4.11. Mandible and dentition

USNM 539939 preserves elements of as few as seven and as many as nine teeth with both roots and crowns, as well as incomplete tooth roots. All of the crown-bearing teeth show thin, raised striations down their length on the lingual surface extended upward from the base of the crown towards but not on the apex. Based on comparison with other toothed mysticetes such as Salishicetus, Aetiocetus, Fucaia and Janjucetus, we conservatively interpret six right lower postcanine teeth and three indeterminate teeth. The morphology of the accessory cusps and the curvature of the tooth roots support this interpretation, as does the presence of right mandibular elements but few rostral elements. However, the possibility that the six identified teeth instead represent left upper postcanine teeth cannot be ruled out. All of the teeth preserving crowns preserve an enamel shelf on the lingual surface, identified here as the entocingulum. There is no evidence of an ectocingulum in any tooth.

Tooth A (figure 5a) was figured by Kiel et al. [12] (figure 2e). This tooth preserves half of the crown, a small portion of one root and a nearly complete second root. The crown preserves the apical cusp, which shows minimal apical wear, and four accessory cusps with heavy occlusal wear. This occlusal wear results in the accessory cusps being blunt and poorly differentiated from one another, reminiscent of those seen in the more heavily worn teeth of S. meadi. The roots are fully separated and bear a strong lingual and slight posterior curvature. As noted by Kiel et al. [12], the root of tooth A preserves at least three distinct Osedax boreholes.

Tooth B (figure 5b) represents a single half of the tooth crown. It preserves the apical cusp with no apical wear, and either four or five accessory cusps with only minimal occlusal wear. The breakthrough of the crown reveals that the enamel layer appears thin.

Tooth C (figure 5c) preserves a crown and both roots which curve labially. The roots are nearly completely fused and probably were preserved in a single alveolus lacking an isthmus. In this, tooth C resembles the possible first premolar of S. meadi. The roots narrow transversely as they near the crown such that the crown is actually transversely wider than the root beneath it. The apical cusp is preserved and worn to a blunt tip, and the accessory cusps preserve heavy occlusal wear such that the number of accessory cusps is uncertain. Tooth C does not preserve any obvious Osedax boreholes.

Tooth D (figure 5d) is the most complete and best-preserved tooth of B. osedax. It was figured by Kiel et al. [12] (figure 2c) and used in the complexity analysis by Peredo et al.[6]. The tooth is double rooted, with the proximal ends of the roots partially broken since they were originally figured by Kiel et al. [12], and the roots have a slight labial and posterior curvature. The two roots are separate for most of their length but fuse just ventral to the crown, though they remain distinguished by a deep groove that continues into the crown and gives it a slight heart-shape in labial view. The crown preserves a complete apical cusp, four anterior accessory cusps and five posterior accessory cusps, all of which demonstrate minimal occlusal wear. The accessory cusps are overall triangular and the divisions between the cusps are shallow relative to other toothed mysticetes such as Coronodon and Salishicetus. As noted by Kiel et al. [12], tooth D preserves several Osedax boreholes throughout its roots.

Tooth E (figure 5e) preserves most of the crown and both roots. As with tooth D, the roots are separate for most of their length and fuse only near the crown. They also exhibit a slight labial and posterior curvature at their proximal terminations. As with tooth C, the accessory cusps show occlusal wear, making it difficult to distinguish the number of anterior accessory cusps, though we interpret four posterior accessory cusps. Though not figured by Kiel et al. [12], tooth E does preserve several boreholes.

Teeth F–I (figure 5f–i) are tooth roots only with no crowns preserved. Of these, only tooth H preserves a double tooth root, but it is unclear if teeth F, G and I are single-rooted teeth or simply single crowns of double-rooted teeth. All four teeth preserve a slight posterior and labial curvature at their proximal terminations. Tooth H was figured by Kiel et al. [12] (figure 2d) and preserves several boreholes throughout both roots. The other three teeth do not preserve any obvious boreholes.

