Tetrapod Zoology

ResearchBlogging.org

Reconstructions of the earliest phases of bird evolution – of ‘near-birds’ or ‘protobirds’, if you will – are pretty familiar, and hypothetical ‘proto-pterosaurs’ are also relatively familiar thanks to their appearance in the mainstream literature (Wellnhofer 1991, Unwin 2006). However, have you ever seen a reconstruction of a proto-bat? Such creatures must have existed, of course, but I’m only aware of one published effort to visualise them. Reproduced here, these reconstructions were originally published in Smith (1977) and were later reproduced in John Hill and James Smith’s excellent Bats: A Natural History (Hill & Smith 1984).

i-cf9079af98b1759e4d86f1a26cad5135-protobats-1-Smith-1977-Mar-2011-redo.jpg

i-bc23bc2ff9f8146ce520f1bc4bf699a4-protobats-2-Smith-1977-Mar-2011-redo.jpg

Smith obviously assumed that bats originated from insectivore-like mammals, and did note that insectivores – as in, shrews, moles and kin – and microbats* perhaps shared an ancestry. We still know all too little about the origins of bats: molecular phylogenies indicate membership of Laurasiatheria, but also show that bats are closer to the clade that includes pangolins, artiodactyls, perissodactyls and carnivorans than to ‘core insectivores’** (Van Den Bussche & Hoofer 2004, Nishahara et al. 2006)). Nishahara et al. (2006) even went as far as proposing the name Pegasoferae for the bat-perissodactyl-carnivoran clade they recovered.

* Hill & Smith (1984) included favourable discussion of the (poorly supported and now generally abandoned) hypothesis that megabats and microbats might descend from different ancestors. It’s now firmly established that bats are monophyletic, but also that ‘microbats’ of tradition are paraphyletic with respect to megabats. For more on this subject please see We flightless primates.

** Or eulipotyphlans.

Exactly what the ancestral pegasoferaen (pegasoferan? Pegasoferine?) looked like is unclear, but it was presumably a small, relatively long-tailed, clawed quadruped – I’m imagining something like a cross between a palaeanodont (viz, a stem-pangolin) and a dichobunid-like artiodactyl. The fact that the key, characteristic chiropteran innovation – those incredibly long, slender hands and arms – might have evolved thanks to changes in a single gene (BMP2) (Sears et al. 2006) indicates that the early stages of bat evolution happened extremely quickly (potentially in less than a few million years). This rapid evolution means that those early ‘protobat’ phases were short-lived. This of course lowers the probability that we’ll find them as fossils.

Smith also produced a series of diagrams showing the hypothetical development of the bat wing from an ancestral, insectivore-like hand. Again, this transition must have happened, but the details might have been slightly different from Smith’s (1977) portrayal: Onychonycteris finneyi from the Early Eocene of Wyoming, for example, has since demonstrated that claws on all five fingers were actually retained into the ‘true bat’ condition (Simmons et al. 2008).

i-37b5984c0f093511064aed1f3746a3d9-protobats-3-Smith-1977-Mar-2011-redo.jpg

Anyway, all very interesting I’m sure you’ll agree. It’s possible that there are more proto-bat reconstructions out there. Let me know if you’re aware of any. I did find this one (below), from a few years ago. It’s by some guy called Naish…

i-4b7fdb12107dc478691ac634b04455cd-Naish-protobat-Mar-2011.jpg

For previous Tet Zoo articles on bats, see…

Refs – -

Hill, J. E. & Smith, J. D. 1984. Bats: a Natural History. British Museum (Natural History), London.

Nishihara, H., Hasegawa, M., & Okada, N. (2006). Pegasoferae, an unexpected mammalian clade revealed by tracking ancient retroposon insertions Proceedings of the National Academy of Sciences, 103 (26), 9929-9934 DOI: 10.1073/pnas.0603797103

Sears, K. E., Behringer, R. R., Rasweiler IV, J. J. & Niswander, L. A. 2006. Development of bat flight: morphologic and molecular evolution of bat wing digits. Proceedings of the National Academy of Sciences 103, 6581-6586.

Simmons, N. B., Seymour, K. L., Habersetzer, J. & Gunnell, G. F. 2008. Primitive Early Eocene bat from Wyoming and the evolution of flight and echolocation. Nature 451, 818-821.

Smith, J. D. 1976. Comments on flight and the evolution of bats. In Hecht, M. K., Goody, P. C. & Hecht, D. M. (eds) Major Patterns of Vertebrate Evolution. NATO Advanced Study Institute Series, Series A, Life Science 14. Plenum Press, New York.

Unwin, D. M. 2006. The Pterosaurs From Deep Time. Pi Press, New York.

Van Den Bussche, R. & Hoofer, S. R. 2004. Phylogenetic relationships among recent chiropteran families and the importance of choosing appopriate out-group taxa. Journal of Mammalogy 85, 321-330.

Wellnhofer, P. 1991. The Illustrated Encyclopedia of Pterosaurs. Salamander Books Ltd., London.

Comments

  1. #1 derek
    March 5, 2011

    We don’t have much in the way of protobirds, and no living protopterosaurs at all, but surely the wonderful thing about flying mammals is that we have so many living prototypes? I wasn’t thrilled with the first illustration at all, because it just looks like another flying squirrel. The second illustration is more interesting, as showing a halfway stage between that and the modern bat.

  2. #2 Noni Mausa
    March 5, 2011

    Can I propose the split from Smith’s (A) to (D, E) would have to be very quick, and not studded with a lot of species displaying intermediate hand web configurations?

    The reason being, (A) is a perfectly useful grasping and climbing hand, while (D, E) are good wings, but (B, C) don’t look very good for either task. Too small for real flight, too clumsy and fragile for grasping.

  3. #3 Brian
    March 5, 2011

    Interesting post!

    I’m probably nitpicking here, but doesn’t Pegasoferae exclude artiodactyls? The imaginative name itself is a clue here: Pegasus for horses (and thus perissodactyls)as well as bats because of Pegasus being a flying horse, and Ferae for the Carnivora + Pholidota clade.

    Anyhow, is the presence of a bird of prey gliding alongside your proto-bat as if they were a team of superheroes of any particular signifance or is it there because of Rule of Cool?

    Smith’s proto-bat reminds me an awful lot of a treeshrew. Is that coincidental or did Smith still envision an Insectivora that included treeshrews? All the more ironic then, considering that treeshrews are of course euarchontoglirans, and thus phylogenetically closer to the creatures from which he believed megabats rather than microbats to descend.

  4. #4 neil
    March 5, 2011

    I propose you get that Dr Witton on the case! I hear hes sitting around doing nothing and in no way really busy ;)

  5. #5 Darren Naish
    March 5, 2011

    Brian: good call – Pegasoferae does indeed exclude artiodactyls (or ‘Cetartiodactyla’). And, yes, the eagle and proto-bat were part of the same team of superheroes (kidding, but let’s go with it).

    Neil: as I learnt the other day, the good Dr Witton is no bat expert :)

  6. #6 schenck
    March 5, 2011

    It looks like Dr. Naish’s reconstruction relied heavily on Smith’s, am I right?

  7. #7 Darren Naish
    March 5, 2011

    Yes. Imitation being flattery and all that :)

  8. #8 Nice Ogress
    March 5, 2011

    Obviously, the racing stripes make it go faster. :)

  9. #9 heteromeles
    March 5, 2011

    The real problem with this is that (from my understanding) it’s hard to go from a stiffly held gliding patagium to a flapping wing.

    Imagine a flying squirrel flapping away. It’s not good for gliding. Additionally, if you add the enlarged hands to the gliding membrane, It’s even harder to move that whole great membrane and try to flap it. That’s sort of like handing a child an adult sized paddle and asking them to move the canoe.

    I’d like to propose the “ski jump” hypothesis: flapping preceded patagium development. The name comes from the videos of how old ski-jumpers used to jump (e.g. http://www.youtube.com/watch?v=T4IM0NrRR2Y&feature=related). Notice how they wave their arms to try to extend their distance? It’s just like the stroke of a bat’s wing.

    I’d propose that bats evolved from leaping insectivores with well-developed, rotatable shoulders like those of modern primates. They rowed through the air in an attempt to gain distance (as human jumpers instinctively do), and had little, if any patagium. Galagos might be a good model. From this stage, most would-be gliders learned to hold their bodies stiff in an aerofoil, but the bat ancestors were stupid/unlucky. Instead of going the stiff glider route, they kept flailing their arms, and finally one of them got lucky with the mutations to BMP2 that grew their hands, and an emergent patagium followed. Since these mutations improved their jumping distance (Flail harder! Fly farther!), there was positive selection pressure for mutations that improved arm size and strength, and eventually bats flew.

    If you look at a bat today, they don’t use their patagiums like stiff aerofoils. Their feet move all over the place in response to the motions of the wings, something that no glider ever does. They’re still flailing away, but now they’re amazing fliers.

