Sallen 2016 presents a fascinating flawed look at fish tails

Sallen 2016 reports,
“The symmetrical, flexible teleost fish ‘tail’ has been a prime example of recapitulation — evolutionary change(phylogeny) mirrored in development (ontogeny).”

Sallan’s cladogram (Fig. 1) lays out the traditional cladogram of fish. Note the position of the bichir (Polypterus), at a basal node and the sturgeon + paddlefish (Acipcenser + Polyodon) near the middle.

Figure 1. Cladogram from Sallan 2016 (above) and young fish tails (below).

Figure 1. Cladogram from Sallan 2016 (above) and young fish tails (below).

Unfortunately,
taxon exclusion mars the cladogram of Sallan 2016 according to the the large reptile tree (LRT, 1704+ taxa; Figs. 2, 5). Due to tradition Sallan has chosen the wrong outgroup. Jawless sturgeons and shark-like paddlefish should be the outgroups here, not lungfish-like bichirs (Polypterus), which are highly derived taxa close to lungfish and tetrapods.

Figure 2. Same taxa as above, but rearranged to fit the LRT tree topology.

Figure 2. Same taxa as above, but rearranged to fit the LRT tree topology. Remember, sturgeons, paddlefish and sharks are basal taxa in the LRT. Esox is a catfish related to placoderms.

Salan reports,
“Paleozoic ray-finned fishes (Actinopterygii), relatives of teleosts, exhibited ancestral scale-coveredtails curved over their caudal fins. For over 150 years, this arrangement was thought to be retained in teleost larva and overgrown, mirroring an ancestral transformation series. New ontogenetic data for the 350-million-year-old teleost relative Aetheretmon overturns this long-held hypothesis.”

By contrast,
in the LRT Aetheretmon nests with Pteronsculus (Figs. 5–7)) far from the base of all bony fish, much closer to lobefin fish and tetrapods.

The Sallan point is still made:
Many fish tails do have two parts, especially when hatchlings.

Unfortunately, Sallan does not understand
the topology of the family tree of fish due to taxon exclusion. This is something the LRT minimizes by testing a wider gamut of taxa. As readers know, we see this same taxon exclusion problem all the time in paleontology.

Figure 2. Muskie (Esox) tail ontogeny from Sallan 2016 (middle row). Top row and photo added here.

Figure 3. Muskie (Esox) tail ontogeny from Sallan 2016 (middle row). Top row (to scale) and photo (below) added here. You might remember, Esox is a derived catfish without barbels.

Salan writes,
These two tails appear at a shared developmental stage in Aetheretmon, (Fig. 4) teleosts and all living actinopterygians. Ontogeny does not recapitulate phylogeny; instead, differential outgrowth determines final morphology.”

That appears to be so, but it still needs a valid tree topology.

Figure 3. Fish tail ontogeny in extinct Aetheretmon and extant Monotrete. Note the upper and lower lobes.

Figure 4. Fish tail ontogeny in extinct Aetheretmon and extant Monotrete. Note the upper and lower lobes. In the LRT these two fish are not closely related. Aetheretmon is basal to lobefins. Monotrete is a puffer fish.

Salan speculates:
“The double tail likely reflects the ancestral state for bony fishes.”

No, the ancestral state for bony fish is the heterocercal tail documented by sturgeons and whale sharks, and this goes back to armored osteostracans according to the LRT (Fig. 5).

Figure x. Subset of the LRT, focusing on fish for July 2020.

Figure x. Subset of the LRT, focusing on fish for July 2020.

Salan speculates,
“Many tetrapods and non-teleost actinopterygians have undergone body elongation through tail outgrowth extension, by mechanisms likely shared with distal limbs.”

Not sure what those ‘mechanisms’ would be, but basal tetrapods and stem tetrapods in the LRT have relatively short, straight tails and elongated bodies with great distances between the fore and hind limbs. Look at Panderichthys.

Figure 5. Aetheretmon is known from the oldest complete growth series for vertebrates.

Figure 6. Aetheretmon is known from the oldest complete growth series for vertebrates.

Figure 6. Pteronisculus, a sister to Aetheretmon in the LRT.

Figure 7. Pteronisculus, a Triassic sister to Early Carboniferous Aetheretmon in the LRT and it is easy to see why.

