Kyawdawia lupina Egi, Holroyd, Aung Naing Soe, Takai & Ciochon, 2005
|
publication ID |
https://doi.org/ 10.5281/zenodo.4665549 |
|
persistent identifier |
https://treatment.plazi.org/id/03D287BC-A703-FFE8-19F3-FB38FB37FB37 |
|
treatment provided by |
Felipe |
|
scientific name |
Kyawdawia lupina Egi, Holroyd, Aung Naing Soe, Takai & Ciochon, 2005 |
| status |
|
Kyawdawia lupina Egi, Holroyd, Aung Naing Soe, Takai & Ciochon, 2005
HOLOTYPE. — NMMP-KU 0042, 0043, 0044, 0784, 0785, and 1661 (acronyms used by Egi et al. 2005): cranial and mandibular fragments with subcomplete permanent dentition and various incomplete postcranial bones (see Egi et al. 2005 for a precise description of the holotype). All of these specimens represent a single individual from Kyawdaw according to Egi et al. (2005). However we add here M3s, which are not listed in the holotype definition (these teeth belong to the type specimen NMMP-KU 0042; see Egi et al. 2005: fig. 1A, C) but in the description of the holotype dentition by Egi et al. (2005: 340).
TYPE LOCALITY. — Kyawdaw, Pondaung Formation, late
middle Eocene of central Myanmar.
Peigné S. et al.
From Thamingyauk: NMM Tmk-500, subcomplete right M 3 in mesial (L) and occlusal (M) views. Scale bar: 1 cm.
Creodonts from the Pondaung Formation, Myanmar gual (B), labial (C) and occlusal (D) views; NMM Pkg-1051, right m 2 in lingual (E), labial (F) and occlusal (G) views. Scale bar: 1 cm.
Peigné S. et al.
NEW REFERRED MATERIAL AND LOCALITIES ( Fig. 1 View FIG ). — From Bahin: NMM Bhn-1040 (left and right isolated M1, Fig. 2A, B View FIG ), NMM Bhn-2000 (fragment of right hemimandible with m3 with broken protoconid, Fig. 3A View FIG ); from Thandaung: NMM Tdg-800 (left M2, Fig. 2 View FIG E-G), NMM Tdg-801 (fragment of left M3, Fig. 2C, D View FIG ), NMM Tdg-802 (fragment of left P4, Fig. 2H, I View FIG ); from Paukkaung 2 (Pk 2): NMM Pkg-1050 (fragment of left hemimandible with m2 talonid and m3 with broken protoconid, Fig. 3 View FIG B-D), NMM Pkg-1051 (isolated right m2, Fig. 3 View FIG E-G); from Kyaukmagyi: NMM Kmi-500 (fragment of right M2, Fig. 2J, K View FIG ).
DISTRIBUTION AND AGE. The material of K. lupina was collected from many localities of the Pondaung Formation , Myanmar ( Fig. 1 View FIG ), which dates to the late middle Eocene. In Mogaung area: Kyawdaw (Kdw), which is the type locality near the village of Kyawdaw ; UCMP locality V 83106 View Materials , located about 5.6 km southwestern from Mogaung ; and 1 unspecified locality near Kyawdaw. In Bahin area : Paukkaung-2 (Pk 2), near the village of Paukkaung, and several unspecified localities near Paukkaung/ Bahin. In Pangan area : Pangan-1 (Pgn 1) and Pangan 2 (Pgn 2). New specimens described here come from the Bahin area, and from Thandaung (Tdg) and Kyaukmagyi (Kmi), both located in the Mogaung area ( Fig. 1 View FIG ) .
