GEOL 431 Vertebrate Paleobiology

Spring Semester 2023

Gnathostomata and Placodermi

Gnathostomata - the jawed vertebrates combine a great range of evolutionary novelties. They appear suddenly, with a large suite of fully-blown derived features.

Chronology:

The earliest well-preserved gnathostomes are a placoderm and an chondrichthyan from the Early Silurian Huixingshao Formation of Chonqing, China. Prior to this, the record consists of individual placoid scales dating as far back as the Late Ordovician that are boldly assigned to Chondrichthyes. (Zhu et al., 2022)

Synapomorphies:

Developmental:
• Four Hox gene clusters (compared with two ancestrally in Vertebrata and one in invertebrates.)
• Enlarged forebrain that is flexed ventrally with respect to hind brain. (Compare with ammocoetes larva of lamprey.)



Placoderm (left) and chondrichthyan (right) neurocrania in dorsal view.
Head:
• Paired nasal capsules communicate with the outside through paired nostrils. Their bony openings are the nares (sing. "naris")
• Hypophyseal duct opens onto roof of mouth
• Postorbital process of neurocranium (right)
• Horizontal (third) semicircular canal
• braincase incorporates occipital arch posteriorly.

Girdles:

• Pectoral fin supported by a pectoral girdle with:
  • Paired endochondral scapulocoracoids
  • Various dermal elements.
• Paired pelvic appendages with internal skeleton and endochondral pelvic girdle. (Only the most basal gnathostomes are slightly odd with dermal pelvic elements and possibly lacking endochondral ones (Zhu et al., in press).)



Schematic shark branchial skeleton from University of the Cumberlands Comparative Anatomy

Branchial arches:

• Branchial skeleton internal to gills. (In contrast to the external branchial basket of lampreys.) Although it was long believed that the gill filaments were ectodermal in gnathostomes and endodermal in lampreys, Gillis and Tidswell, 2017 have demonstrated that they are endodermal so we are confident that they are homologous in all vertebrates.
• The visceral skeleton (gill arches and related structures) consists of jointed elements:
  • epibranchials - dorsal
  • ceratobranchial - ventral
  These elements that meet at a posteriorly pointing V. Additionally:
  • Pharyngobranchials dorsal to the epibranchials
  • Paired hypobranchials ventral to the ceratobranchials
  • Mid-line basibranchials
• Ancestrally there are five or fewer branchial arches.





Squalus skull and arches from British Chalk Fossils

The Big News: The first two arches are specialized

• The jaws made of the mandibular arch consisting of paired:
  • palatoquadrates - upper elements
  • Meckel's cartilages - lower elements.
• The hyoid arch, that helps connect the mandibular arch to the braincase. It breaks down into:
  • Hyomandibulae (sing. hyomandibula) - upper elements that articulate with the braincase
  • Ceratohyals - lower elements that support Meckel's cartilages
  • Basihyal - ventral mid-line element.
  Note: These structures form first as cartilage and give rise to different patterns of ossification in different critters. (Later.)

This list invokes some challenging problems in gnathostome evolution. These are take up below, after an anatomy review.

Jaw anatomy: Up until now, when we have spoken of a "skull" we have meant only a plating of dermal bone around the head, or, in the case of Galeaspids and Osteostraci, that plus in perichondrally ossified neurocranium or braincase. With the appearance of the gnathostome jaws and branchial skeleton, the skull becomes a complex composite structure. The following illustrations show its components using the fossil bony fish Eusthenopteron as an example.

• Start with the neurocranium:

  
  

• Add the branchial arches:

  
  

• Add the hyoid arch:
  

  
  

• Add the mandibular arch - palatoquadrate and Meckel's cartilage:

  Note, whereas the branchial arches are separated from the hyoid arch and from one another by gill slits, the space between the hyoid and mandibular arches is either closed or perforated by the spiracle. (This is particularly conspicuous in rays, who breath in through it.) See also this link.
  

