About the SA Blog Network

Tetrapod Zoology

Tetrapod Zoology

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

Hoatzins are no longer exclusively South American and once crossed an ocean

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

Email   PrintPrint

Like a lot of people interested in birds and weird animals in general, I have a great love of the Hoatzin Opisthocomus hoazin (yes, the binomial name does not have the same spelling as the vernacular one).

Hoatzins photographed in Brazil by Cláudio Dias Timm, from wikipedia.

I’m sure most Tet Zoo readers are already familiar with this species, but – for the few of you that aren’t – this is that famously strange bird of the Amazon and Orinoco basins that typically lives in the trees along waterways and digests the leaves it eats in an enlarged crop (Grajal et al. 1989). This ‘foregut fermentation’ style of digestion is unique. Hoatzins are also unusual in possessing crenulated bill margins that are actually used in chewing (Korzun et al. 2003).

Climbing hoatzin chick. Image by J. Arthur Thomson, from wikipedia.

The one factoid that most people know about the Hoatzin is that its chicks possess obviously clawed fingers and use them in climbing. While it’s often stated or implied that this is a primitive feature that links the Hoatzin with Archaeopteryx and other Mesozoic theropods, the truth is that hand claws are actually widespread in modern birds and not at all unusual. If you’re a regular Tet Zoo reader this should sound somewhat familiar, since I covered the subject of manual claws in birds back in June 2010. Oh, ‘Hoatzin’ is pronounced something like ‘what-seen’ or ‘what-son’.

Anyway, I digress. People who pay attention to my twitter feed (I know that at least two of you do) might have noticed my reference to the ‘Neornithine Holy Grail’ a few weeks ago. What the hell was I referring to? As usual I wasn’t being entirely serious, and was also employing a degree of hyperbole. I was actually referring to a paper – published earlier this week in Naturwissenschaften by Gerald Mayr and colleagues – that provides some fascinating new information on hoatzin history and origins, and also identifies two new fossil hoatzin species.

Foro panarium, from Olson (1992).

We need to note to begin with that fossil hoatzins have been known for a while. In 1953, Miller (1953) named Hoazinoides magdalenae from the Middle Miocene of Colombia. It’s known from a partial skull that’s highly similar to that of the modern Opisthocomus but somewhat larger. I should also mention that an alleged hoatzin relative called Onychopteryx simpsoni (named for a partial tarsometatarsus) was reported from the Eocene of Argentina in 1971 (Cracraft 1971). However, the fossil concerned is so scrappy that few experts have been prepared to accept this as a valid hoatzin record. And then there’s Foro panarium from the Eocene of North America, described by Olson (1992) as superficially hoatzin-like. It’s possible that this bird (represented by fairly substantial remains, as you can see here) is allied with hoatzins, but so far only superficial similarities have been alluded to and detailed evidence for a relationship hasn’t been presented.

Two new fossil hoatzins

Mayr et al. (2011) now give us two additional fossil hoatzins. The first is named Hoazinavis lacustris and is known from a set of wing and chest bones, collected in 2008 from the Upper Oligocene-Lower Miocene Tremembé Formation of São Paulo, Brazil. All of Hoazinavis’s bones are very similar to those of Opisthocomus (though there are enough differences to warrant clear taxonomic distinction), but they come from a rather smaller bird (and the bone surface texture shows that it was an adult at the time of death). Given the problematic status of both Onychopteryx and Foro, Hoazinavis is currently the oldest known hoatzin.

Elements of Namibiavis and Hoazinavis, from Mayr et al. (2011). Images (c) Senckenberg.

Then we come to the second hoatzin. It’s known from three coracoids and six humeri, all of which come from Lower Miocene sediments. It isn’t actually a new discovery: it was named Namibiavis senutae by Cecile Mourer-Chauviré in 2003.

As you might already be guessing from its name, the big deal is that it’s not South American – it’s from Namibia! Mourer-Chauviré (2003) originally identified Namibiavis as a member of the seriema-like Idiornithidae but reanalysis shows that it definitely possesses distinctive hoatzin features (Mayr et al. 2011). The original idiornithid identification was certainly understandable since idiornithids (and seriemas) are hoatzin-like in several respects. Indeed, several authors have previously suggested that all of these birds might be close allies (e.g., Mourer-Chauviré 1983, Olson 1985, Mayr and Clarke 2003). Anyway, what little we know of Namibiavis suggests that it was, again, generally similar to Opisthocomus and presumably similar to it in flight abilities and perhaps in ecology and appearance.

Global range of the extant hoatzin, from wikipedia.

So, it’s official – hoatzins are no longer exclusively South American; they’re also known from south-western Africa. This is an exciting discovery given that hoatzins have always been regarded as quintessentially South American. What might this discovery mean for hoatzin history and evolution? And how on earth could ancient hoatzins be present in both Africa and South America?

“No” to vicariance

For starters, you can forget the idea that this distribution is best explained by vicariance: that is, by the idea that both South American hoatzins and Namibiavis descend from a population that was once continuous across both landmasses, but then evolved into two lineages once the Atlantic formed.

You can forget this because Africa and South America rifted apart in the Cretaceous, and were substantially separated by a young Atlantic by the Late Cretaceous. There are no indications that hoatzins are anywhere near old enough to have pre-dated the opening of the Atlantic. Indeed, the opening of the Atlantic even pre-dates the appearance of Neoaves (the major clade that includes all crown-birds except palaeognaths, waterfowl and gamebirds) and probably Neognathae (waterfowl, gamebirds and neoavians) and Neornithes (palaeognaths and neognaths).