The right dentary of USNM 539939 is in several pieces, including one larger posterior section and one smaller anterior section, with the latter so poorly preserved that pieces of it crumble readily. Posteriorly, the mandible preserves an articular condyle that is round and kobold-like with a deeply pitted and rugose surface. In posterior view, the articular condyle is keeled dorsally and broad transversely. Anterior to the articular condyle, the body of the mandible thins quickly so that the condyle appears to be connected by a thin neck to what probably was the rest of the mandible corpus. Anterior to this level, the lateral surface of the mandible is preserved but is badly damaged, so that any morphology is obscured.

A second anterior piece preserves approximately 156 mm of the alveolar surface. This mandibular element preserves alveoli for at least 12 tooth roots. Based on size, orientation and spacing, we interpret these alveoli to correspond to three single-rooted teeth and three double-rooted teeth. Deep breaks in the body of the mandible expose the internal morphology, revealing that the mandibular canal is low in the body of the mandible and the tooth roots penetrate to the level of the mandibular canal, as in most toothed cetaceans [24].

5. Discussion

5.1. Phylogenetic position and taxonomic identity

The skull of Borealodon is poorly preserved, making phylogenetic coding difficult. In general, the skeletal material for genotypic species of Aetiocetidae are largely non-overlapping: many named taxa lack isolated tympanoperiotics, complete crania, dentition or mandibles. For example, there are no described periotics for Aetiocetusspp.; equally, there is no cranial material aside from the elements directly surrounding the tympanic cavity in Salishicetus. Testing the position of Borealodon relative to mammalodontids is particularly difficult because many of the mammalodontid synapomorphies relate either to the rostrum or to the bizygomatic width, neither of which is preserved nor can be estimated.

Our analysis does not include recently described stem mysticetes such as Mystacodon, Llanocetus, Coronodon and Niparajacetus, which have not been observed by the authors. However, respective phylogenetic analyses recover these taxa basal to Mammalodontidae, and are therefore unlikely to affect the phylogenetic position of Borealodon. Moreover, the size, spacing and overall dental morphology of all of these aforementioned taxa do not correspond to Borealodon.

The most recent analysis of mysticete phylogeny using this matrix [15] recovered Mammalodontidae (Mammalodon + Janjucetus) as a monophyletic clade based on the following 10 synapomorphies: (i) the premaxilla exposed in the palate only anterior to the maxilla (character 11, state 1); (ii) a short rostrum that is less than 40% of the condylobasal length (character 15, state 1); (iii) a posterior portion of the maxilla situated lateral to the nasal (character 20, state 1); (iv) a wide antorbital notch measuring between 40 and 54% of the bizygomatic width (character 42, state 0); (v) an anterior edge of the narial fossa located in the anterior quarter of the rostrum (character 47, state 1); (vi) a broad rostrum overall measuring over 80% of the rostrum length (character 49; state 1); (vii) short nasals measuring less than 40% of the bizygomatic width (character 59, state 1); (viii) a preorbital process of the frontal that is rounded off with an anteriorly convex margin (character 83, state 1); (ix) an anterior process of the periotic preserving a deep pit on the lateral surface (character 166, state 0); (x) a shallow or absent suprameatal fossa of the periotic (character 218, state 1).

Our phylogenetic analysis recovers similar results; we recover a monophyletic Mammalodontidae based on eight of the aforementioned synapomorphies (characters 11, 15, 20, 42, 47, 49, 59 and 218) and one new synapomorphy: a tympanic bulla with a medial and lateral lobe approximately equivalent in width (character 244) (figure 11). Two previous synapomorphies of Mammalodontidae (characters 83 and 166) no longer inform the monophyly of Mammalodontidae because they are shared with Borealodon. Borealodon can be coded for three of the nine diagnostic characters of Mammalodontidae recovered by this analysis; all three (characters 20, 218 and 244) exclude Borealodon from Mammalodontidae (see diagnosis).

The other well-known clade of toothed, stem mysticetes from the Pacific Northwest region are the aetiocetids. Peredo&Pyenson [15] identified seven synapomorphies of Aetiocetidae. Our analysis recovers a monophyletic Aetiocetidae based on four synapomorphies: characters 173, 180, 190 and 244. Of these, Borealodon can be coded for three (characters 180, 190 and 244), and all three exclude Borealodon from Aetiocetidae (see diagnosis).

