how did they spread to continents?

    I do not know where tortoises first evolved. I do suggest that the origin of the tortoises on the Galapagos was by colonizing from a former island arc. What they did before that is beyond my current knowledge.

    * …even though their two closest relatives, Geoemydidae and Emydidae, are native to Asia and North America, respectively.

    That may well point to a Pacific basin ancestral distribution for these groups.

    Ashamed of what? Be precise please.
    You were pointed towards “the multiple lines of data linking the first settlement of America with Northeast Asia”. Instead of starting to read up on them or coming up with any counterargument, you made that quip about 48 chromosomes – which only makes “sense” as the fallacy of “science has been wrong before, so it can all be safely ignored; nothing shall endanger my pet hypothesis”.

    Thank you. But there were no details provided as to why these “multiple lines” had any necessary validity. Given the emphasis on the “multiple lines” I was just pointing out that multiple lines does not give any necessary validity. If details of what constituted supporting evidence was given I might have addressed that if I felt there was any contention. As it is I am OK with there being a lack of corroboration from those studies..

    1) Judging from the paragraph you wrote next, you mean “morphological”. “Morphogenetic” applies to the early development of the embryo.
    I was using it in the sense of morpo genesis – development of genetic molecules into morphology.

    2) Oh dear, you’ve taken a half-digested tidbit of overheard information and are now regurgitating it 20 to 30 years later. Some methods that were used in early analyses – till the early ’00s or so in some cases – were arguably phenetic in that they didn’t try to distinguish plesio- from apomorphic character states or, rather, were distance methods that didn’t look at individual characters at all: DNA/DNA hybridization, WPGMA, UPGMA, and neighbor-joining.

    No, I was not just referring to these methods.

    3) Phonetics is of course something different. Did your spellchecker strike?

    I guess so!

    4) For the last 15 years or so, phylogenetic analyses of molecular data have almost exclusively used maximum likelihood and Bayesian inference, which cannot be called phenetic by any stretch of the imagination.

    Some cladistic theorists disagree.

    Oh, BTW, you’ll find molecular phylogeneticists today who restrict the term “cladistics” to parsimony, which they don’t use (outside of special situations) because they don’t need to. Those people will happily tell you that what they’re doing is not cladistics. It is, however, phylogenetics and not phenetics.

    And some cladists will argue that parsimony is not really cladistics either.

    Of course; just as the stem (trunk) on a tree (ancestor) produces the crown (descendant). The stem is the ancestor.

    In other words, the ancestor is the character/s defining that node.

    The stem group, however, contains this stem and all branches except those that belong to the crown group. I agree that the term is somewhat misleading.

    This would appear to make the stem group paraphyletic since it is a grouping that does not contain all descendants (crown group).

    What constitute vicariance events? Events that fragment a habitat that was continuous until then.
    Which means that vicariance can be more than just continental separation.

    John Grehan

    Link to this
  429. 429. Grehan1896 5:36 pm 07/5/2014

    If you want to be a scientist, you need to recognize that “it’s possible” isn’t enough to make a hypothesis robust. A hypothesis needs to be more probable than all alternatives; if it’s not, it can and should be dismissed as rubbish.

    Maybe. No doubt you would have been one of those putting the boot into McClintock for that very kind of reasoning. What may seem reasonable and straightforward in principle may not be in practice. After all, philosophers of science have been arguing with each other as to what constitutes the rules of science for decades.

    John Grehan

    Link to this
  430. 430. irenedelse 7:31 pm 07/5/2014

    I’m afraid Darren is going to have to close the thread to put the italics out of their misery. Sigh. (Plus the length is starting to get a bit hard on internet browsers…)

    Link to this
  431. 431. irenedelse 7:43 pm 07/5/2014

    Alternately, TetZoo could pull a Pharyngula and open up a new post for general discussion purposes… But beware that it may become high maintenance. (BTW, it’s not that I’m extremely eager to discuss the minutiae of panbiogeography, but Heteromeles’ lichen is fascinating, and so are cryptobranchid salamanders, human dispersion over the globe, island biogeography, etc.)

    Link to this
  432. 432. irenedelse 7:48 pm 07/5/2014

    @Grehan:

    “419. Heteromeles
    “Hard to be sarcastic in here, isn’t it, Irene?”

    Why is it important for you to be sarcastic?”

    Next time, maybe we should try the Turing test.

    Link to this
  433. 433. DavidMarjanovic 10:03 am 07/6/2014

    I do not know where tortoises first evolved. I do suggest that the origin of the tortoises on the Galapagos was by colonizing from a former island arc. What they did before that is beyond my current knowledge.

    No, you can’t use this way out. Science is science all the way down; if you stop your train of thought at any point before the bottom, you’re not doing science.

    You think that tortoises don’t do overseas dispersal except within an archipelago – this means you think they either 1) evolved on the archipelago, or 2) they once had greater dispersal capability but somehow lost it at some point, or 3) the hypothetical island arc was very close to a suitable continent at a suitable time. Now, option 1 just moves the problem instead of solving it. 2 brings up a further question: how did this loss happen? 3 requires you to review the geologic history of the area to test it.

    What you’re doing by just taking your ball and going home is to say that the Galápagos tortoises just magically poofed into being on an oceanic island for all you care. And that’s not science. You have to follow through all the implications of a hypothesis in order to evaluate it by both falsification (if possible) and by parsimony.

    I was using it in the sense of morpo genesis [sic] – development of genetic molecules into morphology.

    Oh. Have any studies of development genetics been done on any paleognaths at all?

    Some cladistic theorists disagree.

    On what basis?

    And some cladists will argue that parsimony is not really cladistics either.

    It has been argued that what Hennig used wasn’t maximum parsimony, but maximum compatibility, a thoroughly extinct method that makes quite unrealistic assumptions. But then, the term “cladistics” is younger than that.

    “Of course; just as the stem (trunk) on a tree (ancestor) produces the crown (descendant). The stem is the ancestor.”

    In other words, the ancestor is the character/s defining that node.

    First, you’re replying to a quote from yourself here. Second, no, an ancestor isn’t a character or several. It’s an individual organism, or a breeding pair, or a population; or, in the other direction, a pair of gametes or a single dividing cell.

    This would appear to make the stem group paraphyletic

    Yes! By definition, all stem groups are paraphyletic!!!

    Say, how does this work? On the one side, you’re acquainted with the work of unspecified “cladistic theorists” that nobody else seems to have heard of – and on the other side, you’re unfamiliar with basic terminology?

    Which means that vicariance can be more than just continental separation.

    …Of course; nobody in this thread has tried to claim otherwise. ~:-|

    “…even though their two closest relatives, Geoemydidae and Emydidae, are native to Asia and North America, respectively.”

    That may well point to a Pacific basin ancestral distribution for these groups.

    It’s a pretty obvious Laurasian distribution, most likely with an Asian origin. Current distributions and the fossil record concur.

    No doubt you would have been one of those putting the boot into McClintock for that very kind of reasoning.

    [PERSONAL INSULT DELETED BY DNAISH: David, please don't]. McClintock demonstrated in detail how her hypothesis explained observations that… didn’t have any other explanation, actually. I’m not sure what was going through the heads of those of her colleagues who didn’t simply dismiss her as a woman; probably it just wasn’t clear to them what even needed an explanation and what didn’t, because McClintock made all those discoveries before 1953 – the year the structure of DNA was published. Figuring out the genetic code took almost to the end of the 1960s, and the transposons McClintock had discovered weren’t cloned and characterized till the 70s.

    Now please explain how you are in an analogous situation: how there is no science of biogeography except for panbiogeography, how all other biogeographers are just stumbling around in the dark, how you have tested your hypotheses (see above!) and people just ignore the results. Go ahead, we’re waiting!

    What may seem reasonable and straightforward in principle may not be in practice. After all, philosophers of science have been arguing with each other as to what constitutes the rules of science for decades.

    Sure, sure. They’ve long stopped arguing about the basics, though. You’re trying to argue about the basics.

    Link to this
  434. 434. irenedelse 10:47 am 07/6/2014

    Could we also refrain from pointless speculation like “No doubt you would have been one of those putting the boot into McClintock for that very kind of reasoning”, please?

    John, when you invoke the example of the greatest names in the field, expect to be held to as high a standard of scrutiny. (And also to a good dose of scepticism, at the very least, when you answer to data-based questions with appeals to personal beliefs.)

    Link to this
  435. 435. Grehan1896 11:11 am 07/6/2014

    433. DavidMarjanovic provided some very thoughtful questions and comments. My responses are below.

    I do not know where tortoises first evolved. I do suggest that the origin of the tortoises on the Galapagos was by colonizing from a former island arc. What they did before that is beyond my current knowledge.
    No, you can’t use this way out. Science is science all the way down; if you stop your train of thought at any point before the bottom, you’re not doing science.

    True. But it’s not a way out. I did not say that the question was off limits. I just said that it was beyond my current knowledge. If others figure things out further then OK.

    You think that tortoises don’t do overseas dispersal except within an archipelago – this means you think they either 1) evolved on the archipelago, or 2) they once had greater dispersal capability but somehow lost it at some point, or 3) the hypothetical island arc was very close to a suitable continent at a suitable time. Now, option 1 just moves the problem instead of solving it. 2 brings up a further question: how did this loss happen? 3 requires you to review the geologic history of the area to test it.

    But geological ‘history’ does not test anything. Any historical reconstruction is just a theory about the past so I don’t see it as a test. One may end up with geological information that is congruent or incongruent with the biogeographic hypothesis. As you say, the island arc is ‘hypothetical’ in the sense that one cannot directly experience its existence. Historical geologists appear to be in agreement about its existence so there is no incongruency there.

    What you’re doing by just taking your ball and going home is to say that the Galápagos tortoises just magically poofed into being on an oceanic island for all you care. And that’s not science. You have to follow through all the implications of a hypothesis in order to evaluate it by both falsification (if possible) and by parsimony.

    Well of course these are your words, not mine.

    I was using it in the sense of morpo genesis [sic] – development of genetic molecules into morphology.
    Oh. Have any studies of development genetics been done on any paleognaths at all?

    I don’t know. But one may compare the morphology of genetic development just as one may compare the morphology of DNA molecules.

    Some cladistic theorists disagree. On what basis?

    That the process is still one of overall similarity because one has to construct a tree in order to know what is primitive or derived (and this on top of those studies that have to match base pairs by overall fit between genes of unequal size). I’m not so worried about this myself, I only mention that it is a contention among some systematic theoriest.

    And some cladists will argue that parsimony is not really cladistics either.
    It has been argued that what Hennig used wasn’t maximum parsimony, but maximum compatibility, a thoroughly extinct method that makes quite unrealistic assumptions. But then, the term “cladistics” is younger than that.

    And I have used parsimony myself. What do you think of three item analysis?

    “Of course; just as the stem (trunk) on a tree (ancestor) produces the crown (descendant). The stem is the ancestor.”
    In other words, the ancestor is the character/s defining that node.
    First, you’re replying to a quote from yourself here. Second, no, an ancestor isn’t a character or several. It’s an individual organism, or a breeding pair, or a population; or, in the other direction, a pair of gametes or a single dividing cell.

    In that sense I agree, but if all taxa are terminals what is at the nodes?

    This would appear to make the stem group paraphyletic
    Yes! By definition, all stem groups are paraphyletic!!!

    Good, I just wanted to be sure I understood. Stem groups are artificial constructions.

    Say, how does this work? On the one side, you’re acquainted with the work of unspecified “cladistic theorists” that nobody else seems to have heard of

    You could not have heard of them since I did not give their names.

    – and on the other side, you’re unfamiliar with basic terminology?

    I probably have a lot of gaps in my knowledge. I have not used the ‘stem group’ in my work and have not needed to. So while I have heard of the term I had no need to look into it.

    Which means that vicariance can be more than just continental separation.
    …Of course; nobody in this thread has tried to claim otherwise. ~:-|

    Well, maybe I got the wrong impression, but nearer the beginning there were indications that a kiwi-elephant bird relationship was incongruous with vicariance because it did not match the continental breakup sequence.

    “…even though their two closest relatives, Geoemydidae and Emydidae, are native to Asia and North America, respectively.” That may well point to a Pacific basin ancestral distribution for these groups.
    It’s a pretty obvious Laurasian distribution, most likely with an Asian origin. Current distributions and the fossil record concur.

    The tortoises are not found all over ‘Laurasia” so I am not sure what ‘Laurasian’ means other than that. Why is an Asian origin for the range ‘most likely’? (And please give a probability for this as you indicate elsewhere that this is your criterion for not being rubbish).

    No doubt you would have been one of those putting the boot into McClintock for that very kind of reasoning.

    Now please explain how you are in an analogous situation:

    Your statement was A hypothesis needs to be more probable than all alternatives; if it’s not, it can and should be dismissed as rubbish. The context of the McClintok reference is that she was not in a position to demonstrate that her model was more probable than all the alternatives so by this criterion I would predict that David would have responded in this way since it is consistent with what he has said. The analogy here is that David is saying that if one cannot provide some kind of probability measurement that is acceptable to him then the proposition should be dismissed as rubbish.

    how there is no science of biogeography except for panbiogeography, how all other biogeographers are just stumbling around in the dark, how you have tested your hypotheses (see above!) and people just ignore the results. Go ahead, we’re waiting!

    Those are some very worthwhile assertions, although with problematic assumptions.
    1/ Panbiogeography is not the only science of biogeography.
    2/ Not all other biogeographers are stumbling around in the dark.
    3/ Hypotheses have been tested in various ways in various papers. Some use geological corroboration, others use biogeographic corroboration. Read Mike’s books for illustration.
    4/ One test is the generation of corroborated novel predictions. For the Galapagos alone there were two novel predictions – that the Galapagos was associated with a major tectonic feature, and there was former land in what is now the Eastern Pacific (they are novel because they had not already been predicted in geology or Darwinian biogeography). The first prediction is corroborated by discovery of the Galapagos Gore, the second corroborated by discovery of allochthonous terranes in the central region of the Americas.

    What may seem reasonable and straightforward in principle may not be in practice. After all, philosophers of science have been arguing with each other as to what constitutes the rules of science for decades.
    Sure, sure. They’ve long stopped arguing about the basics, though. You’re trying to argue about the basics.

    Perhaps so.

    Link to this
  436. 436. irenedelse 11:27 am 07/6/2014

    Italics. Pity for the italics. :-(

    Link to this
  437. 437. naishd 11:29 am 07/6/2014

    John: please try and stick with ‘forward slash’, then ‘i’, when closing italics. If you put the ‘i’ first, the tag doesn’t get closed and we have italics for eternity…

    Link to this
  438. 438. Heteromeles 11:49 am 07/6/2014

    Does it help to put in a ? Let’s see.

    Link to this
  439. 439. Heteromeles 11:50 am 07/6/2014

    Nope, the closed ” didn’t do anything. All yours, Darren.

    Link to this
  440. 440. Grehan1896 12:35 pm 07/6/2014

    434. irenedelse
    Could we also refrain from pointless speculation like “No doubt you would have been one of those putting the boot into McClintock for that very kind of reasoning”, please?

    It was not pointless. It made the point that with the kind of demands being made this would have been the outcome.

    John, when you invoke the example of the greatest names in the field, expect to be held to as high a standard of scrutiny. (And also to a good dose of scepticism, at the very least, when you answer to data-based questions with appeals to personal beliefs.)

    The standards of scrutiny themselves may vary in quality too.

    John Grehan

    Link to this
  441. 441. Grehan1896 12:43 pm 07/6/2014

    John: please try and stick with ‘forward slash’, then ‘i’, when closing italics. If you put the ‘i’ first, the tag doesn’t get closed and we have italics for eternity…

    I did that. Seems not to work. I think I will just avoid italics altogether and space things so my responses are clear.

    John Grehan

    Link to this
  442. 442. irenedelse 1:48 pm 07/6/2014

    “The standards of scrutiny themselves may vary in quality too.”

    And changing the argument all the time is helpful to avoid the kind of scrutiny that would make a hypothesis falsifiable. Which makes it pretty pointless to discuss.

    (About formatting issues, Darren, I’m afraid you’ll have to edit the buggy comments. The SciAm platform is not forgiving to html mistakes.)

    Link to this
  443. 443. naishd 2:36 pm 07/6/2014

    Re: comment 441 — John, it works fine – it’s just that you keep on putting ‘i’ before ‘/’ when adding the closing tag (instead of the other way round). I know because I keep editing it (and I’m about to go do so again, now). Thanks indeed.

    If you don’t want to use html, just put quote marks around the quoted section… it would also be useful if the commenter name and/or comment number were mentioned too, since it’s just about impossible to keep track of who’s quoting who some of the time.

    Link to this
  444. 444. Heteromeles 5:03 pm 07/6/2014

    Apropos of nothing, a friend pointed me towards this article First evidence of bryophyte diaspores in the plumage of transequatorial migrant birds. It should be readable online, and there are links to other bipolar disjunct studies.

    Link to this
  445. 445. MichaelHeads 6:06 pm 07/6/2014

    433: John: Which means that vicariance can be more than just continental separation.

    David: …Of course; nobody in this thread has tried to claim otherwise. ~:-|

    No, but Mitchell et al. 2014 did in their Science paper that started the thread. It’s a very common mistake and creates a lot of confusion.

    Link to this
  446. 446. irenedelse 6:52 pm 07/6/2014

    @Michael Heads #445:

    Let’s see what Mitchell et. al. say on the subject:

    “Perhaps the most enigmatic of the modern palaeognaths are the recently extinct giant Madagascan elephant birds. Africa and Madagascar were the first continental fragments to rift from the supercontinent Gondwana, separating from the other continents (and each other) completely during the Early Cretaceous (~130 to 100 Ma)(15). Consequently, the continental vicariance model predicts that elephant birds and ostriches should be the basal palaeognath lineages (16). Most molecular analyses recover the ostrich in a basal position, consistent with a vicariant model. However, the phylogenetic position of the elephant birds remains unresolved, as cladistic studies of ratite morphology are sensitive to character choice and may be confounded by convergence (17), whereas DNA studies have been hampered by the generally poor molecular preservation of elephant bird remains (18).”

    They don’t say that only continental separation matters in vicariance, but that in the case of elephant birds, the timing of continental separation makes the vicariance model inapplicable since these birds are not the most basal of paleognaths.

    Link to this
  447. 447. MichaelHeads 2:16 am 07/7/2014

    Hi Irene,

    The key sentences in Mitchell et al 2014 are:

    ‘We sequenced the mitochondrial genomes of two elephant birds and performed phylogenetic analyses, which revealed that these birds are the closest relatives of the New Zealand kiwi and are distant from the basal ratite lineage of ostriches. This unexpected result strongly contradicts continental vicariance [no - it only contradicts the breakup sequence]and instead supports flighted dispersal in all major ratite lineages’.

    ‘molecular dating provides limited power for testing hypotheses about ratite biogeography… Thus, topological comparisons [for them, this means comparison between phylogeny and breakup sequence] may be a more robust tool to test hypotheses of vicariance and connection’. This is true only if you assume that breakup is the only possible mechanism for vicariance.

    ‘The phylogenetic placement of the elephant
    bird as sister to the kiwi creates a marked discordance between the order of continental breakup
    (Fig. 3, A and B) and the sequence of palaeognath
    divergences (Fig. 3C)
    . Instead [i.e. instead of vicariance], it appears that the common ancestor of elephant birds and kiwis was probably flighted and capable of long-distance dispersal…’,

    We suggest that flighted dispersal was the primary driver of the distribution of palaeognath lineages and that the discordance between distribution and phylogeny [no - the discordance is between the phylogeny and breakup sequence] is more consistent with lineage turnover in a phylogenetically diverse, flighted, and widespread clade [than with vicariance]‘.

    I’ve added all the italics. Note that they assume, with most other authors, that vicariance could only be the result of breakup. But there was extensive rifting in the Tasman Sea region and between West and East Antarctica for about 15 million years before the crust finally broke completely with seafloor spreading in the Tasman and the Southern Ocean.

    The pre-drift phase of rifting is not mentioned by Mitchell et al., but coincides spatially with the break between the trans-Indian Ocean ratite clade and the trans-Pacific clade (cf. the polar projection figure 3.3 in my 2014 book ‘Biogeography of Australasia’). This occurred after – both phylogenetically and geologically – the development of the break between the ostriches in Africa and their sister group, the Madagascar – Indo-Pacific clade.

    As always, another possible explanation for this very beautiful pattern is chance dispersal.

    Link to this
  448. 448. John Scanlon FCD 4:36 am 07/7/2014

    “This is true only if you assume that breakup is the only possible mechanism for vicariance.” – M Heads

    Surely this applies to any inference that palaeognath distribution has something to do with Jurassic/early-Cretaceous events; the only thing demanding that timescale is the assumption that the kiwi-elephant-bird hypothetical flightless ancestor* had a continuous terrestrial range across the Africa-Madagascar rift.

    If not continental breakup (and overwater dispersal by volant ancestors is not to be contemplated) they must be thinking of pushing ratite diversification earlier, not later than breakup of Pangaea. [OK I won't say what I want to add, to save Darren cleaning it up]

    *I’m not sure, but suspect that our PBG friends (like the pattern cladists of old) may claim not to be making inferences about ancestors not directly observed; but that would be false.

    Link to this
  449. 449. MichaelHeads 7:17 am 07/7/2014

    Hi John,

    You suggest: Surely this applies to any inference that palaeognath distribution has something to do with Jurassic/early-Cretaceous events; the only thing demanding that timescale is the assumption that the kiwi-elephant-bird hypothetical flightless ancestor* had a continuous terrestrial range across the Africa-Madagascar rift.

    I’m certainly not suggesting that the elephant bird-kiwi clade or its latest common ancestor was in Africa. That’s ostrich territory.

    You sugest: I’m not sure, but suspect that our PBG friends (like the pattern cladists of old) may claim not to be making inferences about ancestors not directly observed; but that would be false.

    I’m certainly not a pattern cladist. I’m making lots of inferences about ancestors not directly observed; specifically, inferences about their distribution.

    Mitchell et al suggested: ‘molecular dating provides limited power for testing hypotheses about ratite biogeography… Thus, topological comparisons [for them, this means comparison between phylogeny and breakup sequence] may be a more robust tool to test hypotheses of vicariance and connection’.

    Does the lack of congruence between breakup sequence and phylogeny (reserving judgment on fossil-calibrated time estimates, as they suggest)provide a robust test of vicariance in general, or just a test of vicariance associated with the breakup sequence? I think it’s the latter. Do you disagree?

    Link to this
  450. 450. Heteromeles 10:54 am 07/7/2014

    @MichaelHead: if Africa is precluded from being the last home of the elephant bird stem clade due to the (inferred) presence of ostriches (which could be wrong, incidentally. For all we know, they evolved in China and spread west), then we’re left with a geological impossibility known as the Lost Continent of Lemuria, because you’re postulating a landmass that connected Madagascar and New Zealand, and not so incidentally, throwing Australia (and its rhea) out of the picture entirely.

