All animals can be classified as either vertebrate (those that have a spine) or invertebrate (those that do not). About 98% of all living vertebrate species belong to a group called Osteichthyes, otherwise known as bony fish. Despite the name, this group also includes all four-limbed vertebrates – amphibians, reptiles, birds and mammals – since they evolved from prehistoric bony fish millions of years ago.

The oldest known bony fish can be traced back to around 425 million years. These ancient bony fish are all part of a sub-group called lobe-finned fish. Most modern bony fish, however, are part of a different sub-group called ray-finned fish, which can only be confidently traced back about 390 million years. A species called Ligulalepis was once thought to represent the oldest ray-finned fish. Scientists worked this out by examining a single Ligulalepis skull fossil from around 400 million years ago. However, subsequent studies have disputed its position in the evolutionary tree. So, the early evolution of bony fish remains poorly understood.

To address this, Clement, King, Giles et al. re-examined the original Ligulalepis skull fossil, alongside a newly discovered second skull fossil of the same species. Modern x-ray scanning techniques were used to produce detailed 3D models of both skulls and compare them to other prehistoric bony fish. This allowed Clement, King, Giles et al. to find Ligulalepis’s exact place in the evolutionary family tree.

The experiments identified many previously unknown features of the Ligulalepis skull. These features suggest that this species was not a ray-finned fish; rather, it existed just before bony fish split into two sub-groups (lobe-finned and ray-finned). The analysis also suggests that Ligulalepis was the species most closely related to another group of fish called psarolepids. Overall, these findings clarify our understanding of the evolutionary tree of all vertebrates, including humans. Future research should continue using modern scanning techniques to uncover new information from old fossils and give further insights into the early evolution of vertebrates.

Some 98% of living vertebrate species belong to Osteichthyes, the ‘bony fishes’. Osteichthyans are distinct from living jawless fishes (lampreys and hagfishes) and Chondrichthyes (the ‘cartilaginous fishes’: sharks, rays and chimaeras), and comprise Actinopterygii, or ‘ray-finned’ fishes (such as teleosts, bichirs and gars), and Sarcopterygii, or ‘lobe-finned’ fishes (lungfishes, coelacanths and tetrapods). The earliest osteichthyans are late Silurian (~425 million years ago), and from first appearance included large forms sometimes considered to be sarcopterygians (Choo et al., 2014; Zhu et al., 2009), but the interrelationships of the earliest members remain contentious (Friedman and Brazeau, 2010).

Despite a lack of clarity with respect to relationships, a number of significant discoveries in recent years has resulted in a clearer view of early osteichthyan anatomy. In particular, taxa such as Youngolepis (Chang, 1982), Diabolepis (Chang and Yu, 1984), Kenichthys (Chang and Zhu, 1993), Psarolepis (Yu, 1998), Achoania (Zhu et al., 2001), Styloichthys (Zhu and Yu, 2002), Guiyu (Zhu et al., 2009), Tungsenia (Lu et al., 2012a2012), Megamastax (Choo et al., 2014), Sparalepis (Choo et al., 2017), and Ptyctolepis (Lu et al., 2017) have provided new morphological information regarding the pattern of character acquisition in osteichthyans, in particular sarcopterygians. In contrast, the Silurian–Early Devonian record of actinopterygian evolution is poorly understood, confounded by fewer identified specimens known from typically fragmentary material. Despite the identification of Silurian sarcopterygians (but see Lu et al., 2017) necessitating the presence of contemporaneous actinopterygians (Coates, 2009), the oldest putative actinopterygian is the Lochkovian (~415 Ma) Meemannia (Lu et al., 2016a), with unequivocal ray-finned fishes such as Cheirolepis known only from the Eifelian-Givetian (~393 Ma) and younger deposits. This paucity of specimens may be a reflection of lower abundance and diversity of early actinopterygians compared to sarcopterygians (Cloutier and Arratia, 2004; Friedman and Brazeau, 2010).

Phylogenetic relationships among stem osteichthyans are also poorly resolved. Several taxa have been proposed to branch from the stem, but there is little consensus as to the membership or branching order. Indeed, it has been suggested that ‘stem-group osteichthyans might not be recognized, even when their remains are discovered’ (Friedman and Brazeau, 2010, pg. 38). Dialipina was originally diagnosed as a actinopterygian based on scale morphology (Schultze, 1968), but more recent analyses have resolved it either as an stem actinopterygian (Giles et al., 2015b; Schultze and Cumbaa, 2001) or stem osteichthyan (Choo et al., 2017; Friedman and Brazeau, 2010; Giles et al., 2015c; Lu et al., 2016a; Qiao et al., 2016). Taxa referred to as ‘psarolepids’ (sensu Choo et al., 2017) were originally placed as sarcopterygians: Psarolepis was initially described as a ‘porolepiform-like’ crown sarcopterygian (Yu, 1998) or either a stem osteichthyan or stem sarcopterygian (Zhu et al., 1999), with most subsequent analyses corroborating a stem sarcopterygian position (Brazeau, 2009; Choo et al., 2017; Lu et al., 2016a; Qiao et al., 2016; Zhu et al., 2001; Zhu et al., 2009); and Guiyu was deemed a stem sarcopterygian when first described (Zhu et al., 2009), a position subsequently supported in other analyses (Choo et al., 2017; Lu et al., 2016a). More recent analyses have recovered a stem osteichthyan position for ‘psarolepids’ (Lu et al., 2017: supported under parsimony, but not Bayesian, analyses), corroborating previous suggestions by Zhu et al. (1999) and Choo et al. (2017), as well as evidence from palaeohistological data (Qu et al., 2015). A Bayesian tip-dating approach provides no resolution regarding the phylogenetic position of Guiyu, Achoania and Psarolepis (King et al., 2017). The uncertainty regarding early osteichthyan relationships may be related to the dual problems of missing palaeontological data and difficulty in polarising osteichthyan characters, exacerbated by the discovery of osteichthyan-like anatomy in stem gnathostomes (e.g. Janusiscus: Giles et al., 2015c; Entelognathus: Zhu et al., 2013).

One taxon that may help to elucidate osteichthyan stem group phylogeny is the enigmatic Ligulalepis. Ligulalepis toombsi (Schultze, 1968) was erected and attributed to Actinopterygii byon the basis of isolated scales from the Early Devonian (Emsian) Taemas Limestones of the Burrinjuck area of New South Wales, Australia. A second species, Ligulalepis yunnanensis Wang and Dong, 1989, was erected on the basis of isolated scales from the Silurian (Ludlow) Miaokao Formation of Yunnan, China. Subsequently, other occurrences of isolated scales from Australia were attributed to the genus (Burrow, 1994; Burrow, 1997), including throughout the Bloomfield Limestone member to the Warroo Limestone member at Burrinjuck (Basden and Young, 2001); scales from at least the latter locality appear to belong to a single taxon (C. Burrow pers. comm to GCY). Schultze (2016) referred a jaw from the Early Devonian Trundle Beds of New South Wales to Ligulalepis, although no justification for this is given by Schultze. Histological sections through the jaw show teeth bearing acrodin, a hypermineralised tissue forming a tooth cap that is currently known only in actinopterygians (Ørvig, 1973).

An incomplete braincase and skull roof, AM-F101607, from the same Burrinjuck limestones, was described by Basden et al. (2000) as perhaps the ‘most primitive osteichthyan braincase’ known, emphasizing its unusual combination of morphological characters. These authors used the phylogenetic analysis of Zhu et al. (1999) to consider alternative placements of AM-F101607 as either a stem gnathostome, stem actinopterygian, or stem osteichthyan, the last option being the most parsimonious. Later, Basden and Young (2001) published a more detailed morphological description of this specimen, which they considered might rather represent a member of the actinopterygian stem. Consequently, the specimen was moved to ‘Ligulalepis’ sp., and tentatively referred to Ligulalepis toombsi, on the basis that this was the only actinopterygian taxon known from the Emsian of southeastern Australia.