Phylogenetic relationship of B. osedax relative to other stem Mysticeti. Cladogram shown is a strict consensus tree. Crown Mysticeti has been collapsed into a single terminal unit for visualization, but relationships therein are unchanged from Peredo&Pyenson [15].

• Download figure
• Open in new tab
• Download powerPoint

This combination of character states results in Borealodon representing a distinct lineage of stem mysticete, crown-ward of Mammalodontidae and stem-ward to Aetiocetidae. This result is recovered in all 359 best score trees (best score of 1595 steps). Thus, despite the challenges in phylogenetic coding of stem mysticetes, there is evidence that Borealodon is neither a mammalodontid nor an aetiocetid. Other toothed, stem mysticetes such as Llanocetus, Mystacodon, Coronodon or Niparajacetus [25] are a poor fit for its size and observed morphology.

5.2. Hearing

We tested the low-frequency limit of hearing following the methods of [26]. The lower limit of hearing frequency in cetaceans can be estimated based on the ratio of the radius of the cochlea at its base with the radius of the cochlea at its apex. Notably, we measured this ratio in Borealodon as 6.44; Park et al. [26] report an identical value for NMV P173220—identified by the authors as Mammalodontidae indet. The value for Borealodon is also comparable with the value of 6.77 reported for M. colliveri. Following the methods of [26], we estimate a low-frequency limit of 64.88 Hz; these values are in line with the estimates for NMV P173220 and M. colliveri (65.07 Hz and 53.81 Hz, respectively).

Moreover, these values are notably larger than those reported by Park et al. [26] for NMV P229119 (7.47 radii ratio, 35.90 Hz low-frequency limit): the only aetiocetid included in their analysis. Inner ear morphology is difficult to reconstruct and is therefore rarely coded in phylogenetic analyses. These results suggest that Borealodon may be more similar in its inner ear morphology to mammalodontids than to aetiocetids. Given that Borealodon is not a mammalodontid, this may suggest that Borealodon occupied a bioacoustic niche similar to mammalodontids, which may represent the plesiomorphic condition shared with mammalodontids and distinct from aetiocetids. Alternatively, if additional fossil material reveals Borealodon to instead be an aetiocetid, these results suggest it may have been acoustically convergent with the mammalodontid bioacoustic niche. However, we emphasize that the inner ear morphology has not been described nor measured for any aetiocetid diagnostic to genus level. Accordingly, the bioacoustic diversity within Aetiocetidae remains poorly known.

5.3. Feeding

The most complete tooth of Borealodon was included in the recent analysis by Peredo et al. [6]. These authors found the tooth of Borealodon to be less complex (based on orientation patch count as initially developed by Evans et al. [27]) than the teeth of aetiocetids such as A. cotylalveus and S. meadi. This less complex value may be partially driven by a reduction in size of the accessory cusps, though the method is robust to changes in tooth size overall. In this, the teeth of Borealodon bear little resemblance to the elaborate, complex teeth of J. hunderi or aetiocetids. The cheek teeth of Janjucetushave three or four accessory cusps with deep divisions separating each cusp from the next. The teeth of Borealodon have as many as five cusps with very shallow divisions between the cusps such that the overall shape of the crown is more continuous. Moreover, the teeth of Borealodon lack the extreme curvature of the tooth roots, curved crowns and reduced accessory cusps characteristic of Aetiocetus spp. Instead, the overall smaller teeth, with shallow divisions between the accessory cusps and the heavy apical wear, are more reminiscent of F. buelli.

Moreover, four of the teeth show occlusal wear on the accessory cusps evocative of, but not as pronounced, as the wear seen in Salishicetus. Given the lack of rostral material and the fragmentary nature of the mandible of Borealodon, we are cautious in interpreting its feeding mode. However, based on comparable occlusal wear to Salishicetus and Fucaia, we interpret Borealodon as a raptorial feeder that used its teeth for prey processing [15,28].