“No” to northern or southern over-land routes

Could the earliest hoatzins have dispersed northwards (using North America, Greenland and Europe as stepping stones between Africa and South America), or southwards (using Antarctica)? Neither proposal is parsimonious, for three reasons. Firstly, there aren’t any fossils that might support either model of dispersal (unless Foro really is a stem-hoatzin). Secondly, living and fossil hoatzins are all tropical birds, likely constrained to waterside habitats and a consistent supply of a certain kind of edible foliage. It would be a real stretch to imagine extinct hoatzins surviving in, and travelling through, the cool temperate and sub-Arctic regions of the extreme north and south.

Thirdly, stretches of water separated South America from North America, South America from Antarctica, Africa from Antarctica and Africa from Eurasia during parts of the Eocene, Oligocene and Miocene. Indeed, both South America and Africa were island continents during much of the Oligocene and Miocene. In short, positing over-land dispersals of this kind would potentially involve several over-water dispersal events, and would also require that early hoatzins travel into cool parts of the world that they almost certainly weren’t suited for. Incidentally, these objections also apply to other tetrapod groups (like, apparently, phorusrhacids) found in Africa and South America but not elsewhere.

Palaeomap of the Early Oligocene, from Mayr et al. (2011), showing distribution of the four fossil and modern hoatzins. Image (c) Ron Blakey, Colorado Plateau Geosystems, Inc.

If both vicariance and over-land dispersal are out the window, how else might we explain hoatzin distribution? The only reasonable answer is trans-Atlantic dispersal. Hoatzins must have crossed the Atlantic Ocean some time during the Cenozoic. In view of their limited flight abilities, it’s unlikely that they flew, even if we do allow for the presence of a narrower Atlantic and various island archipelagos. Rafting is the only alternative.

Because we regard hoatzins as South American birds, it might seem natural to assume that the group is ancestrally South American and that its presence in Africa represents a west (South America) to east (Africa) dispersal. The fact that the oldest hoatzin – Hoazinavis from the Late Oligocene or Early Miocene – is South American could be interpreted as support for South American ancestry.

However, while there are a few groups of animals that have moved across the Atlantic from west to east – like manatees (Domning 2005) and thrushes (Voelker et al. 2009) – animals that have rafted across the ocean have otherwise all moved from east to west. Caviomorph rodents, platyrrhine primates, various amphisbaenian and gekkotan lizards, and others, are thought to have rafted from Africa to South America. Remember that these ‘rafts’ shouldn’t be imagined as small, simple structures, but as sometimes enormous, complex messes of live vegetation, tree trunks and sediment. They more resemble floating islands than ‘rafts’, and can be tens of metres long. The only half-decent image I can find of such a structure is this painting of an Amazonian raft [it comes from an encyclopedia series called The Joy of Knowledge, published by Mitchell Beazley in 1987, but I can’t find an artist to credit]. It does a nice job of showing the ‘raft’ to be complicated and with structure and areas of refuge, but ideally I want a picture showing a much larger raft.

There’s some suggestion that, prior to the formation of the Central American landbridge in the Pliocene, winds and currents made it easier for volant animals to move from South America towards Africa (Voelker et al. 2009). However, most data indicates that moving westwards across the Atlantic was easier in the past, as it is now. For these reasons, Mayr et al. (2011) favour the idea of an east to west dispersal and thus support the idea that hoatzins began their history in Africa. As Mayr et al. (2011) note, because hoatzins typically live in or on the vegetation that flanks very large, powerful rivers they might be at special risk of being swept out to sea during times of excessive flooding. Furthermore, their folivorous habits might have proved advantageous as goes transportation on a massive pile of floating vegetation. This is the first proposed instance of trans-oceanic rafting by a group of birds.

The only photo I have of a hoatzin skeleton. Yeah, it's terrible.

As noted earlier in this article, uncertainties over the affinities of hoatzins make it difficult to work out where hoatzin ancestors and close relatives ‘should’ be distributed. Hoatzins have at times been considered close to gamebirds, cuckoos, or pigeons (Sibley & Ahlquist 1973, Hedges et al. 1995, Hughes & Baker 1999, Sorenson et al. 2003), all of which are widely distributed and not clearly associated with any specific ‘ancestral’ area. Hoatzins have also been considered close to the exclusively African turacos (supposed fossil turacos from Europe are likely misidentified: Mayr 2009). They’ve also been suggested to be close to the South American seriemas (e.g., Mourer-Chauviré 1983, Olson 1985, Mayr and Clarke 2003). However, because close fossil relatives of seriemas were present across Europe and North America during the early Cenozoic, it isn’t necessarily clear that a possible link between hoatzins and seriemas would constrain the hoatzin ‘centre of origin’ to South America.

An approximate tree for neognaths (though with some groups in controversial positions). Where do hoatzins go? Take your pick.

And of course I could say a lot more about the possible affinities of hoatzins, since it’s one of the most contentious and debated issues in neornithine systematics (see also Fain & Houde 2004, Ericson et al. 2006, Hackett et al. 2008). Mayr et al. (2011) don’t come down in favour of any one particular hypothesis on hoatzin affinities, but they do note that a close relationship with turacos is most consistent with an African origin.

Given both the enigma and controversy that’s surrounded hoatzin affinities and the various morphological and physiological peculiarities evolved within the hoatzin lineage, any news on the group’s history and evolution is of great interest. I’m very excited to know that hoatzins once inhabited Africa, and the fact that these birds seemingly once rafted across an ocean makes them all the more interesting. As usual, we must hope that – in time – additional discoveries will tell us more about the history of this most remarkable group of birds.

For previous Tet Zoo articles that mention or discuss hoatzins, see…

Refs – -

Cracraft, J. 1971. A new family of hoatzin-like birds (Order Opisthocomiformes) from the Eocene of South America. Ibis 113, 229-233.

Domning, D. P. 2005. Fossil Sirenia of the West Atlantic and Caribbean region. VII. Pleistocene Trichechus manatus Linnaeus, 1758. Journal of Vertebrate Paleontology 25, 685-701.

