General remarks.
Bartassochersis lopezi gen. et sp. nov. presents fine ornamentation consisting of small vascular openings and a network of undulating and branching vascular imprints, which is relatively typical for proterochersids (Broin 1984; Szczygielski and Sulej 2016, 2019; Szczygielski et al. 2018, 2025; Szczygielski and Dróżdż 2026). As such, the ornamentation is finer than in Proganochelys quenstedtii and Ypomonetikochelys euryaspis (see Jaekel (1916); Gaffney (1990); Scheyer et al. (2022); Szczygielski and Dróżdż (2026)) . There is no clear difference between the ornamentation on the carapace and on the plastron.
Due to the incompleteness of the material, accurate identification of carapacial scute rows (e.g. supramarginal, marginal), number of scutes within each row and scute proportions and shapes (e.g. dorsal versus ventral exposition of the marginal scutes) is, in most cases, either difficult or impossible. For that reason, unless there is evidence to conclude otherwise, we tentatively assume that the basic scute pattern was the same as in Proterochersis spp. – this assumption may, of course, be verified with more complete material in the future, but for now, scute identifications should be considered tentative.
Scute sulci are, in most cases, subtly impressed, relatively narrow and straight, with little to no sinuous undulation typical for adult Triassic turtles (Fraas 1899; Jaekel 1914, 1916; Gaffney 1985, 1990; Szczygielski and Sulej 2016, 2019; Hermanson et al. 2024; Szczygielski et al. 2024, 2025, 2018; Szczygielski and Dróżdż 2026), on either the plastron or carapace, except for the caudoanal sulci. It is uncertain whether this is solely a taxonomic trait of Bartassochersis lopezi gen. et sp. nov. or if it is indicative of a young ontogenetic age. The latter interpretation would be consistent with the less distinct, narrower and less undulating sulci observed in younger individuals of Triassic taxa (Broin 1984; Gaffney 1990; Szczygielski and Sulej 2016; Szczygielski et al. 2018) and with the somewhat raised edges of the sulcus in UM-PDE2-16 (compare with the juvenile Proterochersis porebensisZPAL V. 39/34; Szczygielski et al. (2018)). Nevertheless, it seems unlikely that the individual was a young juvenile and more probably was closer to adulthood, considering its relatively large size and well-developed caudal processes of the plastron (UM-PDE2-02). Neither specimen preserves clear open sutures. However, early shell ankylosis was common in proterochersids (Szczygielski and Sulej 2016, 2019; Szczygielski et al. 2018, 2025; Szczygielski and Słowiak 2022) and, therefore, this fact does not unequivocally support maturity of the individual.
Carapace. UM-PDE2-03 (Fig. 2) is a small fragment of the carapace that viscerally preserves broken neural processes of vertebrae and proximal contacts of two ribs. Externally, the element is crossed by a gently bowed sulcus. The position of that sulcus, relative to the neural processes marking the mid-line of the carapace, indicates that it represents an intervertebral sulcus. However, unlike most intervertebral sulci in Triassic turtles, it is not transverse, but angled obliquely relative to the anteroposterior axis. For that reason, it is recognised here as the posterior edge of a bell-shaped first vertebral scute area, similar to that in Chinlechelys tenertesta (presumably; Lichtig and Lucas (2021)), Keuperotesta limendorsa and Proterochersis spp. (Karl and Tichy 2000; Szczygielski and Sulej 2016; Szczygielski et al. 2018). In Proganochelys quenstedtii, all intervertebral sulci are nearly transverse and the first vertebral scute (possibly either consisting of the fused first and second vertebral scute or one of them invading the area of the other) was clearly longer (Fraas 1899; Jaekel 1914, 1916; Gaffney 1985, 1990; Scheyer et al. 2022; Szczygielski et al. 2025). Intervertebral sulci are not preserved in other Triassic turtle taxa (Rougier et al. 1995; Sterli et al. 2007, 2021; de la Fuente et al. 2021; Szczygielski et al. 2025; Szczygielski and Dróżdż 2026). The surface of the presumed first vertebral scute area is gently concave in UM-PDE2-03, but this seems to be an effect of post-mortem deformation rather than a genuine morphological feature. Unlike in Chinlechelys tenertesta, the Owl Rock Member turtle, Proganochelys quenstedtii and some Proterochersis porebensis, it lacks sagittal ridges or grooves (Fraas 1899; Jaekel 1914, 1916; Gaffney 1990; Karl and Tichy 2000; Joyce et al. 2009; Sulej et al. 2012; Szczygielski and Sulej 2016; Szczygielski et al. 2018; Lichtig and Lucas 2021; Scheyer et al. 2022; Klingman et al. 2025). The anterior edge of the specimen is broken and shows no evidence of a sulcus, so it is uncertain how closely it reflects the original anterior border of the first vertebral scute area. Viscerally, the broken rib necks probably correspond to the first and second right dorsal rib. They are inclined anterolaterally and parallel to each other. The row of neural processes is divided into two portions (most probably corresponding to the first dorsal vertebra and one or two subsequent dorsal vertebrae) separated by a rounded notch. Anteriorly to the point of contact for the first vertebra, on the visceral surface of the specimen, there is a poorly and incompletely preserved depression most likely corresponding to the right scapular pit. If correct, then the distance between the scapular pit and the vertebral column was relatively small, similar to that in Cryptochersis paraxene, Keuperotesta limendorsa, and Proterochersis spp. and smaller than in Proganochelys quenstedtii and Ypomonetikochelys euryaspis (see Fraas (1899, 1913); Jaekel (1916); Gaffney (1990); Joyce et al. (2013); Szczygielski and Sulej (2016, 2019); Szczygielski (2017); Szczygielski and Piechowski (2023); Szczygielski and Dróżdż (2026)). Overall, the relative positions of vertebrae and rib attachments appear roughly consistent with those observed in proterochersids (see Fraas (1913); Joyce et al. (2013); Szczygielski and Sulej (2016, 2019); Szczygielski (2017); Szczygielski and Piechowski (2023); Szczygielski and Dróżdż (2026)).
