Glaucidium ireneae sp. n.; Corvus bragai sp. n. (1 Viewer)

[Fred Ruhe]

Marco Pavia, 2020

Palaeoenvironmental reconstruction of the Cradle of Humankind during the Plio-Pleistocene transition, inferred from the analysis of fossil birds from Member 2 of the hominin-bearing site of Kromdraai (Gauteng, South Africa)

Quaternary Science Reviews. 248: Article 106532. doi:10.1016/j.quascirev.2020.106532


Highlights: https://www.sciencedirect.com/science/article/abs/pii/S0277379120304947?via=ihub

The fossil bird association of Kromdraai Member 2 is described in detail.

A great avian diversity (25 taxa, 2 new for the science) has been detected.

The high numbers of rock-dweller birds reinforce the presence of a cliff.

The presented results modify the palaeoenvironmental reconstruction of Kromdraai.

The importance of fossil birds as palaeoenvironmental proxies is confirmed.

Abstract

The results of the analysis of the bird assemblage from Member 2 of the Paranthropus robustus type-locality of Kromdraai are presented here, and the data are used to infer the environmental context of the Plio-Pleistocene transition in the Cradle of Humankind. The study of more than 800 remains revealed the presence of 25 bird taxa, 2 of them described as new to science. For some of the taxa, Kromdraai represents the oldest fossil record so far, if not the first ever. This relatively diverse bird association comprises terrestrial taxa dominated by rock-dwelling birds, such as Geronticus thackerayi and Tyto cf. alba, followed by open grassland birds, such as Francolinus sp. and Turnix sp., freshwater birds, ducks and waders, and forest birds, such as Accipiter melanoleucos and Treron sp.. The presence of these taxa, and their percentages in terms of the number of bones and individuals, point to the presence of a cliff or important rocky outcrop very close to the fossil site with the presence of a waterbody and a gallery forest. The grassland birds and their numbers confirm the presence of open grassland or savannah, which was already indicated by the analysis of large mammals. These results shed new light on the environment of the Cradle of Humankind during the Plio-Pleistocene transition and confirm the validity of the birds as palaeoenvironmental proxies.

Marco Pavia describes two new species: Glaucidium ireneae sp. n. and Corvus bragai sp. n.

Enjoy,

Fred

 

[Fred Ruhe]

Systematic paleontology

Among the 805 bird bones analyzed here, 512 were identified to family or species level (Supplementary 1). They represent a minimum number of 87 birds, with a relative high diversity, since 25 taxa have been recognised with two species described as new to science (Table 1). Here is presented the description of the detected taxa and their occurrence in the fossil record, African and worldwide.

Phasianidae Horsfield, 1821
Francolinus sp. (sensu Louchart, 2011)
Material. 54 bones (Table 1; Fig. 2AeC, H; Supplementary 1).
Remarks. The mid-sized galliform remains found at Kromdraai
can be referred to Phasianidae, rather than Numididae for the characters described by Louchart (2011) and Pavia (in Fourvel et al., 2016). In particular, they can be referred to Francolinus, an essentially African genus which comprises four subgenera (Francolinus, Peliperdix, Scleroptila, and Pternistis, the latter three endemic to Africa). Each of the four subgenera includes a variable number of species, from five (Peliperdix) to 24 (Pternistis) (Del Hoyo and Collar, 2014). From the osteological point of view, the morphology of the different species is rather homogeneous with also a great individual variation and strong sexual dimorphism. Following Crowe (1993), the various species can be separated with osteometrics, but complete skeletons and complete bones are needed. On the other hand, currently only two to three species live in real syntopy, and when this happens, they are in contrasting sizes, which makes their separation slightly less complicated. The Kromdraai material is rather fragmented and does not belong to single individuals collected in anatomical connection, even partially, thus it is not possible to evaluate the relative proportions of the various skeletal parts. The Kromdraai sample is not enough to perform any statistical analysis, like the one proposed by Louchart (2011) for the francolins from Laetoli (Tanzania) on the size of the distal humeri. A more detailed analysis of the francolin remains from the Cradle of Humankind as a whole could help clarify their relationships. For these reasons, the taxonomic attribution of the Kromdraai’s francolins are left here as Francolinus sp. (sensu Louchart, 2011). The francolins are a common presence in the African fossil record since the Pliocene, where they are rarely abundant (Louchart, 2011). Phasianidae referred to Francolinus sp. have been reported in Africa since the Miocene with a distal tibiotarsus from the Early Miocene of Rusinga Island, Uganda (Harrison, 1980) and a fragmentary tarsometatarsus from the Late Miocene of Middle Awash, Ethiopia (Louchart et al., 2008).

