Ichthyornis tener, Marsh, 1880

CLARKE, JULIA A., 2004, Morphology, Phylogenetic Taxonomy, And Systematics Of Ichthyornis And Apatornis (Avialae: Ornithurae), Bulletin of the American Museum of Natural History 2004 (286), pp. 1-179 : 44-48

tener Marsh, 1880 (converted species

name)

DEFINITION: The name ‘‘ tener ’’ Marsh, 1880 is converted and defined as the name for the species that includes YPM 1760 (see the Introduction of the Taxonomic Revision for the definition of ‘‘species’’ used). With its clade address (sensu Cantino et al., 1999), the converted name of this species is ‘‘ Guildavis tener ’’. The name ‘‘ tener ’’ was originally published in the combination ‘‘ Ichthyornis tener ’’ ( Marsh, 1880).

HOLOTYPE SPECIMEN: A specimen indicat­ ed by Marsh as the holotype was never figured, described, or diagnosed. YPM 1760, a sacrum (fig. 13), was specified in the following way: ‘‘The type specimen (number 1760) of the present species [ Ichthyornis tener ] was found in 1879, by Mr. E. W. Guild in Wallace County, Kansas.’’ ( Marsh, 1880: 198). The locality, date of collection, and collector are that of YPM 1760, which is comprised of the sacrum. However, a humerus ( YPM 1738) was figured in Odontornithes ( Marsh, 1880: pl. XXX, fig. 8) with the name ‘‘ Ichthyornis tener ’’ applied to it and this figure is referenced under the heading ‘‘ Ichthyornis tener ’’ where the species is named ( Marsh, 1880: 198). The specimen number of the humerus was not mentioned in the text ( Marsh, 1880) as either the holotype or as referred to Ichthyornis tener , although another specimen (a coracoid; YPM 1766) was mentioned as referred ( Marsh, 1880: 198). The humerus was collected by S. W. Williston in 1877 and, thus, does match the description of the intended type specimen.

Brodkorb (1967) listed YPM 1760, the specimen mentioned by Marsh (1880), as the holotype, but incorrectly described it as an isolated distal humerus. Presumably, Brodkorb (1967) thought that the figured distal humerus ( YPM 1738) in Odontornithes ( Marsh, 1880: pl. XXX, fig. 8), with the name ‘‘ Ichthyornis tener ’’ applied to it, was YPM 1760, as nowhere in the text is it mentioned that YPM 1760 is a sacrum. This confusion points to a fundamental taxonomic question: whether the name ‘‘ tener ’’ is available and, if so, if it is tied to the humerus figured ( YPM 1738) and not to the sacrum ( YPM 1760) Marsh (1880) designated as the holotype. Under the ICZN ( International Commission on Zoological Nomenclature, 1999; the PhyloCode does not yet address species level names) the name ‘‘ Ichthyornis tener ’’ would be available; its publication satisfies the provisions of Articles 11 and 12. However, it is less clear whether the name is tied to the figured humerus (the illustration clearly constituting an ‘‘indication’’ under Article 12.2.7; International Commission on Zoological Nomenclature, 1999) or the sacrum the specimen number of which is referred to explicitly as the holotype specimen and additionally specified by locality, collection year, and collector.

Here, the explicit designation by Marsh (1880) of YPM 1760 as the holotype of tener was followed in the choice of the internal specifier for both ‘‘ Guildavis ’’ and ‘‘ tener ’’ and considered the valid name­bearing specimen under the ICZN ( International Commission on Zoological Nomenclature, 1999). While the humerus is supported by morphological correspondence as a part of Ichthyornis dispar , the sacrum can be differentiated from Ichthyornis dispar and is placed in the phylogenetic analyses as more closely related to Aves than to Ichthyornis dispar .

YPM 1760 consists solely of two fragments of a tiny sacrum. One fragment, from the anterior end, preserves the anterior articular surface and the three anteriormost ankylosed vertebrae. The second fragment is from the middle of the sacral series. It is crushed dorsoventrally and no morphologies could be discerned.

LOCALITY AND HORIZON: Marsh (1880: 198) mentioned that YPM 1760 was ‘‘found in 1879 by Mr. E. W. Guild in Wallace County, Kansas.’’ No locality information other than ‘‘Cretaceous, Kansas’’ is provided on the original label associated with the specimen. The humerus (YPM 1738), figured as Ichthyornis tener in Odontornithes ( Marsh, 1880: pl. XXX, fig. 8), but never mentioned in the text as referred to Ichthyornis tener , was collected by S. W. Williston ‘‘near Monument Rock’’ two years earlier (i.e., in 1877).