UM-PDE2-11 (Fig. 3) includes a small part of the presumed anterolateral left carapace margin. The specimen cannot be convincingly interpreted as originating from the inguinal region due to inconsistency of the angle and size of the attached plastron part with the geometry and posterolateral extent of the hypoplastra in all currently known Triassic turtle taxa and morphology that is incongruent with the preserved dorsal portion of UM-PDE2-08, interpreted here as a part of a hyoplastron. The dorsal and ventral faces of the peripheral meet in UM-PDE2-11 at an acute angle, but, in contrast to other Triassic turtles, the element seems unusually tall, centrifugally shallow (that is, not projecting far outwards compared to the corresponding marginal region in other Triassic turtles) and roughly symmetrical dorsoventrally, with no dorsal or ventral deflection of the edge and no marked concavity or convexity to either of its faces. In other taxa, with the possible exception of the Owl Rock Member turtle (Klingman et al. 2025), the anterior peripherals are deeper centrifugally than dorsoventrally, so they project into a pronounced lappet-like margin, the reach of which exceeds their dorsoventral height (Jaekel 1914, 1916; Gaffney 1985, 1990; Karl and Tichy 2000; Sterli et al. 2007, 2021; Szczygielski and Sulej 2016, 2019; Szczygielski et al. 2018, 2025; Scheyer et al. 2022; Szczygielski and Piechowski 2023; Ferreira et al. 2024; Szczygielski and Dróżdż 2026). The specimen preserves only one, ventrally placed sulcus, interpreted here as the marginoaxillary sulcus. It is parallel to the free (marginal) border. The maginopleural or marginosupramarginal sulcus is either indistinct or was located beyond the preserved area, so it is not possible to establish the relative size of the dorsal and ventral exposition of the corresponding marginal scutes. Viscerally, the specimen is concave and bears an obliquely (lateroventrally) inclined, rounded ridge that is dorsally and ventrally delineated by smaller distinct concavities. This structure is most likely the distal end of the first costal rib. The exact position of the specimen along the marginal/peripheral series is ambiguous, but, based on the position of the rib apex and the proportions of the preserved fragment of the axillary buttress, it seems to correspond to the level of the fourth or fifth marginal scute of Proterochersis spp.
UM-PDE2-19 (Fig. 4) is an incomplete fragment of a bridge partially embedded in a small block of coarse-grained rock matrix. The specimen is exposed viscerally, presenting the V-shaped wall of the bridge and, dorsolaterally, showing partial areas of at least three presumed scutes: two pleurals more dorsomedially and one supramarginal, located more ventrolaterally. Assuming the composition and geometry of the scutes in this region are similar to those of Keuperotesta limendorsa, Thaichelys ruchae and Proterochersis spp. (Karl and Tichy 2000; Szczygielski and Sulej 2016; Szczygielski et al. 2018, 2025), this specimen corresponds with the second and third pleural scutes and the second supramarginal scute. The slight flexure of the interpleural sulcus (assuming the ventral end is deflected posteriorly as in those other taxa) suggests that the piece comes from the right side of the carapace. Most of the bone from the supramarginal scute level is heavily damaged and the periphery supporting the protruding marginal scute row is completely broken off. The ventral surface of the marginal scute areas is partially obscured by rock, but partially impressed in the broken region, with a short traceable portion of one intermarginal sulcus. Interestingly, the broken region exposes a pill-shaped area of yellow sediment embedded in bone, which seems to be an infilling of an internal peripheral cavity, possibly comparable to structures observed previously in the peripherals of the australochelyids (de la Fuente et al. 2021) and Ypomonetikochelys euryaspis (see Szczygielski and Dróżdż (2026)) or between the elements of the dermal carapacial mosaic of Proterochersis porebensis (see Szczygielski and Słowiak (2022)). Medially, a break through the ventral lappet of the bridge region shows two layers of bone separated by a narrow slit, most likely due to some overlap between the peripherals and the associated plastron.