DESCRIPTION OF THE NEW MATERIAL
Compared with the holotype specimen, the study of which is based on a resin cast made by one of us (SD), the new material is both better preserved and less worn, although it is less complete since the holotype documents the major part of the tooth rows. The measurements of the new material described in the present study and those from the hypodigm are presented in Table 1 (upper dentition) and Table 2 (lower dentition). The P4 lacks its mesial and mesiolingual portions but the tooth was probably longer than it is high (height is greater than length on p4; Egi et al. 2005). The cingulum is developed on the talon of P4, which is crest-like. Although the tooth is not completely preserved lingually, it is clear that the protocone lobe was not as developed as in some other proviverrines (e.g., Masrasector Simons & Gingerich, 1974 , Cynohyaenodon Filhol, 1873 ), and that the lingual bulge was poorly developed; these are derived characters of the species ( Egi et al. 2005). The M1 (NMM Bhn- 1040) is small and approximately the size of that in
the holotype (see Table 1). On the contrary to the latter, the M1 from Bahin is very well preserved. The paracone and metacone are about the same height, the former being smaller in diameter/length than the latter. Both cusps are connated at their base, with their upper portion being separated by a deep valley and a shallow notch. The protocone is much lower than the paracone and metacone, but it has a wide and triangular-shaped base and marked crista. It is widely separated from the labial cusps by a deep basin. A small paraconule and a vestigial metaconule are distinct. The metastyle blade is elongated, low and separated from the postmetacrista by a deep notch. The cingulum is distinct but poorly developed all through the lingual face and it is labially more marked. A small parastyle is present in the mesiolabial corner. The M 2 from Thandaung ( NMM Tdg-800) is a large morphological replica of the M1, although both are easily distinguishable. On M 2 the paracone and metacone are still of similar height, with the metacone base larger than the paracone one ; the V-shaped valley that separates the two cusps is, however, more open than in M1. The metaconule is larger and the cingulum is more mesially developed than in M1. The fragmentary M2 from Kyaukmagyi ( NMM Kmi-500) preserves the paracone, the metacone, the protocone and the conules ; it is slightly smaller than NMM Tdg-800, with a slightly stronger cingulum around the protocone. The M 3 from Thandaung ( NMM Tdg-801) compares well with that of the holotype ; both specimens preserved the same labial portion of this tooth and are of similar length. The tooth from Thandaung is, however, a fresh one while that of the holotype is horizontally abraded. Size and morphology of NMM Tdg-801 match those of the corresponding tooth of the holotype, especially in the size of the parastyle and the paracone, which is larger than the metacone, and the two cusps being widely fused except for their apex. Because of the wearing pattern, conules are not distinct in the holotype but, on the M3 from Thandaung, a metaconule and a paraconule are present, the former being the largest .
The lower dentition is documented by fragmentary hemimandibles with m2 talonid, m3 with broken protoconid (NMM Bhn-2000 and
NMM Pkg-1050), and an isolated m2 (NMM
Pkg-1051), which represent three individuals. Both teeth have a cingulid developed on the labial face, which is particularly prominent mesiolabially on the paraconid (“precingulid” of Van Valen 1994). The isolated m2 from Paukkaung is particularly interesting because the tooth was previously known only from the holotype, in which it is much worn. In contrast NMM Pkg-1051 is a fresh, perfectly preserved tooth. The trigonid is tall, with a protoconid much taller than the paraconid, although the two cuspids are about the same mesiodistal length. In addition, the protoconid height is much greater than the trigonid length (see measurements in Table 2). The mesial face of the paraconid is slightly convex, displays a prominent ridge and, at its base, a strong precingulid. The metaconid is reduced and lower than the paraconid ; the two cuspids are separated by a V-shaped valley that is prolongated by a distinct groove on the lingual face of the tooth. The metaconid shows a prominent ridge along its distolingual face. The distal face of the trigonid is nearly transversely oriented, with the distal faces of the metaconid and protoconid being at the same level. The talonid is as wide as the trigonid but slightly shorter; the talonid basin is deep. The cristid obliqua is labial, and its maximum height is located distally at the level of the hypoconid. The entocristid is longer than the cristid obliqua and it is curved lingually. There is no distinction between an entoconid and a hypoconulid. The latter is separated mesially from the metaconid and distally from the hypoconid by a notch. The enamel on the labial face of the protoconid and the talonid, above the cingulid, is clearly wrinkled. NMM Pkg-1051 has a strong cingulid along its labial face. The m2 talonid is also preserved, but very worn, in NMM Pkg-1050, it shows the same shape as in NMM Pkg-1051 and the similarly developed labial cingulid. The m3s in NMM Bhn-2000 and NMM Pkg-1050 do not preserve the protoconid ; the tooth is also more worn in the former than in the latter. Relative to that of m2, the talonid of m3 is distinctly narrower than the talonid, which is mainly due to a more centrally placed cristid obliqua. In addition, compared to m2, m3 has a slightly more reduced metaconid and a more triangular-shaped
talonid, with an entocrista that is clearly less curved
Creodonts from the Pondaung Formation, Myanmar
and distally more prominent relative to the cristid obliqua. Nevertheless, the trigonid/talonid length ratio and the development of the labial cingulid are similar in m2 and in m3.