Coturnix cf. coromandelica (J.F. Gmelin, 1789)
Material. 10 bones (Table 1; Fig. 2I, K; Supplementary 1).
Remarks. These bones show all the typical features of Phasianidae, and their small size refers them to Coturnix, the smallest African Phasianidae. The fossil from Kromdraai are comparable in size with Coturnix coromandelica, to which they are referred with uncertainty, as the presence of an extinct, undescribed species cannot be ruled out. The presence of other bones from Kromdraai (Pockock, 1969) and of very similar remains from the nearly coeval site of Cooper’s Cave (Pavia et al., in prep.) suggests a broader analysis of the Coturnix bones from the Cradle of Humankind is required to clarify their relationships. Despite quails being quite common in Palaearctic fossil record (Tyrberg, 1998), the African fossil record of Coturnix is surprisingly limited to the Early Pliocene of Langebaanweg, South Africa

 

[Fred Ruhe]

Anatidae Vigors, 1825
Tadorna sp.
Material. 18 bones (Table 1; Fig. 2D, E, M; Supplementary 1).
Remarks. Mid-sized Anatidae are comparable in size with Tadorna cana, definitively smaller than Plectropterus gambensis, and bigger than any species of Anas. The smaller size and some morphological details excludes the slightly bigger Alopochen egyptiacus (sic) and refer the bones to Tadorna: 1) very deep and very well-defined sulcus m. supracoracoidei of the proximal coracoid; 2) rounder condylus dorsalis ulnaris and more developed proximad in the distal ulna; 3) condylus medialis of the distal tibiotarsus more developed medially than in Alopochen egyptiacus (sic). The fragmented fossil remains include some differences with the recent Tadorna cana, such as the wider incisura sternocoracoidei, the processus acrocoracoidei less developed in the coracoid, and the incisura intercondylaris deeper in the distal tibiotarsus, suggesting leaving the fossil as Tadorna sp. Tadorna tadorna and Tadorna sp. have been reported from the Plio-Pleistocene locality of Ahl al Oughlam in Morocco (Mourer-Chauviré and Geraads, 2010), and Harrison (1980b) reported Tadorna cf. cana from the Early Pleistocene of Olduvai, Tanzania. These data, together with the Kromdraai record, represent the oldest African records of this taxon ecologically related to lentic water bodies or, more rarely, to brackish or shallow marine waters.

Alopochen cf. aegyptiacus (Linnaeus 1766)
Material. 1 bone (Table 1; Fig. 2F; Supplementary 1).
Remarks. The fossil belongs to a large sized Anatidae, which is smaller than Plectropterus gambensis. It can be separated from Tadorna for its bigger size and for its morphology, with the crista trochanterica less developed than in Tadorna. It is similar in size with recent Alopochen aegyptiacus, even if slightly smaller, with the crista trochanterica not as developed as the recent one and an evident sulcus on the cranial side of the facies articularis antitrochanterica. The presence of a single femur of this taxon, even well preserved, and very similar but not identical to Alopochen aegyptiacus suggests leaving the specific attribution dubitative. The Kromdraai femur represents the first fossil record of Alopochen in South Africa. In fact, Alopochen cf. aegyptiacus has only been reported from the nearly coeval site of Ahl al Oughlam in Morocco (Mourer-Chauvire and Geraads, 2010) and from the Early Pleistocene of Olduvai, Tanzania (Prassack, 2014).

