Myorycteropus, MACINNES, 1956

Lehmann, Thomas, 2009, Phylogeny and systematics of the Orycteropodidae (Mammalia, Tubulidentata), Zoological Journal of the Linnean Society 155 (3), pp. 649-702 : 678-679

GENUS MYORYCTEROPUS MACINNES, 1956

Emended diagnosis: Small-sized genus (50–60% of the general size of the extant aardvark) that is distinct from all other known Orycteropodidae by the shallow lingual groove that is uniquely present on the M 2; the perpendicular orientation, in caudal view, of the upper molars alveoli (except for O. crassidens ); on the humerus, the proximal extension of its brachial crest above the level of the deltoid tuberosity, the presence of a medial crest next to the entepicondylar crest, and the very important breadth of the distal epiphysis (breadth to length ratio of 46.9%); the presence of a cnemial tuberosity on the tibia; the drop-shaped sustentacular facet, as well as the crescent-shaped, medioventrally to laterodorsally oriented, condyle for the navicular, and the vertical and negligible cotyloid facet on the talus. It also shares the developed and laterally projected deltoid crest common to all Orycteropodidae except L. guilielmi and A. abundulafus . Some significant features of this genus appear to be primitive: the very shallow angle between the mandibular rami (inferior to 60°), and the mediodorsal to lateroventral orientation, in proximal view, of the caput femoris.

Remarks: This diagnosis merges and complements the original one given by MacInnes (1956), and the emended one given by Pickford (1975). The general size of the animal is probably between 50 and 60% of that of the extant species, based on the length of the limb bones of the holotype. Pickford (1975: 68) followed MacInnes (1956) with this estimation in his diagnosis. Later in the text, however, Pickford (1975: 79) specified that the size of the animal must have been 66% of that of O. afer without giving any justification. The autopodes of M. africanus are indeed proportionally longer. For instance, the length of the metacarpal V is 84% of that of its extant counterpart, according to MacInnes (1956: 19). Nonetheless, the general size of the animal is more reliably estimated on the basis of the limb bones. As previously discussed, the bilobation of the M 3 and M 3 is now rejected in the diagnosis. Furthermore, MacInnes (1956) and Pickford (1975) suggested that the extent of the third trochanter crest on the diaphysis was a valid character. However, a study of 32 extant aardvark specimens gave values ranging from 8.5 to 15.3%. Although these values are significantly different from the value given for the holotype of M. africanus (22.5%) by MacInnes and Pickford, their range of variation shows that this feature is likely to be very variable in other species as well. Moreover, the difficulty to accurately take this measurement further reduces its significance. Finally, no complete femur of M. africanus is known, so that the proportional extent of the third trochanter can only be estimated. Therefore, I recommend rejecting it from the diagnostic features.

Type species: Myorycteropus africanus MacInnes, 1956 .

Other forms: Myorycteropus cf. africanus Kenya ( MacInnes, 1956; Pickford, 1975), Myorycteropus sp. Arrisdrift ( Pickford, 1996, 2003), as well as possibly aff. Myorycteropus chemeldoi ( Pickford, 1975) , aff. Myorycteropus minutus ( Pickford, 1975) , and aff. Myorycteropus sp. Napak ( Pickford, 1975).

Geographic distribution: Rusinga Island and Mfwangano ( Kenya), as well as possibly Arrisdrift ( Namibia), Napak ( Uganda), and Songhor and Tugen Hills ( Kenya).

Temporal distribution: Early Miocene, between 20 and 17 Mya.

Remarks: The possible aff. Myorycteropus species ‘ O. minutus ’ and ‘ O. chemeldoi ’ are described below. Moreover, specimens from Arrisdrift ( Namibia) and Napak ( Uganda) are also referred to this genus ‘faute de mieux’, and are discussed below.

SPECIES MYORYCTEROPUS AFRICANUS MACINNES, 1956

Emended diagnosis: Same as for the genus.

Holotype: NHM M 21500–NHM M 21538 (formerly field number 1264′50): parts of skull, mandible and associated skeleton of a subadult individual (having reached adult size); housed at the NHM, London.

Remarks: The last piece (NHM M 21538) was not mentioned in the original description made by MacInnes (1956), but does belong to the skeleton. The specimen NHM M 21517 does not apparently exist.

Paratype: Fragment of right mandible ( NHM M 21539 ) ; fragment of right mandible ( NHM M 21540 ) ; fragment of mandible (field number MW 61 ′52); right M 3 or P 4 (?) (field number 369′52); metacarpal V ( KNM RU 5968 ) (the latter specimen is housed at the NMK).

Type locality and age: Rusinga Island ( R.2–4 Series , Kulu-Waregi), Lake Victoria ( GPS: 0°25′S, 34°11′E) ( Kenya). This site is dated to around 17.8 Mya ( Drake et al., 1988) .

Main occurrences: Rusinga Island and Mfwangano ( Kenya).

Additional material: Several isolated basipode elements from Mfwangano, and elements of the hindlimb from Rusinga, described by Pickford (1975), or yet unpublished, housed at the NMK, Nairobi (KNM MW 183; KNM MW 480; KNM RU 8263; KNM RU 8269; KNM RU 8307–8308; field numbers MFW 566′55, MFW 781′55, MFW 64′52).