    Sorry, the science has progressed to the point where you can download a program called Google Earth and study the sea floor all around every landmass. There’s plenty of evidence for plate tectonics all over the ocean floor, but there’s no evidence whatsoever for a land bridge between Madagascar and New Zealand at the time when the ratites were evolving (or likely, for that matter, at any time). Lemuria is a failed hypothesis, as is the lost continent of Mu in the Pacific. Renaming lost continents as panbiogeography won’t resurrect this idea, because anyone with a computer and Google Maps can disprove it in a few minutes.

    Link to this
  451. 451. DavidMarjanovic 11:08 am 07/7/2014

    “No doubt you would have been one of those putting the boot into McClintock for that very kind of reasoning.”

    [PERSONAL INSULT DELETED BY DNAISH: David, please don't]

    Uh, Darren, accusing someone of being the kind of person that was “putting the boot into McClintock” is a much greater insult than me showing how small my vocabulary really is.

    Fact is, you know that.

    The Galileo gambit is bad enough. Accusing specific people, on no evidence, of being on the other side of the Galileo gambit is completely unacceptable – there’s no way to sugarcoat that. That’s what I was trying to get across.

    Back to business:

    True. But it’s not a way out. I did not say that the question was off limits. I just said that it was beyond my current knowledge. If others figure things out further then OK.

    But why do you bother forming an opinion on the topic without first trying to find out if others have figured things out further?

    By postulating that the Galápagos tortoises come from an island arc, you automatically postulate a list of other things with implications for the parsimony of your hypothesis. Instead of exploring that list, you just lean back and say you don’t know. That does not make sense.

    “[...] or 3) the hypothetical island arc was very close to a suitable continent at a suitable time. Now, option [...] 3 requires you to review the geologic history of the area to test it.”

    But geological ‘history’ does not test anything. Any historical reconstruction is just a theory about the past so I don’t see it as a test. One may end up with geological information that is congruent or incongruent with the biogeographic hypothesis. As you say, the island arc is ‘hypothetical’ in the sense that one cannot directly experience its existence. Historical geologists appear to be in agreement about its existence so there is no incongruency there.

    As you quoted, I’m talking about the hypothesis that this island arc was very close to a suitable continent at a suitable time. Do you know a paper about this?

    one may compare the morphology of genetic development just as one may compare the morphology of DNA molecules

    Of course (even though you’re the first to call either of those “morphology”). You made it sound like there was evidence from embryology that the tinamous are not nested among the other paleognaths but form their sister-group; that’s why I asked if any study of these data had ever been done (I’m not aware of any, but it’s not impossible that I’d overlook one).

    That the process is still one of overall similarity because one has to construct a tree in order to know what is primitive or derived

    …You need an outgroup to root the tree = to polarize your characters, except in very special cases like retrotransposon insertions maybe. This is true of parsimony, likelihood and Bayesian inference: they produce unrooted trees (outside of special cases), and then they root them on the outgroup. Distance methods, on the other hand, produce rooted trees.

    So I’m not sure I understand what you’re talking about.

    (and this on top of those studies that have to match base pairs by overall fit between genes of unequal size).

    Alignment is of course a problem, and so is making a morphological data matrix. I don’t see what that has to do with phylogenetics vs. phenetics.

    I’m not so worried about this myself, I only mention that it is a contention among some systematic theoriest.

    I can’t imagine that it has been in the last 10, more likely 20 years. Do you have a citation?

    What do you think of three item analysis?

    Short answer: this. Long answer: I need to read up on it, but it seems to make several assumptions that fail to hold for most empirical datasets; this seems to be why almost nobody uses it.

    Or did you mean if I think it’s phenetics? I’m sure it’s not.

    if all taxa are terminals what is at the nodes?

    Ah, that’s a common misunderstanding. Due to how diverse life is and how incomplete the fossil record is, it’s extremely unlikely that anything we know from the fossil record is a direct ancestor of anything else we know; many large extant taxa famously have no known fossil record at all. For this reason, the computer programs for phylogenetic analysis assumes that everything in the data matrix is a terminal. However, a “terminal” taxon can sit on the end of a zero-length branch! If you find a zero-length branch in your phylogenetic analysis, the taxon at its end is a good candidate for being an ancestor of what the analysis finds to be its sister-group.

    The tortoises are not found all over ‘Laurasia” so I am not sure what ‘Laurasian’ means other than that. Why is an Asian origin for the range ‘most likely’?

    Tortoises (Testudinidae) are found in the warm, not too wet parts of the northern continents (Laurasia), Africa and South America as well as islands around them (including Madagascar). The fossil record supports an Asian origin, and apparently so does molecular phylogenetics (Google just found me this). Geoemydids live today in the warm, wet parts of the northern continents and South America; North America is mostly too dry for them nowadays, but has yielded possible Eocene representatives. Emydids are mostly North American, where they apparently go back to the Late Cretaceous; only Emys has definitely made it out and spread all the way to Europe in the Miocene, a spread via Asia that is fairly well documented in the fossil record. Platysternon is Asian (warm & wet again). The extinct “lindholmemydids” (likely paraphyletic, or so I hear), sinochelyids and haichemydids are exclusively Asian as far as I know, and reach deep into the Cretaceous.

    So, from extant distributions you get Asia and/or North America as the most parsimonious place of origin, and the fossil record tilts this balance toward Asia. New fossils or new analyses of known fossils could overturn this at any time, but so far that’s the hypothesis that requires the smallest number of assumptions about migration and preservation – even though the difference to a few alternatives isn’t very large.

    The context of the McClintok reference is that she was not in a position to demonstrate that her model was more probable than all the alternatives

    Her model was the only explanation that had ever been proposed for her discoveries. It did not contradict any observations. So, it should have been tentatively accepted.

    “Sure, sure. They’ve long stopped arguing about the basics, though. You’re trying to argue about the basics.”

    Perhaps so.

    And you leave it at that? Aren’t you curious to find out if you’re really arguing about the basics? Aren’t you curious to find out if the philosophers of science have stopped arguing about the basis?

    I don’t get this.

    4/ One test is the generation of corroborated novel predictions. For the Galapagos alone there were two novel predictions – that the Galapagos was associated with a major tectonic feature, and there was former land in what is now the Eastern Pacific (they are novel because they had not already been predicted in geology or Darwinian biogeography). The first prediction is corroborated by discovery of the Galapagos Gore, the second corroborated by discovery of allochthonous terranes in the central region of the Americas.

    Good! (Honestly. I’m not being sarcastic.) Next test: does the timing fit? Were those terranes in the right area at the right time? Is there evidence that, say, the Galápagos tortoises came from elsewhere anyway?

    Let’s see what Mitchell et. al. say on the subject:

    The first time they mention vicariance, they say “continental vicariance”. If that doesn’t mean “vicariance by breakup”, what could it mean?

    Later, granted, they say just “vicariance” or “vicariant model”. Given that they spelled it out the first time, I think this is acceptable in something as space-limited as a Science paper.

    But there was extensive rifting in the Tasman Sea region and between West and East Antarctica for about 15 million years before the crust finally broke completely with seafloor spreading in the Tasman and the Southern Ocean.

    The pre-drift phase of rifting is not mentioned by Mitchell et al., but coincides spatially with the break between the trans-Indian Ocean ratite clade and the trans-Pacific clade

    Interesting point. But why would a rift in a forested continent be a barrier to even a flightless bird?

    And why are there both moa and kiwi in New Zealand?

    Link to this
  452. 452. naishd 11:16 am 07/7/2014

    On insults: yes, I get it, David. But I can’t permit the sort of insults that involve profanity. We move on.

    Link to this
  453. 453. John Scanlon FCD 12:08 pm 07/7/2014

    Oh, I’d though from reading bits of Croizat that extant taxa didn’t really descend from ancestral individuals or species, but sort of crystalised out from ancestral collections of characters permanently smeared across the landscape. Decided life was too short to read 10,000 pages of his stuff, and couldn’t see any use for it in my work.

    Pan-biogeographic methodology and synthesis is here to stay.
    – Robin Craw, Tuatara 27: 13 (1984)

    I wonder if he meant here in 1984? But no, here’s progress:

    “As always, another possible explanation for this very beautiful pattern is chance dispersal.” M Heads

    Glad to hear it; yes, individuals actually exist, and they are not all nailed to the bedrock. There are theories and methods that deal with them extensively; Darwin was quite good on some of them. You know, tall chap, balding, long white beard?

    Link to this
  454. 454. DavidMarjanovic 1:56 pm 07/7/2014

    I can’t permit the sort of insults that involve profanity

    Oh, is this a SciAm term of service? The National Geographic blogs are like that.

    Link to this
  455. 455. Grehan1896 2:37 pm 07/7/2014

    OK, I have made a darned good effort to get the italics right this time!
    451. DavidMarjanovic

    Uh, Darren, accusing someone of being the kind of person that was “putting the boot into McClintock” is a much greaterinsult than me showing how small my vocabulary really is.
    Fact is, you know that.

    It was certainly my inference that since McClintock’s theory could not have meet the kind of criteria David was invoking, David would have responded the same way. Of course there is no way to prove that since the event is historical.

    The Galileo gambit is bad enough. Accusing specific people, on no evidence, of being on the other side of the Galileo gambit is completely unacceptable – there’s no way to sugarcoat that. That’s what I was trying to get across.

    Who is on what side?

    By postulating that the Galápagos tortoises come from an island arc, you automatically postulate a list of other things with implications for the parsimony of your hypothesis. Instead of exploring that list, you just lean back and say you don’t know. That does not make sense.

    Perhaps so. My paper has its limits.

    . As you quoted, I’m talking about the hypothesis that this island arc was very close to a suitable continent at a suitable time. Do you know a paper about this?

    I am not so well acquainted with more recent literature. In the 2001 paper I referred to work by Pindell.

    one may compare the morphology of genetic development just as one may compare the morphology of DNA molecules
    Of course (even though you’re the first to call either of those “morphology”).

    I’m not so sure about that, but I cannot cite anything. I’ll ask some colleagues.

    You made it sound like there was evidence from embryology that the tinamous are not nested among the other paleognaths but form their sister-group; that’s why I asked if any study of these data had ever been done (I’m not aware of any, but it’s not impossible that I’d overlook one).

    I’m not so sure that I made it that way, but I understand that it read to you that way.

    …You need an outgroup to root the tree = to polarize your characters, except in very special cases like retrotransposon insertions maybe. This is true of parsimony, likelihood and Bayesian inference: they produce unrooted trees (outside of special cases), and then they root them on the outgroup. Distance methods, on the other hand, produce rooted trees. So I’m not sure I understand what you’re talking about.

    Since I use molecular phylogenies when I have no reason not and a conversation of these matters will be fragmentary in this format I will not pursue this further. I did go into this in some detail in my Grehan & Schwartz (2009) paper if you want to look at it.

    Alignment is of course a problem, and so is making a morphological data matrix. I don’t see what that has to do with phylogenetics vs. phenetics.
    Alighment is made by overall matching. That is phonetics as I understand it.

    Do you have a citation?

    I think I make reference to this problem in the Grehan & Schwartz paper.

    What do you think of three item analysis?
    Short answer: this. Long answer: I need to read up on it, but it seems to make several assumptions that fail to hold for most empirical datasets; this seems to be why almost nobody uses it. Or did you mean if I think it’s phenetics? I’m sure it’s not.

    OK. I have no strong opinions myself. Was just curious as to whether you had any opinion.

    The fossil record supports an Asian origin, and apparently so does molecular phylogenetics (Google just found me this).

    Sure, that is what the abstract claims. What do you regard as the evidence that supports such a view?

    So, from extant distributions you get Asia and/or North America as the most parsimonious place of origin

    What, precisely, do you see as evidence from the extant distributions?

    and the fossil record tilts this balance toward Asia.

    And what is it about the fossil record that you see pointing to this?

    Her model was the only explanation that had ever been proposed for her discoveries. It did not contradict any observations. So, it should have been tentatively accepted.

    Perhaps so.

    “Sure, sure. They’ve long stopped arguing about the basics, though. You’re trying to argue about the basics.”
    Perhaps so.
    And you leave it at that? Aren’t you curious to find out if you’re really arguing about the basics? Aren’t you curious to find out if the philosophers of science have stopped arguing about the basis?
    I don’t get this.

    Only so many hours in the day.

    Next test: does the timing fit? Were those terranes in the right area at the right time? Is there evidence that, say, the Galápagos tortoises came from elsewhere anyway?
    Fair enough questions.

    Link to this
    1. 453. John Scanlon FCD12:08 pm 07/7/2014
    Oh, I’d though from reading bits of Croizat that extant taxa didn’t really descend from ancestral individuals or species, but sort of crystalised out from ancestral collections of characters permanently smeared across the landscape.
    You might want to read more recent literature on ancestors rather than stopping there

    Decided life was too short to read 10,000 pages of his stuff, and couldn’t see any use for it in my work.
    It’s not for everybody (although one did not have to read all 10,000 pages)

    “As always, another possible explanation for this very beautiful pattern is chance dispersal.” M Heads
    Glad to hear it; yes, individuals actually exist, and they are not all nailed to the bedrock. There are theories and methods that deal with them extensively; Darwin was quite good on some of them. You know, tall chap, balding, long white beard?

    Need to be less obscure about the point you want to make.

    John Grehan

    Link to this
  456. 456. MichaelHeads 3:57 pm 07/7/2014

    450 Heteromeles wrote: if Africa is precluded from being the last home of the elephant bird stem clade due to the (inferred) presence of ostriches (which could be wrong, incidentally. For all we know, they evolved in China and spread west), then we’re left with a geological impossibility known as the Lost Continent of Lemuria, because you’re postulating a landmass that connected Madagascar and New Zealand, and not so incidentally, throwing Australia (and its rhea [?!]) out of the picture entirely.

    Sorry, the science has progressed to the point where you can download a program called Google Earth and study the sea floor all around every landmass. There’s plenty of evidence for plate tectonics all over the ocean floor, but there’s no evidence whatsoever for a land bridge between Madagascar and New Zealand at the time when the ratites were evolving (or likely, for that matter, at any time). Lemuria is a failed hypothesis, as is the lost continent of Mu in the Pacific. Renaming lost continents as panbiogeography won’t resurrect this idea, because anyone with a computer and Google Maps can disprove it in a few minutes.

    There’s no need to invoke Lemuria. Ratites, living and fossil, are known more or less globally and no-one would be surprised if they turned up as fossils in India/Sri Lanka or East Antarctica. In a standard reconstruction of Gondwana (e.g. Reeves 2014 J African Earth Sci) these areas give continuous land between Madagascar and New Zealand.

    Link to this
  457. 457. MichaelHeads 4:33 pm 07/7/2014

    451 David wrote

    The first time they mention vicariance, they say “continental vicariance”. If that doesn’t mean “vicariance by breakup”, what could it mean?

    No, that is what they mean. And I’m suggesting that you can have continental vicariance without breakup.

    For example, I wrote: ‘But there was extensive rifting in the Tasman Sea region and between West and East Antarctica for about 15 million years before the crust finally broke completely with seafloor spreading in the Tasman and the Southern Ocean’.

    ‘The pre-drift phase of rifting is not mentioned by Mitchell et al., but coincides spatially with the break between the trans-Indian Ocean ratite clade and the trans-Pacific clade’.

    You wrote: Interesting point. But why would a rift in a forested continent be a barrier to even a flightless bird?

    Barriers producing allopatric clades of birds have arisen somehow, even if the barriers are no longer evident. Despite the fact that they can fly, most birds are ‘extraordinarily sedentary’ (Mayr, 1940). (For this and other, similar quotes from very experienced ornithologists, see my 2012 book ‘Molecular panbiogeography of the tropics’, Univ California Press, p. 278).

    Uplift of mountain ranges such as the Andes, and the development of seaways are often cited as possible mechanisms for allopatric differentiation. But intra-continental rifts, such as the East African rift or the West Antarctic rift between East and West Antarctica, have also generated large-scale topographic features (major valleys, fault-scarps etc.) as well as large-scale volcanism. It’s also likely that these processes induced regional change in the vegetation.

    Link to this
  458. 458. MichaelHeads 4:49 pm 07/7/2014

    451 David wrote And why are there both moa and kiwi in New Zealand?

    Sorry, this is a key question and I forgot to respond. We explain allopatry by vicariance and overlap by dispersal. The trans-Indian basin clade (elephant birds, emus, cassowaries and kiwis) is broadly allopatric with the trans-Pacific clade (moas, rheas, tinamous). The only area for tens of thousands of kilometres in which the two clades overlap (living and fossil) is New Zealand.

    Our model suggests that the break between the two clades has developed in a widesprad Indo-Pacific ancestor, somewhere in or around what is now the New Zealand plateau (Zealandia). After the break developed, there has been limited, local dispersal around the break-zone, and this has blurred the precise break. The proposed dispersal has been by normal population expansion (probably mediated by geological or climatic change), not by a one-off, freak event unrelated to Earth history.

    Link to this
  459. 459. irenedelse 6:24 pm 07/7/2014

    @DavidMarjanovic, #451:

    “Uh, Darren, accusing someone of being the kind of person that was “putting the boot into McClintock” is a much greater insult than me showing how small my vocabulary really is. [...] The Galileo gambit is bad enough. Accusing specific people, on no evidence, of being on the other side of the Galileo gambit is completely unacceptable”

    And that’s why, for my part, I’ve stopped trying to discuss seriously with Mr. Grehan. At this point, it doesn’t matter which side of Hanlon’s Razor his contributions are coming, it’s virtually indistinguishable.

    “”Perhaps so.”

    And you leave it at that? Aren’t you curious to find out if you’re really arguing about the basics?”

    Exhibit B. Maybe we need a Voigt-Kampff test?

    Link to this
  460. 460. ectodysplasin 6:49 pm 07/7/2014

    Ultimately, the hypothesis that island arcs can provide delayed dispersal routes is an interesting one and probably one that is worthy of further study. However, I feel like Michael and John (or other members of their research program) need to do the scientific footwork of developing this as a hypothesis, and establish a set of conditions where this hypothesis is a valid alternative to standard vicariance and dispersal models and a methodology that actually permits rigorous testing between hypotheses under specific circumstances. As it is, I don’t think the methodology presented actually tests anything.

    I also think that the focus on ratites by these workers is not doing them any favors, because their entire argument there comes across as “well, there’s not a lot of data, therefore we are free to make any interpretation we’d like.” They’d be better off stating that ratites are difficult to assess and focusing on taxa for which island arc development presents a more credible dispersal mechanism.

    I don’t think either Michael nor John are bad, duplicitous, or ideological people. I think they’re entrenched in a specific biogeographic model and have not given a lot of thought to alternative models. As the saying goes, when all you have is a hammer, every problem looks like a nail.

    Link to this
  461. 461. irenedelse 7:40 pm 07/7/2014

    @MichaelHeads #458:

    “451 David wrote “And why are there both moa and kiwi in New Zealand?”

    Sorry, this is a key question and I forgot to respond. We explain allopatry by vicariance and overlap by dispersal. The trans-Indian basin clade (elephant birds, emus, cassowaries and kiwis) is broadly allopatric with the trans-Pacific clade (moas, rheas, tinamous). The only area for tens of thousands of kilometres in which the two clades overlap (living and fossil) is New Zealand.”

    I can see what you’re getting at. But what reason, apart from the newly elucidated phylogeny of kiwi and elephant birds, is there to distinguish an Indian Ocean clade and a Pacific basin clade? Or is it just the vocabulary used in the PBG context to describe the observed distribution? If so, I can see problems with implying in the descriptive words used the causal relationships you’re arguing for.

    Also, are Paleognaths the only clade where such a repatriation (one Afro-Australian clade in two distinct areas with a single NZ overlap) occurs? Wouldn’t a way to test your hypothesis be a comparison with other fauna and flora groups with species occurring in this Ocean Indian basin and this Pacific Ocean basin? Do we see the same pattern or is it more complex?

    Testing the hypothesis with terrestrial vertebrates would probably be a better test than birds, bats, vegetation relying on seeds or spores, little invertebrates and marine species, all of which also have a potential for air and/or water dispersion. If the existence of a more or less continuous domain, or even short lived land briges, between the former parts of Gondwana, played a significant part in their distribution, shouldn’t we see more clades of terrestrial vertebrates with a similar pan-Gondwanan distribution? Instead, we have on the one hand the Metatherians in Australia and South America, which is consistent with the fact that West Gondwana broke up during the early Cenozoic… But Australia and Zealandia had very few Eutherians appart from bats until the arrival of humans. Which is consistent with an absence of ecological and geological continuity. As David explained many comments ago, Madagascar got several Eutherian lineages from Africa during relatively short span of time after breaking up from it. South America exchanged fauna with Australia and Africa as long as it was connected to Antarctica. Afterwards, it didn’t completely stop but become rarer… And then there’s the high number of island endemics, both in the Indian and the Pacific Ocean emerged lands. Which argues for evolution in quasi-isolation on these land masses. Galapagos, the Seychelles, the Sub-Antarctic islands, New Caledonia…

    “Our model suggests that the break between the two clades has developed in a widesprad Indo-Pacific ancestor, somewhere in or around what is now the New Zealand plateau (Zealandia). After the break developed, there has been limited, local dispersal around the break-zone, and this has blurred the precise break. The proposed dispersal has been by normal population expansion (probably mediated by geological or climatic change), not by a one-off, freak event unrelated to Earth history.”

    Interesting, but what a complex scenario! There’s a number of assumptions in there that you wouldn’t need with the dispersion hypothesis. Which was an objection since the start of this interminable thread, and the reason Mitchell et al. supported dispersion in their paper.

    Link to this
  462. 462. Halbred 7:45 pm 07/7/2014

    It really bothers me that, time and time again, when evidence contrary to panbiogeography is brought up, John responds in one of two ways (to paraphrase):

    1. “I dunno;” or
    2. “It’s possible.”

    The most recent exchange about island tortoises has provided both examples. The former is used as a defensive or delaying statement, the latter a dismissive one. These two things, in addition to my earlier comment that John’s Jurassic ratite position is not falsifiable and therefore intellectually dishonest, make me extremely uncomfortable arguing anything with him.

    I’ve seen several points worth disputing, but he’s just going to answer in one of three ways: “I don’t know,” “it’s possible,” or “we have different definitions of X.”

    I mean, seriously, look back through his responses. A disturbing number of them can be boiled down to these three sentences. I can’t be only one who’s noticed this.

    Link to this
  463. 463. Grehan1896 7:46 pm 07/7/2014

    460. ectodysplasin
    Ultimately, the hypothesis that island arcs can provide delayed dispersal routes is an interesting one and probably one that is worthy of further study.

    I would be interested to know what ‘delayed dispersal routes’ mean in the context of island arcs.

    However, I feel like Michael and John (or other members of their research program) need to do the scientific footwork of developing this as a hypothesis, and establish a set of conditions where this hypothesis is a valid alternative to standard vicariance and dispersal models and a methodology that actually permits rigorous testing between hypotheses under specific circumstances. As it is, I don’t think the methodology presented actually tests anything.

    I would be very interested in your assessment of Mike’s two recent books.

    I also think that the focus on ratites

    It was a ratite paper that led to this particular conversation. Other groups have also been mentioned.

    by these workers is not doing them any favors, because their entire argument there comes across as “well, there’s not a lot of data, therefore we are free to make any interpretation we’d like.”

    I think that this interpretation is erroneous.