In addition to characters considered actinopterygian-like (dermal ornament, skull roof pattern and overall endocranial proportions; Basden and Young, 2001), the skull of ‘Ligulalepis’ displayed characters found scattered across the gnathostome tree. ‘Primitive’ features included the presence of an eye stalk, myodomes for the attachment of oculomotor-innervated eye muscles, and an opening for the orbital artery (Basden et al., 2000). However, certain features were noted to bear resemblance to sarcopterygians, including the proportions of the (short and broad) telencephalic region, ‘Psarolepis-like’ pit lines on the skull roof, shape of the basisphenoid and the shallow depth of the oticoccipital area (Basden and Young, 2001). The position of the hyomandibular attachment along the anteroposterior axis of the otic capsule was considered intermediate between the posterior placement in chondrichthyans and the anterior placement in osteichthyans, similar to that in Acanthodes (Basden and Young, 2001; Basden et al., 2000; Brazeau and de Winter, 2015; Davis et al., 2012).

Subsequent phylogenetic analyses have recovered contrasting placements for ‘Ligulalepis’. Zhu et al. (2001), in describing the primitive sarcopterygian Achoania from the Early Devonian of China, resolved ‘Ligulalepis’ as a basally-branching actinopterygian. Friedman, 2007 (pg. 311) determined ‘Ligulalepis’ to be a stem osteichthyan, arguing that the ‘actinopterygian affinities of Ligulalepis … have relied upon characters of uncertain polarity.’ Following the discovery of Guiyu from the late Silurian of China, ‘Ligulalepis’ was recovered as a stem sarcopterygian (Zhu et al., 2009), whilst Brazeau (2009), resolved ‘Ligulalepis’ as a stem osteichthyan. Friedman and Brazeau (2010) examined the early osteichthyan radiation and presented a more detailed argument for stem osteichthyan affinity of ‘Ligulalepis’. Since then, ‘Ligulalepis’ has been recovered as a stem osteichthyan in most analyses (Davis et al., 2012; Giles et al., 2015b; Giles et al., 2015c; Giles et al., 2017; Long et al., 2015; Lu et al., 2017; Zhu et al., 2016; Zhu et al., 2013). However, tip dated analyses employed by King et al., 2017 place ‘Ligulalepis’ within Actinopterygii with quite strong support.

Uncertainty surrounding the phylogenetic placement of ‘Ligulalepis’ clearly warrants further investigation. Here, we use micro-CT scanning to reinvestigate the anatomy of the original cranium (AM-F101607), augmented by description of a second, recently-discovered specimen (ANU V3628). The main goals of this work are: (1) to provide a revised account of the anatomy of ‘Ligulalepis’, including the previously unknown anterior region of the skull roof (preserved in ANU V3628); (2) to test the anatomical interpretations produced on the basis of external investigation only (Basden and Young, 2001; Basden et al., 2000); (3) to examine the effect, if any, of this new anatomical data on the phylogenetic position of ‘Ligulalepis’ and understanding of early osteichthyan evolution, based on a revised version of a recent phylogenetic analysis (Lu et al., 2017); and (4) to investigate the implications of our phylogenetic placement of AM-F101607 and ANU V3628 for the taxonomic referral of the skulls to Ligulalepis.

Skull roof

Scans of AM-F101607 reveal for the first time some of the sutures between the skull roofing bones (Figure 1) showing a pattern different in important respects to the previous interpretation (Basden and Young, 2001). Viewing the scan in Drishti reveals a set of parallel bands tracing what we assume to be bone sutures in the posterior part of the skull (Figure 1B). Closer inspection of the scan data reveals these bands to be high-density thickenings in the basal layer of the dermal skull roof bones (Figure 1—figure supplement 1). These show the outline of the postparietals (of sarcopterygians; parietals of actinopterygians) and the posterior edges of the parietals (of sarcopterygians; frontals of actinopterygians) (Figure 1B,C). No midline suture is evident between the postparietals, but a very faint suture between the parietals is suggested. The lateral margin of the postparietal is scalloped in such a way as to provide contact faces for a series of three bones, with faint lines visible demarcating them. The most posterior bone presumably corresponds to the tabular (of sarcopterygians; supratemporal of actinopterygians). Anterior to this is a supratemporal (of sarcopterygians; intertemporal of actinopterygians), and a broad and elongate intertemporal (of sarcopterygians; dermosphenotic of actinopterygians) borders the orbit. Unfortunately, sutures cannot be visualized in the same way in ANU V3628, despite the higher scan resolution, because no high density growth bands are evident. Instead the basal layer of the skull roof dermal bone is of uniform density and thickness. This suggests that they may vary between individuals or growth phases. Further specimens of ‘Ligulalepis’ are required to unambiguously determine the pattern of skull roof bones in this taxon. The presence of middle and posterior pitlines, and the supraorbital canals extending to the posterior edge of the postparietals, is confirmed in ANU V3628.

Figure 1 with 1 supplement see all

Download asset Open asset

ANU V3628 preserves the previously unknown anterior portion of the skull roof (Figure 2). A pineal foramen is preserved, but due to a crack in the specimen it is unclear if a separate pineal plate was present. Sutures in the anterior part of the skull are unclear. The pattern of ornamentation anterior to the pineal opening is suggestive of a median ossification (i.e. a median rostral), but this is not evident from the CT data.

Figure 2

Download asset Open asset

The profile of the snout has a sharply downturned anterior face (Figure 3), as is general for gnathostomes (Gardiner, 1984; Long, 1988; Zhu et al., 2013; Zhu et al., 2009). There is an abrupt change in ornamentation on the snout, from short anteriorly directed ridges to elongate transverse ridges (Figure 2C). A similar pattern is known in Dialipina (Schultze and Cumbaa, 2001).

Figure 3

Download asset Open asset

The incurrent nostrils are large and widely separated from the excurrent nostrils, which appear to lie entirely within the orbits (Figure 3). Basden and Young (2001) also assumed communication of the posterior nostril with the orbit, including a notch for the nostril on the anterior margin of the orbital fenestra. Neither specimen of ‘Ligulalepis’ show evidence for such a notch, although the ventral part of the nostril and orbital margin are unknown. A nostril confluent with the orbit is typically considered an actinopterygian character, but without preservation of the premaxilla and cheek bones in ‘Ligulalepis’ we cannot rule out the possibility that dermal bone separated the external opening of the nostril from the orbit – for example a postero-dorsal process of the premaxilla as in Psarolepis (Yu, 1998), and perhaps Cheirolepis (Gardiner, 1984, Fig. 49). However, in ‘Ligulalepis’ the opening for the posterior nostril in the endocranium lies directly within the orbit (Figure 3A). This is in contrast to the situation in both actinopterygians and sarcopterygians, where an endoskeletal lamina (the postnasal wall) separates the nostril and the orbit (e.g. Gardiner, 1984, Fig. 13). ‘Ligulalepis’ lacks such a lamina, and in this respect more closely resembles some placoderms such as Parabuchanosteus (Young, 1979) and Dicksonosteus (Goujet, 1984).

Braincase

The preservation of the braincase is similar in both AM-F101607 and ANU V3628. It is mostly well ossified, and comprises the basisphenoid, orbitotemporal and otic regions. The ethmoid region is preserved on the left side of AM-F101607 and is more complete, but less well ossified, in ANU V3628. The posterior and ventral portion of the braincase, comprising the occiput and basioccipital, is absent in both specimens. The loss of this region, which is rarely preserved in early osteichthyans (e.g. Yu, 1998; Zhu et al., 2001; Lu et al., 2012a2012), presumably corresponds to the presence of well-developed otoccipital and ventral otic fissures, possibly in conjunction with a vestibular fontanelle. The orbital region is large, comprising nearly half of total braincase length, and the ethmoid region short.

Ethmoid region

The ethmoid region is very short, and is moderately well ossified. It is separated from the orbitotemporal region by a poorly developed postnasal wall. A canal leaves the cranial cavity at the left lateral limit of the pineal opening and extends posterolaterally to open into the orbit (Figures 4C–E and 5, acv). This opening was identified by Basden and Young (2001) as for the trochlear nerve (n.IV), but its anterior position suggests it may have housed the anterior cerebral vein. This canal is present on only the left side, as in some sarcopterygians such as Latimeria (Robineau, 1975), and various early actinopterygians such as Mimipiscis (Giles and Friedman, 2014) and Kansasiella (Poplin, 1974). Anterior to this, a ramifying network of canals (identified previously as for the anterior cerebral vein; Basden and Young, 2001: fig. 1) may have transmitted branches of the profundus nerve from the orbit to the skull roof, but their course is incomplete (Figure 5A–C, br.prof).