Remarks: Specimen KNM MW 183 (relabelled) corresponds to the published specimen KNM MW 83 of Pickford (1975). The phalanx KNM RU 3059 described by Pickford (1975) is not a Tubulidentata .

Discussion: In his original description, MacInnes (1956: 2) suggested that the infraorbital foramen was situated ‘very nearly in the middle of the total vertical height of the skull’, and thus that the skull was less elevated than in the modern aardvark. However, the dorsal deformation of the specimen and the absence of data concerning the development of the frontal and nasal bones hinder such conclusions. MacInnes (1956: plate 1, figs 1, 4) represented the mandible of the holotype with a complete articular condyle in connection with the rest of the mandible. The current condition of that specimen is different. Indeed, the articulation surface itself is damaged now, and the articular condyle is separated from the rest of the mandible, without a perfect contact surface. Thus, it is not possible to reconstruct or to confirm the restoration of the mandible made by MacInnes. Likewise, MacInnes (1956: 6) proposed a reconstruction of the length of the mandible based on an index joining the ‘distance from the posterior border of the M 3 to the anterior tip of the horizontal ramus’ in the modern aardvark. This index is 28.5% according to MacInnes (1956), and has been confirmed by the analysis of a large sample of O. afer (mean = 28.3% ± 2.3; n = 61). Nonetheless, the same index is 31.6% for O. crassidens (holotype), 37.6% for aff. A. pottieri (holotype), and 34.6% for A. abundulafus (holotype). These results show first that the index is not constant for all Orycteropodidae , and second, that the antemolar part of the mandible is relatively longer in the extant aardvark than in the fossil ones. Therefore, the calculated length of the mandible of M. africanus is not confirmed in this study.

On the scapula of M. africanus , the anterior border of the acromion, the edge directly opposite to the metacromion, is slightly expanded. MacInnes (1956: 13) interpreted this structure as suggestive of a ‘strong trapezius in direct opposition to the powerful deltoid’. However, this edge is the insertion surface of the subclavius muscle ( Thewissen & Badoux, 1986), which is a muscle involved in the digging process, and is not an antagonist to the deltoid. This feature strengthens the idea of a powerful digging animal. The two humeri of the holotype have been preserved. However, the diaphysis of the left one is broken and glued together above the deltoid crest. This reconstruction is contested here, as the right humerus has an intact diaphysis and is distinctly longer. I propose that the left humerus, used by MacInnes (1956) throughout his description, is distorted, and is missing a portion of its shaft. The ‘sharp curvature of the shaft’ described by MacInnes (1956: 15), and followed by ( Patterson, 1975: 207), was probably based on this distorted left humerus. The shaft of the right humerus is more similar to those of the other Orycteropodidae . Note that the distal mediolateral breadth to length ratio for M. africanus has been accordingly calculated using the length of the right humerus and the distal breadth of the left humerus (the right distal epiphysis being damaged). Another point about the humeri of the holotype concerns the proportion of the proximal epiphysis. Indeed, the left humerus shows subequal dorsoventral and mediolateral breadths, whereas the right humerus shows a longer dorsoventral breadth. This intraindividual variation probably results again from post-mortem deformations, but no indication of such alteration can be seen on the proximal epiphysis of the humeri. Although we know that the left humerus suffered from distortions (see above), it is the configuration of the right humerus that is out of the ordinary among Orycteropidinae. In consequence, the peculiarities of the proximal humerus of M. africanus will not be taken into account for the diagnosis until further material is discovered.

A specificity that is unique among Orycteropodidae is the important distal development of the medial distal condyle of the femur in M. africanus . In fact, this condyle is the distalmost point of the femur, whereas in all other aardvarks, it is the lateral condyle. This feature has never been highlighted before, and is probably related to another peculiarity displayed by the proximal epiphysis of the tibiofibula. Indeed, in dorsal view, the proximal articulation surface of the tibio-fibula is tilted approximately 45° relative to the diaphysis, which is sharply curved. Patterson (1975: 207) already brought attention to the fact that MacInnes (1956) reconstructed the tibiofibula with such an oblique proximal epiphysis. He concluded that such a tibia would fit a femur with an outward inclination (45°) that is ‘surely an impossible position’. The left hindlimb of the holotype is the only known specimen of M. africanus where the articulation of the knee can be completely observed. Thus, the present configuration could be an isolated pathologic knee joint. Nonetheless, no fracture or gap, suggesting a reconsolidated injury or a post-mortem deformation, can be seen on the tibia and femur. Besides, the left tarsus of the holotype shows no unusual structure that would compensate for such an oblique orientation of the tibia. The distal epiphysis of the tibia is missing in all specimens, so that the ankle joint configuration remains totally unknown for M. africanus . Considering the unique orientation of the caput femoris in M. africanus , and thus the peculiar pelvis/femur articulation, it may not be surprising to also have a remarkable knee, and perhaps also ankle, joint. Although very unlikely, the hypothesis of a 45° angle between femur and tibia must not be rejected until new material has been discovered. These characters must be integrated in the diagnosis of the genus if confirmed.