    They’d be better off stating that ratites are difficult to assess

    But they are not.

    and focusing on taxa for which island arc development presents a more credible dispersal mechanism.

    More credible that what mechanism?

    I don’t think either Michael nor John are bad, duplicitous, or ideological people. I think they’re entrenched in a specific biogeographic model and have not given a lot of thought to alternative models. As the saying goes, when all you have is a hammer, every problem looks like a nail.

    Given the history of literature on panbiogeography I would be interested for the justification for saying that we “have not given a lot of thought to alternative models”.

    John Grehan

    Link to this
  464. 464. Grehan1896 8:07 pm 07/7/2014

    462. Halbred
    It really bothers me that, time and time again, when evidence contrary to panbiogeography is brought up, John responds in one of two ways (to paraphrase):
    1. “I dunno;” or
    2. “It’s possible.”

    This is an assertion. It needs to be specific in order for me to respond. Please refer to some evidence posted that was contrary to panbiogeography where I responded with either or both. I do recall being asked questions about origins for which I did not have an answer and said so, and also some have made assertions about what they believe is true and I have agreed that their assertion may be possible.

    The most recent exchange about island tortoises has provided both examples. The former is used as a defensive or delaying statement, the latter a dismissive one.

    What was the ‘evidence’ contrary to panbiogeography for the tortoises that you are specifically referring to?

    These two things, in addition to my earlier comment that John’s Jurassic ratite position is not falsifiable and therefore intellectually dishonest, make me extremely uncomfortable arguing anything with him.

    One is never under any obligation to argue anything with anyone.

    I’ve seen several points worth disputing, but he’s just going to answer in one of three ways: “I don’t know,” “it’s possible,” or “we have different definitions of X.”

    Sometimes that is all that can be said on some issues.

    I mean, seriously, look back through his responses. A disturbing number of them can be boiled down to these three sentences. I can’t be only one who’s noticed this.

    Perhaps it is in the nature of assertions that are often being made.

    John Grehan

    Link to this
  465. 465. ectodysplasin 8:43 pm 07/7/2014

    @John,

    I would be interested to know what ‘delayed dispersal routes’ mean in the context of island arcs.

    This is, as far as I can tell, the hypothesis that you and Michael have forwarded, i.e. that faunas have arisen and dispersed along island arc systems (Galapagos, Polynesia, etc) without continental influx. In other words, when transoceanic dispersal does occur, it occurs via plate movements with respect to slow movement of hotspots and other such igneous features, rather than via instantaneous events that occur within an individual organism’s lifetime.

    I would be very interested in your assessment of Mike’s two recent books.

    I’m not a biogeographer and don’t have massive amounts of free time (I’m working on revisions of one paper, drafts of two more, and rewrites on a thesis). I’d be happy to read these and comment when I have the time, but that will not be soon.

    It was a ratite paper that led to this particular conversation. Other groups have also been mentioned.

    >>by these workers is not doing them any favors, because their entire argument there comes across as “well, there’s not a lot of data, therefore we are free to make any interpretation we’d like.”

    I think that this interpretation is erroneous.

    >>They’d be better off stating that ratites are difficult to assess

    But they are not.

    Of course ratites are difficult to assess. You have a very small number of data points and very little fossil record in either direction. You yourself have commented on the paucity of ratite fossils when it serves your purposes.

    What I am trying to get across here is that you’ve suggested novel mechanisms, but those mechanisms do not stand or fall on ratites, which are a group with very, very little useful data in general for the testing of any sort of biogeographical hypothesis.

    Given the history of literature on panbiogeography I would be interested for the justification for saying that we “have not given a lot of thought to alternative models”.

    I’m not interested in the history of literature on a specific school of thought. I’m interested in how a group of researchers approaching a specific set of data go about investigating that data. You seem to think that the appropriate first step to any set of occurrence data is to draw minimum spanning trees and compare with plate boundaries. To me, and to many others, this is methodologically deficient. You make no effort to state the assumptions of the method, or how the data meets those assumptions. You make no effort to exclude contributions of other mechanisms to the effect observed. You offer ad hoc justifications for disregarding results of other analyses that contradict your own results. And so on and so forth. You seem very wed to a specific method, even though you do not understand what the method actually does, how it treats occurrence data, or what the “results” actually indicate.

    Link to this
  466. 466. Grehan1896 10:34 pm 07/7/2014

    465. ectodysplasin
    I would be interested to know what ‘delayed dispersal routes’ mean in the context of island arcs.
    This is, as far as I can tell, the hypothesis that you and Michael have forwarded, i.e. that faunas have arisen and dispersed along island arc systems (Galapagos, Polynesia, etc) without continental influx. In other words, when transoceanic dispersal does occur, it occurs via plate movements with respect to slow movement of hotspots and other such igneous features, rather than via instantaneous events that occur within an individual organism’s lifetime.

    Thanks. I think that is a fairly good generalized description (not so much about slow movement of hotspots for example). But keep in mind that this is not just open speculation but based on analysis of biogeographic and tectonic patterns that leads to that model for particular situations and taxa.

    I would be very interested in your assessment of Mike’s two recent books.
    I’m not a biogeographer and don’t have massive amounts of free time (I’m working on revisions of one paper, drafts of two more, and rewrites on a thesis). I’d be happy to read these and comment when I have the time, but that will not be soon.

    Fair enough.

    But they are not.
    Of course ratites are difficult to assess. You have a very small number of data points and very little fossil record in either direction. You yourself have commented on the paucity of ratite fossils when it serves your purposes.

    I would have to go back and see what I have said about paucity of ratite fossils, but what I was referring to about ratites biogeography not being difficult to assess is that the predominance of vicariism (allopatry) is pretty obvious and it is pretty clear that there are two major baseline groups. This pattern along with correlations with some major tectonic features leads to a pretty straightforward vicariance model along with dispersal for the overlap of the two ocean basin groups.

    What I am trying to get across here is that you’ve suggested novel mechanisms

    Need to clarify ‘novel mechanisms’ please.

    but those mechanisms do not stand or fall on ratites, which are a group with very, very little useful datain general for the testing of any sort of biogeographical hypothesis.

    It depends on what constitutes a ‘test’. In the ratite case it may represent a test of the center of origin and dispersal model by showing the predominance of allopatry of genera and higher taxa along with phylogenetic and distributional relationships with tectonic features.

    Given the history of literature on panbiogeography I would be interested for the justification for saying that we “have not given a lot of thought to alternative models”.
    I’m not interested in the history of literature on a specific school of thought. I’m interested in how a group of researchers approaching a specific set of data go about investigating that data. You seem to think that the appropriate first step to any set of occurrence data is to draw minimum spanning trees and compare with plate boundaries.

    Well, there is a little more to it than that, but as you say, you are not a biogeographers and you do not have much spare time for the subject at this time.

    To me, and to many others, this is methodologically deficient.

    As said, there is more to the method than just that (panbiogeography really encompasses more than one method). But even with these other methods you may come to the same conclusion.

    You make no effort to state the assumptions of the method, or how the data meets those assumptions.
    See 202. Grehan1896

    You make no effort to exclude contributions of other mechanisms to the effect observed.

    Please explain.

    You offer ad hoc justifications for disregarding results of other analyses that contradict your own results.

    Please refer me to a specific example.

    You seem very wed to a specific method

    As if no one else on this list is?

    even though you do not understand what the method actually does, how it treats occurrence data, or what the “results” actually indicate.

    Your opinion is acknowledged.

    John Grehan

    Link to this
  467. 467. MichaelHeads 3:17 am 07/8/2014

    461 Irene wrote: I can see what you’re getting at. But what reason, apart from the newly elucidated phylogeny of kiwi and elephant birds, is there to distinguish an Indian Ocean clade and a Pacific basin clade?

    It’s easiest to see on a map of the clades (as far as I know, the only one is in my 2014 book, published before Mitchell et al appeared). Madagascar, Australia, New Guinea and NZ, vs. NZ and South Central America. If there’s a shorter or clearer way of describing the pattern than Indian vs. Pacific, let me know!

    Or is it just the vocabulary used in the PBG context to describe the observed distribution? If so, I can see problems with implying in the descriptive words used the causal relationships you’re arguing for.

    Also, are Paleognaths the only clade where such a repatriation (one Afro-Australian clade in two distinct areas with a single NZ overlap) occurs? Wouldn’t a way to test your hypothesis be a comparison with other fauna and flora groups with species occurring in this Ocean Indian basin and this Pacific Ocean basin? Do we see the same pattern or is it more complex?

    There are lots of examples of Indian – Pacific sister pairs in my 2014 book. The ones I mapped to introduce the concept (it’s not a complete list, which would fill a book!) were: Salvadoraceae (Indian) vs. Bataceae and Koeberliniaceae (Pacific)(eudicot plant families; Fig. 3.19); Afrocarpus + Nageia vs. Retrophyllum (podocarp trees, Fig. 3.20), 3.21 is ratites; Troidini butterflies (including the largest of all butterflies, in New Guinea): Pharmacophagus and Cressida (Indian) vs. Trigonoptera etc. (Pacific) (Fig. 3.22).

    Testing the hypothesis with terrestrial vertebrates would probably be a better test than birds, bats, vegetation relying on seeds or spores, little invertebrates and marine species, all of which also have a potential for air and/or water dispersion.

    Everything disperses. The Indian vs. Pacific thing isn’t really a hypothesis, it’s just a pattern you see in many, many groups if you spend your days looking at distributions and phylogenies. (Yes, I know, everything is ultimately a hypothesis).

    If the existence of a more or less continuous domain, or even short lived land briges, between the former parts of Gondwana, played a significant part in their distribution, shouldn’t we see more clades of terrestrial vertebrates with a similar pan-Gondwanan distribution? Instead, we have on the one hand the Metatherians in Australia and South America, which is consistent with the fact that West Gondwana broke up during the early Cenozoic… But Australia and Zealandia had very few Eutherians appart from bats until the arrival of humans. Which is consistent with an absence of ecological and geological continuity.

    The only terrestrial mammals in New Zealand are bats, which are often ignored because they fly, but if you actually look at their distributions, you find, for example, the following three clades with a phylogeny 1 (2 + 3):

    1. Myzopodidae (sucker-footed bats) (Madagascar and, as fossils, in Tanzania).

    2. Mystacinidae (New Zealand and, as fossils, in eastern Australia).

    3. Phyllostomidae and the much smaller families Mormoopidae, Noctilionidae, Furipteridae and Thyropteridae (northern Argentina to southernmost United States).

    (The group as a whole, Noctilionoidea, also includes an Oligocene fossil clade from Florida, but the affinities are uncertain).

    Clade 2 + 3 is a very common trans-Pacific range, and its sister in East Africa/Madagascar gives a typical trans-Indian link.

    As David explained many comments ago, Madagascar got several Eutherian lineages from Africa during relatively short span of time after breaking up from it.

    One of the most interesting patterns in the mammals of Madagascar is that carnivorans, primates, tenrecs, artiodactyls (extinct hippo) and shrews are all represented there by just a single clade (as with ratites). In the standard theory, they all dispersed just once, but why? E.g. why are there are lemurs but no monkeys, which can supposedly cross the Atlantic? Explaining these is very straightforward in a vicariance model.

    You quoted my post: “Our model suggests that the break between the two clades has developed in a widesprad Indo-Pacific ancestor, somewhere in or around what is now the New Zealand plateau (Zealandia). After the break developed, there has been limited, local dispersal around the break-zone, and this has blurred the precise break. The proposed dispersal has been by normal population expansion (probably mediated by geological or climatic change), not by a one-off, freak event unrelated to Earth history.”

    Interesting, but what a complex scenario!

    It’s not that complex! A break in a widespread ancestor followed by some local dispersal around the break. You can explain all the ratites in the same way, with minor dispersal blurring the boundary between kiwis and moas, and between rheas and tinamous.

    There’s a number of assumptions in there that you wouldn’t need with the dispersion hypothesis.

    The main assumption in our model is an original global ratite ancestor. The proposed vicariance events that break it up mean that you can predict major environmental breaks in the same areas and also breaks in other groups there.

    The dispersion model assumes a centre of origin, followed by one-off chance dispersal events unrelated to other groups or to environmental factors. (If they were related to environmental factors, e.g. wind direction, you would not expect singular events). This is an emerging theme in modern studies of long-distance dispersal, even by non-panbiogeographers. Thus there are no particular predictions, and its hard to test. Of course, this doesn’t mean it’s wrong.

    Carrying on from your last quote: Which was an objection since the start of this interminable thread, and the reason Mitchell et al. supported dispersion in their paper.

    The explicit reason they cited for supporting dispersion was that the phylogeny wasn’t congruent with the breakup sequence. Their model does involve assumptions, though, such as a centre of origin and allopatric differentiation by dispersal.

    Link to this
  468. 468. DavidMarjanovic 11:16 am 07/8/2014

    Alighment is made by overall matching. That is phonetics as I understand it.

    Well, there’s always POY, which tries to do alignment and phylogenetic analysis in one step instead of two; as several papers showed around 2006 (IIRC), however, it has other problems.

    Alignment is the process that generates the dataset; it’s not the phylogenetic analysis of the dataset. Morphological datasets aren’t made in the same step as phylogenetic analysis either.

    Sure, that is what the abstract claims. What do you regard as the evidence that supports such a view?

    What, precisely, do you see as evidence from the extant distributions?

    And what is it about the fossil record that you see pointing to this?

    Same answer to all three: use geographic distribution as an ordered (or more complex) multistate character and optimize it on the tree. What else would you suggest?

    “Her model was the only explanation that had ever been proposed for her discoveries. It did not contradict any observations. So, it should have been tentatively accepted.”

    Perhaps so.

    “Perhaps”, then, your assumption that I’d have tried to silence McClintock was unwarranted…?

    “Next test: does the timing fit? Were those terranes in the right area at the right time? Is there evidence that, say, the Galápagos tortoises came from elsewhere anyway?”

    Fair enough questions.

    As long as they aren’t answered, your hypothesis hasn’t been tested as well as it can reasonably be tested today. You should go do that instead of evangelizing for it anyway.

    We explain allopatry by vicariance and overlap by dispersal.

    Is that it? Is it that simple?

    Because I think it should be obvious that that’s a massive oversimplification of the real world and should not be used as a starting assumption.

    The only area for tens of thousands of kilometres in which the two clades overlap (living and fossil) is New Zealand.

    You’re making a bold statement about a very bad fossil record here. Elephantbirds have no fossil record – Madagascar doesn’t have any known terrestrial fossil record between the end of the Cretaceous and somewhere in the Pleistocene. Kiwis, and moas, have no pre-Pleistocene fossil record either, except for that one Miocene locality (St. Bathans). Antarctica hasn’t yielded any paleognaths at all yet, as far as I know. We currently have no way of testing if either or both of the clades inhabited East Antarctica.

    I repeat: the known fossil record of elephantbirds and kiwis together consists of one partial femur (holotype of Proapteryx micromeros) and possibly one quadrate (questionably referred to Proapteryx).

    Instead, we have on the one hand the Metatherians in Australia and South America, which is consistent with the fact that West Gondwana broke up during the early Cenozoic… But Australia and Zealandia had very few Eutherians appart from bats until the arrival of humans. Which is consistent with an absence of ecological and geological continuity. As David explained many comments ago, Madagascar got several Eutherian lineages from Africa during relatively short span of time after breaking up from it. South America exchanged fauna with Australia [...] as long as it was connected to Antarctica. Afterwards, it didn’t completely stop but become rarer…

    But, you see, as long as you’re willing to assume weirdly selective local extinction, all of this can be explained away if you just stretch your ghost lineages far enough. And that’s exactly what the panbiogeographers in this thread are proposing, ignoring the fact that absence of evidence in the better parts of the fossil record is evidence of absence. Australia boasts a lineage of mice (not rodents in general – specifically mice) since the Miocene and another since the Pliocene? Allopatry must be explained by vicariance rather than dispersal, so the great diversification of Muridae happened in the Cretaceous or Jurassic, and Pinky conspired with the Brain to keep them all out of the fossil record till much more recently. Easy. Madagascar has 5 lineages of placentals today and 0 in the Cretaceous (let alone earlier) fossil record? All of these, including crown-group hippos, must be projected back to the Jurassic, and a miracle happened to keep osteosclerotic hippo bones out of the fossil record worldwide till much more recently while stem hippos litter the fossil record in the Old World from ocean to shining ocean. Easy. Everything is easy if you simply declare negative evidence from the fossil record irrelevant to parsimony. And suddenly nothing is testable anymore.

    And then there’s the high number of island endemics, both in the Indian and the Pacific Ocean emerged lands. Which argues for evolution in quasi-isolation on these land masses. Galapagos, the Seychelles, the Sub-Antarctic islands, New Caledonia…

    Oh, that reminds me. Galápagos is a hotspot, unconnected to any continent except that I need to read up on that island arc. The islands between South America and Antarctica are an island arc that formed very soon after those two continents broke apart. New Caledonia and the Seychelles are continental splinters – and a small part of their fauna must have been there when they separated from India and New Zealand, respectively. The Seychelles in particular have an endemic clade of caecilians (part of Indotyphlidae) and an endemic clade of frogs (Sooglossidae); both have their sister-groups in India, both are bad (though not completely incapable) at overseas dispersal, and there’s no evidence from the very bad fossil record or from molecular dating or from biogeography or from geology to contradict the possibility that they’ve been on the Seychelles since the Seychelles broke off of India in the early Paleocene. Dispersal from India in significantly later times is hard to imagine given distance, currents and, as far as I know, absence of biota with a distribution that needs to be explained by dispersal from India.

    In addition to the sooglossids, however, the Seychelles are home to the hyperoliid frog Tachycnemis. That one I have to explain as the result of dispersal. Why? The clade formed by Sooglossidae and the Indian Nasikabatrachus is either the sister-group to Neobatrachia or to one of its two primary branches (Hyloidea, Ranoidea). Hyperoliidae, on the other hand, is nested deeply within Ranoidea – its sister-group is the African Arthroleptidae, and the sister-group to both together is formed by Brevicipitidae and Hemisus, both likewise African. Today, almost all hyperoliids live in Africa; the only exceptions are Heterixalus from Madagascar and Tachycnemis. The latter two, it turns out, are sister-groups and are deeply nested within Hyperoliidae (they aren’t the sister-group to the rest or most of the rest). Unfortunately the whole large clade has no known fossil record, which isn’t surprising for frogs that live in rainforests or (Hemisus) in soils. Either we have a Jurassic radiation that is somehow entirely confined to Africa, with no trace in South America or India, and with only one little clade that is very deeply nested present in Madagascar and the Seychelles; or we have overseas dispersal to Madagascar and then on to the Seychelles a few million years ago, a total of two inferred events that have been more or less paralleled by other frogs like Ptychadena mascareniensis or the Mayotte species of Mantidactylus and Boophis.

    I mean, seriously, look back through his responses. A disturbing number of them can be boiled down to these three sentences. I can’t be only one who’s noticed this.

    You aren’t the only one, and it annoys the living… profanity out of me. In fact, I’m writing this while sitting on the… whatever. :-)

    “You seem very wed to a specific method”

    As if no one else on this list is?

    Uh, the tu quoque argument is a logical fallacy. If we’re all making the same mistake as you (which remains to be demonstrated), it’s still a mistake.

    “even though you do not understand what the method actually does, how it treats occurrence data, or what the “results” actually indicate.”

    Your opinion is acknowledged.

    Is that all? No attempt to show that you do understand these things? No attempt to show that ectodysplasin has misunderstood something? Just “no comment”?

    That’s bizarre behavior for a scientist.

    The only terrestrial mammals in New Zealand are bats, which are often ignored because they fly, but if you actually look at their distributions, you find, for example, the following three clades with a phylogeny 1 (2 + 3):

    [...]

    Clade 2 + 3 is a very common trans-Pacific range, and its sister in East Africa/Madagascar gives a typical trans-Indian link.

    But can you distinguish middle Late Cretaceous vicariance from much later dispersal to NZ from nearby continents (Australia for Mystacinidae; Australia, with later local extinction, or possibly Antarctica for Myzopodidae)?

    One of the most interesting patterns in the mammals of Madagascar is that carnivorans, primates, tenrecs, artiodactyls (extinct hippo) and shrews are all represented there by just a single clade (as with ratites). In the standard theory, they all dispersed just once, but why? E.g. why are there are lemurs but no monkeys, which can supposedly cross the Atlantic? Explaining these is very straightforward in a vicariance model.

    This is just baffling. They only managed to disperse once because getting to Madagascar from Africa is simply difficult – it’s a very, very rare chance event. So is getting to South America from Africa, which monkeys (or indeed primates as a whole) managed to do a grand total of once, possibly on the same raft as the hystricognath rodents.

    Even ignoring all negative evidence from the fossil record, as you do, and even entertaining, as you do, the possibility that molecular divergence dates are systematically too young instead of too old (all evidence for a bias goes in the other direction, but never mind), vicariance has a much harder time explaining the placental fauna of Madagascar! You would need to assume a very high rate of weirdly specific extinctions in Madagascar and often elsewhere.

    Apparently I need to explain this in detail, so:
    Carnivora: All carnivorans of Madagascar belong to the endemic clade Eupleridae; different euplerids are convergent on herpestids, viverrids and/or felids, but they’re all euplerids. Now, a vicariance scenario would predict that Eupleridae is the sister-group to all other carnivorans, or perhaps the sister-group to an Indian clade. No such thing is the case. Eupleridae is deeply nested within Feliformia. How does that work? If the split between Eupleridae and Herpestidae happened in the Cretaceous or Jurassic, why aren’t there any felids or caniformians in Madagascar, and why aren’t there at least herpestids all over Gondwana?
    Primates: All known primates from Madagascar are lemuroids. Lemuroidea is nested deeper within Primates than most people think; its sister-group is Lorisoidea, today distributed from Africa to southeast Asia but absent from Madagascar. Stem-lorisoids and likely stem-lemuroids are known from the Eocene of northern Africa. Lemuroidea and Lorisoidea belong to Strepsirrhini; stem-strepsirrhines (“adapiforms”) are plentiful in the Cenozoic of the northern continents. – The sister-group of Strepsirrhini is Haplorrhini, and both the oldest monkeys and the oldest tarsiers are from the Eocene of Asia.
    Tenrecoidea/Afrosoricida: All the Malagasy ones form a clade nested within Tenrecidae; among the living at least, it’s the sister-group to Potamogalinae. Its relatives all the way to the base of Afrotheria are African. Why aren’t there potamogalines all over the rest of Gondwana, and why aren’t golden moles, aardvarks, hyraxes, proboscideans and, like, embrithopods all over Gondwana?
    Artiodactyla: Crown-group hippos are pretty deeply nested within Artiodactyla. Hippo intra- and interrelationships are surprisingly confusing, but apparently the oldest crown- or near-crown hippo is Kenyapotamus from the Miocene of, well, Kenya. If hippos are anthracotheres, things are easy: anthracotheres reached Africa in the Oligocene, and the oldest ones are from the Eocene of Asia. Asia is thoroughly unsurprising because Artiodactyla as a whole comes from Asia unless you dismiss all negative evidence from the whole fossil record. Better yet, Africa remained an artiodactyl-free zone till the semiaquatic anthracotheres managed to swim there, and those remained the only artiodactyls in Africa till the Oligocene was over and the Tethys closed! Instead of antelopes, for instance, Africa used to have antelope-shaped and -sized hyraxes. – In short, hippos swam to Africa quite recently, just like crocodiles did: in the Pliocene or Pleistocene.
    (Eu)lipotyphla: Oh. I didn’t even know about the shrew. There’s one species, which may be endemic (in that case it’s called Suncus madagascariensis) or may also occur from Spain to Vietnam and Malaysia (in which case the populations on Madagascar belong to Suncus etruscus). Well, if people can’t even agree if it’s a separate species, it must be very young. I even have to wonder about accidental introduction by humans! Vicariance is a plainly ridiculous hypothesis in this case.
    Rodentia: All rodents native to Madagascar belong to the endemic clade Nesomyinae. The other nesomyids (which seem to form a clade) are all African, and Nesomyidae is fairly deeply nested within Muroidea. Apparently there weren’t any muroids in Africa before the Miocene; in Laurasia they go back to the middle Eocene. The origin of Rodentia (and that of Lagomorpha, its sister-group among the living) is pretty well documented in the Paleocene of Asia. Instead of rodents, Madagascar had gondwanatheres at the end of the Cretaceous.