Figure 4

Download asset Open asset

Figure 5

Download asset Open asset

Basden and Young (2001) identified a number of foramina in the dorsal wall of the orbit as branches of the superficial ophthalmic nerve. However, the main trunk of the superficial ophthalmic nerve does not enter the orbit. It remains within the neurocranium, passing below the supraorbital sensory line (Figures 5 and 6, soph). Thus, the foramina in the orbit more likely carried branches of the profundus nerve to the skull roof (Figures 3–6, br.prof). The internal course of the superficial ophthalmic nerve may be related to the relatively wide interorbital septum in ‘Ligulalepis’.

Figure 6

Download asset Open asset

Below the large opening for the olfactory canal, the posterior face of the nasal capsule is pierced by six foramina in three groups. The two dorsal-most foramina enter the nasal capsule from the orbit, and most likely transmitted branches of the profundus nerve (Figures 4–6, prof). The most ventral three foramina also extends from the orbit, and may have carried branches of the maxillary and buccal nerves (Figures 4–6, br.max). As noted by Basden and Young (2001), the remaining canal originates in the forebrain, but its purpose is unclear. CT scans show that the apparent foramen at the anterior extent of the basisphenoid (Basden and Young, 2001: fig.7, ?fica) is in fact blind.

Figure 7

Download asset Open asset

Orbitotemporal region

The orbitotemporal region is extensive, forming the widest part of the braincase and comprising nearly half the total length. A large opening in the orbital wall of AM-F101607 represents the eyestalk attachment area, and was recognized as such due to its everted rims (Basden et al., 2000). The scan of ANU V3628 reveals delicate endochondral bone laminae filling the eyestalk attachment area (Figure 3A, eys), forming a rough surface as seen on other articular surfaces in many osteichthyans. This ossification may have been present in AM-F101607 but lost during preparation (the area is protected by resin in ANU V3628), or ANU V3628 could reflect a more advanced stage of ossification.

The oculomotor (III) and profundus (prof) nerves, as well as the entry of the jugular canal (jug.c) into the orbit, were correctly identified by Basden and Young ([2001]: fig. 2), although there is no communication between the profundus nerve and the canal described by Basden and Young (2001) as housing the orbital artery. The position of the pituitary vein and ophthalmic artery foramina can also be confirmed (Figures 4C and 5A,C,D, pv, opha).

The pituitary vein is continuous between the orbits, and is connected to the hypophysial chamber by a median hypophysial vein (Figure 5D, hyp.v). This condition is similar to the transverse pituitary vein in placoderms, for example Brindabellaspis (Young, 1980), Parabuchanosteus (Young, 1979), Jagorina (Stensiö, 1969) and probably Romundina (Dupret et al., 2017; Goujet and Young, 2004). Petalichthyid placoderms however lack this character (Castiello and Brazeau, 2018). In some early diverging arthrodire placoderms (e.g. Kujdanowiaspis, Dicksonosteus), the pituitary vein is continuous, but exits the floor of the braincase via the subpituitary fossa. There is no foramen in the hypophysial fossa identified that could have carried a median hypophysial vein in these taxa (Goujet, 1984; Stensiö, 1963b). The condition in ‘Ligulalepis’ contrasts with other crown gnathostomes in which the pituitary vein enters the hypophysial chamber directly (e.g. Maisey, 2005; Maisey, 2007; Holland, 2014; Giles and Friedman, 2014). A continuous transverse pituitary vein canal may be partly linked to the relative position of the forebrain and the angle of the hypophysial chamber in ‘Ligulalepis’.

The trochlear (IV) nerve enters the orbit dorsal to the eyestalk attachment area, some way posterior to the dorsal myodome (my.IV) and anterior cerebral vein (Figures 3 and 4C,D). The canal originally identified for the trochlear (IV) nerve (Basden and Young, 2001) in fact houses the anterior cerebral vein. This revised position of the trochlear (IV) nerve (i.e. posterodorsal to the eyestalk) reflects the general gnathostome condition (Chang, 1982; Gardiner, 1984; Maisey, 2005; Young, 1979), but the position of the myodome anterior to the orbit is more similar to that of osteichthyans.

A large opening on the postorbital process was previously identified (Basden and Young, 2001) as housing the orbital artery, in line with the position of this feature in placoderms. Segmentation of the internal course of this canal shows there is no communication with the profundus nerve canal, contrary to Basden and Young (2001) description. The canal connects with a large opening beneath the cerebellar portion of the cranial cavity, most parsimoniously identified as the root of the trigeminal (V) nerve. As such, the large foramen in the orbit most likely transmitted the mandibular branch of the trigeminal nerve (Figures 4C–E and 5A,C, Vmd). This canal also aligns with a notch in the postorbital process, along which the mandibular nerve would have travelled. This morphology is similar to that seen in chondrichthyans (e.g. Cladodoides: Maisey, 2005; ‘Cobelodus’: Maisey, 2007). Small branches are given off the trigeminal nerve within the braincase. One branch (Figure 6B–C, br.buc.1) enters the posterodorsal part of the orbit at a steep angle and likely carried lateralis fibres to small canals in the roof of the orbit that lead to the dorsal part of the infraorbital canal. A second branch (Figure 6B–C, br.buc.2), previously suggested as carrying the posterior branch of the oculomotor (III) nerve (Basden and Young, 2001), opens onto the postorbital process just dorsal to the opening for the mandibular branch. This may also have carried lateralis fibres to the infraorbital canal.

Posterior to the root of the trigeminal nerve, a canal (r.lat) leaves the anterior face of the utricular region and enters the "trigemino-facialis chamber" (Figure 5B). This is interpreted as the root of the anterior lateral line nerves, in a similar position as in other early osteichthyans (Jarvik, 1980; Chang, 1982; Giles and Friedman, 2014). An additional canal (mcv) exits the cranial cavity from the midpoint of the cerebellum and enters the "trigemino-facialis chamber" at a steep angle (Figure 5C). Due to its position and orientation, this is interpretated as the middle cerebral vein. The jugular canal communicates with the "trigemino-facialis chamber" via an opening in the roof of the canal (Figure 6B, com.V.jug), through which the middle cerebral vein and the maxillary branch of the trigeminal nerve may have been transmitted (Basden and Young, 2001).

The identity of the large foramen in the dorsal portion of the anterior pocket (Figure 4C, ap.f) is hard to discern. Segmentation reveals a cavity that is continuous ventrally and dorsally with the infraorbital canal, and may be related to the spiracle. The cavity is also connected with the otic nerve anteriorly. The openings identified by Basden and Young, ([2001]: fig 2) ventral to this foramen do not continue within the bone.

Further clarifications can be made to the identity of the large foramina on the lateral and ventral face of the otic region (Figure 4). The canal ventral to the hyomandibular facet intersects the ventral portion of the jugular canal and can be traced to the ventral otic fissure. It can be confirmed as the hyomandibular trunk of the facial nerve (VIIhm; Basden and Young, 2001: figs. 2,3).

Young, 1979Dupret et al., 2017Maisey, 2005Maisey, 2007Maisey, 2005Maisey, 2007

The foramen identified by Basden and Young ([2001]: figs 2,3) as for the glossopharyngeal nerve is in fact the posterior exit of the jugular canal (Figure 4B,F, jug.c); the glossopharyngeal nerve presumably exited through the otic-occipital fissure. Giles et al., 2015c

Ventral surface

As outlined by Basden and Young (2001), the internal carotids enter the braincase through two foramina flanking the median hypophysial opening before giving off the efferent pseudobranchial and ophthalmic artery (Figure 4B, epsb, Figure 4C, opha). As in chondrichthyans (Maisey, 2005; Maisey, 2007), but unlike in osteichthyans (Chang, 1982; Gardiner, 1984) and placoderms (Hu et al., 2017; Young, 1980), there is no evidence of a parabasal canal carrying the palatine artery anterior to this point. Basden and Young (2001) identified grooves on the ventral surface of the basisphenoid as for the lateral dorsal aorta. However, since these grooves are anterior to the efferent hyoid artery we prefer to refer to them as the internal carotid arteries (Figure 4B, ica). Although (Basden and Young [2001] : fig. 3) identified foramina for the palatine branch of the facial nerve and the orbital artery in the roof of the canal for the internal carotid (their lateral dorsal aorta), the roof appears to be complete.

