© 

A new large philisid (Mammalia, Chiroptera, Vespertilionoidea) from the late Early Eocene of Chambi, Tunisia

ANTHONY RAVEL, LAURENT MARIVAUX, RODOLPHE TABUCE, MUSTAPHA BEN HAJ ALI, EL MABROUK ESSID, MONIQUE VIANEY-LIAUD

First published: 18 September 2012

 

https://doi.org/10.1111/j.1475-4983.2012.01160.x

Citations: 13

SECTIONS

PDF

TOOLS

 

SHARE

Abstract

Abstract: Among the new dental remains from the late Early Eocene of Chambi (Kasserine area, Tunisia) is a large-sized upper molar of a new bat species, Witwatia sigei nov. sp. (Chiroptera, Vespertilionoidea, Philisidae), described herein. The locality of Chambi has revealed evidence for an early appearance of two modern microchiropteran superfamilies in Africa: Dizzya exsultans, a Philisidae, which is considered to be an archaic Vespertilionoidea, and an indeterminate Rhinolophoidea. In addition to D. exsultans, the new species, W. sigei, is the second representative of the Philisidae in this locality. W. sigei extends back to the late Early Eocene the occurrence of the genus Witwatia, which was previously only reported from the early Late Eocene of the Fayum (BQ-2, Egypt). By analogy with the largest extant microbats, the large size of Witwatia suggests a tendency to the opportunistic diet of this taxon, thereby contrasting with the strict insectivory characterizing primitive bats found in other continents in the same epoch.

From the Afro-Arabian Paleogene, nearly all African bat fossils are limited to the northern regions of the continent and are essentially documented by dental remains and jaw fragments (Sigé 1985, 1991; Gunnell et al. 2003, 2008; Tabuce et al. 2005; Eiting and Gunnell 2009). A notable exception is the Middle Eocene bat Tanzanycteris mannardi (Gunnell et al. 2003) from Mahenge, Tanzania, known by a partial skeleton. Despite the fragmentary nature of the African fossils, the bat record reveals that chiropterans were already well diversified in these regions by the early Paleogene. The oldest occurrence of the order Chiroptera in Africa is an ‘eochiropteran’ (or archaic bat) recently discovered from the middle Early Eocene of El Kohol, Algeria (Ravel et al. 2011). Beyond this archaic bat, the other Paleogene localities in North Africa and Arabia yield only microchiropteran families of modern aspect as early as the Early Eocene. The late Early Eocene locality of Chambi (Fig. 1; Tunisia) documents an indeterminate rhinolophoid and a Philisidae (Dizzya exultans), which is considered to be an archaic vespertilionoid (Sigé 1991). The Early Oligocene of Taqah, Oman, documents also Philisidae and modern families, including members of Nycteridae, Hipposideridae and Emballonuridae (Sigéet al. 1994). The Fayum deposits (Late Eocene – Early Oligocene, Egypt) have yielded Philisidae as well and modern bat families, including Rhinopomatidae, Megadermatidae, Emballonuridae and Vespertilionidae (Sigé 1985; Gunnell et al. 2008). More precisely, from the Quarry BQ-2 of the Fayum, Gunnell et al. (2008) described the philisid genus Witwatia, the largest African Paleogene bat; this genus is represented in the Egyptian deposits by two species, W. schlosseri and W. eremicus.

Figure 1

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

Location map of the Chambi locality in Tunisia, situated in the Djebel Chambi National Park approximately 15 km, West to Kasserine.

In the course of our paleontological field work in Tunisia, and after careful acid etching of freshwater limestone from Chambi, numerous new bat remains have been discovered. Among this material was the particularly remarkable recovery of a large upper molar attributable to the genus Witwatia. This specimen, particularly well preserved, represents a new species (W. sigei sp. nov.), which exhibits distinctive philisid characteristics. The tooth is similar in size and morphology to W. schlosseri and W. eremicus. W. sigei represents the oldest representative of Witwatia in the African fossil record, thereby suggesting an early appearance of the large-sized philisids.