    What alternative to dispersal to you propose?

    If they were related to environmental factors, e.g. wind direction, you would not expect singular events).

    For Madagascar, all proposed dispersal events I know of go in the same direction: from Africa to Madagascar, and then sometimes from Madagascar to the Comoros or Mascarenes or Seychelles. – But I don’t agree with your blanket statement. Whether you should expect a singular event depends on how probable the event is and how much time is available.

    Link to this
  469. 469. Heteromeles 12:36 pm 07/8/2014

    @467: So far as plants go, there’s some evidence that podocarps are vicariant, because they’re an old group, and they disperse primarily by fleshy-fruited seeds, which aren’t the greatest mechanism for long-distance overwater dispersal (reference: Carlquist’s Island Biology, among others. This is endozoochory vs. ectozoochory). Indeed, one reference towards a biogeography of the Podocarpaceae pulled up from the web notes that ” Endozoochory is important for present short- and medium-distance dispersal; however, past long-distance dispersal events, leading to some present disjunct distribution patterns, are more likely to have resulted through rafting on microcontinental blocks and their subsequent accretion on to other land masses. Malesia and Australasia have the greatest diversity of living podocarps but, because podocarps only immigrated to Malesia from the south in relatively recent geological times (principally post-Miocene) and there neither are, nor have been, any endemic genera there, Malesia is regarded as a refugium in which there has been adaptive radiation in response to edaphic and insular factors.”

    Still, there’s nothing that precludes a fleshy-fruited from overwater dispersal. It’s worth remembering that there are two native species of Rubus in Hawaii (R. hawaiensis and R. macraei), so it’s possible for a group that only produces fleshy fruits to colonize the most remote archipelago on the planet. Given that Rubus seeds are known to survive for decades before germinating, this isn’t surprising. Indeed, this is one reason why researchers like Sherwin Carlquist tested seeds to see how long they could survive saltwater immersion, because that’s a better indicator of how well a plant can disperse in the water than is fruit type.

    Batis (Bataceae) is a salt marsh genus of salt marsh species. In general, I’d pitch any forebeach (strand) or salt marsh species out of any argument of vicariance, because their seeds float and they grow in tidal areas connected to oceans. They’re overwater dispersal specialists.

    Koeberlinia spinosa (aka allthorn) is the only member of its monotypic genus from the deserts in Arizona to Texas and Mexico. There’s no vicariance story here.

    Link to this
  470. 470. Heteromeles 12:47 pm 07/8/2014

    @462: It really bothers me that, time and time again, when evidence contrary to panbiogeography is brought up, John responds in one of two ways (to paraphrase):

    1. “I dunno;” or
    2. “It’s possible.”

    Yep. I noticed that a long time ago. That’s why I made a few remarks about “Stultus saxum” that were summarily brushed off by people who were more obsessed with engaging, rather than walking away from what has looked like a baiting exercise for quite some time.

    To me, the best part of TetZoo is when people actually learn stuff. There’s no learning going on here, and there hasn’t been since before comment 150. The only reason I’m staying engaged is that it seems to help Darren somehow (perhaps number of comments=renewed contract with SciAm?), so I’ll keep dutifully piling on.

    Link to this
  471. 471. Grehan1896 12:52 pm 07/8/2014

    468. DavidMarjanovic

    David,

    Some very interesting observations covering a lot of ground so I will address them separate installments as time permits.

    Alignment is made by overall matching. That is phenetics as I understand it.
    Well, there’s always POY, which tries to do alignment and phylogenetic analysis in one step instead of two; as several papers showed around 2006 (IIRC), however, it has other problems.

    Agreed, and it also does not eliminate the phonetic nature of base pair matching.

    Alignment is the process that generates the dataset; it’s not the phylogenetic analysis of the dataset. Morphological datasets aren’t made in the same step as phylogenetic analysis either.

    Agreed on both counts. But the problem is in how the data set is generated in each case. The alignment process creates base pair homologies that do not actually exist in nature, but are theorized to have existed based on one or other best fit models. In morphology it is possible to directly observe character states and to identify those that are absent in the outgroup. I would agree that this process involves judgments about similarity and homology so mistakes may occur (but this does not mean that such phylogenetic information does not stand up as valid in its own right). My overall view is that neither morphogenetic or molecular approaches are infallible, and neither automatically disqualifies the other when they are incongruent for a particular group. Mike’s books are, for me, the best demonstration that molecular approaches, despite their phenetic content, do generate historically informative patterns of relationship.

    Sure, that is what the abstract claims. What do you regard as the evidence that supports such a view?
    What, precisely, do you see as evidence from the extant distributions?

    And what is it about the fossil record that you see pointing to this?
    Same answer to all three: use geographic distribution as an ordered (or more complex) multistate character and optimize it on the tree. What else would you suggest?

    You are describing the generalities of the steps involved, but what I am asking is what exactly do you see in the extant distributions as evidence necessitating an Asian origin, and what in the fossil record necessarily points to this?

    John Grehan

    Link to this
  472. 472. ectodysplasin 1:38 pm 07/8/2014

    John, I think you fundamentally misunderstand how data is generated, how one goes about assessing signal in data, and how similarity methods actually work.

    I am confident that I can propose a dataset for which your method will produce positively misleading results, for example.

    Link to this
  473. 473. Grehan1896 1:57 pm 07/8/2014

    468. DavidMarjanovic
    I’ve since noticed that most of the comments are directed to Mike. Those remaining that are directed to me are:

    “Her model was the only explanation that had ever been proposed for her discoveries. It did not contradict any observations. So, it should have been tentatively accepted.”
    Perhaps so.
    “Perhaps”, then, your assumption that I’d have tried to silence McClintock was unwarranted…?

    No.

    “Next test: does the timing fit? Were those terranes in the right area at the right time? Is there evidence that, say, the Galápagos tortoises came from elsewhere anyway?”
    Fair enough questions.
    As long as they aren’t answered, your hypothesis hasn’t been tested as well as it can reasonably be tested today. You should go do that instead of evangelizing for it anyway.

    Sorry, you did not pose the questions in a way that did not read as rhetorical. So for answers:

    1. Were those terranes in the right area at the right time?

    Apparently – at least here is nothing to suggest that they were not.

    2. Is there evidence that, say, the Galápagos tortoises came from elsewhere anyway?”

    No.

    John Grehan

    Link to this
  474. 474. Grehan1896 3:04 pm 07/8/2014

    468. DavidMarjanovic

    “You seem very wed to a specific method”
    As if no one else on this list is?
    Uh, the tu quoque argument is a logical fallacy. If we’re all making the same mistake as you (which remains to be demonstrated), it’s still a mistake.

    Mistake or not, everyone who has a strong opinion on such matters is wedded to a specific method.

    “even though you do not understand what the method actually does, how it treats occurrence data, or what the “results” actually indicate.”
    Your opinion is acknowledged.
    Is that all? No attempt to show that you do understand these things? No attempt to show that ectodysplasin has misunderstood something? Just “no comment”?
    That’s bizarre behavior for a scientist.

    Well of course I accept your right to have an opinion about what is bizarre or not for a scientist. Same goes for the assertions about what I do or do not understand. It’s just an expression of opinion and I acknowledged that. I did not feel I needed to say anything more. People are entitled to their opinions.

    John Grehan

    Link to this
  475. 475. Stripeycat 3:18 pm 07/8/2014

    Some of the rhetoric problem seems to come from the fact that most people will interpret statements like “your opinion is noted”, “thankyou for your opinion” and “you’re entitled to your own opinion” as silencing tactics in a debate. This is because, de facto, you are declining to engage with them, whatever your reason. Also note that dismissing a statement of known facts as “opinion” because you don’t see the relevence of the facts is disengenious – if you need a further explanation of the argument, ask for it.

    The overall effect translates as somewhere between “shut up and let the grownups talk”, and “go jump in a lake”. And you wonder why it provokes hostility!

    To make matters worse, it’s also a common tactic of known nutters like creationists and alt-medders, and the DP who shall not be named. While technically this is not relevant to your case, using the same rhetorical devices without logically sound arguments will inevitably cause a halo (or, in this instance, devil horns) effect in our minds.

    Rhetoric out of the way, I’ll try to come up with some science comments next.

    Link to this
  476. 476. Stripeycat 3:46 pm 07/8/2014

    John #360
    With respect to moas and kiwis I did not intentionally indicate that one would assign ocean basins without having some prior assumption about their relationships to other groups. But it’s not just their relationships to other groups, but where those other groups are located with respect to major tectonic features that result in the different baselines.

    And Mike #458

    Sorry, this is a key question and I forgot to respond. We explain allopatry by vicariance and overlap by dispersal. The trans-Indian basin clade (elephant birds, emus, cassowaries and kiwis) is broadly allopatric with the trans-Pacific clade (moas, rheas, tinamous). The only area for tens of thousands of kilometres in which the two clades overlap (living and fossil) is New Zealand.

    Our model suggests that the break between the two clades has developed in a widesprad Indo-Pacific ancestor, somewhere in or around what is now the New Zealand plateau (Zealandia). After the break developed, there has been limited, local dispersal around the break-zone, and this has blurred the precise break. The proposed dispersal has been by normal population expansion (probably mediated by geological or climatic change), not by a one-off, freak event unrelated to Earth history.

    Neither of you will explain *why* you assign moas to the Pacific and kiwis to the Indian Ocean Basins. I grasp that you have an MST with a group in each basin, but I do not understand how, unless it’s on the basis of the phylogeny, you choose which goes to which clade. If it is on the basis of the phylogeny, you cannot then claim that the fact that the phylogeny and your conclusions coincide is evidence to strengthen your claim.

    John #360-361 re: galapagos hotspot.

    First of all, thankyou for the link to your paper – I’d got the wrong end of the stick about some things you’d said here, that the paper clarified.

    However, we still don’t have evidence for what’s going on at the hotspot between c76 and c17-20mya, because the relevant bits of sea-floor have probably been destroyed. You simply cannot hypothesise that there was a continuous procession of archipelagoes, use that as support for Galapagos vicariance, and then claim vicariance as evidence for the hotspot’s behaviour. Do you not see why this is circular reasoning?

    Unless there’s something fundamental you’ve forgotten to mention, the moa-kiwi situation exhibits exactly the same fallacy. This is basic logic, not some obscure aspect of philosophy of science; and the fact that you don’t seem to grasp this (and haven’t had it pointed out by reviewers) says scary things about your field and your conclusions as presented here.

    Link to this
  477. 477. ectodysplasin 4:00 pm 07/8/2014

    Ultimately, I think there’s only so much that can be said to people who think that the Pacific Ocean basin opened up in the Jurassic.

    Link to this
  478. 478. Grehan1896 4:50 pm 07/8/2014

    469. Heteromeles

    @467: So far as plants go, there’s some evidence that podocarps are vicariant, …….podocarps only immigrated to Malesia from the south in relatively recent geological times (principally post-Miocene) and there neither are, nor have been, any endemic genera there, Malesia is regarded as a refugium in which there has been adaptive radiation in response to edaphic and insular factors.”

    But none of this addresses the point made by Mike, that podocarps are part of an Indian ocean-Pacific sister group pair. From the quoted section it seems that the authors are not aware of that. What is biogeographically interesting is that the Podocarpacea comprise five basal clades around thePacific and one pantrocial clade. Podocarpus itself is made up of two groups, the trans-Indian Ocean ‘Polypodiopsis clade’ (Afrocarpus and Nageia) and a trans-Pacific clade (Retrophyllum. The two clades overlap only in the Maluki Isands and mainland New Guinea.

    . It’s worth remembering that there are two native species of Rubus in Hawaii (R. hawaiensis and R. macraei), so it’s possible for a group that only produces fleshy fruits to colonize the most remote archipelago on the planet.

    By colonization do you mean that they dispersed across the oceans from some adjacent continent? If so, what evidence do you present for that conjecture?

    Given that Rubus seeds are known to survive for decades before germinating, this isn’t surprising. Indeed, this is one reason why researchers like Sherwin Carlquist tested seeds to see how long they could survive saltwater immersion, because that’s a better indicator of how well a plant can disperse in the water than is fruit type.

    But there are organisms that have trans-oceanic distributions without having the ability to survive saltwater immersion for a long time, so the ability is itself not a predictor of biogeography.

    Batis (Bataceae) is a salt marsh genus of salt marsh species. In general, I’d pitch any forebeach (strand) or salt marsh species out of any argument of vicariance, because their seeds float and they grow in tidal areas connected to oceans. They’re overwater dispersal specialists.

    That might be ok in theory, but then why does the group have the kind of distribution that it does and how do their floating seeds and tidal area growth explain their vicariance with respect to Indian and Pacific clade sister groups?

    Koeberlinia spinosa (aka allthorn) is the only member of its monotypic genus from the deserts in Arizona to Texas and Mexico. There’s no vicariance story here.

    Not by itself of course, but Koeberlinia is mostly vicariant to the Bataceae, and it has a disjunct outlier in Boliva. Why not first take a look at the figures Mike mentioned before offering the critique so you could make sure you understood the spatial structure of the patterns involved?

    John Grehan

    Link to this
  479. 479. irenedelse 5:25 pm 07/8/2014

    @MichaelHeads #467:

    “”Testing the hypothesis with terrestrial vertebrates would probably be a better test than birds, bats, vegetation relying on seeds or spores, little invertebrates and marine species, all of which also have a potential for air and/or water dispersion.”

    Everything disperses. The Indian vs. Pacific thing isn’t really a hypothesis, it’s just a pattern you see in many, many groups if you spend your days looking at distributions and phylogenies. (Yes, I know, everything is ultimately a hypothesis).”

    I should have said “a higher potential for long-distance dispersion”: non volant terrestrial animals which are bigger than a rat are unlikely to get carried on the wind by a storm, or to cling to a vegetation raft borne by sea currents. Plants that don’t disseminate by seeds and spores would also be interesting (that’s why the sweet potato cultiva from South America occurring in Polynesia could only have come there through human actions).

    The idea is to test the dispersion vs vicariance hypothesis and see of the pattern is the same as for birds, bats, aquatic & semi-aquatic animals and various plants. If the patterns were similar for less easily dispersed species, it would plead in favour of biotic exchanges through land connections. (As counter-examples, consider what happened in South and North America when they became connected, or the case of the Bering land bridge between Eurasia and North America during Pleistocene glaciations.) Focusing on Paleognaths or Nothofagus doesn’t help to discriminate between Gondwana relics and dispersion along the trade winds and westerlies.

    I’m glad to see Heteromeles at #469 is on the same page. Thanks for the example of marsh plants adapted to brackish water or saltwater!

    By the way, David Marjanovic discussed at #468 the case of New Caledonia and its separation from Zealandia. Contrary to New Zealand, it does have a fossil record from the Cenozoic. And if it did have big, flightless, recently extinct birds, they were not Moa or any kind of Paleognaths, but Sylviornis, a flightless Galliforme species (even if it was first identified as a ratite, and some websites still have the erroneous classification). There were no Kiwi there, either, and no Emus or Cassowaries or Rheas. That should give one pause when suggesting scenarios for Paleognath evolution relying on Gondwanan refugia and island arcs.

    Link to this
  480. 480. irenedelse 5:34 pm 07/8/2014

    @John #478:

    “That might be ok in theory, but then why does the group have the kind of distribution that it does and how do their floating seeds and tidal area growth explain their vicariance with respect to Indian and Pacific clade sister groups?”

    What *do* floating seeds and growing on the seashore has to do with dispersal over water, indeed…

    Link to this
  481. 481. Heteromeles 6:36 pm 07/8/2014

    @479 and @460: Here you go again.

    Go read up on Rubus and Batis (and podocarps, for that matter) and come back with positive evidence, not weak rhetorical comebacks, that seriously back up your contention that evidence of massively good dispersal ability does not automatically mean that dispersal is the most parsimonious explanation for why we find these species where we find them. Appeals to finding species of limited dispersal ability across wide areas are moot. Of course vicariance happens. So what?

    As a contrast to the pointless rhetorical tricks I’ve gotten tired of, I’d suggest that (following what Irene said above), you’re not human, you’re a Turing Test candidate program, designed to generate indefinite debates on a fringe scientific paradigm, in the hopes that “you” will impress the gullible that”you’re” human so that “you” can charge them to read “your” papers, so that “you” can gain credibility, funding, and tenure for the computer scientists who created “you.” If so, bravo to your creators.

    How are you going to prove you’re human? Anyone can swipe a photo and a biography from the internet. What about what you’re writing here could make anyone think there’s a human at the other end of the conversation?

    Link to this
  482. 482. Grehan1896 7:05 pm 07/8/2014

    475. Stripeycat
    Some of the rhetoric problem seems to come from the fact that most people will interpret statements like “your opinion is noted”, “thankyou for your opinion” and “you’re entitled to your own opinion” as silencing tactics in a debate.

    Yes, when people just state an opinion or assertion all I can do is acknowledge that. If the opinion includes something of substance that I can work with I will.

    This is because, de facto, you are declining to engage with them, whatever your reason.

    Because arguing about opinions does not, in my experience, get very far. Someone might, for example, have the opinion that moas could swim to New Zealand. Without anything else to go on I would not engage with that opinion.

    Also note that dismissing a statement of known facts as “opinion” because you don’t see the relevence of the facts is disengenious – if you need a further explanation of the argument, ask for it.

    I often do.

    The overall effect translates as somewhere between “shut up and let the grownups talk”, and “go jump in a lake”. And you wonder why it provokes hostility!

    I was not aware of that.

    To make matters worse, it’s also a common tactic of known nutters like creationists and alt-medders, and the DP who shall not be named. While technically this is not relevant to your case, using the same rhetorical devices without logically sound arguments will inevitably cause a halo (or, in this instance, devil horns) effect in our minds.

    I sympathize with your plight.

    Rhetoric out of the way, I’ll try to come up with some science comments next.

    Good

    John Grehan

    Link to this
  483. 483. Grehan1896 7:21 pm 07/8/2014

    476 stripycat

    Neither of you will explain *why* you assign moas to the Pacific and kiwis to the Indian Ocean Basins. I grasp that you have an MST with a group in each basin, but I do not understand how, unless it’s on the basis of the phylogeny, you choose which goes to which clade. If it is on the basis of the phylogeny, you cannot then claim that the fact that the phylogeny and your conclusions coincide is evidence to strengthen your claim.

    As said before, sister group of the kiwi is on the other side of the Indian Ocean, sister group of moa is on the other side of the Pacific Ocean. Phylogeny provides the biological relationship; spatial analysis provides the biogeographic relationship. I can appreciate that this may be a bit difficult to absorb if one is not well acquainted with the field of biogeography.

    John #360-361 re: galapagos hotspot.
    First of all, thankyou for the link to your paper – I’d got the wrong end of the stick about some things you’d said here, that the paper clarified.

    It’s probably almost impossible to give a clear rendition of something like this in disjointed postings, so glad the paper was helpful.

    However, we still don’t have evidence for what’s going on at the hotspot between c76 and c17-20mya, because the relevant bits of sea-floor have probably been destroyed.

    Correctly, one may not have the stratigraphic evidence. There is biogeographic evidence.

    You simply cannot hypothesise that there was a continuous procession of archipelagoes, use that as support for Galapagos vicariance,

    Agreed!!! And of course, that was not done.

    and then claim vicariance as evidence for the hotspot’s behaviour. Do you not see why this is circular reasoning?

    It would be if that were the way it was argued. But it is the vicariance evidence that leads to the prediction that there was a continuity of Galapagos archipelagos that allowed for survival of the colonists from earlier landscapes.

    Unless there’s something fundamental you’ve forgotten to mention, the moa-kiwi situation exhibits exactly the same fallacy. This is basic logic, not some obscure aspect of philosophy of science; and the fact that you don’t seem to grasp this (and haven’t had it pointed out by reviewers) says scary things about your field and your conclusions as presented here.

    This is a good example of an opinion being presented without anything to go on. How does the moa-kiwi situation exhibit the same fallacy (especially as the fallacy is non existent since there was no hypothesis of continuous land that was used to support vicariance that was then used as evidence for the geology.

    John Grehan

    Link to this
  484. 484. MichaelHeads 7:29 pm 07/8/2014

    I wrote: ‘We explain allopatry by vicariance and overlap by dispersal’.

    David replied (468): Is that it? Is it that simple?

    Pretty much. Extinction is a another factor, but happens everywhere, all the time, and doesn’t account for the neat, mosaic patterns of allopatry that are observed in practice. E.g. in the ratites, the global allopatry of the clades is the dominant pattern. The comparatively minor areas of overlap can be explained by dispersal. The absence from India and the Antarctica mainland (living and fossil) is presumably the result of extinction. The absence from New Caledonia and all the Caribbean islands is more enigmatic but very interesting (Grenada is just 100 km from Trinidad, where there are tinamous).

    David, carrying on from his last quote, wrote: Because I think it should be obvious that that’s a massive oversimplification of the real world and should not be used as a starting assumption.

    I think the best thing to do with any new method is to try it out and see how it works, on a group that you know well, rather than just writing it off a priori.

    I noted: The only area for tens of thousands of kilometres in which the two clades [trans-Indian and trans-Pacfic] overlap (living and fossil) is New Zealand.

    You’re making a bold statement about a very bad fossil record here.

    No, I’m just stating the facts as we know them. Obviously, most of the data is from the extant record, but I wanted to emphasize that we do not ignore the fossil record, because our critics often suggest that we do. The fact that the groups have such a large, global range but such a small area of overlap is interesting and needs to be explained somehow.

    Elephantbirds have no fossil record – Madagascar doesn’t have any known terrestrial fossil record between the end of the Cretaceous and somewhere in the Pleistocene. Kiwis, and moas, have no pre-Pleistocene fossil record either, except for that one Miocene locality (St. Bathans). Antarctica hasn’t yielded any paleognaths at all yet, as far as I know. We currently have no way of testing if either or both of the clades inhabited East Antarctica.

    You could test it by continuing to look for fossils in Antarctica, which people do. (There’s ratite material from Seymour Island off the Antarctic Peninsula, but apparently nothing more can be said about its affinities).

