Eocypselus geminus sp. nov., Eocypselus grandissimus sp. nov, and Eocypselus paulomajor sp. nov. (1 Viewer)

[albertonykus]

Mayr, G. and A.C. Kitchener (2024)
New fossils of Eocypselus and Primapus from the British London Clay reveal a high taxonomic and ecological diversity of early Eocene swift-like apodiform birds
Ibis (advance online publication)
doi: 10.1111/ibi.13323

We describe new specimens and species of apodiform birds from the early Eocene London Clay of Walton-on-the Naze (Essex, UK). In addition to multiple partial skeletons of Eocypselus vincenti Harrison, 1984, three new species of Eocypselus are described as Eocypselus geminus, sp. nov., Eocypselus paulomajor, sp. nov. and Eocypselus grandissimus, sp. nov. The previously unknown quadrate of Eocypselus shares a characteristic derived morphology with the quadrate of the Aegothelidae, Hemiprocnidae and Apodidae, whereas the quadrate of the Trochilidae is very different. We also report a striking disparity of the shapes of the axis vertebra of apodiform birds, which is likely to be of functional significance. Eocypselus and extant Hemiprocnidae and Cypseloidini (Apodidae) exhibit the plesiomorphic morphology, whereas a derived shape characterizes extant Aegothelidae, Apodini and Trochilidae. Furthermore, we describe the first partial skeleton of the earliest aegialornithid species, Primapus lacki Harrison & Walker, 1975, which was previously only known from the humeri of the type series that stem from different sites of the London Clay. The apodiform birds from Walton-on-the-Naze show a considerable taxonomic and ecomorphological diversity, and whereas Eocypselus may have inhabited forest edges and caught insects by sallying flights from perches, Primapus probably was a fast-flying and more aerial bird.

 

[Fred Ruhe]

SYSTEMATIC PALAEONTOLOGY

Aves Linnaeus, 1758
Apodiformes Peters, 1940
Eocypselidae Harrison, 1984
Eocypselus Harrison, 1984
Typespecies Eocypselus vincenti Harrison, 1984

Emended diagnosis
Characterized and distinguished from other Apodiformes by the combination of the following features: axis with processus spinosus well separated from processus articulares caudales; furcula without laterally projected facies articularis acrocoracoidea; sternum with deeply concave and dorsoventrally narrow sulci articulares coracoidei; tarsometatarsus slightly longer than carpometacarpus.

Eocypselus vincenti Harrison, 1984

Referred specimens
NMS.Z.2021.40.109 (Fig. 1b; partial furcula, both coracoids, left humerus, both ulnae, right radius, distal portion of left radius, both partial carpometacarpi, both phalanges proximales digitorum majorum, right os carpi ulnare); collected in 1980 by M. Daniels (original collector’s number WN 80311B). NMS.Z.202.40.110 (Fig. 3i; right humerus); collected in 1977 by M. Daniels (original collector’s
number WN 77184). NMS.Z.2021.40.111 (Fig. 1c; cranial extremity of right scapula, left coracoid, partial left humerus, proximal and distal portions of left ulna, partial left carpometacarpus, proximal portion of right carpometacarpus); collected in 1996 by M. Daniels (original collector’s number WN 96960). NMS.Z.2021.40.112 (partial left humerus); collected in 1987 by M. Daniels (original collector’s number WN 87569). NMS.Z.2021.40.114 (Fig. 3j; partial right humerus, distal portion of left ulna, partial left carpometacarpus, right phalanx proximalis digiti majoris); collected in 1982 by M. Daniels (original collector’s number WN 82405B). NMS.2021.40.115 (Fig. 1d; tip of rostrum, extremitas omalis of left coracoid, cranial extremity of left scapula, cranial portion of sternum, distal end of left ulna, left os carpi ulnare); collected in 1991 by M. Daniels (original collector’s numberWN 91693B).