    I’d also point out that the Paleocene environment favored this evolution. Conifers like redwoods were quite common (often dominant), so we’re talking about a world of big, straight trees, somewhat like the dipterocarp forests of today. Additionally, many of the fruit sources that megabats currently depend on hadn’t evolved yet. That environment isn’t great for herbivorous arboreal mammals, but it’s okay for insectivores, particularly if they can travel long distances. Perhaps the Paleocene was a glider’s mammals’ paradise. Unfortunately, we don’t have the fossils to prove it.

  10. #10 Gerdien
    March 5, 2011

    The Pegasoferae are only supported by Nishihara et al (2006). The tree by Aranson et al (2008) is nearer consensus.

    U. Arnason, J.A. Adegoke, A. Gullberg, E.H. Harley, A. Janke, M. Kullberg, 2008. Mitogenomic relationships of placental mammals and molecular estimates of their divergences Gene 421:37–51

    Are any bats older than Onychonycteris finneyi known from skulls? How would one recognize a fossil bat if the wing was not preserved?

    I’m always surprised at the lack of interest in the bat shoulder. Bats as birds raise their wings above their backs, and this requires a special shoulder construction. (I could not get from the description of Onychonycteris finneyi whether it could flap or not, but that might as well be my lack of proper anatomy knowledge). The bird shoulder seems to have become very movable before or independent of flying.
    Bats, birds and pterosaurs have 3-part wings with elongated hands. Gliders have small hands, and do not flap – some gliders have very long arms. Would a moveable shoulder be a precondition for evolution of flapping flight with an elongated hand? Is a 3-part hand necessary for flapping flight?

  11. #11 heteromeles
    March 5, 2011

    Hi Gerdien,

    Given that insects fly without hands, I’d suggest that development of a movable shoulder joint (or the equivalent in insects) is a necessary precondition for powered flight. Off-hand, I can’t think of a single glider (not including flightless birds) that has a highly mobile shoulder.

    There might even be some selection against a mobile shoulder in gliders. A relatively stiff shoulder joint could help hold a patagium rigid with minimal muscular effort, while an animal with a mobile shoulder joint would have to expend muscular effort to gain the same effect. Gliding is primarily about efficiency and weight savings.

  12. #12 Raymond
    March 5, 2011

    Here is a linky

    http://onlinelibrary.wiley.com/doi/10.1002/bies.200900053/full

    not bad. Bats of course are Laurasiotherians, no question. WHERE they nest is still in the air excuse the pun. I hope Pegasoferae does turn out to be the correct interpretation.
    It would be a mind spinner if perissiodactyls, carnivorans and bats are sister clades to the exclusion of lypotyphalans and cetartiodactyls.

  13. #13 Anonymous
    March 5, 2011

    Kenneth Oppel once wrote a book called Darkwing, which was set at the very beginning of the Cenozoic era (as in, mere years after the K-T boundary). Most of the main characters of this book belonged to a species of protobat called “Chiropters”, which looked rather like your protobat drawing or the third protobath in the sequence that Smith did. Two of the major characters were abberant individuals that looked more like Onychonycteris than a protobat, and could actually achieve flight.

    On an interesting note, almost all of the non-chiropter characters in the book with speaking roles were carnivorous mammals (three miacids and a Hyaenodon…yes, you heard me right), and with one exception were all evil.

  14. #14 Gerdien
    March 5, 2011

    If gliding is energetically easier with a stiffish shoulder joint, what are the bat data on fossil shoulder joints?

  15. #15 Morgan Churchill
    March 5, 2011

    When I was a kid, I got zoobooks (and the prehistoric zoobooks), and if I recall a protobat was illustrated in one of the issues (I think the one on bats, but I might be mistaken)

    Actually there was a lot of pretty cool wildlife art in Zoobooks

  16. #16 Rajita
    March 5, 2011

    The idea that changes in the expression of a single gene BMP2 resulted in the bat-like state of hands is a seductive albeit rather simplified or even erroneous one. By just elongating the arms you cannot achieve flight. You need the patagium to go hand in hand :-). In fact the same group has shown in other studies that the expression of Fgf8 is critical to prevent cell death in the interdigital tissue to foster patagium formation. Another protein Gremlin is also over expressed just as in the case of the ducks which paradoxically block the effects of the BMPs to prevent cell-death (BMPs accelerate interdigital cell death). Thus it appears that more than one gene expression change accompanied the emergence of the bat wing. Now it conceivable that the interdigital webbing emerged prior to a major elongation phase.

    I think a more recent paper favors this model where in the establishment of Fgf8 expression initiates a wave of signaling via a second signaling molecule Sonic hedgehog that in turn activates BMP2 signaling and finally that activate Gremlin. It appears that here the gene expression is recapitulating the evolutionary process: First selection for an incipient patagium for gliding, then activation of BMP2 by the same process fostered digit elongation and finally this established a larger patagium favoring flight.

    cheers
    Rajita

  17. #17 Riggi
    March 5, 2011

    If I can recall, I remember there was one paper that for the life of me can’t remember. It suggested that, instead of bats being a kin to perissodactyls, carnivorans and pangolins (and if you want to be complicated, creodonts, dinocertans and a paraphyletic Cimolesta). Instead more closer to the ancestry of the artiodactyls/cetartiodactyls based on a single gene (if you think about it, it kind’a makes sense the arctocyonids were aboreal animals that appear around the same time as those “plaeobats”). Is it just me, or does anyone here favor a (Arctocyonidae + Chiroptera + (Mesonychia + Artiodactyla/Cetartiodactyla)), “diversungulates” anyone?

    P.S If anyone could be so kind, can anyone find the name for the document? Thanks.

  18. #18 Simon G
    March 5, 2011

    heteromeles: the ski-jumpers’ arm movements appear more to do with increasing stability than extending distance per se. The direction in which they rotated their arms (forward at the bottom and back over the shoulder) suggests that they were perhaps counteracting a tendency to tip onto their backs, caused by air resistance and an upright stance. Later jumpers achieved stability by leaning forward and despite lack of arm movements managed much longer jumps.

  19. #19 heteromeles
    March 5, 2011

    Hi Simon,

    Yes, I know about that, although I also understand that, as with long jumping, the old-time ski jumpers thought it would extend their range. The bigger point is that, at least with humans, there’s an instinctive tendency to windmill when you perform long jumps. Oddly enough, that motion pretty closely resembles the motion of a bat’s wing. As I found when I explained what I was writing to my wife, you can do a pretty good protobat imitation by windmilling your arms and going “eek, eek, eek” with each rotation. Pretend you’re trying not to fall off a stage.

    The other point is that adding muscle and shoulder flexibility to a glider to turn it into a powered flier is maladaptive. The glider becomes heavier and less stable. It also defeats the proposed purpose of gliding, which is to travel longer distances more efficiently than the animal could climb to the same spot.

    What we need here is a scenario where partially powered flight is superior to whatever preceded it. I don’t think that’s the case with going from glider to flapping glider. Perhaps, instead of the ski-jump hypothesis, we should call this the arm-flail hypothesis.

  20. #20 CS Shelton
    March 5, 2011

    I heard about pegasoferae a few years ago, but a practical upshot of that never fully registered in my brain – we’re not related to bats. Guess science made a liar of me again; I don’t know how many times I’ve told people about that bit of now falsified trivia. *sigh* Amusingly, a cited bit of morphology for the relation of primates and bats was the dangling penis. Bats have creepy primate-like junk flappin’ in the breeze, for flying dog-horse-pangolin kin.

    One question – The only primitive bat I’ve seen represented (name eludes me) lacked patagia between the legs and tail, so I assumed those patagia were developed more recently in bat evolution. Not so, according to these images. And where do New Zealand’s short-tailed bats fit into all of that?

  21. #21 J. S. Lopes
    March 5, 2011

    Could the retention of membrane between fingers imply that proto-bats use their extended, webbed fingers to catch insects in the air?
    Could the presence of proto-bats in Earliest Eocene in Southern Hemisphere sugest a Neogondwanan origin for bats? They were found in Tingamarra (Australia) and Laguna Fria (South Argentina), and maybe in Antarctica. Or did they cross Paleogene Atlantic from Africa?

  22. #22 heteromeles
    March 5, 2011

    Thanks Darren, now I’m really thinking about how an antebat and a protobat might move.

    Most of the attention focuses on the evolution of bat arms, and the rest of it is effectively negative space that we’re projecting our prejudices on.

    For example, if the antebat was some sort of generalized insectivore, much of its movement was with its spine, not its limbs. This is in total contrast with bats, which move almost entirely with their arms and little else.

    Furthermore, we’re assuming that the hands evolved, and somehow dragged the rest of the bat body along to accommodate them. This is weird, because it ignores the power plant (the shoulders), and bats’ highly modified pelvis.

    Does any of this make sense? Or should we start thinking a bit more outside the box?

