Cretaceous amniotes from Angola: dinosaurs, pterosaurs, mosasaurs, plesiosaurs, and turtles

Mateus, O., Polcyn M. J., Jacobs L. L., Araújo R., Schulp A. S., Marinheiro J., Pereira B., & Vineyard D. (2012).  Cretaceous amniotes from Angola: dinosaurs, pterosaurs, mosasaurs, plesiosaurs, and turtles. V Jornadas Internacionales sobre Paleontología de Dinosaurios y su Entorno. 71-105., Salas de los Infantes, Burgos


Although rich in Cretaceous vertebrate fossils, prior to 2005 the amniote fossil record of Angola was poorly known. Two horizons and localities have yielded the majority of the vertebrate fossils collected thus far; the Turonian Itombe Formation of Iembe in Bengo Province and the Maastrichtian Mocuio Formation of Bentiaba in Namibe Province. Amniotes of the Mesozoic of Angola are currently restricted to the Cretaceous and include eucryptodire turtles, plesiosaurs, mosasaurs, pterosaurs, and dinosaurs. Recent collecting efforts have greatly expanded our knowledge of the amniote fauna of Angola and most of the taxa reported here were unknown prior to 2005.


Angola has the best Southern Hemisphere record of Late Cretaceous marine amniotes, including mosasaurs, plesiosaurs and chelonians (all refs). It has also produced pterosaurs and dinosaurs (refs) preserved in marine sediments. This faunal richness is particularly relevant in the context of the mid-Cretaceous opening of the South Atlantic (Jacobs et al., 2006a, 2006b, 2009a, 2009b) and the general paucity of records for that region of the Earth (Jacobs et al., 2011).

The main sedimentary packages of the Angolan Mesozoic can be broadly grouped into three divisions: (1) Karoo-like continental basin deposits (with Triassic fishes), at Baixa de Cassange; (2) Cretaceous marine sediments related to the opening and expansion of the South Atlantic (and rich in marine vertebrates); and (3) continental Cretaceous, with no vertebrate record so far, filling rift valley remnants along the coast and broadly distributed in the interior of the country. The latter are poorly known, in part because of the inaccessibility and rareness of outcrops.

The goal of this article is to provide an overview of the fossil amniotes from the Cretaceous of Angola. Additionally, we provide a list of non-amniote taxa reported from Angola (Appendix).

History of Mesozoic amniote paleontology in Angola

According to Nunes (1991: 317), the first reports of fossils in Angola are attributed to Lang, who in 1839 (cit. in Nunes, 1991, without bibligraphic reference) wrote “…petrifications of Ammon horns, that show the existence of Mesozoic terrains in the country.” The German explorer Eduard Peschuel-Loesche (German, 1849-1913) prospected that part of Africa between 1873 and 1876, including what are now the provinces of Cabinda and Namibe. O. Lenz (1877) wrote notes on fossils from the Cenozoic of Cabinda, including a crocodile tooth collected by Pechuel-Loesche. To our knowledge this is the first report of a fossil amniote from Angola.

Edmond Dartevelle and Edgard Casier worked on the rich fish fauna from Bentiaba, and also reported a “partial tooth of dinosaur, and vertebrae and bone fragments of reptiles” (Dartevelle and Casier, 1941: p.106; and 1943: p.16). The Dartevelle collection is housed in the Royal Museum of Central Africa in Tervuren, Belgium. Their purported dinosaur tooth (RG 2084; Fig. 1) was found to be the proximal portion of a plesiosaur dorsal rib Mateus et al. (2011)  

Soares Gaspar de Carvalho published a major work on the geology of Namibe desert in which mosasaur teeth attributed to Mosasaurus beaugei are first reported and figuredfrom Bentiaba and other localities in Namibe (Carvalho, 1961: p.85, 89, 92, fig. 128).

Mascarenhas Neto of the Angola Mines and Geology Survey (Neto, 1960) reported reptiles from the locality of Iembe later published by Antunes (1961). Miguel Telles Antunes published the first comprehensive work on the Mesozoic-Cenozoic vertebrates of Angola (Antunes, 1964). He has continued to work on the fossil fishes, publishing a report on the Cretaceous selachians of Angola (Antunes and Cappetta, 2002).  Antunes (1964) named the mosasaurs Angolasaurus bocagei (see also Jacobs et al )and Mosasaurus iembensis, later transferred to Tylosaurus iembensis (Lingham-Soliar ref…). He also reported other mosasaurs in Cabinda, Ambrizete, Barra do Dande, Benguela-Cuio, and eight localities in the Namibe basin.

