Saurodektes rogersorum

Hamley, Tim, Cisneros, Juan C. & Damiani, Ross, 2021, A procolophonid reptile from the Lower Triassic of Australia, Zoological Journal of the Linnean Society 192 : -

SAURODEKTES ROGERSORUM

Following the naming of theTriassic owenettid Owenetta kitchingorum by Reisz & Scott (2002), the polyphyly of the South African genus Owenetta was demonstrated by Modesto et al. (2003) and subsequently confirmed by other studies ( Cisneros et al., 2004; Piñeiro et al., 2004; Modesto & Damiani, 2007; Säilä, 2008; Tsuji et al., 2013). However, the attribution of the Triassic owenettid to the genus Owenetta has not been questioned since Modesto et al. (2003) and was, in fact, reinforced after the description of the owenettids Candelaria barbouri Price, 1947 from Brazil ( Cisneros et al., 2004) and Ruhuhuaria reiszi Tsuji et al., 2013 from Tanzania.

The taxon Saurodektes rogersorum Modesto et al., 2003 ( Modesto et al. 2004) was erected to accommodate a new owenettid specimen from South Africa that was anatomically similar to ‘ Owenetta ’ kitchingorum , but lacked the autapomorphies of the latter. However, two of the three diagnostic characters of ‘ O ’. kitchingorum , namely (1) the presence of a caniniform region in the maxilla and (2) medial contact between prefrontals anterior to frontal ( Reisz & Scott, 2002), cannot be observed in the holotype of S. rogersorum due to the incompleteness of its skull. The third diagnostic character of ‘ O ’. kitchingorum is the presence of postparietals ( Reisz & Scott, 2002). However, postparietals are also recorded in the owenettid Barasaurus ( Meckert, 1995; Cisneros, 2008a), making that character an invalid autapomorphy for ‘ O.’ kitchingorum . The fact that postparietals are present in only some Barasaurus individuals, a taxon for which a large number of specimens is available, suggests that these minute bones are easily detached post-mortem from the skull roof and lost in many specimens.

The first diagnostic character for Saurodektes rogersorum , a quadratojugal anterior process that constricts the ventral cheek emargination (Modesto et al., 2003), is problematic because the condition is present in some specimens of ‘ O. ’ kitchingorum ( Fig. 25). Furthermore, all owenettids and several procolophonids, including Eo. yurrgensis , possess a quadratojugal with an anterior process that delimits the ventral embayment to some degree. Another diagnostic character of Saurodektes rogersorum , namely the acute ventral process of the squamosal that overlays the quadratojugal (Modesto et al., 2003), is also visible in the holotype of ‘ O.’ kitchingorum (pers. obs. on BP/1/4195a by JCC), although not correctly depicted in one illustration presented by Reisz & Scott (2002: fig 2d; but see fig. 2g, where the squamosal acute ventral process is shown). This would leave S. rogersorum with only one valid diagnostic character, namely a squamosal with an anteroventral margin that is deeply embayed within the otic notch (Modesto et al., 2003). However, this character is actually correlated with the last one. The acute ventral process of the squamosal is, in part, responsible for the architecture of a wide embayment in this area, a fact that is acknowledged by Modesto et al. (2003).

Following from the above considerations, ‘ Owenetta ’ kitchingorum becomes virtually indistinguishable from Saurodektes rogersorum . In addition, both taxa occur at the same stratigraphic level in the South African Beaufort Group. Moreover, the type locality of the S. rogersorum (Barendskraal Farm, Inxuba Yethemba Local Municipality, Eastern Cape Province, South Africa) has also produced ‘ O ’. kitchingorum specimens ( Damiani et al., 2004). Therefore, the species ‘ O’. kitchingorum is here considered a subjective senior synonym of S. rogersorum and the genus name Saurodektes becomes available as a replacement for ‘ Owenetta ’.