Quercygale angustidens ( Filhol, 1872 )
|
publication ID |
https://doi.org/ 10.5281/zenodo.13625323 |
|
persistent identifier |
https://treatment.plazi.org/id/03B3C348-FF87-FFF4-FCC5-0199FD2E9FAA |
|
treatment provided by |
Felipe |
|
scientific name |
Quercygale angustidens ( Filhol, 1872 ) |
| status |
|
Quercygale angustidens ( Filhol, 1872)
Figs. 1, 2 View Fig .
Age and locality: upper Eocene, Quercy area, France.
Description of basicranium
The auditory region has been central to understanding carnivoramorphan phylogeny for more than 150 years ( Flower 1869; Hunt 1974; Turner 1848; Wang and Tedford 1994; Wesley−Hunt and Flynn 2005; Wesley and Flynn 2003; Wyss and Flynn 1993). In recent years, the discovery and study of new material of basal carnivoramorphan taxa, Tapocyon and Oodectes , have significantly added to the cranial material known and the understanding of phylogenetic relationships among Viverravidae , “ Miacidae ” and Carnivora ( Wesley−Hunt and Flynn 2005) . Cranial material of Quercygale has been known for almost 100 years, but has yet to be incorporated into a modern phylogenetic framework. Nor has the auditory region been described in the context of basicranial evolution as we understand it today.
The following is a description of the auditory region of NRM−PZ M2329 (henceforth M2329) with comparisons and discussions of MNHN QU 17287 and MNHN QU 8755 based on personal observations of original specimens. Comparisons and observations based on text or illustrations are cited in the text. The auditory region of M2329 is in excellent condition with minimal breakage and no distortion. The specimen is a young adult with all molars fully erupted but the cranial sutures are not fully fused and can be plainly observed in the auditory region.
Petrosal and surrounding bones.—The basioccipital is relatively narrow and is ventrally convex. The lateral edges are smooth. The anteriomedial corner of the promontorium contacts the basioccipital ventrally just posterior to the basioccipital–basisphenoid suture. Anteriorly, the promontorium contacts the alisphenoid. Posteriorly, there is contact between the basioccipital and promontorium as the posterior basicapsular commissure, like that described by Wible (1983), or posterior “roof” to the inferior petrosal sinus ( Fig. 2 View Fig ). A groove for the inferior petrosal sinus is present along the medial surface of the promontorium. It is evident that the inferior petrosal sinus was not enlarged, and exited the cranium through the basicapsular fenestra (the space posterior to the anteromedial contact between the promontorium and the basioccipital), ran under the posterior basicapsular commissure, and then joined the internal jugular vein (illustrated in Wible 1983: 285, fig. 3b).
The promontorium is elongate anteriorly and rounded, and similar in its general shape to that of Oodectes . The pro−
WESLEY−HUNT AND WERDELIN—BASICRANIUM OF CARNIVORAMORPHAN QUERCYGALE 839
P3 P4 M1 alisphenoid canal foramen ovale glenoid process glaserian fissure basisphenoid promontorium 20 mm basioccipital paroccipital process
montorium of Quercygale is distinctive from other “miacids” in that there is a ventral ridge that runs along the length of the medial edge. This ridge is strongest anteriorly and curves over itself slightly ( Fig. 2 View Fig ). The lateral side of the ridge is rugose and appears to be an area of tympanic attachment. The ridge is not in contact with the basioccipital. No other “miacid” has this morphology. This ridge differs from the ventral petrosal process and lateral edge of the promontorium of Proailurus and other early feliforms in being elongate and lacking any contact with the basioccipital (see Hunt 1998 for anatomical comparison).
A sulcus for the internal carotid artery and promontory artery ( Fig. 2 View Fig ) is clearly visible, running anteroposteriorly along the ventral surface of the promontorium. The sulcus begins just anterior to the medial extent of the fenestra cochleae, runs along the anterior lip of the fenestra then turns anteriorly and continues to the middle lacerate foramen. There is no sulcus for a stapedial artery present in M2329; however this morphology appears to be variable, as it is extremely distinctive in MNHN QU 8755. The stylomastoid foramen is posterolateral to the fenestra cochleae, tightly defined and anteriorly bordered by the mastoid tubercle. The mastoid tubercle is formed by the petrosal and lies laterally to the fenestra cochleae. The mastoid process is small, not distinctive and similar to that of Prohesperocyon . A shallow suprameatal fossa is present on the anterior face of the mastoid process similar to that in Oodectes . The mastoid shelf is smooth, similar in size to that of Vulpavus and relatively large compared to that of Oodectes . The paroccipital process is a distinct narrow process and projects posteroventrally.
