taxonID	type	description	language	source
276487C62739FF9DC8A79839FB577FFF.taxon	type_taxon	TYPE SPECIES. — Chilcacetus cavirhinus Lambert, Muizon & Bianucci, 2015 by original designation. OTHER REFERRED SPECIES. — Chilcacetus ullujayensis n. sp.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62739FF9DC8A79839FB577FFF.taxon	diagnosis	EMENDED DIAGNOSIS. — The differential diagnosis focuses primarily on differences with taxa that were found to be closely related to Chilcacetus in our phylogenetic analysis and comparison (other members of the Chilcacetus clade, Eoplatanistidae Muizon, 1988, Eurhinodelphinidae, and Squaloziphiidae (Muizon, 1991 )). Chilcacetus is a medium-size (bizygomatic width ranging from about 240 to 280 mm), longirostrine (ratio between rostrum length and condylobasal length from 0.68 to 0.73), and homodont dolphin genus differing from all other odontocetes except Eoplatanista Dal Piaz, 1916, Eurhinodelphinidae, Physeteroidea, Yaquinacetus Lambert, Godfrey & Fitzgerald, 2019, and Ziphiidae Gray, 1850 in the tubercule of the malleus being shorter than the articular head; from all other odontocetes except ‘ Argyrocetus ’ joaquinensis, Ninoziphius Muizon, 1983, Olympicetus Vélez-Juarbe, 2017, Simocetus Fordyce, 2002, Squaloziphius Muizon, 1991, and Ziphiodelphis abeli Dal Piaz, 1908 in having a finger-like posterior projection of the hamular process of the pterygoid; from all other odontocetes except Amphidelphis bakersfieldensis n. comb., ‘ Argyrocetus ’ joaquinensis, Crisocetus Gaetán, Paolucci & Buono, 2023, Dolgopolis Viglino, Gaetán, Cuitiño & Buono, 2021, Eoplatanista, Eurhinodelphinidae, Perditicetus Nelson & Uhen, 2020, Squaloziphius and Yaquinacetus in the postglenoid process of the squamosal being anteroposteriorly long and transversely thick (this process being even anteroposteriorly longer in Crisocetus, Dolgopolis, Squaloziphius, and Yaquinacetus); from Dolgopolis, Squaloziphius and Yaquinacetus in the dorsal opening of the mesorostral groove being more gradual anterior to the bony nares; from most members of other longirostrine to hyper-longirostrine homodont extinct families (Allodelphinidae, Eoplatanistidae, Eurhinodelphinidae, and Platanistidae) in the absence of a deep lateral groove along the rostrum, the posterior process of the periotic being elongated in a posterolateral direction and transversely broad, and in the absence of ankylosis for the symphysis of the mandibles. It further differs from Eurhinodelphinidae in lacking an extended edentulous anterior premaxillary portion of the rostrum, in the mandible being roughly as long as the rostrum, in the nasals partly overhanging the bony nares, in the more anteriorly elongated zygomatic process of the squamosal (ratio between the length and the height of the process ≥ 1.10), in the neurocranium being distinctly longer than wide, in the lesser transverse widening of the occipital shield (ratio between maximum width of supraoccipital at the lateral corners of the nuchal crest and postorbital width <0.70), in the posterior margin of the postorbital process being vertical, in the top of the temporal fossa being nearly as high as the nuchal crest, in the palatines being separated anteromedially for a long distance at rostrum base, and in the longer and more laterally directed posterior process of the periotic. It differs from Eoplatanistidae in the premaxillary foramen being roughly at the level of the antorbital notch, in the thinner and flatter antorbital process, in the acute anterior margin of the nasal partly overhanging the bony nares, in the transversely concave and less anteriorly projected anterodorsal portion of the supraoccipital shield, in the deep anterior bullar facet of the periotic, and in bearing a conspicuous median furrow on the tympanic bulla. It differs from Argyrocetus patagonicus in the premaxillae being transversely wider at rostrum base, in lacking a wide dorsal opening of the mesorostral groove, in the premaxillary foramen being roughly at the level of the antorbital notch, in the angle formed by the basioccipital crests in ventral view <50 °, in the top of the temporal fossa being nearly as high as the nuchal crest, and in the absence of ankylosis for the symphysis of the mandibles. It differs from Amphidelphis bakersfieldensis n. comb. in its larger size, in lacking a deep sulcus anterior to the main dorsal infraorbital foramen at rostrum base, and in the palatines being separated anteromedially for a long distance at rostrum base. It differs from ‘ Argyrocetus ’ joaquinensis in the dorsal opening of the mesorostral groove anterior to the rostrum base being narrower than the premaxilla, in the presence of more than one dorsal infraorbital foramen at rostrum base, in lacking a deep sulcus anterior to the main dorsal infraorbital foramen at rostrum base, in the proportionally shorter and wider nasal, in the nasal partly overhanging the bony nares, in the wide exposure of the frontal on the vertex, in the posterior margin of the postorbital process being vertical, in the palatines being separated anteromedially for a longer distance at rostrum base, and in the ventral margin of the postglenoid process of the squamosal being approximately at the same level as the ventral margin of the exoccipital in lateral view. It differs from Caolodelphis Godfrey & Lambert, 2023 in its larger size, the frontals not being separated anteromedially on the vertex, and the basioccipital crests being transversely thinner. It differs from Macrodelphinus in its smaller size, in the premaxillary portion of the rostrum making less than 10 % of its total length, in the premaxillary foramen being roughly at the level of the antorbital notch, in the nasal proportionally longer compared to the frontal on the vertex, in the palatines being separated anteromedially for a longer distance at rostrum base. It differs from Perditicetus in in the premaxillary foramen being roughly at the level of the antorbital notch and the zygomatic process of the squamosal being dorsoventrally more slender. It differs from Papahu Aguirre-Fernández & Fordyce, 2014 in the rostrum being proportionally dorsoventrally thicker in its proximal part, in the single premaxillary foramen being roughly at the level of the antorbital notch, in the dorsal exposure of the premaxilla being wider than the dorsal exposure of the maxilla at rostrum base, in the proportionally wider ascending process of the premaxilla, in the anterodorsal elevation of the dorsal surface of the nasal, in the posterolateral projection of the nasal, in the elongate postorbital process of the frontal, in the longer and deeper anterior bulla facet of the periotic, and in the posterior elevation of the dorsal margin of the mandible being progressive.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C6273BFF91C9759D94FD907EB8.taxon	description	(Figs 2 - 4)	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C6273BFF91C9759D94FD907EB8.taxon	materials_examined	TYPE MATERIAL. — Holotype. Peru • 1 specimen (nearly complete cranium with the associated mandibles and the manubrium; the anterior portion of the rostrum and of the right mandible, and the left mandible were re-discovered in the MNHN collection after the initial publication of this specimen); East Pisco Basin; 14 ° 34 ’ 40 ” S, 75 ° 38 ’ 40 ” W; Chilcatay Formation; Burdigalian (late Early Miocene); MNHN. F. PRU 11. TYPE LOCALITY. — Precise locality unknown, but close to Ullujaya, East Pisco Basin, Peru. Approximate geographic coordinates: 14 ° 34 ’ 40 ” S, 75 ° 38 ’ 40 ” W (Fig. 1 A, B). TYPE HORIZON AND AGE. — In Ullujaya, cetacean fossil remains only occur, abundantly, in the widely exposed Lower Miocene Chilcatay Formation. More specifically, all cetaceans originate from the Ct 1 a facies association, a sub-horizontal 36 m package of interbedded sandstones, sandy siltstones, and siltstones punctuated by conglomerate levels (Bianucci et al. 2018; Di Celma et al. 2018). The age of the entire Chilcatay Formation exposed at Ullujaya can be restricted between 19.28 and 17.95 Ma (Burdigalian) based on radiometric dating of two volcanic ash layers positioned respectively 1.6 m below the contact between the Chilcatay and overlying Pisco Formation (40 Ar / 39 Ar: 18.02 ± 0.07 Ma) and 2 m above the base of the exposed section (40 Ar / 39 Ar: 19.0 ± 0.28 Ma) (Di Celma et al. 2018; Bosio et al. 2020 b) (Fig. 1 D). This age is consistent with the 87 Sr / 86 Sr stratigraphy and silicoflagellate and diatom biostratigraphy (Di Celma et al. 2018; Bosio et al. 2020 a, 2022). REFERRED SPECIMEN. — MUSM 4692, nearly complete cranium including a few teeth in situ with the associated partial right and left mandibles, the atlas, and two sternum elements. This specimen was reported in the geological map of Zamaca provided by Di Celma et al. (2019) with the field number ZM 12 and referred to Chilcacetus cavirhinus. Geographic coordinates: 14 ° 38 ’ 20.2 ” S, 75 ° 38 ’ 26.1 ” W. Zamaca is a locality a few kilometers south of Ullujaya where the same Burdigalian Chilcatay Formation is extensively