COMPARISONS AND DISCUSSION
Our comparisons are based on a direct study of a great number of proviverrine hyaenodontids by at least one of us (SP and/or MM). Egi et al. (2005) support a relationship of Kyawdawia Egi et al., 2005 to the proviverrines Masrasector and Paratritemnodon (but see the following discussion on character misinterpretations in that analysis), partly based on the presence of a distally extended labial cingulid on m3, which was said to be an autapomorphy of this clade. However, an extended labial cingulid is also known from: 1) the European monospecific Proviverra Rütimeyer, 1862 (but not Lesmesodon Morlo & Habersetzer, 1999 , even if both genera were coded as a single taxon by Egi et al.); 2) the North American Sinopa and monospecific Proviverroides Bown, 1982 (Morlo & Habersetzer 1999), taxa close to Prototomus Cope, 1874 (see Gunnell 1998; Morlo & Gunnell 2003); 3) Laurasian Arfia Van Valen, 1965 (see Gingerich 1989); and 4) the apterodontine hyaenodontid Apterodon Fischer, 1880 (see Lange-Badré & Böhme 2005). The character description, however, is correct in stating that the labial cingulid surrounding the hypoconid extends to the distal tooth margin. This is not the case in Proviverra , but it is in Sinopa , Proviverroides , Arfia , Paratritemnodon , Yarshea Egi et al., 2004 , and Masrasector , and also in Apterodon . The character thus cannot be interpreted as an autapomorphy of the proposed Afro-Asian proviverrine clade (see Fig. 4A View FIG ). The character is present on the specimens described and referred to here as Kyawdawia ; these remains differ from Masrasector and Paratritemnodon by their much larger size.
The new material described here reveals a great range of size variation. Compared to most of the known proviverrines, Kyawdawia lupina is a large species, especially after Dissopsalis Pilgrim, 1910 has been removed from Proviverrinae and included in Hyainailourinae (Polly 1996), partly based on the tooth eruption sequence of its sister-taxon Buhakia
Morlo, Miller & El-Barkooky, 2007. Body size
Peigné S. et al.
TABLE 1. — Measurements (in mm) of the upper dentition of Kyawdawia lupina Egi et al., 2005 . Abbreviations: L, labial length; W, labiolingual width; LL, lingual length; ML, metastylar length; *, mean of right and left teeth.
is one of the characters used by Egi et al. (2005: character 1) in their character matrix. According to these authors, m1-3 length is greater than 47 mm in K. lupina (character state and threshold previously used by Barry 1988) which is based on the holotype. Adding each of the molar length gives 47 mm while (due to overlapping of the molars) m1-m3 length is 44.9 mm only in the holotype right hemimandible NMMP-KU 1661 (see Egi et al. 2005: 342); lower dentition measurements of the holotype in Egi et al. (2005: table 1) should refer to NMMP-KU 1661, not NMMP-KU 0042, which corresponds to the maxilla, cranial and postcranial fragments. In the material described here, the upper teeth are larger (M1 and M2) or slightly smaller (M3) than the holotype (Table 1). Compared to the size variation of, e.g., the proviverrine-like hyainailourine (Polly 1996; Morlo et al. 2007) Dissopsalis carnifex Pilgrim, 1910 ( Barry 1988: table 1), this size difference is not large, however. On the contrary, the size difference between the newly referred lower dentition and the material previously described by Egi et al. (2005) is much greater (the m3 NMM Bhn-2000 is only 67% the size of the holotype), although the proportion of lower molars (L/Wtrigo, Ltal/Ltrigo in Table 2) is not much different if the large variation observed in the hypodigm is taken into account. Among other early to middle Eocene hyaenodontids, a similar size variation of lower molar length is present in:
the North American proviverrines Arfia shosho- niensis (Matthew, 1915), where m3 length varies between 8.2 mm (AMNH 15747) and 12.5 mm (AMNH 15743), and Prototomus phobos Gingerich & Deutsch, 1989 , where m2 length varies between 5.3 mm (UM 92614) and 7.4 mm (UM 79358) (MM unpublished data); and the limnocyonine Thinocyon velox Marsh, 1872 (Morlo & Gunnell 2003). Large size variations of about 30% are also known from Oligocene Hyaenodon (Mellett 1977; Lange-Badré 1979; Morlo & Nagel 2006). Various modern carnivorans may also display such a great range, such as Felis sylvestris , where m1 length variation reaches 72% (N = 9; SP unpublished data) or Meles meles where m1 and m2 length variations are up to 74 and 66%, respectively (N = 131; Abramov & Puzachenko 2005).
