3.1 Description of the adult

The adult consists of a well-preserved and articulated postcranium including presacral, sacral, and approximately 23 caudal vertebrae, cervical and abdominal ribs (generally 5mm thick in the midshaft region), as well as both complete fore- and hind limbs with pectoral and pelvic girdle (Fig. 1). The skull is absent (Fig. 1). All observable aspects of the postcranial anatomy conform to that of Stereosternum tumidum, as described by Modesto (1999, 2010). Specifically, the unfused intermedium and lateral centrale, the presence of a pisiform, and a diamond-shaped interclavicle characterize this specimen as Stereosternum rather than Mesosaurus (Fig. 1; Modesto, 1999, 2010). The fore- and hind limbs are elongate, characteristic for the aquatic lifestyle proposed for mesosaurs (Modesto, 2006; Piñeiro et al., 2012a, b; Villamil et al., 2016). Size and appearances of the fore- and hind limb stylopodia are similar, although the femora are slightly shorter. The ratio of the midshaft diameter to total length of the femur is around 9.2 %, which lies between an earlier erected 11 % in Stereosternum as compared to around 8 % for Mesosaurus (Modesto, 1999, 2010). Zeugopodial and autopodial elements of the hind limbs are wider in the midshaft region and the proximal epiphyses compared to the ones in the forelimbs (Table 1). All limb measurements are listed in Table 1.

Figure 2Detailed stippling of postcranium of an isolated femur (a) and the mostly articulated juvenile (b), based on the latex cast (MB.R.2089.B). Adult specimen is in grey. Scale bars both equal 10 mm. For abbreviations see Material and Methods section.

Download

3.2 Description of the juvenile material

The juvenile material preserves one mostly complete individual and an isolated long bone element. Although there is considerable intraspecific variation in limb element length and shape in sub-adult mesosaurs (Rossmann and Maisch, 1999), we believe the isolated element belonging to a second individual of Stereosternum is most likely a femur based on size and appearance of the element compared to that of the mainly complete individual (Fig. 2a; Table 1).

The almost complete individual consists of elements of the postcranium, including vertebral column, ribs, gastralia, and fore- and hind limbs (Fig. 2b). The skull is not visible; it may be present in the rock in an underlying layer, or may have been separated from the postcranial on the missing counterpart (Fig. 2b). Body size, if measured from the approximate position of last cervical to first sacral rib, is approximately 30 % of the size of the adult (Table 1). The taphonomic orientation of the skeleton cannot be attributed, as the elements are typically in close association but not completely articulated, but the axis of the skeleton is oriented in the same direction as in the adult and is “stretched out”, rather than curled into a foetal position (Fig. 2b). The mostly complete juvenile is located close to the right hind limb of the adult skeleton and does not appear to have been greatly disturbed pre or post burial, in contrast to the isolated femur, which lies close to the adult left forelimb. Although the skeletal elements are not completely articulated as can be seen in the adult, they are very closely associated and in their approximate anatomical positions.

Cervical vertebrae, located proximal to the forelimbs, are clearly diagnosable. However, dorsal and caudal vertebrae are distinguishable due only to their location in situ (Fig. 2b). While cervical vertebrae are fully developed, the neural arch and pleurocentra of both dorsal and caudal vertebrae are not yet fully fused and are preserved in two adjacent lines, though the lack of definitive anatomy in these bones makes it difficult to identify any individual element (Fig. 2b). Concerning length, the small skeleton retains what appears to be an almost complete tail, and at least 25 vertebrae can be recognized (Fig. 2b). Further caudally, vertebral elements are more difficult to discern, yet there appears to be the continuation of the outline of the soft parts of the tail with occasional bone fragments visible. Adult Mesosaurus and Brazilosaurus typically have more than 60 caudal vertebrae (Modesto, 2010), so it is reasonable to conjecture that the caudal vertebrae in this juvenile Stereosternum numbered more than what can be directly observed here. The ribs are already diagnostically thickened (pachyostotic; most are 2 mm thick in the midshaft) in the juvenile and are scattered throughout the thoracic region (Fig. 2b). Slender gastralia lie in the abdominal region (Fig. 2b).

Identifiable juvenile limb elements preserved in MB.R.2089 include two humeri, three femora, one indeterminate forelimb zeugopodial element, one indeterminate hind limb zeugopodial element, and three metatarsals. Due to the extreme delicacy of the juvenile skeleton, other limb elements may be preserved but obscured under those we can identify, or not differentiated enough from the “matrix” to definitively name. In the forelimb, both humeri do not exhibit characteristics such as the deltopectoral crest, ectepicondylar groove, entepicondylar foramen, and an enlarged entepicodyle that are present in the adult (Figs. 1, 2b). Instead, they retain metaphyseal compression and epiphyseal ridges (Fig. 2b) (sensu Hugi and Scheyer, 2012), indicating low maturity similar to what is seen in pachypleurosaur juveniles (Hugi and Scheyer, 2012). In the hind limbs, all femora show reduced ontogenetic compression (sensu Hugi and Scheyer, 2012) compared with the humeri, yet also lack adult phenotypic features such as patellar groove and muscle attachment sites for trochanter (Fig. 2a, b). One of either the fibula or tibia can be identified per side (Fig. 2b, Table 1). Due to the close association of the zeugopod and the preserved metatarsals, we hypothesize that tarsals are not yet ossified (Fig. 2b), a condition also seen in other juvenile mesosaur skeletons (i.e. BSP 1979, fig. 4 from Rossman and Maisch, 1999). Two faint metatarsals can be identified in the latex cast (MB.R.2089.B). According to their location, we think these might be the middle ones (II & III; Fig. 2b, Table 1). On the other side, one indeterminate metatarsal is preserved (Fig. 2b, Table 1).

Table 2Comparison of growth ratios in the limbs of the two distantly related Stereosternum tumidum (MB.R.2089; this study) and “younginiform” Hovasaurus boulei (sub-adult MNHN 1908-21-8 and adult SAM 6231; after Currie, 1981, table 1).

Download Print Version | Download XLSX

3.3 Growth Rates

We calculated growth rates using limb measurements in Table 1. We find that the juvenile hind limb zeugopod, which is longer in the hind limb as compared to the forelimb in the adult, is already slightly larger (37.5 % of the adult size) than all other limb elements (ranging from 26.5 % in the metatarsal III to 35 % in the forelimb zeugopod), indicating a faster growth rate or an earlier onset of ossification (Table 2).

Further, we compared the established growth rates in MB.R.2089 to earlier published limb growth data from another secondarily aquatic reptile, Hovasaurus boulei from the Upper Permian of Madagascar (Currie, 1981). Hovasaurus is a “younginiform” diapsid reptile (sensu Bickelmann et al., 2009), for which, fortunately much ontogenetic material is available (Currie, 1981). It also displays enlarged hind limbs compared to the forelimbs, which are potentially useful when swimming (Currie, 1981). Comparing the ossification of limb elements in sub-adults and adults of Stereosternum and Hovasaurus, we find that the humerus, forelimb zeugopod, femur, and metatarsals show growth rates or onsets of ossification to be comparable (Table 2). However, the growth rate of the hind limb zeugopod is higher in Stereosternum compared to Hovasaurus (Table 2).