In addition to mosasaurs, Antunes (1964, 1970) also reported vertebrae and girdle elements of two plesiosaur specimens from the Turonian of Iembe, isolated plesiosaur teeth and vertebrae from the Maastrichtian of Cambota, Cabinda, Barra do Dande, Ambrizete, and Bentiaba, and teeth from Fazenda dos Cavaleiros (Bero River)  Neto (1964: 221) reports the occurrence of reptile teeth from the Maastrichtian of Bentiaba (former São Nicolau) and later, Cooper (1972) reported Mosasaurus beaugei from that locality.

South of Sumbe (formerly Novo Redondo), Lapão (1972) reported the presence of one tooth and vertebrae from Maastrichtian sediments, attributing them to Mosasaurus and Plesiosaurus respectively. Referral of Maastrichtian material to Plesiosaurus is doubtful since that is restricted to the Lower Jurassic (Groβmann 2007).  Lapão 1972) also reported the occurrence of a mosasaur skull but it is now lost or was never collected.  
Contributions on other aspects of the history of the geology and paleontology are provided by Andrade and Andrade (1957), Antunes (1964, 1970), Antunes et al. (1990), Nunes (1991) and Brandão (2010).

Angola's war of independence in the early 1970s, and subsequent civil war persisting until 2002, halted field research during that period. After reaching a peace agreement between the warring factions in 2002, fieldwork became feasible once more, and thus the initiation of the current research cycle. All current paleontological work is conducted under the auspices of the PaleoAngola Project, a scientific collaboraton between researchers from Universidade Agostinho Neto (Angola), Southern Methodist Universty (USA)  Universidade Nova de Lisboa (Portugal), and the Natuurhistorisch Museum Maastricht (Netherlands). The current research cycle commenced in May of 2005, and has continued with expeditions in May and July 2006, July of 2007, August of 2009, February of 2010,  July of 2010, and is planned for July 2011. All coastal provinces have been visited with the exception of Zaire province in the extreme northwest (Fig. 2).

Field work in Angola has been extremely productive. The first visit resulted in the discovery of a new genus and species of sea turtle, Angolachelys mbaxi (Mateus et al. 2009), the dinosaur Angolatitan adamastor  (Mateus et al. 2011), the mosasaur Prognathodon kianda (Schulp et al., 2008), and new specimens of the mosasaurs Angolasaurus bocagei, and Tylosaurus iembensis  (Antunes 1964). Later expeditions allowed excavation of new specimens including additional forms never reported from Angola.Publication by Jacobs et al. (2006a;2006b, 2009a, 2009b, 2010a, 2010b), (Polcyn et al., 2007a, 2007b, 2007c, 2009, 2010), Schulp et al., ( 2006a, 2006b, 2008), Mateus et al. (2006, 2008, 2009, 2011), and Araújo et al. (2010) are beginning to document the richness of the Late Cretaceous amniote fauna of Angola. The Projecto PaleoAngola set of contributions represents an increase up to five high-rank clades of Mesozoic tetrapods. Prior to 2005 only the genera Angolasaurus, Mosasaurus, Tylosaurus, Globidens, and an indeterminate plesiosaur material were known. Here we report 21 different taxa. The fossils are being prepared in the laboratories of the participating institutions: ML, SMU, and NHMM (see in acronyms), where also replicas of the most important specimens will be made prior to the return of the material to University Agostinho Neto, in Luanda.

ML - Museu da Lourinhã, Portugal.
SMU - Southern Methodist University, Dallas, USA.
NHMM - Natuurhistorisch Museum Maastricht, Maastricht, The Nethertlands.
MGUAN-PA: Museu de Geologia da Universidade Agostinho Neto, Luanda, Angola (PaleoAngola Collection).
MRAC:  Musée Royal d'Afrique Central, Tervuren, Belgium.

Projecto PaleoAngola has worked mostly in coastal marine Cretaceous rocks, obtaining abundant fishes (both Chondrichthyes and Osteichthyes, see Antunes and Cappetta, 2002), mosasaurs, plesiosaurs, and marine turtles as well as terrestrial animals including isolated bones of pterosaurs and dinosaurs and the articulated forelimb of a sauropod dinosaur (Mateus et al., 2011). No mammals, amphibians, or birds have been dicovered thus far. Of the localities explored by the PaleoAngola Project (Fig. 2), two deserve special mention due to their richness: Iembe, in the province of Bengo and Bentiaba, in the province of Namibe. Being situated in the Southern Hemisphere, which has a relatively poor Late Cretacous marine vertebrate record when compared to Northern Europe, North America, and Morocco (Bardet et al., 2010), add to the importance of these localities.

Iembe (Turonian)
The locality of Iembe (Fig. 2) is late Turonian in age and yielded fishes, mosasaurs, plesiosaurs and remains of a dinosaur (Fig. 3). The most productive formation in the Iembe area is the Itombe Formation (see Mateus et al., 2011: fig 1 for the Mesozoic formations in the Cuanza Basin). This is the type locality of the mosasaurs Angolasaurus bocagei Antunes, 1964, Tylosaurus iembeenses Antunes 1964, and the turtle Angolachelys mbaxi Mateus et al., 2009. It also yielded a forelimb of a sauropod dinosaur that represents the first non-avian dinosaur discovered in Angola, Angolatitan adamastor (Mateus et al., 2011).