    For example, were antebats and protobats four-legged hoppers, like what was proposed for the multituberculate Kryptobataar (Kielan-Jaworowska et al.)? Perhaps the protobat flew a bit like Rhacophorus (the modern gliding frog).

    Hill and Smith’s reconstruction assumes that hand evolution drove all the torso adaptations. I’d argue that it’s equally possible that antebats were already weird little beasties with the mobile, shoulder and hip joints that allowed powered flight to evolve. For all we know, they even advertised their territories with ultrasonic screams.

  23. #23 Allen Hazen
    March 6, 2011

    Thanks, Darren: if you’d asked me for a list of topics I would particularly like to see you blog about, the origin of bats would have been near the top of the list!

    I’m with a couple of the other commenters on one thing: the first thing I thought when looking at the drawing of a pre-protobat was TREE SHREW, which (whether or not the delightfully named Pegasoferae hypothesis holds up) seems phylogenetically inappropriate. But then, tree shrews have often been thought of the best extant analogues of basal unspecialized Eutherians. Thought experiment: somebody shows you a fully fleshed and furred Eomaia and doesn’t tell you it’s from the Cretaceous. Suppose you try to assign it to some extant Eutherian order: which one? Most Eulipotyphlans are pretty specialized: Eomaia doesn’t really look like a shrew. When it opens its mouth you can see it doesn’t have a rodent’s front teeth, so you know it isn’t a mouse. You’re being pushed for a guess. Anything more plausible than tree shrew come to mind?

    So maybe it isn’t too bad to think of the ancestors of bats LOOKING like tree shrews, even if we know they WEREN’T tree shrews.

    —-”Exactly what the ancestral pegasoferaen (pegasoferan? Pegasoferine?) looked like is unclear, but it was presumably a small, relatively long-tailed, clawed quadruped”
    —Certainly.
    —-”I’m imagining something like a cross between a palaeanodont (viz, a stem-pangolin) and a dichobunid-like artiodactyl” (or maybe Hyracotherium instead of the dichobunid)
    —–No, both of these are WAY too derived. What did the common ancestor of Perissodactyla and Carnivora (to leave out the WEIRD Pegasoferans) look like? Something pretty small and generalized.

  24. #24 Andreas Johansson
    March 6, 2011

    Darren wrote:

    ancestral pegasoferaen (pegasoferan? Pegasoferine?)

    “Pegasoferan” makes the most sense. The base word here is fera “wild animal”.

  25. #25 David Marjanović
    March 6, 2011

    Yay! :-)

    The Pegasoferae are only supported by Nishihara et al (2006). The tree by Aranson et al (2008) is nearer consensus.

    U. Arnason, J.A. Adegoke, A. Gullberg, E.H. Harley, A. Janke, M. Kullberg, 2008. Mitogenomic relationships of placental mammals [...]

    Firstly, Nishihara et al. (2006) used LINE insertion sites as data. Arnason et al. (2008) used mitochondrial genome sequences as data. LINE insertion sites win every day of the week and twice on Mondays, Tuesdays, Wednesdays, Thursdays, Fridays, and Sundays.

    Secondly, Arnason is the first author of the 2004 paper that found monophyletic fish. That’s right, a clade composed of Chondrichthyes, Actinopterygii, Dipnoi and Latimeria but not Tetrapoda. Just saying he has been wrong before.

    Are any bats older than Onychonycteris finneyi known from skulls?

    Nope. A few isolated teeth from about the same age might belong to bats, but obviously don’t tell us anything about flight; that’s it.

    How would one recognize a fossil bat if the wing was not preserved?

    From lots of other features throughout the skeleton.

    If gliding is energetically easier with a stiffish shoulder joint, what are the bat data on fossil shoulder joints?

    None. Onychonycteris already was a flapping flier.

    Could the presence of proto-bats in Earliest Eocene in Southern Hemisphere sugest a Neogondwanan origin for bats? They were found in Tingamarra (Australia) and Laguna Fria (South Argentina), and maybe in Antarctica. Or did they cross Paleogene Atlantic from Africa?

    Or they crossed the early Eocene Caribbean from North America. That’s where Onychonycteris lived, and AFAIK all contemporaneous laurasiatheres come from the northern continents as well.

    Hill and Smith’s reconstruction assumes that hand evolution drove all the torso adaptations. I’d argue that it’s equally possible that antebats were already weird little beasties with the mobile, shoulder and hip joints that allowed powered flight to evolve.

    Indeed, I wonder if bats went straight from grasping at branches to grasping at prey while jumping from branch to branch, without a gliding stage.

    The joint between wing and body in bats isn’t the joint between upper arm and shoulder, as it is in birds. It’s the joint between clavicle and sternum. That’s the joint I use when I try to imitate a flight stroke. Maybe bats had primate-like (monkey-like… ape-like?) shoulder mobility before they started to fly.

    For all we know, they even advertised their territories with ultrasonic screams.

    No — Onychonycteris lacked the modern bat ear apparatus.

    Thought experiment: somebody shows you a fully fleshed and furred Eomaia and doesn’t tell you it’s from the Cretaceous. Suppose you try to assign it to some extant Eutherian order: which one? [...] When it opens its mouth you can see it doesn’t have a rodent’s front teeth

    When it opens its mouth, it leads the thought experiment ad absurdum, because then you can see it has five premolars per jaw quarter. That only occurs in some afrotherians anymore, and, well, Eomaia just isn’t a hyrax.

    “Pegasoferan” makes the most sense. The base word here is fera “wild animal”.

    And its plural is ferae.

  26. #26 David Marjanović
    March 6, 2011

    Forgot to mention how obviously the ginkgo in the first picture is based on several Archaeopteryx restorations (well, one that was copied again and again).

    Also forgot to thank Raymond for this link! I’m reading the paper right now. :-)

  27. #27 David Marjanović
    March 6, 2011

    Yeah… Arnason et al. (2008) found Dermoptera and Anthropoidea as sister-groups; that’s nowhere near consensus and contradicts a lot of molecular and morphological evidence.

  28. #28 J. S. Lopes
    March 6, 2011

    | Or they crossed the early Eocene Caribbean from North America. That’s where Onychonycteris lived, and AFAIK all contemporaneous laurasiatheres come from the northern continents as well.

    Is there any phylogenetic study of Australonycteris? Is it somehow closer to Wasatchian proto-bats? Until now there’s no known Paleocene American bat, what seems to point that Wasatchian bats came from another region that Western North America.

  29. #29 Darren Naish
    March 7, 2011

    Thanks, everyone, for great discussion. In view of some of the comments here, I’ve been thinking a lot about bat shoulders. The entire pectoral region of bats is pretty complicated and even the stem-bats we know of have mostly modern anatomy – where they seem ‘primitive’ (e.g., in narrow muscular fossae on scapula, absence of suprascapular process, clavicle articulating with acromion process) they merely possess the conditions widespread in crown-bats (yinpterochiropterans especially). A dorsal articular facet on the scapula (part of a sort of locking mechanism in the shoulder) is present in some stem-bats, though apparently not in Onychonycteris and Icaronycteris; anterior ‘laminae’ (flange-like structures for increased muscle attachment) are present on the ribs of Palaeochiropteryx at least. The point is: there is indeed no evidence from early bat skeletons that they were doing anything other than active flapping of the sort practised by crown-bats. For an extension discussion of character distribution in stem-bats, see…

    Simmons, N. B. & Geisler, J. H. 1998. Phylogenetic relationships of Icaronycteris, Archaeonycteris, Hassianycteris, and Palaeochiropteryx to extant bat lineages, with comments on the evolution of echolocation and foraging strategies in Microchiroptera. Bulletin of the American Museum of Natural History 235, 1-182.

    As for early biogeography (comment 28), see Teeling et al (2005). They conclude:-

    “Geographic ancestral reconstructions suggest that bats originated in the Laurasian land masses, possibly in North America during the early Paleocene, and fail to support a Gondwanan origin for bats, even with the inclusion of Australonycteris in the analyses”.

    The discovery of Onychonycteris (Paleocene of North America!) is in obvious agreement with their suggestion.

    Ref – -

    Teeling, E. C., Springer, M. S., Madsen, O., Bates, P., O’Brien, S. J. & Murphy, W. J. 2005. A molecular phylogeny for bats illuminates biogeography and the fossil record. Science 307, 580-584.

  30. #30 Gerdien
    March 7, 2011

    If bats use the joint between clavicle and sternum for flight stroke, does that rule out a glider stage? How do flying squirrels and colugos spread their arms?

    Nishihara et al 2006 have incongruence in their tree from L1 insertions, Chiroptera + (peris +carn) 4 versus Chiroptera + (peri+carn, cetart)1.
    This suggests LINEs are not the last word, and it all has to be done again. The reference Raymond gave lists alternative trees.

  31. #31 johannes
    March 7, 2011

    (three miacids and a Hyaenodon…yes, you heard me right), and with one exception were all evil.