Sallan is ‘Pulling a Larry Martin’
by putting too much emphasis on one trait without testing all the traits on many more taxa. Only after a valid phylogenetic context is established can one begin to figure out if trait A came before trait B or not.

Sallan goes into great detail describing
the successive stages of growth in Aetheretmon, but this is problematic because the cladogram is invalid. “First things first” is a motto all paleontologists should ascribe to. First get the phylogeny correct. Fish workers are relying on an invalid family tree. The LRT is here to fix that.

Its worth remembering,
many fish on the other branch of bony fish (perch, anglers, etc., Fig. 5, orange right column) bring the pelvic fins beneath the pectoral fins, shortening the gut cavity and elongating the tail to extremes in some cases (oarfish). This is all distinct from the longer torso, shorter tail trend in the stem tetrapod branch of bony fishes (Fig. 5, yellow left column).


References
Sallan 2016. Fish ‘tails’ result from outgrowth and reduction of two separate ancestral
tails. Current Biology 26, R1205–R1225.
White EI 1927. The fish fauna of the Cementstones of Foulden, Berwickshire. Transactions of the Royal Society Edinburgh 55:255–287.

https://www.the-scientist.com/the-nutshell/a-tale-of-two-tails-32394

Batrachognathus tail – Rio Ptero Symposium

This post continues a series on Rio Ptero Symposium abstracts found earlier here, here, here and here.

Costa et al. (2013) were surprised
to discover a long stiff tail on an undescribed specimen they attributed to Batrachognathus associated with the holotype of Sordes (the first and only incidence of two distinct pterosaur genera on one slab). The tail was nearly half again as long as the humerus and included filiform hemapophyses, pre- and post-zygopophyses that stiffened the tail. They provided no figures or photos but measurements of the new humerus and tail indicate the former was shorter and the later was similar in length.

Neither the original description
of the holotype of Batrachognathus (Fig. 1) by Rjabinin (1948) nor the subsequent redescription in an abstract by Bakhurina (1988) reported the presence of a tail, but assumed if a tail were present it would be short based on the situation in Anurognathus, the only other anurognathid known at the time. Unfortunately both workers simply overlooked the tail, along with several other oversights. A long tail is indeed present on the holotype, but it doesn’t match the description by Costa et al.

Figure 1. Batrachognathus in situ.

Figure 1. Click to enlarge. Holotype of Batrachognathus in situ.

DGS (digital graphic segregation)
helped to find the tail (Fig. 2, identified by CAV = caudal vertebrae).

Figure 2. Batrachognathus in situ as traced using DGS. The tail is identified by CAV, rimming the right side of the slab.

Figure 2. Clickd to enlarge. Holotype of Batrachognathus in situ as traced using DGS. The tail is identified by CAV, rimming the right side of the slab and originating on the sacrals. The tip of the tail contributes to the ptero-hair (pycnfibers) reported earlier.

Roadkill reconstruction
The DGS tracing (Fig. 2) found most of the parts of Batrachognathus. Even so, this chaotic roadkill specimen required a reconstruction (Fig. 3) to make sense of it. Note the tail is not stiffened by bony spines, nor is it half again as long as the humerus.

Figure 3. Batrachognathus volans recontructed. Note the tail is not half again as long as the humerus and not provided with stiffening spines, casting doubt on the identification of the Costa specimen.

Figure 3. Holotype of Batrachognathus volans recontructed. Note the tail is not half again as long as the humerus and not provided with stiffening spines, casting doubt on the identification of the Costa specimen.

Too different?
These differences cast doubt on the affinity of the new specimen to Batrachognathus. Without seeing the specimen either in photos or figures, one can only speculate as to the identify of this new specimen.

Anurognathid tails
Anurognathids were derived from a sister to Dimorphodon and the earliest proto-anurognathids, like Peteinosaurus, retained a short yet robust tail. The tail of the MCSNB 8950 specimen (formerly attributed to Eudimorphodon) is broken after just a few vertebrae, but is likewise robust. Both of these tails show greatly reduced stiffening elements.

The tail of the IVPP embryo shows the correct length (half again as long as the humerus), but also demonstrates a complete lack of bony stiffening elements on a series of disc-like vertebrae. The correct (not the chimaera) tail of Dendrorhynchoides is also sufficiently long, but lacking in stiffening elements. The same goes for the GLGMV 0002 specimen attributed Dendrorhynchoides, as well as the holotype of Jeholopterus.