Fain, M., & Houde, P. (2004). PARALLEL RADIATIONS IN THE PRIMARY CLADES OF BIRDS Evolution, 58 (11), 2558-2573 DOI: 10.1111/j.0014-3820.2004.tb00884.x

Grajal, A., Strahl, S. D., Parra, R., Dominguez, M. G. & Neher, A. 1989. Foregut fermentation in the hoatzin, a Neotropical leaf-eating bird. Science 245, 1236-1238.

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.

Hedges, S. B., Simmons, M. D., van Dijk, M. A. M., Caspers, G.-J., de Jong, W. W. & Sibley, C. G. 1995. Phylogenetic relationships of the hoatzin, an enigmatic South American bird. Proceedings of the National Academy of Sciences 92, 11662-11665.

Hughes, J. M. & Baker, A. J. 1999. Phylogenetic relationships of the enigmatic hoatzin (Opisthocomus hoazin) resolved using mitochondrial and nuclear gene sequences. Molecules and Biological Evolution 16, 1300-1307.

Korzun, L. P., Erard, C., Gasc, J.-P. & Dzerzhinsky, F. J. 2003. Biomechanical features of the bill and jaw apparatus of cuckoos, turacos and the hoatzin in relation to food acquisition and processing. Ostrich 74, 48-57.

Mayr, G. 2009. Paleogene Fossil Birds. Springer, Berlin.

- ., Alvarenga, H. & Mourer-Chauviré, C. 2011. Out of Africa: Fossils shed light on the origin of the hoatzin, an iconic Neotropic bird. Naturwissenschaften DOI:10.1007/s00114-011-0849-1

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

Miller, A. H. 1953. A fossil hoatzin from the Miocene of Colombia. Auk 70, 484-489.

Mourer-Chauviré, C. 1983. Les Gruiformes (Aves) des Phosphorites du Quercy (France). 1. Sous-order Cariamae (Cariamidae et Phorusrhacidae). Systématique et biostratigraphie. Palaeovertebrata 13, 83-143.

- . 2003. Birds (Aves) from the Middle Miocene of Arrisdrift (Namibia). Preliminary study with description of two new genera: Amanuensis (Accipitriformes, Sagittariidae) and Namibiavis (Gruiformes, Idiornithidae). Geological Survey of Namibia, Memoir 19, 103-113.

Olson, S. L. 1992. A new family of primitive landbirds from the Lower Eocene Green River Formation of Wyoming. Natural History Museum of Los Angeles County, Contributions in Science 36, 137-160.

- . 1985. The fossil record of birds. In D. S. Farner, J. R. King, and K. C. Parkes (eds.), Avian Biology 8: 79-238. Orlando, Fla.: Academic Press.

Sibley, C. G. & Ahlquist, J. E. 1973. The relationships of the hoatzin. The Auk 90, 1-13.

Sorenson, M. D., Oneal, E., García-Moreno, J. & Mindell, D. P. 2003. More taxa, more characters: the hoatzin problem is still unresolved. Molecular Biology and Evolution 20, 1484-1499.

Voelker, G., Rohwer, S., Outlaw, D. C. & Bowie, R. C. K. 2009. Repeated trans-Atlantic dispersal catalysed a global songbird radiation. Global Ecology and Biogeography 18, 41-49.

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 He has been blogging at Tetrapod Zoology since 2006. Check out the Tet Zoo podcast at! Follow on Twitter @TetZoo.

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

Rights & Permissions

Comments 56 Comments

Add Comment
  1. 1. BrianL 10:07 am 10/5/2011

    This is very exciting indeed!

    I’m also glad you’re so dismissive of attempts to explain neornithean distribution patterns using vicariance. Far too often molecular clock estimates and calibration points rely on assumptions of divergence through vicariance. Given how great birds tend to be at oversea dispersal assuming them to remain confined to landmasses for many millions of years in order to explain present day distributions is rather absurd, to me. Every time I see things like the split between strigopids and other crown group parrots or the split between various ratites being calculated using continental drift events like New Zealand seperating from Australia or South America splitting from Africa, I have to overcome a very strong desire to facepalm (and sometimes I can’t).

    Is there any good reason why molecular phylogeny relies so often on vicariance events for its divergence estimates and denies the dispersive abilities of neornitheans?

    By the way, I’m quite surprised that a supposed African idiornithid and thus member of Cariamae was already known since 2003. I imagined that African phorushacid to have been the first one described from the continent.

    Given the apparently easier way from Africa to South America compared to vice versa, I have to wonder about that supposed African phorusrhacid though. While I suppose it might have rafted in either direction, the fact we’re dealing with a large, terrestrial, flightless bird here makes rafting somewhat doubtful though not impossible to me. Moreover, it’s apparent great resemblance to specific, derived phorusrhacids would certainly suggest such a rafting event to have been from South America to Africa. I’m much more in favour of the resemblance between these African and South American being convergent, with both evolving from bathornithid-grade Cariamae.

    Link to this
  2. 2. Heteromeles 11:00 am 10/5/2011

    Don’t forget that mangrove forest once stretched up to London (albeit in the Eocene) and mangrove forests edged the borders of the Tethys all around the world (including in the Eocene Sahara). I’m not sure when Hoatzins became Hoatzins (all fossils are Miocene), but if they are mangrove and riverine specialists, there would have been more forests for them back further in the Cenozoic than there are now.

    Link to this
  3. 3. Jerzy New 1:29 pm 10/5/2011

    I think it is easier to assume that some ancestors of hoatzins were more volant than today.

    Actually, I also never seen a picture of these vegetation rafts, which are so important concept in zoogeography. Pretty strange. Do ocean-going ships sometimes hit mats of vegetation, with or without a jaguar resting on top? How long would such a raft cross the Atlantic – or proto-Atlantic? Wouldn’t any animal starve or be killed by sea salt? Was this concept ever examined critically?

