Cretoxyrhina mantelli (Agassiz, 1843)

Siverson, Mikael & Lindgren, Johan, 2005, Late Cretaceous sharks Cretoxyrhina and Cardabiodon from Montana, USA, Acta Palaeontologica Polonica 50 (2), pp. 301-314 : 303-307

Cretoxyrhina mantelli (Agassiz, 1843)

Fig. 2.

For synonyms.—See Siverson (1992).

Material.—61 teeth (WAM 04.10.3–04.10.63).

Description.—For a general description of Cretoxyrhina mantelli teeth, see Eastman (1895), Welton and Farish (1993), and Shimada (1997a). Below we describe more specific characters possibly unique to Fairport−age populations.

The collection includes 18 small to moderate−sized anterior teeth (three complete), of which 13 are from larger juveniles (and a few possibly from small adults) and five from very young sharks, possibly yearlings (original height of the smaller anterior teeth ranging from 5–7 mm). The gap in tooth size between the two groups of juvenile teeth is large. This may be a result of different collecting techniques (surface collecting versus ant hill exploration).

Five different types of anterior teeth are present, presumably corresponding to the first, second and third lower anterior files and the first and second upper anterior files. Two imperfectly preserved teeth may belong to the third upper anterior files (the teeth from the third upper anterior files are reduced in size in Cretoxyrhina mantelli ; Siverson 1999). The root is relatively symmetric (as far as can be determined) and the cusp is slightly distally bent in two of the five identified anterior tooth−types. One of these two tooth−types is characterized by having a flat basal portion of the labial side of the cusp and cutting edges reaching the base of the cusp. In the other type, the base of the labial face of the cusp is convex and the cutting edges terminate about two millimetres above the base of the cusp ( Fig. 2E). Comparison with the dentally similar modern short fin mako ( Isurus oxyrinchus Rafinesque, 1810 ) and crocodile shark— Pseudocarcharias kamoharai ( Matsubara, 1936) indicates that the first type represents the first upper anterior file while the second type represents the first lower anterior file.

Of the remaining three anterior tooth−types, one has a relatively tightly curved root with a markedly longer mesial lobe ( Figs. 2F, G). The cutting edges reach the base of the cusp. We refer this type to the second lower anterior file. Another morphology is represented by a complete tooth showing features typical for the third, most distal lower anterior file in makos ( Fig. 2H). The cusp is distally bent and slightly lingually curved. The cutting edges reach the base of the cusp. The root is asymmetric, with a short but broad and angular distal lobe in contrast to a narrow but much longer and acute mesial lobe. The fifth type has a cusp curved in the same manner as the second upper anterior tooth in juvenile short fin makos ( Fig. 2D). The basal two thirds are distally bent whereas the apical third is mesially curved. Both the mesial and the distal cutting edges terminate some distance above the base of the cusp. Unlike the teeth assigned to the lower anterior files, the cusp has no lingual curvature in profile view. The root is markedly asymmetric with a longer mesial lobe.

None of the anterior teeth at hand have any traces of lateral cusplets. Our material of anterior teeth does not include large specimens. These are more generalised in morphology (i.e., not as differentiated as in juveniles) and sturdier (see Williamson et al. 1993: figs. 6.10, 6.11; Antunes and Cappetta 2002: pl. 11: 2).

We have not been able to separate upper lateroposterior teeth from lower lateroposterior teeth with much confidence, as they are much alike. In all lateroposterior teeth, the crown and the root are flushed in the same plane labially and the whole labial surface of the tooth is very flat (see Fig. 2O 1 ). The angle between the labial faces of the root−lobes and the cusp is close to 180 ° in both upper and lower lateroposterior teeth. In living macrophagous lamniforms this angle is usually markedly less than 180 ° in upper lateroposterior teeth. The cusp is thick and terminates relatively abruptly at its apex. The heels are fairly short and oblique. Lateral cusplets are present on posteriorly situated lateroposterior teeth. There is no clear correlation between the presence of lateral cusplets and the ontogenetic stage of the tooth. The only ontogenetic difference, other than size, includes a broader more triangular cusp on larger teeth from a similar position.

