Allosauridae Marsh, 1878 sensu Sereno 1998

Malafaia, Elisabete, Mocho, Pedro, Escaso, Fernando, Narvaéz, Ivan & Ortega, Francisco, 2024, Taxonomic and stratigraphic update of the material historically attributed to Megalosaurus from Portugal, Acta Palaeontologica Polonica 69 (2), pp. 127-171 : 161-164

Allosauridae Marsh, 1878 sensu Sereno 1998 Genus Allosaurus Marsh, 1877b

Type species: Allosaurus fragilis Marsh, 1877 , from the Kimmeridgian-Tithonian Brushy Basin member of the Morrison Formation at Garden Park, Colorado, USA .

Allosaurus sp.

Fig. 25 , SOM 5: fig. S4.

Previous identifications: MG 8775 is associated with a label with the attribution to Megalosaurus pombali probably corresponding to the specimen from Ribamar mentioned (but not figured) by Lapparent and Zbyszewski (1957) in the description of the species ( Lapparent and Zbyszewski 1957: 25). MG 4817 is associated with a label with the identification Megalosaurus insignis from Paimogo, but it is also part of a set with another tooth fragment collected in Atalaia. None of the teeth attributed to this taxon by Lapparent and Zbyszewski (1957) were referred as having been collected at Paimogo so it is possible that this specimen came from Atalaia. MNHN/UL.EPt.019 was described and figured by Sauvage (1897 –1898: pl. 5: 3), who referred it to M. insignis . MG 8781 is labeled as pertaining to Megalosaurus , but we have not found any reference to this specimen in previous publications. Finally, MNHN/UL.EPt.004 does not have any information about the identification and we could not find any reference to it on previous publications. Any information about the specimen could also be found in the database of the museum, probably because it is part of the collections that were partially destroyed in the fire of 1978.

Material.—Five tooth crowns (MG 8781, SOM 5: fig. S4B; MG 8775, Fig. 25B ; MG 4817, Fig. 25A ; MNHN/UL.EPt.019, Fig. 25C ; MNHN/UL.EPt.004, SOM 5: fig. S4A). MG 8775 is from Outeiro do Seixo, near Ribamar, in the municipality of Lourinhã ( Portugal), probably coming from upper Kimmeridgian–lowermost Tithonian levels of the Praia da Amoreira-Porto Novo Formation ( Fig. 2 ) ( Manuppella et al. 1999a). MG 4817 is associated with a label saying that it came from Paimogo, but it is also part of a set with another tooth fragment with the location Atalaia, both from the municipality of Lourinhã ( Portugal). In both cases, the specimen was probably collected in sedimentary deposits of the Sobral Formation ( Fig. 2 ) which is KimmeridgianTithonian in age ( Manuppella et al. 1999a). MG 8781 is labeled as coming from Pedras Muitas, near Baleal, in the municipality of Peniche ( Portugal) ( Fig. 2 ). This area is near the contact between the Alcobaça and Bombarral formations. However, the vertebrate fossil site of Pedras Muitas is marked in the Bombarral Formation on the geological map ( Camarate França et al. 1960). If MG 8781 came from the same locality, it would have been collected from sedimentary rocks of the Bombarral Formation, which is Tithonian in age ( Manuppella et al. 1999a). MNHN/UL.EPt.019 came from Pombal ( Portugal), but there is no more specific information about where it was found. Finally, MNHN/UL.EPt.004 is part of the collections of MUHNAC and is associated with a label saying that it was probably collected from Lower Cretaceous levels at the locality of Cabo Espichel (Sesimbra, Portugal). However, its preservation is very different from other material from these levels, and it is a dental morphotype that has not yet been identified in the Lower Cretaceous of the Lusitanian Basin. It is more likely that this specimen came from an unknown Late Jurassic locality.