    Link to this
  4. 4. David Kelly 3:08 pm 10/5/2011

    Coincidentally a query has just been raised on Birdforum about a Ruddy Ground Dove Columbina talpacoti in the environs of Cape Town. This also discusses a few other Neotropical vagrant birds to Africa.



    Link to this
  5. 5. naishd 4:48 pm 10/5/2011

    Thanks for comments.

    Heteromeles (comment 2): even if mangroves were more extensive across the regions you discuss, I’m not sure that it makes hoatzin distribution any easier to explain. To get, hypothetically, from tropical Paleogene Europe to tropical Paleogene America, you have to go via cool Greenland/Ellesmere or whatever – not exactly an ideal route for a group of birds restricted (so far as we know) to lush, tropical, riverine habitats.


    Link to this
  6. 6. naishd 5:13 pm 10/5/2011

    I’m always a bit surprised when people express incredulity or scepticism about the existence of the floating mega-rafts required to explain biogeographical ‘rafting’. Fact is, such rafts are well recorded and not all that unfamiliar if you know the literature on such things. Most of these are reported floating out of big rivers in South America and from the Caribbean, but there are others seen emerging from the Congo and from the Sepik on New Guinea. Some of the New Guinea ones were about 100 m wide and ‘floating islands’ seen carrying animals have been seen 240 km out at sea, off the Atlantic coast of Africa. As for collisions with ships, in the 1905 flood of the Río de la Plata in Buenos Aires, rafts as much as 800 m across were said to collide with moored ships and tear them from their anchor points. Animals seen on these rafts included a puma, a deer, monkeys and snakes. Photos confirm observations that some of these rafts had standing trees, often very large ones. So – - is it really so hard to believe that such occurrences have been occurring throughout history, sometimes carrying animals and plants to new homes far away? In cases you might only need a single pregnant individual to make a successful crossing for a population to become established – and this has been observed in modern times too (in 1995, iguanas made the 320 km crossing from Guadeloupe to previously iguana-free Anguilla, and a breeding population is now established there).

    Virtually all of this information comes from Paul V. Heinrich and Mike Everhart – thanks, guys.


    Link to this
  7. 7. KarMannJRO 6:14 pm 10/5/2011

    Another point to stress would be the timescales involved. It’s kind of like how creationists seem to have trouble grasping what can happen with millions or billions of years for changes to accumulate. So, even if these rafts were what we would consider rather rare events, say once per millennium on average, that would imply something like 18,000 rafting events during the Miocene, just to pick an epoch at random. Even the very rare things can happen often with enough time. (See also: Infinite monkeys, infinite typewriters, infinite patience.)

    Link to this
  8. 8. adzebill 6:51 pm 10/5/2011

    Rafting would indeed be something, as there’s precious little evidence of a poorly-flighted or flightless bird dispersing over water. I’d be interested in seeing an analysis of the humerus and corocoid that establishes these fossil hoatzin were as bad at flying as the present-day species. Given avian flying abilities are pretty plastic over relatively short geological time scales, I would be cautious about deciding this had to be rafting; and the Namibian species dates from after the dispersal event if we assume Mayr’s scenario.

    Imagine if the hoatzin was flightless, and so were all its fossil ancestors, the best explanation for this pattern might still be flying dispersal and independent flightlessness events (as seems to have occurred with the ratites). Perhaps we shouldn’t assume all these birds have always been poor fliers.

    Link to this
  9. 9. vdinets 6:56 pm 10/5/2011

    If trans-Atlantic rafting was such a popular way of moving between continents, how come all those islands in the mid-Atlantic have zero non-volant mammals, large reptiles or amphibians? The Azores, Cabo Verde, Saint Helena, Ascension? Some of them are a bit arid nowadays, but were much better forested before goat introduction.

    Link to this
  10. 10. smonkey 7:46 pm 10/5/2011

    Here’s a photo of a not-so-mega raft 9 nautical miles out to sea.

    So there ya go.

    Link to this
  11. 11. victorg 7:47 pm 10/5/2011

    @vdinets: It’s much easier to hit Africa or South America with something you send floating in the ocean than hitting the tiny (by comparison) islands. All that biogeography stuff about ease of colonization and resilience to extinctions being correlated to area would also apply, right?

    Being Brazilian, I’ve heard about how people walk, go fishing and sometimes set camp on (rather static) floating vegetation mats (balseiro, baceiro, camalote, ilha flutuante, etc.), that these mats create enough soil for trees to grow on them and that on occasion these are carried by floods. It’s rather common knowledge, even if in school I was rather taught the “floods take bits of river shore with trees into the ocean” version (which is also true, but floating mats are way more common).

    There’s a lot of literature about ecological succession on these aquatic macrophyte mats, fish assemblages associated with them, etc., specially in the Pantanal and Amazon biomes (both landscapes shaped by floods).

    So it’s a bit maddening to be unable to find good pictures of these. Here’s a video of fishing in one such mat[1], and a picture of a tiny mat in the Amazon[2].


    Link to this
  12. 12. victorg 7:55 pm 10/5/2011

    Some interesting photos and descriptions from Florida. Imagine this in a much larger scale (huuge mats, rivers kilometers wide, flash floods aka repiques, large >2m waves coming from the ocean into the Amazon river aka pororoca).

    Link to this
  13. 13. KarMannJRO 8:12 pm 10/5/2011

    @vdinets: This might seem to be somewhat contrary to my earlier comment, but I would think it’s just that an island, or even an archipelago, is a much smaller target than a continent. And you just need one pregnant female or breeding pair to introduce a species, regardless of whether it’s an entire vast continent or just a speck of an islet.