The tegmen tympani is not fully ossified and a piriform fenestra is present between the petrosal and alisphenoid. (It could be argued that the piriform fenestra in this specimen should not be referred to as such, and is actually an extension of the middle lacerate foramen.) The middle lacerate foramen and piriform fenestra are separate openings in M2329, but they are joined into one vacuity in MNHN QU 17287 and MNHN QU 8755. However, in M2329 the middle lacerate foramen is not a distinct foramen as in more advanced carnivoramorphans, but a small vacuity. The canal for the facial nerve is floored by very thin bone anteromedially, and is open only at the lateral end of the canal. On the left side of M2329 the bony floor has been broken during preservation or preparation. This raises concerns, as expressed in earlier studies, about observations of fully open canals for the facial nerve and whether the open morphology is merely an artifact of preservation or preparation damage to the fragile bony floor ( Wesley−Hunt and Flynn 2005; Wesley and Flynn 2003).
The epitympanic recess is shallowly excavated into the squamosal. The suture between the petrosal and the squamosal is very evident on the posterior wall of the recess. A foramen is present at this suture as in Miacis parvivorus Cope, 1872 and paroccipital process
Oodectes . It is probable that this canal held the superior ramus of the stapedial artery as it exited the auditory region ( Novacek and Wyss 1986; Wang and Tedford 1994).
The Glaserian fissure is narrow and almost closed over at its posterior dorsal extent. Lateral to the fissure, the postglenoid foramen is open, not reduced, and medial on the squamosal ridge bordering the auditory region. The alisphenoid rises to the tegmen tympani at a shallow angle, similar to that in Tapocyon .
Bulla.—A firmly attached ossified bulla is not present in Quercygale , and no ossified bulla is preserved in any specimens. However, there is evidence on the surrounding bones that some form of bulla was present, whether ossified but weakly attached, or fully or partially cartilaginous. On both the left and right side of the M2329 there is an indentation at the suture between the squamosal, petrosal and alisphenoid, anteromedial to the epitympanic recess and medial to the posterior extent of the Glaserian fissure. This indentation is rugose and the point of contact with the anterior crus of the ectotympanic as it is in modern carnivores and other “miacids”. There is no evidence of a smooth indentation on the alisphenoid caused by the anterior face of the ectotympanic as a result of anterior expansion of the bulla, as observed in other taxa (see Hunt 1998). Therefore, Quercygale probably did not have an anteriorly expanded bulla. The medial ridge of the promontorium is the most definitive evidence of contact with the tympanic, most likely the caudal entotympanic due to its linear anteroposterior extent. This ridge ( Fig. 2 View Fig ) is smooth on the medial face and rugose on the lateral face. The rugosity extends on to the ventral surface but is limited to the medial edge of promontorium. A similar morphology can be observed on Miacis sylvestris ( Marsh, 1872b) (AMNH 129284), however, no distinct ridge is present, just a wide roughened surface on the medial edge of the promontorium. This ridge morphology is unique to Quercygale among the “miacids” and Carnivora . On the ventral surface of the promontorium, medial to the anterior edge of the foramen cochlea there is a small divot or dimple that stands out from the surrounding attachment surface. This divot is of itself not noteworthy on Quercygale , as it is just part of the larger entotympanic attachment surface, but this feature is also present on Oodectes , Miacis parvivorus , and Tapocyon . These specimens do not have the extensive entotympanic attachment surfaces observed in Quercygale , so the more concrete association of this divot with an entotympanic attachment in Quercygale bolsters the argument for an entotympanic attachment in the other specimens. The roughened surface anteromedially on the promontorium supports the inferred presence of a rostral entotympanic. The medial extent of the bulla appears to be the medial ridge of the promontorium: there is no evidence of a medially expanded bulla. The apron of the promontorium posterior to the fenestra cochleae is rugose and suggests an attachment surface for the caudal entotympanic as is observed in Nandinia and Tapocyon . The mastoid shelf is smooth, as is the paroccipital process, and therefore the entotympanic was probably not expanded posteriorly past the promontorium. As observed in Tapocyon , Oodectes and Nandinia , there is no evidence that the caudal entotympanic was attached to the paroccipital process in Quercygale .
