Skip to main content
NCBI home page
Journal of Anatomy logoLink to Journal of Anatomy
. 2021 Mar 19;239(2):374–390. doi: 10.1111/joa.13432

On the structure and development of the parasphenoid and the ventral rostral bones in Acipenser baerii and Polyodon spathula (Actinopterygii, Acipenseriformes)

Alexey A Tsessarsky 1,✉
PMCID: PMC8273586  PMID: 33742445

Abstract

Acipenseriformes (sturgeons and paddlefishes) are currently recognized as sister‐group of Neopterygii (bowfin, gars and teleosts) and along with Polypteriformes (bihirs) constitute the two most basal taxa among living ray‐finned fishes. Acipenseriforms uniquely possess a large preoral snout which distinguishes them from other actinopterygians. It is covered ventrally by a longitudinal series of exoskeletal elements which extends along the middle part of the snout from the parasphenoid to the very anterior tip of the head. These cranial elements, highly variable in size, number and proportions, are generally referred to as ventral rostral bones. The homologies of these bones remain unresolved. The issue is getting even more complicated because of vague nature of the parasphenoid of acipenseriforms, with which the ventral rostral series is in a contact. Paradoxically, the homology of this bone of acipenseriforms has never been subjected to thorough survey based on the early development and morphology of this bone. Here, the development of the parasphenoid and the ventraI rostral bones in Siberian sturgeon Acipenser baerii and American paddlefish Polyodon spathula is investigated based on a large sample of specimens of both species ranging from larvae just posthatching to juveniles of 50 days posthatching. Data obtained in this study allowed to establish primary homologies of the parasphenoid and the ventral rostral bones of Acipenseriformes and to address the evolutionary history of the snout in these fishes.

This study provides a detailed description of the development of the dermal bones underlining the cranial base in Acipenser baerii and Polyodon spathula. The data obtained in this research allowed to reconsider the homologies of the parasphenoid and the ventral rostral series of acipenseriforms. Based on these findings, it is suggested that the developmental mechanism responsible for the diversity of these bones in acipenseriforms involved heterochrony.

graphic file with name JOA-239-374-g008.jpg

1. INTRODUCTION

Thought of as “living fossils” (Gardiner, 1984a) sturgeons and paddlefishes are currently classified in two families (Acipenseridae and Polyodontidae), which along with fossil relatives constitute the crown‐group Acipenseriformes (Hilton & Forey, 2009). It is generally agreed that Acipenseriformes is the sister‐group of Neopterygii (gars, bowfins and teleosts), and together with Polypteriformes (bichirs) represent the two most basal groups of extant ray‐finned fishes (Betancur et al., 2017; Cloutier & Arratia, 2004; Coates, 1999; Gardiner et al., 2005; Patterson, 1982). Given their phylogenetic position, acipenseriforms appear to be of key importance for understanding the evolutionary history of actinopterygians. The detailed studies of their morphology will have vital effect on assigning plesiomorphic states and polarity of characters transformations within ray‐finned fishes, and in bony fishes in general.

As is well known (Grande & Bemis, 1991; Hilton et al., 2011), the snout in acipenseriforms is covered ventrally by a longitudinal series of exoskeletal elements which extends along the middle part of the snout from the parasphenoid to the very anterior tip of the head. These cranial elements, highly variable in size, number, and proportions, are generally referred to as ventral rostral bones (Findeis, 1997; Hilton et al., 2011; Jollie, 1980).

As discussed in more detail below, the homologies of the ventral rostral bones of acipenseriforms to the bones of the snout of other fishes remain unresolved, as are the correspondences of these bones of sturgeons to those of the paddlefishes. Along with many other traits of acipenseriforms, which have no clear homologues in other actinopterygians, the ventral rostral bones impose severe constraints on the studies of the systematics and evolution of the group (e.g. Hilton & Forey, 2009; Hilton et al., 2011). The problem is further complicated by uncertain nature of the parasphenoid of acipenseriforms, with which this series is in a contact. Surprisingly, the homology of the parasphenoid in sturgeons and paddle‐fishes has never been subjected to thorough survey based on early development and morphology of this bone.

The account which follows is an attempt to trace the development of the parasphenoid and the ventral rostral bones in sturgeon and paddlefish in the hope that it may help to establish the primary homologies (de Pinna, 1991) of these bones and to address the evolutionary history of the snout in Acipenseriformes.

2. MATERIAL AND METHODS

The larvae (1 dph) of Acipenser baerii Brandt, 1869 and Polyodon spathula Walbaum, 1792 were obtained from commercial fishery and transferred to the laboratory, where the larvae were held in aquaria at 19–21°C and fed with Artemia nauplii and chironomid larvae. Each day, from 5 to 10 specimens of both species were euthanized with phosphate‐buffered MS‐222 (Sigma Chemical Co.) overdose and fixed in 4% buffered formalin. Prior to fixation the euthanized specimens were photographed and the total length (TL) of each specimen was measured using measuring software tools of the digital optical microscope Keyence VHX‐1000 (see Figure 1 for growth patterns of A. baerii and P. spatula). Full developmental series from 1 to 50 dph comprised 282 specimens of P. spathula (8.7–81.4 mm TL) and 285 specimens of A. baerii (9.8–66.3 mm TL). Additionally, several larger specimens of both species (80–110 mm TL) and two subadult Siberian sturgeons (550 and 583 mm TL) have been used for the present study.

FIGURE 1.

FIGURE 1

Open in a new tab

Total lengths (mean; whisker: mean ±2 × SD) of sampled Acipenser baerii and Polyodon spathula during 50 days posthatching. ◊—A. baerii, △—P. spathula

Development of the bones was studied on whole‐mounts stained with Alizarin Red S (Fluka Chemie AG) and Alcian Blue 8GX (Sigma Aldrich) according to standard protocol (Depew, 2008). However, in most cases, only alizarin staining was used to prevent decalcification by acetic acid and to contrast the minute details of bone appearance.

Specimens were studied, photographed, and sketched under Leica MZ6 stereomicroscope. The confocal images of the early stages of bones development (alizarin staining) were obtained from Zeiss Ism 880 microscope.

Transverse serial sections of larval and early postlarval stages of both species were prepared by embedding in paraffin and sectioning at 8 µm with a rotary Reichert microtome. Sections were stained according to Mallory protocols (Barnard, 1987). Also, in certain cases (see Figure 4b), the alizarin stained and cleared specimens were decalcified with 5% HNO3, embedded in paraffin, sectioned transversely at 10 µm and contrasted with Toluidine blue O (Sigma Aldrich).

FIGURE 4.

FIGURE 4

Open in a new tab

Acipenser baerii, development of the medial palatobasal process. Transverse sections through the snout at the level of the subrostral. (a) 24.1 mm TL, 16 dph; scale bar—0.1 mm; (b) 24.5 mm TL, 17 dph; scale bar—0.1 mm; (c) 36.0 mm TL, 26 dph; scale bar—0.2 mm. pr.pb.m.—medial palatobasal process; Sr—subrostral; arrowheads point to the outgrowths of developing palatobasal process

3D‐reconstruction from serial paraffin sections of the skull of A. baerii (20 mm TL, 10 dph) was made using the ImageJ software. The sections were photographed and composed into stack, which was manually aligned with the help of the TrackEm2 plugin (Cardona et al., 2012). The Volume Viewer plugin (https://imagej.nih.gov/ij/plugins/volume‐viewer.html) was used to control the accuracy of sections alignment. The final visualization of the model was performed in Imaris V 7.1.1.

3. RESULTS

Besides the parasphenoid and the ventral rostral series, the exoskeleton of the ventral side of the neurocranium of acipenseriforms includes the lateral‐line bones of the infraorbital canals, the upper pharyngeal dental plates and the so‐called stellate bones of Polyodon (see Grande & Bemis, 1991). All these ossifications are not the subject of the present account and will be dealt with elsewhere.

In the description below, the development of the bones is divided into a number of readily discernable developmental stages, such as the first appearance of ossification, fusion of the rudiments, etc. The age and size limits for each stage are specified in the text.