    Ironically, the neutral-tending to good miacid was named *Patriofelis*

  32. #32 David Marjanović
    March 7, 2011

    If bats use the joint between clavicle and sternum for flight stroke, does that rule out a glider stage?

    It means there weren’t any adaptations to stiffen the shoulder. Had there been any, mobility of the shoulder relative to the ribcage wouldn’t have reevolved, and bats would be more like birds or at least pterosaurs (the shoulder girdles of pterosaurs were mobile forwards and backwards only).

    Nishihara et al 2006 have incongruence in their tree from L1 insertions, Chiroptera + (peris +carn) 4 versus Chiroptera + (peri+carn, cetart)1.
    This suggests LINEs are not the last word

    It rather suggests what the paper says: incomplete lineage sorting. It means Scrotifera must have diversified extremely quickly. (The filling of a world that was emptied by the K-Pg boundary mass extinction is an obvious scenario.)

    (It’s also significant that none of the LINE insertions they found supports “Euungulata” — Perisso- and Artiodactyla as sister-groups.)

  33. #33 Dartian
    March 7, 2011

    Darren:

    It’s now firmly established [...] that ‘microbats’ of tradition are paraphyletic with respect to megabats.

    About that ‘firmly established’ part… The most recent study on bat phylogeny that I know of (Agnarsson et al., 2011) reports conflicting results; some of their analyses find support for the traditional megabat-microbat split (and others for the ‘Yinpterochiroptera’-'Yangochiroptera’ split). It might still be a bit early to declare consensus on this issue.

    CS Shelton:

    where do New Zealand’s short-tailed bats fit into all of that?

    Have you already clicked on the above-listed link to Darren’s old article ‘The most terrestrial of bats’?

    Reference:

    Agnarsson, I., Zambrana-Torrelio, C.M., Flores-Saldana, N.P. May-Collado, L.J. 2011. A time-calibrated species-level phylogeny of bats (Chiroptera, Mammalia). PLoS Currents 3, doi: 10.1371/currents.RRN1212.

  34. #34 Darren Naish
    March 7, 2011

    Thanks, Dartian, for the mention of Agnarsson et al. (2011) – I was unaware of it until now (hey, it was only published on February 2nd!). Very interesting, but some of their trees strongly contradict many relationships widely supported across other, independent studies, plausibly because their study does not incorporate data from enough sources: it’s all based on a single gene. Indeed, the authors do note a few times “we stress that a single-gene analysis of such a large and old group cannot be interpreted as more than a crude estimate of the bat species tree”. As a crude estimate, one can agree that it’s congruent with other phylogenies, but as a detailed estimate… well, it’s a bit of an oddity.

  35. #35 Bryan Ibeas
    March 7, 2011

    Great post! I’m pretty fond of the “wings as modified raptorial arms” theory myself – at least as it applies to protobats. AFAIK it was proposed for the evolution of avian flight (does that still hold water? or is everyone more in favour of incline flapping?), but the image of little protobats hanging upside down from tree branches or at cave mouths snatching at passing insects with their big webbed Gollum-fingers is too irresistible.

  36. #36 Gerdien
    March 7, 2011

    No molecular study of any size supports Euungulata; only one of the two first 2001 studies. Lineage sorting seems a weak proposal: at least, how fast a split and what size population would that imply?

  37. #37 heteromeles
    March 7, 2011

    @35 Brian: Gee, and here I was thinking of the “microgorilla” as the antebat. More seriously, maybe we should be playing with different environments for bat ancestors.

    One I’d propose is branch tips: the protobats of Hill and Smith seem to be tree trunk animals, as are many flying squirrels today. Around the K-T split and the early Palaeocene, there weren’t any true primates, in the sense that there weren’t any species with well-developed grasping hands that could readily exploit the thinnest branches.

    Perhaps the protobats were out there, clambering around on the thinner branches, and launching themselves between trees with near-ultrasonic screeches of “Cowabunga!” I don’t know if the scenario works for development of wide-spread grabbing arms turning into wings, but it’s not impossible.

    That said, the protobat-as-Gollum image is quite striking. If that one turns out to be true, the protobat should be named Gollumraptor.

  38. #38 Bryan Ibeas
    March 7, 2011

    @heteromeles

    I completely forgot about echolocation, but hey, that works for Gollumraptor! I imagine that some sort of prey-detection system would be advantageous for the mammalian equivalent of an upside-down nocturnal hangingfly.

    I doubt any of this is supportable by anything but hey, speculation is fun.

  39. #39 John Harshman
    March 7, 2011

    What about using as a model some of the frogs that use their webbed feet for gliding? Not all gliders do in fact have tiny hands. Hand-web first, patagium last?

  40. #40 Bill Davis
    March 8, 2011

    Nice try guys. But it is all but obvious that bat wings could not have evolved in a step by step fashion. What’s more, there are countless organs, instincts, morphological changes,–not to mention cells themselves–which one can hardly even fantacise, even while on drugs, developing in an incremental fashion.

    I know what you are thinking, but I’m right.

  41. #41 CS Shelton
    March 8, 2011

    Harshman- I like that idea. Might help show why they have more finger bones in the wing than other tetrapod fliers- By the time they developed patagia, the other finger bones were too important to the wing structure to lose.

  42. #42 raymond
    March 8, 2011

    Is Bill Davis the same as Cal King? Hmmm I agree with Christopher Taylor, it’s ALWAYS fun to see DM tear these idiots a new one.

  43. #43 Owlmirror
    March 8, 2011

    The comments about bat shoulder mobility did get me wondering — would it make sense to propose a semi-aquatic origin for protobats, using webbed hands to rapidly swim?

    I mean, as long as we’re speculating and all.

  44. #44 heteromeles
    March 8, 2011

    @Owlmirror,

    Well, aside from the shoulders, what other evidence do we see of aquatic origins of bats?…

    …right. It’s actually an interesting idea, since we do have things like sea lions which (I think) have developed some of the same shoulder mobility and hip restructuring as that found in bats.

    That said, there’s no other structural evidence in bats for an aquatic origin, and primates, for one, have similarly flexible shoulders and are arboreal. The simplest explanation is that the antebats were arboreal. What they looked like, and what the trees they lived in looked like, well that’s a different question entirely.

  45. #45 Owlmirror
    March 8, 2011

    But it is all but obvious that bat wings could not have evolved in a step by step fashion.

    Quite right, quite right. An Eocene wizard did it . . . !!

    Hm. An avian Eocene wizard.

    In fact, an avian Eocene wizard . . . trying to win a bet. And succeeding beyond his/her wildest dreams.

    “Look!!! I’ve made mammals — that fly! They’ll be great at parties!!”

    Yeah, that’s the ticket.

    What’s more, there are countless organs, instincts, morphological changes,

    Moar wizards . . . !!

    not to mention cells themselves

    Crap. Where are we going to find an acellular wizard? No, wait . . .

    A mineral wizard did it . . . !!

    which one can hardly even fantacise, even while on drugs, developing in an incremental fashion.

    Because fantasizing about wizards is so much moar phun.

    Wingardium Leviosa . . . !!

    I know what you are thinking,

    Because you’re telepathic!

    but I’m right.

    And you’re a wizard, too!

  46. #46 Zach Miller
    March 8, 2011

    Is it possible that antebats went through some kind of brachiating stage which would account for the primate-like shoulder mobility? Gettin’ on the spec-train…

  47. #47 Allen Hazen
    March 8, 2011

    David M:
    Re:25. I was thinking of a scenario in which Eomaia gives you only a quick grin: long enough for you to see it doesn’t have mouse-incisors, not long enough for you to count premolars (or upper incisors, for that matter). (Grin!)
    But thank you for TRYING to keep me honest in my thought experiments!

  48. #48 Andrew Wright
    March 8, 2011

    Hey, completely off-topic (and also completely missed by Cryptomundo), but the giant turtle of Hoan Kiem lake in Hanoi, has surfaced again and is bobbing about in the water, looking rather the worse for wear:

    http://news.yahoo.com/s/ap/20110304/ap_on_re_as/as_vietnam_sacred_turtle

    Quick question: Is Rafeteus swinhoei a valid taxon for this incredible beast?

    cheers

    Andy

  49. #49 James P.
    March 9, 2011

    So does anybody have any insight on the status of the “flying primate” hypothesis? I’m pretty curious on the characters the author used for that one.

  50. #50 Dartian
    March 9, 2011

    Bill Davis is probably a Poe. Real creationists don’t use mealy-mouthed expressions such as ‘it is all but obvious’; they would say ‘it is obvious’. The comment also contains too few grammatical and spelling errors – the one misspelling there is (‘fantacise’) seems out of place and is presumably intentional – and there are no exclamation marks. (Note to wannabe trolls: All wackaloony comments should include incorrect usage of their/there/they’re, it’s/its, or you’re/your.)

    James:

    So does anybody have any insight on the status of the “flying primate” hypothesis? I’m pretty curious on the characters the author used for that one.