exposed (Fig. 1 B). In detail, the MUSM 4692 skeleton was collected 10.7 m above the contact with the underlying Otuma Formation, near the top of the of the Ct 1 c facies association (sandstones and conglomerate beds) of the Ct 1 allomember (Fig. 1 E). The age of Ct 1 c is well defined by two dates: 1) a 19.25 ± 0.05 Ma 40 Ar / 39 Ar radiometric dating of a volcanic ash layers near the base of Ct 1 c (Di Celma et al. 2018); and 2) a 19.1 - 18.7 Ma 87 Sr / 86 Sr dating obtained with the Strontium Isotope Stratigraphy analyzing oyster shells collected a few meters below the MUSM 4692 skeleton (Bosio et al. 2020 a, 2022). Therefore, the age of MUSM 4692 is roughly 19 Ma (lower Burdigalian).	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C6273BFF91C9759D94FD907EB8.taxon	diagnosis	EMENDED DIAGNOSIS. — Chilcacetus cavirhinus differs from Chilcacetus ullujayensis n. sp. in its slightly smaller size (bizygomatic width <260 mm); the rostrum being proportionally shorter (ratio between rostrum length and condylobasal length ≤ 0.7 and ratio between preorbital width and rostrum length> 0.4); 34 maxillary teeth per side; the mesorostral groove being dorsally open for a shorter distance (dorsomedial contact between premaxillae> 190 mm long); the vertex of the cranium being proportionally anteroposteriorly longer and narrower; the occipital shield being transversely broader (minimum posterior distance between temporal fossae clearly higher than width of premaxillae in facial region); the long axis of the zygomatic process of the squamosal being closer to the horizontal plane; and the higher number of alveoli (10) in the post-symphyseal region of the mandible.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C6273BFF91C9759D94FD907EB8.taxon	discussion	COMMENTS The specimen MUSM 1401 was originally referred to Chilcacetus cavirhinus, but several differences with the holotype were already noted (Lambert et al. 2015). With the addition of two new Chilcacetus specimens in the sample, some of these differences were observed between two subsets of specimens, leading to the proposal that two species of Chilcacetus are recorded in the Chilcatay Formation. Interestingly, the two specimens referred to Chilcacetus ullujayensis n. sp. originate from close levels of the Ct 1 allomember of the Chilcatay Formation, both in the facies association Ct 1 a, while C. cavirhinus MUSM 4692 was collected in a lower level, from facies association Ct 1 c (Fig. 1). The latter can thus be considered geologically older. These considerations support the hypothesis that the two Chilcacetus species were not sympatric, with C. cavirhinus preceding C. ullujayensis n. sp. Such scenario would require that the holotype of C. cavirhinus, whose exact locality and stratigraphic provenance are unknown, also came from the oldest layers of the Chilcatay Formation exposed in the Ullujaya-Zamaca area.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C6273BFF91C9759D94FD907EB8.taxon	description	BRIEF DESCRIPTION This description focuses on the skull parts of the holotype MNHN. F. PRU 11 recently re-discovered at the MNHN, as well as on some diagnostic regions of the newly referred specimen MUSM 4692. Ontogenetic stage Both MNHN. F. PRU 11 and MUSM 4692 have robust cranial and mandibular bones, and neither display detached cranial elements or broadly open cranial sutures. Altogether, these observations point to subadult to adult individuals. Cranium The addition of the anterior part of the rostrum to the cranium of the holotype (MNHN. F. PRU 11; Fig. 2) allows for the measurement of the condylobasal length (cbl) and rostrum length (rl) (Table 1). With a rl / cbl ratio of 0.7, corresponding to a longirostrine condition (seeMcCurry & Pyenson 2019; Lambert & Goolaerts 2022), the holotype has a rostrum proportionally slightly shorter than in Chilcacetus ullujayensis n. sp. MUSM 2527 (0.73) and MUSM 1401 (0.72). With a ratio estimated to 0.68, MUSM 4692 (Fig. 3) has a similarly shorter rostrum, as confirmed by the ratio between postorbital width and rl (0.48 and 0.49 for the holotype and MUSM 4692, respectively), which is greater than in MUSM 2527 (0.42). On the complete rostrum of the holotype, 41 alveoli are counted on the left side, a number that falls in the estimated range of C. ullujayensis n. sp. The premaxilla-maxilla suture is visible on both lateral surfaces of the holotype’s rostrum, reaching the lateral margin of the alveolar groove at 65 mm from the tip of the rostrum. The premaxillary part thus makes 12.5 % of