• Finally, add the dermal bone of the skull roof:

  
  

Six Gnathostome Issues:

1. The Origin of Jaws:

Many gnathostome synapomorphies are continuations of longstanding vertebrate trends toward increased skeletal ossification, brain enlargement, and improvement of swimming. Big exceptions are the changes to visceral arches.

Oddly, fossil intermediaries between jawed and jawless vertebrates just don't seem to exist. What's up?

Three historic hypotheses:

Serial homology of visceral arches: Carl Gegenbaur in 1872 proposed that the jaws and hyoid arch represented modified anterior branchial arches. The earliest versions would have come complete with respiratory gills. Gegenbaur responsibly suggested future fossil evidence that would support this in a specific way:

• In many modern gnathostomes, the mandibular and hyoid arches enclose an opening called the spiracle (=eustacean tube in land verts). In living aquatic vertebrates, the spiracle can be invested with pleated gill-like tissue called the pseudobranch. To Gegenbauer, this was the vestigial remnant of what was originally a full gill occupying a full gill slit between the mandibular and hyoid arches, prior to their becoming specialized. Gegenbaur predicted that fossil gnathostomes would be found with a complete gill slits where the spiracle sits.

Alas, but after a century of careful scrutiny, no such transitional forms have been found. Ironically, Gegenbaur's appealing hypothesis was so widely accepted that in many modern textbooks it is still treated as revealed wisdom.

During the 20th century, two new hypotheses have arisen:



Jon Mallatt
The New Mouth: Jon Mallatt, 1984, noted that the branchial arches of gnathostomes and lampreys can't be homologous, as they are medial to and lateral to the gills, respectively. (Indeed, sharks retain extrabranchial cartilages lateral to their gills.) He agreed with Gegenbaur's opinion that the mandibular arch had originated as anterior (front) branchial arches. Under selective pressure to improve ventilation of the gills, the anterior arch enlarged, moved anteriorly, and became more mobile. In 1996, he noted that although lampreys and fossil jawless vertebrates may lack jaws, they have cheeks and lips that enclose a large oral cavity. He proposed that the original mandibular arch had been behind the "old mouth," at the front of the pharynx, and had functioned in ventilation of the gills and sucking in prey. As the jaws became specialized for feeding, they moved forward, crowding the "old mouth" into the small space between the jaws and lips, and creating a "new mouth" - the oral cavity between the jaws and pharynx. The connection between breathing and feeding came from the use of the pharynx in suction feeding. Starting with the arches having had a secondary food manipulation function, they became specialized as primary feeding structures. This tied up one loose end, the presence of labial cartilages in sharks. Still, paleontological evidence of the transition has not been forthcoming.


Janvier, Philippe. 1993. Early Vertebrates
The Velar skeleton: During the 20th century, the combination of genomic and comparative anatomical study of the issue in lampreys and gnathostomes has led to the general acceptance that the homolog of the mandibular arch, in lampreys, is the velum that separates the oral and pharyngeal chambers. We've seen this in Branchiostoma, however it is present in both ammocoetes larvae and adult lampreys. Lampreys preserve this arrangement, but their velum is supported and moved by paired skeletal elements - the velar cartilages. Phillippe Janvier, 1996 proposed that the skeleton of the velum was the precursor to the mandibular arch. Like jaws, the cartilagenous velar elements of lampreys have upper and lower segments articulating at a posteriorly facing hinge. By this hypothesis, the developmental patterning identity of a visceral arch was applied to the velar cartilage to produce the mandibular arch.

The 2023 Synthesis: Everyone is right!

• Recent developmental studies have established that anterior and posterior surfaces of the lamprey velum and anterior and posterior regions of the mandibular arch receive homologous neural crest cells from different regions of the neural tube, and show similar gene expression. Finally, both the velum and mandibular arch are innervated by the posterior branch of the trigeminal nerve. (Yokoyama et al., 2020)
• Establishing this homology allows us to identify homologs of the hyoid arch in lamprey embryos by position and by innervation by the facial nerve, as in gnathostomes.
• Finally, Gai et al., 2022 use Synchrotron X-ray Tomographic Microscopy to identify corresponding structures in the galeaspid Shuyu. In Shuyu, however, the hyoid arch and pouch resemble the branchial arches behind them. Patterns of blood vessels around it indicate that it contained functional gills, as well.