3.1. Development of the ventral rostral bones and the parasphenoid in A. baerii

First bony rudiments of the parasphenoid are found at the beginning of the exogenous feeding (9–10 dph) in the specimens of 19.0–21.4 mm TL. The ossifications arise in the posterior ends of the thin sheet of fairly dense mesenchymous tissue (Figure 2, p.mes.), which underlies the cranial base. The anterior border of this mesenchymous anlagen lies behind the hypophysis (Hyp) and the foramens for the internal carotid arteries (a.ci.), while posteriorly it spreads into the parachordal part of the skull and reaches the level of the first infrapharyngobranchials (Iph1). In front of the later, the anlagen send on both sides the lateral extensions (pr.asc.), which point toward the lateral commissures (lc) of the neurocranium. These outgrowths correspond topographically to the ascending processes of the parasphenoid. The rostral part of the anlagen is shaped as a broad plate, while posteriorly, from the level, where the cranial base becomes split by the canal for the notochord (nch.c.), the anlagen bifurcates into the left and the right arms, which stretch posteriorly on both sides of the notochord. It is here, in the posterior portions of these arms, at the bases of the future ascending processes, that the first ossification of the parasphenoid takes place (Prsp).

FIGURE 2.

FIGURE 2

Open in a new tab

Acipenser baerii (20 mm TL, 10 dph), 3D‐reconstruction of the neurocranium. (a) Ventral view; with close up (bottom) to show the position of the ossification of the parasphenoid, and transverse sections of the neurocranium (right) at the levels shown by dotted lines. (b) Sagittal section of the neurocranium to show the position of the anterior and posterior mesenchymous anlagens of the parasphenoid. a.ci.—internal carotid artery; a.mes.—anterior mesenchymous anlagen; a.pb.e.—efferent pseudobranchial artery; f.bc.—basicranial fenestra; Hm—hyomandible; Hyp—hypophysis; ibr1—infrapharyngobranchial 1; lc—lateral commissure; nc—nasal capsule; nch.c.—canal for notochord; p.mes.—posterior mesenchymous anlagen; pr.bp.—basipterygoid process; Prsp—parasphenoid ossification

No other ossification is developed at this stage on the ventral side of the chondrocranium. However, one more mesenchymous structure is found in front of the hypophysis (Figure 2, a.mes.). Posteriorly, it begins at the level of the anterior edges of the basipterygoid (basitrabecular—de Beer, 1925) processes (pr.bp.), just behind the posterior border of the remnant of the basicranial fenestra (f.bc.). This anlagen is shaped as a narrow strip of loose mesenchymous tissue, which passes beneath the basicranial fenestra and reaches up to the middle of the nasal capsules (nc). As revealed by further development, this rudiment occupies the area in which the ossifications of the ventral rostral series will arise at later developmental stages.

The process of ossification of the parasphenoid proceeds mostly in the anterior direction and at 19.8–21.8 mm TL (12 dph) the bone (Figure 3a) is represented by the ascending processes (pr.asc.), which are readily recognized in front of the first infrapharyngobranchials, and the narrow bony strips stretching rostrad from the bases of these processes along the anterior part of the notochord (Prsp). Also, in the specimen shown on Figure 3a, thin bony spicule can be seen on one side of the head posteriorly to the ascending process (Figure 3a, arrowhead), representing the early rudiment of the posterior extension of the parasphenoid.

FIGURE 3.

FIGURE 3

Open in a new tab

Acipenser baerii, early development of the parasphenoid and the subrostral. (a, b, e) Confocal images (left) of the early parasphenoid ossification, and the diagrammatic representations (right) based on the same specimens. Ventral views of the skull with visceral skeleton removed. (a) 21.0 mm TL, 10 dph; scale bar 0.3 mm, arrowhead points to the posterior extension of the parasphenoid. (b) 22.3 mm TL, 10 dph; scale bar 0.5 mm. (e) 27.2 mm TL, 17 dph; scale bar 0.5 mm. (c, d) Confocal images of the parasphenoid in ventral view. (c) 24.8 mm TL, 15 dph; scale bar 0.2 mm. (d) 26.0 mm TL, 16 dph; scale bar 0.2 mm. Hyp—hypophysis; Nch—notochord; pr.bp.—basipterygoid process; Prsp—parasphenoid, Sr—subrostral. Barbles are drowned by dotted outlines

At 22.3–24.3 mm TL (12–13 dph) the narrow posterior extensions are developed on both sides of the notochord (Figure 3b, Prsp). Anteriorly the ossification expands to the anterior tip of the notochord, where the left and the right rudiments of the parasphenoid become interconnected by bony trabecles. From this stage on, more trabecles develop between the parasphenoid rudiments (Figure 3c,d) producing bony network which unites the rudiments in a single bone. At the same time this bony network growths in rostral direction expanding beyond the tip of the notochord.

At 24.5–30.3 mm TL (17–19 dph), the parasphenoid is much densely ossified and deeply forked posteriorly. New ossification appears at this stage anteriorly to the parasphenoid (Figure 3e, Sr). It is represented by a narrow bony splint situated in front of the hypophysis, at the level of the postnasal walls. Given its position and the splinter‐like shape, this ossification fully corresponds to the bone which Sewertzoff (1926) described in A. ruthenus under the name of the subrostral.

Remarkably, in A. baerii, roughly at the time when the subrostral begins to ossify the underlying cartilage produces a pair of ridges or swellings, which rise on both sides of the bone (Figure 4a). Later on, these swellings curve medially under the subrostral (Figure 4b) and eventually fuse below the bone enclosing the latter within the canal inside the cartilage (Figure 4c). In this way, the medial palatobasal process (Sewertzoff, 1928; central trabecular process of Findeis, 1993) is being formed (Figure 4c, pr.pb.m.).

It is noteworthy at this juncture, that the outlines of the subrostral at early stages of its development distinctly vary in different specimens. In some it remains very narrow (Figure 5b), while in others the bone appears notably broader (Figure 5a,e). There are also specimens in which the subrostral displays intermediate condition, with narrow rostral or hind part and broadened opposite portion (Figure 5c,d). In larger specimens these differences become more or less levelled off due to the overall growth of the fish.

FIGURE 5.

FIGURE 5

Open in a new tab

Acipenser baerii, shape variability of the subrostral and the limits of the medial palatobasal process in the specimens of 24.7–28.4 mm TL (16–17 dph). Upper row (a, b, c, d)—Ventral views of the subrostral (anterior to the top) in alizarin stained specimens with the contours of the medial palatobasal process (black lines). Bottom row (a′, b′, c′, d′)—diagrammatic representations based on the same specimens. a, e—palatobasal process just commences overgrowing the subrostral (arrowheads). b—The whole subrostral is enclosed within the palatobasal process. c, d—palatobasal process is formed around the caudal (c) or rostral (d) ends of the subrostral. Sr—subrostral

Closer examination indicates that the shape of the subrostral correlates with the state of development of the medial palatobasal process. In those cases where the subrostral is relatively broad (Figure 5a′,e′), the palatobasal process is just beginning to form. In the specimens on Figure 5c,b the palatobasal process encloses only the narrow part of the subrostral, while the broader end of the bone remains free (Figure 5c′,d′). Finally, in the specimens with narrow needle‐shaped subrostral (Figure 5b) the whole bone is enclosed within the cartilage of this process. These patterns indicate that the width of the subrostral is controlled by the palatobasal process: once the cartilaginous swellings of the developing palatobasal process raised on the sides of the subrostral, further broadening of the bone appears constrained by the cartilage.

From the above it can be supposed, that the shape of the subrostral in sturgeon most likely depends on the mutual timing of this bone and the palatobasal process development. The later the growth of the palatobasal process begins relative to the onset of the subrostral ossification, the more time remains for the bone to expand laterally. Further discussion of this model and its bearing on acipenseriforms evolution is provided in Section 4.5.

Soon after the appearance of the subrostral, at 31.2–36.0 mm TL (18–26 dph), another unpaired ossification arises in front of the latter, at the level of the barbles (Figure 6a, Br1). Contrary to the narrowed subrostral, this bone from the very beginning is represented by a rounded or roughly triangular plate. After Sewertzoff (1926) this ossification is here referred to as the medial basirostral.

FIGURE 6.