Bentiaba (Campanian-Maastrichtian)
The main vertebrate bearing layers at the locality of Bentiaba (Fig. 4) are Late Campanian and Maastrichtian in age, although older rocks are present. Most of the specimens have been collected from the mid Maastrichtian. Bentiaba is one of the most important localities for marine vertebrate fossils by virtue of the: (i) high concentration and abundance, (ii) excellent preservation, (iii) completeness and (iv) taxonomic diversity (see faunal list below). Numerous fish species (Dartevelle, 1942), at least twelve taxa of mosasaurs, three taxa of plesiosaurs, marine turtles, dinosaurs and pterosaurs are present. Hundreds of specimens have been located or collected including partial or complete skeletons and skulls, in an exposure covering less than two square kilometers. Thus far, more than 200 marine reptile specimens were and are being unearthed from a two-meter thick bonebed. Bentiaba is the type locality of the mosasaur Prognathodon kianda Schulp et al., 2008 and has provided the most complete Globidens phosphaticus skeletons to date (Polcyn et al. 2010). Work in progress shows the occurrence of new taxa of plesiosaurs, turtles, and mosasaurs. Also several isolated bone attributed to dinosaurs were found at this locality.

The presence of such a large number of top predators in the Late Cretaceous fossil beds of Angola suggest the high primary productivity of the Benguela upwelling system (Shannon 1985) extends back to the Cretaceous (Jacobs et al. 2009a,b).



Although the picture is changing quickly, chelonians are poorly known in the Cretaceous of Angola. The first taxon identified to species level was the recently described and named Angolachelys mbaxi (Mateus et al. 2009) from the Turonian of Bengo Province. To date, all turtle remains collected in the Mesozoic of Angola are Cretaceous eucryptodirans, including angolachelonians (Mateus et al., 2009), chelonioids, and a possible dermochelyid, which would be the oldest record for that clade.

Eucryptodira Gaffney, 1975

Angolachelonia Mateus et al., 2009

Angolachelys mbaxi Mateus et al. 2009

Material:MGUAN-PA2, nearly complete skull (fig. 5), dentary, fragments of vertebrae, carapace, one ungual phalanx.
Locality and horizon: North of Iembe (Bengo Province), Turonian.
Comments: Angolachelonians have mandibular articulation aligned with or posterior to the occiput, and basisphenoid not visible or visibility greatly reduced in ventral view.
Basal eucryptodires and angolachelonians originated in the northern hemisphere (Mateus et al., 2009), thus Angolachelys represents one of the first marine amniote lineages to have invaded the South Atlantic after separation of Africa and South America, as shown by the phylogenetic affinities of angolachelonians (Mateus et al, 2009) with marine eucryptodires from Europe and from Glen Rose in Texas (Mateus et al., 2009; Vineyard, 2009; Vineyard et al., 2009).

Chelonidae Bonaparte 1832

Euclastes Cope 1867

Euclastes sp.

Material: MGUAN-PA14, an incomplete anterior portion of a skull (fig. 6) and MGUAN-PA157, a new complete skull and mandible, cervical vertebrae, peripheral plates, forelimb (collected in 2010, still mostly unprepared, fig. 7).
Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian, Mucuio Formation.
Comments: Vineyard et al. (2009) reported the presence of an Euclastes-like turtle from Bentiaba based on specimen MGUAN-PA14. Additional material collected at Bentiaba since the work of Vineyard et al. (2009) improves the quality of the sample of Maastrichtian angolachelonians and will allow evaluation of diversity and relationships among marine eucryptodires.

With respect to MGUAN-PA14, the ventral surface of the palate is flat, smooth and vascularized with a low marginal rim (fig. 6). In ventral view, the premaxillae are well developed paired bones that form a large contact with the vomer and the maxilla laterally. Sutures are present that form a contact with the palatines, although part of the left palatine is missing. Orbits are anterodorsally positioned on the skull. The vomer is large and roughly rectangular in shape. The dorsal surface floors the fossa nasalis, and foramina praepalatina are present. In anterior view, the premaxillae are large, and form the ventral margin of the external narial opening. The maxilla forms the broad and posteriorly inclined margin of the nares. In lateral view the maxilla is large and forms the anteroventral margin of the fossa orbitalis. In posterior view, the vomer pillar is strong, forming the septum between the nasal passages. The posterior extent of the pillar is in line with the posterior margin of the horizontal portion of the vomer. Also in posterior view, a foramen presumed to be the foramen supramaxillare, is present on the maxilla.