UM-PDE2-14 represents the posterior carapacial margin (Fig. 5) as an elongated element with one natural, deflected border, so the specimen is comma-shaped in the anteroposterior aspect. Based on the preserved fragment of another scute area located at the base of the presumed marginal, the free edge is deflected dorsally, which is consistent with the middle (bridge level) and posterior carapace periphery in proterochersids (Szczygielski et al. 2018, 2025; Szczygielski and Sulej 2019; Szczygielski and Dróżdż 2026). The ventromedial surface of the specimen is damaged, but still retains a layer of cortical bone and bears indistinct longitudinal relief suggesting attachment of the body wall, which indicates a lack of osseous contact with the plastron at this location. As there is no evidence for a ligamentous bridge in any Triassic turtle and UM-PDE2-11 suggests a fully ossified contact, the interpretation of UM-PDE2-14 as a posterior part of the carapace is supported. The free edge is very gently scalloped and bears two intermarginal sulci, the direction of which (presumed posterolateral based on Proterochersis spp.; Szczygielski et al. (2018)) suggests that the specimen represents the right side of the carapace. This seems consistent with the location of the presumed pleural or supramarginal scute area, suggesting anteriorly constricted and posteriorly more dorsally exposed marginal scutes in the area immediately behind the bridge (Szczygielski et al. 2018, 2025; Szczygielski and Sulej 2019). Outside of that scute area, the proximal (delineating the base of the scute) sulci are not preserved in the dorsal aspect and no sulci are traceable ventrally, but, based on the presumed body wall attachment, the marginals had a somewhat smaller exposition ventrally than dorsally. Overall, the carapace margin is more reminiscent of proterochersid turtles rather than the spiky and more dorsoventrally symmetrical carapace periphery of Proganochelys quenstedtii (Fraas 1899; Jaekel 1914, 1916; Gaffney 1985, 1990; Szczygielski and Sulej 2016, 2019; Szczygielski et al. 2018, 2025; Czepiński et al. 2021; Scheyer et al. 2022; Szczygielski and Dróżdż 2026).
UM-PDE2-12 (Fig. 6) is a relatively large fragment of the carapace and includes three partial scute areas separated by a Y-shaped sulcus. Due to the size of the largest scute area, it most likely corresponds to partial areas of a vertebral and two pleural scutes, but identification as a pleural and two supramarginal scutes cannot be definitively refuted. The direction of bending of the shorter preserved sulcus, considered here as an interpleural sulcus (posteriorly concave), suggests that the specimen may represent the left side of the shell (compare with, for example, Szczygielski et al. (2018, 2025)), but, because only a small portion of that sulcus is preserved, this designation is not certain. In contrast to most other Triassic turtles (Gaffney 1985, 1990; Jaekel 1914, 1916; Karl and Tichy 2000; Sterli et al. 2007, 2021; Szczygielski and Sulej 2016, 2019; Szczygielski et al. 2018, 2025; Scheyer et al. 2022; Szczygielski and Dróżdż 2026), the presumed pleural scute areas are flat and show no presence of pleural bosses, although only small parts adjacent to the pleurovertebral sulcus are preserved. Only in Proganochelys quenstedtii the pleural bosses are located relatively far from that sulcus and often poorly marked or even absent (SMNS 17203), either due to ontogenetic or intraspecific variability or preservational factors (Gaffney 1990; Scheyer et al. 2022; Szczygielski et al. 2025). Viscerally, the specimen is featureless, with the exception of imprints of intercostal vasculature or nerves (Bojanus 1821; Gaffney 1990; Szczygielski and Sulej 2019). The direction of those imprints is roughly perpendicular to the longer (presumed pleurovertebral) sulcus and, thus, consistent with the proposed interpretation of the scute pattern.
UM-PDE2-13 (Fig. 7) consists of two flat fragments with no natural borders that fit together. Their external surfaces bear five narrow and straight sulci. Three of those meet around the same area close to one of the edges of one of the plates, but with a slight offset, so none of them can be identified as a possible mid-line sulcus of the plastron. Two of these five sulci, including the longest one, are parallel to each other, whereas the remaining two form an obtuse angle. The longest sulcus continues on to the other plate, which also bears short portions of another two sulci, subperpendicular to the longest one, in its two opposing ends. One of those short sulci is set at a slight angle relative to the longest one and appears to be a continuation of one of the shorter sulci of the previous plate. This asymmetrical layout, revealing an outline of a polygonal scute and lacking a single identifiable mid-line sulcus, suggests that the fragment represents a part of the carapace, but it is difficult to establish its precise position. In Proterochersis spp. and Keuperotesta limendorsa, a comparable forked combination of sulci may be present either in the anterior part (involving the areas of the third and fourth marginal, the first vertebral and the first pleural scutes) or in the posterior part (involving the areas of the twelfth and thirteenth marginal, fifth vertebral, and fourth pleural scute) of the carapace (Szczygielski et al. 2018, 2025; for example, Szczygielski and Sulej (2016)). The thinness, uniform, and featureless visceral surfaces of UM-PDE2-13 argue against its interpretation as an anterior carapacial part, because this region is expected to be thickened and geometrically complex in the area of the body wall attachment (Szczygielski and Sulej 2016, 2019; Szczygielski and Piechowski 2023; Szczygielski et al. 2025; Szczygielski and Dróżdż 2026). In Proganochelys quenstedtii, a potentially similar layout of sulci may be also present in the posterior part of the carapace, between the areas of the third and fourth pleural and the ninth and tenth marginal scute (Fraas 1899; Jaekel 1914, 1916; Gaffney 1985, 1990; Scheyer et al. 2022; Szczygielski et al. 2025). Although the interpleural sulcus in Proganochelys quenstedtii always seems to precede the intersupramarginal sulcus (Fraas 1899; Gaffney 1985, 1990; Jaekel 1914, 1916; Scheyer et al. 2022; Szczygielski et al. 2025), the relative position of the corresponding pleurovertrebral and intermarginal sulci in Proterochersis spp. is variable (Szczygielski et al. 2018) and, thus, cannot be used as an indicator of body side. However, the presence of a short perpendicular sulcus at the medial corner of UM-PDE2-13, presumed to be a pleurovertebral sulcus, would suggest that UM-PDE2-13 comes from the right side of the carapace, if the scute layout were comparable to that in Proterochersis spp. and Keuperotesta limendorsa. Similar to UM-PDE2-12, the specimens are flat and show no evidence for scute bosses. The bones gradually become thicker towards the presumed centre of the carapace.