There are additional, morphological differences between the description of Egi et al. (2005) and our observations of the holotype and new material described here. According to these authors, the paracone of the M1 and M2 of K. lupina (represented only by the holotype in their study) is slightly lower than the metacone (character 11), but slightly larger in basal diameter (character 12 in Egi et al. 2005: 340, table 5); for the latter character, Kyawdawia is coded “0: paracone basal diameter relatively larger than the metacone diameter” (character 12 in Egi et al. 2005: table 5). However, the relative height or size of these cusps on the only preserved holotype M1 cannot be estimated. In the dentition of
the holotype, left and right M1 and right M2 are
Creodonts from the Pondaung Formation, Myanmar
TABLE 2. — Measurements (in mm) of the lower dentition of Kyawdawia lupina Egi et al., 2005 . Abbreviations: L, maximum length; Ltrigo, maximum trigonid length; Wtrigo, trigonid width; Wtal, talonid width; Ht proto, protoconid height. Measurements of Egi et al. (2005) are based on NMMP-KU 1661 (holotype) and NMMP-KU 1288; number in parentheses is the sample size if greater than 1.
far too heavily worn and/or damaged for a precise measurement of height and basal diameter, with the paracone being even more damaged than the metacone on the M2; the left M2 is well preserved and displays the metacone slightly longer than the paracone.The upper molars (M1 and M2) described here (NMM Bhn-1040, NMM Tdg-800, NMM Kmi-500) show that the height of the two cusps is roughly similar and that the metacone is more elongated than the paracone, which mainly results from the distal development of the postmetacrista on the former. The basal diameter of the paracone is rather lesser than that of the metacone on M1 and M 2 in Kyawdawia lupina . The relative height and size of the paracone and metacone of M1-2 is therefore similar to that stated in Paratritemnodon and Masrasector , contra Egi et al. (2005). The heavy wear and/or poor preservation obscure other details on the upper molars of the holotype such as the presence or absence of the conules in M1 and M2. Egi et al. (2005) state that the paraconule is “indistinct or absent” on M1 and that the metaconule is “present only on M2” in the holotype (characters 20 and 23 respectively; Egi et al. 2005), which is exactly the same as in Masrasector . While from our observations, the molars on the holotype are much too worn or too damaged for determining whether
the conules are present or absent. The M1 and M2 described here have small but distinct conules, the paraconule being larger than the metaconule. Character states are therefore not confirmed for both teeth, Kyawdawia having the primitive state for the two characters. In Paratritemnodon, Kumar (1992) mentions a paraconule on M1 and M2, and a tiny metaconule only on M2 (contra Egi et al. 2005); Masrasector also has distinct conules at least on M1 ( Crochet et al. 1990), which contrasts with the definition of the character states for this genus in Egi et al. (2005). Unfortunately, there are no published data for upper teeth of Masrasector from the Fayum.
The lower dentitions referred to here agree with the previous description of K. lupina . Thus, as in the holotype, the m2 of NMM Pkg-1051 has a metaconid very reduced and smaller than the paraconid, and a paraconid and protoconid subequal in length. Such a reduction of the metaconid is known in Paratritemnodon , Masrasector , Kyawdawia and distinguishes the m2 described here from that of Yarshea cruenta . However, it is known from a couple of other proposed proviverrines, e.g., Alienetherium Lange-Badré, 1981 , Eurotherium Polly & Lange- Badré, 1993, Prodissopsalis Matthes, 1952 , and Matthodon Lange-Badré & Haubold, 1990 (Morlo & Habersetzer 1999), and, thus cannot easily be
used as a diagnostic feature. In addition, in NMM
Peigné S. et al.