Parham (2005) placed Euclastes among the Pancheloniidae (however, see also Jalil et al., 2009 for alternate interpretation).  Parham’s (2005) diagnosis of the genus, based on Euclastes wielandi is: (1) V-shaped basisphenoid; (2) secondary palate; (3) closely positioned foramina for the exits of the anterior carotids; and (4) rod-shaped rostrum basisphenoidale, but noted a high variability within the group. Euclastes meridionalis has a more extensive secondary palate, Euclastes planimenta has a wider robust head, Euclastes platyops has a shallow tomial ridge, dorsally directed orbits, and non-concave triturating surface; and Euclastes roundsi has a less developed secondary palate.  The complete skull and jaw, incomplete postcrania and carapace collected during the 2010 expedition will test and refine the attribution of these specimens to Euclastes.

Chelonioidea Baur, 1893

Protostegidae Cope, 1873

Protostega sp. 

Material: MGUAN-PA158, two humeri, three costal plates, one xiphiplastron (fig 8), and one peripheral plate (identified as the first or second peripheral) of a single animal.
Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.

Comments: The PaleoAngola expedition of July 2010 to the Maastrichtian of Namibe Province collected bones of a very large turtle. More bones are still in the field for future excavation. The turtle is identified as a protostegid cryptodire based on its very large size (humerus length is 63 cm), fenestrated carapace, finger-like terminations and curvature of the xiphiplastron, and long and thin costal plates (e.g. Zangerl, 1953, but see Hooks, 1998, for the systematics of Protostegidae).

The long costal plate (98cm long) found in Bentiaba is virtually identical to the holotype of Protostega dixie (CNHM P27314) figured by Zangerl (1953: fig. 39). The xiphiplastron is long, with an acute curve of the xiphiplastron body, forming a distinct L-shape of the lateral rim, rather than the smooth posterolateral curve for most chelonioids. The anterior end is pointed, while the posterior end is digitiform. These features are also similar to those observed in the xiphiplastron of Protostega dixie (see Zangler, 1953: fig. 45 and 46). The humeri show abundant bite marks, probably caused by sharks. The articular ends of the humerus are broad. The lateral process is moderately projected, not beyond the shaft edge in ventral view.

Protostegidae Cope, 1873


Material: MGUAN-PA 167, neural scute (fig. 9).
Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.

Comments: An isolated bone (fig. 9) found in Bentiaba is here identified as a neural scute, possibly from a Calcarichelys-like protostegid, which would indicate the presence of a second protostegid taxon at Bentiaba.

The base has a saddle-like shape, with the conical dorsal face culminating in an acute thorn-like structure. The bone resembles the ankylosaur osteoderms but does not have the typical keel forming a sharp point, and the edges are thin. The surface texture is smooth, contrary to the typically rugose osteoderms of thyreophorans.
Calcarichelys gemma Zangler 1953 is known from the Upper Cretaceous of Alabama (Hooks, 1998). Hooks (1998) diagnosed Calcarichelys by the mid-dorsal keel composed of alternating, laterally compressed conical and saddle-shaped elements. The Angolan specimen is known from the above described scute, which is not enough for a detailed attribution. It differs from Calcarichelys because the Angolan scute is totally conical without an elongate midline keel. The protostegid Chelosphargis advena (Hay, 1908) also has sharp scutes (see Hay, 1908: fig 257), but they are more keel-like and not as prominent as Calcarichelys.

Toxochelyidae Baur, 1896

Toxochelys sp.

Material: MGUAN-PA168, Hyoplastron, peripherals and costals plates (fig. 10).
Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.

Comments: Contrary to most cheloinoids, the toxochelyids have rectangular peripheral plates without digitiform terminations, with a socket-like cavity for the rib, and often fenestrated carapace. Nicholls (1988) lists two characters as diagnostic of the genus Toxochelys that are visible in the Angolan material: the presence of costoperipheral fontanelles in carapace and the peripheral border smooth, without notches or serrations. The hyoplastron outline is virtually identical to that of Toxochelys moorevellensis (see Zangler, 1980: fig.3c). We tentatively assign this Angolan form to Toxochelys sp.


The expeditions of 2007 and 2010 yielded significant plesiosaur material. Several partial skeletons, including cranial material were collected in 2010. Although some plesiosaur vertebrae have been recovered from the Turonian of Iembe, the most productive locality for plesiosaurs is Bentiaba (fig. 11)

Sauropterygia Owen, 1861

Plesiosauria de Blainville, 1835

Plesiosauroidea Welles, 1943

Elasmosauridae Cope, 1869

cf. Tuarangisaurus Wiffen and Moisley, 1986

Material: MGUAN-PA85, MGUAN-PA106, MGUAN-PA120, propodials, cervical and pectoral vertebrae and pectoral and pelvic girdle elements.
Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.
Comments: Referral to this taxon is based on possession of unfaceted distal ends of the propodials, short, straight and robust propodials, presence of low longitudinal ridges along the distal borders of the propodials, and short and narrow ischia. The Tuarangisaurus-like specimens are important to better understand the phylogenetic status and the ontogeny of this genus, previously reported from Argentina and New Zealand (Gasparini et al. 2003, Wiffen and Moisley, 1986).