    Click on the link to Darren’s earlier Tet Zoo article ‘We flightless primates’. There you’ll also find listed most of the relevant literature.

  51. #51 Owlmirror
    March 9, 2011

    Maybe yes and maybe Poe.

    Still, it inspired me to imagine a phorusrhacid with a pointy hat and a wand held in a hand-claw.

  52. #52 Vladimir Dinets
    March 9, 2011

    Andy” it’s Rafetus. The turtle has never been identified, but I’m not aware of any other Asian softshell reaching this size. If confirmed, it would be one of the last 3-4 turtles of this species known to exist. I hope it will still be around when I go to Hanoi in June.

  53. #53 Andrew Wright
    March 9, 2011

    Thanks Vladimir, my bad! (and sorry to drag this off-topic again!) I’ve got an article from 2008 that says the Cleveland Metroparks Zoo id’d it as R. swinhoei, but it seems way bigger and stranger than those (based on what I’ve seen of the video of two of those turtles in that Chinese zoo).

    Anyway, it keeps slipping the net, so chances it’ll hopefully still be there in June. If not, there’s a stuffed one in the Ngoc Son temple you can check out. Some nice pubs by the lake to chill out it – I’ve just done a short blog on the whole thing here:

    http://fantasygamebook.blogspot.com/2011/03/giant-turtle-from-lake-of-restored.html

    Apologies again!

    cheers

    Andy

  54. #54 Owlmirror
    March 9, 2011

    Well, if we’re going to speculate, let us speculate… with boldness.

    Consider the forests of the mid-to-late Palaeocene. The climate is hot and humid. Rainfall is heavy. Vegetation is thick and lush and fast-growing. Hence, the tips of tree-branches are long, soft, slick, and very springy.

    Along the branch of one tree, a small mammal comes running. Its paws are well-suited for grasping said tree limbs, and its forepaws are wide and webbed. It comes to the tips of one branch, and wishes to reach the branches of a neighbouring tree. Leaping directly is made difficult by the slickness and springyness of the tree branch.

    Some mammals might brachiate at this point, but not this one. Rather, instead, it grasps the branch that it is on with its hind paws, and hangs downward. It then begins swinging back and forth, using its outstretched forepaws to help “pump”, in addition to curving and arching its body and tail. Back and forth it swings, and then, at just the high point closest to the target branch, it lets go with its rear paws, flying through the air, reaching out with its wide webbed forepaws, and grasps the slick springy target branch tightly, and clambers on, hauling the rest of its body onto the branch, and then unconcernedly scampers along it.

    (Perhaps, if feeling particularly energetic and flashy, it may do a flip or a twist or other unnecessary manoeuvre).

    Some might call this mammal a protobat. Others might prefer antebat.

    But I say, let this mammal be called… an acrobat.

  55. #55 Owlmirror
    March 9, 2011

    (It occurred to me that the semi-aquatic hypothesis might be called the aqua-bat conjecture. But perhaps that is going just a little bit Too Far.)

  56. #56 Zach Miller
    March 9, 2011

    Brilliant. I’m all for this idea, in part because Owlmirror has–knowingly or not–referenced a horrible SNES platformer called Aero the Acrobat (developed by Sunsoft) in which a circus-performing bat runs around collecting things and using death-defying methods to do so.

  57. #57 valagos
    March 9, 2011

    On virtually all these speculative scenarios one still gets a somewhat vulnerable transitional creature. I wonder if there was any island endemism phase in early bat evolution that allowed it to evolve relatively free of predation and competitive pressures during that critical stage of vulnerability. Of course, with the island endemism scenario, the chances of ever finding fossils of the transitional phases dwindle even further as the hypothetical location is probably long lost back into the ocean waters…

  58. #58 Bryan Ibeas
    March 9, 2011

    @56: That’s actually something I wonder about, not just in this case but with other evolutionary scenarios. Is it possible that a few of the truly weird or extraordinary adaptations persist because they’ve been allowed to “incubate” in geographically isolated areas – so that when the particular organism finds itself in a more volatile ecosystem, it can thrive, whereas a less derived transitional form would not?

  59. #59 heteromeles
    March 9, 2011

    Effectively, much of the Paleocene was an “island.” Most of the big herbivores and carnivores disappeared, the carbon cycle was (apparently) far more chaotic than it is now–it wasn’t just warm and humid, there were some biggish droughts in there–and the biggest predators were a variety of crocs and terrestrial terror birds.

    You don’t particularly need to invoke an island. There’s plenty of room on such a planet for all sorts of weird adaptive radiations, and the local isolation caused by a wildly fluctuating climate would presumably also favor speciation.

    Additionally, as I pointed out above, there were apparently some big trees (Sequoia relatives) around, so we can expect forests with lots of big, semi-isolated trees, especially in the paratropics (you know, Scotland, north Antarctica). This is a perfect environment to favor extreme gap crossing.

  60. #60 Christopher Taylor
    March 10, 2011

    On virtually all these speculative scenarios one still gets a somewhat vulnerable transitional creature. I wonder if there was any island endemism phase in early bat evolution that allowed it to evolve relatively free of predation and competitive pressures during that critical stage of vulnerability.

    I don’t see that that’s immediately necessary. After all, the world is full of large numbers of seemingly ungainly, vulnerable animals that still manage to get by just fine. Toucans, sword-billed hummingbirds, bombyliid flies (at least, the ones that seem to go through life staring at their own bums) might all seem too ungainly to survive in competitive environments. Squirrel monkeys and sloths might seem to have too little in the way of defenses to survive. But they all do, because being fast and lean with great big teeth and spines is not the only way to avoid predation.

    Also, consider what sort of predators there were in the Palaeocene (as ‘Heteromeles’ has already alluded to). My impression is that mammals, at least, were generally slower than they are today (at least partially because wide open spaces would have been rarer, if I understand correctly). So proto-bats wouldn’t have been worrying about cats and mustelids, they’d be worrying about arctocyonids and the like. It’s possible that agility rather than simple speed would have been more important in predator avoidance; Owlmirror’s ‘acrobat’ may not have been able to outrun a cat, but it may have been able to move in directions not open to an arctocyonid.

  61. #61 Dartian
    March 10, 2011

    Owlmirror:

    protobat [...] antebat [...] acrobat [...] aqua-bat

    Hmm. As both ‘wombat’ and ‘numbat’ are already taken, what would you call a hypothetical terrestrial chiropteran?

    Valagos:

    On virtually all these speculative scenarios one still gets a somewhat vulnerable transitional creature.

    Be careful with making such assumptions; as Christopher says, many seemingly vulnerable organisms are, in fact, perfectly capable of making a successful living.

    Christopher:

    proto-bats wouldn’t have been worrying about cats and mustelids, they’d be worrying about arctocyonids and the like

    Snakes would surely have been among the most serious predators of small-to-medium-sized terrestrial and arboreal mammals already in the early Paleogene (there were various boids around by then).

  62. #62 Christopher Taylor
    March 10, 2011

    Snakes would surely have been among the most serious predators of small-to-medium-sized terrestrial and arboreal mammals already in the early Paleogene (there were various boids around by then).

    Very good point. I had been wondering whether I needed to qualify my speculations about inefficient mammalian predators to account for avian predators, not being aware whether there were any significant raptorial birds in the Palaeocene, but I confess that I had completely overlooked the squamate factor.

    Which does inspire my stupid question for the day (though I’ve been told that there are no stupid questions, only stupid people): how exactly does the booid method of capturing prey work in an arboreal habitat? Do arboreal non-venomous snakes just grab and swallow, or do they still constrict their prey? And if the latter, then they are presumably restricted to branches robust enough that they’re not going to fall off them while subduing prey? Or is the whole constriction trope only characteristic of a small number of non-venomous snakes and not at all representative of normal snake behaviour?

  63. #63 David Marjanović
    March 10, 2011

    Lineage sorting seems a weak proposal: at least, how fast a split and what size population would that imply?

    No idea in absolute numbers. It implies a large population and a fast split…

    I don’t think that’s at all unrealistic, however. I actually predict we’ll see more of this, because the world was so empty in the Paleocene.

    What about using as a model some of the frogs that use their webbed feet for gliding?

    Not gliding — parachuting.

    Colugos (“flying” “lemurs”) have webbed hands, however.

    Is Bill Davis the same as Cal King?

    Nope. Cal King is a BANDit; Bill Davis is a creationist if we take him literally (as mentioned above, we probably shouldn’t — Poe’s Law knows no exceptions).

    And if the latter, then they are presumably restricted to branches robust enough that they’re not going to fall off them while subduing prey?

    I suppose they simply restrict themselves to prey small enough that they don’t need their entire length to constrict it and can still hold on to a branch.

    Most snake prey does need to be killed before swallowing to avoid injury to the snake.