Description.—MG 4817 ( Fig. 25A ) and MG 8775 ( Fig. 25B ) are relatively well-preserved tooth crowns, which are triangular in lateral view and quite elongated ( CHR = 2.25 for MG 4817 and 2.13 for MG 8775). The basal section is subcircular, with a CBR close to one in MG 8775. The crown is slightly recurved, with the mesial surface somewhat convex, the distal surface is straight, and the apex is placed at the level of the mesial carina. In MG 8775 the lingual surface is weakly concave in mesial view, whereas it is apically convex in MG 4817 ( Fig. 25A 4 , B 2 ). The labial surface is strongly convex, whereas the lingual surface has a convex centrally placed surface, but it is mostly flat adjacent to the distal margin and has a well-marked longitudinal concavity adjacent to the mesial carina. This longitudinal concavity is bounded mesially by the carina that projects labially (particularly in MG 4817). The enamel has a relatively well-marked irregular ornamentation and subtle transverse undulations are visible on both lingual and labial surfaces. In MG 8775, a well-developed spalled surface is present on the apex and extends on both lingual and mesial surfaces. The mesial and distal carinae are serrated and the denticles extend to the cervix in the distal carina and at least to the basal part of the mesial carina (the mesial surface is incomplete to the base). The distal carina is straight and placed on the distal surface, whereas the mesial carina is strongly twisted, being placed on the labial surface near the apex, but strongly twists to the base, being placed almost entirely in the lingual surface in the basal half of the crown. In MG 4817 there are 11 denticles per 5 mm in the apical and central sectors of the mesial carina and approximately 10 denticles per 5 mm in the basal sector of the mesial and distal carinae. In MG 8775 there are 10.5 and 12 denticles per 5 mm in the central and basal sectors of the mesial carina and 10, 10.5, and 11 denticles per 5 mm in the apical, central, and basal sectors of the distal carina, respectively (see SOM 2). The denticles of both mesial and distal carinae of MG 4817 are subquadrangular, with mostly straight to slightly convex external margins, whereas in MG 8775 they are rectangular, slightly longer mesiodistally than the apicobasal width. In MG 4817 there are neither interdenticular sulci between the denticles nor marginal and transverse undulations visible on the lingual and labial surfaces of the crown. However, in MG 8775 well-marked interdenticular sulci curving towards the base of the crown are present adjacent to the distal denticles in the lingual surface, but not between the mesial denticles.

MNHN/UL.EPt.04 is an almost complete crown, missing the apex and a fragment of the lingual surface (SOM 5: fig. S4A 1 –A 8). It has a similar morphology to the previously described specimens but is slightly shorter apicobasally ( CHR = 1.58). The labial surface is strongly convex (at least the preserved part, all the surface adjacent to the distal margin is missing) and the lingual surface has a longitudinal convexity that is centrally positioned and delimited distally by a well-marked concave area adjacent to the distal carina. Due to this morphology of the lingual surface the base of the crown most likely had a J-shaped cross-section. The distal carina is centrally positioned on the distal surface of the crown in its apical portion, but strongly twists into the lingual surface to the base so it mostly faces lingually. The distal denticles are poorly preserved and only the morphology of those from the basal part of the crown can be described. They are rectangular with asymmetrically convex external margins (slightly pointing apically). The mesial denticles are subquadrangular, becoming more rectangular (larger mesiodistally than apicobasally) to the base of the crown. They have mostly symmetrically convex external margins, except those from the basal sector, which have somewhat apically pointed external convexities. No marginal or transverse undulations are visible, but the enamel has a well-marked ornamentation consisting of an irregular (non-oriented) texture that is more visible in the lingual surface adjacent to the distal carina. Finally, MG 8781 corresponds to a fragment of the apical part of a tooth crown (SOM 5: fig. S4B 1 –B 5). It is poorly preserved but has a similar morphology to the previously described specimens, including the presence of a convex longitudinal surface in the central part of the lingual surface. The specimens MG 4817, 8775, 8781 and MNHN/ UL.EPt.04 are interpreted as mesial tooth crowns based on the poor labiolingual compression, the strongly twisted mesial carina with a well-marked longitudinal groove, and the presence of a convex longitudinal surface in the central part of the lingual surface (see Hendrickx et al. 2020a).