   Elasmosauridae indet.

Material: MGUAN-PA113, basicranium, MGUAN-PA126, symphyseal region of mandible, various unnumbered specimens, teeth.

Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.
Comments: The recovered Elasmosauridae mandible portion is ascribed on the basis of straight dentary rami, short symphysis and presence of four symphyseal teeth. Several isolated teeth have been found ascribable to Elasmosauridae on the basis of the recurvature of the tooth crown, thin enamel, fine striae that do not anastomose, and some teeth slightly labiolingually flattened.

Antunes (1964) describes fragmentary plesiosaur material coming from coastal basins except Benguela Basin. The most complete plesiosaur remains, referred to “Cimoliasauridae”, comprise twelve vertebrae and a tooth from Cambota in the Cabinda Enclave. As the “Cimoliasauridae” family is not supported in recent phylogenetic hypotheses (O’Keefe and Street 2009, Ketchum and Benson 2009) the above mentioned material should be referred to Elasmosauridae indet., following Vincent et al 2010).


With the exception of two Turonian taxa described by Antunes (1964), the mosasaur record of Angola was limited to isolated tooth crowns. Since 2005, a collection of relatively complete and well preserved material of several taxa from the late Turonian through the Maastrichtian deposits, now represents the largest collection of Southern Hemisphere mosasaurs. It also provides new information on poorly known taxa and is still producing numerous new species. Mosasaurs are the most common amniote found in all Late Cretaceous localities visited in Angola, with dozens of specimens collected to date and many more still in situ. They are also the most taxonomically diverseamniote in the Upper Cretaceous of Angola, with at least ten species identified to date.

Squamata Oppel, 1811

Mosasauridae Gervais, 1853

Mosasaurinae Gervais, 1853

Plotosaurini Russell 1967

Mosasaurus sp. aff. M. hoffmanni Mantell, 1829

Material:  MGUAN-PA35, Isolated shed teeth, complete and partial humeri.
Locality and horizon: Baba, Bentiaba, Fazenda Dos Cavaleiros (Namibe Province); middle to upper Maastrichtian.
Comments: The teeth we have referred to Mosasaurus are strongly D-shaped in cross section and medially recurved, with minor faceting of the otherwise smooth enamel surface. Previous discoveries of teeth of this genus from Angola have been referred to Mosasaurus beaugei Arambourg 1952 (Carvalho 1961, Antunes 1964), known from the Maastrichtian of Morocco; however, the isolated teeth reported here are indistinguishable from specimens of Mosasaurus hoffmanni from northern Europe and Mosasaurus maximus from North America. The European and North American forms likely represent the same species (Russell, 1967; Mulder, 1999). Fernandez et al. (2008) reported a contemporaneous Mosasaurus from the upper Maastrichtian of northern Patagonia, referring it to Mosasaurus sp. aff. M. hoffmanni.

Globidensini Russell, 1967

Prognathodon kianda Schulp et al. 2008

Material: MGUAN-PA129 (holotype), MGUAN-PA128, MGUAN-PA149, MGUAN-PA150, MGUAN-PA151. Several specimens of different ontogenetic stages. Shed teeth, isolated bones, partial skulls and skeletons.
Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.
Comments: Prognathodon kianda (Fig. 12 A & B, D) is the most abundant taxon at the Bentiaba locality and previous reports of Liodon (Antunes, 1964; Jacobs et al., 2006a) from Bentiaba are referred to that taxon. A new specimen collected in 2010 consists of a largely articulated skeleton and skull, but lacks the portion of the tail posterior to the second caudal vertebra.

Prognathodon saturator Dortangs et al. 2002

Material: MGUAN-PA 169.Fragmentary dentary including tooth crown.
Locality and horizon: “Bentiaba 2” locality (Namibe Province); middle Maastrichtian.

Comments: The specimen is composed of badly weathered fragments of the right dentary (fig. 12 C). Collectively the fragments preserve three relatively complete alveoli and a single tooth crown. The broad tooth base is elevated above the dorsal margin of the dentary. The tooth crown is extremely robust and ~60 mm in basiapical length. Albeit fragmentary and despite that the only tooth crown lacks the enamel, the specimen preserves some diagnostic characters. The bicarinate tooth is slightly recurved posteriorly, and possesses no medial curvature. It is swollen above the constricted region at its base. This combination of characters allows referral to the genus Prognathodon (Bell, 1997; Schulp, 2006). The only Late Maastrichtian species of Prognathodon possessing the robustness and proportions seen in the new specimen is P. saturator Dortangs et al., 2002, allowing tentative referral to that taxon. This is the first report of this taxon from the South Atlantic region. Antunes (1964, plate 26, fig 4) reported a tooth from the Maastrichtian of Cabinda that he referred to Mosasauridae indet. That tooth may represent an early ontogentic stage of P. saturator.