FIGURE 6

Open in a new tab

Acipenser baerii, development of the basirostrals. (a) Ventral view of the skull (31.2 mm TL, 18 dph) with visceral skeleton removed (right) and the close up of the snout (left) to show the early ossification of the basirostral. Scale bar 1 mm. (b, c) Ventral views of the medial basirostral series in 34.2 mm TL, 20 dph (b) and 38.0 mm TL, 21 dph (c) specimens; scale bars 1 mm. (d, e) The medial basirostrals in lateral (top) and ventral (bottom) views (anterior to the left) in 51.0 mm TL, 45 dph (d) and 550 mm TL (e) specimens. Br1, Br2, Br3—basirostrals 1, 2, 3; Vr—ventrorostral; vpBr—ventral process of the basirostral

The appearance of the medial basirostral may be followed by the second unpaired basirostral bone in front of it (Figure 6b, Br2), and sometimes, by the third basirostral in front of the second (Figure 6c, Br3). From the total of 125 specimens at the age of 26–50 dph (32.2–63.8 mm TL), 59 specimens had only one medial basirostral, 61 specimens developed two, and in five specimens there were 3 basirostrals.

During further development, the central sculptured part of each basirostral grows downwards to produce large ventrally directed process (Figure 6d,e, vpBr). The shaft of this process is embedded into the dermal layers of the skin, while its flattened distal end projects to the epidermis and is sculptured with irregular pits and ridges. The presence of similar protrusions were reported in A. ruthenus and A. fulvescens by Jollie (1984) and in A. ruthenus, A. baerii and Huso by Hilton et al. (2011). Homology of these processes is discussed below, in Section 4.3.

At 34.0–42.1 mm TL (28–36 dph), new ossifications are found in front of the anterior basirostral (Figure 7a, Vr1). In contrast to the basirostral, they arise not in the midline, but on the sides to the sagittal plane of the head and display distinctly paired nature. The term ventrorostrals is used here to distinguish these bones from the invariably unpaired basirostrals. Noteworthy, in most specimens, these ossicles appear asynchronously, one after another, with slight delay. For instance, in the specimen shown on Figure 6c only one bone of the pair began to ossify (Vr), while the place of its counterpart is still empty. As a result of such asymmetry in timing, one bone of the pair turns out somewhat larger than the other (see for instance, Figure 7a).

FIGURE 7.

FIGURE 7

Open in a new tab

Acipenser baerii, development of the ventrorostrals. (a) Ventral view of the head (right) and the close up of the snout (left) to show the early appearance of the first pair ventrorostrals in the specimen of 34.0 mm TL, 28 dph, scale bar 1 mm. (b–e) Snouts in ventral view. b—61.2 mm TL, 48 dph; c—56.9 mm TL; 47 dph; d—64.5 mm TL, 48 dph; e—96.5 mm TL, 65 dph. b–d—scale bar 1 mm; e—scale bar 2 mm. Br1; Br2—basirostrals 1, 2; Vr1; Vr2; Vr3; Vr4—ventrorostrals 1, 2, 3, 4

This difference in the size of the first pair ventrorostrals seems to affect the arrangement of the new ventrorostrals, which develop farther rostrally. They appear shortly after the first pair and ossify one by one in from the posterior to the anterior direction in parallel with the elongation of the rostrum. In most specimens, these ventrorostrals exhibit almost symmetrical bilateral pattern (Figure 7d,e). However, in some individuals the symmetrical lay‐out of the ventrorostral series is distorted and the bones are arranged in alternating chessboard order (Figure 7b).

It appears that which of the two patterns will be established depends on how much the ventrorostrals of the first pair differ from each other in size. In the cases when this difference is insignificant, the more anterior ventrorostrals develop in more or less symmetrical mode, resulting in symmetrical bilateral pattern (Figure 7d,e). Conversely, when the first pair ventrorostrals display more drastic disproportion, the larger one invades the space of the ventrorostrals of the next anterior pair, in which case the bones of this pair develop not opposite each other, but diagonally. In this way the initial bilateral arrangement of the ventrorostrals is transformed into alternating chessboard pattern (Figure 7b).

Soon after the ventrorostrals began to ossify, the adjacent ends of the parasphenoid and the subrostral are getting closer and eventually become interconnected by bony trabecles (Figure 8a). In this way the subrostral fuses with the parasphenoid and becomes the so called “median anterior process” (Bemis et al., 1997; Findeis, 1997; Grande & Bemis, 1991; Hilton et al., 2011) of the parasphenoid (Figure 8b).

FIGURE 8.

FIGURE 8

Open in a new tab

Acipenser baerii, development of the median anterior process of the parasphenoid. (a) Ventral view of the head (left) and the close up (right) to show the subrostral connecting the parasphenoid by bony trabecle (arrowhead) in 46.7 mm TL, 38 dph specimen. Scale bar 2 mm. (b) Ventral view of the parasphenoid and its median anterior process in 550 mm TL specimen (head length 183 mm). Scale bar 8 mm. Br—basirostral; map—median anterior process (the subrostral fused to the parasphenoid); Prsp—parasphenoid; Sr—subrostral

The general pattern of the parasphenoid and the ventral rostral series in A. baerii appears to be completed at 45–50 dph, when the fish reaches 60–65 mm TL. Further development is confined to the addition of small ventrorostral ossicles at the tip of the rostrum and on the sides of the series.

3.2. Development of the parasphenoid and the ventral rostral bones in P. spathula

First ossified rudiment of the parasphenoid in the paddlefish was found in the specimen of 32.0 mm TL (15 dph). It is represented by a bony network underlining the parachordal part of the neurocranium (Figure 9a, Prsp). Anteriorly it ends at the level of the rostral tip of the notochord, where the left and the right parts of the rudiment are interconnected by transverse bony bridge. The ascending processes are present, as well as the posterior extensions of the bone on both sides of the notochord (somewhat asymmetrically developed). All these features correspond to rather derived condition of the parasphenoid development in sturgeon (comp. Figure 3a). Curiously, no earlier developmental stages of the parasphenoid was detected in either of the sampled paddlefishes, neither in the whole‐mounts or on the sections. This might indicate, that the ossification of the parasphenoid in the paddlefish proceeds very quickly and more frequent sampling is needed to detect the earlier stages of its development.

FIGURE 9.

FIGURE 9

Open in a new tab

Polyodon spathula, development of the parasphenoid and the ventral rostral bones. (a–c) Confocal images (left) of the parasphenoid, and the diagrammatic representations (right) based on the same specimens. Ventral views of the skull with visceral skeleton removed, barbles shown by dotted lines. (a) 32.0 mm TL, 15 dph, scale bar 0.6 mm; (b) 32.6 mm TL, 19 dph, scale bar 2 mm; (c) 35.1 mm TL, 23 dph, scale bar 0.7 mm. (d) Ventral view of the rostrum with ventrorostrals, 33.4 mm TL, 19 dph, scale bar 1 mm. (e–g) Ventral views of the subrostral and the rostral end of the parasphenoid, scale bars 0.5 mm. (e) 39.0 mm TL, 24 dph, confocal image; (f) 43.0 mm TL, 26 dph, photo of alizarin stained specimen; (g) 42.2 mm TL, 26 dph, confocal image. (h) Ventral view of the skull (visceral skeleton removed) of alizarin stained specimen of 72.4 mm TL, 47 dph, scale bar 3 mm. Hm—hyomandible; Hyp—hypophysis; Ifo—infraorbitals; nc—nasal capsule; Nch—notochord; pr.asc.—ascending process of parasphenoid; pr.bp.—basipterygoid process; Prsp—parasphenoid; Sr—subrostral; Vr—ventrorostral; udp—upper pharyngeal dental plates. Barbles are drowned by dotted outlines

In the specimens of 32.6–36.3 mm TL (19–20 dph) the parasphenoid is much enlarged, but still retains its reticulate lacy pattern (Figure 9b, Prsp). Anteriorly the bone reaches the hypophysis. At this stage new ossifications appear in the rostrum area. In the specimen shown on Figure 9b, they are represented by a pair of elongate ossicles (Vr) situated just in front of the barbles. A slightly larger specimen of the same age (Figure 9d) has two pairs of similar bony rudiments in the rostrum area. Given their paired arrangement and the location in front of the barbles and the nasal capsules, all these ossifications should be considered the members of the ventrorostral series.