Beyond Chiroptera, the locality of Chambi has yielded a very diversified faunal and floral assemblage including indeterminate amphibians and squamates, a possible peradectid marsupial (Kasserinotherium tunisiensis; Crochet 1986) and several eutherians such as a putative erinaceomorph (Chambilestes foussanensis; Gheerbrant and Hartenberger 1999), a strepsirhine primate (Djebelemur martinezi; Hartenberger and Marandat 1992), a zegdoumyid rodent (Zegdoumys sbetlai; Vianey-Liaud et al. 1994), a stem macroscelidid (Chambius kasserinensis; Hartenberger 1986), a hyracoid (Titanohyrax tantulus; Court and Hartenberger 1992) and charophytes (Raskyella cf. sahariana; Mebrouk et al. 1997). The charophytes and some mammals including Zegdoumys sbetlai, Titanohyrax tantulus and Chambius kasserinensis are very similar to those from the middle member of Glib Zegdou (Gour Lazib) in Algeria dating from the late Early to early Middle Eocene (Adaci et al. 2007). In the Chambi sedimentary sequence, the fossiliferous lacustrine limestone represents a distinct layer, situated at the base of the continental formation. The stratigraphic position of the fossiliferous level and the comparison of the charophytes and mammalian fauna of Chambi with that of the other Paleogene localities of Algeria (El Kohol and Gour Lazib) and Morocco (Ouarzazate Basin and Ouled Abdoun) suggest that Chambi dates probably from the late Early Eocene (Hartenberger et al. 2001; Seiffert 2010).

Systematic palaeontology

Order CHIROPTERA Blumenbach, 1779
Superfamily VESPERTILIONOIDEA Van Valen, 1973
Family PHILISIDAE Sigé, 1985
Genus WITWATIA Gunnell, 2008
Witwatia sigei, sp. nov.
Figure 2

Figure 2

Open in figure viewerPowerPoint

Left upper molar Witwatia sigei CB1-230 from Djebel Chambi (Early Eocene). The upper part is an annotated drawing of the upper molar in occlusal view A, by © Laurence Meslin CNRS – 2010. Abbreviations: ectlph, ectoloph; cgl, cingulum; lg, lingual; ms, mesostyle; mtc, metacone; mts, metastyle; mtcst, metacrista; pc, paracone; pcst, paracrista; prc, protocone; prcst, protocrista; prf, protofossa; ps, parastyle. On the lower part, pictures of the upper molar of Witwatia sigei in B, occlusal, C, buccal, D, lingual, E, mesial, and F, distal views. The different tooth views B–F, are the results of the fusion of multi-focus images obtained with an optical stereomicroscope Leica M205C. Scale bar = 1 mm.

Differential diagnosis. Differs from the other species of Witwatia in having a protofossa distally closed by the junction of the postprotocrista, the postcingulum and the lingual cingulum, no talon basin, and a deeper ectoflexus. Upper molar smaller than W. schlosseri but slightly larger than W. eremicus.

Derivation of name. Dedicated to Bernard Sigé, who first extensively described Paleogene bat remains including those from Chambi.

Type locality and age. Chambi, late Early Eocene, Kasserine area, Tunisia (Fig. 1).

Holotype. Left upper molar M2 (Fig. 2); specimen CB1-230. Material housed in the Museum of the Office National des Mines (ONM), in Tunis, Tunisia.

Description. The specimen is a large-sized left upper molar (3.8 × 4.3 mm; Fig. 2). The rectangular outline of the crown, which is relatively elongated transversely, suggests a M2 rather than a M1. The development of the ectoloph occupies the main part of the buccal crown margin and is twice as wide as the lingual part.

 

The crests of the ectoloph are subparallel. Buccally, the preparacrista makes a mesial flexion and joins a well-developed parastyle. The mesostyle is as buccally situated as the parastyle and is connected to the long postparacrista. A deep notch, in distobuccal position to the mesostyle, makes a buccal opening to the wide protofossa.