Globidens phosphaticus Bardet and Pereda Suberbiola, 2005b

Material: MGUAN-PA23, MGUAN-PA24, partial skulls, vertebrae and limb material.
Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian, Mocuio Formation.

Comments: Globidens phosphaticus was originally described and named on the basis of isolated tooth crowns from Morocco by Bardet et al. (2005b). Bardet (2005b) also indicated the presence of that taxon in Angola, based on illustration of an isolated tooth crown in Antunes (1964). Polcyn et al. (2010) reported the first skeletal material of this taxon, providing a preliminary description and confirming the taxonomic validity of the species. Referral of the Angolan materal to G. phosphaticus was on the basis of the tooth characters given by Bardet et al. (2005b) That study also presented character data establishing Globidens phosphaticus as the sister-taxon to the late Campanian G. schurmanni from North America.

Plioplatecarpinae Russell, 1967

Angolasaurus bocagei Antunes, 1964

Material: MGUAN-PA 001, Partial skull, MGUAN-PA 063, MGUAN-PA 065,articulated skull, partial postcrania, and  nearly complete forelimbs.

Locality and horizon: Iembe (Bengo Province); upper Turonian.

Comments: The Turonian section at Iembe has yielded new material of Angolasaurus bocagei (fig. 13) allowing an updated phylogenetic analysis and confirming its taxonomic validity. The phylogenetic analysis supports Angolasaurus as the sister-taxon of Selmasaurus, and along with Ectenosaurus, form a clade that apparently diverged from the Platecarpus-Plioplatecarpus lineage in the Turonian (Polcyn and Everhart, 2008; Polcyn et al. 2009).

“Platecarpus” ptychodon Arambourg, 1952

Material: MGUAN-PA 160  Two partial skulls and skeletons including limb material, isolated tooth crowns.
Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.

Comments: This species was erected by Arambourg (1952) on the basis of isolated tooth crowns from the Maastrichtain phosphates of Morocco. Antunes (1964) reported isolated tooth crowns (Antunes, 1964; plate XXVI figures 11 and 11a) that he referred to the same taxon. The new specimens collected in 2010 represent most of the skull elements, the presacral vertebral column, and the pectoral girdle and forelimb. The new material does not support referral to the genus Platecarpus, or any named genus of mosasaur.

Tylosaurinae Williston, 1895

Tylosaurus iembeensis Antunes 1964

Material: MGUAN-PA64, fragmentary skull elements.
Locality and horizon: Iembe (Bengo Province); upper Turonian.

Comments: Tylosaurus iembeensis remains poorly known and the holotype was lost in a fire in Lisbon (Jacobs et al, 2006); however, a fragmentary new specimen has been recovered. The preserved quadrate has a poorly developed infrastapedial process, similar to that seen in T. kansasensis, suggesting a relatively basal divergence within Tylosaurinae. However, T. iembeensis is significantly larger than T. kansasensis, approaching the size of T. nepaeolicus and T. proriger.

Halisaurinae Bardet et al., 2005a

Halisaurus sp.

Material:MGUAN-PA18, Two partial skeletons including many elements of the skulls, vertebrae, ribs, and limb material,MGUAN-PA 83, fragmentary skull found with a single vertebrae.

Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.
Comments: These specimens (Fig. 14 A and B) will be described and named elsewhere; however, preliminary phylogenetic analysis supports its referral as a new species, closely related to Halisaurus arambourgi and H. platyspondylus (Polcyn et al, 2007a).

Phosphorosaurus sp.

Material: MGUAN-PA52, isolated partial frontal.
Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.

Comments This isolated partial frontal (fig. 14 C and D ) allows description and comparison with the type material of Phosphorosaurus ortliebi from the Maastrichtian of Belgium (IRSNB R34). The frontal is damaged, missing a small portion of the posterior border, the posterolateral edges, and a significant portion anteriorly. The lateral margins, dorsal and ventral surfaces are largely intact. MGUAN-PA52 belongs to a relatively small animal, the frontal measuring 35 mm wide at its widest point anterior to the orbits, and 61 mm long as preserved, thus it belongs to an individual that was approximately 1.5 to 2 meters long.
The dorsal surface bears a strong, tall median ridge along the entire length of the frontal, terminating posteriorly in a prominent triangular boss. There is a significant supraorbital constriction, the preserved lateral margins are strongly convex antorbitally. The dorsal surface is striated, with strong longitudinal ridges near the midline, and more laterally, weaker ridges occur, radiating anterolaterally to meet the lateral margins. Ventrally, the frontal has a broadly open olfactory canal, the descending processes are prominent, forming blunt parallel ridges anteriorly and grading to finer narrow ridges interorbitally, where they trend toward but do not meet the midline, then diverge as low blunt ridges posteriorly. Two sulci are present on this posterior portion, corresponding to the structures labeled by Russell (1967, his text-figure 4) as accommodating the cerebral hemispheres. The articulation for the postorbitofrontal is missing due to breakage. The prefrontal articulation is only weakly developed antorbitally, slightly incising the lateral surface of the descending processes and forming a simple, fibrous lap-joint with ventral surface of the frontal but with no corresponding excavation.