MNHN/UL.EPt.019 is an almost complete and well-preserved crown (only the apical end is missing), interpreted here as a transitional tooth ( Fig. 25C ). The crown is relatively low (CHR = 2.14) and slightly compressed labiolingually (CBR = 0.61). In lateral view, it is triangular and slightly recurved distally, but the distal margin is mostly straight, and the mesial margin is slightly convex. The labial surface is flat for most of its width and convex adjacent to the mesial margin. The lingual surface is slightly convex but is weakly concave transversely because it projects somewhat lingually. The crown base section is oval to subcircular. The mesial and distal carinae are denticulated and both extend to the base of the crown (the mesial carina is somewhat incomplete, and the exact extension cannot be determined). The distal carina is centrally positioned in the distal margin and is mostly straight in distal view, but projects slightly lingually to the base of the crown. The mesial carina is centrally positioned on the mesial margin in the apical end of the crown but progressively twists to the lingual surface to the base. There is a narrow longitudinal sulcus on the lingual surface adjacent to the mesial carina, which projects for most of the height of the crown but becomes shallower apically. Transverse undulations are well marked in both lingual and labial surfaces, being slightly more evident lingually in the central part of the crown. A shallow longitudinal concavity is visible in the lingual surface adjacent to the distal carina, but it is restricted to the basal end of the crown ( Fig. 25C 1 ). The enamel has a subtle and irregular (non-oriented) ornamentation. There are 17.5 and 12 denticles per 5 mm in the central section of the mesial and distal carinae, respectively (see SOM 2). The distal denticles are mostly quadrangular (mesiodistally as long as the apicobasal width), with symmetrically convex external margins ( Fig. 25C 6 , C 7 ). They are separated by narrow interdenticular sulci that project only slightly into the surface of the crown. The denticles decrease in size to the base of the crown and the external margin becomes asymmetrically convex (pointing apically). The mesial denticles are rectangular (slightly larger apicobasally than mesiodistally). The external margin is also symmetrically convex, and the denticles become smaller towards the base of the crown ( Fig. 25C 8 , C 9 ). The mesial denticles are separated by narrow interdenticular sulci, which do not extend on the surface of the crown.

Remarks.—The discriminant analysis classifies two specimens (MG 8775, 4817) as belonging to Allosaurus, MNHN / UL.EPt.019 as Genyodectes and MNHN/UL.EPt.004 as Dromaeosaurus (see Table 1). The plot shows similar results for the distribution of these specimens, but three of the specimens fall within the morphospace of Allosaurus and only MNHN/UL.EPt.004 is placed with Dromaeosaurus Fig. 16 ). The cladistic analysis of the dentition-based data matrix with constraints and including all studied morphotypes recovered the specimens here grouped in Morphotype 3 at the base of the Allosauroidea clade in polytomy with Sinraptor and Erectopus ( Fig. 18A ). The results of the analysis pruning a priory all the morphotypes from the Lusitanian Basin but Morphotype 3 found 2 MPTs (CI = 0.204, RI = 0.464, L = 1316). The consensus tree from these MPTs (CI = 0.203, RI = 0.463, L = 1318) shows a better resolved Allosauroidea clade and similarly to the previous analysis, Morphotype 3 is allied to Sinraptor and Erectopus Fig. 21C ). However, this result should be interpreted with caution due to the similarity on both discrete and quantitative features of the dentition of Sinrapto r and Allosaurus that show almost indistinguishable mesial and lateral tooth morphology ( Hendrickx et al. 2020a). The synapomorphies found in the analysis supporting the position of Morphotype 3 within a group with Sinraptor and Erectopus include the straight distal margin of the crown in lateral view and the presence of short interdenticular sulci, which are features also present in Allosaurus (see Hendrickx et al. 2020a, b).