the rostrum length, a ratio that is close to that of MUSM 1401 (see Lambert et al. 2015). Seven well-defined alveoli are counted on each side of this premaxillary portion. In both the holotype and MUSM 4692, the mesorostral groove is dorsally open on a shorter distance in the anterior part of the rostrum (Figs 2; 3; Appendix 1) when compared to specimens of C. ullujayensis n. sp. Previously observed in the holotype, the relatively long region of contact between the two premaxillae above the mesorostral groove is confirmed in MUSM 4692 (at least 200 mm). As for the holotype, the bony nares of MUSM 4692 are narrower than in C. ullujayensis n. sp. MUSM 2527 and MUSM 1401. The posterior apex of each premaxilla reaches farther along the nasal in MUSM 4692 compared to MUSM 1401, but no clear difference is noted with MUSM 2527. Due to the incompleteness of the anterior part of the nasals and the complex topology of the nasal-frontal sutures (see below), proportions of the vertex in MUSM 4692 cannot be precisely quantified. Still, it was originally proportionally anteroposteriorly longer than in C. ullujayensis n. sp. MUSM 2527 and MUSM 1401, with longer nasals, as in the holotype. Sutures between nasals and frontals appear to be interdigitated in MUSM 4692, with many longitudinal grooves and ridges, as in the holotype. The interpretation of the outline of these sutures is complex, as several lines are observed that may correspond to the suture lines (see Fig. 3 B), a condition reminiscent of the holotype of Yaquinacetus meadi Lambert, Godfrey & Fitzgerald, 2019, where two different interpretations were provided for the nasal-frontal suture outline (Lambert et al. 2019). The challenging interpretation of vertex sutures in various groups of odontocetes could be partly due to the complex evolutionary and developmental history of the bones making this region (Roston et al. 2023). Noteworthily, several ziphiids have been described with an unusual configuration of vertex bones, sometimes showing extreme ossification (Bianucci et al. 2013) and in other cases including a prominent medial bone identified as the interparietal (e. g., Bianucci et al. 2007). The upper part of the occipital shield of MUSM 4692 is transversely and dorsoventrally concave. The temporal crests are robust and project far posteriorly, more so than in C. ullujayensis n. sp. MUSM 2527. Mandible With the anterior tip now added to the previously described posterior portion, the right mandible of the holotype is preserved on its whole length (Fig. 2 A, B, D). The mandibular condyle is shifted about 30 mm anterior to its original position in the mandibular fossa, and the anterior tip of the mandible reaches about 40 mm more anterior than the rostrum. The rostrum was thus not longer than the mandible, a major difference with the eurhinodelphinids for which a mandible was found associated to the rostrum (e. g., Kellogg 1925; Pilleri 1985). Thirty-eight alveoli are counted on the nearly complete left mandible of the holotype (Fig. 4); the total count was most likely close to the upper count. As in the holotype, the mandible of MUSM 4692 bears 10 alveoli in the post-symphyseal region, a number that is higher than in Chilcacetus ullujayensis n. sp. MUSM 2527. The symphysis is not ankylosed in MUSM 4692, a feature shared with the holotype. Atlas Preserved upside down attached to the left side of the occipital shield, the atlas of MUSM 4692 is 118 mm wide, 73 mm high, and 30.5 mm long on the ventral side along the sagittal plane (Fig. 3). There is no indication of an ankylosis, even partial, with the axis, a difference with most, but not all, ziphiids (Ramassamy et al. 2018). The lower transverse process is moderately elongated posterolaterally. The base of the ventral tubercle is broad, but this process only extends for a short distance posterior to the ventral edge of the posterior articular facets. The ventral tubercle is for example longer in the eurhinodelphinids Xiphiacetus cristatus and Ziphiodelphis sigmoideus Pilleri, 1985 (Pilleri 1985; Lambert 2005 a) and several early platanistoids (Tanaka & Fordyce 2015; Kimura & Barnes 2016; Bianucci et al. 2020).	