 

A second minute mesostyle occurs distobuccally to the notch but remains isolated to the premetacrista. This latter is the longest crest of the ectoloph and has a quite similar orientation to the others, but it is curved distally at its buccal extremity. The postmetacrista is slightly shorter than the premetacrista and extends to the metastyle, which is less developed than the parastyle. Deep para- and metafossa overhang buccally the paracone and the metacone, respectively. The paracone is thinner and slightly more lingual than the metacone. The two lingual cusps are inclined buccally. The protocone is strongly developed mesiodistally and compressed buccolingually. It is slightly higher and narrower than both paracone and metacone. The preprotocrista, shorter than the posprotocrista, joins a wide precingulum. The postprotocrista is slightly curved buccally and extends to the postcingulum, which is narrower than the opposite lateral cingulum. A thick lingual cingulum, situated at the disto-lingual edge of the protocone, ends near the most lingual part of the cusp. Another short and thin part of the lingual cingulum extends on the mesiolingual edge of the protocone.

Comparison. Philisid chiropterans are characterized by their great size, upper molars exceeding generally 2 mm wide, with a rectangular outline, the buccal opening of the protofossa separating the postparacrista and premetacrista, the absence of both conules and lophs, the lateral crests of the ectoloph subparallel, the absence of hypocone and an undeveloped talon. The new specimen from Chambi displays all of these features and allows us to refer this tooth to the Philisidae.

​

The oldest representative of this family is Dizzya exsultans (Fig. 3), which was also discovered from Chambi (Sigé 1991). This species is the smallest philisid known thus far. Compared with D. exsultans, Witwatia sigei shares a protofossa distally closed, a broad distolingual cingulum and the absence of development of the talon basin. Beyond the strong difference in size, the new species from Chambi differs also from D. exsultans in having the ectoloph more transversely developed, the crests of the ectoloph subparallel, the postmetacrista and the metastyle more projected buccally, a lingual cingulum interrupted but stronger on the distolingual part of the crown. Furthermore, in contrast to Dizzya, W. sigei has neither conules nor lophs at the base of the lingual cusps.

Figure 3

​

Occurrence and relationship hypothesis for Philisidae. Vampyravus orientalis is not considered here due to the ambiguity of its family attribution and the lack of provenance. The ‘eochiropteran’ from the El Kohol Formation and the Rhinolophoidea from Chambi are included, to emphasize the large difference of body size in the Early Eocene African bats. The short dashed lines correspond to the uncertain relationships. The long dashed lines represent the uncertain stratigraphic position of the localities. The question marks point out the gap in the fossil record during the beginning of the Eocene, the Middle Eocene and the Late Oligocene to the Early Miocene. Scale bar = 2 mm.

The Philisidae are well diversified in the Fayum deposits of Egypt. Philisis sphingis is the most representative species of this group (Sigé 1985; Fig. 3). This species occurs in the late Early Oligocene of the Fayum (Quarry I). The upper molar of Witwatia sigei is wider and slightly longer than the second upper molar of P. sphingis. It differs also in having a broad discontinuous lingual cingulum, a protofossa distally closed, premetacrista and styles more developed, and a deeper ectoflexus. The genus Philisis is also present in the Early Oligocene of Taqah (Sultanate of Oman; Sigéet al. 1994). P. sevketi and other indeterminate species are only documented by fragments, without well-preserved upper molars. Only half of a worn upper molar was illustrated in Sigéet al. (1994) but without detailed description. This specimen seems to have a continuous lingual cingulum and an opened protofossa contrary to W. sigei.

The genus Witwatia was first described in the Fayum deposits from the earliest Late Eocene (BQ-2), by two species: W. eremicus and W. schlosseri (Gunnell et al. 2008; Fig. 3). These two Egyptian species of Witwatia are relatively large (>3 mm) as is W. sigei and show long and curved premetacrista, a parastylar inflexion, a protofossa open buccally, the mesostyle more developed on the postparacrista than on the premetacrista, a wide para- and metafossa and mesiodistally extended protocone. The upper molar of W. sigei is slightly wider and more elongated than the upper molar of W. eremicus, but it is smaller than W. schlosseri, which is the largest species among the Philisidae. Furthermore, the Chambi species differs from W. eremicus and W. schlosseri in lacking the development of a talon basin, the protofossa closed by protocrests, deeper ectoflexus and a wider notch between mesostyles.