So everyone is right:

• Fossil fish do occur with full gill slits between the mandibular and hyoid arches, just not jawed fish!
• The original of the mandibular arch was at the rear of the original oral cavity
• The velar cartilage was the evolutionary basis for the mandibular arch
• Gai et al., 2022 point out that although no gnathostome has a fully open spiracular gill-slit the acanthodians Cheiracanthus and Mesacanthus appear to have retained both anterior and posterior pseudobranchs in their spiracles. Thus, the evolution of the gnathostome spiracle proceeded in two steps:
  • Reduction of the full gill opening with functioning gills in front and in back to a narrow spiracular passage with paired pseudobranchs
  • Elimination of one of the pseudobranchs.

2. New Branchial Arches:

The lamprey velar skeleton bears on the issue of gnathostome branchial arches, too.

Note:

• Lampreys' branchial arches are external to the gills
• Gnathostomes' are internal to them
• As gnathostomes, the principal arches of chondrichthyans (cartilaginous fish) are internal, however they also have thin extrabranchial cartilages lateral to their gills.

For fun, the lamprey velar skeleton (right) actually has both medial and lateral components.

As with jaws, Mallatt and Janvier square off:

• Mallatt proposes that the ancestors of vertebrates had both internal and external branchial arches. In lampreys the internal ones were lost and in gnathostomes, the external ones were lost.
• Janvier regards the internal gnathostome arches as evolutionary novelties of gnathostomes. However, he considers them to be based on the developmental pattern of the medial elements of the lamprey velar skeleton that has been developmentally shifted back into the branchial chambers.

Ontogeny: Developmentally, lampreys and gnathostomes are working with the same material: The gills form from blocks of endoderm that are surrounded top, bottom, medially, and laterally by a layer of migrating neural crest ectoderm. In gnathostomes, the neural crest cells develop into branchial arches medially because of an inductive relationship with adjacent gut-tube endoderm. In lampreys, neural crest apparently is triggered to form branchial arches by a similar relationship with regular ectoderm. Thus, the evolution of medial gill arches really reflects the evolution of a change in the way in which neural crest is signaled to make cartilage.

But note: Conventional wisdom holds that branchial arches are formed by neural crest and paired fins are formed from mesodermal mesenchyme. Sleight and Gillis, 2020, however, report that in the skate Leucoraja erinacea neural crest and mesenchyme collaborate in the formation of branchial arches, with neural crest predominating anteriorly and mesenchyme posteriorly. How taxonomically widespread this pattern is is a question for the future.

3. Teeth:

Until recently, proper teeth - odontodes lining the oral and pharyngeal cavities employed in feeding - were not recognized in any gnathostomes outside the crown Group of Gnathostomata. Recently, however, they have been recognized in several basal groups, especially members of Acanthothoracidae, enabling us to say:

• They occur in tooth-families in which new teeth are added on the lingual (medial) side as the animal grows.
• Older teeth are not shed in the most basal gnathostomes
• Tooth families tend to be continuous with and grade into rows of dermal odontodes (placoid scales) on the outer surface (Vaškaninová et al., 2020)

This pattern seems to be plesiomorphic for gnathostomes but is greatly modified in more derived groups.

4. The Pelvic Fins:

The evolution of distinct pectoral and pelvic fins is prefigured in the fin-folds of anaspids and thelodonts.

• Lobe-shaped suprabranchial folds above the gill openings. (In osteostraci these have a proper skeleton.)
• Long uniform postbranchial folds stretching from posterior to the gill openings to near the anus.

In some cases, both pairs appear in the same animal. When we next encounter these structures in gnathostomes, both sets are proper fins with internal skeletons, however these differ:

• Pectoral fins originate higher on the body and attach by means of a small number of basal elements.
• Pelvic fins are often still elongate, and their skeleton consists of a long array of parallel fin rays.

