[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: *Fruitafossor*

  Throughout this post, I will be trying a small test. Jon Wagner finally got
on my nerves about using the term "basal" much as I do lately (given that he
hates it with some vehemence), to which some of you may have gotten an earful
every time the words came out of your mouths last year in Denver in his
presence ;). So I am using two test terms to replace "basal" and "derived" when
using a phyllogram, and instead am using "root" and "stem" in that this is both
more aesthetic with the idea of a phyllogram (of leaves and bushes) or a
cladogram (of branching), and it carries no "worse than/better than" baggage
(his big complaint). Root and rootward = at or towards the root of a phyllogram
or a given node or stem; stem and stemward = at or towards an operational
taxonomic unit or group of such units; i.e., any clade, taxon, or rank you may
chose to determine. I use stemward and such without reference to a root, since
it's definition usually involves the given cladogram (and I will be referencing
only one such tree in this post); rootward is always used in reference to a
clade or taxon, and is included within that clade or taxon; i.e., following its
evolutionary origins but not as a terminal branching (if there are further
branchings of the stem it's included on but less from the referred node, it's
more rootward than stemward). I hope this helps. Jon doesn't subscribe to the
list currently, so this is being sent to him as well to get on his nerves even
more since he so LOVES mammals! I will likely get more replies on this
terminology than on the ideas presented on the phylogeny and features.

  Anyways, on with the show: I had written (in double braces):

<<As for the coracoid bone, I beleive it's fusion and/or transformation into a
coracoid process has ALSO been proposed.>>

to which David Marjanovic (david.marjanovic@gmx.at) wrote (in single braces):

<Which one? The procoracoid or the metacoracoid? Monotremes retain both. In
Trechnotheria (spalacotheriids + us), at least, the procoracoids form the caps
on the medial ends of the clavicles, while the metacoracoids form the processus
coracoidei on the scapulae.>

  These may not be the same osseous centers as in the basal amniote coracoid. I
am far more familiar with reptile coracoids. However, the protothere coracoids
I have seem appear homologous to the metacoracoid of therians, yet the coracoid
process appears to be different from the clavicular ossifications.
Sánchez-Villagra and Maier have some papers out on the topic and I am working
on getting a hold of them now.

<<Minimalization of the coracoid is recorded during mammalian evolution,>>

<Yes, once AFAIK>

  As far as the current data suggests, there aren't that many
exquisitely-preserved shoulders to tell us for true one way or another, though
what we do have seems to imply a single loss. I don't hedge bets (thus I
typically don't take bets).

<In short, no others are so strongly adapted to digging and myrmecophagy,

  Moles and chrysochlorids are just so adapted in the limbs and lack such
xenarthrous vertebrae and non-enamellated teeth. This is both dietary and
locomotory in distinction. While an animal that habitually burrows is one
thing, another that lives underground and locomotes beneath the ground requires
a separate set of adaptations. Moles and chrysochlorids are small-bodied,
insectivorous and flexible, so require stronger teeth relative to prey size and
more flexible spines than do just digging, anteater or armadillo-like animals,
which are surface foragers. So yes, there are no other mammals yet known so
strongly adapted, and this raises a few red flags in calling these convergent
when 20% of the skeleton including the basicranium is unknown. As I said
before, I like being thourough.

<Does it have epipubes? I bet it does... and I bet palaeanodonts don't...>

  I can only hazard a guess that any rootward member of the crown Eutheria is
more likely to retain epipubes than stemward, given their retention in "basal"
eutherians like *Eomaia.* Being suspected of them and preserving them are two
different stories, and we have no astoundingly complete basal palaeanodonts to
clue us in on the resolution.

<On the other hand, I'd certainly love to find out what palaeanodonts are.
Related to Xenarthra? To Pholidota? To some other placentals? >

  I would try to resolve wether palaeanodonts are a natural grouping, first,
given some ideas that they may be a paraphyletic with regards to xenarthrans,
that xenarthrans are paraphyletic with regards to palaeanodonts, any
relationship with pholidotes that are NOT based on parallel or convergent
evolution, etc. This is a problem I was trying to point out with regards to the
presence of tubular teeth lacking enamel, since it HAS developed twice in
living mammals, given tubulidentates and xenarthrans. It may have developed
convergently WITHIN xenarthrans, for that matter. I won't put it past those
wascally teeth to try things out repeatedly among unrelated groups throughout
time. What works, works, and if successful, you're likely to see it again.
Gotta love cladistic homoplasy, too, another problem I was trying to bring up
with regards to Luo and Wible's analysis.