    Link to this
  14. 14. victorg 8:33 pm 10/5/2011

    Nice example from a great paper on West Indian colonization (Hedges, 2006):

    Hedges, S. Blair. 2006. Paleogeography of the Antilles and Origin of West Indian Terrestrial Vertebrates. Annals of the Missouri Botanical Garden 93 (2), 231-244. (Available from the author, along with many sweet rafting-related papers: )

    Link to this
  15. 15. josimo70 8:46 pm 10/5/2011

    Is it plausible that transatlantic proto-hoatzin route was the same New World monkeys and caviomorphs crossed the sea? I think South American Geochelone and Amphisbaenids (and perhaps Cathartids too).

    Link to this
  16. 16. John Harshman 8:56 pm 10/5/2011

    We may not know what hoatzins are, but we know quite a bit about what they aren’t. They clearly aren’t “land birds” or charadriiforms, and they are neoavian. The only locus that seems to care about them at all is beta-fibrinogen, and it puts them in different places depending on details of analysis, among the mess that is “water birds” sensu lato and “Metaves”, if there is anything resembling that group. There really needs to be a hoatzin genome project so we can get us some transposon insertions relevant to the case.

    Link to this
  17. 17. ggarbino 10:03 pm 10/5/2011

    Very cool post ! The african “rafters” are one of my favorite themes in Biogeography ! When you cited manatees and thrushes a case can be made also for the Angiosperms: the Bromeliads are restricted to the Americas but there is a single species in west Africa !
    One very intriguing and often forgotten group of mammals that crosses from Africa to America are the Emballonuridae bats. I don’t know many papers that deal with the theories regading how this family colonized South America.

    Link to this
  18. 18. vdinets 11:19 pm 10/5/2011

    KarMannJro: there is a difference in target size, of course. But only a small section of a continent is accessible by drifting down a current at any given time. Right now, for example, the only way to float from tropical Africa to S America is with South Equatoiral Current. It is about 100 km wide. Ascension Island is smack in the middle of it, and is 10 km north to south. It was probably much larger in the past, as it is slowly sinking. Still, prior to recent introductions, it had no land birds, mammals or herps, despite being forested. In fact, no island in the Mid-Atlantic has a single species of possible “raft” origin, even though their total north-to-south span is about 300 km.
    By the way, has anyone ever seen a large vegetation raft more than 500 miles from land? It has to be really huge, with enough fresh water for a band of monkeys to survive many months of sea crossing.

    Link to this
  19. 19. Therizinosaurus 3:03 am 10/6/2011

    Surely being known from a complete skeleton, it wouldn’t be hard to check Foro for those characters used to assign scrappy wing bone based taxa to Opisthocomidae?

    Link to this
  20. 20. Dartian 6:24 am 10/6/2011

    How long would such a raft cross the Atlantic – or proto-Atlantic?

    See Houle (1998), especially Table 5. In his calculations, he took account of paleocurrents and paleodistances, and assumed a starting point in modern-day Liberia and a landing point in Paraíba (westernmost part of Brazil). 50 million years ago, when that distance would have been ca. 1000 km, the crossing would have taken 5-8 days; 40 Mya, when the distance was ca. 1400-1500 km, it would have taken 7-11 days; and 30 Mya, when the distance was ca. 1900-2000 km, it would have taken 10-15 days.

    a case can be made also for the Angiosperms: the Bromeliads are restricted to the Americas but there is a single species in west Africa

    There are actually several groups of angiosperms that seem to have made the transatlantic crossing. For example, one species of cactus, the epiphytic Rhipsalis baccifera, has dispersed across the Atlantic to the Old World, and apparently relatively recently too (all other species of Rhipsalis are restricted to the Americas). Incidentally, among plants, dispersal by water seems to have happened in both directions, but dispersal by wind has more frequently happened from South America to Africa than the other way around (Renner, 2004).

    many months of sea crossing

    Not months. About two weeks or less; see above.

    Houle, A. 1998. Floating islands: a mode of long-distance dispersal for small and medium-sized terrestrial vertebrates. Diversity and Distributions 4, 201-216.

    Renner, S. 2004. Plant dispersal across the tropical Atlantic by wind and sea currents. International Journal of Plant Sciences 165, S23-S33.

    Link to this
  21. 21. josimo70 7:38 am 10/6/2011

    Perhaps cariamid-like Old World birds like idiornithids and bathornithids (and maybe the mysterious putative ratite Eremopezus)were closer to these African hoatzin stock, and their cariamid-like traits were convergent.What would be the neoavian equivalent of mammalian Afrotheria? I meant to say: any chance of an endemic African Paleogene avian clade?

    Link to this
  22. 22. David Marjanović 8:35 am 10/6/2011

    Is there any good reason why molecular phylogeny relies so often on vicariance events for its divergence estimates and denies the dispersive abilities of neornitheans?

    No, just ignorance of paleontology (paleobiology, paleobiodiversity, history of the Earth, paleogeography, everything).
    I feel your forehead pain.

    If trans-Atlantic rafting was such a popular way of moving between continents, how come all those islands in the mid-Atlantic have zero non-volant mammals, large reptiles or amphibians? The Azores, Cabo Verde, Saint Helena, Ascension?

    How old are these islands? Aren’t they small targets in time as well as in space? And the younger they are, the farther they have always been from the nearest continent…

    Link to this
  23. 23. KarMannJRO 8:41 am 10/6/2011

    Well, shoot, I was going to bring up the African/Asian cactus, but our esteemed host beat me to it. Just as well; my fuzzy memory had me thinking it was only in Madagascar, of all places, and I certainly didn’t remember the species.

    And for the record, I’d like to point out that victorg’s comment hadn’t shown up yet when I made mine, presumably because of the two links. I wouldn’t have bothered being that redundant if I’d known!

