Pathological caudal skeleton of an ichthyodectiform fish from the Upper Cretaceous Niobrara Formation of western Kansas, USA

Description

YPM VP 42619 consists of a series of 12 contiguous vertebral centra, two of which are free (i.e., not fused to adjacent centra) and ten of which are coossified, most of which have broken bases of neural and haemal spines preserved in articulation with the centra (Figs. 1 and 2), and four small neural and haemal spine fragments. The posteriormost vertebra is interpreted as the first ural, the other nine vertebrae of the coossified series are interpreted as preural vertebrae 1-9, and the two free vertebrae are preurals 10 and 11. The entire series of 12 vertebrae exhibits extensive overgrowth of pathological bone tissue. The anteriormost free vertebra is 34.4 mm wide × 27.7 mm high × 18.7 mm long, the posterior free vertebra is 35.6 × 31.0 × 18.4 mm, and the two free vertebrae have a combined length of 36.9 mm. The series of ten coossified vertebrae is 172.8 mm long, and is broken into three sections: an anterior section of three vertebrae, 59.9 mm long (∼20 mm/centrum), 34.4 × 36.5 mm at the anterior end, and 34.2 × 29.6 mm at the posterior end, with little taper; a middle section of four vertebrae 69.3 mm long (∼17.3 mm/centrum) and 25.7 × 23.0 mm at the posterior end, with somewhat more taper; and a posterior section of three vertebrae 43.6 mm long (∼14.5 mm/centrum) and 13.4 mm × 16.2 mm at the posterior end, with significant taper and a marked upward curvature posteriorly. The combined length of all 12 vertebrae is 209.7 mm. At the breaks between the sections of the coossified centra, calcite and/or bone tissue can be seen filling the biconical space between the cotyles of the adjacent vertebrae where the intervertebral disk would have been. The posteriormost vertebra has a small broken and distorted posterior cotyle, 12.6 × 9.8 mm as preserved, which indicates that there was another vertebra posterior to it, which is why the posteriormost vertebra is interpreted as the first ural vertebra and the subsequent vertebra would have been the second ural. The fragments of neural and haemal spines not articulated with the vertebral centra consist of three 2−2.5 cm long sections of individual spines and one ∼3 cm long section of ∼5 adhering spines, which show little evidence of pathology.

The bone is mostly brown, though paler in places, presumably as a result of weathering, and parts, particularly the dorsal surfaces, are encrusted with orange ferruginous deposits. Masses of somewhat darker ferruginous particulate matter with an appearance that resembles sintered porous bronze are present within some depressions on the centra (Fig. 3A). The material seems not to be osseous, and presumably was derived from the surrounding matrix, but it resisted removal by the air abrasive when the specimen was prepared.

Anterior and posterior views of preural vertebra 11 show that the centrum is amphicoelous with an hourglass shaped cross-section typical of fishes (Fig. 1). In lateral view, the contours of the external surface of the centrum are evident despite the bony overgrowth, and include a prominent lateral longitudinal ridge bounded above and below by longitudinal grooves. That morphology is typical of ichthyodectiform fishes (Bardack, 1965; Cavender, 1966), which are represented in the Smoky Hill Chalk by Ichthyodectes, Gillicus, and Xiphactinus, and the morphology of the caudal skeleton of YPM VP 42619 compares well with the ‘Ichthyodectes type’ caudal skeleton described and illustrated by Cavender (1966), Fig. 1.

The outer margin of the anterior cotyle of preural vertebra 11 is 22.4 mm wide × 21.2 mm high, but the centrum proper is surrounded by ∼5 mm of bony overgrowth. That bone has a dense surface with an intricate reticulate pattern of ridges and grooves and many small pores opening onto the surface (Fig. 3A). Preural vertebra 10 is similar to preural 11 and is shown articulated with preural 11 in Fig. 1C. There is a gap between the outer parts of the overgrowing bone at the posterior end of preural 11 and the anterior end of preural 10, and the bony overgrowth on either side of the gap is thickened and coarsely textured with projections and pits. Because there was no contact between the bony overgrowth of the two vertebrae, the bony overgrowth would not have prevented movement between the two vertebrae. However, some bone tissue is present in the intervertebral space between preurals 10 and 9, which suggests that movement between those vertebrae was reduced.

The centra of the anterior section of the coossified series of vertebrae also exhibit gaps between the outermost bony overgrowth of adjacent centra (Fig. 2). However, the gaps are narrower than those between the free vertebrae, ∼2 mm or less, and at their bottoms there are sinuous grooves where the posterior surface of the bony overgrowth of the anterior vertebra meets the anterior surface of the overgrowing bone of the posterior vertebra. The grooves presumably are sinuous because they are between the bony overgrowths of the centra rather than the centra themselves and the rate of deposition of the bony overgrowth was not uniform. In some places, the bony overgrowth of adjacent vertebrae is fused and the groove obliterated, but despite the fact that in most places the bony overgrowth is not fused, the close correspondence of the shapes of the surfaces of the adjacent vertebrae meeting at the grooves suggests that no significant movement was possible (Fig. 3C). The base of the haemal arch of preural vertebra 9 is better preserved than any other and exhibits some bony overgrowth, though to a lesser extent than the vertebral centra.