There is no evidence of flutes in the lingual surface, which is a feature present in the mesial teeth of some ceratosaurians, such as Masiakasaurus and Ceratosaurus Madsen and Welles 2000 ; Carrano et al. 2012; Soto and Perea 2008; Rauhut 2011). The crowns are poorly compressed labiolingually (CBR close to 1, in MG 8775, or higher than 0.6), with both mesial and distal carinae extending to the cervix. The mesial carina strongly twists lingually to the base of the crown and the lingual surface has a longitudinal centrally positioned convexity, resulting in a characteristic J-shaped cross-sectional outline. This combination of features is exclusive to the teeth from the mesial dentition of Allosaurus and Sinraptor ( Hendrickx et al. 2015b, 2020a). Based on this combination of features, together with their paleogeographic and stratigraphic distribution, the specimens of Morphotype 3 are here interpreted as a mesial teeth of Allosaurus . In three of the described specimens (MG 8775, 4817 and MNHN/UL.EPt.019), the mesial carina is strongly twisted lingually to the base of the crown (particularly marked in MG 4817), defining a longitudinal groove in the lingual surface adjacent to the carina that extends for most of the height of the crown. On the other hand, the mesial carina of MNHN/UL.EPt.019 is much less twisted lingually and the crown is more compressed labiolingually, with an oval cross-section of the base. These features suggest that this specimen is a transitional tooth from a position between the most mesial and lateral teeth, possibly a first or second maxillary tooth).

Figures

Fig. 25. Mesial tooth crowns attributed to allosauroid theropod Allosaurus sp. from different Upper Jurassic localities of Portugal. A. MG 4817,?Atalaia Lourinhã region). B. MG 8775, Ribamar (Lourinhã region). C. MNHN/UL.EPt.019, unknown locality (Pombal region). Lingual (A1, B1, C1), labial (A2, B4, C5), distal (A3, B3, C2), and mesial (A4, B2, C4), views; basal cross-section (B5, C3); detail of the distal denticles from the central and basal sector of the crown respectively (C6, C7); detail of the mesial denticles from the central and basal sector of the crown respectively (C8, C9).

Fig.2. Geological and stratigraphical context of fossil sites in the Consolação and Bombarral sub-basins, Portugal. A. Geological map of coastal region of Lourinhã and Peniche with localities yielding fossil remains studied in this work (modified from Camarate França et al. 1960; Manuppella et al. 1999a). B. Map of the Lusitanian Basin with the main structural elements and the different sub-basins (sensu Fürsich et al. 2021). C. Simplified stratigraphy of the Upper Jurassic and Lower Cretaceous sequences in the Torres Vedras and Lourinhã costal region and Consolação-S.Martinho do Porto coastal region based on Hill (1988), Manuppella et al. (1999a) and Azerêdo et al. (2010).Abbreviations:Assen., Assenta; Chron., Chronostratigraphy; Fm., Formation; mb., member; l., lower; m., middle; Oxfor., Oxfordian; sb., sub-basin; u., upper.

Fig. 4. Geological and stratigraphical context of fossil sites in the Mamede sub-basin. A. Geological map of the region of Batalha and Ourém with localities yielding fossil remains studied in this work (modified from Teixeira et al. 1968; Manuppella et al. 2000). B. Map of the Lusitanian Basin with the main structural elements and the different sub-basins (sensu Fürsich et al. 2021). C. Simplified stratigraphy of the Upper Jurassic sequences in the Mamede sub-basin based on Kullberg et al. (2013) and Fürsich et al. (2021).Abbreviations: Chron., Chronostratigraphy; Fm., Formation; mb., member; l ., lower; m., middle; sb., sub-basin; u., upper.

Fig. 1. Geological map (adapted from Oliveira et al. 1992) showing the Mesozoic levels and the main structures that delimit the different sectors (sensu Ribeiro et al. 1996 and Kullberg et al. 2006) and sub-basins (sensu Fürsich et al. 2021) of the Lusitanian Basin, Portugal.