Locality and horizon
Walton-on-the-Naze, Essex, UK; Walton Member of the London Clay Formation (previously Division A2; Jolley 1996, Rayner et al. 2009, Aldiss 2012), early Eocene (early Ypresian, 54.6–55 Ma; Collinson et al. 2016).

Measurements (maximum length in mm)
NMS.Z.2021.40.109: left coracoid, 11.6; right coracoid, 11.4; left humerus, 12.4; left ulna, 17.2; right ulna, 17.2; right radius, 16.1; left carpometacarpus, 11.0; right carpometacarpus, 11.0. NMS.2021.40. 111: left coracoid, 10.4; left humerus, 11.6; left carpometacarpus, 10.6.

Fred


Figure 1. Specimens of Eocypselus (Eocypselidae) and Primapus (Aegialornithidae) from the early Eocene London Clay of Waltonon-
the-Naze (Essex, UK). (a–d) Eocypselus vincenti (a: holotype, NHMUK A 5429; b: NMS.Z.2021.40.109; c: NMS.Z.2021.40.111; d:
NMS.Z.2021.40.115). (e, f) E. geminus, sp. nov. (e: holotype, NMS.Z.2021.40.116; f: NMS.Z.2021.40.117). (g–k) Eocypselus sp. (g:
NMS.Z.2021.40.122; h: NMS.Z.2021.40.113 (the arrow denotes an enlarged detail of the processus lateralis); i: NMS.Z.2021.40.130;
j: NMS.Z.2021.40.128 (the dotted frame indicates an ulna that probably does not belong to the specimen); k: NMS.Z.2021.40.129
(the caudalmost portion of the neurocranium, which is preserved in a separate piece of matrix, is not shown)). (l) E. paulomajor, sp.
nov. (holotype, NMS.Z.2021.40.118). (m) E. grandissimus, sp. nov. (holotype, NMS.Z.2021.40.119). Primapus lacki (NMS.Z.2021.40.133). pla, processus lateralis. Scale bars: 5 mm. (Colour online).

 

[Fred Ruhe]

Eocypselus geminus, sp. nov.

Holotype
NMS.Z.2021.40.116 (Fig. 1e; partial rostrum and mandible, axis, furcula (broken into three pieces), right scapula, right coracoid in a piece of matrix, partial sternum, partial left humerus, left radius, proximal portion of right radius, right phalanx proximalis digiti majoris, both carpometacarpi, fragments of the pelvis, right tarsometatarsus lacking proximal end, a few pedal phalanges and an ungual phalanx); collected in 1991 by M. Daniels (original collector’s number WN 91680).

Differential diagnosis
Distinguished from Eocypselus vincenti in that the processus procoracoideus of the coracoid projects less strongly towards the processus acrocoracoideus; the furcula has wider omal sections of the scapi clavicularum; the humerus has a stouter shaft and a dorsoventrally wider distal end (ratio distal width to humerus length 0.27 in E. vincenti vs. ~ 0.30 in E. geminus). Differs from E. rowei in that the tarsometatarsus is longer (> 11.0 mm in the holotype vs. 10 mm in E. rowei, whereas the humerus and carpometacarpus of the holotype are of similar length to the corresponding bones of E. rowei; Table 1); the incisura intertrochlearis medialis of the tarsometatarsus is narrower (compare Fig. 5b,c with Ksepka et al. 2013: fig. 1f).

Etymology
The species epithet is derived from geminus (Lat.), twin, in reference to the close similarity of the new species to E. vincenti.

Type locality and horizon
Walton-on-the-Naze, Essex, UK; Walton Member of the London Clay Formation, early Eocene (early Ypresian).

Referred specimen
NMS.Z.2021.40.117 (Fig. 1f; left coracoid, omal extremity of right coracoid, left humerus, proximal portion of right humerus, left ulna, proximal portion of right ulna; a mediolaterally narrow premaxilla associated with the specimen belongs to a different, non-apodiform species); collected in 1992 by M. Daniels (original collector’s number WN 92727A).