UM-PDE2-21 (Fig. 8) is another fragment that lacks natural borders and has a single, straight sulcus on its external surface. There is a recessed area in one of the corners with a bowed, convex perimeter. This area is lined with relatively smooth compact bone with no obvious signs of damage, but also no clear sutural features. It is uncertain whether it represents a peculiar form of bone delamination or compaction (rather unlikely because there is no evidence for trabecular bone in that area), an embayment for a neighbouring bone (unlikely due to the lack of a clear sutural surface, which is present in such cases) or a slot for an element of the dermal carapacial mosaic (Szczygielski and Słowiak 2022; more likely, because these areas are typically relatively smooth; Szczygielski and Sulej (2019)). The structure does not show reactive bone formation, increased vascularity, exposed trabecular bone or depth variation, so it appears to be non-pathological. The specimen approximates to the morphology observed in Proterochersis porebensisZPAL V. 39/419, which is a costal fragment with a slot presumably for an element of the dermal carapacial mosaic (Szczygielski and Sulej 2019), supporting this interpretation. Unfortunately, the walls of the recessed area are obscured by sediment that obscures their morphology.
UM-PDE2-04 (Fig. 8) is a thin plate that is subtly convex externally and concave viscerally. Viscerally, it bears a subtle groove reminiscent from imprints of intercostal vasculature or nerves (Bojanus 1821; Gaffney 1990; Szczygielski and Sulej 2019) and, thus, may represent a fragment of two joined costal plates. Alternatively, the groove may be a post-mortem surface scratch, although its smooth morphology and straightness seem more consistent with a non-damage-related origin.
Plastron. Unfortunately, the material of Bartassochersis lopezi gen. et sp. nov. lacks the anteriormost part of the plastron, which is amongst the most diagnostic regions of the shell in Triassic turtles (e.g. Broin (1984); Szczygielski et al. (2025); Szczygielski and Dróżdż (2026)). The posterior entoplastral process is, however, represented by UM-PDE2-01 (Fig. 9), but is only preserved within a short distance around the level of the axillary notch. Nevertheless, it exhibits a distinct posterior taper, as in other Triassic turtles, except australochelyids (Palaeochersis talampayensis and Waluchelys cavitesta), where that process has a more consistent width along most of its length (Rougier et al. 1995; Sterli et al. 2007, 2021). Despite a distinct, longitudinal ridge present on its dorsal (visceral) face, it is noticeably shallow (dorsoventrally flattened and wider than tall), which contrasts with the typical relatively tall posterior processes of the entoplastron in other Triassic turtles (Jaekel 1914, 1916; Broin 1984; Gaffney 1990; Sterli et al. 2007, 2021; Szczygielski and Sulej 2016, 2019; Szczygielski and Piechowski 2023). Due to the fragmentary nature of the material, it is uncertain whether the posterior epiplastral process is reduced or altered by compaction.
The hyoplastra are, likewise, represented by UM-PDE2-01, which includes the area around the junction of the paired humeropectoral sulci and a substantial part of the left pectoral scute area, including most of its contribution to the left axillary buttress and UM-PDE2-11, which preserves an anterodorsal portion of the axillary buttress. Based on comparisons with other taxa, the posterior edge of the element preserved in UM-PDE2-01 was likely close to the hyomesoplastral suture and the pectoroabdominal sulcus, but neither of these structures is identifiable in the specimen. The anterior plastral lobe portion preserves no natural edges, so its original width and taper cannot be established. The anterior border of the axillary notch is slightly damaged, obscuring any scute borders, except for the pectoroaxillary sulcus. The axillary buttress meets the main plastral plate at a distinct, but obtuse angle, with both facets separated by a rounded eminence that becomes more prominent anteriorly. In that region, the two facets separated by the eminence seem to approach a right angle. Unfortunately, this area is partially damaged, but it seems more similar to the axillary region of Odontochelys semitestacea and proterochersids rather than the more obtuse and less angular axillary area in proganochelyids or the more anteriorly facing facet in australochelyids (Jaekel 1916; Broin 1984; Gaffney 1985, 1990; Karl and Tichy 2000; Sterli et al. 2007, 2021; Li et al. 2008; Lyson et al. 2014; Szczygielski and Sulej 2016, 2019; Szczygielski et al. 2018, 2024, 2025; Szczygielski and Dróżdż 2026). Only short portions of relatively indistinct sulci are preserved on the external surface of the plastron. The mesial (interhumeral and interpectoral) sulcus is straight and relatively narrow. The humeropectoral sulci meet at a distinctly acute angle anteriorly to the axillary notch. Their arrangement is, therefore, more consistent with most proterochersids (except Keuperotesta limendorsa) than the more posteriorly reaching sulci in Odontochelys semitestacea and Proganochelys quenstedtii. However, the angle between them, at least in their medial parts, is even more acute than in Odontochelys semitestacea and Proganochelys quenstedtii and, thus, quite unique (Gaffney 1985, 1990; Karl and Tichy 2000; Li et al. 2008; Lyson et al. 2014; Szczygielski and Sulej 2016, 2019; Szczygielski et al. 2018, 2024, 2025; Szczygielski and Dróżdż 2026). The preserved anterodorsal part of the axillary buttress seems comparable in size to the one present in the corresponding region in Proterochersis spp. and is, therefore, proportionally larger than in Proganochelys quenstedtii, but smaller than in Palaeochersis talampayensis and Waluchelys cavitesta (compare with Quenstedt (1889); Fraas (1899, 1913); Jaekel (1916); Gaffney (1990); Karl and Tichy (2000); Sterli et al. (2007, 2021); Szczygielski and Sulej (2016); Werneburg et al. (2022); Szczygielski and Piechowski (2023)). An elongate axillary scute seems to have been present.