Anas sp.
Material. 9 bones (Table 1; Fig. 2G; Supplementary 1).
Remarks. All the specimens belong to a small-sized species of Anatidae, clearly smaller than the others found at Kromdraai. The specimens show the morphology of the genus Anas and can be separated from the similar-in-size Dendrocygna for the following characters: 1) shallower sulcus m. supracoracoidei and less developed facies articularis clavicularis in the proximal coracoid; 2) flat corpus scapulae not as rounded as in Dendrocygna; 3) a wide femur head and rounded and not protruding mediad; 4) relatively shallow sulcus intercotylaris and smooth distal outlines of the distal condyles in the distal femur; 5) distal epiphisis of the tibiotarsus tilted mediad with thin condylus medialis, and wide incisura intercondylaris; 6) proximal tarsometatarsus with a small cotyla lateralis with a well-marked small accessorial facet on the dorsal side of the bone, just distally of the cotyla lateralis, and medial margin of the cotyla medialis not as sharp as Dendrocygna. The osteological characteristics of the various bones excludes the very similar genus Spatula, but the preserved material is not enough to improve the taxonomic attribution. The Kromdraai record represents the oldest African fossils of the genus Anas, as the only other evidence is from the Early Pleistocene of Olduvai (Harrison, 1980).

 

[Fred Ruhe]

Columbidae Illiger, 1811
Treron sp.
Material. 21 bones (Table 1; Fig. 2OeP; Supplementary 1).
Remarks. The Columbidae remains from Kromdraai more likely belong to a single species of middle size, smaller than any recent African Columba but similar to some Streptopelia species. In addition to the morphological characters listed by Louchart (2011), the morphology of the Kromdraai bones more recalls that of Treron rather than the other Columbidae in: 1) different shape of angulus medialis and deep pneumatized fossa in the medial side of the impression m. sternocoracoidei of the distal coracoid; 2) small and pointed tuberculum ventrale and well-marked tuberculumdorsale in the proximal humerus; 3) pointed and well evident processus supracondylaris dorsalis and a tuberculum between the fossa m. brachialis and the processus supracondylaris dorsalis in the distal humerus; 4) pointed olecranon and narrow cotyla dorsalis not projecting distally in the proximal ulna; 6) rounded ventral outline of condylus dorsalis ulnaris and
very pointed tuberculum carpale in the distal ulna; 7) papillae remigalis caudalis very prominent; 8) crista trochanteris very low and fossa trochanteris virtually absent in the proximal femur; 9) distal end of the tibiotarsus medio-laterally wide and narrow in the distal view with condylus medialis proximo-distally low and projected mediad; 10) crista medialis hypotarsi very protruding plantarly and fossa infracondylaris dorsalis well marked in the proximal tarsometatarsus; 11) three distal trochleae almost on the same line with the trochlea metatarsi III only slightly more projected distally than the others. The comparative material available shows a certain homogeneity among the African species of Treron, thus the Kromdraai bones are here referred to Treron sp. The fossils from Kromdraai represent the oldest occurrence of this taxon worldwide and the fourth record worldwide, after the findings of Okinawa, Japan (Matsuoka, 2000), Ankilitelo, Madagascar (Goodman et al., 2013), and Jerimalai, Timor Leste (Mejier et al., 2019) [sic, must be Meijer et all, 2019], all dated to the Late Pleistocene/Holocene.

Gruidae Vigors, 1825
Anthropoides cf. paradiseus (Lichtenstein, 1793)
Material. 1 bone (Table 1; Fig. 2N; Supplementary 1).
Remarks. This tarsometatarsus has already been described in Fourvel et al. (2016). It has been referred to Anthropoides rather than Balearica for different morphological characters: 1) wider
eminentia intercotylaris and not protruding dorsally; 2) concave lateral side of the cotyla lateralis; 3) foramen vasculare proximale mediale opened inside a vascular groove; 4) hypotarsus extended more distally than the tuberositas musculi tibialis cranialis. In respect to the recent Anthropoides paradiseus, the Kromdraai specimen shows the eminentia intercotylaris narrower and more pointed, thus its attribution is left dubitative. This specimen represents the oldest record of Anthropoides so far and, if confirmed, the first fossil record of A. paradiseus.