  64. #64 Adam
    March 10, 2011

    To answer Chris’s question I’ve seen quite a few photographs and videos of Antaresia maculosa (Spotted Python)eating bats while perched on a wall or roof of a bat cave. Not exactly arboreal but close enough to work the same. Anyway the snake is anchored to its spot by the posterior part (about 2/3) of its body while the front part does the constricting. This seems to be done dangling upside down from its perch. Once dead the snake needs a good grip on the prey wth its mouth before the constricting coils are released. It then seems to be able to swallow the prey against gravity.

  65. #65 David Marjanović
    March 10, 2011

    I suppose ScienceBorg isn’t quite back up yet… because… this is currently the 5th most active post! :-)

  66. #66 David Marjanović
    March 10, 2011

    Annnnnnnd this is now the 4th most active ScienceBlogs post!

  67. #67 David Marjanović
    March 10, 2011

    And comment 65 probably turned it into the 3rd most active one, which it now is. ~:-| Something is non-tetrapod-gnathostomey here.

  68. #68 heteromeles
    March 10, 2011

    To the squamates, I’d also add lizards. Monitors and/or necrosaurs were certainly around in the Paleocene. I’d even argue they might have been as smart as the arctocyonids, too.

  69. #69 Dartian
    March 10, 2011

    David:

    Most snake prey does need to be killed before swallowing to avoid injury to the snake.

    Indeed, even bats may be capable of inflicting injuries on a snake. Inside a cave in Tamaulipas, Mexico, that housed a colony of vampire bats, Villa & Lopez-Forment (1966) found two Mexican night snake Elaphe flavirufa individuals that had strange wounds on their bodies. The authors interpreted these as having most likely been caused by the teeth of vampire bats trying to defend themselves against the snakes. (One of these Mexican night snakes was caught by the researchers and kept in captivity for some months – and during this time the snake swallowed the stuffed skin of a vampire bat.)

    Speaking of snake predation on bats: for reviews on this topic, see Schätti (1984) and Esbérard & Vrcibradic (2007).

    References:

    Esbérard, C.E.L. & Vrcibradic, D. 2007. Snakes preying on bats: new records from Brazil and a review of recorded cases in the Neotropical Region. Revista Brasileira de Zoologia 24, 848-853.

    Schätti, B. 1984. Fledermäuse als Nahrung von Schlangen. Bonner Zoologische Beiträge 35, 335-342.

    Villa, B. & Lopez-Forment, W. 1966. Cinco casos de depredación de pequeños vertebrados en murciélagos de México. Anales del Instituto de Biología de la Universidad de Mexico 37, 187-193.

  70. #70 Kilian Hekhuis
    March 10, 2011

    So the main question here is: if a proto/antebat was a glider, how did it start flapping? As flapping is seen to be counterproductive to gliding. However, I’m not sure the latter statement is true: e.g. calugos, though not flapping their arms, are flapping their tail. Someone should try to find out whether flapping arms is indeed that counterproductive. Also, pterosaurs: they must have evolved in a similar fashion.

  71. #71 J. S. Lopes
    March 10, 2011

    Maybe protobats used their webbed hands before reaching the stage of gliders. Why did birds lose all fingers and bats didn’t? Pterosaurs’ finger remain unwebbed and too small, unlike bats’.

  72. #72 Owlmirror
    March 10, 2011

    what would you call a hypothetical terrestrial chiropteran?

    We have all, I hope, seen the bounding gait of a [vampire] bat running on a treadmill. Clearly, this sort of movement is optimised for pursuing insect prey that also moves in leaps and bounds.

    Thus, this hypothetical creature would no doubt be a cricket bat.

    (Well, you did ask…)

    PS: I have never played any SNES games whatsoever, FWIW.

  73. #73 heteromeles
    March 10, 2011

    @70: I disagree Killian, as posted at 9 and following. The big question isn’t whether they started gliding and then started flapping, it’s whether they glided that way at all. As I pointed out above, moving from gliding flying-squirrel style to flapping is maladaptive, because the aerofoil becomes unstable, and because the animal has to add weight to its arms, it also become proportionally heavier, meaning it falls faster.

    I suppose it’s possible that protobats were somewhat like flying frogs, using their enlarged hands as primary glding membranes. However, I do favor some scenario where they start off strengthening their arms, and basically beat their way into the air, becoming stronger and stronger fliers to the point where they could actually lift their own weight, rather than being aerial throughout and somehow totally rebuilding their anatomy to accommodate powered flight.

    The basic point is that only four lineages have ever become powered fliers (insects, pterodactyls, birds, and bats), but many more have become gliders (spiders, multiple lineages of fish, frogs, multiple lineages of lizards, snakes, multiple lineages of marsupials, and at least four lineages of placentals). Aside from the fact that an underpowered flier technically has to glide (because it can’t lift its own weight), there’s actually no reason to assume that powered flyers evolved from unpowered gliders.

    The group we SHOULD be focusing on are is the fish family gastropelecidae, the freshwater hatchetfishes. They are considered “gliders,” but anatomically, they are underpowered, short-distance flyers. They are the only extant group I know of (not counting secondarily flightless birds and insects) that flaps its forelimbs and doesn’t fly.

  74. #74 Owlmirror
    March 11, 2011

    Has anyone ever compared the development of gliding mammals and bats?

    I did have another antebat notion, about the uropatagium evolving first, between the legs and tail (and maybe partway down the torso?), but no more than that at first. This would allow some gliding, but leave the arms free to steer, and eventually evolve the shoulder mobility for flapping movement, perhaps with webbed hands.

    I originally thought of calling this the half-gliding scenario, but the part of my brain responsible for bat bad wordplay offered up the idea of calling it the bat-assward conjecture.

  75. #75 Dartian
    March 11, 2011

    Heteromeles:

    The group we SHOULD be focusing on are is the fish family gastropelecidae, the freshwater hatchetfishes. They are considered “gliders,” but anatomically, they are underpowered, short-distance flyers. They are the only extant group I know of (not counting secondarily flightless birds and insects) that flaps its forelimbs and doesn’t fly.

    Unfortunately, anecdotal claims notwithstanding, there is no good evidence that hatchetfish actively flap their pectoral fins while jumping out of the water (Wiest, 1995). Thus, gasteropelecids are probably not particularly good proto-bat (or -bird) analogues.

    Reference:

    Wiest, F.C. 1995. The specialized locomotory apparatus of the freshwater hatchetfish family Gasteropelecidae. Journal of Zoology, London 236, 571-592.

  76. #76 heteromeles
    March 11, 2011

    Thanks Dartian,

    I can only see the abstract, and (correct me if I’m wrong), what they’re saying is that the hatchetfish use their pectoral fins while in the water to launch themselves into the air, but that they do not use the fins in the air to propel themselves.

    So yes, hatchetfish are not proto-flying fish. Nonetheless, they may be useful for understanding bat evolution, because they are still using the pectoral region for launch. As Owlmirror noted above, antebats and protobats may have used something like the vampire’s gallop, especially for emergency escape. I still think (subject to a reality check) that they were arboreal rather than terrestrial, simply because a minority of modern bats can move effectively on the ground.

    The nice thing about this model is that it concentrates bat evolutionary development in the chest and arms, where it belongs. An antebat that invests more in jumping farther with its forelimbs has a competitive advantage, as does a protobat. This appears to avoid the problem of the flapping glider, where arm muscular development is disadvantageous until the protobat is close to powered flight.

    Of course, now I’m thinking of Spike the Bulldog from the Tom and Jerry cartoons. Another inappropriate metaphor, I’m sure.

  77. #77 Owlmirror
    March 11, 2011

    [Still thinking about my bat-asswards conjecture... sorry]

    I note that bats use their uropatagium to catch the insects they eat.

    Consider some small primitive (in the technical sense) arboreal insectivorous mammal of the Palaeocene (or perhaps even the Cretaceous). It’s small enough that it can fall any distance without injury (low terminal velocity), but as yet, has no obvious patagium at all.

    One fine day, it slips and falls through a cloud of flies, and catches a few of them in the folds of the skin of its nether portions. “Yum!” it thinks, and not being in the least fastidious, eats them up.

    Then it thinks “Aha!”, and climbs up the tree, and hurls itself downward again, trying to aim for the cloud of flies again.

    It’s one small step for an ur-bat, and one giant leap for batkind.

  78. #78 heteromeles
    March 11, 2011

    Thanks Darren, this is turning into a tremendous time waster. But a fun one!

    I started thinking about what a protobat would look like, and I came across this picture: http://upload.wikimedia.org/wikipedia/commons/9/99/Rousettus_egypticus.jpg. The genus Rousettus are thought to be primitive megabats, and this particular picture, if given a 90 or 180 degree rotation, is believable as a protobat, at least on a dim monitor with a couple of beers.

  79. #79 heteromeles
    March 11, 2011

    Oops. Try this link: http://en.wikipedia.org/wiki/File:Rousettus_egypticus.jpg. While you can go from the Wikipedia Rousettus page to the image, you can’t go directly to the image I linked above.