UM-PDE2-10 (Fig. 10) is an irregular, plate-like fragment with two sulci on its external surface: a wider, less sharply defined one identified here as the mesial sulcus and another that is more perpendicular, narrow, sharp and slightly undulating. The contact point of the sulci is not preserved. The plate is uniformly thin and bears no structures, such as the posterior entoplastral process, viscerally. This combination of features suggests that it comes from the middle part of the plastron and probably represents conjoined mesoplastra or an anterior part of conjoined hypoplastra, which is less likely due to its uniform and low thickness. In that case, the transverse sulcus would be either the interabdominal (if more than one pair of abdominal scutes were present) or the femoroabdominal sulcus. The mesial (interabdominal) sulcus is wider and less sharply defined than in UM-PDE2-01 and UM-PDE2-02, but a less pronounced anteroposterior variation of the mid-line sulcus width and depth is also present in other Triassic turtle plastra (e.g. Szczygielski et al. (2018, 2024, 2025)). For that reason, and because all three specimens are comparable in thickness, we suggest that this difference is better explained by the position of the fragment along the anteroposterior axis and, perhaps, preservational factors than by the presence of another, larger individual in the sample. The lateralmost portion of the right mesoplastron (or mesoplastra) is likely present in UM-PDE2-19, but its external face is apparently shielded by the overlapping ventral portion of the peripherals, whereas the visceral face is irregular and largely uninformative.
UM-PDE2-17 (Fig. 11) is a heavily damaged fragment preserving a mostly concave visceral surface with two portions of the external surface, set at an obtuse angle to each other. It may represent a middle part of the right bridge and, thus, an anterior continuation of UM-PDE2-08 and a ventral continuation of UM-PDE2-19, but damage to the edges prevents a direct, unambiguous articulation of those specimens. The small, presumably dorsal portion of the external face is crossed by a roughly horizonal groove, which may be a margino-inframarginal sulcus. The larger presumed ventral portion bears several grooves, one of which may be a sulcus separating the inframarginal and the abdominal scute areas. The external cortex in the middle part of the specimen is destroyed and the damage reaches the visceral surface in the centre. It is uncertain whether this represents pre-burial trauma or if it is more recent. The fragment most likely corresponds to the mesoplastron (or mesoplastra, if paired), but bone borders are not discernible.
UM-PDE2-08 (Fig. 12) constitutes a partial right hypoplastron from the inguinal area. This identification is somewhat tentative, because the element has a partially eroded external surface, but clearly represents a part of the bridge. The specimen could alternatively represent the left hyoplastron, but such an interpretation would conflict with UM-PDE2-01, the identification of which is unambiguous. Nevertheless, UM-PDE2-08 preserves a short portion of a transverse sulcus, just anterior to the presumed inguinal notch, which is consistent with the expected location of the abdominofemoral sulcus (Gaffney 1985, 1990; Jaekel 1914, 1916; Karl and Tichy 2000; Szczygielski et al. 2024, 2025; Szczygielski and Dróżdż 2026). The specimen, unfortunately, provides no information regarding whether Bartassochersis lopezi gen. et sp. nov. had one or two pairs of abdominal scutes. It reveals, however, that the inguinal buttress was relatively shallow, more reminiscent of Proganochelys quenstedtii and Ypomonetikochelys euryaspis or, at most, Thaichelys ruchae, rather than Cryptochersis paraxene, Palaeochersis talampayensis, Keuperotesta limendorsa and Proterochersis spp. (Fraas 1913; Jaekel 1916; Gaffney 1990; Sterli et al. 2007; Szczygielski and Sulej 2016, 2019; Werneburg et al. 2022; Szczygielski and Piechowski 2023; Szczygielski et al. 2025; Szczygielski and Dróżdż 2026). The posterior surface of the axillary notch is deeply incised by a medially deepening groove. This groove is probably an anterior continuation of the trough separating the scute-covered and visceral surfaces on the dorsal side of the posterior plastral lobe (see below). A weak groove is also present around the same area in some specimens of Proterochersis spp. (e.g. Proterochersis robustaSMNS 18440; Proterochersis porebensisZPAL V. 39/48, ZPAL V. 39/49) and in Keuperotesta limendorsa, but is less distinct (Szczygielski and Sulej 2016, 2019; Szczygielski and Piechowski 2023).