5.4. Taphonomy

First, the cortical surface of nearly every element of USNM 539939 preserves small (less than 0.1 mm) circular boreholes. Previous authors have reported and attributed the boreholes to Osedax, a bone-eating annelid that consumes cetacean skeletons on the abyssal seafloor today [11,12]. These boreholes are preserved on the skull, mandibular elements, rib elements and tooth roots. They are not preserved on the ear bones nor on the tooth enamel. Previous authors speculated that the osteosclerotic earbones may have been too hard or lacked sufficient nutrients for Osedax to colonize [11]. The lack of boreholes on the tooth enamel supports the former hypothesis; Osedax seems to preferentially favour the less dense material—at least in USNM 539939. In particular, teeth A and B (figure 5a,b) demonstrate Osedax activity in the exposed dentin but not the enamel.

Second, the extent of bioerosion of the underlying cancellous bone, deep to the cortical bone across each element, is so substantial that it has jeopardized the structural integrity of the specimen. This bioerosion is visible along the posterior border of the supraorbital process of the right frontal and the preserved parts of the mandible (figures 1, 2 and 4), which are both adjacent to nearby boreholes on cortical bone. We infer that Osedaxbioerosion reduced much of the cranial material of USNM 539939 prior to burial.

Third, we note the presence of a single, incomplete bivalve in direct association with ventral surface of the right parietal, along the internal surface of the preserved braincase (figure 2). The bivalve preserves an incomplete external surface that is approximately 2 mm wide and 3 mm long; the overall oval morphology and size closely matches that of other vesicomyid bivalves that are reported in direct association with fossil mysticete skeletal material from fossil whale-falls, such as the one reported by Pyenson&Haasl [29] from the Monterey Formation of California. The addition of a likely chemosymbiotic bivalve taxon to the invertebrate fauna associated with this specimen provides more support for interpreting the type specimen of Borealodon as representing a sulfophilic-stage whale-fall community. Today's whale-fall communities advance through ecological stages that culminate in a sulfophilic stage where Osedax is present along with other invertebrate (although no other bone-boring) taxa [30,31]. Interestingly, Goedert&Squires [32] reported fossil chemosynthetic bivalves belonging to the vesicomyid genus Calyptogena from LACMIP locality 6295 (= LACMVP locality 5412), also located on the west side of Murdock Creek in Washington State (see type locality information), indicating that fossil whale-fall communities probably harbour richer associated invertebrate communities than more broadly appreciated.

Fossil cetacean skeletal material that comprise examples of chemosymbiotic fossil whale-falls have been described from Oligocene and Miocene localities along the North Pacific basin [11,12,33]. The presence of Osedax with clear taxonomic identification for the host skeleton (Borealodon, in this case) is unique for Osedax associations in the fossil record thus far, which includes a range of marine tetrapod host taxa from the Cretaceous to the Cenozoic from deep-sea rock deposits [34]. As Kiel et al. [11,12] proposed, Osedax itself may be a major contributor to an unknown (and potentially unknowable) component of the removal of skeletal material after death but prior to burial and incorporation into the geologic record.

Lastly, other authors have reported that during preparation of USNM 539939, five teeth of the shark Somniosus gonzalezi were discovered in the surrounding matrix [35]. These authors, as well as previous authors [11], speculated that the shark teeth may have been shed while scavenging the carcass. While we agree that it is possible, there is no evidence of scavenging directly preserved on the skeletal material of USNM 539939 itself, contra Kiel et al. [11].

6. Conclusion

Borealodon osedax represents a new lineage of toothed, stem mysticete crown-ward to mammalodontids and stem-ward to aetiocetids. Borealodon preserves elaborate, multi-cusped teeth with apical and occlusal wear reminiscent of Salishicetus and Fucaia. MicroCT scans of the inner ear indicate that Borealodon had a lower frequency limit of hearing similar to mammalodontids. The preservation of Osedax burrows on the type material of Borealodon is unique in that the host taxon now has a clear taxonomic identification. Borealodon offers more insights into the extent to which Osedaxscavenging may remove skeletal material and more broadly impact the cetacean fossil record, as well as into the evolution of these deep-sea whale-fall communities overall.

Borealodon osedax, a new stem mysticete (Mammalia, Cetacea) from the Oligocene of Washington State and its implications for fossil whale-fall communities

• B. K. Shipps
, 
• Carlos Mauricio Peredo
 and 
• Nicholas D. Pyenson

Royal Society Open ScienceVolume 6, Issue 7

Published:24 July 2019 https://doi.org/10.1098/rsos.182168