Tim Williams (twilliams_alpha@hotmail.com) wrote:

<Overall, the modern placentals are severely underrepresented in the matrix. 
There's a lipotyphlan, two carnivorans, one rodent, one lagomorph and six
xenarthrans (the old "edentates").  There's no ungulates or subungulates, or
afrotherians, or primates.  Even the aardvark (_Orycteropus_) was left out,
despite being mentioned in the text.>

  Unfortunately this limited extant sampling may be enforcing the homoplasy of
less derived clades that would share common features with *Fruitafossor*, and
the absence of *Orycteropus* and *Manis* are rather glaring as they had been
mentioned in the text. The authors have, as David noted, focused on their
previous analyses of the *Sinodelphys* and *Eomaia* matrices


 Finally, the next email by David Marjanovic prompts a more complete response
on some fine details of anatomy. Please skip if you're bored at this point. YOU
HAVE BEEN WARNED, I am getting a tad pedantic here, and it's not David's fault.

<Through dark, albeit not hideous, channels I have acquired a pdf of the
description of *Fruitafossor*. It's almost cute how strongly the authors
emphasize _twice at every occasion_ that *F.* is not a xenarthran or anything
remotely similar; that's because *F.* falls between the error margins of the
ludicrous divergence time estimate for Xenarthra vs. other placentals
calculated by Kumar & Hedges in 1998 (129 +- 18.5 Ma ago).>

  Yes, driving a point this hard can be very easy (and misleading), one fails
to see the offshoots of the path because the avenue is so fascinating. Hence my
tagline. To illustrate this, the actual diagnostic differences listed (and NOT
the full diagnosis for it's length) are [my comments are dispered throughout
within brackets, like this one]:

  "*Fruitafossor windscheffeli* differs from all Cenozoic fossil therians
    (including palaeanodonts) and extant placentals (including tubulidentates
    and xenarthrans) in retaining a long list of plesiomorphies including:
    presence of a broad Meckel?s sulcus on the mandible [no comment, but
    to be related to a basal middle-ear complex, lacking even in highly derived
    digging mammals which use the jaw to "tune in," like naked mole rats]; in
    the presence of widely separated and spherical radial and ulnar condyles of
    the humerus [this is locomotory in nature], in the absence of a spherical
    head of humerus [less generalized movement of the humerus relative to the
    shoulder, referring to a more strict range of movement in the arm], in a
    strong torsion of the humerus with hypertrophied deltopectoral and teres
    crests separated by a broad intertubercular groove [very mole-like and so
    far indicative of hyperfossorial digging habitus]; in the presence of a
    transversely broad and laterally directed greater trochanter of the femur
    [see above]; in the presence of a cynodont-like, broad peroneal shelf of
    calcaneus [lacking in more rootward crown mammals], and in the vertical
    orientation and medial placement of the astraglar and sustentacular facets
    on the calcaneus [restricted range of ankle movement could very well be
    related to a hindlimb component to digging].

  "*Fruitafossor windscheffeli* differs from all xenarthran and pholidotan
    placentals in additional pelvic characters: no fusion of the proximal
    vertebrae to the sacral vertebrae [supposedly such fusion is synapomorphic
    of a Pholidota + Xenarthra clade,m but some morph trees and new gene trees
    place pholidotes and xenarthrans a LONG distance from one another, thus
    would be convergent between the two and says nothing for referring to *F.*
    as a basal morph of either, not that I am calling it either; can't make an
    evolutionary claim off of this, however]. It differs from all Cenozoic and
    extant xenarthrans in that the caudal vertebrae are not fused to the
    to form the synsacrum [see above]. It differs from palaeanodonts,
    xenarthrans, and several Early Cenozoic fossorial mammals in that its
    deltopectoral crest is crest-like, not as broad mediolaterally and distally
    extended as in palaeanodonts [plesiomorphy], xenarthrans, *Eurotamandua*,
    and *Ernanodon*. Although *Fruitafossor* has xenarthrous articulation
    (presence of lateral zygapophyseal joint in addition to the medial
    zygapophyseal joint) in four of the five lumbar vertebrae, its thoracic
    vertebrae lack the xenarthrous intervertebral articulation, as in Cenozoic
    and extant xenarthrans [so this doesn't really say much, and we all think
    xenarthry was progressive and not spontaneous in the lumbar spine]. *F.
    windscheffeli* and Early Tertiary placental palaeanodonts both have single,
    and openrooted cheek teeth with elliptical cross-sections; but
    retains only one or two cheek teeth (Simpson, 1931; Emry, 1970), far fewer
    than the six cheek teeth of *F. windscheffeli* [plesiomorphy; the authors
    should know better, but this IS a full differential diagnosis]."

<Pangolins, palaeanodonts, *Eurotamandua* and *Ernanodon* (a particularly
mysterious Paleocene mammal from China that's sometimes suspected to be a
xenarthran) are mentioned in the supplementary information. They all lack the
tremendous list of plesiomorphies that *F.* retains. (On the other hand, it
isn't mentioned that basal palaeanodonts, like *Arcticanodon* which was
mentioned onlist today, have a more plesiomorphic dentition than *F.*, for
example retaining two-rooted cheek teeth!)>

  Though this says something, Ross et al. refer *Articanodon* to the
Palaeanodonta based on the relative shape of the mandibular ramus and the shape
of the canine. We all know how rare convergent jaw shape and tooth shape is in
Mammalia. Enough said. Based on the double rooted postcanines, I would likely
have been a LOT more cautious about systematic assignment of this taxon.
Evidence of the Meckelian groove is lacking.