The posterior plastral lobe (Fig. 13) is represented by UM-PDE2-20 (the region of the left femoral bulge sensuSzczygielski and Dróżdż (2026) – that is, the thickening and lateral projection of the femoral scute area around the level of the hypoxiphiplastral suture) and UM-PDE2-02 – the right and the medial part of the right xiphiplastron with attached bases of the right lateral pubic process and ischia and preserved intercaudal scute area and the base of the right caudal process; associated with a more complete natural impression covering more of the left side of the specimen, including the whole left caudal process and with an embedded distal part of the right caudal process. UM-PDE2-20 reveals that the femoral scute had a substantial, spindle-shaped dorsal exposition, with a nearly flat, posterodorsolaterally facing surface. The medial border of that surface is damaged anteriorly, but prominently and sharply raised in the posterior part. Medially, it is accompanied by a trough-like groove separating it from the more medially placed visceral region. Posteriorly, the specimen preserves a groove identified here as the femoroanal sulcus and an impression of the anteromedially directed femoroanal sulcus is traceable on the partially exposed natural mould of the ventral surface. In the anterior part, the anal scute seems to have had a distinct dorsal exposition, as well, but the lateral border of the bone is damaged. More posteriorly in UM-PDE2-02, however, the dorsally exposed area of the anal scute seems to be completely reduced. Instead, the corresponding surface becomes vertical, gently concave and faces posterolaterally. The distinct ridge along the medial edge of the scute area continues along the length of the anal scute area, where it becomes the dorsalmost projection of the scute and, at the same time, the lateral border of the dorsal surface of the posterior plastral lobe. Likewise, the trough separating the scute-covered surface from the visceral surfaces and the attachment site for the pelvis is distinct throughout. Ventrally, the surface of the femoral scute area is mostly flat, whereas the anal scute area is gently concave. The femoroanal sulcus is directed posterolaterally and becomes more sinuous towards its lateral end. The interanal sulcus is very gently sinuous along its length. The anal scute was roughly trapezoidal and longer medially than laterally. The only suture between plastral bones recognisable in the specimen is located between the contralateral xiphiplastra.
As typical for proterochersids and unlike Odontochelys semitestacea, proganochelyids, nd australochelyids (Fraas 1913; Gaffney 1985, 1990; Li et al. 2008; Lyson et al. 2014; Szczygielski and Sulej 2016, 2019; Szczygielski et al. 2018, 2024; Szczygielski and Dróżdż 2026), Bartassochersis lopezi gen. et sp. nov. had a set of postanal scutes and two caudal processes. The caudoanal sulci are sinuous and nearly transverse, with a gentle anterolateral deflection. The caudal processes were triangular, relatively wide relative to their length, directed posterolaterally and apparently covered by a single caudal scute each. Dorsally, the caudal scute region bears a lateral ridge and trough continuing from the anal scute region, a minor dorsal exposition of the caudal scute and the attachment of the ischia medially. There was a distinct and longer than wide pentagonal intercaudal scute that wedged between the anal scutes. Posteriorly, the intercaudal scute area tapers and terminates in a narrow, but distinct transverse sulcus. Posterior to that, there is a posteriorly widening field of bone roughly level with the external surfaces of neighbouring scute areas. This region is not preserved posteriorly in the osteological specimen, but its likely posterior perimeter is revealed by an associated natural imprint, suggesting a trapezoid and posteriorly gently convex outline. The surface of that area is pitted with vascular openings, similar to the surface of the neighbouring scute-covered bone, but unlike the visceral surfaces of the plastron and the pelvis. Based on these features, the structure is best represented as a postintercaudal scute area. A set of at least two postintercaudal scutes was present in Cryptochersis paraxene, but not in any other proterochersid thus far (Szczygielski and Dróżdż 2026). A posterior extension of the ischium beyond the area of the intercaudal scute is present in some specimens of Proterochersis porebensis, but in those cases, the ventral surface of the posterior ischial process is deeper (more dorsal) than that of the scute-covered surfaces and there is no evidence for the presence of postintercaudal scutes (Szczygielski et al. 2018; Szczygielski and Sulej 2019). In contrast to Cryptochersis paraxene (see Szczygielski and Dróżdż (2026)), there is no evidence for postcaudal scutes in Bartassochersis lopezi gen. et sp. nov. It should be noted that, while the shape, size and proportions of the postanal scutes were highly variable in Proterochersis spp., no variability regarding their number has been observed (e.g. Szczygielski et al. (2018, 2024)).
UM-PDE2-09 (Fig. 14) is a plate with no natural edges that has gently concave external and gently convex visceral surfaces and no other notable features. Based on its curvature, it is interpreted as a plastral rather than carapacial fragment, but it is otherwise uninformative.