Internal carotid artery.—The carotid artery follows a course very similar to its condition in Miacis sylvestris . When the internal carotid artery enters the middle ear cavity it splits into a stapedial and a promontory branch. The presence of the stapedial artery is problematic. There is no evidence of a sulcus for a stapedial artery in the specimen M2329, nor in MNHN QU 17287. However, in MNHN QU 8755 the sulcus is very clear. In addition, although no sulcus is present for the stapedial artery in M2329, a canal is present, entering the posterior wall of the epitympanic recess, presumably the course for the superior ramus of the stapedial artery to exit the middle ear cavity ( Novacek and Wyss 1986; Wang and Tedford 1994). The sulcus for the promontory artery is, however, clearly defined in all the specimens observed. The simplest conclusion is that the stapedial artery is present in Quercygale , but that the sulcus for the stapedial artery is variably developed. Quercygale is one of the few genera of “Miacoidea” in which it is possible to study variation in basicranial anatomy from multiple specimens.
The internal carotid artery is transpromontorial ( Wible 1983, 1986) running along the promontorium on the ventrolateral surface toward the middle lacerate foramen. The artery passes anterior to the foramen and turns around on itself, forming an anterior loop before entering the brain cavity through the middle lacerate foramen ( Fig. 2 View Fig ). This anterior loop is present in more derived “miacids” such as Tapocyon , Miacis sylvestris , and Prohesperocyon .
Phylogenetic analysis
Methods and data.—The phylogenetic analysis is based on Wesley−Hunt and Flynn (2005) to whom we refer for detailed information on the character matrix (cf. Appendix 1). There are some minor alterations made to the matrix for the purposes of this paper. Characters 43 and 44 of Wesley−Hunt and Flynn (M1 metastyle projection and M1 parastyle projection, respectively) have been replaced by a composite character with three states, 0 = metastyle projecting further labially, 1 = metastyle and parastyle with equal projections, 2 = parastyle projecting further labially. This was done because of the realization that characters 43 and 44 of Wesley−Hunt and Flynn (2005), due to their definition, are only semi−independent, with state 1 of character 43 being nearly the same as state 0 of character 44. This change has, in fact, not affected the analysis. Despite this minor change, we have retained the character numbering of Wesley−Hunt and Flynn (2005). Thus, our characters are numbered 1–99, but exclude #44. In addition, one stem−group taxon, Viverravus acutus Matthew and Granger, 1915 , was added here (Polly et al. in press), and six crown−group taxa removed [ Otarocyon macdonaldi Wang, Tedford, and Taylor, 1999 , Pteronarctos goedertae Barnes, 1989 , Otariidae , Mephitis sp. , Ailurus fulgens F. Cuvier, 1825 , and Procyon lotor ( Linnaeus, 1758) ], because these are redundant to the issues addressed in the present paper.
A total of 36 taxa were thus analysed using a matrix of 98 characters, of which one, placement of the middle lacerate foramen, was ordered (on the basis of analyses carried out by Hunt (1987, 1998). Tree topologies were evaluated using the heuristic search algorithms of PAUP* version 4.0b10 for Macintosh ( Swofford 1998), with ACCTRAN optimization. The heuristic searches carried out 100 random addition sequence iterations. Bremer support values were calculated using TreeRot, version 2 ( Sorenson 1999) and tree manipulation was carried out with Mesquite, version 1.05 ( Maddison and Maddison 2004).
Didymictis vancleveae Protictis schaffi
Viverravus acutus Viverravus minutus Viverravus cf. gracilis
Oodectes herpestoides Vulpavus ovatus
Vulpavus profectus Tapocyon robustus Prohesperocyon wilsoni Miacis cf. M. sylvestris Quercygale angustidens Dinictis felina
Hoplophoneus sp. Zodiolestes daemonelixensis Gulo gulo
Hesperocyon gregarius Canis latrans
Nandinia binotata Palaeoprionodon lamandini Stenogale julieni
Proailurus lemanensis Herpestides antiquus Civettictis civetta
Results.—Of the 98 characters, 89 were parsimony−informative for the present set of taxa. The parsimony analyses using PAUP* found 12 most parsimonious trees, each of length 386, consistency index 0.339, retention index 0.657, and rescaled consistency index 0.223. The heuristic searches found these 12 trees about half the time, but only one tree island was recovered, with other searches stopping at solutions known to be suboptimal for this data set. The strict consensus tree of these 12 trees is shown in Fig. 3 View Fig , which also shows the Bremer support values for each ingroup node.