Cranial endocast

A comparison of the two cranial endocasts is shown in Figure 7. Differences in appearance largely relate to the presence of extensive rock matrix surrounding ANU V3628, in contrast to the acid-prepared cranium of AM-F101607. The external walls of the endocranial cavity are largely complete in both specimens, although as the parachordal plate of the braincase is not preserved the ventral extent is uncertain. Overall, the endocast of ‘Ligulalepis’ is short and broad, particularly the otic region (Figure 7A,B). The proportions occupied by different regions are similar to early chondrichthyans, with the forebrain section comprising less than 20% of the total length, the midbrain section around 15%, and the hindbrain section some 65%.

Description of the endocast allows the identity of features within the cranial cavity to be revised. A distinct depression in the roof of the cranial cavity, medial to the otic capsule, was considered by (Zhu et al. ([2010], fig. 4c) to be evidence of a lateral cranial canal. This embayment is in fact the crus commune of the anterior and posterior semicircular canal (Figure 4B). The groove anterior to this is somewhat shallower in the braincase and indicates where the roof of the utricular region joins the rest of the cranial cavity (the groove for anterior and posterior semicircular canals of Basden and Young 2001: fig. 3).

Forebrain

The region of the endocast corresponding to the forebrain comprises space for the olfactory bulbs, telencephalon and diencephalon. This region in ‘Ligulalepis’ is relatively wide (Figure 7A,B), comparable to the forebrain in placoderms such as Macropetalichthys (Stensiö, 1963a) and chondrichthyans such as Orthacanthus (Schaeffer, 1981). However, it is still only half the width of the cerebellum. The short, wide olfactory tracts leave the anterolateral corners of the telencephalic region in separate tracts and connect to the bulbous nasal capsules, preserved in ANU V3628 (Figure 7B, n.cap). The short olfactory tracts are similar to those of placoderms, for example Buchanosteus (Young, 1979) and Kujdanowiaspis (Stensiö, 1963a), as well as chondrichthyans such as Cladodoides (Maisey, 2005) and Orthacanthus (Schaeffer, 1981), but also some sarcopterygians such as Tungsenia (Lu et al., 2012a2012) and Qingmenodus (Lu et al., 2016). A small canal for the terminal nerve (0) exits from the anterior face of the forebrain, between the olfactory tracts, in AM-F101607.

The telencephalon is the widest and highest portion of the forebrain. At its widest point it measures 6 mm in AM-F101607. It is developed into slight lobes dorsolaterally; these may represent olfactory bulbs (Figure 7A, olf.b). The dorsal roof of this region is preserved in ANU V3628, as is the canal to the pineal opening (Figure 7B,D, pin). The oblique crack across ANU V3628 intersects the pineal opening, so it is unclear whether or not a parapineal organ was present. The margin between the telencephalic and diencephalic regions is marked by a gentle constriction in the endocranial cavity.

The region of the endocast corresponding to the diencephalon is short and narrow in dorsal view. Ventrally, the diencephalic region extends to the floor of the cranial cavity, being continuous with the hypophyseal fossa (hyp), and projects posteriorly some way under the mesencephalon (Figure 7C). This region is unfinished posteriorly, and it is unclear whether a saccus vasculosus was present as in actinopterygians (Giles and Friedman, 2014). The lateral wall of the diencephalic region of the endocast is unfinished for the eyestalk attachment area in AM-F101607 (Figure 4C,D; this area is not preserved in ANU V3628). The optic nerves (II) enter the orbit through a large foramen at the anterolateral limit of the diencephalon (Figures 6A and 7C). Beneath this opening, a vertical ridge on the side of the hypophysial chamber likely shows the course of the internal carotid artery after it enters the braincase. The efferent pseudobranchial artery joins the internal carotid at the point of entry into the braincase (Figure 4B, epsb, f.ica), and internally the ophthalmic artery branches from the same point and enters the orbit (Figure 5D, opha). The hypophysis is oriented ventrally, in agreement with the generalised osteichthyan condition (e.g. Youngolepis: Chang, 1982; Mimipiscis: Giles and Friedman, 2014), but unlike the posterodorsally-oriented hypophysis seen in Cladodoides (Maisey, 2005).

Midbrain

Posterior to the diencephalic portion of the endocast, the region corresponding to the midbrain (mesencephalon) widens slightly in AM-F101607. The midbrain cavity is not differentiated into separate recesses for each optic lobe (Figure 7A, opt.l), which appears to be the general gnathostome condition. There are similarly slight bulges in chondrichthyans (e.g. Cladodoides, Xenacanthus), whereas highly distinct optic lobes are seen in actinopterygians crownward of Mimipiscis (Coates, 1999; Giles and Friedman, 2014). A narrow, dorsally positioned canal leaves the cranial cavity and enters the orbit (Figure 5A–C, IV). This foramen was illustrated, but not identified, by Basden and Young [2001]: fig. 2b, the opening posterior to that labeled IV and dorsal to the eyestalk). The position of the canal strongly suggests it housed the trochlear nerve (IV), given a similar placement in crown gnathostomes (Chang, 1982; Giles and Friedman, 2014; Maisey, 2005). More ventrally, the oculomotor (III) nerve leaves the midbrain and enters the orbit (Figure 5A,C); there is no evidence that this nerve bifurcated along its course. The oculomotor nerve does not typically bifurcate in chondrichthyans (e.g. Cladodoides, Maisey, 2005) or sarcopterygians (e.g. Eusthenopteron, Jarvik 1980; Youngolepis, Chang 1982, ), and is variably developed in actinopterygians such as Mimipiscis (Giles and Friedman, 2014) and Lawrenciella (Hamel and Poplin, 2008).

Hindbrain

The hindbrain is the widest portion of the endocast and would have housed the metencephalic and myelencephalic brain regions in life. The cerebellum extends anterior to the labyrinth (Figure 7A,B, cer), as in chondrichthyans (e.g. Cladodoides, Maisey, 2005) and, to a lesser extent, sarcopterygians (e.g. Eusthenopteron, Jarvik 1980). Although the dorsal surface bears a slight suggestion of two lobes, these can hardly be compared to the distinct cerebellar auricles of actinopterygians such as Mimipiscis (Giles and Friedman, 2014). Similarly, there is no obvious protrusion housing the cerebellum corpus.

The profundus nerve (prof) leaves the cranial cavity separately from the trigeminal nerve (V) and enters the orbit (Figure 6A). Northcutt and Bemis, 1993 made a case that the profundus should be considered a separate nerve rather than a branch of the trigeminal, based on developmental evidence and the separation of these nerves in chondrichthyans, early actinopterygians and Latimeria. The cranial nerve configuration seen in ‘Ligulalepis’ adds to a growing body of evidence from fossil endocranial studies that the separation of the trigeminal and profundus nerves is indeed the plesiomorphic state for crown gnathostomes (Chang, 1982; Giles and Friedman, 2014; Maisey, 2005).

Posterior to the cerebellum, the dorsal part of the hindbrain narrows and drops in height, before the endocast broadens again at the midpoint of the labyrinth. The entire dorsal surface of the hindbrain is smooth, and does not rise as high dorsally as the crus commune of the anterior and posterior semicircular canals (Figure 7C, cc). The posterior dorsal fontanelle is trapezoidal in outline (Figure 7A, pdf). A ridge on the dorsal surface at the lateral edge of the hindbrain may indicate the path of the endolymphatic ducts within the cranial cavity into the posterior dorsal fontanelle (Figure 4B, g.dend).

ANU V3628 appears to have a lateral cranial canal (Figure 7B,D, lcc?), as in actinopterygians (Giles et al., 2018). Basden and Young ([2001]: Fig. 3) identified a groove for the posterior cerebral vein in AM-F101607, in a corresponding position to a similar groove in Mimipiscis and Moythomasia (Gardiner, 1984). In ANU V3628, the dorsal part of this groove contains a large foramen, most clearly developed on the left side (Figure 8C,D), in the same position to the opening for the lateral cranial canal in Moythomasia (Gardiner, 1984, fig. 27) and Mimipiscis (Gardiner, 1984, fig. 11). Segmentation reveals that this foramen opens into a large unossified space (Figure 8E), as expected for a lateral cranial canal (Gardiner, 1984; Patterson, 1975), although this cavity does not have a continuous perichondral lining.