Indeterminate shell elements. UM-PDE2-15A (Fig. 15) is arguably the most difficult to interpret potentially diagnostic element of the Bartassochersis lopezi gen. et sp. nov. sample. It is a relatively large, subtly bowed fragment with a long, gently curved portion of a natural rim. This natural margin is thickened relative to the neighbouring part of the plate; it increases in thickness towards one of its broken edges and produces a distinct ridge along the body wall attachment that makes it triangular in cross section. One of its three faces is continuous with the external surface of the plate. The second face is narrow and set at a right angle to the first where the rim is at its thinnest, but forms an increasingly acute angle and expands in exposition with greater thickness. It is also crossed by at least three supposed sulci, but these do not visibly continue on to the first face. The third face is visceral and it is separated from the second by a sulcus. Externally, the element bears only one distinct sulcus, which is oblique to the natural border. It is also crossed by a T-shaped layout of extremely subtle and shallow, groove-like concavities, which may either represent extremely inconspicuous sulci, scute growth marks, superficial deformation or some other indistinct surface features. The top bar of the “T” shape is subparallel to the natural edge, approaching it towards its thinner extent, whereas the perpendicular bar converges with the definitive sulcus, but does not meet it within the area of the preserved fragment. The visceral surface bears a rounded ridge, which is subparallel to the ridge for the body wall created by the second and third face of the peripheral thickening and gradually broadens as the latter increases in width and depth. Therefore, both structures (the viscerally located ridge and the body wall ridge) are separated by an elongated fossa. In one corner of the plate, where the periphery is thin with and external oblique sulcus, the visceral surface forms a recessed area with a jagged edge and a lining of smooth, cortical bone. The natural break across the thickest region of the plate is occupied by an irregular cavity. Unlike in UM-PDE2-02, this cavity is not walled with cortical bone, but, instead, exposes bone trabeculae embedded in matrix in the deeper part of the cavity. Thus, it is most likely artificial and unrelated to the ossification process or various peripheral cavities observed in some Triassic turtles (Gaffney 1990; Sterli et al. 2021; Szczygielski and Słowiak 2022; Szczygielski and Dróżdż 2026). The presence of a natural, scute-covered rim and scute sulci indicates that the element belongs to a marginal shell region (i.e. anterior or posterior perimeter of the carapacial aperture or a border of a plastral lobe). However, its precise location remains unclear because none of these four possibilities perfectly matches morphologies observed in other turtle taxa.
The interpretation of UM-PDE2-15A as a part of the posterior plastral lobe is the easiest to reject, based on its incongruence with UM-PDE2-20 and UM-PDE2-02, particularly due to thinness of the main part of the plate, different morphology of the rim, lack of a distinct trough, lack of pelvic attachments and inconsistent size. Likewise, UM-PDE2-15A is most likely not a fragment of the posterior part of the carapace, because it lacks an attachment for the ilium, its visceral surface does not conform geometrically to the morphology of other Triassic turtles, it shows no evidence for the presence of a caudal notch and its single well impressed sulcus would necessitate a lack of lateral symmetry or an extreme elongation of the posterior part of the carapace.
A more compelling interpretation is that the element represents the left edge of the anterior plastral lobe, anterior to UM-PDE2-01. In that case, the oblique sulcus would likely represent the anterolateral part of the humeropectoral sulcus or the gularopectoral sulcus, the visceral depressed area a socket for the entoplastron and the poorly-defined grooves representing either deformation or scute growth marks; in fact, similar irregularities of the external surface are also present in UM-PDE2-01. The main issue with this interpretation is the presence of multiple sulci on the dorsally exposed part of the border. It seems possible that they may delineate areas of small preaxillary scutes, analogous to the postinguinal or periinguinal scutes in some testudinids (e.g. Chroust et al. (2025)). Several small peri-axillary scutes were present in the holotype of Keuperotesta limendorsa (SMNS 17757) along the edge of the axillary buttress. The presence of additional, pre-axillary scutes on the edge of the anterior plastral lobe was, thus far, not documented in Triassic turtles, but the lateral, raised, dorsally scute-covered edge of the anterior plastral lobe is, in general, rarely preserved intact in proterochersids and, thus, poorly documented (Szczygielski and Piechowski 2023; Szczygielski and Dróżdż 2026). There is no evidence for its presence in either proganochelyids or australochelyids (Broin 1984; Jaekel 1916; Gaffney 1990; Sterli et al. 2007, 2021; Szczygielski et al. 2025; Szczygielski and Dróżdż 2026). Nevertheless, the morphology of that structure, especially its large thickness in the presumed posterior part, would likely be autapomorphic for Bartassochersis lopezi gen. et sp. nov.
Another interpretation of UM-PDE2-15A is that it represents the anterior right part of the carapacial border, anterolateral to UM-PDE2-03. In this case, the distinct oblique sulcus on the external surface of the specimen would likely divide the areas of the first vertebral and the first pleural scute and the sulci on the edge would separate marginal scutes, whereas the indistinct linear concavities could represent interpleural and pleuromarginal sulci. This interpretation, however, is not preferred for several reasons: (1) it would imply an anteriorly narrowing shape for the first vertebral scute, which would be very unusual for a proterochersid; (2) while the depth and width of the sulci between different parts of the shell is common, there is no precedence for such a striking difference around a single scute area; (3) the visceral surface of the element does not correspond well with the morphology of the nuchal region in proterochersids or any other Triassic turtle (Jaekel 1916; Gaffney 1990; Sterli et al. 2007, 2021; Joyce et al. 2013; Szczygielski and Sulej 2016; Szczygielski 2017; Szczygielski and Piechowski 2023; Szczygielski and Dróżdż 2026). Although it is also possible that the shallow groove-like structures correspond to scute sulci, delineating a more proterochersid-normal, bell-shaped first vertebral scute and the distinct oblique groove is a pathology, abnormality or a scute growth mark, the latter is morphologically more consistent with sulci, it shows no pathological features and its position and direction do not fit an exaggerated growth mark (compare with Proterochersis robustaSMNS 17930; see Szczygielski et al. (2018)). Obviously, this would also necessitate a highly autapomorphic condition of a marked reduction of anterior marginal scutes and, presumably, the cervical scute. In juvenile individuals of Proterochersis porebensis, the carapacial margin lacks serration and the nuchal embayment, but the scute sulci are distinct and even the anteriormost marginals had a larger ventral exposition (Szczygielski and Sulej 2016; Szczygielski et al. 2018; Szczygielski and Piechowski 2023). Therefore, even though UM-PDE2-15A may possibly represent an anterior border of the carapace, we find that possibility more problematic than its identification as a part of the anterior plastral lobe. Either way, the specimen is morphologically distinct from all other Triassic turtles known thus far.