Vampyravus orientalis is an enigmatic microbat, known only by a right humerus that was discovered in the Jebel Qatrani Formation of the Fayum (late Eocene or Oligocene; Sigé 1985). Unfortunately, this specimen has no type locality as its provenance remains uncertain. Sigé (1985) did not exclude the possibility that Vampyravus orientalis and Philisis sphingis belong to the same taxon, estimating their size to be nearly the same. In addition, Gunnell et al. (2008) classified V. orientalis with uncertainty among Philisidae. The genus Vampyravus was recently discussed by Gunnell (2010), who considers this taxon to be a primitive stem Emballonuridae or Rhinopomatidae. The body mass estimated from the humerus of V. orientalis is the highest among the bat species from the Fayum (higher than the body mass estimated from the large specimens of Witwatia). Unfortunately, the lack of associated dental material does not allow for a better familial attribution.

The latest representative of the Philisidae occurs in the Early Miocene of Jebel Zelten (Libya; Horáček et al. 2006). This taxon, Scotophilisis libycus (Fig. 3), is documented by fragments of a dentary bearing two molars. The lower dentition is more similar to Philisis sphingis than to the other philisids in having the strong protoconid complex, a high lingual crown wall and a shallow protofossid. The upper molars of this species are unknown.

Discussion

If the Philisidae are considered to be archaic Vespertilionoidea, the phylogenetic position of this extinct family within this superfamily remains uncertain. Sigé (1985) noticed several dental features of Philisis sphingis that are shared with the Vespertilionidae and the Molossidae. He concluded that the Philisidae could have originated from a generalized vespertilionoid after the split between the Molossidae and Vespertilionidae, probably during the Late or Middle Eocene in North Africa. Unfortunately, there is not enough fossil documentation to properly assess this hypothesis. After the discovery of Dizzya exsultans, Sigé (1991) argued for an early differentiation of the philisid lineage, probably during the Early Eocene. Witwatia sigei greatly strengthens this hypothesis (Fig. 3).

Although Witwatia sigei appears as the oldest member of Witwatia, the specimen exhibits very advanced, somewhat highly specialized features that characterize more recent philisid from the Fayum deposits: development of the crests and styles of the ectoloph, absence of conules and lophs, buccal opening of the protofossa separating the postparacrista and premetacrista and large size. These features are associated with the closed protofossa and a broad buccodistal cingulum, plesiomorphies found in primitive bats and also Dizzya exsultans. This suggests the basal position of this species within the Witwatia radiation, but also that philisids have rapidly acquired, probably as early as the Early Eocene, a very autapomorphic morphology (Fig. 3). The two philisids in Chambi are clearly distinct. The strongly marked dental differences reflect a more ancient diversification of the group. The upper molars of Philisis from the Fayum are more similar to the upper molar of D. exsultans from Chambi than those of W. sigei in having a more rectangular outline, crests and styles of the ectoloph less developed, labial cusps more buccally displaced, the postcingulum less wide and a continuous lingual cingulum. As such, Philisis could be more closely related to Dizzya than to Witwatia.

The genus Witwatia includes some of the largest Paleogene bat fossils. Gunnell et al. (2009) estimated the weight of the Fayum bats based on tooth areas ranging from 57 to 116 g for W. schlosseri, and 40 to 89 g for W. eremicus. W. sigei is medium-sized, appearing to be between the two other species, which allows us to estimate its body size to around 100 g. The body size of these bats is comparable to one of the largest extant microbats Macroderma gigas (Gunnell et al. 2008), the body mass of which is 150 g. These large-bodied taxa contrast substantially with the small-sized primitive Eocene bats (‘Eochiroptera’). Among the latter, Icaronycteris index, Archaeonycteris trigonodon, Palaeochiropteryx tupaiodon and Palaeochiropteryx spiegeli have a weight estimated by several body measurements ranging from 6 to 58 g (Simmons and Geisler 1998; Simmons and Conway 2003). Only Hassianycteris magna, the largest ‘eochiropteran’ bat from Messel (early Middle Eocene, Germany), could have a body mass that exceeded 90 g (Smith and Storch 1981; Simmons and Geisler 1998; Simmons and Conway 2003). The size, the effective echolocation and the stomach contents indicate that these archaic bats were essentially insectivorous (Habersetzer and Storch 1992; Simmons and Geisler 1998).