Figure 8

Download asset Open asset

However, the situation regarding a lateral cranial canal in AM-F101607 is less clear. Basden and Young (2001) identified foramina in the posterior cerebral vein groove, and identified them as anterior tributaries of the posterior cerebral vein. Although the foramina on the left hand side are indeed small (Figure 8B), on the right hand side there is a larger, more distinct foramen in the same position (Figure 8A). However, there is no obvious connection between the unossified space and the dorsal part of the lateral endocranial wall, and furthermore this cavity appears interconnected with much of the remaining interperichondral space in the otic region of the braincase. The lateral cranial canal may have been variable in its development, as has been suggested for Mimipiscis (Gardiner, 1984). It seems likely, however, that the foramina in ANU V3628 are far too large to be identified as tributaries of the posterior cerebral vein. As the occipital portion of the braincase is missing, the posterior extent of the hindbrain cannot be described.

Labyrinth

The labyrinth region in ‘Ligulalepis’ is well preserved in AM-F101607 (Figure 7A,C), with three complete, slender semicircular canals present, and all carrying small expansions for ampullae.

The anterior semicircular canal (asc) is anteroposteriorly long, but does not extend far ventrally. In contrast, the posterior semicircular canal (psc) is tall dorsoventrally, but anteroposteriorly very short. A short portion of preampullary canal separates the posterior ampulla from the cranial cavity. The posterior semicircular canal curves back underneath the external semicircular canal (esc) to meet the cranial cavity far ventrally. This ventral position of the posterior canal is reminiscent of that in placoderms (e.g. Dicksonosteus; Goujet, 1984), chondrichthyans (Schaeffer, 1981; Maisey, 2007), early sarcopterygians (e.g. Youngolepis; Chang, 1982) and, to a slightly lesser extent, in the early actinopterygian Mimipiscis (Giles and Friedman, 2014).

Strikingly, the external canal is positioned obliquely at an angle of about 30 degrees from the cranial cavity, and completes nearly a full circle before re-entering the vestibule (Figure 7A,C). The posterior connection with the cranial cavity is swollen, almost giving the appearance of an ampulla like that at the anterior extent of the canal (ext.amp).

Other notable features of the vestibular system are the relatively shallow superior sinus (Figure 7C, s.su) situated below the crus commune, seen elsewhere in Cladodoides, Youngolepis and Kansasiella (Chang, 1982; Maisey, 2005; Poplin, 1974), but not in Mimipiscis (Giles and Friedman, 2014) or Acanthodes (Davis et al., 2012). As well as the crus commune, a portion of the sinus superior, anterior and posterior semicircular canals project dorsally above the endocranial roof. The same condition is found in chondrichthyans and early actinopterygians (Giles and Friedman, 2014).

Although incompletely known ventrally, the sacculus is not laterally extensive and appears to have been shallow (Figure 7C, sac). The general morphology of the labyrinth, including the dorsoventrally extensive posterior canal, which projects above the endocranial roof as well as below the cerebellar floor, and the inclined external canal, recalls that of an early chondrichthyan such as Cladodoides (Maisey, 2005) or perhaps even Acanthodes (Davis et al., 2012). The labyrinth is far removed from that seen in Mimipiscis (Giles and Friedman, 2014), or sarcopterygians such as Eusthenopteron and Devonian lungfishes (Clement and Ahlberg, 2014; Clement et al., 2016; Jarvik, 1980). No otoliths were recovered from the specimen, although their absence is most likely due to either a failure to be preserved or a consequence of acid preparation.

A life reconstruction of ‘Ligulalepis’ based on the skull morphology of AM-F101607 and ANU V3628 (other features remain hypothetical) is presented in Figure 9.

Figure 9

Download asset Open asset

Phylogenetic analysis

AM-F101607 and ANU V3628 were coded into an updated phylogenetic analysis modified from Lu et al., 2017. As well as changes to anatomical scores for ‘Ligulalepis’, codes for several taxa were updated and some characters were deleted or reformulated to give a total of 282 characters coded for 94 taxa (for full details see the ‘phylogenetic methods’ section). This dataset was analysed using both parsimony and Bayesian inference. The parsimony analysis retrieves Dialipina, ‘Ligulalepis’, and ‘psarolepids’ as successively branching sister taxa to the osteichthyan crown node (Figure 10A). However, support for the clade that comprises crown osteichthyans (as retrieved from this analysis) is low, with Bremer support of 1 and a bootstrap of just 4. This is very weak support, although we note that bootstrap values obtained from TNT are likely to be much more conservative than those produced by PAUP*: bootstrap values in TNT are calculated from the strict consensus trees found in each replicate (Goloboff et al., 2008), whereas PAUP* uses all the shortest trees from each replicate, weighted by the reciprocal of the number of trees found in that replicate (Swofford, 2003).

Figure 10

Download asset Open asset

There are six unambiguous character state changes on the branch leading to crown osteichthyans. These are #78 (enameloid on teeth gained), #110 (shape of parashenoid splint shaped), #116 (olfactory tracts long), #130 (eyestalk absent), #184 (median dorsal plate absent), #211 (dorsal fin spines absent). Of these, only the olfactory tracts and eyestalk are known in 'Ligulalepis'.

Alternative phylogenetic placements under parsimony were tested using two constrained searches, one with ‘Ligulalepis’ constrained within actinopterygians and another with ‘psarolepids’ constrained within sarcopterygians. A stem actinopterygian position for ‘Ligulalepis’ requires a single additional step, and the grouping of ‘Ligulalepis’ and actinopterygians was found in 18% of the bootstrap replicates. Enforcing this topology also resulted in ‘psarolepids’ being resolved as stem sarcopterygians (Figure 10B). A single additional step is required to place ‘psarolepids’ on the sarcopterygian stem, and this grouping is found in 16% of bootstrap replicates. When this grouping is enforced it leads to ‘Ligulalepis’ falling into a polytomy with actinopterygians and sarcopterygians (Figure 10B).

The Bayesian analysis retrieves ‘psarolepids’ on the sarcopterygian stem with moderately strong support (pp = 0.94, Figure 11). ‘Ligulalepis’ is resolved as a stem osteichthyan in the 50% majority rule tree (Figure 11), although the crown osteichthyan clade has weak support (0.61). However, an actinopterygian position for ‘Ligulalepis’ has a posterior probability of 0.22.

Figure 11

Download asset Open asset

1. 1. Basden AM
   2. Young GC
   3. Coates MI
   4. Ritchie A

   (2000) The most primitive osteichthyan braincase?

   Nature 403:185-8.

   https://doi.org/10.1038/35003183

   • PubMed
   • Google Scholar
2. 1. Basden AM
   2. Young GC

   (2001) A primitive actinopterygian neurocranium from the Early Devonian of southeastern Australia

   Journal of Vertebrate Paleontology 21:754–766.

   https://doi.org/10.1671/0272-4634(2001)021[0754:APANFT]2.0.CO;2

   • Google Scholar
3. 1. Bjerring HC

   (1972) Morphological observations on the exoskeletal skull roof of an osteolepiform from the carboniferous of Scotland

   Acta Zoologica 53:73–92.

   https://doi.org/10.1111/j.1463-6395.1972.tb00575.x

   • Google Scholar
4. 1. Brazeau MD
   2. de Winter V

   (2015) The hyoid arch and braincase anatomy of Acanthodes support chondrichthyan affinity of 'Acanthodians'

   Proceedings. Biological Sciences 282:20152210.

   https://doi.org/10.1098/rspb.2015.2210

   • PubMed
   • Google Scholar
5. 1. Brazeau MD

   (2009) The braincase and jaws of a devonian 'acanthodian' and modern gnathostome origins

   Nature 457:305–308.

   https://doi.org/10.1038/nature07436

   • PubMed
   • Google Scholar
6. 1. Bridge TW

   (1878) On the osteology of Polyodon folium

   Philosophical Transactions of the Royal Society of London 169:683–733.

   https://doi.org/10.1098/rstl.1878.0021

   • Google Scholar
7. 1. Burrow CJ

   (1994)

   Form and function in scales of Ligulalepis toombsi Schultze, a palaeoniscoid from the Early Devonian of Australia

   Records of the South Australian Museum 27:175–185.

   • Google Scholar
8. 1. Burrow CJ

   (1997)

   Microvertebrate assemblages from the Lower Devonian (pesavis/sulcatus) of central New South Wales, Australia

   Modern Geology 21:43–77.