In fact, the pelvic fin is often in line with various other structures also derived from the postbranchial fin-fold, such as the series of spines in Kathemacanthus (right. Other examples below). Indeed, the pectoral fin is also part of an array of spines arranged perpendicular to the spines of the postbranchial fin-fold.

5. The Horizontal Semicircular Canal:

A true enigma. We see no intermediate forms. However:

• In mice, mutations of the otx1 gene results in the failure of the canals to form.
• A stunning reversal: Maklad et al., 2014 have shown that two sets of horizontal semicircular canals are present in the lamprey Petromyzon marinus but that they form medially rather than laterally and are not homologous to the gnathostome canals.

6. Braincase Development:

The braincase of adult gnathostomes ossifies in a variety of patterns, however the early development of the cartilages that give rise to it is relatively uniform. The first cartilages to appear give rise to the floor of the braincase or form the special sense capsules. From front to back we see:

• Ethmoid region:
  • Nasal capsules: Enclosing the olfactory organs.
  • Ethmoid plate: (Neural crest ectoderm) Midline structure supporting the nasal capsules.
• Sphenoid region:
  • Trabeculae: (Neural crest ectoderm) Neural crest ectodermal - Paired rods giving rise to the floor of the front of the braincase between the optic capsules
  • Eyes
• Occipito-otic region:
  • Otic capsules: (Mesoderm) Enclosing the membranous labyrinth.
  • Parachordals: (Mesoderm) Medial to otic capsules and lying alongside the notochord.
  • Occipital arch: An upside-down U forming the posterior end of the braincase enclosing the foramen magnum through which the spinal cord exits the braincase.

These define the basic landmarks of the braincase. For example, the hypophysis forms on the midline between parachordals and trabeculae. During growth, these cartilages grow together to form a trough-shaped braincase.

All vertebrates have the special sense capsules (albeit only one nasal capsule for some basal vertebrates) and parachordals. Only gnathostomes definitely have the remainder (maybe galeaspids and osteostracans). The gnathostome braincase encloses the exits of the twelve cranial nerves. The occipital arch is believed to have originated from a short series of arcualia or vertebrae that became incorporated into the rear of the skull. Note that the otic capsules and occipital arches do not fuse well, resulting in a persistent lateral otic fissure.

Gnathostome diversity:

Traditionally, three major groups of unknown relationships were recognized:

• Placodermi - Silurian - Devonian forms with extensive bony armor
• Acanthodii - Silurian - Permian forms with extensive bony armor
• Chondrichthyes - living cartilaginous fish and their fossil relatives
• Osteichthyes - living bony fish and their fossil relatives

Placodermi:

(Silurian - Devonian) Lovely and entertaining armored gnathostomes and poster-children for the Devonian. Experienced a rapid worldwide diversification at the beginning of the Devonian and abrupt decline and extinction at its end.

Identifying features:

• Distinct cranial and thoracic armor with unique ossification pattern. The cranial and thoracic shields often meet at a mobile joint.
• Jaws lined with gnathals, self-sharpening occluding bony plates.
• Teeth? Recent investigations have identified smaller odontodes in the gnathal plates of some placoderms:
  • Acanthothoracans - (basal placoderms) (Rücklin and Donoghue, 2015; Vaškaninová et al., 2020), where odontodes are added during growth to the peripheral margin of broad gnathal plates.
  • Arthrodires - (derived placoderms) (Rücklin et al., 2012), in which the odontodes form a linear array.
  Caveat: Although some placoderms add oral odontodes, none show any sign of shedding them in a system of regular tooth replacement as we see in crown-group gnathostomes.
• Palatoquadrate attaches to the dermal bones of the skull roof, causing adductor muscles (that close the jaw) to pass medial to (inside) palatoquadrate, in stark contrast to other gnathostomes.
• The ethmoid region tends to ossify separately from the remainder of the braincase.
• Compared to other gnathostomes, the otico-occipital region takes up much more of the length of the braincase (a plesiomorphy!).