Like the largest modern microbats, which are small vertebrate hunters, Witwatia could have been more opportunistic than the strictly insectivorous primitive bats. Vertebrate predation (amphibians, fishes, reptiles, birds and small mammals) is practised notably in the largest extant forms of phyllostomids, megadermatids, nycterids, noctilionids and vespertilionids (Freeman 1984, 2000; Norberg and Fenton 1988). The dental morphology does not seem to change radically with the type of predation (Freeman 1984; Swartz et al. 2003). Freeman (1984, 1988) attempted to delineate some dental morphological characteristics that would discriminate between strict insectivorous bats and flesh-eating bats. On the upper molars, the length and more anteroposterior orientation of the postmetacrista appear to be associated with vertebrate carnivory because of the resultant increase in shearing potential. However, the buccolingually elongated postmetacrista is not exclusive to the carnivorous bat. The insectivorous Megadermatidae and the Nycteridae have a generally distorted W-shaped ectoloph by the anteroposterior orientation of the postmetacrista. The postmetacrista of the species of Witwatia is subparallel to the other crests of the ectoloph, and it is more buccolingually elongated. The orientation of the postmetacrista, the broad crests of the ectoloph, the wide cusps and the large fossae, which characterize the upper molars of Witwatia, are more efficient for crushing than cutting. However, a crushing function of the molars does not exclude carnivory. For instance, the large, extant vespertilionid Scotophilus gigas, which exhibits molars presumably adapted for crushing (crests of the ectoloph subparallel, broad cuspids and large fossae), can hunt vertebrates including other bats (Freeman, 1984). Additionally, Freeman (1984) has established a relation between the interloph (ridge between upper molar in contact with protoconid) and the intraloph (internal ridge in contact with hypoconid) to discriminate among carnivorous taxa. The vertebrate predators have principally a low intraloph/interloph ratio on the upper tooth row (except for Phyllostomus hastatus and Scotophilus gigas). This ratio is also low for the upper tooth row of W. schlosseri (M1-2 ratio < 1). On the lower molars, the carnivorous bats generally have a smaller talonid than trigonid and a strong reduction in the talonid of the m3, which is associated with the absence of the metacone and metacrista on the M3 (Freeman 1984). The lower molars of W. schlosseri have the talonid as large as the trigonid, the talonid of the m3 slightly reduced and the M3 with a minute metacone and a reduced premetacrista. This suite of dental features does not support the hypothesis of a carnivorous diet. But the strength of the mandible (approximately 145% of the height of the m1), the strong lower canine (twice as high as the lower molar) and the high and broad buccal cusps of the lower molars are features well adapted to flesh-eating diet. In sum, the large size of the Philisidae, most notably Witwatia, is not alone sufficient to support the notion that these bats were obligate vertebrate predators, but some dental features seem to favour an opportunistic diet, which may have included some small vertebrates.

​

​

​

Acknowledgments

​

Acknowledgements. The authors are very grateful to S. Jiquel (ISE-M), B. Marandat (ISE-M) and G. Merzeraud (Geosciences Montpellier) for their field assistance during the expedition assistance in the Kasserine area. Special thanks are extended to A.-L. Charruault for her help both in the preparation of fossils specimens and in the field. We are thankful to B. Sigé for his helpful remarks and discussion. Many thanks to L. Meslin (CNRS) for artworks. This research was supported by the French ANR-ERC PALASIAFRICA Program (ANR-08-JCJC-0017) and CS-UM2 grants. This is ISE-M publication 2012-042.

​

.......................................................................................................................................

T

​

 

 

​

​

​

​

​

…