    On bats, I mentioned that clade 2 (Mystacinidae) + 3 (Phyllostomidae etc. in America) is a very common trans-Pacific range, and its sister (Myzopodidae) in East Africa/Madagascar gives a typical trans-Indian link.

    But can you distinguish middle Late Cretaceous vicariance from much later dispersal to NZ from nearby continents (Australia for Mystacinidae; Australia, with later local extinction, or possibly Antarctica for Myzopodidae)?

    Based on fossil calibrations, Mystacina and its sister group diverged at 35-68 Ma (Lloyd, 2003) or 41-51 Ma (Teeling et al., 2005), and as minimum dates these are compatible with Cretaceous vicariance.

    The dipteran family Mystacinobiidae has a single, wingless genus, Mystacinobia, which is only found in association with Mystacina and feeds mainly on the guano. The sister-group of Mystacinobia is the undescribed ‘McAlpine’s fly’ from south-eastern Australia, and the two form a Tasman basin clade that is sister to a worldwide complex, the family Sarcophagidae; see p. xxx).
    The two trans-Tasman groups – the bat family Mystacinidae and the pair of flies – can both be derived by simple vicariance in widespread ancestors. In the case of the bats the ancestor was panaustral, while in the case of the flies the ancestor was global. If the ancestors attained their wide global range during mid-Cretaceous phases of high sea-level, they could then have been fractured during the pre-breakup rifting of Gondwana.

    As with the fly Mystacinobia, the mite Chirophagoides (Sarcoptidae) is known only in association with Mystacina. Its sister-group comprises Chirnyssoides of tropical America, plus Notoedres, a diverse group found worldwide (but known in New Zealand only from introduced mammals) (Klompen, 1992). As with Mystacinobia and ‘McAlpine’s fly’, Chirophagoides cannot be derived from anywhere outside the Tasman region without invoking extra ad hoc hypotheses, but its affinities are consistent with simple vicariance between the Tasman region and the rest of the world.

    The simplest explanation for the affinities of the three Tasman region groups (Tasman vs. world, or Tasman vs. panaustral)is that these occurred before Gondwana breakup, consistent with a break during pre-drift rifting (for the dozens of other ‘globally basal’ groups in the Tasman region, see my 2014 Australasia book). This is also consistent with the trans-Tasman affinities being the result of Tasman rifting.

    On Madagascar I wrote: One of the most interesting patterns in the mammals of Madagascar is that carnivorans, primates, tenrecs, artiodactyls (extinct hippo) and shrews are all represented there by just a single clade (as with ratites). In the standard theory, they all dispersed just once, but why? Explaining these is very straightforward in a vicariance model.

    [I should have said single *monophyletic* clade on Madagascar - sorry for the mistake, but you got my drift].

    David replied: This is just baffling. They only managed to disperse once because getting to Madagascar from Africa is simply difficult – it’s a very, very rare chance event.

    It’s not just very, very rare, it’s unique – it only happened once in tens of millions of years (or however old you accept that artiodactyls, tenrecs, primates etc. are). It happened precisely once in each group, despite their very different ecologies and means of dispersal. What ecological event only happens once in tens of millions of years? Even a very, very rare ecological event, such as a once-in-a-thousand-years storm, would have occurred millions of times. Apparently there were even currents flowing to Madagascar until, from memory, the Eocene(Ali and Aitchison in Nature). So why no monkeys on Madagascar, even though they could cross the Atlantic? Why no strepsirrhines (lemurs etc.) in America, even though they could cross to Madagascar? Why have no non-human primates – monkeys, strepsirrhines or tarsiers – ever crossed the stretch of sea (Salue Timpaus Strait, just 20 km wide!) that separates Sulawesi from the Sula Islands and Australasia? (For much more detail on the primates, including their fossil record, see my 2012 book ‘Molecular panbiogeography of the tropics’).

    Link to this
  485. 485. Grehan1896 7:31 pm 07/8/2014

    481. Heteromeles
    @479 and @460: Here you go again.
    Go read up on Rubus and Batis (and podocarps, for that matter) and come back with positive evidence, not weak rhetorical comebacks, that seriously back up your contention that evidence of massively good dispersal ability does not automatically mean that dispersal is the most parsimonious explanation for why we find these species where we find them. Appeals to finding species of limited dispersal ability across wide areas are moot. Of course vicariance happens. So what?

    So what precisely is problem you see with my statements about Rubus, Batis, and podocarps?

    As a contrast to the pointless rhetorical tricks I’ve gotten tired of, I’d suggest that (following what Irene said above), you’re not human, you’re a Turing Test candidate program, designed to generate indefinite debates on a fringe scientific paradigm, in the hopes that “you” will impress the gullible that”you’re” human so that “you” can charge them to read “your” papers, so that “you” can gain credibility, funding, and tenure for the computer scientists who created “you.” If so, bravo to your creators.
    How are you going to prove you’re human? Anyone can swipe a photo and a biography from the internet. What about what you’re writing here could make anyone think there’s a human at the other end of the conversation?

    I guess this is just another example of a series of assertions that I do not see the point of engaging with. I can understand that what I provide in the way of explanations may not be seen as satisfactory to some, but the converse may also be true. As Patton once said (maybe slightly paraphrased), if everyone is thinking alike then no one is thinking.

    John Grehan

    Link to this
  486. 486. MichaelHeads 7:44 pm 07/8/2014

    Ectodysplasin (477) wrote: ‘Ultimately, I think there’s only so much that can be said to people who think that the Pacific Ocean basin opened up in the Jurassic’.

    The Pacific basin is formed, for the most part, by the Pacific plate. This originated at a point (a triple junction, located in the Cook Islands region) in the Jurassic. Since then it has expanded dramatically by seafloor spreading to occupy the entire region between New Zealand and California. (For reconstructions, see Fig. 6-1 in my 2012 book ‘Molecular panbiogeography of the tropics’. Univ. California Press).

    Link to this
  487. 487. MichaelHeads 8:15 pm 07/8/2014

    Heteromeles wrote (469): ‘It’s worth remembering that there are two native species of Rubus in Hawaii (R. hawaiensis and R. macraei), so it’s possible for a group that only produces fleshy fruits to colonize the most remote archipelago on the planet’.

    The archipelago is remote now, but it is surrounded by flat-topped seamounts (e.g. Musicians to the north, Lines to the south) that are submerged islands. Obviously, fresh lava is colonized from somewhere else, but in Hawaii it’s not necessarily from the continents (as E.O. Wilson thought) or even from what are now the nearest islands.

    Batis (Bataceae) is a salt marsh genus of salt marsh species. In general, I’d pitch any forebeach (strand) or salt marsh species out of any argument of vicariance, because their seeds float and they grow in tidal areas connected to oceans. They’re overwater dispersal specialists.

    So why are Batacaeae restricted to the Pacific Ocean and why do they show such precise allopatry with their Indian Ocean sister group, Salvadoraceae? Why, in turn, is this Indo-Pacific clade almost entirely allopatric with its sister group, Koeberliniaceae (minor overlap in NW Mexico)?

    The general question is: Why do groups with excellent means of oceanic dispersal show such similar patterns to groups with very poor means of dispersal? For example, albatrosses and their relatives are some of the best dispersers, but the breeding records show clear patterns of allopatry that are repeated in many other groups. Their natal philopatry is key here, and if, say, an island arc with breeding petrels is rifted apart, the group may well differentiate (see Fig. 8-10 of the Pacific petrels in my 2012 book Molecular panbiogeography. On the biogeography of marine groups, see my review in Biol J Linn Soc 84: 675. 2005).

    Link to this
  488. 488. MichaelHeads 8:39 pm 07/8/2014

    476 stripeycat wrote: Neither of you will explain *why* you assign moas to the Pacific and kiwis to the Indian Ocean Basins. I grasp that you have an MST with a group in each basin, but I do not understand how, unless it’s on the basis of the phylogeny, you choose which goes to which clade. If it is on the basis of the phylogeny, you cannot then claim that the fact that the phylogeny and your conclusions coincide is evidence to strengthen your claim.

    My interpretation was based on the molecular phylogeny and fossil record of ratites, as currently known, and on comparison with the other groups I’m familiar with. The ratite phylogeny and fossil record indicate a trans-Indian clade sister to a trans-Pacific clade (as far as I’m aware this has only been mentioned in my 2014 book).

    This is a common pattern in other groups and so it supports the idea that ratite distribution conforms to a general pattern. It’s not just some random phenomenon for which the only conceivable explanation is chance.

    PS. I’ve just realised that once-in-a-thousand-years storms blowing from Africa to Madagascar will ‘only’ occur ten thousand times in 10 million years, not millions, as I suggested (484). Sorry. I think the point is clear enough though.

    Link to this
  489. 489. LeeB 1 10:24 pm 07/8/2014

    Natal philopatry of Albatrosses is certainly not absolute; albatrosses fledged in the Hawaiian Islands are apparently part of the increasing populations breeding on the islands offshore of Mexico (Clarion and Revillagigedo Islands).

    LeeB.

    Link to this
  490. 490. MichaelHeads 1:57 am 07/9/2014

    Lee (489): thanks for the information. I didn’t mean to suggest that albatrosses etc. never expand their range; the ranges of some albatross species overlap, which suggests that they do. But their well-documented philopatry can help explain their differentiation and general allopatry.

    Laysan albatrosses were wiped out in many places in the 1800s by feather hunters and have only recolonised some of these recently. Young et al 2009 wrote: ‘Laysan Albatross began colonizing islands from Japan to Mexico in the 1970s, despite their reportedly high natal philopatry. Some sites were re-colonizations of islands where the species had been extirpated, such as Mukojima, Japan… Wake Island in the Western Pacific… , and Kauai in the main Hawaiian Islands… In addition to re-colonization of historical nesting sites, Laysan Albatross also recently expanded their range to several more islands, including Lehua Islet in the main Hawaiian Islands, and into the Eastern Pacific on Isla Guadalupe, Clarion, and San Benedicto in Mexico’.

    Is it certain that these new areas were not occupied in, say, the early 1800s, with the populations then being eliminated? On Guadalupe, ‘We do not know when albatrosses started nesting on the offshore islets…’ (Pitman et al., 2004). (It reminds me of the case in New Zealand in which groups first reported in the early 1800s are inferred to have been absent before then, even though ornithology in NZ only started at that time).

    As I said, though, it’s not really important for the argument – no-one is denying that groups can expand their range.

    Link to this
  491. 491. irenedelse 3:36 am 07/9/2014

    So, what _is_ the record length for a SciAm blog post comment thread?

    Link to this
  492. 492. Dartian 4:31 am 07/9/2014

    Irene:
    So, what _is_ the record length for a SciAm blog post comment thread?

    It’s a pretty good guess that whatever will be the final number in the current thread is it.

    Shame that so many of the comments here are absolute tosh though.

    Link to this
  493. 493. irenedelse 6:49 am 07/9/2014

    @Heteromeles #469:

    “As a contrast to the pointless rhetorical tricks I’ve gotten tired of, I’d suggest that (following what Irene said above), you’re not human, you’re a Turing Test candidate program, designed to generate indefinite debates on a fringe scientific paradigm in the hopes that “you” will impress the gullible that”you’re” human so that “you” can charge them to read “your” papers, so that “you” can gain credibility, funding, and tenure for the computer scientists who created “you.” If so, bravo to your creators.
    How are you going to prove you’re human? Anyone can swipe a photo and a biography from the internet. What about what you’re writing here could make anyone think there’s a human at the other end of the conversation?”

    At this point, I think it’s more a scenario of “Who are you and what did you do with the real John Grehan??”

    :-/

    Link to this
  494. 494. irenedelse 6:55 am 07/9/2014

    P.S. In the spirit of lightening the mood a bit: there’s now panbiogeography memes inspired by this thread… :-)

    Link to this
  495. 495. Grehan1896 8:26 am 07/9/2014

    396. keesey@gmail.com

    David Marjanovic already addressed this. I’ll add a reference for a soft-tissue analysis that also supports the usual (Pongo, (Gorilla, (Pan, Homo))) topology:
    http://www.ncbi.nlm.nih.gov/pubmed/21689100
    And I’ll also point at that it’s not just “similarity”, it’s shared derived states. (They happen to be similar as well.)

    I’m still interested to know how the character states they use to support the Pan-Homo relationship are cladistically valid – that these character states represent (uniquely) shared derived conditions.

    John Grehan

    Link to this
  496. 496. Heteromeles 10:30 am 07/9/2014

    So why are Batacaeae restricted to the Pacific Ocean and why do they show such precise allopatry with their Indian Ocean sister group, Salvadoraceae? Why, in turn, is this Indo-Pacific clade almost entirely allopatric with its sister group, Koeberliniaceae (minor overlap in NW Mexico)?

    Sorry, you’re babbling again. “Precise allopatry” is a term that needs precise definition, because it looks like total obfuscation, ignorance, or a combination of both. So far as I can tell, you’re enthralled that three minor groups are united by genes do not live in overlapping regions of the globe. This is hardly a unique pattern. I’d say that the simple answer and null hypothesis is that the “pattern” you’re observing is a Black Swan Effect, aka random history of their ancestors, most of which was not preserved in the fossil record. This is the same reason why people see constellations in the mostly random patterns of stars in the sky. Your first job as a scientist is to demonstrate that the patterns you’re interested in are not random, not to get trapped in “oooh, look, pretty patterns, they must have some deep meaning.” Chase the meaning after you’ve eliminated the randomness. Otherwise, all you’re doing is something akin of astrology, which is based on the meaning people see in the mostly random patterns of stars, as coupled with the beguiling periodicity of planetary motion across this mostly random background.

    This may sound like a blow-off, but there are many weird
    non-patterns in the plant world. For example, plants also have groups like the Malpighiales, which are united on good genetic characteristics, but to my knowledge, share few, if any, structural synapomorphies (for the strict zoologists, the Malpighiales include everything willows, violets, euphorbias, coca, passionfruit, flax, and many others, most of which are small families in the tropics). The null explanation for the Malpighiales is that, unlike other groups, every physical character–stems, leaves, flowers, roots, chemistry, etc.–has been under differential selection for whatever reason. As a result, they don’t have much structurally in common. They are probably the most extreme example of this, but they are other plant orders that have a lot of structural diversity (Asparagales, for example). Conversely, there are some plant orders that have very comparatively little structural diversity. If you could figure out an index of diversity within orders, the Malpighiales would just be the most extreme data point on a long continuum, not something that’s so wildly different that it needs to be explained.

    The general question is: Why do groups with excellent means of oceanic dispersal show such similar patterns to groups with very poor means of dispersal? For example, albatrosses and their relatives are some of the best dispersers, but the breeding records show clear patterns of allopatry that are repeated in many other groups. Their natal philopatry is key here, and if, say, an island arc with breeding petrels is rifted apart, the group may well differentiate (see Fig. 8-10 of the Pacific petrels in my 2012 book Molecular panbiogeography. On the biogeography of marine groups, see my review in Biol J Linn Soc 84: 675. 2005).

    I see that other people assaulted your interpretation of albatrosses, but the equally simple answer is that they don’t. We don’t see native oaks in Hawaii, for example, although we do see some species of tarplants that did make it from California turning into small forest trees. Hawaii doesn’t exclude trees; the Pacific Ocean kept oaks away.

    To pick on another example, albatrosses and seals need to raise young on isolated areas away from terrestrial predators, and since these places are pretty rare across the world, it’s unsurprising that they tend to turn up in the same areas. That’s simply utilization of a scarce resource, not some sign that all the islands in they share were once connected to a common continent.

    I’m surprised that you’re not doing the same analysis for things like salt licks and caves. After all, ungulates tend to congregate at salt licks. Why aren’t you claiming that all salt licks were once one grand salt lick, because widely separated salt licks attract animals descended from a common ancestor? Similarly, a few groups of animals tend to evolve into troglophiles (think salamanders and crickets), while many others don’t. Why aren’t you making an argument that all the cave systems in the world were once connected into one vast underworld that has since been severed by continental drift? It’s the same type of evidence that you’re proposing to poo-poo dispersal and claim vicariance. You should be claiming vicariance for all these cases, not cherry-picking a few that sound good.

    Link to this
  497. 497. SeanMcCabe 11:55 am 07/9/2014

    “Thanks to all for continuing comments, great stuff! Did you all know that Tet Zoo is (so far as I can tell) far and away the biggest bringer-in of comments on the whole SciAm network? Hmm. Anyway…”

    You had no idea… sorry if somebody’s already pointed this out, I stopped reading this somewhere int the late 200s…

    Link to this
  498. 498. Grehan1896 11:57 am 07/9/2014

    496. Heteromeles

    So why are Batacaeae restricted to the Pacific Ocean and why do they show such precise allopatry with their Indian Ocean sister group, Salvadoraceae? Why, in turn, is this Indo-Pacific clade almost entirely allopatric with its sister group, Koeberliniaceae (minor overlap in NW Mexico)?

    Sorry, you’re babbling again. “Precise allopatry” is a term that needs precise definition, because it looks like total obfuscation, ignorance, or a combination of both.

    OK, no need for the ‘precise’ here if that bothers you.

    So far as I can tell, you’re enthralled that three minor groups are united by genes do not live in overlapping regions of the globe. This is hardly a unique pattern.

    There was no claim for uniqueness. In fact to the contrary, it is a pattern shared by other groups.

    I’d say that the simple answer and null hypothesis is that the “pattern” you’re observing is a Black Swan Effect, aka random history of their ancestors, most of which was not preserved in the fossil record. This is the same reason why people see constellations in the mostly random patterns of stars in the sky. Your first job as a scientist is to demonstrate that the patterns you’re interested in are not random, not to get trapped in “oooh, look, pretty patterns, they must have some deep meaning.” Chase the meaning after you’ve eliminated the randomness. Otherwise, all you’re doing is something akin of astrology, which is based on the meaning people see in the mostly random patterns of stars, as coupled with the beguiling periodicity of planetary motion across this mostly random background.

    Thanks. You’re view is that the pattern is uninformative and random as a null hypothesis. There have been quantitative analyses of such patterns suggesting that they are not random.

    This may sound like a blow-off, but there are many weird
    non-patterns in the plant world. For example, plants also have groups like the Malpighiales, which are united on good genetic characteristics, but to my knowledge, share few, if any, structural synapomorphies (for the strict zoologists, the Malpighiales include everything willows, violets, euphorbias, coca, passionfruit, flax, and many others, most of which are small families in the tropics). The null explanation for the Malpighiales is that, unlike other groups, every physical character–stems, leaves, flowers, roots, chemistry, etc.–has been under differential selection for whatever reason. As a result, they don’t have much structurally in common. They are probably the most extreme example of this, but they are other plant orders that have a lot of structural diversity (Asparagales, for example). Conversely, there are some plant orders that have very comparatively little structural diversity. If you could figure out an index of diversity within orders, the Malpighiales would just be the most extreme data point on a long continuum, not something that’s so wildly different that it needs to be explained.

    If I understand the above correctly you are saying that the case of the Batacaeae may represent a false phylogeny.

    The general question is: Why do groups with excellent means of oceanic dispersal show such similar patterns to groups with very poor means of dispersal? For example, albatrosses and their relatives are some of the best dispersers, but the breeding records show clear patterns of allopatry that are repeated in many other groups. Their natal philopatry is key here, and if, say, an island arc with breeding petrels is rifted apart, the group may well differentiate (see Fig. 8-10 of the Pacific petrels in my 2012 book Molecular panbiogeography. On the biogeography of marine groups, see my review in Biol J Linn Soc 84: 675. 2005).

    I see that other people assaulted your interpretation of albatrosses, but the equally simple answer is that they don’t. We don’t see native oaks in Hawaii, for example, although we do see some species of tarplants that did make it from California turning into small forest trees. Hawaii doesn’t exclude trees; the Pacific Ocean kept oaks away.

    Debate over the absence of oaks from Hawaii does not negate the assertion that there are many groups with excellent means of oceanic dispersal that show similar patterns with very poor means of dispersal.

    To pick on another example, albatrosses and seals need to raise young on isolated areas away from terrestrial predators, and since these places are pretty rare across the world, it’s unsurprising that they tend to turn up in the same areas. That’s simply utilization of a scarce resource, not some sign that all the islands in they share were once connected to a common continent.

    Can you support that hypothesis with some kind of analytical evidence?

    I’m surprised that you’re not doing the same analysis for things like salt licks and caves. After all, ungulates tend to congregate at salt licks. Why aren’t you claiming that all salt licks were once one grand salt lick, because widely separated salt licks attract animals descended from a common ancestor? Similarly, a few groups of animals tend to evolve into troglophiles (think salamanders and crickets), while many others don’t. Why aren’t you making an argument that all the cave systems in the world were once connected into one vast underworld that has since been severed by continental drift?

    Because I have not seen any biogeographic evidence for such.

    It’s the same type of evidence that you’re proposing to poo-poo dispersal and claim vicariance. You should be claiming vicariance for all these cases, not cherry-picking a few that sound good.

    It is not a matter of claiming vicariance for anything at all or cherry picking, but understanding the origin of vicariant distributions that are observed. Comparative spatial analysis does lead to the conclusion that a vicariance origin for vicariant distributions is more often the case than not, and that claims for dispersal origins are either in the nature of speculations extrapolated from presumed dispersal ability or from transmogrification of minimal molecular divergence as actual or maximal.

    By the way, there is a nice vicariance example of the fish Mileyringa in ground waters and anchialine seawaters of NW Australia that has its sister group Typhleotris in subterranean freshwaters of Madagascar.

    John Grehan

    Link to this
  499. 499. irenedelse 12:04 pm 07/9/2014

    @John Grehan #495:

    “I’m still interested to know how the character states they use to support the Pan-Homo relationship are cladistically valid – that these character states represent (uniquely) shared derived conditions.”

    Hint: the paper Keesey cited is open access. It’s not hard to see for yourself what the authors say…

    @Heteromeles #496:

    “So far as I can tell, you’re enthralled that three minor groups are united by genes do not live in overlapping regions of the globe. This is hardly a unique pattern. I’d say that the simple answer and null hypothesis is that the “pattern” you’re observing is a Black Swan Effect, aka random history of their ancestors, most of which was not preserved in the fossil record. This is the same reason why people see constellations in the mostly random patterns of stars in the sky. Your first job as a scientist is to demonstrate that the patterns you’re interested in are not random, not to get trapped in “oooh, look, pretty patterns, they must have some deep meaning.” Chase the meaning after you’ve eliminated the randomness.”

    This reminds me there is a famous pity quote attributed to Carl Sagan (who else…) which goes: “Randomness is clumpy.”

    Link to this
  500. 500. irenedelse 12:04 pm 07/9/2014

    Yay, 500! ;-)

    Link to this
  501. 501. irenedelse 12:07 pm 07/9/2014

    @Heteromeles #496:

    “Why aren’t you making an argument that all the cave systems in the world were once connected into one vast underworld that has since been severed by continental drift? It’s the same type of evidence that you’re proposing to poo-poo dispersal and claim vicariance.”

    Oh, nice. Hollow earth anyone?