Comparisons.—We have adopted a conservative taxonomic approach by referring the Fairport sample of Cretoxyrhina to the type species C. mantelli , but recognise that this sample differs significantly from late Coniacian to middle Santonian samples of Cretoxyrhina teeth from the Smoky Hill Chalk Member of the Niobrara Formation in western Kansas (the Niobrara specimens correspond closely to the syntypes of C. mantelli ). Besides a marked difference in maximum size (early middle Turonian teeth reach a considerably smaller maximum size than do late Coniacian to middle Santonian ones; Shimada 1997d and personal observations), the cusp is, on average, narrower at its base on lateroposterior teeth from late Coniacian to middle Santonian strata than it is on the corresponding teeth from middle Turonian deposits. This results in relatively longer and more horizontal heels on a typical tooth from the Smoky Hill Chalk. There is considerable overlap between middle Turonian and late Coniacian to middle Santonian samples in these characters but the number of lateroposterior teeth needed to demonstrate the difference is still rather small, probably in the order of a dozen lateroposterior teeth. Shimada (1997d) noted a significant increase in tooth size, coupled with a decrease in the crown−height/ crown−width ratio for anterior files in Niobrara C. mantelli compared to Cenomanian–Turonian populations.

In two different types of small, presumably juvenile anterior teeth of Cretoxyrhina from the Fairport Member, here interpreted as the first lower ( Fig. 2E 2) and the second upper teeth ( Fig. 2C 2 and D 1 ), the distal cutting edge terminates well above the base of the cusp. The character is present on first lower anterior teeth up to a maximum slant height of at least 25 mm and on second upper anterior teeth up to a maximum slant height of at least 28 mm. This feature may have had a relatively short temporal distribution in Cretoxyrhina . The distal cutting edge is complete in latest Albian Cretoxyrhina from Mangyshlak, Kazakhstan (DJW collection) and in early–middle Santonian C. mantelli from Kansas, USA (collected by us). The temporal distribution of an incomplete distal cutting edge on small, presumably juvenile anterior teeth is thus within the early Cenomanian to late Coniacian interval. This amounts to a maximum of 10.6 Ma using the data in Kauffman et al. (1994).

Another feature in Fairport−age Cretoxyrhina that we believe to be of importance from a biostratigraphic point of view is the very flat labial side of both upper and lower lateroposterior teeth. In latest Albian Cretoxyrhina , the angle (in profile view) between the labial side of the root−lobes and the labial side of the cusp, is typically 150–170 ° in upper lateroposterior teeth. This is in contrast to the corresponding teeth in Fairport−age Cretoxyrhina which have a labial angle very close to 180 ° ( Fig. 2M 2).

Antunes and Cappetta (2002) illustrated seven teeth of C. mantelli from the upper Turonian of Angola. None of these teeth represents anterior tooth−files from smaller juveniles, preventing a comparison with the juvenile anterior teeth from the Fairport Member at Mosby.

Discussion.— Cretoxyrhina extends back in time to the late Albian, where it is present in Texas (Pawpaw Formation, MS unpublished data), Mangyshlak ( Zhelezko 2000; described as Pseudoisurus ; a nomen dubium; see Siverson 1999) and Stary Oskol in eastern Russia (personal observations of material in DJW’s collection). The three teeth illustrated by Zhelezko (2000: pl. 1: 3–5) as Pseudoisurus vraconensis , is a mixture of Cretoxyrhina vraconensis ( Zhelezko, 2000) and Cretalamna appendiculata (see Zhelezko 2000: pl. 1: 5). The Stary Oskol selachian fauna includes three large selachian apex predators, including Cretoxyrhina vraconensis , Cardabiodon sp. , and Dwardius siversoni ( Zhelezko, 2000) . In contrast, the late Albian selachian faunas from Mangyshlak (DJW collection) and Texas (Scott Kelley private collection, Dallas, Texas, USA, in addition to unregistered specimens collected by MS) lacked both Cardabiodon and Dwardius Siverson, 1999 .

There is currently nothing to suggest that there was ever more than one species of Cretoxyrhina present at any given time during the late Albian to early Campanian interval. Any sub−division of the Cretoxyrhina −lineage should be based on well preserved and well illustrated material collected from a narrow stratigraphic interval. Zhelezko (2000) described a new species of Cretoxyrhina from the Turonian of Mangyshlak, Kazakhstan, and named it Pseudoisurus sulukapensis (= Cretoxyrhina sulukapensis ). He relied on a total of 13 teeth (of which he illustrated two) lacking published information on their stratigraphic provenance within the Turonian stage. The presumed rarity of the nominal species at Sulukapy in Mangyshlak, Kazakhstan, from where the holotype was collected (not all of the 13 teeth are from this locality) indicates that it would be difficult to collect additional material of this nominal species from the (unpublished) type stratum. The holotype is probably early or late Turonian in age, as the middle Turonian is missing at Sulukapy (see Marcinowski et al. 1995).