UM-PDE2-15B (Fig. 16) is a plate fragment with no natural edges and a single sulcus that is partially obscured by sediment residue. Obliquely to the sulcus, the specimen quite rapidly increases in thickness from approximately 5 to 15 millimetres (mm). This may indicate association with the bridge or the body wall attachment, but it is otherwise uninformative.
UM-PDE2-16 (Fig. 16) is an elongated element with no natural edges. It bears a single, straight, longitudinal sulcus with raised edges and otherwise shows a gently concave external surface and a gently convex, smooth and featureless visceral surface. It slightly thickens along the axis of the sulcus. The sulcus may potentially correspond to one of the transverse edges of an abdominal scute (or separate the first and the second abdominal pair, if two were present), but it may also be a deformed carapace fragment with a part of an intervertebral sulcus. Raised edges of the sulci are present in some juvenile specimens of Proterochersis porebensis (see Szczygielski et al. (2018)).
Scapula. UM-PDE2-05 (Fig. 17) is a partial left scapula. The specimen consists almost exclusively of the base and middle portion of the dorsal scapular process with small parts of the glenoid and the acromioscapular lamina. As preserved, the dorsal scapular process is anteroposteriorly compressed and mediolaterally expanded more than in Cryptochersis paraxene, Keuperotesta limendorsa, Proterochersis spp. and the australochelyids, resembling Proganochelys quenstedtii (Jaekel 1914, 1916; Gaffney 1990; Sterli et al. 2007, 2021; Joyce et al. 2013; Szczygielski and Sulej 2016; Szczygielski and Piechowski 2023). The acromioscapular lamina seems orientated at an angle similar to that of Proterochersis porebensis, suggesting a medially rather than anteriorly directed acromion (Szczygielski and Sulej 2016; Szczygielski and Piechowski 2023), but it is incomplete and might have been subjected to deformation. The glenoid region terminates in an uneven and rather jagged structure instead of a well-defined glenoid facet. Its surface is mostly obscured by a thick layer of glue, locally revealing bony trabeculae and a thin layer of probable calcified cartilage. It seems possible, but not certain, that it was not yet completely ossified at the moment of death. Note, however, that both Proganochelys quenstedtii and Proterochersis porebensis exhibit rather well defined and smooth glenoid facets even in juvenile individuals (Gaffney 1990; Szczygielski and Piechowski 2023).
Pelvis. UM-PDE2-02 (Fig. 13) preserves bases of the right lateral pubic process and ischia. The outline of the attachment of the lateral pubic process is comma-shaped. In UM-PDE2-02, the widest portion of that attachment bears a deep, ovoid pit devoid of trabecular bone. It is unknown if that structure reflects the imperfect ossification of the pubis, resulting in retention of a void in the place of a cartilaginous core of the lateral pubic process or if it is an unusual form of damage or weathering. The ischia attach to the plastron across a pentagonal area. Mesially, the connection appears to be completely fused and the elements cannot be differentiated even along the break surface, but, laterally, it retains a sutural character. The dorsal surface of the ischiadic plate is nearly parallel to the plastron and recessed posteromesially, with a gentle sagittal ridge in the central part and it ascends laterally and anteriorly. Its posterior limit is not preserved. Overall, the attachment between the pelvis and the plastron is very similar as in Keuperotesta limendorsa and Proterochersis spp. (Fraas 1913; Joyce et al. 2013; Szczygielski and Sulej 2016, 2019; Szczygielski and Piechowski 2023).
Ungual. UM-PDE2-06 (Fig. 18) is an isolated ungual. It is generally consistent morphologically with unguals of other Triassic turtles (Gaffney 1990; Sterli et al. 2007; Szczygielski and Piechowski 2023). Its proximal articular surface is concave and subtly bipartite, the proximodorsal and proximoventral processes are moderately sized and the flexor tubercle is relatively flat. The specimen is relatively straight, which is more consistent with its identification as a manual rather than pedal ungual (Gaffney 1990; Sterli et al. 2007). It is also somewhat asymmetrical, which is most notable in the subdivision of its dorsal surface into two variably slanted faces. This feature may potentially be related to its position within the autopodium or side or may be a deformation or a pathology, because we did not observe asymmetry of that kind in the autopodia of Palaeochersis talampayensis or Proganochelys quenstedtii.