   • Google Scholar
9. 1. Castiello M
   2. Brazeau MD

   (2018) Neurocranial anatomy of the petalichthyid placoderm Shearsbyaspis oepiki Young revealed by X-ray computed microtomography

   Palaeontology 61:369–389.

   https://doi.org/10.1111/pala.12345

   • Google Scholar
10. 1. Chang MM
    2. Yu X

    (1984)

    Structure and phylogenetic significance of Diabolichthys speratus gen. et sp. nov., a new dipnoan-like form from the lower devonian of eastern Yunnan, China

    Proceedings of the Linnean Society of New South Wales 107:171–184.

    • Google Scholar
11. 1. Chang MM
    2. Zhu M

    (1993)

    A new middle Devonian osteolepid from Qujing, Yunnan

    Memoirs of the Association of Australasian Palaeontologists 15:183–198.

    • Google Scholar
12. Book

    1. Chang MM

    (1982)

    The Braincase of Youngolepis, a Lower Devonian Crossopterygian From Yunnan, South-Western China PhD. University of Stockholm and Section of Palaeozoology

    Swedish Museum of Natural History.

    • Google Scholar
13. 1. Choo B
    2. Zhu M
    3. Qu Q
    4. Yu X
    5. Jia L
    6. Zhao W

    (2017) A new osteichthyan from the late Silurian of Yunnan, China

    PLoS One 12:e0170929.

    https://doi.org/10.1371/journal.pone.0170929

    • PubMed
    • Google Scholar
14. 1. Choo B
    2. Zhu M
    3. Zhao W
    4. Jia L
    5. Zhu Y

    (2014) The largest Silurian vertebrate and its palaeoecological implications

    Scientific Reports 4:5242.

    https://doi.org/10.1038/srep05242

    • PubMed
    • Google Scholar
15. 1. Clement AM
    2. Ahlberg PE

    (2014) The first virtual cranial endocast of a lungfish (Sarcopterygii: Dipnoi)

    PLoS One 9:e113898.

    https://doi.org/10.1371/journal.pone.0113898

    • PubMed
    • Google Scholar
16. 1. Clement AM
    2. Challands TJ
    3. Long JA
    4. Ahlberg PE

    (2016) The cranial endocast of Dipnorhynchus sussmilchi (Sarcopterygii: Dipnoi) and the interrelationships of stem-group lungfishes

    PeerJ 4:e2539.

    https://doi.org/10.7717/peerj.2539

    • PubMed
    • Google Scholar
17. Book

    1. Clément G
    2. Ahlberg PE

    (2010)

    The endocranial anatomy of the early sarcopterygian Powichthys from Spitsbergen, based on CT scanning

    In: Elliott D. K, Maisey J. G, Yu X, Miao D, editors. Morphology, Phylogeny and Paleobiogeography of Fossil Fishes: Honoring Meemann Chang. Munich: Dr. Friedrich Pfeil. pp. 363–377.

    • Google Scholar
18. Book

    1. Cloutier R
    2. Arratia G

    (2004)

    Early diversification of actinopterygians

    In: Arratia G, Wilson M. V. H, Cloutier R, editors. Recent Advances in the Origin and Early Radiation of Vertebrates. Munchen, Germany: Verlag Dr. Friedrich Pfeil. pp. 217–270.

    • Google Scholar
19. 1. Coates MI

    (1999) Endocranial preservation of a Carboniferous actinopterygian from Lancashire, UK, and the interrelationships of primitive actinopterygians

    Philosophical Transactions of the Royal Society B: Biological Sciences 354:435–462.

    https://doi.org/10.1098/rstb.1999.0396

    • Google Scholar
20. 1. Coates MI

    (2009) Beyond the age of fishes

    Nature 458:413–414.

    https://doi.org/10.1038/458413a

    • Google Scholar
21. 1. Davis SP
    2. Finarelli JA
    3. Coates MI

    (2012) Acanthodes and shark-like conditions in the last common ancestor of modern gnathostomes

    Nature 486:247–250.

    https://doi.org/10.1038/nature11080

    • PubMed
    • Google Scholar
22. 1. Dupret V
    2. Sanchez S
    3. Goujet D
    4. Ahlberg PE

    (2017) The internal cranial anatomy of Romundina stellina Ørvig, 1975 (Vertebrata, placodermi, acanthothoraci) and the origin of jawed vertebrates-Anatomical atlas of a primitive gnathostome

    Plos One 12:e0171241.

    https://doi.org/10.1371/journal.pone.0171241

    • PubMed
    • Google Scholar
23. 1. Friedman M
    2. Brazeau MD

    (2010) A reappraisal of the origin and basal radiation of the osteichthyes

    Journal of Vertebrate Paleontology 30:36–56.

    https://doi.org/10.1080/02724630903409071

    • Google Scholar
24. 1. Friedman M

    (2007) Styloichthys as the oldest coelacanth: Implications for early osteichthyan interrelationships

    Journal of Systematic Palaeontology 5:289–343.

    https://doi.org/10.1017/S1477201907002052

    • Google Scholar
25. 1. Gardiner BG

    (1984)

    Relationships of the palaeoniscoid fishes, a review based on new specimens of mimia and Moythomasia from the Upper Devonain of Western Australia

    Bulletin of the British Museum of Natural History 37:173–428.

    • Google Scholar
26. 1. Giles S
    2. Coates MI
    3. Garwood RJ
    4. Brazeau MD
    5. Atwood R
    6. Johanson Z
    7. Friedman M

    (2015a) Endoskeletal structure in Cheirolepis (Osteichthyes, Actinopterygii), an early ray-finned fish

    Palaeontology 58:849–870.

    https://doi.org/10.1111/pala.12182

    • PubMed
    • Google Scholar
27. 1. Giles S
    2. Darras L
    3. Clément G
    4. Blieck A
    5. Friedman M

    (2015b) An exceptionally preserved Late Devonian actinopterygian provides a new model for primitive cranial anatomy in ray-finned fishes

    Proceedings of the Royal Society B: Biological Sciences 282:20151485.

    https://doi.org/10.1098/rspb.2015.1485

    • PubMed
    • Google Scholar
28. 1. Giles S
    2. Friedman M
    3. Brazeau MD

    (2015c) Osteichthyan-like cranial conditions in an Early Devonian stem gnathostome

    Nature 520:82–85.

    https://doi.org/10.1038/nature14065

    • PubMed
    • Google Scholar
29. 1. Giles S
    2. Friedman M

    (2014) Virtual reconstruction of endocast anatomy in early ray-finned fishes (Osteichthyes, actinopterygii)

    Journal of Paleontology 88:636–651.

    https://doi.org/10.1666/13-094

    • Google Scholar
30. 1. Giles S
    2. Rogers M
    3. Friedman M

    (2018) Bony labyrinth morphology in early neopterygian fishes (Actinopterygii: Neopterygii)

    Journal of Morphology 279:426–440.

    https://doi.org/10.1002/jmor.20551

    • PubMed
    • Google Scholar
31. 1. Giles S
    2. Xu GH
    3. Near TJ
    4. Friedman M

    (2017) Early members of 'living fossil' lineage imply later origin of modern ray-finned fishes

    Nature 549:265–268.

    https://doi.org/10.1038/nature23654

    • PubMed
    • Google Scholar
32. 1. Goloboff PA
    2. Catalano SA

    (2016) TNT version 1.5, including a full implementation of phylogenetic morphometrics

    Cladistics 32:221–238.

    https://doi.org/10.1111/cla.12160

    • Google Scholar
33. 1. Goloboff PA
    2. Farris JS
    3. Nixon KC

    (2008) TNT, a free program for phylogenetic analysis

    Cladistics 24:774–786.

    https://doi.org/10.1111/j.1096-0031.2008.00217.x

    • Google Scholar
34. Book

    1. Goujet D
    2. Young GC

    (2004)

    Placoderm anatomy and phylogeny: new insights

    In: Arratia G, Wilson M. V. H, Cloutier R, editors. Recent Advances in the Origin and Early Radiation of Vertebrates. München: Verlag Dr. Friedrich Pfeil. pp. 109–126.