  80. #80 Andrew O'Donnell
    March 11, 2011

    @ heteromeles:
    Try http://upload.wikimedia.org/wikipedia/commons/9/99/Rousettus_egypticus.jpg

    I think it’s incorporating the period as a part of the URL.

  81. #81 abadidea
    March 12, 2011

    I woke up at 3am last night with my own idea for this thread. Suppose that instead of gliding, the protobats jumped out of trees to catch bugs and fell straight down, relying on their membrnanes to slow them down with air resistance. I think flapping behavior would make more sense in this situation? (I am not a biologist)

  82. #82 David Marjanović
    March 12, 2011

    the snake swallowed the stuffed skin of a vampire bat

    :-o

    and because the animal has to add weight to its arms, it also become proportionally heavier, meaning it falls faster.

    That would make it glide faster — it wouldn’t do anything to the distance it covered.

  83. #83 kad
    March 12, 2011

    Weight has no effect on how fast things fall. Wind resistance can but simply adding weight to the arms would not make the bat fall faster. Gravity is a constant and objects fall on earth at a rate of 32 ft/sec/sec regardless of how much they weigh.

  84. #84 heteromeles
    March 12, 2011

    @83: Irrelevant. What you say about weight is true only in a vacuum, and mammals cannot survive in a vacuum, let alone flap their wings to fly in one. Therefore, you can’t divorce weight from surface area.

  85. #85 Owlmirror
    March 12, 2011

    “Wind resistance” does indeed take into account the mass of the object, as well as its shape. See “Drag (physics)” in wikipedia for the formula (not linking because that always gets comments sent to moderation for me now), or any advanced physics textbook.

    Or do you think that a lead parachute would fall at exactly the same rate as a nylon one, all other things being equal?.

    ======

    “Dipterids are a super-nutritious and cowardly lot. . . . I shall become– a bat!”
      — Things a protobat might say.

  86. #86 kad
    March 12, 2011

    Of course you can’t divorce weight from surface area. A feather and a bowling ball would not fall at the same rate outside of a vacuum. However a bowling ball and a slightly heavier bowling ball would fall at the same rate, and that’s what we’re talking about in regards to adding weight to a bat’s arm.

  87. #87 heteromeles
    March 12, 2011

    @86: not really. Note that by adding weight to a protobat’s arm, you’re also shifting its center of mass forward (towards nosedive), unless you also make its butt heavier. Additionally you need to increase respiratory rates to keep all that extra muscle alive.

    If you want to actually do the experiment, make a paper airplane with a rectangular wing, then start adding paperclips behind the nose and see what effect this has on performance.

  88. #88 Dartian
    March 14, 2011

    Heteromeles:

    I can only see the abstract, and (correct me if I’m wrong), what they’re saying is that the hatchetfish use their pectoral fins while in the water to launch themselves into the air, but that they do not use the fins in the air to propel themselves.

    That is correct. While airborne and about to re-enter the water, hatchetfish do move their pectoral fins to some extent (apparently for stabilising purposes) but these movements can not be considered flapping.

    they are still using the pectoral region for launch

    And their tails, which the hatchetfish keep flicking from side to side while they are in the air. (Incidentally, Wiest suggests that this tail movement may explain the ‘buzzing’ sound that some eyewitnesses have claimed to have heard while observing these fish in nature.)

  89. #89 Riggi
    March 14, 2011

    Y’know what I just realize: Why not, instead of speculating on what these paleobats looked like in life, how about if we look into the embryonic development of these animals? True it won’t tell us the whole story of their evolution, at least it provides us with clues.

  90. #90 heteromeles
    March 14, 2011

    @89 Riggi: Here you go, bat development pictures on a scienceblog even:
    http://scienceblogs.com/pharyngula/2006/04/bat_development.php.

  91. #91 Riggi
    March 14, 2011

    @90 heteromeles: Oh wow, thanks mate! The pictures are quite stunning. Now I am not implying that we had some sort of “paleo-surfbat”, but it looks to me that the stages when the wings and feet are at, it reminds me of the webbed feet of semi-aquatic mammals (like the Yakpok).

  92. #92 Alan
    March 14, 2011

    Going back to the snake discussions – the other feature bats are famous for is their upside down method of resting. I wonder if this has any connection with a similar habit among Loriculus hanging parrots, who sleep that way to keep out of the way of snakes and other predators travelling along the upper surface of branches.

  93. #93 heteromeles
    March 14, 2011

    Do Loriculus hang to avoid snakes, or to feed? The pictures certainly make it look like can they feed upside-down too. Are you proposing the sloth-bat hypothesis for bat evolution?

    On a side note, it would be great if the molecular biologists or evo.devo. gurus would chime in about what’s going on with that embryonic bat development. I’d love to know what, if anything, the series tells us about bat evolution.

  94. #94 Riggi
    March 15, 2011

    @93 heteromeles: Yeah me too, I am actually curious to see what they have to say of it in terms of evolution. I wouldn’t be surprise they purposes a aquatic origin for bats. If so, is it even possible for a aquatic animal to evolve flight? I mean if you take the body plan of a sea lion, the front limbs would be excellent power for wing movement in the air. If they were smaller and had lighter bones, I can almost see a small, seal-like bat swiming one day and then propel itself for a short distance to escape pesky crocodiles and other swamp nasties. If true, then the wing is really a modified flipper in which the arm increases in length and the webbing increased in realtion to that as well. And ta-dah, wings! Of course, there would be some flaws (is it even possible to have an animal with the body plan of a aquatic animal to adapt flight?)

    @92 Alan: But would a predator be smart enough to check under the branch? I can’t imagine a snake not sensing prey on a branch while it slithers around the branch? It just doesn’t seem right to me. Maybe there is another reason why that some things just chill upside down (thermoregulation, metabolism or just no reason at all would be my pick).

  95. #95 Dartian
    March 15, 2011

    Alan:

    I wonder if this has any connection with a similar habit among Loriculus hanging parrots, who sleep that way to keep out of the way of snakes and other predators travelling along the upper surface of branches.

    That’s an interesting idea… but is it actually supported by empirical data? Are there any observations of sleeping/roosting Loriculus parrots (preferably in the wild) that have avoided detection by snakes or other predators by hanging underneath a branch? And if that seemingly simple method really is an efficient anti-predator strategy, it begs the question why no other birds (AFAIK) have adopted it.

  96. #96 David Houston
    March 15, 2011

    As a layperson looking at those embryos, it seems obvious the webbing developed first – and then the extended fingers.

    How that maps to evolution/ecological niches, I’ll leave to others to guess.

  97. #97 David Marjanović
    March 15, 2011

    I repeat: the colugos (“flying” “lemurs”, Dermoptera) have webbed fingers that participate in the wing.

    The development genetics stuff tells us that finger length can increase very drastically in one step.

  98. #98 Alan
    March 15, 2011

    Loriculus prefer to sleep hanging from the roof wires rather than larger perches – my guess is that they use the tip of thin branches as roost sites in the wild. Most of the parrots are very acrobatic and can feed from any position. I was wondering whether feeding from branch tips might be a niche conducive to developing powered flight – gliders tend to launch from branches or main trunks as these provide a more stable launch platform.

  99. #99 heteromeles
    March 19, 2011

    @99: great coincidence: see post #9.

  100. #100 heteromeles
    March 22, 2011

    Since I spent waaaay too much time thinking about this, I decided to go all the way and actually design an antebat and protobat and post them on my blog.

    Comments are welcome, if anyone is still interested in the topic.

  101. #101 Jason S.
    June 12, 2011

    The proto-bat is no longer a figment of the imagination:

    http://www.reptileevolution.com/protictis.htm

    New cladistic analysis with suggests that the poorly known Middle Paleocene laurasithere (or Protictis) haydenianus – formerly split into three genera and allied with the vivveravid miacoids – has skeletal similarities with Onychonycteris and is close to the ancestry of bats. The analysis, part of an ungoing project yet to be published, was headed by David Peters. (Don’t worry, he’s cleaned up his act and has stopped having hallucinations when comes to pterosaurs!)

  102. #102 P. Patrick
    June 12, 2011

    Jason – that’s very interesting thanks. But I must remind you that there’s a big difference between an “ongoing project yet to be published” and a published work that has made it through peer review. Despite what you say, the reptileevolution.com site (which uses lots of other people’s work -uncredited – and is hard to navigate and is highly, shall we say, idiosyncratic) does not in the least demonstrate that Mr Peters has “cleaned up his act” nor “stopped having hallucinations”. If anything, things are worse than ever.

  103. #103 David Marjanović
    June 12, 2011

    That analysis wasn’t “headed” but entirely done by David Peters, who still has trouble interpreting line drawings and photos of fossils, still has loads of correlated characters in his matrices, and still gives up every time reviewers reject one of his manuscripts — instead of following the reviewers’ suggestions, he retracts the manuscript and puts the results online.