Pkg-1051, the length of the m2 trigonid relative to the height of the m2 protoconid, which equals 0.72, corresponds to that found in the genera of the so-called “Afro-Asian proviverrines” (see Egi et al. 2005; Fig. 4A View FIG ); this character cannot be measured in the holotype. Finally, like in the holotype of K. lupina , the labial cingulid is continuous and present on the whole labial side in NMM Pkg-1051. The main difference observable on the lower dentition between the holotype and the new material is the relative greater width of the talonid of m2. It is subequal to the trigonid width in NMM Pkg- 1051 while the talonid is clearly narrower than the trigonid in the holotype of Kyawdawia lupina , based on measurements in Egi et al. (2005). On m2, however, the talonid is wider than in m3 and it is less triangular in occlusal shape in both the holotype and in NMM Pkg-1051. Proviverrine taxa in which the m2-talonid width is subequal or greater than its trigonid width are numerous and present on all continents, e.g., Arfia , Proviverra , Lesmesodon , Proviverroides , Allopterodon Ginsburg in Ginsburg et al., 1977, Cynohyaenodon trux Van Valen, 1965 (but not Cynohyaenodon cayluxi Filhol, 1873 , the only species of the genus Egi et al. 2005 included into their analysis), Metasinopa Osborn, 1909 or Anasinopa Savage, 1965 . This character cannot be used to establish a clade Masrasector - Paratritemnodon - Kyawdawia even if Allopterodon , Cynohyaenodon , Proviverra , and Lesmesodon (all are closely related to each other) are much smaller than Kyawdawia and display, e.g., completely separate paracone and metacone and a shorter metastyle on M1, and an unreduced m2 metaconid (which is reduced in Cynohyaenodon cayluxi and Cynohyaenodon ruetimeyeri (Depéret, 1917)) (Morlo & Habersetzer 1999) . Metasinopa and Anasinopa differ by lacking the complete labial cingulid on m3, which is diagnostic of the suggested clade Masrasector - Paratritemnodon - Kyawdawia , and by having a less mesiolingually projected protocone on M1-2 and a narrower talonid with a lower entoconid crest on m3. Moreover, both genera (and Buhakia ) are now included in Hyainailourinae (Morlo et al. 2007) as was their close relative Dissopsalis already (Polly 1996). The morphology of the m3 of NMM Bhn-
2000 and NMM Pkg-1050 is very similar to that of the m3 of the hypodigm specimens of Kyawdawia lupina . The proportion of the trigonid compared to the total length (Ltrigo/L) and the proportion of the tooth (L/Wtrigo) are exactly the same (Table 2). In contrast, the talonid/trigonid proportion in NMM- Bhn 2000 and NMM-Pkg 1050 is much different from that in Yarshea cruenta . An additional difference is that the labial cingulid is restricted to the paraconid in the latter, while it is extended distally in K. lupina and in NMM Bhn-2000 and NMM Pkg-1050. The talonid of m3 is shorter than the trigonid, with a talonid/trigonid length proportion greater than 0.5, as in diverse North American proviverrines like Arfia , Prototomus , Proviverroides , and partly Sinopa Leidy, 1871 , European Allopterodon , Proviverra , Lesmesodon , but not Cynohyaenodon (only very occasionally in C. trux ; MM unpublished data), and Masrasector ( Egi et al. 2005) . The talonid of m3 is distinctly narrower than its trigonid (like in nearly all proviverrines). The m2 and m3 of this new material share the absence of distinction between the hypoconulid and the entoconid and the presence of a notch separating the hypoconid and the hypoconulid with African and Asian proviverrines. However, these characters are typical as well for the “North American” proviverrines (see Morlo & Gunnell 2003). It could serve as an argument that African and Asian, but not European proviverrines, root in taxa related to Prototomus or Arfia rather than those close to Proviverra .
| UCMP |
University of California Museum of Paleontology |
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.