Otididae Rafinesque, 1815
Ardeotis cf. kori (Burchell, 1822)
Material. 1 bone (Table 1; Fig. 2Q; Supplementary 1).
Remarks. This wing phalanx is morphologically comparable with the Otididae and comparable in size and morphology with the males of Ardeotis kori, the biggest African species, to which the specimen is tentatively referred. This bone represents the oldest occurrence of Ardeotis and the second record of Ardeotis kori so far, after the radius found in the late Early Pleistocene of Buia, Eritrea (Delfino et al., 2018).

Threskiornithidae Richmond, 1817
Geronticus thackerayi Pavia, 2019
Material. 231 bones (Table 1, Fig. 2SeT; Supplementary 1).
Remarks. Geronticus thackerayi is an extinct species of bald ibis, recently described from Kromdraai and Bolt’s Farm in the Cradle of Humankind. For the detailed description and comparison, see Pavia (2019). Geronticus thackerayi is the most common bird species at Kromdraai during the Plio/Pleistocene transition where there was a breeding colony, as testified by high percentage (32.6%) of its remains belonging to juvenile/immature birds or not-fledged individuals (Pavia, 2019). This fact has a great palaeoenvironmental
significance, as all the Geronticus species are obligate cliff or rock nesters (Matheu and del Hoyo, 1992).

 

[Fred Ruhe]

Charadriidae Vigors, 1825
Charadriidae indet.
Material. 1 bone (Table 1; Fig. 2R; Supplementary 1).
Remarks. This bone can be referred to Charadriidae among the various Charadriiformes on the basis of the general shape of the humerus and, in particular, of the processus supracondylaris dorsalis, also following Ballmann (2004). The humerus is comparable in size with various species of both Vanellus and Pluvialis, but the lack of the proximal end prevents any further taxonomic attribution.

Scolopacidae Vigors, 1825
Scolopacidae indet.
Material. 2 bones (Table 1; Supplementary 1).
Remarks. These two bones belong to a small-sized Charadriiformes. The morphology of the processus supracondylaris dorsalis of the humerus and of the pointed shape of the crista trochanteris of the femur refer them to a Scolopacidae. The size of the bones recalls those of Tringa glareola, but their fragmentary status refers them only to Scolopacidae indet. The bones are morphologically different from those of Gallinago nigripennis, the only Scolopacidae currently breeding in Africa (del Hoyo and Collar, 2014), thus the remains most probably belong to a Palaearctic migrant.

Turnicidae Gray, 1840
Turnix sp.
Material. 10 bones (Table 1; Fig. 2J, L; Supplementary 1).
Remarks. The bones from Kromdraai can be referred to Turnicidae for the following characters: 1) very big fossa pneumotricipitalis ventralis in the proximal humerus; 2) small and pointed processus supracondylaris dorsalis in the distal humerus; 3) crista trochanteris well developed and tilted mediad in the proximal femur; 4) two condyles of the distal tibiotarsus not as different from each other as in the other Charadriiformes, but the condylus medialis proximo-distally narrower than the c. laterialis with a notch in the distal outline; 5) canal for the tendon of musculus flexor digitorum longus (Mayr, 2016) aligned plantarly to the eminentia intercotylaris in the proximal tarsometatarsus. The Turnix remains here analyzed are bigger than T. sylvatica and T. hottentotta and is here referred to Turnix sp., as the presence of more material from Kromdraai (Pockock, 1969) and Cooper’s Cave (Pavia et al. in prep.) suggests a broader analysis of the Turnix bones from the Cradle of Humankind. Button-quails of the genus Turnix are reported as fossils only from six localities worldwide, including Kromdraai (Goodman et al., 2013; Matsuoka, 2000; Meijer et al., 2013, 2019; Olson, 1994; Pockock, 1969; Zelenkov et al., 2016), despite their broad distribution in the Old World (Del Hoyo and Collar, 2014).