Measurements (maximum length in mm)
NMS.Z.2021.40.116: left humerus, length as preserved, 11.3, estimated total length, ~11.8; left carpometacarpus, 10.9; right tarsometatarsus, length as preserved, 11.0, estimated total length, ~11.2. NMS.Z.2021.40.117: left coracoid, 11.1; left humerus, 12.4; left ulna, 17.0.

Fred

 

[Fred Ruhe]

Eocypselus paulomajor, sp. no

Holotype
NMS.Z.2021.40.118 (Fig. 1l; partial furcula, left coracoid, left ulna); collected in 1997 by M. Daniels (original collector’s number WN 97987).

Differential diagnosis
Distinctly larger than E. vincenti and E. geminus (Table 1); furthermore, differs from E. geminus in that the scapi clavicularum of the furcula are narrow throughout their lengths; the processus lateralis of the coracoid is better developed and has a more strongly convex lateral margin (Fig. 3a–d).

Etymology
The species epithet is derived from paulum major (Lat.), somewhat larger, and refers to the size of the new species in comparison with Eocypselus vincenti

Type locality and horizon
Walton-on-the-Naze, Essex, UK; Walton Member of the London Clay Formation, early Eocene (early Ypresian).

Measurements (maximum length in mm)
Left coracoid, 12.4; left ulna, 19.0.

Fred

 

[Fred Ruhe]

Eocypselus grandissimus, sp. nov.

Holotype
NMS.Z.2021.40.119 (Fig. 1m; a few vertebrae, both coracoids, both scapulae, cranial portion of sternum, distal end of right humerus, proximal end of left humerus, partial left carpometacarpus, distal
end of left femur, one pedal phalanx; a distal left humerus associated with the specimen belongs to a non-apodiform species); collected in 1994 by M. Daniels (original collector’s number WN 94825A).

Differential diagnosis
Distinguished from all other Eocypselus species by its distinctly larger size (Table 1).

Etymology
The species epithet is the superlative of grandis (Lat.), large, and refers to the fact that the new species is the largest known representative of Eocypselus.

Type locality and horizon
Walton-on-the-Naze, Essex, UK; Walton Member of the London Clay Formation, early Eocene (early Ypresian).

Measurements (maximum length in mm)
Left coracoid, 13.8; left carpometacarpus, 13.0.

Remarks
The holotype includes only wing and pectoral girdle bones, and no similar-sized leg bones of an Eocypselus-like apodiform bird could be identified in the Daniels collection at NMS. A possible candidate specimen is the holotype of Parvulivenator watteli Harrison 1982, which is of appropriate size and superficially resembles the Eocypselus tarsometatarsus. The holotype of P. watteli stems from Walton-on-the-Naze and consists of the distal portion of the tarsometatarsus (Fig. 5i,j,q) with a few associated pedal phalanges. The fossil was originally identified as a tiny falcon (Falconiformes; Harrison 1982) but there exists no reasonable basis for this classification. If Parvulivenator watteli is conspecific with the largest eocypselid species from Walton-on-the-Naze, the taxon Eocypselus Harrison 1984 would be a junior synonym of Parvulivenator Harrison 1982. However, there appear to be subtle differences between that tarsometatarsus of Parvulivenator and Eocypselus that conflict with a potential synonymy of these taxa. Compared with Eocypselus, the plantar rim of the fourth trochlea of
Parvulivenator extends farther proximally (Fig. 5m–q) and, as shown in the drawing of the holotype by Harrison (1982), it was plantarly more pronounced (the plantar portion of the rim is now broken in the specimen). Therefore, we maintain Parvulivenator and Eocypselus as separate taxa, but the possibility of synonymy should be scrutinized once further material becomes available.

Fred

 

[Fred Ruhe]

Eocypselus sp.