A Primitive aspect: It's worth noting that although they display the synapomorphies of Gnathostomata, many placoderms retain plesiomorphies reminiscent of their jawless relatives:

• Many have pectoral skeletons sporting corneal processes like those of galeaspids or osteostracans.
• Among Cyclostomi, Galeaspida, and Osteostraci the nasal capsules (or capsule) are crowded in front of a relatively short forebrain, while the upper lip is expansive. Basal placoderms including Acanthothoraci and Antiarchi show similar facial proportions whereas more derived gnathostomes have longer forebrains and nasal capsules that lie directly adjacent to the upper lip. (Dupret et al., 2014).

Placoderms were very diverse and occupied many ecological roles. During the Devonian, they greatly outnumbered other marine fish (right). Their specializations included:

• Shellfish feeding with crushing dental plates.
• Bottom feeding.
• Predation: Some of the largest predatory fish ever, such as Dunkleosteus. Many seem suited to resting on the bottom and engulfing prey by rapidly raising their head and opening jaws. Some localities, however, preserve more streamlined placoderms.
• Deposit feeding.
• Ram suspension-feeding: Titanichthys, although the size of Dunkleosteus has a much more slender jaw. A morphometric analysis by Coatham et al., 2020 indicates that ecologically, it was specialized for suspension-feeding in the manner of a basking shark.
• Some placoderms are known to have practiced internal fertilization (the earliest known instance). Long suspected because some individuals had structures looking like "claspers" - the intromittant organs of chondrichthyans.
  • In 2009, Long et al. described the fossil of a pregnant placoderm confirmed this.

  
  
  
  Placoderm claspers after Trinajstic et al.2015
  During the 20th century, this led to confusion about the identity of clasper-bearing fossils - placoderm or chondrichthyan?
• Trinajstic et al.2015 showed that truth is more interesting than fiction by describing in detail the intromittant organs of placoderms. Whereas:
  • the claspers of sharks are definitely enhancements of their pelvic fins,
  • those of placoderms are separate and distinct paired structures behind the pelvic fin.
  Yet another interesting feature arguably derived from the ancestral postbranchial fin-fold.

Fossil record: Fragmentary record begins in the Early Silurian, followed by a rapid diversification. During the Devonian, placoderms were the dominant vertebrate group. Both marine and fresh water forms are recorded with a worldwide distribution except for absence in South America, which hosted south polar glaciers. Placoderm diversity was greatly reduced by the Late Devonian Frasnian extinction event, and they were completely extinguished by the Fammenian event at the end of the Devonian. Thus, entire radiation took up only about 77 million years, but while it lasted, it was spectacular. But maybe -

Placoderm diversity:

Placoderms look different and distinct, and for over a century, were assumed to be monophyletic, so it came as a surprise when Brazeau, 2009 and others began recovering them as paraphyletic. Today we regard them as basal gnathostomes. Recent phylogenies differ in detail but generally agree on which placoderm groups are closer to the base of Gnathostomata and which are closer to the crown. (Though King et al., 2016 find paraphyletic and monophyletic Placodermi to be equally parsimonious.) A sampling of some placoderm high-points:

Xiushanosteus mirabilis: (Early Silurian) The oldest known placoderm. Phylogenetically basal, but not the most basal. Abundant where present, this creature is similar to acanthothoracid placoderms, but is tiny.. Note 5 mm scale bar in image at right!

Stensioella: (Early Devonian) Among the most basal gnathostomes. Although poorly known, gnathostome synapomorphies including paired nares and pelvic fins are apparent. Features:

• The head and thoracic shields are seen to be an array of individual tesserae that have not firmly coalesced into characteristic placoderm head plates.
• Pectoral fins are large.
• Tail is long and whip-like.

Janvier has suggested that Stensioella is actually a chondrichthyan close to ratfish - a minority opinion.