Phosphorosaurus ortliebi was described and named by Dollo (1889) and subsequently redescribed and referred to the genus Halisaurus by Lingham-Soliar (1996). Some characters do unite Phosphorosaurus with Halisaurus, including the configuration of the quadrate but significant differences also exist, warranting retention of Dollo’s (1889) genus (Polcyn et al., in press). These include the strong median ridge along the entire length of the dorsal surface, terminated posteriorly by a triangular boss and the pineal foramen location on the frontoparietal suture. A weak posteriorly placed triangular boss is visible in H. platyspondylus and H. arambourgi, but in those taxa, the median ridge is restricted to the anterior part of the frontal and the pineal foramen rests within the parietal table (Holmes and Sues, 2000; Bardet et al., 2005a).

Halisaurinae sp.

Material: MGUAN-PA 070 Cervical vertebra, two dorsal vertebrae and four caudal vertebrae (fig. 14 E – I).

Locality and horizon: Iembe (Bengo Province); ?Santonian.

Comments: The four caudal vertebrae were found together and two are still articulated. The remaining vertebrae were found as isolated specimens. The single cervical vertebra (fig. 14 I), is an axis and has an oblique intervertebral articulation, a broadly oval condyle, and relatively large hypapophysis. The dorsal vertebra (fig. 14 H) bears an oblique intervertebral articulation, has a weak constriction anterior to condyle, its condyle is wider than high, and the synapophyses originate anteriorly. The caudals (fig. 14 E, F and G) have a roughly symmetrical hexagonal condyle, relatively large haemal arches, a ventral sulcus between haemal arch bases, and large prezygopophyses. Although no cranial material has been collected, the combination of characters present in the vertebral elements supports the referral to the Halisaurinae (Polcyn et al., 2009). This specimen is diminutive and judging from the size of individual centra, would have been no more than about 1.5 meters in length. It represents the oldest halisaurine outside North America and presuming the Santonian date as reliable, is approximately as old as the holotype of Eonatator sternbergi (See Bardet et al 2005a) from the Santonian of Kansas.


So far, pterosaur remains have only been recorded from the Maastrichtian of Bentiaba. Several bones were collected, mainly incomplete and isolated. Here we present the preliminary data on the first report for pterosaurs in Angola.

Pterosauria Kaup 1834

Pterodactyloidea Plieninger 1901

Ornithocheiroidea Seeley 1876

Material: Left femur (MGUN-PA163; fig. 15).  

Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.

Comments: Attribution of this specimen to the Ornithocheiroidea is due the femoral head with distinct stout neck and steeply directed caput (see Unwin, 2003: character 38). However, this referral is tentative. 


Dinosaurs were unknown in Angola until the publication of Angolatitan adamastor Mateus et al., 2011.  Here we report two additional bones from Bentiaba (Mocuio Formation, middle part of the Maastrichtian) belonging to distinct individuals. Material includes the distal end of a propodial element and a phalanx.

Dinosauria Owen, 1842

Saurischia Seeley, 1888

Sauropoda Marsh, 1878

Eusauropoda Upchurch, 1995

Neosauropoda Wilson and Sereno, 1994

Somphospondyli Wilson and Sereno 1998      

Angolatitan adamastor Mateus et al. 2011

Material: MGUAN-PA3, The only material known is the forelimb and scapular girdle, including the scapula, humerus, ulna, radius, and metacarpals I, III, and IV.

Locality and horizon: North of Iembe (Bengo Province); Turonian.
Comments: A forelimb of the sauropod dinosaur (Angolatitan adamastor Mateus et al. 2011 (fig. 16) from the late Turonian of Iembe, represents the first dinosaur discovery in Angola, and is one of the few occurrences of sauropod dinosaurs in sub-Saharian Africa with a reliable geochronological dating. The marginal marine sediments yielding the specimen are reported to be late Turonian in age (see more on the geology of Iembe in Antunes, 1964, Jacobs et al., 2006), thus it is a non-titanosaurian sauropod in sub-Saharian Africa at a time supposed to be dominated by titanosaurians. Moreover, Angolatitan adamastor is the only basal somphospondyl known in the late Cretaceous, which implies the existence of relict forms in Africa. 
Angolatitan is more derived than Giraffatitan but less derived than Euhelopus, which is notable given its relatively late appearance in the sauropod fossil record. Its life is thought to have been an arid setting.