    Link to this
  24. 24. KarMannJRO 8:42 am 10/6/2011

    *sigh* And further self-correction: That wasn’t ‘our esteemed host’, that was Dartian. I’ll just shut up now.

    Link to this
  25. 25. David Marjanović 8:44 am 10/6/2011

    Surely being known from a complete skeleton, it wouldn’t be hard to check Foro for those characters used to assign scrappy wing bone based taxa to Opisthocomidae?

    It’s preserved as a slab. Perhaps the characters are all on the wrong side of the animal.

    But it’s more probable that nobody has looked. Given his record, I’m sure Mayr will get to it sometime. :-)

    Link to this
  26. 26. Dartian 9:02 am 10/6/2011

    How old are these islands?

    Dunno about the others, but Ascension is only about one million years old.

    There are, incidentally, 19th century reports of a native species of skink (probably a Mabuya species) living on Ascension island; unfortunately, it’s almost certainly extinct now.

    That wasn’t ‘our esteemed host’, that was Dartian.

    Well, to be confused with Darren is hardly the worst thing that can happen to someone… ;)

    Link to this
  27. 27. naishd 9:04 am 10/6/2011

    KarMannJRO: yeah, sorry, victorg’s comments were quarantined for a few hours (while I was asleep) due to the links. On Foro, let me ask Gerald Mayr…

    Great discussion, by the way. I really appreciate you all taking the time and effort to leave comments. Things ARE changing behind the scenes.. it’s just that these things take time. Lots and lots of time.


    Link to this
  28. 28. naishd 9:54 am 10/6/2011

    Ok, I asked Gerald Mayr about Foro and the possibility of it being allied to hoatzins, and this is what he had to say…

    “I do not know of features in the postcranial skeleton of Foro that would support identification as a stem lineage representative of Opisthocomidae. It is mainly the skull that shows similarities to the Hoatzin (with the shape of the upcurved retroarticular process being a possible synapomorphy mentioned by Olson). As far as comparisons are possible, the humerus is quite different from that of Hoazinavis, Namibiavis, and Opisthocomus and lacks the derived shape of the deltopectoral crest we described in the Naturwissenschaften study. The coracoid of Foro is not well visible, but what is exposed (part of the omal extremity) is different from opisthocomiforms (including the new fossils). The poorly resolved phylogenetic affinities of Foro are in some way connected with the unknown relationships of Opisthocomus. If there exist no strongly supported phylogenies of the extant species, with all the anatomy available (soft issue etc.), it may be difficult if not impossible to place a weird Eocene fossil in a phylogenetic context. Foro may well be “closely related” (whatever this means) to Opisthocomus, but whether it is indeed the sister taxon of a clade including Hoazinavis, Namibiavis, and Opisthocomus is another issue. Because Foro is morphologically very different from Namibiavis, Hoazinavis, and Opisthocomus (whose osteology is quite similar), it does not bear on the distribution issues discussed in the paper. To show that opisthocomiforms dispersed across the Northern Hemisphere, one would need fossils of Hoazinavis or Namibiavis-like birds.”

    So, there you go. My thanks to Gerald for providing this pers. comm.


    Link to this
  29. 29. Heteromeles 9:56 am 10/6/2011

    If it hasn’t been done already, someone needs to do with vegetation rafts is to start attaching small GPS buoys to them, preferably with a camera attached, so that their movements and changes can be documented over time.

    Right now, rafts seem to be in the same category as the botanist’s myth of the constipated seagull, which is invoked to explain amphitropical distributions of plant genera like Larrea. Gulls don’t eat creosote, but something has to have moved them…

    Link to this
  30. 30. naishd 11:00 am 10/6/2011

    Woo-hoo, this article is now at the top of the ‘most commented posts’ list. Thank you all :)


    Link to this
  31. 31. victorg 12:11 pm 10/6/2011

    I bet that if you write an article about rafting and the major (and/or most interesting) distributions it explains, it’d be a reddit TIL hit. I could even resume my search for photos ;)

    Another strategy would be to do an AMA (Ask Me Anything, see on reddit.

    Link to this
  32. 32. vdinets 1:40 pm 10/6/2011

    Dartian: St. Helena is 15 mln years old, Tristan group islands 5 to 30. As for Cabo Verde Islands, Wiki claims 180 mln years – older than the Atlantic Ocean. Does anybody know what their real age is?

    Link to this
  33. 33. vdinets 1:44 pm 10/6/2011

    Heteromeles: the problem with rafting theory is that there is no way (or at least no obvious way) to falsify it. So I am a bit suspicious and trying to play Devil’s advocate here :-)

    Link to this
  34. 34. josimo70 2:31 pm 10/6/2011

    Monkeys and caviomorphs reached South America in the end of Eocene, probably through same way (whatever it was), is it possible the emmergence of an archipelag in this epoch to make ocean-crossing easier?

    Link to this
  35. 35. Heteromeles 2:44 pm 10/6/2011

    @vdinets: Agreed. In fact, I wrote a SF novel (self-published) where I assumed that the dominant biome was composed of oceanic floating rafts, not that it’s explicit in the story, simply to goof around with the concept.

    Nonetheless, if there’s a system (the Amazon?) that semi-predictably coughs up vegetation rafts and sends them out to sea, it would be nice to attempt to study some of them, to see where they go and how they fall apart. If there are any regularities in such a process (production, dispersal, and demise), it would add quite a lot to models of dispersal. All we need is a grad student with a sailboat who likes following them around in the central Atlantic…

    Link to this
  36. 36. vdinets 6:25 pm 10/6/2011

    Heteromeles: a bunch of GPS transmitters and an agreement with someone running an appropriate satellite network (to get photos) would be enough. People track icebergs that way.

    josimo70: there are two undersea ridges (Walvis Ridge and Rio Grande Rise) that look like the remains of a landbridge between Namibia and S Brazil, now separated by the Mid-Atlantic Ridge. I don’t know if this idea has ever been discussed by geologists.