<I can't hazard a guess about whether the middle ear was still connected to the
lower jaw (as the authors infer) or if the Meckelian groove just housed a big
splenial. However, the vestigial angular process is interesting --  that's
exactly how the missing link between the australosphenidan grade (bigger
angular process) and the triconodont-multituberculate grade (no trace of an
angular process) is supposed to look like! :-)>

  The Meckelian fossa is very large and there is a distinct crest separating it
and the mandibular trough, a feature not present in mammals which retain a
large mandibular component to the middle ear (as in *Morganucodon* and
*Docodon*) [yes, I am using Rich et al.'s work on *Teinolophos* here]. The
dorsal Meckelian shelf is not broad as in root mammaliaforms and root
mammalians, but narrow and defining an extremely large component of the medial
jaw to the pterygoid fossa, rather than the likely now-incorporated middle-ear
bones. If nothing else, this resembles the triconodont jaw structure more than
it does any therian condition. I do not see more substantial evidence to
support a complex of middle ear bones medial to the mandible. 

  The authors do not elaborate on the inferrence in the main thesis of the
paper, and details are not clear except in the analysis, in which of 16
included characters (#'s 350-365) of the middle ear, none are scored as present
or absent, simply unknown. Middle-ear related features of the mandible (3's
4--5 & 12) are scored as: 2 [? -- unknown scars for prearticular and
surangular], 3 [? -- unknown overhanging ridge above Meckel's sulcus (groove);
I call this the subdental platform when it is further anterior in reptiles, but
in this case I called it the Meckelian ridge (see above)], 4 [0 --
well-developed Meckel's sulcus (groove) {plesiomorphic}], 5 [0 -- Meckel's
sulcus (groove) parallel to ventral mandible margin {plesiomorphic}]; and 12 [0
-- mandibular foramen in posterior Meckel's sulcus (groove) {plesiomorphic}].
2, 4, and 5 are scored as unknown in all included xenarthrans (such traces are
absent, as the bone has undergone such dynamic evolution as to be elongated
oval rods with a few flanges for muscle insertion in the anteaters and without
much variation in armadillos and sloths); 3 is coded as (1) for all xenarthrans
save *Euphractus,* where it is coded as unknown; and 12 is coded as 4 in all

  Because the lower jaw has undergone such transformation in xenarthrans, they
can be compared only to character 12 favorably, in which *Fruitafossor* has
state 4 (mandibular foramen not associated with either Meckel's sulcus {groove}
or the postdentary trough), and xenarthrans have state 0. Comparatively,
monotremes are coded as 01000 (relatively) in *Ausktribosphenos*, 1?100 in
*Bishops*, 10023 in *Teinolophos*, ??000 in *Steropodon*, ??2?3 in both
*Obdurodon* and *Ornithorhynchus*, and *??2?4 in *Tachyglossus*. In more
stemward therians, *Jeholodens* has ??101, *Zhangeotherium* has ??102,
*Prokennalestes* and *Eomaia* share ??111. and *Zalambdalestes* and
*Erinaeceus* share ??2?4. Based on all characters associated with the middle
ear, it can be inferred that *Fruitafossor* is closest in morphology to
*Ausktribosphenos* (3 out of 5) and *Steropodon* (the same, and for the same
characters). Poor character selection or lack of discussion of the derived or
"obliterated" condition in slender-jawed mammals like *Ornithorhynchus* and
xenarthrans appears to also affect the scoring of jaw features. *Teinolophos*
is the only taxon that could be scored for all characters without scoring for ?
at any state. The large margin of unknown cranial and some mandibular features
due to character selection may result is less than a complete prognosis of the
mandibular variation and similarities among all therians included. Nonetheless,
there is no reason to suspect that *Fruitafossor* had a very primitive ear
structure, but likely one as developed as any root ornithorhynchid or even root
therian or eutherian, given the relative placement of elements of the jaw

<To answer my own questions, yes, it's the metacoracoid, and yes, epipubes are
at least reconstructed...>

  As noted above, the epipubes may not be statistically very important, present
as they are throughout crown mammals up into rootward crown eutherians.


Jaime A. Headden

  Little steps are often the hardest to take.  We are too used to making leaps 
in the face of adversity, that a simple skip is so hard to do.  We should all 
learn to walk soft, walk small, see the world around us rather than zoom by it.

"Innocent, unbiased observation is a myth." --- P.B. Medawar (1969)

Yahoo! Messenger 
Show us what our next emoticon should look like. Join the fun.