Ornithopoda Marsh, 1881

?Iguanodontia Baur 1891

Material: MGUAN-PA 176, pedal phalanx II-2 (fig. 17).

Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.
Comments: A single isolated phalanx from Bentiaba is here interpreted as the left pedal phalanx II-2 of an ornithopod, because its the proximodistally short compared to width (being wider than long), it has two assymetrical distal condyles and assymetrical proximal pits, and trapezoidal outline in distal view.
Is cannot be a theropod because of the lack of pneumatic cavity, the shallow colateral pits and greater width and length; and a possible hadrosauroid because the  (measurements: 19 mm wide, 17 mm long, 11mm high). The ventro-lateral and ventro-medial margins have a expanded lip, which is seen in the pedal phalanx II-2 of hadrosaurs but unsual in other dinosaurs. Prieto-Márquez and Wagner (2009: character 295) consider that subsquared proportions of pedal phalanx II2, i.e. only slightly shorter proximodistally than it is wide mediolaterally, is an apomorphy common in non-lambeosaurine hadrosauroids.
The Hadrosauroidea clade is not known in the fossil record of Africa, so if this interpretation is correct, this phalanx represents the possible first Hadrosauroidea in that continent.  In any event, it is the first record of dinosaur from Bentiaba.

Dinosauria indet.

Material: MGUAN-PA 175, Distal part of ?humerus.

Locality and horizon: Bentiaba (Namibe Province); middle Maastrichtian.

Comments: The specimen is the distal end of a large propodial, likely a humerus. Collected in 2010, the specimen is unprepared and identification is preliminary. Assignment to Dinosauria is based on its large size, well formed distal condyles, and long shaft. The very thick medullary region excludes the possibility it belongs to a theropod dinosaur.

Associated fauna (Chondrichthyes, Osteichthyes, and invertebrates)

Although the associated fauna (Chondrichthyes, Osteichthyes and invertebrates) is outside the focus of this article, the Mesozoic of Angola has proven to be very productive of some clades, namely on the Chondrichthyes and marine molluscs. 
In the Appendix 1, we provide the compilation of Mesozoic fossil animal species (excluding amniotes), with the systematic, chronostratigraphic and geographic data, and bibliographic source, compiled using the available scientific publications, including geological maps explanations and recent PhD theses (for example, Tavares, 2006). We recognize that much of the taxonomy might require revision, so we have excluded reports that seem doubtful and pre-1960 citations of genera not existing in the Paleobiology Database ( That decision reduces the list of taxa, but guarantees that most taxonomy is updated. This should be seen as a historical list of species reported to Angola.

To date, over 700 animal species have been reported from the Cretaceous of Angola. Over 80% of those taxa is represented by molluscs, mainly ammonites, which represent around 70% of the total mollusca species . Ammonites alone provide more than half of the taxa found and were as such an important component of the Angolan Late Cretaceous marine ecosystem.

More than fifty species of Chondrichthyes have been reported in the literature (Antunes and Cappetta, 2002) and represent the most diverse vertebrates in the study area. Other important groups include the Osteichthyes, with 18 species, and the Echinodermata with 60 taxa.

Summary taxa checklist

Checklist of the Mesozoic amniotes of Angola (see Appendix 1 for invertebrates and non-amniote vertebrates list):

Angolachelys mbaxi Mateus et al. 2009 (Turonian)
?Euclastes sp. (middle part of the Maastrichtian)
Protostega sp. (middle Maastrichtian)
Protostegidae indet aff. Calcarichelys (middle part of the Maastrichtian)


Toxochelys sp. (middle Maastrichtian)
cf.Tourangisaurus (middle part of the Maastrichtian)
Elasmosauridae indet. (middle part of the Maastrichtian)




Mosasaurus sp. aff. hoffmanni Mantell 1829 (middle part of the Maastrichtian-late Maastrichtian)


Globidens phosphaticus Bardet and Pereda Suberbiola, 2005b (middle part of the Maastrichtian)
Prognathodon kianda Schulp et al. 2008 (Maastrichtian)
Prognathodon cf. saturator Dortangs et al. 2002 (Maastrichtian)


Angolasaurus bocagei Antunes, 1964 (Turonian)

“Platecarpus” ptychodon Arambourg, 1952 (Maastrichtian)


Halisaurus sp. (Maastrichtian)

Phosphorosaurus sp. (Maastrichtian)

Halisaurinae indet. (Santonian)


Tylosaurus iembeensis Antunes 1964 (Turonian)

Ornithocheiroidea indet. (middle part of the Maastrichtian)
Angolatitan adamastor Mateus et al. 2011 (Turonian)
Dinosauria indet. (middle part of the Maastrichtian)