Referred specimens
NMS.Z.2021.40.113 (Fig. 1h; partial furcula, right coracoid, right scapula lacking caudal end, partial sternum, left os carpi ulnare); collected in 1992 by M. Daniels (original collector’s number WN
92750A). NMS.Z.2021.40.120 (Fig. 2o–r; right quadrate); collected in 2001 by M. Daniels (original collector’s number WN 011029A). NMS.Z. 2021.40.121 (distal portion of left tibiotarsus, ungual phalanx); collected in 1975 by M. Daniels (original collector’s number WN 75061). NMS.Z. 2021.40.122 (Fig. 1g; left tarsometatarsus); collected in 2004 by M. Daniels (original collector’s number WN 041064B). NMS.Z.2021.40.123 (partial
sternum); collected in 1990 by M. Daniels (original collector’s number WN 90668). NMS.Z. 2021.40.124 (fragmentary margo costalis of sternum, fragmentary proximal end of right humerus, phalanx distalis digiti majoris); collected in 1978 by M. Daniels (original collector’s number WN 78229). NMS.Z.2021.40.125 (left scapula lacking caudal portion, fragmentary proximal end of right carpometacarpus); collected in 1992 by M. Daniels (original collector’s number WN 92730B). NMS.Z. 2021.40.126 (omal extremity of left coracoid, cranial extremity of left scapula); collected in 1976 by M. Daniels (original collector’s number WN 76160). NMS.Z.2021.40.127 (right coracoid, left scapula lacking caudal end); collected in 1979 by M. Daniels (original collector’s number WN 79267B). NMS.Z. 2021.40.128 (Fig. 1j; fragment of furcula, both humeri (left one on a block of matrix), partial left carpometacarpus; an ulna associated with the specimen is not from an apodiform bird); collected in 1987 by M. Daniels (original collector’s number WN 87551A). NMS.Z.2021.40.129 (Fig. 1k; tip of rostrum, neurocranium, processus oticus of right quadrate, distal end of right tibiotarsus); collected in 1982 by M. Daniels (original collector’s number WN 82385). NMS.Z.2021.40.130 (Fig. 1i; partial furcula, fragmentary right coracoid, right scapula, extremitas cranialis of left scapula, left ulna, right os carpi ulnare; a pedal phalanx associated with the specimen is too large and belongs to a different species); collected in 1984 by M. Daniels (original collector’s number WN 84488A). NMS.Z.2021.40.131 (fragmentary left coracoid, extremitas omalis of right coracoid, fragment of proximal end of left humerus); collected in 1986 by M. Daniels (original collector’s number WN 86543). NMS.Z.2021.40.132 (extremitas omalis of right coracoid, fragments of the proximal ends of both humeri, and a few other bone fragments); collected in 1994 by M. Daniels (original collector’s number WN 94841).

Locality and horizon
Walton-on-the-Naze, Essex, UK; Walton Member of the London Clay Formation, early Eocene (early Ypresian).

Measurements (maximum length in mm)
NMS.Z.2021.40.122: left tarsometatarsus, 11.7. NMS.Z.2021.40.130: left ulna, 16.2. NMS.Z. 2021.40.128: left humerus, 11.4; left carpometacarpus, length as preserved, 10.3.

Remarks
These specimens are from species the size of E. vincenti and E. geminus but cannot be unambiguously assigned to either species. Furthermore, some specimens show features that may indicate a taxonomic distinctness. For example, the coracoid of NMS.Z.2021.40.113 differs from other specimens in that the extremitas sternalis exhibits a pointed processus lateralis (Fig. 1h).
The humeri of NMS.Z.2021.40.128 (Fig. 1j) closely correspond to those of the E. vincenti holotype, but the ulna and carpometacarpus associated with the specimen show different morphologies. The ulna is much shorter relative to the humerus than the ulna of E. vincenti, the olecranon is narrower than in E. vincenti, and the tuberculum ligamenti collateralis ventralis is smaller. Because the bone is so different from the E. vincenti ulna, whereas the humerus is nearly identical to that of E. vincenti, we consider it likely that the ulna associated with NMS.Z.2021.40.128 does not belong to the same species as the humerus and is not from an apodiform bird. The carpometacarpus of NMS.Z.2021.40.128 (Fig. 3y) has a proportionally larger trochlea carpalis than that of E. vincenti and E. geminus and the processus alularis is not well defined.
The tarsometatarsus NMS.Z.2021.40.122 (Fig. 5d–f) differs from that of E. geminus sp. nov. in its proportions (see description above), but otherwise has a very similar morphology. The bone is assigned to Eocypselus because there is no other similar-sized small bird known from the London Clay of Walton-on-the-Naze, which may have had a similar tarsometatarsus morphology. It is associated
with a partial tarsometatarsus of a similar-sized zygodactylid species (see Mayr & Kitchener 2023 for a review of the Zygodactylidae from Waltonon- the Naze) and a few other fragmentary bones (distal ulna, distal radius, distal femur).