DISCUSSION: Guildavis tener is differentiated from the holotype of Ichthyornis dispar by the presence of a parapophysis visible on the left side of the completely fused first sacral vertebra. The widths of the iliosynsacral sulci appear greater than in the holotype of Ichthyornis dispar . The subcircular anterior articular surface and lateral excavations in the first sacral are shared with the holotype of Ichthyornis dispar and referred specimens. YPM 1760 is significantly smaller than the holotype of Ichthyornis dispar ; the diameter of the anterior articular surface of the first sacral is approximately two­thirds that of the Ichthyornis dispar holotype. It is also smaller than the holotype of Apatornis celer . However, YPM 1760 cannot otherwise be differentiated relative to Apatornis celer ; these specimens cannot be compared directly, as the holotype of Apatornis celer lacks at least one vertebra from the anterior end of the sacrum and is strongly crushed mediolaterally.

REFERRED SPECIMENS: Marsh (1880: 198) referred a coracoid (YPM 1766) to this taxon and figured a humerus (YPM 1738; Marsh, 1880: pl. XXX, fig. 8) with the name ‘‘ Ichthyornis tener ’’ applied to it. These referrals are considered unsupported as neither specimen can be compared to the holotype. The only feature that the humerus and the holotype share is that they are both conspicuously smaller than Ichthyornis dispar . By contrast, the coracoid is approximately the same size as that of Ichthyornis dispar .

Furthermore, the holotype is not placed as part of Ichthyornis when included in analysis (Part II, Results), while the humerus and coracoid are referred to Ichthyornis dispar based on the correspondence of preserved morphologies. In the case of the humerus, several correspondent features include: hypertrophied muscle insertions on the dorsal supracondylar process and a narrow intercondylar groove consistent with referral to Ichthyornis .

Apatornis Marsh, 1873b (converted clade name)

DEFINITION: ‘‘ Apatornis ’’ Marsh, 1873b is here converted to a stem­based name ( de Queiroz and Gauthier, 1992) for the clade including all taxa/specimens more closely related to YPM 1451 ( celer Marsh, 1873a holotype specimen) than to Ichthyornis dispar or the following internal specifiers of Aves ( Gauthier and de Queiroz, 2001; see Introduction): Vultur gryphus , Struthio camelus Linnaeus, 1758 , Tetrao [ Tinamus ] major Gmelin, 1789, and Vultur gryphus Linnaeus, 1758 .

celer Marsh, 1873a (converted species name)

DEFINITION: The name ‘‘ celer ’’ Marsh, 1873a is converted and defined as the species that includes YPM 1451 (see the Introduction of the Taxonomic Revision for the definition of ‘‘species’’ used). With its clade address (sensu Cantino et al., 1999), the converted name of this species is ‘‘ Apatornis celer ’’. This name was originally published in the combination ‘‘ Ichthyornis celer ’’ ( Marsh, 1873a) and later in the combination, ‘‘ Apatornis celer ’’ ( Marsh, 1873b), when the species was identified as the type species of the genus Apatornis ( Marsh, 1873b) .

HOLOTYPE SPECIMEN: YPM 1451 is a sacrum lacking its proximal end ( Marsh, 1880; fig. 14).

LOCALITY AND HORIZON: Marsh (1880: 192) specified that he collected the specimen in October 1872, from Butte Creek, a locality discussed with reference to Ichthyornis agilis . Butte Creek, also known as Twin Butte Creek, is a well­known locality of Logan County, Kansas, in the Smoky Hill Chalk Member, Niobrara Formation ( Stewart et al., 1990), representing ( Bennett, 1990) beds between Marker Units 15 and 19 of Hattin (1982). Stewart (1990) estimated that the interval containing Hattin’s (1982) Marker Units 8–10 was upper Santonian in age. Thus, Marker Units 15–19 must represent, at the earliest, a period of time later in the late Santonian through the early Campanian. The top of the Smoky Hill Chalk Member, close to Marker Unit 23 of Hattin (1982), is in the early Campanian ( Stewart et al., 1990).

DISCUSSION: Marsh (1873a) originally differentiated the holotype (YPM 1451) from Ichthyornis dispar by more slender proportions and a more deeply concave posterior articular surface on the last vertebra of the sacrum. Neither of these characters was confirmed to differ discernibly in the holotype of Ichthyornis dispar . Marsh (1880), however, noted three additional differences from Ichthyornis dispar that are confirmed here. First, while both holotypes preserve 10 remnants of transverse processes, indicating the presence of 10 fused sacral vertebrae, that of Apatornis celer is incomplete anteriorly ( Marsh, 1880: 162). The sacrum of Apatornis celer had a least one more fused sacral vertebrae than Ichthyornis dispar . Spinal nerve openings visible on the anterior edge of the first preserved vertebra indicate that the series continued anteriorly.