At 34.1–38.9 mm TL (20–23 dph) a fine mesh of bony trabecles appears in front of the parasphenoid (Figure 9c, Sr). It is situated between the caudal parts of the nasal capsules, rostrally to the hypophysis (Hyp) and the basipterygoid processes (pr.bp.). The topographic position of this rudiment indicates that it corresponds to the subrostral of sturgeon.

At 24–25 dph (39.0–43.1 mm TL) the subrostral in Polyodon is much enlarged (Figure 9e, Sr). In the specimen shown on Figure 9e it is connected to the parasphenoid by bony trabecle (arrow). Later, more trabecles develop between the two bones (Figure 9f), and finally, at 25–26 dph (40.8–44.0 mm TL), the subrostral fuses with the parasphenoid in a unified bone (Figure 9g).

Noteworthy, the shape of the subrostral of Polyodon from the very beginning differs strikingly from the split‐like subrostral of Acipenser. The most likely explanation seems to be the lack of the medial palatobasal process in Polyodon (see, however below, Section 4.5). As discussed above, in sturgeon, this process restricts the lateral expansion of the subrostral at the very onset of its ossification (see Figures 4 and 5). In the paddlefish there is no such impediment and the ossification of the subrostral can freely expand in lateral directions. Consequently, at later stages, the subrostral in Polyodon attains almost the same width as the main body of the parasphenoid (Figure 9h).

Development of the ventrorostral series proceeds through the significant lengthening of the bones and by addition of new ossifications farther rostrally (Figure 10a). As mentioned above, the posterior ventrorostrals arise in front of the barbles, pretty far from the parasphenoid. Consequently, a distinct gap in bony covering appears between the latter and the ventrorostrals. This discontinuity persists through the later stages markedly contrasting close juxtaposition of the ventrorostrals in the rostrum area. In most specimens this gap is being obliterated on 40–43 dph at 60–70 mm TL, whereas there are also fishes of 100 mm TL and even more, in which the empty space between the parasphenoid and the posterior ventrorostrals is still present (Figure 10b). Whatever the size and the age of the specimens, the closure of this space results mainly from the backward expansion of the posterior ventrorostrals, which grow out between the barbles and eventually meet the rostral end of the parasphenoid (the subrostral) at the prenasal level (Figures 9h and 10a), just where the posterior basirostral and the median anterior process of the parasphenoid (the subrostral) meet in sturgeon (see Figure 8a). Thus, based on the comparison with sturgeon, it is evident that in Polyodon the medial basirostrals are altogether wanting and that the place of these bones is occupied by the posterior ventrorostrals.

FIGURE 10.

FIGURE 10

Open in a new tab

Polyodon spathula, ventral views of the skull, visceral skeleton removed. (A) Photo of double‐stained skull (left) of the specimen of 70.6 mm TL, 46 dph (scale bar 4 mm), and the diagrammatic representation (right) based on of the same specimen. (b) Alizarin stained specimen of 100.6 mm TL, 65 dph (scale bar 4 mm) showing the broad gap between the parasphenoid and the posterior ventrorostrals. nc—nasal capsule; Nch—notochord; pr.asc.—ascending process of the parasphenoid; Prsp—parasphenoid; Sr—subrostral part of the parasphenoid; upd—upper pharyngeal dental plates; Vr—ventrorostrals

4. DISCUSSION

4.1. The composition of the parasphenoid and the ventral rostral series in A. baerii and P. spathula

Since the paper on osteology of Polyodontidae by Grande and Bemis (1991) it is generally agreed that Polyodon differs from sturgeons in the paired nature of the posterior bones in the ventral rostral series and by the parasphenoid without the median anterior process. These differences, although important in phylogenetic reconstructions (e.g., Findeis, 1997; Hilton et al., 2011), have never been the subject of developmental studies and never been addressed in terms of homology.

Data obtained in the present study allow to establishing primary homologies of the constituting parts of the parasphenoid and the ventral rostral series between A. baerii and P. spathula (see Figure 11). First of all, the position of the developing parasphenoid in Polyodon is identical to that of Acipenser and clearly demonstrates the homology of this bone in two species. Also, in both species there is a separate unpaired ossification, the subrostral, which arises in front of the hypophysis and later fuses with the anterior end of the parasphenoid. The only difference is that in the paddlefish it becomes almost as wide as the parasphenoid, while in sturgeon the subrostral remains narrow and elongate.

FIGURE 11.

FIGURE 11

Open in a new tab

Composition of the parasphenoid in Polyodon spathula (a) and A. baerii (b). Posterior division of parasphenoid—red; anterior division of parasphenoid—blue; ventrorostral series—green; the subrostral part of parasphenoid is marked by oblique hatching

The unpaired basirostral series of sturgeon is lacking in the paddlefish, and its space is occupied in the latter species by the enlarged posterior ventrorostrals. In both species posterior ventrorostrals are paired bones which arise in front of the barbels and the nasal capsules and likewise should be considered homologous. The same is true for the more anterior ventrorostrals, which in both species may lose their regular paired arrangement (see figures 7b and 6 in Grande & Bemis, 1991).

Thus, the main differences between sturgeons and the paddlefish are the lack of the basirostrals in Polyodon and the greater width of the subrostral in the latter species. These differences are considered below, in the section on polyodontids evolution (Section 4.5).

4.2. Homology of the parasphenoid of acipenseriforms

Since Bridge (1878) and Parker (1882) first described the parasphenoid in Polyodon and Acipenser respectively, the homology of this bone of acipenseriforms has never been questioned (see de Beer, 1937; Findeis, 1993; Hilton et al., 2020, 2011; Jollie, 1980; Pehrson, 1944; Sewertzoff, 1926, 1928; Warth et al., 2017).

To address the homology of parasphenoid and ventral rostral bones of acipenseriforms, it is necessary first to consider the development of the parasphenoid in other actinopterygians.

In Amia calva, as described by Pehrson (1922, 1940) the parasphenoid develops from the unpaired anterior and the paired posterior rudiments (see Figure 12), which later fuse in a single bone. The anterior rudiment (Ps.a.) ossifies beneath the basicranial (hypophyseal) fenestra (f.bc.), in front of the hypophysis (Hyp) and the notches in the trabecles (i.ci.), through which the internal carotids enter the neurocranium. The posterior rudiment (Ps.p.) develops behind the hypophysis on the sides of the notochord (Nch).

FIGURE 12.

FIGURE 12

Open in a new tab

Development of the parasphenoid in Amia calva (adopted from Pehrson, 1940; figures 38, 39). Ventral views of the skull of the specimens of 9.5 mm (a) and 10.4 mm (b). Dpl—dermopalatine; f.bc.—basicranial fenestra; Hyp—hypophysis; i.ci.—notch for internal carotid artery; Nch—notochord; Ps.a.—anterior rudiment of the parasphenoid parasphenoid; Ps.p.—posterior rudiment of the parasphenoid; Vo—vomer

A similar pattern was reported also for Lepisosteus (Hammarberg, 1937, p. 319).

In Polypterus the earliest stages of the parasphenoid development are yet to be described, however Pehrson (1958) definitely figured and designated the paired posterior rudiments of this bone on his reconstruction of the skull of 8 mm Polypterus (Pehrson, 1958, Figure 6, ps). Topographically, they correspond to the posterior rudiment of this bone of Amia. In 13.5 mm TL larvae of Polypterus described by Wacker et al. (2001) the parasphenoid “is a broad, thin bony plate surrounding the hypophyseal opening (Rathke's pouch). It shows two distinct posterior (occipital) processes that reach behind the occipital arches” (Wacker et al., 2001, p. 40).

Thus, in Amia and Lepisosteus, and most likely in Polypterus, the parasphenoid consists developmentally from the anterior unpaired and the posterior paired parts, of which the anterior part arises beneath the basicranial fenestra, while the posterior one develops in the parachordal division of the neurocranium, on the sides of the notochord. The hypophysis and the foramens for the internal carotids provide the reference point of demarcation between these two constituting parts of the parasphenoid.