    • Google Scholar
35. Book

    1. Goujet D

    (1984)

    Les poissons placodermes du Spitsberg : arthrodires dolichothoraci de la formation de Wood Bay (Dévonien inférieur) Cahiers de Paléontologie, Section Vertebrés

    Paris: CNRS.

    • Google Scholar
36. 1. Hamel M-H
    2. Poplin C

    (2008) The braincase anatomy of Lawrenciella schaefferi, actinopterygian from the Upper Carboniferous of Kansas (USA)

    Journal of Vertebrate Paleontology 28:989–1006.

    https://doi.org/10.1671/0272-4634-28.4.989

    • Google Scholar
37. 1. Holland T

    (2014) The endocranial anatomy of Gogonasus andrewsae Long, 1985 revealed through micro CT-scanning

    Earth and Environmental Science Transactions of the Royal Society of Edinburgh 105:9–34.

    https://doi.org/10.1017/S1755691014000164

    • Google Scholar
38. 1. Hu Y
    2. Lu J
    3. Young GC

    (2017) New findings in a 400 million-year-old devonian placoderm shed light on jaw structure and function in basal gnathostomes

    Scientific Reports 7:e7813.

    https://doi.org/10.1038/s41598-017-07674-y

    • Google Scholar
39. 1. Jarvik E

    (1972)

    Middle and Upper Devonian porolepiformes from east Greenland with special reference to Glyptolepis groenlandica n. sp. and a discussion on the structure of the head in the Porolepiformes

    Meddelelser Om Gronland 187:1–307.

    • Google Scholar
40. Book

    1. Jarvik E

    (1980)

    Basic Structure and Evolution of Vertebrates

    London: Academic Press.

    • Google Scholar
41. 1. King B
    2. Qiao T
    3. Lee MSY
    4. Zhu M
    5. Long JA

    (2017) Bayesian morphological clock methods resurrect placoderm monophyly and reveal rapid early evolution in jawed vertebrates

    Systematic Biology 66:499–516.

    https://doi.org/10.1093/sysbio/syw107

    • PubMed
    • Google Scholar
42. 1. Lewis PO

    (2001) A likelihood approach to estimating phylogeny from discrete morphological character data

    Systematic Biology 50:913–925.

    https://doi.org/10.1080/106351501753462876

    • PubMed
    • Google Scholar
43. 1. Limaye A

    (2012) Drishti - Volume exploration and presentation tool

    SPIE Developments in X-Ray Tomography 8506:85060X.

    https://doi.org/10.1117/12.935640

    • Google Scholar
44. 1. Long JA
    2. Mark-Kurik E
    3. Johanson Z
    4. Lee MS
    5. Young GC
    6. Min Z
    7. Ahlberg PE
    8. Newman M
    9. Jones R
    10. den Blaauwen J
    11. Choo B
    12. Trinajstic K

    (2015) Copulation in antiarch placoderms and the origin of gnathostome internal fertilization

    Nature 517:196–199.

    https://doi.org/10.1038/nature13825

    • PubMed
    • Google Scholar
45. 1. Long JA

    (1988)

    New palaeoniscoid fishes from the Late Devonian and early Carboniferous of Victoria

    Memoirs of the Association of Australasian Palaeontologists 7:1–64.

    • Google Scholar
46. 1. Lu J
    2. Giles S
    3. Friedman M
    4. den Blaauwen JL
    5. Zhu M

    (2016a) The oldest actinopterygian highlights the cryptic early history of the hyperdiverse ray-finned fishes

    Current Biology 26:1602–1608.

    https://doi.org/10.1016/j.cub.2016.04.045

    • PubMed
    • Google Scholar
47. 1. Lu J
    2. Giles S
    3. Friedman M
    4. Zhu M

    (2017) A new stem sarcopterygian illuminates patterns of character evolution in early bony fishes

    Nature Communications 8:1932.

    https://doi.org/10.1038/s41467-017-01801-z

    • PubMed
    • Google Scholar
48. 1. Lu J
    2. Zhu M
    3. Ahlberg PE
    4. Qiao T
    5. Zhu Y
    6. Zhao W
    7. Jia L

    (2016) A devonian predatory fish provides insights into the early evolution of modern sarcopterygians

    Science Advances 2:e1600154.

    https://doi.org/10.1126/sciadv.1600154

    • PubMed
    • Google Scholar
49. 1. Lu J
    2. Zhu M
    3. Long JA
    4. Zhao W
    5. Senden TJ
    6. Jia L
    7. Qiao T

    (2012) The earliest known stem-tetrapod from the lower devonian of China

    Nature Communications 3:1160.

    https://doi.org/10.1038/ncomms2170

    • PubMed
    • Google Scholar
50. 1. Maisey JG

    (2005) Braincase of the Upper Devonian shark Cladodoides wildungensis (Chondrichthyes, elasmobranchii), with observations on the braincase in early chondrichthyans

    Bulletin of the American Museum of Natural History 288:1–103.

    https://doi.org/10.1206/0003-0090(2005)288<0001:BOTUDS>2.0.CO;2

    • Google Scholar
51. 1. Maisey JG

    (2007) The braincase in Paleozoic symmoriiform and cladoselachian sharks

    Bulletin of the American Museum of Natural History 307:1–122.

    https://doi.org/10.1206/0003-0090(2007)307[1:TBIPSA]2.0.CO;2

    • Google Scholar
52. 1. Nielsen E

    (1942)

    Studies on Triassic Fishes from east Greenland

    1–309, I. Glaucolepis and Boreosomus, Studies on Triassic Fishes from east Greenland, Vol. 146, Palaeozoologica Groenlandica.

    • Google Scholar
53. 1. Northcutt RG
    2. Bemis WE

    (1993) Cranial nerves of the coelacanth, Latimeria chalumnae [Osteichthyes: sarcopterygii: actinistia], and comparisons with other craniata

    Brain, Behavior and Evolution 42 Suppl 1:1–69.

    https://doi.org/10.1159/000114175

    • PubMed
    • Google Scholar
54. Book

    1. Northcutt RG

    (1989) The phylogenetic distribution and innervation of craniate mechanoreceptive lateral lines

    In: Coombs S, Görner P, Münz H, editors. The Mechanosensory Lateral Line. New York: Springer. pp. 17–78.

    https://doi.org/10.1007/978-1-4612-3560-6_3

    • Google Scholar
55. 1. Ørvig T

    (1973)

    Fossila fisktänder i svepelektronmikroskopet: gamlafrågeställningar i ny belysning

    Fauna Och Flora 68:166–173.

    • Google Scholar
56. 1. Patterson C

    (1975) The braincase of pholidophorid and leptolepid fishes, with a review of the actinopterygian braincase

    Philosophical Transactions of the Royal Society B: Biological Sciences 269:275–579.

    https://doi.org/10.1098/rstb.1975.0001

    • PubMed
    • Google Scholar
57. Book

    1. Poplin CM

    (1974)

    Etude de quelques paleoniscides pennsylvaniens du kansas

    Paris: CNRS.

    • Google Scholar
58. 1. Qiao T
    2. King B
    3. Long JA
    4. Ahlberg PE
    5. Zhu M

    (2016) Early gnathostome phylogeny revisited: multiple method consensus

    Plos One 11:e0163157.

    https://doi.org/10.1371/journal.pone.0163157

    • PubMed
    • Google Scholar
59. 1. Qu Q
    2. Haitina T
    3. Zhu M
    4. Ahlberg PE

    (2015) New genomic and fossil data illuminate the origin of enamel

    Nature 526:108–111.

    https://doi.org/10.1038/nature15259

    • PubMed
    • Google Scholar
60. 1. Rayner DH

    (1948) The structure of certain Jurassic holostean fishes with special reference to their neurocrania

    Philosophical Transactions of the Royal Society B: Biological Sciences 233:287–345.

    https://doi.org/10.1098/rstb.1948.0006

    • Google Scholar
61. 1. Robineau D

    (1975)

    The jugular vein system and its homologies in Latimeria chalmunae (Pisces, Crossopterygii, Coelacanthidae)

    Comptes Rendus Hebdomadaires Des Seances De l'Academie Des Sciences. Serie D: Sciences Naturelles 281:45–48.