Available data indicate that the extinction datum of Cretoxyrhina mantelli , the last representative of the genus, is diachronous. In Western Australia, the youngest occurrence is an undescribed tooth from the late Santonian Marsupites testudinarius Zone in the Gingin Chalk, Molecap Hill Quarry, Perth Basin (see also Siverson 1996: 822).

In the Western Interior of the USA, Cretoxyrhina mantelli persisted well into the early Campanian. On a field trip in 1999, Jerome (Pete) Bussen, Wallace (Kansas), collected a C. mantelli tooth (WAM 04.10.101) from within the uppermost metre of the Smoky Hill Chalk Member of the Niobrara Chalk in Logan County, western Kansas (locality data on file at the WAM). This level is 30–45 m above the Santonian–Campanian boundary, as defined by nannofossils, which lies somewhere within a 15 m thick interval bounded biostratigraphically by the First Appearance Datum of regular Calculites obscurus (Deflandre) and the FAD of Broinsonia parca parca (Stradner) , as determined by Watkins et al. (1994).

In northwestern Europe, the last known occurrence of Cretoxyrhina mantelli is about two million years younger than that in the Western Interior of the USA. Siverson (1992) recorded the species from the local Belemnellocamax mammillatus zone in the Kristianstad Basin in southern Sweden. This informal biozone correlates with the uppermost lower Campanian in the German zonation; the Gonioteuthis quadrata gracilis / Belemnitella mucronata senior Zone of Schultz et al. (1984). The species was recorded from three localities within the Kristianstad Basin (i.e., Ullstorp 1 sensu Erlström and Gabrielson 1986, Ignaberga and Ivö Klack localities). The C. mantelli −yielding horizon in the Ullstorp 1 quarry (Bed C; see Erlström and Gabrielson 1986: fig. 3) was originally thought to be latest early Campanian in age ( Belemnellocamax mammillatus zone) but Siverson (1993) indicated that it was older. This suspicion was later confirmed by the finding of a diverse belemnite assemblage placing the bed in the oldest biozone of the lower Campanian ( Gonioteuthis granulataquadrata Zone ; Siverson and Lindgren, unpublished data). Likewise, the geology at Ignaberga allows for the possibility that the C. mantelli teeth from this locality are reworked from an upper middle Santonian sandstone (see Erlström and Gabrielson 1992). At Ivö Klack, however, there is no evidence of marine strata or fossils of pre− B. mammillatus −age and the eight teeth of C. mantelli listed in Siverson (1992) show no obvious sign of being reworked (unlike the teeth from Ignaberga) and have the same state of preservation as selachian teeth of other species from Ivö Klack typical for the B. mammillatus zone.

Cretoxyrhina mantelli was also reported from the lowermost upper Campanian by Siverson (1992: 526) but this record rests on a single, much worn tooth (in the private collection of Peter Cederström, Eslöv, Sweden) from the B. balsvikensis −conglomerate at the Balsvik quarry (see Christensen 1998: fig. 2). Its state of preservation indicates extensive reworking.

During the latest Albian to middle Turonian interval, Cretoxyrhina and Cardabiodon were in the process of loosing the lateral cusplets on their teeth. In both genera this process commenced in the anterior files. The loss of cusplets on anterior teeth was well under way by the latest Albian in Cretoxyrhina (see Zhelezko 2000: pl. 1: 3) and by the middle Cenomanian in Cardabiodon (see Siverson 1999: fig. 9.1a). As is evident from the illustrations in this paper, Fairport−age Cretoxyrhina had reached a more advanced stage of cuspletreduction than had the sympatric Cardabiodon .

Figures

Fig. 1. Map showing the location of studied area in Montana (A) and location of Mosby (B). All selachian remains were collected on private land. Precise locality data is on file at the Department of Earth and Planetary Sciences, Western Australian Museum, Perth.