In teleosts, the parasphenoid develops from the unpaired ossification, which arises in the basicranial fenestra, while the posterior part of the bone is said to be partially or completely replaced by the endochondral basioccipital (Vandewalle et al., 2005 and the references therein).

As described above, the so‐called parasphenoid of sturgeon and paddlefish arises from the paired rudiments situated posteriorly to the hypophysis on the sides of the notochord. Thus, topographically and in general appearance they fully correspond to the posterior rudiments of this bone in Amia, Lepisosteus, and Polypterus. This leaves no other options than to conclude that the so‐called parasphenoid in sturgeon and the paddlefish is actually not the parasphenoid sensu stricto, but that it corresponds to the posterior part of this bone only.

However, in sturgeons, there is a series of unpaired bones, which ossify anteriorly to the hypophysis, in the area, where the basicranial fenestra existed at earlier developmental stages (see Figure 2, f.bc.). These are the subrostral and the basirostrals. Consequently, both topographically and by their unpaired makeup, these bones en masse correspond to the anterior division of the parasphenoid of Amia and other actinopterygians.

In Polyodon, the basirostrals are altogether missing (see above) and the only unpaired ossification which develops in front of the hypophysis is the subrostral. Thus, in Polyodon the anterior division of the parasphenoid is represented by the subrostral only (see Figure 11a).

Unlike the parasphenoid, the ventral rostral bones of acipenseriforms have been the subject of some discussion in the literature. Bridge (1878) considered the posterior ventrorostrals of Polyodon to be the vomers. However, according to Allis (1919) the “vomer” in Polyodon lies wholly anteriorly to the buccal cavity, “and if it actually be a vomer, its anomalous position needs explanation” (Allis, 1919, p. 372). Jollie (1980) used the same argument and concluded, that “this homology cannot be supported by any real facts” (Jollie, 1980, p. 232; see also Patterson, 1975, p. 513; Findeis, 1993). Grande and Bemis (1991) followed Bridge in naming the posterior ventral rostral bones of Polyodon but provided no discussion of their homology. At present, the ventral rostral series is generally considered the synapomorphy of Acipenseridae +Polyodontidae without homologs outside the acipenseriforms (Findeis, 1997; Grande et al., 2002; Hilton et al., 2011).

In actinopterygians, the vomers are situated immediately in front of the parasphenoid in the ethmoid region of the skull. They are plesiomorphically paired in Amia, Lepisosteus and many non‐teleost fossils and secondarily unpaired in most teleosts and their fossil allies (for thorough reviews see Gardiner, 1984b; Grande, 2010; Wacker et al., 2001 for conditions in Polypterus).

Given the subrostral+basirostrals in sturgeons and the subrostral alone in paddlefish represent the anterior division of the parasphenoid, the posterior ventrorostrals, which are definitely paired and which are situated in front of this division, should be the vomers, as originally proposed by Bridge (1878).

The bones which are situated in front of the posterior ventrorostrals are more difficult to interpret because of their more unstable arrangement. As discussed above, in some specimens of A. baerii the pairwise order of this series is disrupted and the bones are arranged in an alternating chessboard pattern (Figure 7b). The same is observed in the specimens of Polyodon, illustrated in Grande and Bemis (1991, Figure 6). Moreover, in Polyodon, this series comprises the medial row of bones (“p”‐bones of Grande & Bemis, 1991), which are evidently unpaired. Nevertheless, all these bones belong to the same series as the invariably paired posterior ventrorostrals and, so far, it seems logical to consider them as vomerine series until further information becomes available to test this hypothesis.

Before continuing further discussion, it should be noted that proposed homologies imply disintegration of the ancestral parasphenoid (its anterior part) and vomers into numerous smaller ossifications. Such a breakdown of a single bone into smaller ones, although not a widespread pattern among actinopterygians, still occurs, for instance, in the fragmented lacrimomaxillary bones in fossil and recent gars (see Grande, 2010). In acipenseriforms, this process was apparently associated with significant elongation of the snout, which accompanied the origin of this group (see Hilton et al., 2011), and was facilitated by the liberation of the parasphenoid and vomers from the functional constraints inherent to the bones of the palate.

The general scheme of the homologies of the parasphenoid and the ventral rostral series suggested in the present account is summarized in the scheme in Figure 13. This scheme clearly shows, that despite fragmentation into separate ossifications, the general topological arrangement of the anterior division of the parasphenoid (subrostral+basirostrals) and vomers (ventrorostrals) in acipenseriforms fits well into the common actinopterygian pattern.

FIGURE 13.

FIGURE 13

Open in a new tab

Homologies of the parasphenoid and vomers across actinopterygians based on the present study. Ventral views of the skull (visceral skeleton removed). (a) Acipenser baerii (based on the present study and Tsessarsky, 2019, Figure 10). (b) Amia calva (based on Grande and Bemis, 1998, figures 38 and 50). (c) Saurichthys ornatus (based on Stensio, 1925, text‐fig. 65). Posterior division of the parasphenoid—red; anterior division of the parasphenoid—blue; vomers—green; autopalatine part of the palatoquadrate—yellow; arrowheads point to the premaxillare (single) and the maxillare (double) in Amia calva and Saurichthys ornatus. See explanations in the text

4.3. On the ventral process of the basirostrals in sturgeons

As discussed above, each basirostral in sturgeon bears a very prominent ventral process of a quite unusual shape (see Figure 6d,e, vpBr). It is built of a rounded basal stem, which is embedded into the thick skin, underlining the snout, and the flattened distal pad, which projects to the epidermis and is sculptured with irregular ridges. Hilton and Forey (2009) and Hilton et al. (2011) reported the presence of this process exclusively in three sturgeon species: A. baerii, A. ruthenus, and Huso huso, and suggested these prominences to be a synapomorphy of these species (Hilton et al., 2011, p. 147). However, quite similar protrusions (oval in shape with a flattened ornamented ventral surface) are found in other acipenserid species as well. They were reported in A. fulvescens and A. ruthenus by Jollie (1980, p. 232) and in A. sinensis by Hilton et al. (2015, p. 7). In the material at the author's disposal, distinct ventral prominences are present on the basirostral in A. gueldenstaedtii, A. baerii, A. mikadoi, A. naccarii, A. ruthenus, A. stellatus and Huso. Also, the drawings of the skulls of A. fulvescens and A. stellatus in Hilton et al. own paper (Hilton et al., 2011, fig. 120b, 121b; see also figure 6C in Hilton et al., 2020) suggest that these protrusions are present in these species too, although less developed than in other sturgeons. In any case, the distribution of this character among sturgeons is much wider than expected by Hilton et al. (2011), and requires further study.

Whether the presence of the ventrorostral prominences is synapomorphic to acipenserids, or whether they were independently acquired in different sturgeon lineages, they represent an interesting case of an evolutionary novelty. To the author's knowledge, no hypotheses were proposed concerning the origin and the primary homology of these prominences. Meanwhile, in many actinopterygians the anterior division of the parasphenoid (to which the basirostrals are supposedly homologous) bears the toothed pad which is often lifted above the bone level. Such a lifted toothed pad is very prominent in cladistians (Allis, 1922; Claeson et al., 2007; Clemen et al., 1998), in Mesozoic leptolepids (Patterson, 1975, p. 530), and Triassic saurichthiid Yelangichthys (Wu et al., 2013). It is also present outside the actinopterygians, as in the Upper Devonian Eusthenopteron (Jarvik, 1954). Although the relations of the parasphenoid toothed pad to the surrounding tissues have not been described in detail, it seems obvious that its toothed surface protrudes to the outer epidermal layers, while the base of the pad is embedded within the surrounding dermis. This is exactly the condition found in the ventral process of the basirostral in sturgeons. To this it might be added, that the skin, which in sturgeons surrounds this process, develops in ontogeny from the endoderm (Minarik et al., 2017 and see below), as does the mucosa surrounding the toothed pad of the parasphenoid of other fishes.

Thus, the ventral process of the basirostrals is very likely a modified parasphenoid toothed pad, which lost teeth due to the loss of its initial function in the palatal makeup.