 

[Fred Ruhe]

Tytonidae Ridgway, 1914
Tyto cf. alba (Scopoli, 1769)
Material. 46 bones (Table 1; Fig. 4AeD; Supplementary 1).
Remarks. All the bones belong to a Tytonidae rather than to a Strigidae (for the diagnostic characters see Louchart, 2002; Pavia and Mourer-Chauviré, 2011). The bones of Kromdraai are more similar in size and morphology to Tyto alba than to T. capensis, the former being slightly bigger in all the bones especially for the very elongated tarsometatarsus. However, all the skeletal elements from Kromdraai show some morphological differences with recent T. alba and are smaller, making their attribution uncertain, as was also suggested by Pockock (1969). A similar-in-size Tyto is common in other sites of the Cradle of Humankind, such as, Cooper’s Cave, Swartkrans, Sterkfontein, and Bolt’s Farm (Pavia pers. obs. 2019; Pavia et al. in prep.; Val and Stratford, 2016), and their analysis is recommended to understand the relationship of this taxon, which is probably mainly responsible for the accumulation of the smalland medium-sized vertebrate remains in all the sites. The presence of bones of young individuals, even of not-fledged ones, testifies to the breeding status of the species during the Plio-Pleistocene transition in the Cradle of Humankind. The barn owls are quite common in the fossil record, also in the African one (Pavia et al., 2015). In the Pliocene of Langebaanweg and Aramis, two extinct species have been reported, one still undescribed (Louchart et al., 2009; Pavia et al., 2015), but their relationships with the extant African species and with the Palaearctic ones, both fossil and recent, remain unclear. Moreover, Tyto alba is now a widespread species in Eurasia and Africa, but the time of its origin and the steps of its spread remain unknown, as also its closer relatives. In addition, T. capensis has never been found as fossil, despite the great number of South African fossil localities, suggesting its very recent evolution or immigration, even if its ancestor remains unknown. A wide analysis of the barn owl remains from Africa and their comparison with the Eurasian ones would clarify the tempo and mode of the evolution of those taxa and their relationships.

Strigidae Vigors, 1825
Glaucidium ireneae nov. sp.

Holotype. Right tarsometatarsus, complete KW 7976 (Fig. 3C).

Paratypes. Right coracoid, almost complete KW 8518a (Fig. 3A); right humerus, distal KW 7925 (Fig. 3B); right femur, complete KW not numbered (Fig. 3D).

Referred Material. One distal left humerus (KW 9046); one proximal right humerus (KW 7175 b); one complete left tarsometatarsus (KW 7175a); one distal right tarsometatarsus (KW 9155).

Derivation of name. The species is named after Irene Pellegrino, Italian specialist of phylogeography of birds, owlets in particular, the main supporter of all my scientific and research activities, and the beloved mother of our daughter Clara.

Type Locality and Horizon. Kromdraai site (260004100S, 27 4406000E) in the Cradle of Humankind, Gauteng, South Africa (Fig. 1). The fossil here described came from Kromdraai Member 2 (following Bruxelles et al., 2016), referred to the Plio-Pleistocene transition (Fourvel, 2018).

Fig. 1. Remains of Glaucidium ireneae nov. sp. from the Kromdraai Member 2. Paratype right coracoid KW 8518a (juvenile), dorsal view (A); paratype right humerus KW 7925, cranial view (B); holotype right tarsometatarsus KW 7976, dorsal view (C); paratype right femur KW not numbered, cranial view (D). The scale bar represents 10 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Diagnosis. Large Glaucidium species (Table 2), distinctly larger than all the African species except G. sjostedti, tarsometatarsus relatively slender with wide sulcus extensiorius and deep fossa infracotylaris, and humerus with a deep and well-defined fossa musculi brachialis and condylus dorsalis separated from the processus supracondylaris dorsalis by a deep groove.