    I can by no means exclude the possibility that viverravids and bats are closely related (IIRC, viverravids and carnivorans aren’t as close as people used to think), but that analysis is very poor evidence indeed.

    And Protictis is a laurasiathere. Laurasiatheria is the cover term for bats, perissodactyls, carnivorans, pholidotans, artiodactyls (incl. whales), and eulipotyphlans (Real True Insectivores).

  104. #104 Allen Hazen
    June 12, 2011

    Re: #101-103.
    Whatever one thinks of David Peters’s phylogenetic methodology and results, he certainly puts a lot of work into it!
    In the present case… I wouldn’t want to bet the farm on his speculative reconstruction of Protictis’s postcrania, given thqat he says the critter is known from a partial skull: look at the first published reconstruction of Pakicetus for an example of how far wrong that kind of speculation can go!
    I note that he suggests that Protictis is closely related to Chriacus: an Arctocyonid thaqt was thought to be close to the ancestry of artiodactyls until post-crania showing possibly arboreal and definitely NON-cursorial adaptations were found.

  105. #105 Allen Hazen
    June 13, 2011

    Re: David Marjanovic (#103)

    Sorry, I haven’t been doing my homework, so this may just be ignorant. What are our grounds for confidence that Protictis really is a Laurasiathere? Geography may not be enough: there have been a number of proposals over the last few years that various northern-continent early Cenozoic types (including Phenacodus, for crying out loud!) might actually be Afrotheres, and that Afrotheria may actually have originated outside Africa. And the last I heard, morphological traits distinguishing Afrotheres from Boreutheria were still hard to pick out. (Not to mention the possibility that some Paleocene Eutherians may be outside of crown Placentalia altogether, or the embarrassing fact that the earliest known Xenarthrans are so derived it’s hard to tell WHAT their sister group might have been like.)

    … I see David Peters suggests that Protictis might be close to the ancestry of both bats (Laurasiatherian) and Scandentians (tree shrews are in Euarchontoglires, aren’t they?)! Do we have enough of Protictis to be confident which of these it’s more likely to be related to?

    … The basic problem, I take it, is that the non-bat ancestor of bats was probably a pretty generalized small Eutherian and in the Paleocene … everybody’s ancestors looked pretty much the same!

  106. #106 Darren Naish
    June 13, 2011

    Incidentally, I’ll be following this article with a ‘protobats pt II’ some time soon.

  107. #107 Allen Hazen
    June 13, 2011

    Re #106–
    (Broad smile, and sense of great gratitude.)

  108. #108 Peter Nyikos
    July 13, 2011

    I am a latecomer to this blog, having learned about it in the Usenet newsgroup talk.origins from one of the participants. I had been talking about how I had long been mystified by what the intermediates between the second, colugo-like picture and modern bats might have looked like.

    Neither a decade and a half ago, when I first posted on this subject in talk.origins, nor this time around, could anyone suggest a sequence whereby the gap could have been bridged by intermediates that were at least as “fit” as their immediate ancestors.

    This blog is the first place where I have seen an intelligent effort to depict intermediates that show some promise in that direction. For the present, though, I lean towards the comments of Noni Mausa right in the second post to this blog. I don’t see where anyone has addressed those so far.

    The very first comments, by derek, do not seem to have been addressed either. My criticism is that there are no known intermediates, either living or fossil, which represent the transition between the colugo-like (“another flying squirrel”) second picture and full fledged bats.

    At least with birds and pterosaurs, a smooth transition from a terrestrial creature is quite easy to imagine. Even though we have nothing like Archaeopteryx to guide us wrt pterosaurs, the gradual lengthening of the outermost digit is does not produce intermediates that seem clumsier than their immediate hypothetical ancestors.

  109. #109 David Marjanović
    July 14, 2011

    At least with birds and pterosaurs, a smooth transition from a terrestrial creature is quite easy to imagine. Even though we have nothing like Archaeopteryx to guide us wrt pterosaurs, the gradual lengthening of the outermost digit is does not produce intermediates that seem clumsier than their immediate hypothetical ancestors.

    Recent work on development genetics suggests that the lengthening of the fingers in bats doesn’t need to have been gradual at all. You overexpress one gene, and the fingers grow like crazy — it’s more like a switch than like a dial.

    Remember that Onychonycteris retained claws on all 5 fingers per hand and probably was still able to use them all for climbing.

  110. #110 Peter Nyikos
    July 14, 2011

    Is overexpression of genes inheritable? Referring the appendages in Fig. 3.3: would a jump from B to D be possible from one generation to the next, with D or E what would be found in future offspring? To me, C looks like the clumsiest of the arrangements.

    How would the claws of Onychonycteris be actually used in climbing? The picture with the bats upside down on the pine tree suggests that all except the pollex would primarily be used in climbing down trees, not up. But at that early stage, before true flight was possible, one would think that it was more important to be able to climb up, not down.

    Come to think of it, the picture with the bats climbing down shows a folding arrangement of the other digits that seems to be aimed at some future winged stage. What would account for it appearing so early in the evolution of bats?

  111. #111 Owlmirror
    July 15, 2011

    Come to think of it, the picture with the bats climbing down shows a folding arrangement of the other digits that seems to be aimed at some future winged stage. What would account for it appearing so early in the evolution of bats?

    I have the impression that the bats climbing down are in fact already at a winged stage (it looks like the map to “D” in Fig 3.3), and “facing down” is simply the launching position.

    Yes?

  112. #112 David Marjanović
    July 15, 2011

    Is overexpression of genes inheritable?

    Depends on its cause… but usually yes.

    Referring the appendages in Fig. 3.3: would a jump from B to D be possible from one generation to the next

    I think what Sears et al. (2006) did to their mice was a jump from A to D. But I haven’t read the paper.

    How would the claws of Onychonycteris be actually used in climbing? The picture with the bats upside down on the pine tree suggests that all except the pollex would primarily be used in climbing down trees, not up.

    So what? A picture of a bat climbing down doesn’t mean the bat was unable to climb up. And a picture of a bat on a tree trunk doesn’t mean the bat was unable to climb among branches.

    I have the impression that the bats climbing down are in fact already at a winged stage

    Yes; in fact, Onychonycteris is not even in that picture — in the next picture, it’s closer to stage E than to stage D.

  113. #113 Peter Nyikos
    July 15, 2011

    Owlmirror, the protobats on the tree trunk in figure 3.3 look to me like they are at Stage C or just a little of the way from C to D. The airborne one looks like a Stage B. It seems a lot better adapted to climbing upwards than the ones on the trunk, while they are obviously better adapted to long glides, but not yet to sustained flight.

    Trivia: your handle is a direct translation of Eulenspiegel and I love the horn solo introduction to Tyl Eulenspiegel . To me it seems to capture the essence of the French horn just as the opening of Rhapsody in Blue captures the essence of the clarinet.

    One of my special joys is that my youngest daughter can play the French horn well enough to do justice to that solo.

  114. #114 Peter Nyikos
    July 22, 2011

    I think what Sears et al. (2006) did to their mice was a jump from A to D. But I haven’t read the paper.

    That would be amazing. Were the mice able to fly after a fashion?

  115. #115 Peter Nyikos
    July 22, 2011

    So what? A picture of a bat climbing down doesn’t mean the bat was unable to climb up.

    Look at the way the claws are curved. Those bats climbing down get great purchase the way the digits are folded back. Now try to imagine them climbing up with the digits folded back that way.

    Those proto-bats would have to turn their digits a completely different way to climb up. By the time they are at stage D, that seems very difficult, while with Onychonycteris that seems impossible.

  116. #116 David Marjanović
    July 23, 2011

    Were the mice able to fly after a fashion?

    I don’t know if, uh, the experiment was continued beyond the embryonic stage necessary to show the effect… But I’m sure they couldn’t have flown. Their muscles can’t have been up to it. Overexpression of BMP-2 at the right time in the right place gives you wings, not flight.

    Look at the way the claws are curved.

    As drawn, their tips point upwards, which means the poor animals look very, very unstable.

    In reality, which is three-dimensional, they’d have pointed inwards, pretty much horizontally into the tree trunk. Claws always point toward the palm of the hand/sole of the foot.

  117. #117 Peter Nyikos
    August 18, 2011

    Sorry to be so late in responding, David. I’ve been very busy.

    I concur with you about the mice not being able to fly. Presumably the transition in bat wing from B to F in Fig. 3.3 took thousands of generations, with the musculature keeping pace with the lengthening of the digits. But it is still a mystery to me what the capabilities of stage C and D were, and how the loss in climbing agility was compensated for by increased time in the air.

    About the claws: it’s not just a question of how the tips point, but also of the direction in which the claws are curved. Cats, for example, cannot turn their paws around; hence, because of the way the claws are curved, they have to climb down from high up in a tree by backing down instead of going forward the way squirrels can.

    The bats hanging upside down in the picture seem to be in the opposite predicament: if they tried to climb up, their claws would slip unless they could turn their manuses around completely.