    Link to this
  37. 37. Heteromeles 9:07 pm 10/6/2011

    @32: The geological history of Maio, Cape Verde Islands
    C. J. Stillman, H. Furnes, M. J. LeBas, A. H. F. Robertson and J. Zielonka

    The oldest igneous rocks on Maio are pillow lavas of Mid-Ocean Ridge pillow basalts character which have been tilted and uplifted about 4 km from the ocean floor to outcrop as a partial ring, dipping steeply away from a central plutonic complex made up of pyroxenites, essexites, syenites and carbonatites. The ocean floor volcanic rocks are overlain conformably by a stratigraphically continuous pelagic carbonate succession which demonstrates a shallowing depositional environment from the Upper Jurassic to Upper Cretaceous times, when tuffaceous beds indicate renewed volcanism. The tuffs are associated with rudites demonstrating the emergence of the island and amongst the clasts are plutonics indicating Upper Cretaceous magmatism and the unroofing of the volcano to a substantial depth. Deformation under compressive stress resulted in the folding and local repetition by thrusting of this sedimentary cover, which, together with the plutonic core, had been intensively injected by major sills.

    The Mesozoic succession has been planed off and overlain with marked unconformity by a largely Neogene sequence of volcanic and terrestrial sedimentary rocks. There is a hiatus throughout the Palaeogene, and constructional activity appears to recommence with ankaramitic hyaloclastite and lava deltas and subaerial ankaramitic flows. These are overlain by fluvial sediments and tuffs.

    Stratigraphically above these is an extensive plateau of silica-undersaturated lavas, olivine-melilitites and nephelinites, which rest on a planed and locally lateritized surface. At topographically higher levels in the eastern part of the island there are thick ankaramitic lavas and pyroclasts which evidently flowed eastward through valleys cut down into the Mesozoic strata, and appear to be of Pliocene age.

    The subsequent history of the island appears to be non-volcanic.


    Link to this
  38. 38. ChasCPeterson 12:55 am 10/7/2011

    Somebody ought to mention one of the great biological puns here. An article on the unique digestive system of this species was titled ‘Alimentary, my dear hoatzin’.

    Link to this
  39. 39. Dartian 4:45 am 10/7/2011

    Regarding the survival of large terrestrial mammals on rafts at sea; here are a few such (admittedly anecdotal) cases that may or may not have some relevance to this general discussion:

    -After the 2011 Japanese tsunami, a dog survived for three weeks at sea amongst floating debris (including the remains of house) before it was rescued.

    -Human beings are also known to have survived for remarkably long times under similar circumstances. For example, after the 2004 Indonesian tsunami, at least three people were rescued alive from the sea after a period of several days: these were a pregnant woman who survived for five days by clinging on to an uprooted palm tree (later, after her rescue, she gave birth to a healthy baby), a man who clinged to a tree trunk for eight days, and another man who survived for two weeks by first clinging on to a log and, later, a makeshift raft. This man ate old coconuts for 12 days, and nothing at all for the last three days of his ordeal, before he was rescued.

    (-There are, of course, even more dramatic survival stories than these; for example, the Chinese sailor Poon Lim, who, after the ship he was serving on had been torpedoed (this was during the Second World War), survived for an astonishing 133 days on a raft before being rescued. For the purposes of the present discussion, however, it’s probably not entirely fair to include cases like Poon Lim, because on his raft he had some food supplies from the ship, and these helped him survive for the initial days and weeks.)

    -As for non-human primates being found rafting at sea, I’m aware of one such case. S.J. Hickson, in his 1889 book A Naturalist in North Celebes, says that in 1883, “some days” after the volcanic eruption of Krakatau, and the tsunami that followed it, a female monkey (species not certain; it was called a “green monkey” in the original source, but most likely it was a crab-eating macaque Macaca fascicularis) was found floating on drifting timber in the Sunda Straits. The monkey, which was picked up by a passing ship, was “terribly scorched” but alive and later made a complete recovery.

    Link to this
  40. 40. David Marjanović 5:41 am 10/7/2011

    yeah, sorry, victorg’s comments were quarantined for a few hours (while I was asleep) due to the links.

    What, two links trigger moderation? On ScienceBlogs, it’s four, and on FtB Pharyngula, PZ himself was able to set it to six. *whine*

    If there exist no strongly supported phylogenies of the extant species, with all the anatomy available (soft issue etc.), it may be difficult if not impossible to place a weird Eocene fossil in a phylogenetic context.

    On the other hand, it may not be possible to place the extant taxa without a taxon sample that includes enough Eocene fossils – if the branches of the extant taxa are long enough to cause long-branch attraction as long as there are no fossils to break them up.

    Livezey & Zusi (2007) strikes me as such a case: lots of extant taxa, lots of characters, almost no fossils, result disturbingly similar to Wetmore’s (1960) classification.

    Dartian: St. Helena is 15 mln years old, Tristan group islands 5 to 30.

    But does any particular island from that time survive? Or do they sink, and new ones rise to take their place?

    As for Cabo Verde Islands, Wiki claims 180 mln years – older than the Atlantic Ocean. Does anybody know what their real age is?

    The eastern Cabo Verde islands belong to the African continent. I suppose that’s what “180 Ma” is meant to mean. Only the western ones are volcanoes that stand on oceanic crust.

    Link to this
  41. 41. kilianh 10:35 am 10/7/2011

    Here’s a nice page with a hoatzin skeleton:

    Link to this
  42. 42. Zoovolunteer 2:40 pm 10/7/2011

    I can see the turaco/hoatzin connection being possible – although they are usually described as frugivores, turacos can eat a lot of foliage as well. That does raise a small question about their need to cross to South America close to the equator though – some turacos live at quite high altitudes and are surprisingly cold tolerant – a friend recently had an escaped one arrive on her roof south of Bristol. Further enquires showed it had been loose in the UK for over a year, and had survived quite heavy snow.