    Link to this
  502. 502. Grehan1896 2:15 pm 07/9/2014

    499. irenedelse@John Grehan #495:
    “I’m still interested to know how the character states they use to support the Pan-Homo relationship are cladistically valid – that these character states represent (uniquely) shared derived conditions.”
    Hint: the paper Keesey cited is open access. It’s not hard to see for yourself what the authors say…
    @Heteromeles #496:

    You will have forgotten that I have already said that I have seen the paper. I know its content. Now I am interested to know on what basis 396. keesey@gmail.com thinks the character states in that paper are cladistically valid for the chimp-human relations (and I am not even asking him to make this assessment in relation to evidence published for the human-orangutan relationship).

    John Grehan

    Link to this
  503. 503. ectodysplasin 2:19 pm 07/9/2014

    @Michael,

    The Pacific basin is formed, for the most part, by the Pacific plate. This originated at a point (a triple junction, located in the Cook Islands region) in the Jurassic. Since then it has expanded dramatically by seafloor spreading to occupy the entire region between New Zealand and California. (For reconstructions, see Fig. 6-1 in my 2012 book ‘Molecular panbiogeography of the tropics’. Univ. California Press).

    This is flat-out incorrect. The fact that the oldest Pacific Plate rocks are Jurassic is due to subduction of older pieces of oceanic plate in the Pacific (Nazca, Farallon, Juan de Fuca, etc) beneath the continental margins with the expansion of the Pacific Plate in the middle of the Pacific Ocean. These older subducted slabs of oceanic plate are still around, deep in the mantle, and can be visualized in high resolution using various tomographic methods.

    In actuality, it is not the Pacific Ocean that opened up in the Jurassic. It is the Atlantic Ocean. And we don’t see record of subducted oceanic plate along the Atlantic Ocean because there is no subduction zone along the Atlantic Rim. This is why there’s essentially no volcanism along the Atlantic Rim (only really along the mid-Atlantic Ridge i.e. Iceland) and why New York doesn’t get hit by regular earthquakes (but LA does).

    In terms of paleogeography, the Pacific is descended from Panthalassa, without any record of closure. It’s been around for the vast majority of metazoan (and even eukaryote) evolution, dating back at least 700 million years. I somehow doubt you want to argue that ratites have been around for 700 million years.

    I’m glad you guys are doubling down on this, because it really underlines how little your school of thought actually interacts with related disciplines.

    Link to this
  504. 504. irenedelse 3:55 pm 07/9/2014

    @ectodysplasin #503:

    Wow. In all the hair splitting, I hadn’t noticed this gem. The Pacific Ocean opening in the Jurassic! Maybe we can now explain the (in)famous ropen by Gondwanan relict faunas and refugia…

    Link to this
  505. 505. ectodysplasin 4:29 pm 07/9/2014

    I noted it a bit further up-thread, and John backpedaled and hypothesized the existence of wandering island archipelagos, which is more or less equally ridiculous from a geological perspective. But now Michael is right-out proposing it again and citing his own writing, so it’s going to be more difficult for them to backpedal again.

    Whether panbiogeography can provide an alternate method for studying biogeography is one question. The history of geological events is a completely different one. John and Michael’s conflation of Pacific Ocean vs. Pacific Basin vs. Pacific Plate is straight-up amateur and really hurts their claim that they’re looking at correspondence between geographic occurrence and tectonic features, because they clearly have not looked at tectonic features at all. The rest of this discussion is rhetorical water-muddying meant to cast just enough doubt on phylogenetics, paleontology, and mainstream biogeography to deflect attempts to falsify their hypothesis, but this critical oversight of theirs about tectonics is pretty damning.

    Link to this
  506. 506. MichaelHeads 4:58 pm 07/9/2014

    Ectodysplasin (477) wrote: ‘Ultimately, I think there’s only so much that can be said to people who think that the Pacific Ocean basin opened up in the Jurassic’.

    I replied 486: The Pacific basin is formed, for the most part, by the Pacific plate. This originated at a point (a triple junction, located in the Cook Islands region) in the Jurassic. Since then it has expanded dramatically by seafloor spreading to occupy the entire region between New Zealand and California. (For reconstructions, see Fig. 6-1 in my 2012 book ‘Molecular panbiogeography of the tropics’. Univ. California Press).

    Ectodysplasin 503 This is flat-out incorrect. The fact that the oldest Pacific Plate rocks are Jurassic is due to subduction of older pieces of oceanic plate in the Pacific (Nazca, Farallon, Juan de Fuca, etc) beneath the continental margins with the expansion of the Pacific Plate in the middle of the Pacific Ocean.

    These are other plates. When do you think the Pacific plate and the basin it forms was created? You don’t accept that it formed at the Tongareva triple junction at/near the present Cook Islands in the mid-Jurassic? Do you have any references?

    For the early history of the Pacific plate, have a look at Fig. 2a in A.D. Smith 2007(A plate model for Jurassic to recent… J. Geol Soc Amer Special Papers 430: 471. It’s available on the net). Or Fig 2c in Utsunomiya et al., 2008 Preserved paleo-oceanic plateaus… Gondwana Research 14: 115. (The young Pacific plate is the small plate bounded by spreading ridges near the bottom of the figure).

    In terms of paleogeography, the Pacific is descended from Panthalassa.

    Descended from, not the same as. It’s like saying humans are Precambrian because they are descended from Precambrian groups. Why do you think geologists call it Panthalassa and not Pacific? Because geologically they’re not the same thing – the Pacific Ocean occupies a new basin, formed from new crust.

    Link to this
  507. 507. MichaelHeads 5:05 pm 07/9/2014

    504 Irene Wow. In all the hair splitting, I hadn’t noticed this gem. The Pacific Ocean opening in the Jurassic! Maybe we can now explain the (in)famous ropen by Gondwanan relict faunas and refugia…

    We’re not talking about the Pacific Ocean (which is obviously continuous with all the other oceans – there has only ever been one ocean). We’re talking about the Pacific Ocean basin, the geological structure, not the water in it.

    I’m not sure what your last sentence meant (typo?). But this may be relevant: in the Cretaceous, after the Pacific plate formed, the largest volcanic eruption in Earth history emplaced very large igneous plateaus in the central Pacific. These are known to have been emergent, and include fossil wood. The plateaus broke up and drifted to the Pacific margins.

    Link to this
  508. 508. Grehan1896 5:20 pm 07/9/2014

    I suspect my earlier wording about the Pacific was inadequate given the equating of my reference to the Pacific as ocean. Whenever I am referring to the Indian, Atlantic, Tasman etc., I am not referring to the water, but the tectonic feature.

    John Grehan

    Link to this
  509. 509. Grehan1896 5:28 pm 07/9/2014

    505. ectodysplasin
    I noted it a bit further up-thread, and John backpedaled

    Backpedaled from what? Please be specific.

    and hypothesized the existence of wandering island archipelagos, which is more or less equally ridiculous from a geological perspective.

    This is not accurate. I did not hypothesize “island archipelagos” (not sure, are you referring to former Galapagos archipelagos or to island arcs?). With respect to hypothesizing their existence, it is geologists who have been doing this almost since the acceptance of plate tectonics. If its “ridiculous from a geological perspective” then you have to ask why some of the most prominent tectonic geologists have been responsible.

    But now Michael is right-out proposing it again and citing his own writing, so it’s going to be more difficult for them to backpedal again.

    I see that Mike has already provided further clarification and sources.

    Whether panbiogeography can provide an alternate method for studying biogeography is one question.

    It can, obviously because it does. Whether one accepts that alternative or not is a separate issue.

    The history of geological events is a completely different one. John and Michael’s conflation of Pacific Ocean vs. Pacific Basin vs. Pacific Plate is straight-up amateur and really hurts their claim that they’re looking at correspondence between geographic occurrence and tectonic features, because they clearly have not looked at tectonic features at all.

    Mike has since addressed the plate and basin issue. Since you speak so authoritatively that we “ clearly have not looked at tectonic features at all” I ask you to cite our published work in support of that claim.

    The rest of this discussion is rhetorical water-muddying meant to cast just enough doubt on phylogenetics, paleontology, and mainstream biogeography to deflect attempts to falsify their hypothesis, but this critical oversight of theirs about tectonics is pretty damning.

    Of course there is no oversight.

    John Grehan

    Link to this
  510. 510. ectodysplasin 5:50 pm 07/9/2014

    @Michael:

    I replied 486: The Pacific basin is formed, for the most part, by the Pacific plate. This originated at a point (a triple junction, located in the Cook Islands region) in the Jurassic. Since then it has expanded dramatically by seafloor spreading to occupy the entire region between New Zealand and California. (For reconstructions, see Fig. 6-1 in my 2012 book ‘Molecular panbiogeography of the tropics’. Univ. California Press).

    The Pacific plate originated at the center of a larger oceanic plate system that included the JdF, Farallon, and Nazca plates, all of which are not mostly subducted. It did not originate at a spreading center between two continents. The expansion of the Pacific plate does not, then, represent the opening of the Pacific Basin, because the basin itself has not changed. It does not represent a vicariance event between the Americas and Eurasia/Oceania, because the oceanic barrier already existed long, long before the Jurassic. You’re engaging in a false equivocation here by implying that the origin of the Pacific plate has anything to do with the formation of barriers to dispersal.

    These are other plates. When do you think the Pacific plate and the basin it forms was created? You don’t accept that it formed at the Tongareva triple junction at/near the present Cook Islands in the mid-Jurassic? Do you have any references?

    For the early history of the Pacific plate, have a look at Fig. 2a in A.D. Smith 2007(A plate model for Jurassic to recent… J. Geol Soc Amer Special Papers 430: 471. It’s available on the net). Or Fig 2c in Utsunomiya et al., 2008 Preserved paleo-oceanic plateaus… Gondwana Research 14: 115. (The young Pacific plate is the small plate bounded by spreading ridges near the bottom of the figure).

    Miscitation, at best. Go back and read those papers….you’ll find that they discuss the triple-point as occurring within oceanic crust as well, not as a triple-point between continental crust. All you’re discussing here is replacement of oceanic crust by oceanic crust. The presence of an oceanic barrier to dispersal has not changed, and did not change in the past 700 million years.

    Descended from, not the same as. It’s like saying humans are Precambrian because they are descended from Precambrian groups. Why do you think geologists call it Panthalassa and not Pacific? Because geologically they’re not the same thing – the Pacific Ocean occupies a new basin, formed from new crust.

    The basin has been in place continuously throughout the past 700 million years, since the breakup of Rodinia. Large slabs of oceanic crust have been subducted beneath Asia and the Americas, but the ocean and the oceanic basin have been there the entire time, serving as an oceanic barrier to dispersal for as long as we have record of metazoan life. Your attempt to conflate basins, oceans, and plates is duly noted, but that does not make these things equivalent. For the sake of discussions about vicariance vs dispersal, what matters is that there’s a large chunk of oceanic crust between the Americas and Oceania. When you imply there has not been, you are incorrect.

    By the way, it is called Panthalassa as opposed to Pacific for the same reason that the supercontinent Pangaea is not called North-South-America-Europe-Asia-Africa-Australia-Kazakhstan-Antarctica-India-etc. However, a lot of folks working on tectonics will call it the Paleo-Pacific, so you’re wrong on that rhetorical point as well.

    Link to this
  511. 511. ectodysplasin 5:52 pm 07/9/2014

    Relevant reading:

    http://www.nature.com/ngeo/journal/v5/n3/full/ngeo1401.html

    Link to this
  512. 512. MichaelHeads 6:06 pm 07/9/2014

    505 Ectodysplasin: But now Michael is right-out proposing it [opening of the Pacific Ocean] again and citing his own writing,

    I never suggested that the ocean opened, I referred explicitly to the basin. I cited a figure in my book showing this, and the figure is a simplified version from a tectonics paper, Utsunomiya et al 2008. You haven’t cited any references.

    John and Michael’s conflation of Pacific Ocean vs. Pacific Basin vs. Pacific Plate is straight-up amateur…

    I think you’re the one confusing the ocean with the basin. Hence your reference to Panthalassa. Also, see 183-185 where you challenged John on what he thinks the age of the Pacific *basin* is. He said the *basin* was Jurassic. You replied (186) ‘The Pacific *Ocean* wasn’t formed in the Jurassic…’. The basin and the ocean are not the same though.

    they [John and myself] clearly have not looked at tectonic features at all.

    ‘Clearly’? But you admitted (465) that you haven’t read my books. They’re full of information on many tectonic features in North and South America, Africa, Asia, Australasia and Antarctica, as well as in the ocean basins.

    Link to this
  513. 513. keesey@gmail.com 6:14 pm 07/9/2014

    John Grehan wrote:

    “You will have forgotten that I have already said that I have seen the paper. I know its content. Now I am interested to know on what basis 396. keesey@gmail.com thinks the character states in that paper are cladistically valid for the chimp-human relations”

    Why are you asking me when the support is in the paper? If you have a problem with the paper, go ahead and tell us.

    By the way, since that paper only covers the muscles of the upper body, here’s another reference that covers a wider variety of characters:

    Gibbs, Collard & Woord (2002). Soft-tissue anatomy of the extant hominoids: a review and phylogenetic analysis. Journal of Anatomy 200(1):3–49. http://dx.doi.org/10.1046/j.0021-8782.2001.00001.x

    Link to this
  514. 514. irenedelse 6:15 pm 07/9/2014

    @Michael Heads #507:

    “We’re not talking about the Pacific Ocean (which is obviously continuous with all the other oceans – there has only ever been one ocean).”

    This is getting embarrassing. Seriously, now you try to switch between talking about the Pacific ocean as a geographical feature, and the water in it? “Well, all oceans were connected once”: this is earth sciences passed through a primary school level distorting filter.

    “I’m not sure what your last sentence meant (typo?).”

    No, No. I did write ropen. Since we can’t swear, I’m going for the next best thing. *sigh*

    Link to this
  515. 515. irenedelse 6:29 pm 07/9/2014

    @Michael Heads #512:

    “Also, see 183-185 where you challenged John on what he thinks the age of the Pacific *basin* is. He said the *basin* was Jurassic. You replied (186) ‘The Pacific *Ocean* wasn’t formed in the Jurassic…’. The basin and the ocean are not the same though.”

    Considering how often John said things in this thread that he later modified, I’d be careful with making assertions about what he meant. :-/

    “‘Clearly’? But you admitted (465) that you haven’t read my books. They’re full of information on many tectonic features in North and South America, Africa, Asia, Australasia and Antarctica, as well as in the ocean basins.”

    Oh, dear. *Read my book, it’s all in there*… /Rhetorical question/ Where have we heard this already? /End rhetorical question/

    Link to this
  516. 516. ectodysplasin 6:39 pm 07/9/2014

    Michael, please explain how changes in seafloor plate composition can be expected to affect ratite distribution in a manner that would make some South American and some New Zealand taxa more closely related to each other than to other ratites.

    Link to this
  517. 517. Grehan1896 7:04 pm 07/9/2014

    513. keesey@gmail.com
    John Grehan wrote:
    “You will have forgotten that I have already said that I have seen the paper. I know its content. Now I am interested to know on what basis 396. keesey@gmail.com thinks the character states in that paper are cladistically valid for the chimp-human relations”
    Why are you asking me when the support is in the paper? If you have a problem with the paper, go ahead and tell us.

    But how do you know that there is support in these papers? Or are you saying there is support because the authors claim there is? Or are you saying that you have assessed the papers and agree with their claims because ……… Before I comment on what I think might be wrong about the papers I need to know from you what you think is right.

    John Grehan

    Link to this
  518. 518. MichaelHeads 7:39 pm 07/9/2014

    In the Nature paper (link in Ectodysplasin, 511) the incipient Pacific plate is indicated in Fig. 1b as a mere point at the triple junction of three ridges (three grey lines just east of the central Panthalassa Ocean). At this stage it is surrounded by the Izanagi, Phoenix and Farallon plates. As the Pacific plate and its basin expanded, the Phoenix and Izanagi plates and their basins were destroyed by subduction. This is exactly the same model that I illustrated in my book and cited above.

    When I cited Utsunomiya etc. showing the same model as the Nature paper does, Ectodysplasin replied (510):

    Miscitation, at best. Go back and read those papers….you’ll find that they discuss the triple-point as occurring within oceanic crust as well, not as a triple-point between continental crust.

    Yes, I know that.

    All you’re discussing here is replacement of oceanic crust by oceanic crust.

    Yes, that’s right. The old basin is replaced with a new basin (including islands, emergent plateaus etc.) that has spread out. ‘All I’m discussing’ is one of the most profound revolutions in the Earth’s tectonic history. The Pacific plate, after starting at a point, is now the largest of all the Earth’s plates. You seem to think that this had no paleogeographic or biogeographic effects.

    The presence of an oceanic barrier to dispersal has not changed, and did not change in the past 700 million years.

    No-one is suggesting otherwise.

    510: The basin has been in place continuously throughout the past 700 million years, since the breakup of Rodinia.

    ‘the basin’: there has been ‘a basin’ there, in the Pacific region, but not the same basin.

    510: The expansion of the Pacific plate does not, then, represent the opening of the Pacific Basin, because the basin itself has not changed.

    The area that you refer to as ‘the basin’ has undergone a complete tectonic revolution. Most of the old basin (all the Izanagi and Phoenix plates, most of the Farallon plate) has been destroyed. The Pacific basin, formed by the Pacific plate, is a completely new basin.

    …the ocean and the oceanic basin have been there the entire time, serving as an oceanic barrier to dispersal for as long as we have record of metazoan life. Your attempt to conflate basins, oceans, and plates is duly noted,

    There has always been ocean there, but now it’s in a different basin. The ocean and its basin are not the same thing. (For your conflation of the Pacific ocean and the Pacific basin, see 183-186).

    For the sake of discussions about vicariance vs dispersal, what matters is that there’s a large chunk of oceanic crust between the Americas and Oceania. When you imply there has not been, you are incorrect.

    I’m not implying or even hinting anything of the sort, and I never have. What I suggest matters for biogeography is that all the crust between New Zealand and California (i.e. the Pacific plate, bearing thousands of flat-topped seamounts – former islands – as well as large, formerly emergent plateaus) has formed by seafloor spreading since the Jurassic. At the same time, the spreading ridge creating it (East Pacific rise) has migrated eastwards, eventually hitting North America.

    Link to this
  519. 519. DavidMarjanovic 7:44 pm 07/9/2014

    Question to the panbiogeographers: how do you explain the Neogene ostriches of Eurasia?

    Agreed, and it [POY] also does not eliminate the phonetic nature of base pair matching.

    How so? It… can hardly be said to match base pairs. (And please do something about your spellchecker – perhaps just turn it off.)

    The alignment process creates base pair homologies that do not actually exist in nature, but are theorized to have existed based on one or other best fit models. In morphology it is possible to directly observe character states and to identify those that are absent in the outgroup.

    In morphology, too, homology is a hypothesis. Hence the distinction between “primary homology” (in the data matrix) and “secondary homology” (hypotheses of primary homology confirmed by phylogenetic analysis of the data matrix).

    My overall view is that neither morphogenetic [you did it again] or molecular approaches are infallible, and neither automatically disqualifies the other when they are incongruent for a particular group.

    That’s the current consensus. (At least among us morphologists, LOL.)

    You are describing the generalities of the steps involved, but what I am asking is what exactly do you see in the extant distributions as evidence necessitating an Asian origin, and what in the fossil record necessarily points to this?

    …I don’t understand what you mean, because I answered that question in quite some detail in a long and a shorter paragraph somewhere in the middle of comment 451 – except that “necessitating” and “necessarily” are your additions; I’ve been much more careful.

    I’ve since noticed that most of the comments are directed to Mike.

    They’re not really directed at any particular person; they’re directed at the quotes they reply to. Feel free to comment on anything you feel competent to comment on. :-|

    “2. Is there evidence that, say, the Galápagos tortoises came from elsewhere anyway?”

    No.

    Yes, there is. It’s what I wrote about testudinoid paleobiogeography somewhere in the middle of comment 451.

    everyone who has a strong opinion on such matters is wedded to a specific method

    *facepalm*

    People are entitled to their opinions.

    People are not, however, entitled to their own facts. ectodysplasin made a claim of fact, and so did I; claims of fact are right or wrong – they aren’t matters of taste.

    By the way, David Marjanovic discussed at #468 the case of New Caledonia and its separation from Zealandia. Contrary to New Zealand, it does have a fossil record from the Cenozoic.

    I didn’t discuss it, I barely mentioned it; and I’m not aware of a terrestrial fossil record there, except maybe from the Pleistocene. You’re right, though, that there’s no evidence of paleognaths in New Caledonia.

    [...] arguing about opinions does not, in my experience, get very far. Someone might, for example, have the opinion that moas could swim to New Zealand. Without anything else to go on I would not engage with that opinion.

    Why? This “opinion” would be a testable hypothesis – testable against parsimony: how likely is it that a non-aquatic bird could swim that far? Even given magical food supply for the full duration, it’s too unlikely to contemplate. There, engaged with the hypothesis and dismissed it based on observable evidence in five seconds. :-|

    Grenada is just 100 km from Trinidad, where there are tinamous

    Trinidad is part of the South American continent, so it’s connected to the mainland every ice age. Wait for an ice age, and you can just walk to Trinidad. Grenada is volcanic, it’s part of an island arc; getting there is transoceanic dispersal. Given this alone, we should absolutely expect tinamous on Trinidad, while we should consider it unlikely (not impossible, but unlikely) that tinamous would reach Grenada.

    The fact that the groups have such a large, global range but such a small area of overlap is interesting and needs to be explained somehow.

    What if the explanation is just that Antarctica has iced over?

    You could test it by continuing to look for fossils in Antarctica, which people do.

    Sure. Trouble is, you’re acting as if the search had already gone on for long enough to confirm your expectations – even though most of the continent hasn’t been sampled.

    “But can you distinguish middle Late Cretaceous vicariance from much later dispersal to NZ from nearby continents (Australia for Mystacinidae; Australia, with later local extinction, or possibly Antarctica for Myzopodidae)?”

    Based on fossil calibrations, Mystacina and its sister group diverged at 35-68 Ma (Lloyd, 2003) or 41-51 Ma (Teeling et al., 2005), and as minimum dates these are compatible with Cretaceous vicariance.

    That’s a “no”, then.

    The two trans-Tasman groups – the bat family Mystacinidae and the pair of flies – can both be derived by simple vicariance in widespread ancestors. In the case of the bats the ancestor was panaustral, while in the case of the flies the ancestor was global. If the ancestors attained their wide global range during mid-Cretaceous phases of high sea-level, they could then have been fractured during the pre-breakup rifting of Gondwana.

    Don’t you think the flies travel in the fur of the bats? If the bats crossed the Tasman Sea at some point in the Cenozoic, they can just have taken the flies with them, and you don’t need to assume anything as miraculous as Cretaceous bats that refuse to leave a single tooth in the fossil record. (Possible bat teeth are pretty common in the late Paleocene of at least North America. Onychonycteris, known from a whole skeleton, shows up in the early Eocene, and by the middle Eocene bats are found all over the world.)