    • PubMed
    • Google Scholar
62. 1. Ronquist F
    2. Teslenko M
    3. van der Mark P
    4. Ayres DL
    5. Darling A
    6. Höhna S
    7. Larget B
    8. Liu L
    9. Suchard MA
    10. Huelsenbeck JP

    (2012) MrBayes 3.2: efficient bayesian phylogenetic inference and model choice across a large model space

    Systematic Biology 61:539–542.

    https://doi.org/10.1093/sysbio/sys029

    • Google Scholar
63. 1. Sakellariou A
    2. Sawkins TJ
    3. Senden TJ
    4. Limaye A

    (2004) X-ray tomography for mesoscale physics applications

    Physica A: Statistical Mechanics and Its Applications 339:152–158.

    https://doi.org/10.1016/j.physa.2004.03.055

    • Google Scholar
64. 1. Schaeffer B

    (1981)

    The xenacanth shark neurocranium, with comments on elasmobranch monophyly

    Bulletin of the American Museum of Natural History 169:3–66.

    • Google Scholar
65. Book

    1. Schultze H-P
    2. Cumbaa SL

    (2001)

    Dialipina and the characters of basal actinopterygians

    In: Ahlberg P. E, editors. Major Events in Early Vertebrate Evolution: Palaeontology, Phylogeny, Genetics and Development. London: Taylor and Francis. pp. 315–332.

    • Google Scholar
66. 1. Schultze H-P

    (1968)

    Palaeoniscoidea-Schuppen aus dem Unterdevon Australiens und Kanadas und aus dem Mitteldevon Spitsbergens

    Bulletin of the British Museum of Natural History 16:343–368.

    • Google Scholar
67. 1. Schultze H-P

    (2016) Scales, enamel, cosmine, Ganoine, and early osteichthyans

    Comptes Rendus Palevol 15:83–102.

    https://doi.org/10.1016/j.crpv.2015.04.001

    • Google Scholar
68. Book

    1. Stensiö E

    (1963a)

    Anatomical Studies on the Arthrodiran Head

    Stockholm: Almqvist & Wiksell.

    • Google Scholar
69. Book

    1. Stensiö E

    (1963b)

    The Brain and the Cranial Nerves in Fossil, Lower Craniate Vertebrates

    Oslo: Universitetsforlaget.

    • Google Scholar
70. Book

    1. Stensiö EA

    (1969)

    Elasmobranchiomorphi Placodermata Arthrodires

    In: Piveteau J, editors. Traité De Paléontologie. Paris: Masson. 71-692.

    • Google Scholar
71. Software

    1. Swofford DL

    (2003)

    PAUP*: phylogenetic analysis using parsimony, version 4.0 b10

    PAUP*: phylogenetic analysis using parsimony.
72. 1. Wang NZ
    2. Dong ZZ

    (1989)

    Discovery of late Silurian microfossils of agnatha and fishes from Yunnan, China

    Acta Palaeontologica Sinica 28:192–206.

    • Google Scholar
73. 1. Watson DMS

    (1937) The acanthodian fishes

    Philosophical Transactions of the Royal Society B: Biological Sciences 228:49–146.

    https://doi.org/10.1098/rstb.1937.0009

    • Google Scholar
74. 1. Young GC

    (1979) New information on the structure and relationships of Buchanosteus (Placodermi: Euarthrodira) from the Early Devonian of New South Wales

    Zoological Journal of the Linnean Society 66:309–352.

    https://doi.org/10.1111/j.1096-3642.1979.tb01912.x

    • Google Scholar
75. 1. Young GC

    (1980)

    A new Early Devonian placoderm from New South Wales, Australia, with a discussion of placoderm phylogeny

    Palaeontographica 167:10–76.

    • Google Scholar
76. 1. Yu X

    (1998) A new porolepiform-like fish, Psarolepis romeri, gen. et sp. nov. (Sarcopterygii, Osteichthyes) from the Lower Devonian of Yunnan, China

    Journal of Vertebrate Paleontology 18:261–274.

    https://doi.org/10.1080/02724634.1998.10011055

    • Google Scholar
77. 1. Zhu M
    2. Ahlberg PE
    3. Pan Z
    4. Zhu Y
    5. Qiao T
    6. Zhao W
    7. Jia L
    8. Lu J

    (2016) A Silurian maxillate placoderm illuminates jaw evolution

    Science 354:334–336.

    https://doi.org/10.1126/science.aah3764

    • PubMed
    • Google Scholar
78. Book

    1. Zhu M
    2. Wang W
    3. Yu X

    (2010)

    Meemannia eos, a basal sarcopterygian fish from the Lower Devonian of China–expanded description and significance

    In: Elliott D. K, Maisey J, Yu X, Miao D, editors. Morphology, Phylogeny and Paleobiogeography of Fossil Fishes. München: Verlag Dr. Friedrich Pfeil. pp. 199–214.

    • Google Scholar
79. 1. Zhu M
    2. Yu X
    3. Ahlberg PE
    4. Choo B
    5. Lu J
    6. Qiao T
    7. Qu Q
    8. Zhao W
    9. Jia L
    10. Blom H
    11. Zhu Y

    (2013) A Silurian placoderm with osteichthyan-like marginal jaw bones

    Nature 502:188–193.

    https://doi.org/10.1038/nature12617

    • PubMed
    • Google Scholar
80. 1. Zhu M
    2. Yu X
    3. Ahlberg PE

    (2001) A primitive sarcopterygian fish with an eyestalk

    Nature 410:81–84.

    https://doi.org/10.1038/35065078

    • PubMed
    • Google Scholar
81. 1. Zhu M
    2. Yu X
    3. Janvier P

    (1999) A primitive fossil fish sheds light on the origin of bony fishes

    Nature 397:607–610.

    https://doi.org/10.1038/17594

    • Google Scholar
82. 1. Zhu M
    2. Yu X

    (2002) A primitive fish close to the common ancestor of tetrapods and lungfish

    Nature 418:767–770.

    https://doi.org/10.1038/nature00871

    • PubMed
    • Google Scholar
83. 1. Zhu M
    2. Zhao W
    3. Jia L
    4. Lu J
    5. Qiao T
    6. Qu Q

    (2009) The oldest articulated osteichthyan reveals mosaic gnathostome characters

    Nature 458:469–474.

    https://doi.org/10.1038/nature07855

    • PubMed
    • Google Scholar

Author details

1. Alice M Clement

   1. College of Science and Engineering, Flinders University, Adelaide, Australia
   2. Department of Organismal Biology, Evolutionary Biology Centre, Uppsala University, Uppsala, Sweden
   3. Department of Sciences, Museum Victoria, Melbourne, Australia

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing—original draft, Writing—review and editing

   Contributed equally with

   Benedict King and Sam Giles

   For correspondence

   alice.clement@flinders.edu.au

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0380-7347

2. Benedict King

   1. College of Science and Engineering, Flinders University, Adelaide, Australia
   2. Naturalis Biodiversity Center, Leiden, Netherlands

   Contribution

   Investigation, Formal analysis, Data curation, Methodology, Visualization, Writing—original draft, Writing—review and editing

   Contributed equally with

   Alice M Clement and Sam Giles

   "This ORCID iD identifies the author of this article:" 0000-0002-9489-8274

3. Sam Giles

   Department of Earth Sciences, University of Oxford, Oxford, United Kingdom

   Contribution

   Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing—original draft, Writing—review and editing

   Contributed equally with

   Alice M Clement and Benedict King

   "This ORCID iD identifies the author of this article:" 0000-0001-9267-4392

4. Brian Choo

   College of Science and Engineering, Flinders University, Adelaide, Australia

   Contribution

   Visualization, Writing—review and editing

5. Per E Ahlberg

   Department of Organismal Biology, Evolutionary Biology Centre, Uppsala University, Uppsala, Sweden

   Contribution

   Resources, Software, Writing—review and editing

6. Gavin C Young

   1. Department of Applied Mathematics, Research School of Physics & Engineering, Australian National University, Canberra, Australia
   2. Australian Museum Research Institute, Sydney, Australia

   Contribution

   Resources, Writing—review and editing

7. John A Long

   1. College of Science and Engineering, Flinders University, Adelaide, Australia
   2. Department of Sciences, Museum Victoria, Melbourne, Australia

   Contribution

   Conceptualization, Formal analysis, Resources, Software, Visualization, Writing—review and editing

   "This ORCID iD identifies the author of this article:" 0000-0001-8012-0114

• 4,458

  views

• 431

  downloads

• 26

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.