    Link to this
  43. 43. Heteromeles 5:39 pm 10/7/2011

    @40: David, check out #37. The eastern Cape Verde islands have rocks that are ~180 myo.

    Link to this
  44. 44. Christopher Taylor 6:38 pm 10/7/2011

    Livezey & Zusi (2007) strikes me as such a case: lots of extant taxa, lots of characters, almost no fossils, result disturbingly similar to Wetmore’s (1960) classification.

    On the other hand, according to Mayr’s (2008) critique of Livezey & Zusi (2007), a number of taxa are potentially miscoded: they seem to have been coded for some characters according to what they ‘should’ have as opposed to what they do have (e.g. all ‘Caprimulgiformes’ [excl. Apodiformes] coded as possessing a tapetum lucidum when in reality only some of them do).

    Link to this
  45. 45. vdinets 10:26 pm 10/7/2011

    David: I think all age estimates refer to extant islands, not to sunken ones.

    Link to this
  46. 46. llewelly 10:14 am 10/8/2011

    One peculiar constraint on “rafting” theories is the mechanism needs to be unlikely. There are perhaps a dozen, maybe two dozen Atlantic crossings by land tetrapods, spread across tens of millions of years. Strangely enough, if there were good records of trans-Atlantic “rafting” events in our meagre observational record of a few hundred years, scientists would have a very different problem; they would need to explain why there weren’t tens of thousands of cases of land tetrapods carrying out successful trans-Atlantic invasions.

    Link to this
  47. 47. Heteromeles 12:50 pm 10/8/2011

    @46: It’s not the crossing, it’s the establishment of a viable breeding population. We could have hundreds of thousands of solo crossings, but unless the passenger is a pregnant female, a pair, or an appropriate minimum breeding population, none of these crossings will result in establishment. Furthermore, many of the breeding populations established will die out on the other side. This has been documented even in the 20th Century (appearance and disappearance of bush tits on Catalina Island).

    The basic point is that we don’t know much about rafting. We know a bit more about how plants cross oceans, thanks to Sherwin Carlquist among others (going back to Darwin), but even in botany, there are many puzzles.

    Link to this
  48. 48. vdinets 2:19 pm 10/8/2011

    llewelly: birds from North America show up in Europe and Asia every year, and vice versa. Even some bats and butterflies cross the Atlantic. Still, there is a big difference in ornithofauna, and neither hoary bats nor monarch butterflies have established in Europe.

    Link to this
  49. 49. David Marjanović 2:54 pm 10/9/2011

    On the other hand, according to Mayr’s (2008) critique of Livezey & Zusi (2007), a number of taxa are potentially miscoded: they seem to have been coded for some characters according to what they ‘should’ have as opposed to what they do have

    Aaaaaaaargh. Another of my pet peeves. :-)

    Link to this
  50. 50. Dartian 3:34 am 10/10/2011

    There are perhaps a dozen, maybe two dozen Atlantic crossings by land tetrapods, spread across tens of millions of years.

    Only two crossings, in fact, if we restrict ourselves to land mammals. Molecular data (e.g., Poux et al., 2006) even hint at the possibility that the ancestors of the extant platyrrhine primates and hystricomorph rodents, respectively, might have made the crossing pretty much at the exact same time. The result of the same ‘paleo-natural disaster’, perhaps?

    Poux, C., Chevret, P., Huchon, D., de Jong, W.W. & Douzery, E.J.P. 2006. Arrival and diversification of caviomorph rodents and platyrrhine primates in South America. Systematic Biology 55, 228-244.

    Link to this
  51. 51. Dartian 3:40 am 10/10/2011

    the ancestors of the extant [...] hystricomorph rodents

    …and just to be clear, by that I meant of course the ancestors of the South American hystricomorphs (that is, the caviomorphs).

    Link to this
  52. 52. Ranjit Suresh 6:39 pm 10/10/2011

    I wonder what species if any (marsupials?) dominated frugivorous and folivorous niches in South America prior to the arrival of platyrrhine monkeys and hystricomorph rodents in the late Eocene/early Oligocene.

    Link to this
  53. 53. David Marjanović 9:38 pm 10/11/2011

    I wonder what species if any (marsupials?) dominated frugivorous and folivorous niches in South America prior to the arrival of platyrrhine monkeys and hystricomorph rodents in the late Eocene/early Oligocene.

    How about nothing? It had only been 30 million years since the end of the Cretaceous. Not all ecological niches get filled in this kind of time.
    How about the polydolopoid marsupials?

    Link to this
  54. 54. victorg 4:44 pm 10/13/2011

    Wouldn’t small sloths be the first suspects on the list for filling the niches Ranjit mentioned? (Or would the timing be wrong?)

    Link to this
  55. 55. John Scanlon FCD 2:10 pm 10/17/2011

    “the problem with rafting theory is that there is no way … to falsify it”

    That’s why there is Vicariance Theory (null hypothesis, frequently falsified) and dispersal, the trivial alternative hypothesis that explains most of the patterns that actually exist.

    The only patterns explained by vicariance are products of slow processes like continental drift, mountain-building, formation of deserts, course-shifts of major rivers etc. Transoceanic dispersal, in contrast, only takes a few days and it happens ALL THE TIME. As a process theory of biogeography, it’s far too powerful; what’s truly amazing is that endemism occurs at all.

    Link to this
  56. 56. vdinets 2:29 am 10/22/2011

    John: Wait a minute! So, dispersal theory predicts that transoceanic dispersal should be a common event. But it obviously isn’t. Doesn’t it falsify the theory?

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article