The differences between the topology obtained in this study and that of Wesley−Hunt and Flynn (2005) are small but interesting. Within the Carnivora (crown−group), the topologies are identical, with two exceptions. The Nimravidae , which in their topology was the sister taxon to Feloidea, here becomes sister taxon to the remaining Carnivora . This reflects the historically unstable placement of Nimravidae ( Hunt 1987; Neff 1983; Werdelin 1996; Wyss and Flynn 1993) and others. It is worth noting that both nimravid taxa used herein belong to the Nimravidae sensu stricto, and that this variable placement therefore does not reflect the debate regarding whether the barbourofelids represent a family distinct from the Nimravidae ( Morlo et al. 2004) . In addition, the present analysis moves the Amphicyonidae and “ Miacis ” cognitus Gustafson, 1986 into successive sister−group positions relative to Ursidae and interchanges the positions of Mustelidae and Canidae . These differences can in part be ascribed to the reduced number of caniform taxa included in the analysis, but the closeness of Amphicyonidae and Ursidae has been observed in other analyses of carnivoran phylogeny ( Wyss and Flynn 1993).
Among the stem lineage taxa, Viverravus acutus is placed together with other members of Viverravus in the Viverravidae of Wesley−Hunt and Flynn (2005). Resolution of the cladogram is reduced above the Viverravidae , and Miacis , Oodectes , and Vulpavus form an unresolved polychotomy. In our analysis, Prohesperocyon and Miacis cf. M. sylvestris form a clade together with Tapocyon , while in the analysis of Wesley−Hunt and Flynn (2005), the latter taxon was placed one step higher in the phylogeny. Finally, Quercygale , our target taxon in this analysis, is placed between the Tapocyon clade and the Carnivora , as the most derived “ Miacidae ” in the formulation of Wesley−Hunt and Flynn (2005).
A number of characters are of particular importance to the placement of Quercygale in the phylogeny (see Table 1). Character #7, proportion of frontal and parietal midline length, excludes Quercygale from crown−group Carnivora (including Nimravidae ); #16, hypoglossal foramen position, separates Quercygale (and Tapocyon , Prohesperocyon , and Miacis cf. M. sylvestris ) from Vulpavus and less derived taxa; #18, fenestra cochleae position, on the other hand, unites Quercygale with Tapocyon and other “ Miacidae ” of Wesley−Hunt and Flynn (2005) and excludes them from crown−group Carnivora . The state of the crown−group represents a reversal to the primitive condition exhibited by the outgroups and Viverravidae ; #23, anterior loop of internal carotid artery, unites Quercygale and the Tapocyon clade with nimravids and caniforms; #25, position of the internal carotid artery, is another character that excludes Quercygale and other “ Miacidae ” from crown−group Carnivora ; #26, apron shelf on promontorium, on the other hand, unites Quercygale (as well as Tapocyon and Vulpavus ovatus Matthew, 1909 ) with crown−group Carnivora ; #32, squamosal/ alisphenoid contact from anterior crus, also excludes Quercygale from crown−group Carnivora , as does #34, extent of flange on basioccipital lateral edge; #45, parastyle direction, however, unites Quercygale and the Tapocyon clade with crown−group Carnivora ; #52, M1 size, also unites Quercygale and Tapocyon with crown−group Carnivora , as does #53, presence/absence of M3; #56, P4 protocone size, also unites Quercygale and the Tapocyon clade with crown−group Carnivora , while #66, position of postorbital constriction, excludes them from crown−group Carnivora ; finally, #88, presence/absence of m3, is again a character that unites Quercygale (but not the Tapocyon group) with crown−group Carnivora . In summary, the analysis points to a number of characters that are important to placing Quercygale near crown−group Carnivora (and Tapocyon ), but other characters definitely exclude Quercygale from the crown−group.
No known copyright restrictions apply. See Agosti, D., Egloff, W., 2009. Taxonomic information exchange and copyright: the Plazi approach. BMC Research Notes 2009, 2:53 for further explanation.