    As with the fly Mystacinobia, the mite Chirophagoides (Sarcoptidae) is known only in association with Mystacina. Its sister-group comprises Chirnyssoides of tropical America, plus Notoedres, a diverse group found worldwide (but known in New Zealand only from introduced mammals) (Klompen, 1992). As with Mystacinobia and ‘McAlpine’s fly’, Chirophagoides cannot be derived from anywhere outside the Tasman region without invoking extra ad hoc hypotheses, but its affinities are consistent with simple vicariance between the Tasman region and the rest of the world.

    I’m really surprised that you seem to believe a mite associated with bats is independent evidence, independent of the bats. It’s almost self-explanatory that the mite was carried to New Zealand in the fur of Mystacina.

    I should have said single *monophyletic* clade on Madagascar -

    What – no. “Monophyletic clade” is redundant. “Clade” doesn’t mean “taxon”, and it doesn’t mean “branch on a tree-shaped diagram”; it means “an ancestor and all its descendants”. It’s a synonym of “monophylum”.

    It’s not just very, very rare, it’s unique – it only happened once in tens of millions of years (or however old you accept that artiodactyls, tenrecs, primates etc. are). It happened precisely once in each group, despite their very different ecologies and means of dispersal. What ecological event only happens once in tens of millions of years? Even a very, very rare ecological event, such as a once-in-a-thousand-years storm, would have occurred millions of times.

    A once-in-a-million-years storm, however, has occurred a bit less often.

    A storm that blows a male and a female across, or a female pregnant with at least a male and a female, has occurred less often still.

    It’s really easy to come up with further factors that restrict dispersal. Our lucky breeding pair must remain lucky all the way across the sea; it mustn’t die from dehydration, starving, or random passing sharks, and it mustn’t be swept somewhere past the promised land. Once that is accomplished and the pair walks ashore, it must establish a population; the population could die out very quickly from inbreeding, and it might face predation or competition.

    Why do you act as if this had never occurred to you? I honestly can’t imagine that it hasn’t.

    Mammals are much more prone to dying from thirst and hunger than squamates are. This explains very neatly why so many oceanic islands are populated by lizards but not by terrestrial mammals.

    Why have no non-human primates – monkeys, strepsirrhines or tarsiers – ever crossed the stretch of sea (Salue Timpaus Strait, just 20 km wide!) that separates Sulawesi from the Sula Islands and Australasia?

    They only reached Sulawesi quite recently. Wait a few more million years, and the Wallace line will continue to blur.

    I’m still interested to know how the character states they use to support the Pan-Homo relationship are cladistically valid – that these character states represent (uniquely) shared derived conditions.

    WTF, they don’t need to be unique at all. Convergence is rampant; if a big morphological data matrix gives you a tree with a consistency index much above 0.3, you have a strong reason to suspect that the authors cherry-picked the data!

    Finding out that a character state is homologous between terminal taxa, however, is what phylogenetic analysis does; and the same generally holds for finding out which condition(s) is/are derived.

    You’re engaging in a false equivocation here by implying that the origin of the Pacific plate has anything to do with the formation of barriers to dispersal.

    I think the claim is a quite different one: that the origin and expansion of the Pacific plate has spread basaltic plateaus around (Ontong Java et al.) that were originally close together and perhaps – though I haven’t noticed anything explicit in this thread – close to a continent. This way, the expansion of the Pacific plate would have carried terrestrial organisms from one side of the ocean to the other… assuming the plateaus didn’t get eroded away during their long journey.

    Link to this
  520. 520. MichaelHeads 7:51 pm 07/9/2014

    514: Irene wrote: This is getting embarrassing. Seriously, now you try to switch between talking about the Pacific ocean as a geographical feature, and the water in it?… this is earth sciences passed through a primary school level distorting filter.

    It’s important to distinguish between the basin and the ocean in it.

    “I’m not sure what your last sentence meant (typo?).”

    No, No. I did write ropen. Since we can’t swear, I’m going for the next best thing. *sigh*

    Sorry for my profound ignorance. I lived in PNG for six years working on biogeography and I don’t remember hearing about the ropen, but there are lots of stories up there…

    Link to this
  521. 521. Stripeycat 8:03 pm 07/9/2014

    OK [pretends this is an introductory informal logic exercise, and takes a deep breath].

    John #483, replying to my ?#376
    However, we still don’t have evidence for what’s going on at the hotspot between c76 and c17-20mya, because the relevant bits of sea-floor have probably been destroyed.

    Correctly, one may not have the stratigraphic evidence. There is biogeographic evidence.

    Citation needed.

    You simply cannot hypothesise that there was a continuous procession of archipelagoes, use that as support for Galapagos vicariance,

    Agreed!!! And of course, that was not done.

    and then claim vicariance as evidence for the hotspot’s behaviour. Do you not see why this is circular reasoning?

    It would be if that were the way it was argued. But it is the vicariance evidence that leads to the prediction that there was a continuity of Galapagos archipelagos that allowed for survival of the colonists from earlier landscapes.

    As I understand it, your argument boils down to “an assumption of vicariance requires continuous succession of archipelagos”. Since there’s no evidence presented for either vicariance or islands, this is irrelevant to the truth values of either “vicariance happened” or “there was no temporal gap in the islands”. (And note that the latter is not sufficient to imply the former.) The only “evidence” you’ve offered for Galapagos vicariance in this thread is the existence of the Carnegie ridge. If you have better arguments, present them!

    John #483
    As said before, sister group of the kiwi is on the other side of the Indian Ocean, sister group of moa is on the other side of the Pacific Ocean. Phylogeny provides the biological relationship; spatial analysis provides the biogeographic relationship. I can appreciate that this may be a bit difficult to absorb if one is not well acquainted with the field of biogeography.

    Mike #488
    My interpretation was based on the molecular phylogeny and fossil record of ratites, as currently known, and on comparison with the other groups I’m familiar with. The ratite phylogeny and fossil record indicate a trans-Indian clade sister to a trans-Pacific clade (as far as I’m aware this has only been mentioned in my 2014 book).

    This is a common pattern in other groups and so it supports the idea that ratite distribution conforms to a general pattern. It’s not just some random phenomenon for which the only conceivable explanation is chance.

    You both say there is a correlation between the ocean basin in which these species occur, and their phylogeny. You also claim that other species show a similar pattern. What you do not anywhere present evidence for is that this pattern is better explained by vicariance than by transoceanic dispersal. (ie you note that x and y are correlated, but do not proffer an argument as to why.) Therefore you are presenting an untested hypothesis of vicariance. (I’d slightly misunderstood, in that I thought you were claiming that phylogeny+distribution supported vicariance.) Untested hypotheses are also called Just So Stories, and unless you can find a way to test them, are ignored regardless of their truth value.

    The trouble is that almost anything can be presented as an unevidenced hypothesis, and if it’s untestable (because the necessary evidence is lacking) it’s, /de jure/, impossible to disprove.

    Link to this
  522. 522. LeeB 1 8:11 pm 07/9/2014

    The western Mexican islands could have been inhabited by albatrosses in the early 1800s but this would have nothing to do with their reappearance in the 1980s unless they can live a very long time indeed; besides the rapid growth in the populations is far more than can be sustained by such slow breeding birds, they have to be immigrating from elsewhere and the Hawaiian Islands are closer than the southern Japanese Islands.

    LeeB.

    Link to this
  523. 523. ectodysplasin 8:16 pm 07/9/2014

    Michael,

    Please explain precisely how the plate composition of the floor of the largest ocean in the world has any impact whatsoever on terrestrial faunas occupying continental crust.

    Link to this
  524. 524. MichaelHeads 9:11 pm 07/9/2014

    David 519: ‘Question to the panbiogeographers: how do you explain the Neogene ostriches of Eurasia?’

    Global ratite ancestor underwent vicariance at Mozambique Channel and also somewhere between China (ostrich fossil) and New Guinea (cassowaries), isolating the ostriches.

    I asked: Why no ratites in the Antilles? Grenada is just 100 km from Trinidad, where there are tinamous…

    Trinidad is part of the South American continent, so it’s connected to the mainland every ice age. Wait for an ice age, and you can just walk to Trinidad. Grenada is volcanic, it’s part of an island arc; getting there is transoceanic dispersal. Given this alone, we should absolutely expect tinamous on Trinidad, while we should consider it unlikely (not impossible, but unlikely) that tinamous would reach Grenada.

    Yes, but tinamous can fly and ancestral ratites could all fly. Given enough time, unlikely dispersal events such as dispersal from Madagascar to New Zealand, let alone Trinidad to Grenada, or Mexico to Cuba, become certain (!). Does anyone know of any ratite fossils in the Greater or Lesser Antilles? I couldn’t find any in a quick search.

    I wrote: The fact that the groups have such a large, global range but such a small area of overlap is interesting and needs to be explained somehow.

    David replied: What if the explanation is just that Antarctica has iced over?

    Yes, they may have overlapped in Antarctica, or they may have been allopatric there.

    I suggested ‘You could test it by continuing to look for fossils in Antarctica, which people do’.

    Sure. Trouble is, you’re acting as if the search had already gone on for long enough to confirm your expectations – even though most of the continent hasn’t been sampled.

    No, I was just suggesting that testing the idea was not impossible in principle.

    David wrote: I’m really surprised that you seem to believe a mite associated with bats is independent evidence, independent of the bats. It’s almost self-explanatory that the mite was carried to New Zealand in the fur of Mystacina.

    I didn’t say it was independent evidence. It’s just more evidence. And why assume that Mystacina dispersed to NZ?
    (Yes, I have read all the Icarops papers by Hand et al., but they do not include any explicit justifications for their ideas on the Australian fossils being ‘primitive’ etc.). The fact that the parasites are globally basal groups needs to be explained. If they resulted from events pre- or at Gondwana breakup (which seems likely), there is no need to assume that their Tasman region host dispersed to NZ.

    Storms, Madagascar etc.: so why are there no monkeys on Madagascar? And no strepsirrhines on America? This interesting symmetry can be explained by vicariance, followed by overlap in Africa and Asia.

    I wrote: Why have no non-human primates – monkeys, strepsirrhines or tarsiers – ever crossed the stretch of sea (Salue Timpaus Strait, just 20 km wide!) that separates Sulawesi from the Sula Islands and Australasia?

    They only reached Sulawesi quite recently.

    Do you have any evidence? Nasalis is an excellent swimmer, probably the best primate swimmer in the sea, but it has never reached Sulawesi from Borneo. Do you think that Wallace’s line itself, as seen in many primates, is the the result of geology? If so, couldn’t the other lines be as well, given that they coincide with major tectonic breaks and also western limits in eastern groups?

    Link to this
  525. 525. MichaelHeads 9:21 pm 07/9/2014

    523 Ectodysplasin wrote:

    Please explain precisely how the plate composition of the floor of the largest ocean in the world has any impact whatsoever on terrestrial faunas occupying continental crust.

    E.g. accretion of Pacific plate oceanic crust (with islands, plateaus etc.) to continental margins, e.g. in New Zealand, New Caledonia, California. For Farallon plate derivatives and Colombia/caribbean, have a look at Kerr and Tarney 2005, Geology, Fig. 3(Google scholar, look under ‘all 6 versions’ for an accessible version).

    Link to this
  526. 526. ectodysplasin 11:52 pm 07/9/2014

    You still have to account for how those birds got onto those islands in the first place.

    Really, what you’re saying is that you’re hyopthesizing both transoceanic dispersal (from the mainland to these islands) and then vicariance (between islands), while simultaneously hypothesizing a lost fossil record of 70+ million years.

    But you haven’t actually explained why the Pacific plate is substantively different from the Farallon etc. Why do we not see ratites in North America or Asia? Why do we not see ratites in Hawaii or Guam or anywhere else in Polynesia? Why are the other endemic avian faunas of the Pacific devoid of ratite relatives?

    Furthermore, how does classifying terrestrial communities by adjacent oceanic plate provide any benefit above classifying them by the land mass they exist on? You still haven’t shown the need for this extra inferential step.

    Link to this
  527. 527. Dartian 1:38 am 07/10/2014

    Irene:
    we can’t swear

    Actually, we can use swear words here on Tet Zoo, at least in small doses* (I have used the f-word once or twice, I recall). But it depends on the context. I think what Darren meant was that we’re not allowed to swear at each other (which is a reasonable request).

    * Or English swear words, at any rate. We’re a pretty educated and international crowd here, so if one absolutely wishes to circumvent the restrictions on the use of profanities that can be done by cursing in Foreignese. Not that I’m recommending that – I’m just saying. ;)

    Fortunately, however, there are more creative ways to express annoyance and frustration than the use of traditional profanities. For example, if someone says something really silly and nonsensical here on Tet Zoo, you can from now on respond with “Now that’s just a load of panbiogeography”. ;)

    Link to this
  528. 528. MichaelHeads 5:59 am 07/10/2014

    526 Ectodysplasin But you haven’t actually explained why the Pacific plate is substantively different from the Farallon etc. Why do we not see ratites in North America or Asia? Why do we not see ratites in Hawaii or Guam or anywhere else in Polynesia? Why are the other endemic avian faunas of the Pacific devoid of ratite relatives?

    There are fossil ratites in Asia (ostriches, east to China) and North America (lithornithids). But I agree, it’s very interesting that the ratites are not known from the Pacific islands apart from New Guinea and New Zealand. They are present, but apparently not native, on New Britain, but there are no records at all (living or fossil) from the rest of Melanesia (including New Caledonia), or Micronesia or Polynesia. This makes the apparent absence of fossils from all the Antilles (??? is this correct??) even more intriguing, as the Caribbean plate probably formed in the Pacific and there are many biogeographic similarities between the two regions.

    Perhaps the ratites were in all or some of these regions and went extinct, with no fossil record. There are plenty of bird fossils on the islands, but these are mainly quite young, and the group could have gone extinct in the region long ago. Alternatively, ratites may have never been there. The same options apply to the region between China and New Guinea. I guess we can agree on this, so far.

    The idea that early ratites were volant makes their absence from the regions just cited even more intriguing, IF it is in fact real and not just the result of sampling error in the fossil record. I’ve written quite a bit on the great biogeographic break between the New Guinea mainland and the large islands of New Britain and New Ireland etc., just 50 km off the mainland. E.g. no birds of paradise on the islands (but major diversity of Platymantis frogs etc.).

    So it wouldn’t surprise me TOO much if the apparent absence of ratites from the Pacific islands (and perhaps even India, SE Asia and the Caribbean) was real, and fossils don’t turn up. (Any fossils from these areas would be of fantastic interest though, and I’m definitely not predicting that they won’t turn up!).

    Of course, ratites do have relatives all through these areas – their sister group. So the question is: how to explain the great apparent gap in the otherwise global range of ratites?

    Here is one possibility: the original split in crown group birds was between one group, the ratites, found everywhere except the Pacific, and another group, its sister (carinates s.str., = neognaths) allopatric with it, and restricted to the Pacific. (This would be an interesting parallel to the crown group frogs, with the basal group, Leiopelmatidae only known (living and fossil) from New Zealand and the western US (the Patagonian records are discounted by recent authors), straddling the Pacific). Subsequently, the ‘Pacific’ neognaths dispersed to everywhere in the world, while the ratites were much less successful, with little speciation anywhere except South America, and complete extinction in many areas (North America, Europe, Asia).

    The great overlap betwen ratites and carinates would probably have occurred before much of the modern differentiation within the ratites and within the carinates took place, as many of the distribution patterns within the two are duplicated (e.g. trans-Indian affinities in elephant birds and in ducks; trans-Pacific affinities in moas-tinamous and in procellariiforms).

    This very speculative example is really just to illustrate how our method works, using the available data. I don’t know of any other ideas on the geography of the ratite-carinate split, and I’d be very interested to hear of any.

    Furthermore, how does classifying terrestrial communities by adjacent oceanic plate provide any benefit above classifying them by the land mass they exist on? You still haven’t shown the need for this extra inferential step.

    It’s just a shorthand terminology/concept that is useful in practice. If you track down the affinities of thousands of groups and their distributions you find that an awful lot of them straddle the Indian, Pacific, Atlantic or Tethys basins, or combinations of these.

    E.g. Everyone knows the Australasia – Patagonia connection, but the tropical West Pacific – tropical east Pacific one is just as common, although it’s hardly ever mentioned. Neither one is expected if you use Wallace’s regions, i.e. the continents. So when people find, say, a New Guinea – Colombia group they always go – ‘how weird, how anomalous, how puzzling’ etc.

    Link to this
  529. 529. irenedelse 7:39 am 07/10/2014

    @Dartian #527:

    Heh. Maybe I should brush up on my Latin…

    Link to this
  530. 530. irenedelse 7:51 am 07/10/2014

    @David #519:

    “[IT]“it [POY] also does not eliminate the phonetic nature of base pair matching.”

    How so? It… can hardly be said to match base pairs. (And please do something about your spellchecker – perhaps just turn it off.)”

    The phonetics of evolution? I’m smelling a new meme…

    BTW, I tip my metaphorical hat to everyone who managed to keep up on this thread riddled with formatting issues, spellchecker mishaps and other sorts of “good fun”!

    Link to this
  531. 531. irenedelse 7:54 am 07/10/2014

    Er… And so the spellchecker strikes again. In my previous comment, it should have been [JG] for John Grehan, not [IT].
    *hangs my head in shame*

    Link to this
  532. 532. Grehan1896 8:28 am 07/10/2014

    526. ectodysplasin

    You still have to account for how those birds got onto those islands in the first place.?

    Those birds may have ‘got onto those islands’ the same way that they got onto those continents. They had a widespread ancestor that encompassed areas that are today both islands and continents. The ancestor established its distribution in the same way that any ancestral species does – by its means of dispersal acting in relation to the ecology and geography of the time (ecology is often overlooked in historical reconstructions as if geography were the only limiting factor as expressed in the view that a taxon should be present throughout an area if that area was contiguous).

    John Grehan

    Link to this
  533. 533. Grehan1896 9:21 am 07/10/2014

    521. Stripeycat
    Thank you for your continuing interest in exploring the Galapagos. It’s certainly a biogeographically interesting center.

    Correctly, one may not have the stratigraphic evidence. There is biogeographic evidence.
    Citation needed.

    See the 2001 Galapagos paper.

    As I understand it, your argument boils down to “an assumption of vicariance requires continuous succession of archipelagos”. Since there’s no evidence presented for either vicariance or islands, this is irrelevant to the truth values of either “vicariance happened” or “there was no temporal gap in the islands”.

    The vicariance evidence is presented in the form of vicariant (allopatric) distributions of taxa. There are distributions that are vicariant between the Galapagos and other localities. In addition there are patterns (tracks) that are allopatric with respect to each other except at the Galapagos (and I think these were even evident to JD Hooker). So there is no assumption of vicariance. Rather there is the analysis that demonstrates the existence of vicariant patterns. One might hypothesize that these patterns are the result of individual chance dispersal over the current geography (at various times in the past) or hypothesize that these patterns are the result of historical vicariance of ancestral distribution ranges (some of these ancestral ranges being brought together at the Galapagos.

    So the biogeographic evidence is in the form of distributions. A vicariance hypotheses generated from that biogeographic evidence leads to the predication that the geography of the eastern Pacific was different from what it is today. Two specific predications have been corroborated (see 435. Grehan1896). That evidence is not limited to the existence of the Carnegie ridge. There is also the Cocos ridge (and of that the section now being subducted has been hypothesized by geologists to represent a former Galapagos archipelago), not to mention the terranes evidence that also includes the geological models for the origin of the Caribbean plate within the Pacific.

    You both say there is a correlation between the ocean basin in which these species occur, and their phylogeny. You also claim that other species show a similar pattern.

    This is not just a claim. It has been demonstrated in the panbiogeographic literature numerous times. If you disagree I would be interested in what you would present as evidence that these spatial correlations do not exist because within the field of biogeography (those publishing on biogeographic method and theory) no one has yet done that.

    What you do not anywhere present evidence for is that this pattern is better explained by vicariance than by transoceanic dispersal. (ie you note that x and y are correlated, but do not proffer an argument as to why.)

    This may depend on what you regard as a “better explanation”. I consider vicariance as a better explanation because it does not present any anomalies with respect to dispersal ability (‘good’ and ‘poor’ dispersers sharing the same kinds of biogeographic pattern), provides a cause effect relationship between the correlation of distributions of differentiated entities and tectonics (rather than having to say that the match is historically meaningless coincidence), does not raise contradictions between supposed dispersal ability and distribution (e.g. 484 MichaelHeads Apparently there were even currents flowing to Madagascar until, from memory, the Eocene(Ali and Aitchison in Nature). So why no monkeys on Madagascar, even though they could cross the Atlantic? Why no strepsirrhines (lemurs etc.) in America, even though they could cross to Madagascar? Why have no non-human primates – monkeys, strepsirrhines or tarsiers – ever crossed the stretch of sea (Salue Timpaus Strait, just 20 km wide!) that separates Sulawesi from the Sula Islands and Australasia?), and perhaps most interestingly the patterns have been successfully used to predict geological structures unknown to geologists of the time, or show relationships to geology that have not been previously recognized (such as the relationship between distributions and some major oil fields).
    Of course, just because I think these qualities provide ‘better explanations’, others may have different criteria for what is ‘better’.

    John Grehan

    Link to this
  534. 534. irenedelse 9:59 am 07/10/2014

    @David #519:

    “I’m really surprised that you [John] seem to believe a mite associated with bats is independent evidence, independent of the bats. It’s almost self-explanatory that the mite was carried to New Zealand in the fur of Mystacina.”

    And of course it’s a phenomenon observed today. Cf. the colonization of the newly formed volcanic island Surtsey, where various insects have been introduced by birds.

    “I didn’t discuss it [biogeography of New Caledonia], I barely mentioned it; and I’m not aware of a terrestrial fossil record there, except maybe from the Pleistocene. You’re right, though, that there’s no evidence of paleognaths in New Caledonia.”

    Point taken. And you are right about only Pleistocene (late Pleistocene and Holocene, even) terrestrial fossil record, at least for animals. These include Megapodes, a Meiolanid turtle, and a Mekosuchine crocodile. No ratites, but giant Megapodes occupied a Moa-like niche…

    The palynological record of NC extends into the Miocene though, which is of interest to panbiogeographers: Nothofagus and so on… *But* earlier in the Cenozoic, there’s only marine deposits. That the archipelago was mostly or entirely submerged at that time is not controversial (cf. Grandcolas et al., 2008), although some researchers argue that a few elevations stayed emerged and served as refugia. Which is still disputed.

    P. Grandcolas, J. Murienne, T. Robillard, L. Desutter-Grandcolas, H. Jourdan, E. Guilbert & L. Deharveng, “New Caledonia: a very old Darwinian island?”, Phil. Trans. R. Soc. B, 27 October 2008, vol. 363 no. 1508, 3309-3317.

    The authors strongly argue against the out-of-Gondwana explanations for New Caledonian endemism and criticize the methodology of researchers who defend them:

    “Ancient radiations or repeated dispersal?
    Answering this classical question is fundamental to understanding the evolution of biodiversity in New Caledonia and makes it possible to distinguish regional endemism (groups restricted to the New Caledonia mainland as a whole) from local endemism (groups restricted to certain locations in New Caledonia), an often confused issue for this island (Murienne 2006). Most studies citing the high rates of endemism (often close to 100%) in many groups of New Caledonian organisms actually refer to regional endemism examined in the context of large-scale phylogenetic studies (e.g. Morat et al. 1986; Chazeau 1993; Lowry 1998). Such studies often reveal that within certain New Caledonian groups, multiple species are nested within larger clades with taxa from Australia, New Zealand or New Guinea, calling for explanations in terms of recent dispersal. [...]