Fig. 6. Geological and stratigraphical context of fossil sites in the Setúbal sub-basin. A. Map of the Lusitanian Basin with the main structural elements and the different sub-basins (sensu Fürsich et al. 2021). B. Geological map of the Sesimbra to Cabo Espichel region with localities yielding fossil remains studied in this work (modified from Manuppella et al. 1999b). C. Simplified stratigraphy of the Lower Cretaceous sequences in the Setúbal sub-basin based on Manuppella et al. (1999b). The stratigraphic position of the Papo Seco Formation from which the studied fossils were collected is marked in green. Abbreviations: Chron., Chronostratigraphy; Fm., Formation; mb., member; l., lower; m., middle; sb., sub-basin; u., upper.

Fig. 7. Geological and stratigraphical context of the Figeira de Foz and Aveiro region. A. Map of the Lusitanian Basin with the main structural elements and the different sub-basins (sensu Fürsich et al. 2021). B. Geological map of Figueira da Foz and Aveiro region (Aveiro sub-basin) with localities yielding fossil remains studied in this work (modified from Rocha et al. 1981 and Barbosa et al. 2008). C. Simplified stratigraphy of the Upper Cretaceous sequences in the Figueira da Foz and Aveiro sub-basins based on Barbosa et al. (2008).Abbreviations: Chron., Chronostratigraphy; Fm., Formation; mb., member; l., lower; m., middle; sb., sub-basin; u., upper.

Fig. 8. Cervical vertebrae attributed to a thalattosuchian crocodyliforms from the Kimmeridgian Alcobaça Formation, Fervença (Alcobaça region, Portugal). A. MG 4829a. B. MG 4829b. Right lateral (A1, B1), left lateral (A2, B2), anterior (A3, B3), dorsal (anterior to the right, A4, B4), ventral (anterior to the left, A5, B5), and posterior (A6, B6) views.

Fig. 9. Cervical (MG 8798a) and dorsal (MG 8798b) vertebrae attributed to indeterminate stegosaurian dinosaurs from the Kimmeridgian–Tothonian Sobral Formation, Porto das Barcas (Lourinhã region, Portugal). A. MG 8798a. B. MG 8798b. Right lateral (A1, B1), left lateral (A2, B2), anterior (A3, B3), dorsal (anterior to the left, A4, B4), ventral (anterior to the right, A5, B5), and posterior (A6, B6) views.

Fig. 16. Graphical results of the Linear Discriminant Analysis (LDA) of 581 tooth crowns belonging to twenty-two theropod taxa and the isolated specimens from the Lusitanian Basin studied in this work. Eigenvalue of axis 1 = 8.2989 (which accounts for 67.28% of the total variation) and eigenvalue of axis 2 = 1.129 (which accounts for 9.152% of the total variation). AL, apical length; CBW, crown base width; CH, crown height; DA, distoapical denticle density; DB, distobasal denticle density; DC, distocentral denticle density; MA, mesioapical denticle density; MB, mesiobasal denticle density; MC, mesiocentral denticle density; MCL, mid-crown length; MCW, mid-crown width; MDE, mesiobasal denticles extent.

Fig. 18. Strict consensus trees resulting from a cladistic analysis performed on a dentition-based data matrix and forcing the constrains defined by Hendrickx et al. (2020a) with the five tooth morphotypes from the Lusitanian Basin (A), simplified strict consensus tree resulting from the cladistic analysis pruning a priori all morphotypes from the Lusitanian Basin but Morphotype 1 (B), and simplified strict consensus tree resulting from the analysis pruning a priori all morphotypes from the Lusitanian Basin but MNHN/UL.EPt.023 (C).

Fig. 21. Simplified strict consensus trees resulting from a cladistic analysis performed on a dentition-based data matrix and forcing the constrains defined by Hendrickx et al. (2020a) pruning a priori all morphotypes from the Lusitanian Basin but Morphotype 2 (A), simplified strict consensus tree resulting from the analysis pruning a priori all morphotypes from the Lusitanian Basin but MG 15 (B), and simplified strict consensus tree resulting from the analysis pruning a priori all morphotypes from the Lusitanian Basin but Morphotype 3 (C).