If not an anurognathid, then what?
The type of tail described by Costa et al. can be found in the no. 110 specimen of Scaphognathus (Fig. 4) and related taxa. Further speculation without seeing any data on the specimen is unwarranted.

 

Figure 4. The Maxberg specimen attributed to Scaphognathus, no. 110 in the Wellnhofer 1975 catalog. Note the tail is greatly reduced yet retains bony stiffening spines, as described by Costa et al.

Figure 4. Figure 4. The Maxberg specimen attributed to Scaphognathus, no. 110 in the Wellnhofer 1975 catalog. Note the tail is greatly reduced yet retains bony stiffening spines, as described by Costa et al.

Yet one wonders,
based on the sparse description of Costa et al.) if the new specimen is indeed an anurognathid at all. A close relative of Sordes, Changchengopterus, has a relatively short tail, but not the hyper-robust humerus of Batrachognathus.

References
Bakhurina NN 1988. [On the first rhamphorhynchoid from Asia: Batrachognathus volans Riabinin 1948, from Tatal, western Mongolia]. Abstract of paper in Bulletin of the Moscow Society for the Study of Natural History, Geological Section 59(3): 130 [In Russian].
Costa FR, Alifanov V, Dalla Vecchia FM and Kellner WA 2013.
On the presence of an elongated tail in an undescribed specimen of Batrachognathus volans (Pterosauria: Anurognathidade: Batrachognathinae). Rio Pterosaur Symposium 54-56.
Rjabinin AN 1948. Remarks on a Flying Reptile from the Jurassic of Kara-Tau. Akademia Nauk, Paleontological Institute, Trudy 15(1): 86-93.

wiki/Batrachognathus

Rhamphorhynchus. Growth Series? Or Speciation?

One of the biggest mistakes I found in paleontology was the unwarranted lumping of all Rhamphorhynchus specimens under one species based on long bone measurements and statistics. Forsaking phylogenetic analysis, Dr. S. Chris Bennett introduced this hypothesis in 1995 and it has been followed and referenced ever since (Unwin 2005) without confirmation (more below). Phylogenetic analysis was not attempted then (or since).

Figure 1 shows the Rhamphorhynchus clade to scale and in roughly phylogenetic order (left to right) based on the large pterosaur study here. A long list of Rhamphs have never been included in a phylogenetic analysis before, so this is a first. One look (at Figure 1) is all it takes to see the morphological variety present here to say nothing of the phylogenetic size variation. The annotated Nexus file is available on request.

The family tree of the Rhamphorhynchus.

Figure 1. Click to enlarge. The family tree of Rhamphorhynchus to scale. That’s Campylognathoides batting first. The largest of the bunch, no. 81, phylogenetically followed the smallest, No. 10. This clade is ripe for a great dissertation. 

From Large to Small to Giant to Medium-Sized
The genus Rhamphorhynchus is led off by the C3 (Pittsburgh) specimen of Campylognathoides, the phylogenetic ancestor. The basal taxon, R. intermedius (No. 28) was the one closest to Campylognathoides in trait similarity. Continuing the size trend, a smaller series of Rhamphs follow, including R. longicaudus (see below). The giant of the bunch, R. longiceps was followed by a series of medium-sized Rhamphs with longer first wing phalanges and nares set further back on the skull.

One of the Littlest 
Rhamphorhynchus longicaudus (Smith-Woodward 1902, B St 1959 I 400, no. 10 of Wellnhofer 1975, Fig. 2), Late Jurassic ~155 mya, was considered a juvenile by Bennett (1995). Actually it is just another tiny species with a distinct morphology nesting close to other tiny species. Similar in size to and derived from a sister to the BMM specimenno. 10 phylogenetically preceded the giant Rhamphorhynchus longiceps no. 81. Another R. longicaudus specimen, No. 11, actually had proportions more typical of R. longiceps and R. muensteri. It has not been included yet in phylogenetic analysis.

rhamphorhynchus_longicaudus-no10

Figure 2. Rhamphorhynchus longicaudus no. 10. Click for more info.