    “None of [the existing] studies provides clear evidence for old local diversification since most dates inferred from molecular phylogenies do not pre-date the Oligocene (figure 2). Several emblematic groups such as Araucaria and Nothofagus have even undergone more recent radiations or colonizations of New Caledonia. [...]

    Microendemism is extremely high on this medium-sized island (16 890 km2) and should not be neglected by emphasizing only larger-scale regional endemism (Murienne 2006). Along this 500 km long island, many related species are each often restricted to a very small area (often less than a few square kilometres). [...]

    A contrary interpretation [to recent diversification after recolonization of the islands after Oligocene uplifting] would require invoking either unknown or extinct mainland species closely related to those on the more recent islands, a presumption that prevents any further logical biogeographic reasoning. However, following this presumption, some authors have hypothesized that palaeo-islands, pre-existing in the same place as recently uplifted islands (a geologically plausible assumption), could have harboured a member of the same clade, thus allowing for an older age (Heads 2005; Ladiges & Cantrill 2007). Such an assumption is, however, not warranted, as there is no actual evidence for recently uplifted islands occurring in conjunction with those palaeo-islands; and in any case, such a scenario would require several dispersal events among those islands in the past.”

    Well, well.

    Link to this
  535. 535. Grehan1896 11:53 am 07/10/2014

    534. irenedelse
    Such an assumption is, however, not warranted, as there is no actual evidence for recently uplifted islands occurring in conjunction with those palaeo-islands; and in any case, such a scenario would require several dispersal events among those islands in the past.”
    Well, well.

    Well, well indeed! You can see here how Murienne is using geological information to delineate what is possible in biogeography on the basis that the geological information is complete and entire. Such an assumption is not warranted. What Murienne fails to recognize is that distributions of taxa within New Caledonia do track tectonic formation, allochthonous terranes and major transform faults.

    John Grehan

    Link to this
  536. 536. Grehan1896 1:47 pm 07/10/2014

    530. irenedelse

    “[IT]“it [POY] also does not eliminate the phenetic nature of base pair matching.”
    How so? It… can hardly be said to match base pairs. (And please do something about your spellchecker – perhaps just turn it off.)” BTW, I tip my metaphorical hat to everyone who managed to keep up on this thread riddled with formatting issues, spellchecker mishaps and other sorts of “good fun”!

    Certainly the formatting problems are mostly my fault – especially the phonetic instead of phenetic (I’ve added that to my dictionary so hopefully that will fix it). I have not experienced this kind of venue very much and with the italic formatting I find I have to be extremely careful. I have this insane proclivity for putting the i before the / instead of after.

    Anyway, about matching base pairs. This is what has to be done to compare different species. The base on position X for species 1 has to be matched with its homologous position Y on species 2. This would not be problematic if the gene section for each species was of the same length – same number of bases. Then one could simply line them up and say X was the same as Y. Apparently it is often the case the homologous genes (and identifying that can have its own challenges) are of different lengths so there is no obvious way to know which of the shorter gene line up with which bases of the longer. So one has to ‘stretch’ out the shorter to maximize the number of matches relative to the number of ‘gaps’ that have to be created. There are algorithms to make a choice, but the choice is arbitrary as there is really no way to be sure how to balance out the number of gaps (in a sense these are ad hoc hypotheses to accommodate the length discrepancy) and number of matches. And having said that, I am not saying that morphogenetic comparisons are not without their problems as well.

    John Grehan

    Link to this
  537. 537. Grehan1896 2:02 pm 07/10/2014

    519. DavidMarjanovic
    [...] arguing about opinions does not, in my experience, get very far. Someone might, for example, have the opinion that moas could swim to New Zealand. Without anything else to go on I would not engage with that opinion.
    Why? This “opinion” would be a testable hypothesis – testable against parsimony: how likely is it that a non-aquatic bird could swim that far? Even given magical food supply for the full duration, it’s too unlikely to contemplate. There, engaged with the hypothesis and dismissed it based on observable evidence in five seconds.

    Ah – OK. Now I understand what you mean by ‘test’ – the test is made against your opinion of what is likely or unlikely.

    My overall view is that neither morphogenetic [you did it again] or molecular approaches are infallible, and neither automatically disqualifies the other when they are incongruent for a particular group.
    That’s the current consensus. (At least among us morphologists, LOL.)

    Consensus means everyone (at least to my understanding). You are speaking authoritatively about this which leaves me confused as there are quite a number of morphologists in primate systematics who have said otherwise – so who is wrong about that? These individuals (just about every hominid systematist who works with living or fossil taxa – but David Pilbeam provides a nice example) regard morphogenetic evidence to be subordinate to molecular evidence. Every time something crops up in morphogenetics that contradicts the human-chimpanzee theory (e.g. orangutan apomorphies in australopiths) they are supposed to be ‘wrong’ – so they say. Same goes for living hominoids. Even though there are about 30 unique or near uniquely shared features among humans and orangutans, and perhaps only one (that I have seen as verifiable so far), this level of shared derived similarity is supposed to mean nothing whereas molecular similarity is seen to be absolute. As Ruvolo has stated – the chimpanzee relationship to humans is a scientific fact.

    John Grehan

    Link to this
  538. 538. Heteromeles 2:03 pm 07/10/2014

    Hmmm. Anyone want to quickly whip up some “I’m a panbiogeographer” t-shirts for TetZooCon? If you kept the quantities low, I’ll bet they’d sell out pretty fast. Extra points for doing it in italics or with spell-checking issues.

    Link to this
  539. 539. Halbred 2:18 pm 07/10/2014

    Not to play defense here, but it’s not like the SciAm platform is doing us any favors when it comes to formatting. ;-)

    Link to this
  540. 540. irenedelse 2:51 pm 07/10/2014

    @John Grehan:

    “You can see here how Murienne is using geological information to delineate what is possible in biogeography on the basis that the geological information is complete and entire. Such an assumption is not warranted. What Murienne fails to recognize is that distributions of taxa within New Caledonia do track tectonic formation, allochthonous terranes and major transform faults.”

    1) Murrienne is not the only or even the main author of the paper I quoted.

    2) Re-read the second paragraph I quoted:

    “None of [the existing] studies provides clear evidence for old local diversification since most dates inferred from molecular phylogenies do not pre-date the Oligocene (figure 2). Several emblematic groups such as Araucaria and Nothofagus have even undergone more recent radiations or colonizations of New Caledonia.”

    So it’s not just an issue of unknowns from geology. The authors are saying that there are more than one line of evidence supporting the scenario of recolonisation of the archipelago in Oligocene and later. Molecular clocks, distribution and phylogenies of extant species, plate tectonics of the region, known fossil deposits…

    By the way, quite a lot is known about the geology of New Caledonia, as mining interests ensured thorough prospection,. But, for obvious historical reasons, a lot of it, especially field data from the 19th and 20th centuries, is in French, so may not be very accessible outside of the Francophone world. Still, I’ve found for example this overview: A.R. Lillie & R.N. Brothers, New Zealand Journal of Geology and Geophysics, Volume 13, Issue 1, 1970, pages 145-183. (PDF link to the complete article.)

    Link to this
  541. 541. irenedelse 3:05 pm 07/10/2014

    @Heteromeles #538:

    Great, why not a t-shirt with “Biologists who disagree with me are orthodox Darwinian evolutionists but I’m not a crank, nuh-uh”?

    Link to this
  542. 542. Heteromeles 3:26 pm 07/10/2014

    @541: That would work too.

    Whatever someone puts on the front, the back should say something like “*Facepalm* — David Marjanovic.”

    Ideally, this could be done with sharpie on plain t-shirts, just to encourage people to think of the shirts as individual works of art, rather than just profiteering off of minor memes.

    Link to this
  543. 543. Grehan1896 3:31 pm 07/10/2014

    540. irenedelse
    Right – Grandicolas et al

    None of [the existing] studies provides clear evidence for old local diversification since most dates inferred from molecular phylogenies do not pre-date the Oligocene (figure 2). Several emblematic groups such as Araucaria and Nothofagus have even undergone more recent radiations or colonizations of New Caledonia.”

    Here’s a commentary by Heads (2014, p. 284):
    In the fauna, Bauer et al (2006) dated New Caledonian gecko lineages back to the Late Cretaceous and ruled out the possibility that the whole island was submerged in the Paleogene.

    Murienne (2009a) also accepted the possibility of refugium islands. In another study Murienne (2009b: 1433) concluded that the different hypotheses – island refugia during flooding, versus long-distance dispersal after flooding – cannot be separated on the basis of the phylogenies alone and he approached the problem using molecular dating studies. These indicate that most New Caledonian groups are of Cenozoic age (cf. Grandicolas et al., 2008, Fig. 2) and so Murienne (2009b) proposed that the ‘only valid interpretation, for these groups is long-distance dispersal. Nevertheless, the two methods employed in the dating studies (treating the oldest fossils and clock dates derived from these – minimum ages – as maximum ages; using the age of the young islands now emergent on the Old Loyalty Ridge to date the taxa there) will both give underestimates of clade age (Chapter 1; Heads, 2008b). In the same way, a review of the molecular studies on New Caledonian plants found that all lineages were regarded as very young (< 37Ma) (Pillon, 2012), but the dates were all calculated using the same method – treating fossil-calibrated youngest possible clade dates as oldest clade dates.

    John Grehan

    Link to this
  544. 544. Grehan1896 4:00 pm 07/10/2014

    540. irenedelse
    Here’s a bit more on marine transgression in New Caledonia

    Heads 2014 p. 284

    “Deep-water sedimentary rocks dates as Paleogene occur in parts of New Caledonia. Some geologists (Pelletier, 2007; Cluzel et al., 2012) and biologists (Pole, 1994; Espeland and Murienne, 2011) have inferred from this that the entire land mass of New Caledonia was submerged from the start of the Paleocene until the mid-Eocene (65-45 Ma). These authors have supported an ‘entirely long-distance dispersal’ explanation for the island’s biota. Nevertheless, without stratum of a single age covering the island, total submersion is difficult to prove from the geological data, and in such a dynamic region tectonic uplift and subsidence are often quite localized”

    John Grehan

    Link to this
  545. 545. Heteromeles 4:37 pm 07/10/2014

    @544: So, if New Caledonia was never entirely submerged, why are there no native amphibians on the island? Why are the only native mammals bats? Why are nearly all the reptiles members of two groups (geckos and skinks) that are known to be good dispersers? The two snakes are also widespread genera that range over to the Moluccas, Australia, and Palau.

    That’s a pattern that looks like pure dispersal, not one that looks like an old island.

    Link to this
  546. 546. Grehan1896 5:07 pm 07/10/2014

    545. Heteromeles

    @544: So, if New Caledonia was never entirely submerged, why are there no native amphibians on the island? Why are the only native mammals bats? Why are nearly all the reptiles members of two groups (geckos and skinks) that are known to be good dispersers? The two snakes are also widespread genera that range over to the Moluccas, Australia, and Palau.
    That’s a pattern that looks like pure dispersal, not one that looks like an old island.

    The pattern of inhabitants that you mention would conform to expectations of pure dispersal (i.e. organisms with good means of dispersal). Two points on that – one is that there may be other groups that are more difficult to ‘explain’ in terms of means of dispersal or theorized dispersal ability (of course that does not mean to say that some way might be imagined to make it feasible), the other is that the means of dispersal may not be a predictor of how groups came to be in what is now New Caledonia, but how they survived. If the geological upheavals have been so profound as to give the impression of complete inundation, then it would be necessary that the organisms present would have to have the means of dispersal that would allow them to colonize new habitats as old ones were inundated. Such upheavals may result in considerable extinction of groups over time (and sometimes positively indicated by fossils).

    Of personal interest, NC has one species of ghost moth. It has up to a 5-6 inch wingspan. These are clumsy fliers and have no mouthparts. The usually do not survive more than a night, having burned up their fat reserves. Females are particularly poor fliers as they are filled with several hundred eggs and the fly so slowly (with the abdomen hanging down almost vertically) that one may almost grab them out of the air as they pass by. This insect does not have apparent good means of dispersal, but it is in NC (and yet its genus is not on the nearby Solomon Islands, Vanuatu or even Fiji (where there is an endemic monotypic genus of possible Australian and South American affinity).
    John Grehan

    Link to this
  547. 547. keesey@gmail.com 5:19 pm 07/10/2014

    “Before I comment on what I think might be wrong about the papers I need to know from you what you think is right.”

    No you don’t.

    Link to this
  548. 548. ectodysplasin 5:31 pm 07/10/2014

    @John,

    Those birds may have ‘got onto those islands’ the same way that they got onto those continents. They had a widespread ancestor that encompassed areas that are today both islands and continents. The ancestor established its distribution in the same way that any ancestral species does – by its means of dispersal acting in relation to the ecology and geography of the time (ecology is often overlooked in historical reconstructions as if geography were the only limiting factor as expressed in the view that a taxon should be present throughout an area if that area was contiguous).

    Interesting that you are willing to invoke transoceanic dispersal when it is required by your approach, but not when it provides a more parsimonious explanation of biogeography that doesn’t involve migration of island arcs across the Pacific and crown birds in the Jurassic.

    Link to this
  549. 549. MichaelHeads 5:42 pm 07/10/2014

    Replying to 545: In small islands such as New Caledonia, Fiji and the Hawaiian islands that have been exposed to major disturbance, such as flooding, uplift, subsidence, and volcanism, there has been a huge amount of extinction. Groups with suitable ecology have survived all the dynamism, others haven’t. Interpreting absence is always difficult – the groups may have been there and gone extinct, or they may never have been there.

    So, focusing on presence rather than absence, why is Amborella, the basal angiosperm (i.e. sister to all the others) only known from New Caledonia? It’s a very old group. There are no fossil records (pollen or macrofossils) from anywhere else on Earth. Just down the road on the same block of crust, in New Zealand, are the basal parrots (Strigopidae), the basal passerines (Acanthisittidae), etc. etc., again, with no extralimital fossils. The New Zealand-New Caledonia plant Xeronema is basal to a huge global clade including agaves, daffodils etc.

    You can explain all of these by chance dispersal, but it’s not simple – for all of these globally basal groups (there are dozens) you have to bring in ad hoc hypotheses of extinction elsewhere leaving no fossils.

    You mentioned the snakes in New Caledonia. Their distribution within the archipelago is especially interesting. There are no indigenous snakes on the main island, Grande Terre, but two families, Boidae and Typhlopidae, are represented on the Loyalty Islands (100 km off the mainland). It’s difficult to explain this if their dispersal is the result of chance, but it’s a common pattern.

    The Loyalties are an emergent part of an old, extinct arc (probably Cretaceous) that has crashed into New Caledonia from the north-east and has a biota that is quite distinct from the mainland. In many ways it is closer to that of Vanuatu, Fiji, etc.,and this is the case with the snakes.

    Link to this
  550. 550. ectodysplasin 6:08 pm 07/10/2014

    @Michael,

    There are fossil ratites in Asia (ostriches, east to China) and North America (lithornithids). But I agree, it’s very interesting that the ratites are not known from the Pacific islands apart from New Guinea and New Zealand. They are present, but apparently not native, on New Britain, but there are no records at all (living or fossil) from the rest of Melanesia (including New Caledonia), or Micronesia or Polynesia. This makes the apparent absence of fossils from all the Antilles (??? is this correct??) even more intriguing, as the Caribbean plate probably formed in the Pacific and there are many biogeographic similarities between the two regions.

    This can either be seen as “intriguing” or it can be seen as a potential strike against your hypothesis.

    Perhaps the ratites were in all or some of these regions and went extinct, with no fossil record. There are plenty of bird fossils on the islands, but these are mainly quite young, and the group could have gone extinct in the region long ago. Alternatively, ratites may have never been there. The same options apply to the region between China and New Guinea. I guess we can agree on this, so far.

    These are all more-or-less general statements that have little to do with your specific hypothesis. As I said before, there’s not a lot of occurrence data to work with when it comes to ratites, because there’s really low diversity in general.

    The idea that early ratites were volant makes their absence from the regions just cited even more intriguing, IF it is in fact real and not just the result of sampling error in the fossil record. I’ve written quite a bit on the great biogeographic break between the New Guinea mainland and the large islands of New Britain and New Ireland etc., just 50 km off the mainland. E.g. no birds of paradise on the islands (but major diversity of Platymantis frogs etc.).

    It’s not a particularly big issue if paleognaths have never been particularly speciose nor abundant.

    Of course, ratites do have relatives all through these areas – their sister group. So the question is: how to explain the great apparent gap in the otherwise global range of ratites?

    I think the mistake here is that you are vastly, vastly, vastly underestimating the importance of competitive exclusion. The reality is, the regions we see living ratites today are regions we see other “archaic” faunas: New Zealand, Australia, Madagascar, etc. These are areas where faunal isolation preserved taxonomic groups that were competitively excluded in other regions. This includes rhynchocephalians, monotremes, lemurs, and all sorts of other animals.

    Here is one possibility: the original split in crown group birds was between one group, the ratites, found everywhere except the Pacific, and another group, its sister (carinates s.str., = neognaths) allopatric with it, and restricted to the Pacific. (This would be an interesting parallel to the crown group frogs, with the basal group, Leiopelmatidae only known (living and fossil) from New Zealand and the western US (the Patagonian records are discounted by recent authors), straddling the Pacific). Subsequently, the ‘Pacific’ neognaths dispersed to everywhere in the world, while the ratites were much less successful, with little speciation anywhere except South America, and complete extinction in many areas (North America, Europe, Asia).

    I fail to see why neognaths would necessarily be Pacific in distribution; there’s basically zip-all evidence for that. What I do see evidence of is substantially higher rates of diversification in neognath birds in comparison with paleognaths, which would suggest that neognaths simply outcompeted paleognaths in areas where paleognaths couldn’t squeeze into niches on the edge of neognath nichespace. Incidentally, this seems to have been in areas without substantial competition from laurasiathere mammals, and in regions where large flightless neognaths also co-occurred (e.g. phorusrhachids and mihirungs).

    As for “leiopelmatids,” Ascaphus is not a leiopelmatid, so don’t appeal to Ascaphus. Origin of the Ascaphus+Leiopelmatidae clade occurred very early (probably earliest Jurassic) but most likely on mainland Pangaea. Archaeobatrachians are widespread in continental deposits throughout Pangaea. Current archaeobatrachians are relatively taxonomically depauperate, with highly localized distributions. Comparing archaeobatrachians with, say, bufonids or ranoids makes it pretty clear that the absence of living archaeobatrachians in most parts of the world is the result of competitive exclusion, not biogeography.

    Link to this
  551. 551. ectodysplasin 6:20 pm 07/10/2014

    @Michael,

    You mentioned the snakes in New Caledonia. Their distribution within the archipelago is especially interesting. There are no indigenous snakes on the main island, Grande Terre, but two families, Boidae and Typhlopidae, are represented on the Loyalty Islands (100 km off the mainland). It’s difficult to explain this if their dispersal is the result of chance, but it’s a common pattern.

    The historical happenstance behind current boid distribution is a lot clearer now that folks are looking at boid fossils. What seems to have happened is that boids were in fact globally widespread in the Paleocene and Eocene, and were then systematically wiped out by glaciation in Europe, northern Asia, and North America, and by competition from pythons in Africa, India, and Australia. We see boids today in local refugia (Charina along climatically mild coastal North America and erycines in desert regions of Africa and Eurasia) or isolated islands (sanziniids in Madagascar and Candoia in Polynesia), but these taxa all used to be much more widespread.

    Link to this
  552. 552. Grehan1896 6:38 pm 07/10/2014

    547. keesey@gmail.com

    “Before I comment on what I think might be wrong about the papers I need to know from you what you think is right.”
    No you don’t.

    From my point of view you do, otherwise I don’t know what to respond to with respect to what you see as validating evidence.

    John Grehan

    Link to this
  553. 553. DavidMarjanovic 7:08 pm 07/10/2014

    Global ratite ancestor underwent vicariance at Mozambique Channel and also somewhere between China (ostrich fossil) and New Guinea (cassowaries), isolating the ostriches.

    …Wow, so you’re postulating Jurassic paleognaths now. Move over, Anchiornis, get out of the way of that ostrich!

    It’s not just China, BTW, but also Ukraine and probably everywhere in between.

    Does anyone know of any ratite fossils in the Greater or Lesser Antilles?

    The Lesser Antilles are volcanic, they simply aren’t old enough. The Greater Antilles are far away from South America.

    I didn’t say it was independent evidence. It’s just more evidence.

    That’s the exact same thing! Do you even listen to yourself? If it’s not independent evidence, it’s part of the same piece of evidence, for the sake of vertical gene transfer! Counting it separately is like having correlated characters in a phylogenetic analysis!

    globally basal groups

    What do you mean by that?

    Storms, Madagascar etc.: so why are there no monkeys on Madagascar? And no strepsirrhines on America?

    Ultimately, random chance. Dispersal of such large mammals over such distances is simply an extremely improbable event.

    This interesting symmetry can be explained by vicariance, followed by overlap in Africa and Asia.

    Only if you completely and utterly ignore all the data from the fossil record.

    As I’ve said before, Strepsirrhini was all over the three northern continents in the Eocene, then showed up in Africa where the oldest known crown-strepsirrhines – the oldest known stem-lorisoids and perhaps the oldest known stem-lemuroids – are found, with Madagascar lacking a terrestrial fossil record in the Cenozoic but having a primate-free fossil record in (small parts of) the Jurassic and Cretaceous. I need to find the paper with the latest phylogenetic analysis of early haplorrhines and possible haplorrhines, because there’s been a lot of confusion; anyway, both monkeys (Simiiformes) and apparently tarsiers (Tarsiiformes) show up in southeast Asia before Africa, where the Eocene/Oligocene of Africa (IIRC) has yielded close relatives of Anthropoidea (the simiiform crown-group); the oldest platyrrhine (a stem-platyrrhine) is from the Oligocene of South America. Primates other than platyrrhines (and humans) are unknown from South America, and platyrrhines are only known from South and Central America including various Caribbean islands. Your hypothesis appears to predict a Cretaceous, if not Jurassic fossil record of platyrrhines in North America that plainly isn’t there.

    “They only reached Sulawesi quite recently.”

    Do you have any evidence?

    Tarsier phylogeny and the intriguing geologic history of Sulawesi.

    Nasalis is an excellent swimmer, probably the best primate swimmer in the sea, but it has never reached Sulawesi from Borneo.

    Maybe it’s precisely because it can swim so well that it doesn’t get blown out to sea involuntarily.

    Do you think that Wallace’s line itself, as seen in many primates, is the the result of geology? If so, couldn’t the other lines be as well, given that they coincide with major tectonic breaks and also western limits in eastern groups?

    Wallace’s line is simply the end of the Asian continent. Every ice age you can walk from here to Wallace’s line, which is where the shore is. Likewise, Lydekker’s line is the end of the Australian continent. Anything between these two lines, these two continents, is only accessible by overseas dispersal even if it’s only 20 km. That means all the restrictions kick in tha