Description and comparisons. Glaucidium ireneae nov. sp. is described on the basis of eight well preserved bones. It can be referred to Glaucidium on the basis of the osteological characters here described: coracoid: 1) processus procoracoideus straight in proximo-distal view and not curved, 2) processus acrocoracoideus rounded and poorly developed, 3) sulcus m. supracoracoidei shallow with small pneumatized foramen; humerus: 4) processus supracondylaris dorsalis very small, 5) squared cranial outline of the condylus ventralis in distal view, 6) distal end of the fossa m. brachialis very close to the condylus ventralis, 7) incisura intercondylaris deep and well defined; femur: 8) crista trochanteris not proximally developed resulting in a flat outline of the proximal end, 9) proximal impressio obturatoria well developed; tarsometatarsus: 10) squared eminentia intercotylaris, 11) pointed dorsal outline of the cotyla medialis, 12) sulcus extensorius shallow and short. The morphological characters described above excludes the attribution of Kromdraai bones to various African Otus and Ptilopsis species, which also are all smaller and bigger respectively. The extinct Athene inexpectata from the Pliocene of Langebaanweg (Pavia et al., 2015) shows morphological and morphometrical differences and is therefore excluded. The comparison with Glaucidium perlatum and G. sjostedti reveals the bigger size of the new species in all the known bones in respect of the former, with the tarsometatarsus more slender. G. sjostedti is similar in size with G. ireneae nov. sp., but shows a stouter tarsometatarsus (Table 2). Morphologically, the bones of the new species show differences with the two recent species, in having the humerus with a deeper and distally better-defined fossa musculi brachialis, a deeper sulcus between the condylus dorsalis and the processus supracondylaris dorsalis, and the tarsometatarsus with a wider sulcus extensorius, a deeper fossa infracotylaris dorsalis and a more evident impressio ligamenti collateralis lateralis. The other African species of Glaucidium are similar in size with G. perlatum with the tarsometatarsus shorter than G. ireneae nov. sp. (Weick, 2006), thus they can be
ruled out.

Remarks. The Strigidae are quite common in the fossil record, including in the African one (Pavia et al., 2015) with both extinct and extant species. Glaucidium ireneae nov. sp. represents the first unambiguous record of Glaucidium in Africa, as a skull of an owlet from the Pliocene of Taung, South Africa, was only tentatively referred to Glaucidium perlatum by Cooke (1990). This latter specimen should be re-analyzed in order to clarify its relationships within Strigidae and with the new species.

 

[Fred Ruhe]

Accipitridae Vigors, 1824
Accipiter melanoleucos Smith, 1830
Material. 5 bones (Table 1; Fig. 4MeN; Supplementary 1).
Remarks. Middle-sized Accipitridae, more recalling Accipiter or Melierax than the other Accipitridae. More in detail, the bones can be referred to Accipiter rather than to Melierax for the following characters: 1) tuberculum ventrale well developed, fossa musculi brachialis well developed proximally and whole humerus not slender; 2) ventral side of the facies articularis major protrudes ventrally in the distal carpometacarpus; 3) short tibiotarsus with straight condylus medialis and not protruding medially.
The size and morphology of the Kromdraai bones fit well with Accipiter melanoleucus, the African bigger species of Accipiter, which is here reported as a fossil for the first time.

Accipiter sp.
Material. 1 bone (Table 1; Supplementary 1).
Remarks. This remain belongs to a small-sized Accipiter, similar in size to Accipiter badius. As only the proximal end is preserved and there is a great amount of diversity among the mid- and small-sized African Accipiter species in terms of size and morphology, this remain is here referred to Accipiter sp.

cf. Gyps sp.
Material. 2 bones (Table 1; Fig. 4J; Supplementary 1).
Remarks. The ungueal phalanx more recalls the morphology of Aegypiinae vultures, rather than Gypaetinae or other Accipitridae, following Manegold et al. (2014). In particular, the morphology of the proximal end is more similar to Gyps than to Aegypius. The vertebra is referred to this taxon on the basis of the morphology and the huge size. Vultures are not commonly reported in the Neogene and
Pleistocene of Africa (Louchart, 2014; Manegold et al., 2014; Prassack et al., 2018), and this record, if confirmed, represents the first occurrence of Gyps in Africa as a fossil.

cf. Milvus sp.
Material. 1 bone (Table 1; Supplementary 1).
Remarks. This phalanx 1 of digit I shows the general morphology of Accipitridae with a flat and enlarged proximal end. This remain is almost indistinguishable from the same element of
the recent Milvus migrans, making the Kromdraai bone, if confirmed, the oldest record of Milvus so far and the second from Africa, after the Late Pleistocene of Madagascar (Goodman et al., 2013).