4.4. On the extraoral position of the anterior parasphenoid and vomers in acipenseriforms

As mentioned above, Allis (1919) was concerned by the anomalous extraoral position of the vomers in Polyodon. But he would have been much more perplexed to know that not only the vomers, but also the anterior division of the parasphenoid (the basirostrals) in acipenseriforms lay outside the oral cavity. Moreover as discovered by Minarik et al. (2017), the dermal covering of the barbles and the whole middle part of the underside of the snout in sturgeons develop not from the ectoderm, as it would be expected for the external craniofacial structures, but from the endoderm. It is therefore not only the preoral parasphenoid and vomers that need to be explained in acipenseriforms, but also the unique endodermal origin of the external preoral skin.

Such an explanation has recently been proposed by Tsessarsky (2019). Based on the structure and development of the skeletal components and associated ligaments and nerves of the snout and jaws in sturgeon, Tsessarsky (2019) suggested, that acipenseriforms have undergone paedomorphic underdevelopment of the lower jaw, which resulted in the posterior shift of the oral margin and the exclusion of the whole ethmoid region from the limits of the mouth cavity. During these transformations (see Tsessarsky, 2019, Figure 10) the “autopalatine” parts of the palatoquadrates have been detached from the jaws and moved forward to become the supporting cartilages of the barbles, the tentacular cartilages (colored in yellow on Figure 13a). The remaining parts of the palatoquadrates have turned midward to meet each other in a symphysis, retaining in this way the occlusion with the shortened mandibles. It is noteworthy, that besides the extraoral vomers and anterior parasphenoid this scenario also predicts the endodermal origin of the skin, which in acipenseriforms covers the ventral side of the snout and the barbles. This is exactly what was discovered in the study by Minarik et al. (2017) mentioned above.

Thus, the paedomorphic scenario in conjunction with the homologies proposed in the present account provide the very explanation that Allis claimed a century ago.

4.5. On the median anterior process of parasphenoid in acipenseriforms

In many acipenseriforms, the parasphenoid has a pronounced anterior projection, commonly referred to as the median anterior process (Grande & Bemis, 1991). This process is present in all living acipenserids (Hilton et al., 2011), in basal fossil acipenseriform Chondrosteus (Hilton & Forey, 2009) and in fossil Priscosturion (Grande & Hilton, 2006), probable sister‐group of Acipenseridae (Grande et al., 2002). It is absent in Polyodon, and probably absent in Psephurus and fossil Palaeopsephurus (Grande & Bemis, 1991), but present in fossil polyodontids Crossopholis (Grande & Bemis, 1991) and Protopsephurus (Grande et al., 2002). The occurrence of this process in fossil Peipiaosteus is doubtful (Hilton & Forey, 2009; but see Zhou, 1992, Figure 3; Jin, 1984, Figure 7), although it is present in Early Cretaceous Yanosteus (Hilton et al., 2020, Figure 4).

The width of the median anterior process also varies among acipenseriforms. It is long and narrow in Chondrosteus (Hilton & Forey, 2009, Figure 9), Protopsephurus (Grande et al. 2002, Figure 7) and all acipenserids except Huso, in which it is broad (Findeis, 1993: 313, fig. 56D). In Crossopholis it is very broad and only slightly differentiated from the parasphenoid (Grande et al 2002; see Grande & Bemis, 1991, figure 52). In Priscosturion the process in question is broad and truncated (Grande & Hilton, 2006, figures 12, 18). The median anterior process of the parasphenoid in Yanosteus is said to be “broader than that found in Acipenseridae” (Hilton et al., 2020, p. 7).

According to Hilton et al. (2011, p. 128), the distribution and variability of the median anterior process in acipenseriforms is rather controversial and needs further study.

In Polyodon, the median anterior process is said to have been phylogenetically lost (Grande et al., 2002; Hilton et al., 2011). However, as shown in the present study, the rostral end of the parasphenoid in this species develops from the separate bone, the subrostral, which in Acipenser becomes the median anterior process of parasphenoid (see Figure 11). The only difference is that in Polyodon the subrostral is almost of the same width as the parasphenoid itself. Consequently, the median anterior process of the parasphenoid has not been lost in Polyodon, but merely became as broad as the rest of the parasphenoid.

Occurrence of the subrostral fused to the parasphenoid in Acipenser and Polyodon sheds light on the conditions found in other acipenseriforms. Given the overall topological similarity of the parasphenoids across the acipenseriforms, it is safe to assume that the anterior part of this bone is developed by fusion with the subrostral not only in Acipenser and Polyodon, but in other acipenseriforms as well. Then, it follows that the presence or absence of the median anterior process in acipenseriforms depends on the width of the subrostral part of the parasphenoid.

As shown above, the width of the subrostral in A. baerii depends on the developing medial palatobasal process of the chondrocranium (Figure 4, arrowheads), which constrains lateral expansion of this bone. Therefore, those fossils in which the parasphenoid has the median anterior process (Chondrosteus, Priscosturion, Crossopholis, Protopsephurus), most probably possessed the cartilaginous medial palatobasal process as well. Conversely, no palatobasal process developed in the polyodontids Psephurus and Paleopsephurus, in which the anterior process of parasphenoid is lacking. Most likely, the development of the subrostral in these forms was similar to that of Polyodon (see, however, below on the rudimentary palatobasal process in this species).

Since in sturgeons, the medial palatobasal process is known to be linked to jaws protrusion (see Carroll & Wainwright, 2003; Meinel, 1962; Stengel, 1962), the above assumptions may be of some interest from a perspective of feeding biomechanics in extinct acipenseriforms. In particular, it follows, that Chondrosteus, Priscosturion and polyodontids Crossopholis and Protopsephurus possessed the protractile jaws similar to living sturgeons.

The last point to be made in the present discussion concerns the probable mechanism that appears to underlie the observed variation of the subrostral width among acipenseriforms.

The individual variability of the subrostral in larval A. baerii (Figure 5) indicates that at early developmental stages the shape of this bone correlates with the development of the medial palatobasal process. As shown on Figure 5, the subrostral is contracted in those parts where the medial palatobasal process has already formed, while it is distinctly broader where the palatobasal process has not yet developed. This suggests, that the developing palatobasal process restricts the lateral expansion of the subrostral.

This assumption gets support in the conditions found in Huso, in which the anterior process of the parasphenoid (the subrostral) is said to be distinctly broader than in other sturgeons (Findeis, 1993: 313, fig. 56D), while the medial palatobasal process is lesser developed leaving the anterior process of parasphenoid partially uncovered (Findeis, 1993, p. 375, figure 60A,B; Findeis, 1997, p. 101, figure 17b).

In Polyodon the situation is even more illustrative. In this species the medial palatobasal process is said to be lacking (Findeis, 1993; Grande & Bemis, 1991). In the present study, this process has not been observed in Polyodon neither in larval stages, nor in the specimens of 70–80 mm TL. Accordingly, nothing prevents the subrostral in the paddlefish to expand laterally and to attain almost the same width as the parasphenoid itself.

However, as it turns out, the conditions in Polyodon are not so straightforward. Bridge (1878) in his description of the skeleton of the subadult Polyodon (head length 220 mm) reported, that the lateral margins of the anterior part of the parasphenoid in this species “are slightly overlapped by the adjacent cartilage” (Bridge, 1878, p. 692). The accompanying figure (Bridge, 1878, Plate 56, Figure 4) clearly shows that the cartilaginous excrescences, which overlap the parasphenoid margins in Polyodon, are situated at the level of the nasal capsules, exactly where the medial palatobasal process develops in sturgeon. They are, therefore, well comparable with the cartilaginous swellings overlapping the margins of the subrostral at the initial stage of this process development in sturgeon. Since there is no reason to doubt the accuracy of Bridge's description and drawings, it is to be concluded that contrary to common views, Polyodon still possesses the medial palatobasal process, although in a rudimentary underdeveloped state. Thus, the broad subrostral of this species (“lack of the median anterior process of parasphenoid”) results from the extreme postdisplacement and underdevelopment of the medial palatobasal process.

The situation in Polyodon along with the above‐mentioned correlation between the subrostral shape and the extent of the palatobasal process development in Acipenser and Huso suggest that the width of the subrostral is a matter of timing: the later the palatobasal process encloses the subrostral, the more time remains for this ossification to expand laterally.