Antiarchi: (Silurian - Devonian) Diverse and successful weirdos (second only to Arthrodira in diversity) common both in sea and fresh-water environments. Key features:

• Extensive thoracic armor forming a box from which the tail (scaled or naked) protruded. The thoracic armor regularly contains elements including a posterior medial dorsal that other placoderms lack.
• Cranial armor articulates with thoracic by a condyle (knob) that fits into a thoracic socket.
• Orbits, nares, and pineal foramen are all confined to a small orbital fenestra in the top of the head.
• Strangest of all, the pectoral fins are armored with dermal bone plates. In derived taxa, these have an articulated joint at mid-length. Older and more basal antiarchs like Parayunnanolepis have shorter unjointed pectoral fins.
• Anal fin absent.
• The presence or absence of pelvic fins in antiarchs has long been debated. Zhu et al.'s 2012 examination of Parayunnanolepis indicate that pelvic fins, attached to a pelvic girdle with both dermal and endochondral elements are definitely present. Note: Zhu et al., in press cast some doubt on the presence of pelvic endochondral elements in this taxon.

Plesiomorphies: While definitely basal gnathostomes, Antiarchi show suggestive similarities to Osteostraci:

• The clustering of orbits, openings to the nasal capsule(s) and pineal foramen on top of the head is similar in both groups. (See osteostracan)
• Patterns of vascularization (blood supply) and nerves in the thoracic armor where these articulate with pectoral fins are very similar to those of osteostraci.

Have antiarchi convergently evolved these traits or simply inherited them?

Rhenanida: (Devonian) Secondarily flattened bottom dwellers with large pectoral fins. We have seen the "bottom-dweller" ecomorph before and will see it again. Rhenanids are the first gnathostomes to take up that life style through the enlargement of their pectoral fins. Features:

• Enlarged pectoral fins
• Reduction of bony plates to small tesserae (obscuring the homologies of all but a few.)
• Jaw plates flattened and covered with small denticles

Brazeau, 2009 recovered Rhenanida as the sister-taxon of the next group.

Ptyctodontida: (Devonian) Strongly similar to living ratfish. Features:

• Significantly reduced cranial and thoracic armor, in which the bone was mostly covered with skin.
• Tooth plates have a spongy microscopic texture
• These are the placoderms in which we see both the sexual dimorphism of the pelvic region mentioned above and evidence for viviparity.

Arthrodira: (Devonian) The most diverse and speciose placoderm group, including both deposit feeders and predators, including giant predators like Dunkleosteus. Features:

• Mobile joint between socket on cranial armor and condyle on thoracic.
• A large nuchal gap between the two shields allows the head to be elevated to increase the gape.

Potential synapomorphy of Arthrodira and Osteichthyes:

• The parasphenoid - a dermal element of the palate that forms beneath the sphenoid region of the neurocranium. (Which other placoderm-grade dermal bones are homologous with those of Ostiechthyes is an area of active research.)

Entelognathus primordialis: (Silurian) Zhu et al., 2013. described this unremarkable creature with a extensive thoracic shield. Entelognathus looks like an arthrodire without the hinged articulation of head and thoracic armor or nuchal gap.

What makes Entelognathus interesting is its possession of potential synapomorphies with Osteichthyes:

• Three dermal bones of the face:
  • Maxilla
  • Premaxilla
  • Dentary
• Much of the adductor muscle of the jaw is lateral to the palatoquadrate.

And introducing...:

• Qilinyu rostrata, (Silurian) a second "maxillate" placoderm-like fish. Based on Qilinyu, Zhu et al conclude that the premaxilla, maxilla, and dentary are homologous to placoderm gnathal plates. This identification facilitates the further identification of the jugal and lacrimal in typical placoderms. (Zhu et al, 2016)
• Bianchengichthys micros, (Silurian) a third. In this one, the premaxilla, maxilla, and dentary retain arthrodire-like teeth. (Li et al., 2021)

Entelognathus and Qilinyu, with their extensive armor, is not what we expect of a close relative of living gnathostomes, whose ancestral body covering, we long presumed, consisted of small scales and plates. We take them up next.