Fred

 

[Fred Ruhe]

Aegialornithidae Lydekker, 1891
Primapus Harrison & Walker, 1975
Typespecies Primapus lacki Harrison, 1975

Emended diagnosis
Furcula with robust scapi clavicularum and extremitas omalis lacking a laterally protruding facies articularis acrocoracoidea; processus acrocoracoideus of coracoid narrow in sterno-omal direction; humerus short and stocky, with well-developed processus supracondylaris dorsalis.

Remarks
The previous assignment of Primapus to the Aegialornithidae was
based on the humerus (Harrison & Walker 1975), which closely resembles the humerus of the aegialornithid taxon Aegialornis (Fig. 7m,n) in combining a relatively slender shaft (a plesiomorphic feature) and a well-developed processus supracondylaris dorsalis (a derived trait). The new specimens show that P. lacki also agrees with aegialornithids in the morphologies of the coracoid and carpometacarpus.
Olson (1985: 135) noted that the distal tarsometatarsus, which forms the holotype Procuculus minutus Harrison & Walker, 1977, may actually stem from Primapus lacki. Procuculus minutus was initially described as a cuckoo, which it definitely is not (Olson 1985), and the holotype and only known specimen comes from the Fish Tooth Beds of Bognor Regis (Sussex, UK), which also yielded the holotype of P. lacki. The distal tarsometatarsus of P. minutus (Fig. 5k,l,r) is of similar size to that of Eocypselus vincenti/E. geminus (Fig. 5b,c,m), but has a proportionally larger foramen vasculare distale. As far as comparisons are possible, the tarsometatarsus of P. lacki corresponds well to that of Aegialornis gallicus from the Quercy fissure fillings in France (Mourer-Chauvire 1988: pl. 1l, m). As in Ae. gallicus, the trochlea metatarsi II reaches farther distally
than the tr. mt. IV. We therefor concur with Olson (1985) that the Procuculus minutus tarsometatarsus is actually from Primapus lacki and synonymize Procuculus Harrison & Walker, 1977 and Procuculus minutus Harrison & Walker, 1977 with Primapus Harrison & Walker, 1975 and Primapus lacki Harrison &Walker, 1975, respectively.

Primapus lacki Harrison, 1975

Referred specimen
NMS.Z.2021.40.133 (Figs 1n and 7; furcula, right coracoid, left and right humeri lacking distal ends, left radius, partial right carpometacarpus, pedal phalanx); collected in 1993 by M. Daniels (original collector’s number WN 93766).

Locality and horizon
Walton-on-the-Naze, Essex, UK; Walton Member of the London Clay Formation, early Eocene (early Ypresian).

Measurements
Right coracoid, 9.2; left humerus, length as preserved, 9.9; right humerus, length as preserved, 10.0; left radius, 18.1; right carpometacarpus, 12.9.

Remarks
This specimen was erroneously identified as a ‘Scaniacypselus-like apodiform bird’ by Mayr (2022: 111). However, it differs from Scaniacypselus and agrees with Primapus in that the shaft of the humerus is slenderer, the crista bicipitalis forms a more pronounced convexity and the carpometacarpus is shorter relative to the humerus length (the bone is about as long as the humerus in Primapus, but it is much longer than the humerus in Scaniacypselus). The new fossil is somewhat older than the previous records of P. lacki from Bognor Regis and the Isle of Sheppey.

Fred