If the preceding is correct, then it follows, that in acipenseriforms with broad anterior process of the parasphenoid (e.g., Huso, Crossopholis, Priscosturion) the palatobasal process was retarded and underdeveloped relative to those with narrow anterior process (Acipenser, Chondrosteus and Protopsephurus).

The median anterior process of parasphenoid is considered the plesiomorphic feature of acipenseriforms lost in polyodontid lineage (Hilton et al., 2011). Since this process appears to be causally linked to the development of the medial palatobasal process of the chondrocranium, the latter process should also be considered the ancestral trait of the group. The tentative developmental mechanism responsible for the variability of the anterior parasphenoid process within acipenseriforms and for its disappearance in polyodontid evolution was the heterochrony: the postdisplacement and underdevelopment of the medial palatobasal process.

5. SUMMARY

The present study provides a detailed description of the development of the dermal bones underlining the cranial base in Acipenser baerii and P. spathula, the representatives of both extant families of Acipenseriformes. The data obtained in this research allowed the author to reconsider the composition of the parasphenoid of acipenseriforms and propose primary homologies of the ventral rostral bones. In particular, the early development of the parasphenoid in A. baerii and P. spathula clearly shows that this bone of acipenseriforms corresponds only to the posterior (posthypophyseal) division of the parasphenoid of other actinopterygians. The anterior division of the parasphenoid in acipenseriforms is represented by the unpaired posterior members of the ventral rostral series, while the bones situated farther anteriorly should be considered vomerine series. These results indicate that acipenseriforms are unique among actinopterygians and bony fishes in general in possessing the extraoral anterior division of the parasphenoid and the vomers and provide necessary morphological grounds for further studies of this striking novelty. Also, the developmental data on A. baerii and P. spathula allowed to address the ingroup diversity of the median anterior process of the parasphenoid, which is considered the plesiomorphic feature of acipenseriforms. It is shown in the present study, that this process develops from a separate ossification, the subrostral, which becomes fused to the parasphenoid. The width of this ossification is constrained by the cartilaginous outgrowths of the chondrocranium, which enclose the bone within the cartilage and eventually produce the medial palatobasal process. Based on these findings and on the variability of the median anterior process across acipenseriforms it is suggested that developmental mechanism responsible for the diversity of this character among acipenseriforms involved heterochrony: postdisplacement and underdevelopment of the medial palatobasal process.

This study demonstrates once again the morphological uniqueness of acipenseriforms and importance of this group not only for better understanding of the evolution of bony fishes, but also as a valuable model for studying morphological novelties.

ACKNOWLEDGMENTS

The author thanks Fedor Shkil (Severtsov Institute of Ecology and Evolution RAS) for assistance provided in sturgeon and paddlefish larvae raising. The author also thanks Elena Voronezhskaya from the Institute of Developmental Biology RAS for their help in obtaining confocal images using equipment of the Core Centrum of Institute of Developmental Biology RAS. Additionally, the author thanks two anonymous reviewers for their useful comments and suggestions. The study was supported by Russian Foundation for basic research, grant No 18‐04‐01301.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the author upon reasonable request.