The extent of fusion of the centra increases posteriorly, and by the middle section of the coossified series of 10 vertebrae the gaps have disappeared and the narrow grooves between the bony overgrowth are at the surface of the bony overgrowth on each centrum (Fig. 3B). On the posterior section of the coossified series, the vertebrae are fully fused and the grooves obliterated. The external surface of the bony overgrowth on the posterior section appears very dense, but retains a little of the reticulate surface structure seen on the more anterior vertebrae. Examination of the intervertebral spaces between the three sections of coossified centra found that there was bone tissue within the spaces.

Longitudinal structures are preserved on the dorsal and ventral midlines of the centra, and are presumably associated with the neural and haemal canals (Figs. 2 and 3). The ventral structure passes between the articular surfaces for, or the bases of, the haemal spines, and is best seen on the ventral midline of preural vertebrae 7–9 (Fig. 3D). The structure is ∼2.5 mm wide over the middle of the centrum and ∼4.0 mm wide over the intervertebral joint. At the anterior end of preural vertebra 9, the structure has a flattened suboval cross-section and does not exhibit a lumen. A similar, but somewhat smaller dorsal longitudinal structure is preserved passing between the articular surfaces for, or the bases of, the neural spines. The structures cannot be interpreted as any normally osseous part of the vertebral column proper. It seems likely that they represent calcified or ossified soft tissues associated with the haemal and neural canals, though the fact that the ventral structure widens and narrows makes it unlikely that it is the caudal vein that passes through the haemal arches in fishes (Schultze & Arratia, 1989). However, an alternative interpretation would be that they are pathological bone tissue that preferentially invaded and grew along the canals. If the longitudinal structures represent soft tissues, their calcification or ossification presumably was related to the osseous pathologies of the caudal skeleton, but it is not clear whether the structures were calcified or ossified, and if so, how.

Thin section

Figure 4 shows a horizontal thin section of the left side of the centrum of preural vertebra 10 from the midline to the external surface. The conical anterior cotyle is intact and apparently was unaffected by pathology, whereas the medialmost ∼25% of the posterior cotyle is missing. Between the cotyles, normal cancellous bone infilled with calcite occupies much of the centrum’s volume (Fig. 4: cb), but the lateral part of the centrum is occupied by denser bone. The anterior cotyle exhibits fine striations parallel to its anterior surface that presumably are growth lines reflecting cyclical changes in the rate of deposition of bone tissue (Fig. 5A). A linear feature exhibiting similar striations that presumably were formed in a similar manner extends posteriorly at least four mm from the anterior cotyle’s lateral margin and is interpreted as the former external surface of the centrum upon which pathological bone tissue was deposited. Posteriorly, there are additional features farther back that may also represent the original external surface. Based on that interpretation, the bony overgrowth surrounding the centrum ranged from 4.0 to 5.3 mm thick. The bone tissue of the overgrowth is quite dense, but exhibits various bone textures. Much of the outermost ∼1 mm of bone tissue exhibits closely spaced slender subparallel canals, ∼30 µm in diameter with occasional rounded expansions, and subperpendicular to the external surface such that the sectioned tissue resembles a palisade (Fig. 5B). Examination of preural vertebra 10 reveals that the palisade-like tissue is associated with the presence of the reticulate pattern on the external surface (Fig. 3A), and the small pores that are part of the reticulate pattern of that surface may be openings of the slender canals. A dark line at the base of the palisade-like layer and subparallel to the external surface may be a growth? line. Deep to the palisade-like layer, the bone tissue is generally quite dense and without any discernible pattern, but there are many spaces that are interpreted as large vascular canals. In places there are also slender subparallel canals that may represent a second palisade-like layer, and deep to them a second dark growth? line. Traces of a third layer of similar tissue with subparallel canals are visible to the left in Fig. 5B, deep to the second layer. It is not clear whether the specimen preserves successive layers of palisade-like tissue with the deeper layers being less distinct because of remodeling or the specimen records increasing expression of the palisade-like character in successive layers deposited sequentially.

The thin section also shows bone tissue in the biconical intervertebral space where the intervertebral disk would have been. The tissue is cancellous, but denser than the normal cancellous bone within the vertebral centrum that was described above. The intervertebral bone appears to be continuous with bone tissue that extends through the gap in the posterior cotyle and into the centrum near the midline (Fig. 6). The latter bone tissue exhibits a rather linear character with subparallel bands of bone and several rather large roughly longitudinal vascular canals that converge anteriorly, extend through the gap, and diverge within the centrum, where they appear to be continuous anteriorly with rather dense cancellous bone tissue similar to that near the middle of the intervertebral space. It is not clear which direction the pathological bone may have grown, but it seems more likely that it was growing into the centrum from the intervertebral space, in which case, pathological bone may have been growing anteriorly through successive centra and intervertebral spaces, with what is seen in preural vertebra 10 being the leading end of what might have continued to grow into preural vertebra 11 if the fish had lived longer. One last feature of note is a large suboval void surrounded by a ring of rather dense dark appearing bone and filled with fractured calcite, which is lateral to the pathological bone passing through the gap in the posterior cotyle and just anterior to the remaining part of the cotyle (Figs. 4, 6). Presumably the void represents an abscess that resulted from an infection that destroyed all trabeculae of the cancellous bone within the ring of dense bone, but it is not clear how that process was related to the deposition of the pathological bone of the ring and elsewhere.