Distinct from the BMM specimen, the skull of R. longicaudus had a longer, thinner rostrum and a relatively larger skull with a narrower lateral temporal fenestra. No. 10 had a hooked lower jaw longer than its upper, the opposite of most other Rhamphs. It had a low hard crest and a high soft crest on its skull. The anterior teeth were longer and sharper. The cervicals were longer relative to the dorsals. The caudals were more gracile and longer. The sternal complex was somewhat cardiod in shape and reduced in size. The forelimb elements were all more gracile. The posterior is unknown in no. 10, but reconstructed here based on similar specimens. The pubis and ischium were close if not joined. The hind limb elements were all more gracile, including the metatarsals and toes.

Rhamphorhynchus longiceps

Figure 3. Rhamphorhynchus longiceps (Smith-Woodward 1902) BMNH 37002, no. 81 in Wellnhofer 1975. Click for more info.

The Giant of the Bunch
Rhamphorhynchus longiceps (Smith-Woodward 1902, BMNH 37002, no. 81 in Wellnhofer 1975, Fig. 3), was the largest known Rhamphorhynchus. Derived from one of the smallest known species, R. longicaudusR. longiceps phylogenetically preceded R. muensteri.

Distinct from R. longicaudus, the skull of R. longiceps was more robust and longer than the torso. The rostrum was pointed and probably sharpened with a keratinous extension. The orbit was only twenty percent of the skull length. The premaxillary teeth were reduced and bunched together. The anterior dentary was concave dorsally. The cervicals decreased in length anteriorly. Seven sacrals were present. The tail was robust but unknown in length. The dorsal ribs were more robust. The sternal complex was rectangular but gently rounded both anteriorly and posteriorly. The humerus was robust. The posterior ilium was as long as the anterior. The pubis and ischium were separate. The prepubic perforation was filled in. The The pedal digits were longer than the metatarsus.


Growth Series? Or Speciation?
Dr. Peter Wellnhofer (1975) continued the traditional labeling of various Rhamphorhynchus  morphotypes as distinct species. Twenty years later, using statistics measured from long bones, Bennett (1995) envisioned a growth series in Rhamphorhynchus with dramatic morphological changes during maturation. This is a blunder. These specimens are morphologically distinct down to the phalangeal proportions (Peters 2011, Fig. 4) and so represent a phylogenetic sequence. The largest specimen is followed phylogenetically by smaller specimens. We also know from pterosaur embryos that hatchlings greatly resembled their parents and therefore did not go through great morphological changes during maturation. The “juvenilization” during size reduction goes back to accelerated developments at the embryonic stage. Read more about the speciation of Rhamphorhynchus here.

 

Figure 4. Click to enlarge. Rhamphorhynchus pedes.Figure 4. Click to enlarge. Rhamphorhynchus pedes.

Figure 4. Click to enlarge. Note the variety in Rhamphorhynchus pedes. These are not conspecific.

Just Like Pteranodon
A similar phylogenetic blunder without phylogenetic analysis occurred when Bennett (1991, 2001) considered all specimens of Pteranodon restricted to just two species. That hypothesis was challenged here.

An Encouraging Note to Any Future Pterosaur Workers
I hope someone takes this lead and runs with it. A darn good dissertation could be written using two to three dozen Rhamphorhynchus specimens, lumping and separating them.

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

References
Bennett SC 1995. A statistical study of Rhamphorhynchus from the Solnhofen Limestone of Germany: Year-classes of a single large species. Journal of Paleontology 69: 569–580.
Maisano JA 2002. Terminal fusions of skeletal elements as indicators of maturity in squamates. Journal of Vertebrae Paleontology 22: 268–275.
Peters D 2011. A Catalog of Pterosaur Pedes for Trackmaker Identification. Ichnos 18(2):114-141. http://dx.doi.org/10.1080/10420940.2011.573605
Smith-Woodward A 1902. On two skulls of the Ornithosaurian  Rhamphorhynchus. Annals and Magazine of Natural History, London, (7) 9: 1-5.
Unwin DM 2005. The Pterosaurs: From Deep Time. New York, Pi Press, 1-352.
Wellnhofer P 1975a-c. Teil I. Die Rhamphorhynchoidea (Pterosauria) der Oberjura-Plattenkalke Süddeutschlands. Allgemeine Skelettmorphologie. Paleontographica A 148: 1-33.Teil II. Systematische Beschreibung. Paleontographica A 148: 132-186. Teil III. Paläokolgie und Stammesgeschichte. Palaeontographica 149: 1-30.

wiki/Rhamphorhynchus