Accipitridae indet. sp. 1
Material. 2 bones (Table 1; Fig. 4I; Supplementary 1).
Remarks. These phalanges belong to an Accipitridae of big size, but their morphology refers them to a large-sized eagle, either Aquila, Polemaetus, or Stephanoaetus, but the attribution is left uncertain due to the general similarities of those phalanges among the different Accipitridae.

Accipitridae indet.
Material. 2 bones (Table 1; Supplementary 1).
Remarks. These two ungueal phalanges of Accipitridae probably belong to one of the previously described taxa but cannot be referred with certainty to any of them.

 

[Fred Ruhe]

Coraciidae Rafinesque, 1815
Coracias sp.
Material. 1 bone (Table 1; Fig. 4H; Supplementary 1).
Remarks. The carpometacarpus from Kromdraai is similar in size and morphology to Coracias garrulus and overlaps in size with different African species. The analysis of the available comparative material did not reveal any solid morphological characteristics to clarify the affinities of this remain. This record represents the oldest occurrence of Coracias worldwide and the first occurrence for the genus in the African fossil record.

Falconidae Vigors, 1824
Falco sp.
Material. 31 bones (Table 1; Fig. 4E, L; Supplementary 1).
Remarks. The fossil remains from Kromdraai are attributable to a small-sized Falco, slightly bigger than F. tinnunculus and smaller that F. rupicoloides. The osteological morphology of the member of the genus is rather homogeneous (Solti, 1996) and the skeletons of the African species available for comparisons were not enough to better the attribution below the generic level. As for Tyto cf. T. alba, this rupicolous species bred in the area during the Plio-Pleistocene, as demonstrated by the bones of young individuals.

Psittaculidae Vigors 1825
Agapornis sp.
Material. 1 bone (Table 1; Fig. 4G; Supplementary 1).
Remarks. This small-sized humerus is attributed to Agapornis on the basis of the morphological characters described by Manegold (2013). Agapornis has already been reported from Kromdraai by Pockock (1969) and Stidham (2009) and is quite common at Cooper’s Cave (Pavia et al. in prep.). The attribution of this humerus is left to the generic level, as a comprehensive analysis of the parrot remains from the Cradle of Humankind and their relationship with those from the Pliocene of Langebaanweg (Manegold, 2013) is required.

 

[Fred Ruhe]

Corvidae Vigors, 1825

Corvus bragai nov. sp.

Holotype. Right humerus, complete KW 9104 (Fig. 5A).

Paratypes. Right coracoid, complete KW 6727 (Fig. 5B); right ulna, complete KW 9072 (Fig. 5D and E); right carpometacarpus, complete KW 9161 (Fig. 5C), left tibiotarsus, distal (KW 6636); one posterior phalanx 1-I, complete (KW 8435).

Referred Material. One proximal left coracoid (KW 10712e); one complete right coracoid (not numbered); one proximal left humerus (KW 7086); two distal left humeri (KW 6649, KW 7915); one proximal right humerus (KW 8011); two distal right humeri (KW 7048, KW 8023); one proximal left ulna (KW 6706); one distal left ulna (KW 9049 b); one proximal right ulna (KW 7004a); one distal right ulna (KW 10720 b); one proximal left carpometacarpus (KW 7008a); two distal right carpometacarpi (KW 6592, KW 8552); one proximal left femur (KW 8513); three complete posterior phalanges (KW 8516).

Derivation of name. The species is named after Jose Braga, French anthropologist and leader of the Kromdraai Research Project, for his essential contribution to the palaeontology and palaeoanthropology of Kromdraai and of the Cradle of Humankind, and for believing in the importance of birds in the analysis of fossil vertebrate assemblages.

Type Locality and Horizon. Kromdraai site (26
0004100S, 27
4406000E) in the Cradle of Humankind, Gauteng, South Africa (Fig.1). The fossil here described came from Kromdraai Member 2 (following Bruxelles et al., 2016), referred to the Plio-Pleistocene transition (Fourvel, 2018).