REFERENCES

  1. Allis, E.P. (1919) The homologies of the maxillary and vomer bones of Polypterus . American Journal of Anatomy, 25, 348–394. [Google Scholar]
  2. Allis, E.P. (1922) The cranial anatomy of Polypterus, with special reference to Polypterus bichir . Journal of Anatomy, 56, 189–294. [PMC free article] [PubMed] [Google Scholar]
  3. Bemis, W.E. , Findeis, E.K. & Grande, L. (1997) An overview of Acipenseriformes. Environmental Biology of Fishes, 48, 25–72. [Google Scholar]
  4. Betancur, R. , Wiley, E.O. , Arratia, G. , Acero, A. , Bailly, N. , Miya, M. et al. (2017) Phylogenetic classification of bony fishes. BMC Evolutionary Biology, 17, 162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bridge, T.W. (1878) On the osteology of Polyodon Folium . Philosophical Transactions of the Royal Society of London, 169, 683–733. [Google Scholar]
  6. Barnard, N. (1987) Routine cytological staining techniques: Theoretical background and practice. Postgraduate Medical Journal 63, 158–158. [Google Scholar]
  7. Cardona, A. , Saalfeld, S. , Schindelin, J. , Arganda‐Carreras, I. , Preibisch, S. , Longair, M. et al. (2012) TrakEM2 software for neural circuit reconstruction. PLoS One, 7, e38011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Carroll, A.M. & Wainwright, P.C. (2003) Functional morphology of prey capture in the sturgeon, Scaphirhynchus albus . Journal of Morphology, 256, 270–284. [DOI] [PubMed] [Google Scholar]
  9. Claeson, K.M. , Bemis, W.E. & Hagadorn, J.W. (2007) New interpretations of the skull of a primitive bony fish Erpetoichthys calabaricus (Actinopterygii: Cladistia). Journal of Morphology, 268, 1021–1039. [DOI] [PubMed] [Google Scholar]
  10. Clemen, G. , Bartsch, P. & Wacker, K. (1998) Dentition and dentigerous bones in juveniles and adults of Polypterus senegalus (Cladistia, Actinopterygii). Ann Anat, 180, 211–221. [DOI] [PubMed] [Google Scholar]
  11. Cloutier, R. & Arratia, G. (2004) Early diversification of actinopterygians. In: Arratia, G. , Wilson, M.V.H. and Cloutier, R. (Eds.) Recent advances in the origin and early radiation of vertebrates. München, Germany: Verlag Dr. F. Pfeil, pp. 217–270. [Google Scholar]
  12. Coates, M.I. (1999) Endocranial preservation of a Carboniferous Actinopterygian from Lancashire, UK, and the interrelationships of primitive Actinopterygians. Philosophical Transactions: Biological Sciences, 354, 435–462. [Google Scholar]
  13. de Beer, G.R. (1925) Contributions to the development of the skull in sturgeons. Quarterly Journal of Microscopical Science, s2–69, 671–687. [Google Scholar]
  14. de Beer, G.R. (1937) The development of the vertebrate skull. New York: Oxford University Press. [Google Scholar]
  15. de Pinna, M.G. (1991) Concepts and tests of homology in the cladistics paradigm. Cladistics, 7, 367–394. [Google Scholar]
  16. Depew, M. (2008) Analysis of skeletal ontogenesis through differential staining of bone and cartilage. Methods in Molecular Biology, 461, 37–45. [DOI] [PubMed] [Google Scholar]
  17. Findeis, E.K. (1993) Skeletal anatomy of the North American shovelnose sturgeon Scaphirhynchus platorynchus (Rafinesque 1820) with comparisons to other Acipenseriformes. PhD dissertation. Ann Arbor: University of Massachusetts Amherst, pp. 1– 495. [Google Scholar]
  18. Findeis, E.K. (1997) Osteology and phylogenetic interrelationships of sturgeons (Acipenseridae). In: Birstein, V.J. , Waldman, J.R. & Bemis, W.E. (Eds.) sturgeon biodiversity and conservation. Dordrecht: Springer Netherlands, pp. 73–126. [Google Scholar]
  19. Gardiner, B.G. (1984a) Sturgeons as living fossils. In: Eldredge, N. & Stanley, S.M. (Eds.) Living fossils. New York/Berlin: Springer‐Verlag, New York Inc, pp. 148–152. [Google Scholar]
  20. Gardiner, B.G. (1984b) The relationships of the palaeoniscoid fishes, a review based on new specimens of Mimia and Moythomasia from the Upper Devonian of Western Australia. Bulletin of the British Museum of Natural History (Geology), 37(4), 173–428. [Google Scholar]
  21. Gardiner, B.G. , Schaeffer, B. & Masserie, J.A. (2005) A review of the lower actinopterygian phylogeny. Zoological Journal of the Linnean Society, 144, 511–525. [Google Scholar]
  22. Grande, L. (2010) An empirical synthetic pattern study of gars (Lepisosteiformes) and closely related species, based mosTLy on skeletal anatomy. The resurrection of Holostei. Copeia, 2010, 1–871. [Google Scholar]
  23. Grande, L. & Bemis, W.E. (1991) Osteology and phylogenetic relationships of fossil and recent Paddlefishes (Polyodontidae) with comments on the interrelationships of Acipenseriformes. Journal of Vertebrate Paleontology, 11, 1–121. [Google Scholar]
  24. Grande L, & Hilton EJ. (2006) An exquisitely preserved skeleton representing a primitive sturgeon from the upper cretaceous judith river formation of Montana (Acipenseriformes: Acipenseridae: n. gen. and sp.). Journal of Paleontology, 80, 1–39. [Google Scholar]
  25. Grande, L. , Jin, F. , Yabumoto, Y. & Bemis, W.E. (2002) †Protopsephurus liui, a well‐preserved primitive paddlefish (Acipenseriformes: Polyodontidae) from the Lower Cretaceous of China. Journal of Vertebrate Paleontology, 22, 209–237. [Google Scholar]
  26. Hammarberg, F. (1937) Zur Kenntnis Der Ontogenetischen Entwicklung Des Schädels Von Lepidosteus platystomus . Acta Zoologica, 18, 209–337. [Google Scholar]
  27. Hilton, E.J. , Dillman, C.B. , Paraschiv, M. & Suciu, R. (2020) Cranial morphology of the Stellate Sturgeon, Acipenser stellatus Pallas 1771 (Acipenseriformes, Acipenseridae), with notes on the skulls of other sturgeons. Acta Zoologica, 00, 1–21. [Google Scholar]
  28. Hilton, E.J. & Forey, P.L. (2009) Redescription of †Chondrosteus acipenseroides Egerton, 1858 (Acipenseriformes, †Chondrosteidae) from the lower Lias of Lyme Regis (Dorset, England), with comments on the early evolution of sturgeons and paddlefishes. Journal of Systematic Palaeontology, 7, 427–453. [Google Scholar]
  29. Hilton, E.J. , Grande, L. & Bemis, W.E. (2011) Skeletal Anatomy of the Shortnose Sturgeon, Acipenser brevirostrum Lesueur, 1818, and the Systematics of Sturgeons (Acipenseriformes, Acipenseridae). Fieldiana Life and Earth Sciences, 3, 1–168. [Google Scholar]
  30. Hilton, E.J. , Dillman, C.B. , Zhang, T. , Zhang, L. , Zhuang, P. (2015) The skull of the Chinese sturgeon, Acipenser sinensis (Acipenseridae). Acta Zoologica (Stockholm), 97, 1–14.
  31. Hilton, E.J. , Grande, L. , Jin, F. (2020a) Redescription of †Yanosteus longidorsalis Jin et al., 1995 (Chondrostei, Acipenseriformes, †Peipiaosteidae) from the Early Cretaceous of China. Journal of Paleontology, 95(1), 1–14. [Google Scholar]
  32. Jarvik, E. (1954) On the visceral skeleton in E̲u̲s̲t̲h̲e̲n̲o̲p̲t̲e̲r̲o̲n̲ with a discussion of the parasphenoid and palatoquadrate in fishes. Kungliga Svenska Vetenskapsakademiens Handlingar, Ser. 4, 5, 1–104. [Google Scholar]
  33. Jin, F. (1999) Middle and late mesozoic acipenseriformes from northern hebei and western liaoning, China. In: Chen, P.‐J. and Jin, F. (Eds.) Hefei, China: Press of University of Science and Technology of China, pp. 188–260. [Google Scholar]
  34. Jollie, M. (1980) Development of head and pectoral girdle skeleton and scales in Acipenser . Copeia, 1980(2), 226–249. [Google Scholar]
  35. Jollie, M . (1984) Development of the head and pectoral skeleton of Amia with a note on the scales. Gegenbaurs morphologisches Jahrbuch Leipzig, 130, 315–351. [PubMed] [Google Scholar]
  36. Meinel, W. (1962) Zur Morphologie Und Funktion Des Splanchnocranium Von Acipenser . Zeitschrift für Morphologie und Ökologie der Tiere, 51, 1–62. [Google Scholar]
  37. Minarik, M. , Stundl, J. , Fabian, P. , Jandzik, D. , Metscher, B.D. , Psenicka, M. et al. (2017) Pre‐oral gut contributes to facial structures in non‐teleost fishes. Nature, 547, 209–212. [DOI] [PubMed] [Google Scholar]
  38. Parker, W.K. (1882) On the structure and development of the skull in sturgeons (Acipenser ruthenus and A. sturio). Philosophical Transactions of the Royal Society of London, 173, 139–185. [Google Scholar]
  39. Patterson, C. (1975) The braincase of Pholidophorid and Leptolepid Fishes, with a review of the Actinopterygian braincase. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 269, 275–579. [DOI] [PubMed] [Google Scholar]
  40. Patterson, C. (1982) Morphology and Interrelationships of primitive Actinopterygian fishes. American Zoologist, 22, 241–259. [Google Scholar]
  41. Pehrson, T. (1922) Some points in the cranial development of Teleostomian fishes. Acta Zoologica, 3, 1–63. [Google Scholar]
  42. Pehrson, T. (1940) The development of dermal bones in the skull of Amia calva . Acta Zoologica, 21, 1–50. [Google Scholar]
  43. Pehrson, T. (1944) Some observations on the development and morphology of the dermal bones in the skull of Acipenser and Polyodon . Acta Zoologica, 25, 27–48. [Google Scholar]
  44. Pehrson, T. (1958) The early ontogeny of the sensory lines and the dermal skull in Polypterus . Acta Zoologica, 39, 241–258. [Google Scholar]
  45. Sewertzoff, A.N. (1926) Studies on the bony skull of fishes. I. Structure and development of the bony skull of Acipenser ruthenus . Quarterly Journal of Microscopical Science, 70, 451–540. [Google Scholar]
  46. Sewertzoff, A.N. (1928) The head skeleton and muscles of Acipenser ruthenus . Acta Zoologica, 9, 193–319. [Google Scholar]
  47. Stengel, F.F. (1962) Untersuchungen am Kopf, besonders am Bänderapparat, des Sterlets (Acipenser ruthenus). Revue Suisse De Zoologie, 69, 513–557. [Google Scholar]
  48. Stensio, E.A. (1925) Triassic fishes from Spitzbergen. Part II. Kungl. Svenska Vetenskapsakademiens Handlingar, 2(1), 1–261. [Google Scholar]
  49. Tsessarsky, A. (2019) The role of the paedomorphosis in the emergence of the bauplan of the skull in Acipenseriformes (Actinopterygii). Journal of General Biology, 80, 334–347.[in Russian] [Google Scholar]
  50. Vandewalle, P. , Germeau, G. , Besancenet, P. , Parmentier, E. & Baras, E. (2005) Early development of the head skeleton in Brycon moorei (Pisces, Ostariophysi, Characidae). Journal of Fish Biology, 66, 996–1024. [Google Scholar]
  51. Wacker, K. , Bartsch, P. & Clemen, G. (2001) The development of the tooth pattern and dentigerous bones in Polypterus senegalus (Cladistia, Actinopterygii). Annals of Anatomy, 183, 37–52. [DOI] [PubMed] [Google Scholar]
  52. Warth, P. , Hilton, E.J. , Naumann, B. , Olsson, L. & Konstantinidis, P. (2017) Development of the skull and pectoral girdle in Siberian, sturgeon, Acipenser baerii, and Russian sturgeon, Acipenser gueldenstaedtii (Acipenseriformes: Acipenseridae). Journal of Morphology, 278(3), 1‐25. [DOI] [PubMed] [Google Scholar]
  53. Wu, F. , Chang, M.‐m. , Sun, Y. & Xu, G. (2013) A new Saurichthyiform (Actinopterygii) with a crushing feeding mechanism from the Middle Triassic of Guizhou (China). PLoS One, 8, e81010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zhou, Z. (1992) Review on Peipiaosteus based on new material of P. pani . Vertebrata PalAsiatica 30, 85–101. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data that support the findings of this study are available from the author upon reasonable request.

Articles from Journal of Anatomy are provided here courtesy of Anatomical Society of Great Britain and Ireland

RESOURCES