Marsh also commented on a difference in the number of vertebrae with a particular morphology that occurs in approximately the middle of the sacral series ( Marsh, 1880: 162–163). There is a difference in the number of sacral vertebrae appearing to lack transverse processes in the middle of the series. The transverse processes appear to be oriented directly dorsally (appendix 1, character 62). Apatornis celer has four such sacrals between a short, blunt, morphologically distinct transverse process anteriorly and a costal strut at the level of the acetabulum, while Ichthyornis (YPM 1450; YPM 1732) has three. In a preliminary survey of extant birds, this count appeared even more conservative than total sacral number.

Finally, Marsh (1880: 163) noted that the Apatornis celer holotype does not have co­ ossified tendons expanding ‘‘posteriorly over as broad a region [of the dorsal surface of the sacrum] as in Ichthyornis .’’ The last not­ ed difference could also be described as the absence of a fan of ossified tendons in Apatornis celer that covers the posterior half of the dorsal surface of the Ichthyornis dispar holotype sacrum.

Based on these three characters from Marsh (1880), Apatornis celer can be differentiated from Ichthyornis dispar . However, the few characters preserved, rather than suggesting that Apatornis celer is particularly close to Ichthyornis dispar as originally proposed ( Marsh, 1873b), place Apatornis celer more closely to Aves than to Ichthyornis (Part II, Results).

REFERRED SPECIMENS: One specimen (YPM 1734) was referred to Apatornis celer ( Marsh, 1880) . This specimen, because of its completeness, has formed the basis for nearly all of (1) the description of Apatornis celer in Odontornithes ( Marsh, 1880), and (2) the subsequent treatment of Apatornis celer in the literature. As no part of YPM 1734 currently overlaps the holotype, there is no evidence to support referral of YPM 1734 to Apatornis celer ( Clarke, 2000a) . YPM 1734 was collected five years after the holotype. The two specimens were found by different collectors in different localities near the Smoky Hill River in western Kansas; the holotype was collected by Marsh from Butte Creek in what is currently called Logan County, while YPM 1734 was collected by F. H. Williston five years later from Gove County ( Marsh, 1880: 192).

The pelvis from YPM 1734 is missing and cannot be compared to the holotype. Furthermore, it appears that it was never able to be compared because it is not clear that sacral vertebrae were ever preserved with the pelvis (a situation that occurs with some frequency in basal avialans). The depicted pelvis is shown in lateral view ( Marsh, 1880: pl. XXXII, fig. 1) with no part of the sacral vertebrae visible. Thus, even this representation (fig. 15) cannot be compared to the holotype of Apatornis celer . Two characters, however, can be compared for the two specimens. Ilium length in YPM 1734 can be roughly compared with the length of the pelvis in the Apatornis celer holotype and ap­ pears notably longer than the holotype of Apatornis celer . The Apatornis celer holotype was differentiated from that of Ichthyornis dispar by the presence of at least one additional anterior sacral, and Marsh (1880) identified the preacetabular ilium as longer in the pictured pelvis of ‘‘ Apatornis ’’ (i.e., the referred specimen, YPM 1734) than in Ichthyornis dispar (YPM 1732) . However, even elongation of the preacetabular ilium or anterior part of the sacrum (if these measures are justifiably compared) would not be an apomorphy allying YPM 1734 with Apatornis celer ; an increase in preacetabular ilium length appears to be an apomorphy of Aves relative to Ichthyornis and more basal avialan taxa.

Furthermore, measurements were given only from the pelvic bones of YPM 1734, and no sacral vertebrae are mentioned in the text ( Marsh, 1880). Whether there was a basis for Marsh’s referral of YPM 1734 to Apatornis celer cannot be assessed. Marsh (1880) did not list characters shared by the holotype and YPM 1734 to identify this specimen as a part of Apatornis celer . Given that nearly all of the referrals Marsh (1880) made were of specimens that could not be compared to the holotypes of the species to which they were referred, YPM 1734 might, similarly, not have had any preserved elements in common with the holotype of Apatornis celer .

The referral of YPM 1734 to Apatornis celer is here considered unsupported, and this specimen is named as the holotype of a new species, Iaceornis marshi.