Diagnosis. Small Corvus species (Table 3), slightly smaller than C. albus and C. capensis, coracoid with deep processus m. supracoracoidei and very reduced processus procoracoideus, humerus with squared crista bicipitalis and rounded caput humeri, and ulna with thin olecranon with globular tip and pneumatized impression brachialis.

Description and comparisons. The remains from Kromdraai agree with Corvus in the morphology of all the preserved bones (Tomek and Bochenski, 2000). This new species is markedly smaller than Corvus crassirostris, C. albicollis, C. ruficollis, C. ripidhurus, which are not further considered for comparison. Corvus bragai is slightly smaller than C. albus and C. capensis with which it has been compared and can be separated for the following characters: coracoid: 1) sulcus m. supracoracoidei deep, 2) processus procoracoideus very reduced, 3) accessory facies articularis sternalis on ventral side absent; humerus: 4) crista bicipitalis with squared ventral outline, 5) caput humeri rounded and well pronounced proximally, ulna: 6) thin olecranon with globular tip, 7) proximal end of the impressio brachialis deep and slightly pneumatized, 8) pointed tuberculum carpale; carpometacarpus: 9) trochlea carpalis little developed proximally, 10) processus alularis faintly separated from processus extensorius with a very small step. The morphological characteristics detailed above and the dimensional differences will describe the Kromdraai Corvus as a new species, of which relationships with the extant African species could be clarified through the analysis of more skeletal elements from the Cradle of Humankind (Brain, 1993, Val and Stratford, 2016; Pavia, 2019 pers. obs.) or East Africa (Prassack et al., 2018).

Remarks. Despite the Corvidae are known in the Palaearctic fossil record since the Late Miocene and became very common from the Late Pliocene (Mlíkovský, 2002; Tyrberg, 1998), their African fossil record is very poor and still not described in detail. Corvidae and Corvus sp. are reported from Early Pleistocene of Swartkrans (Brain, 1993; Val and Stradford, 2015) and Olduvai (Prassack et al., 2018) and in other localities of the Late Pleistocene and Holocene of sub-Saharan Africa and Madagascar (Goodman et al., 2013; Plug and Clark, 2008; Prassack et al., 2018; Val, 2016; Val and Stratford, 2016). Corvus bragai is thus the oldest occurrence of the genus Corvus, and the family Corvidae, in all of Africa.

Fig. 2. Remains of Corvus bragai nov. sp. from the Kromdraai Member 2. Holotype right humerus KW 9104, caudal view (A); paratype right coracoid KW 6727, dorsal view (B); paratype right carpometacarpus KW 9161, ventral view (C); paratype right ulna KW 9072, cranial (D) and dorsal (E) views. The scale bar represents 10 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

 

[Fred Ruhe]

Hirundinidae Rafinesque, 1815
Hirundinidae indet
Material. 2 bones (Table 1; Fig. 4F; Supplementary 1).
Remarks. The two humeri show the peculiar morphology of Hirundinidae (Manegold, 2010; Zelenkov and Kurochkin, 2012). They belong to a mid- or large-sized species, but the great diversity of African swallows and the lack of enough comparative material prevent their better attribution. Swallow remains have already been reported from Kromdraai (Pockock, 1969) which, together with Langebaanweg (Manegold, 2010), represent the only African fossil locality bearing Hirundinidae.

Sturnidae Rafinesque, 1815
Sturnidae indet
Material. 20 bones (Table 1; Fig. 4K; Supplementary 1).
Remarks. All the examined humeri belong to middle-sized Passeriformes. The presence of the two deep and clearly separated fossae in the proximal end refer the Kromdraai remains to the Sturnidae in respect to the other Passeriformes families of the same size. The morphology of the humeri of the various Sturnidae species is very similar one to each other and the comparative material examined didn’t reveal solid differences among taxa, thus the remains have been identified as Sturnidae indet. The Sturnidae are reported in different African fossil localities since the Pliocene of Langebaanweg (Manegold et al., 2013), even in South Africa (Brain, 1993; Harrison, 1980; Plug and Clark, 2008; Pockock, 1969; Prassack et al., 2018; Val, 2016).

Fred