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62736FF8EC91F9BBEFD6C7BBA.taxon	description	(Figs 5 - 15) urn: lsid: zoobank. org: act: 7 CFE 27 F 3 - EDC 1 - 4572 - B 5 DD-F 1662483258 D	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62736FF8EC91F9BBEFD6C7BBA.taxon	materials_examined	TYPE MATERIAL. — Holotype. Peru • 1 specimen (nearly complete cranium including the two in situ periotics, the detached left malleus, left tympanic bulla, and several teeth, the posterior part of mandibles, the left humerus, three thoracic vertebrae, and a few rib fragments; this specimen has been previously attributed to Chilcacetus cavirhinus, in a paper including a field photo of the specimen, as well as photos of one rib and the humerus (Bianucci et al. 2018: figs 8, 13 )); East Pisco Basin, Ullujaya locality; 14 ° 34 ’ 2.2 ” S, 75 ° 38 ’ 54.8 ” W; Chilcatay Formation; Burdigalian (late Early Miocene); MUSM 2527. TYPE LOCALITY. — Ullujaya, East Pisco Basin, Peru (Fig. 1 A, B). Geographic coordinates: 14 ° 34 ’ 2.2 ” S, 75 ° 38 ’ 54.8 ” W. This specimen was reported in the geological map of Ullujaya provided by Di Celma et al. (2018) with the field number O 5. TYPE HORIZON AND AGE. — Lower Miocene Chilcatay Formation. More precisely, the holotype MUSM 2527 was collected 9.7 m above the base of the exposed section in the sandstones of the Ct 1 c facies association of the Ct 1 allomember (Fig. 1 E). The age of the type horizon is constrained to the 18.9 - 18.5 Ma interval based on the 87 Sr / 86 Sr values calculated from barnacle and pectinid samples collected just below the holotype skeleton (Bosio et al. 2020 a, 2022). This dating is consistent with the roughly 19 - 18 Ma age interval for the whole Chilcatay Formation exposed at Ullujaya as indicated by 40 Ar / 39 Ar radiometric dating, 87 Sr / 86 Sr stratigraphy, and silicoflagellate and diatom biostratigraphy (Di Celma et al. 2018; Bosio et al. 2020 a, b, 2022). REFERRED SPECIMEN. — MUSM 1401, nearly complete cranium with associated fragments of both mandibles, eight partly preserved vertebrae, and rib fragments. This specimen was previously tentatively attributed to Chilcacetus cavirhinus (Lambert et al. 2015). It was reported in the geological map of Ullujaya provided by Di Celma et al. (2018) with the field number O 6 (see also Bianucci et al. 2018: table 1, fig. 13). Geographic coordinates: 14 ° 34 ’ 36.00 ” S, 75 ° 38 ’ 38.30 ” W. MUSM 1401 was collected 15 m above the base of the exposed section and 5.3 m above the level of the holotype of Chilcacetus ullujayensis n. sp., in the same sandstones of the Ct 1 c facies association of the Ct 1 allomember of the Lower Miocene Chilcatay Formation (Fig. 1 D). Two meters above MUSM 1401, sampled barnacles and oyster shells gave a 87 Sr / 86 Sr dating between 18.8 Ma and 18.6 Ma (Bosio et al. 2020 a, 2022). Therefore, both the holotype and referred specimen of C. ullujayensis n. sp. fall in the 18.9 - 18.4 age interval, i. e. lower Burdigalian.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62736FF8EC91F9BBEFD6C7BBA.taxon	diagnosis	DIAGNOSIS. — Chilcacetus ullujayensis n. sp. differs from Chilcacetus cavirhinus in its slightly greater size (bizygomatic width> 270 mm); the rostrum being proportionally longer (ratio between rostrum length and condylobasal length> 0.71 and ratio between preorbital width and rostrum length <0.4); 37 - 39 maxillary teeth per side; the mesorostral groove being dorsally open for a longer distance (dorsomedial contact between premaxillae absent to very short); the vertex of the cranium being proportionally anteroposteriorly shorter and wider; the occipital shield being transversely narrower (minimum posterior distance between temporal fossae approximates maximum width of premaxillae in facial region); the long axis of the zygomatic process of the squamosal being farther to the horizontal plane; and the lower number of alveoli (8) in the post-symphyseal region of the mandible.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62736FF8EC91F9BBEFD6C7BBA.taxon	etymology	ETYMOLOGY. — The species name refers to the East Pisco Basin fossil-rich locality of Ullujaya, where both the holotype MUSM 2527 and referred specimen MUSM 1401 of this new species were collected.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62736FF8EC91F9BBEFD6C7BBA.taxon	description	DESCRIPTION This description is mostly based on the holotype MUSM 2527, focusing on parts that are not preserved or differently shaped in the previously described MUSM 1401 (Lambert et al. 2015; at that time referred to Chilcacetus cavirhinus). Ontogenetic stage All epiphyses of the three preserved thoracic vertebrae and left humerus of MUSM 2527 are fully ankylosed, indicating an adult individual (Galatius & Kinze 2003; Moran et al. 2015). A similar interpretation was proposed for MUSM 1401, based on a series of 8 vertebrae (Lambert et al. 2015). This is in agreement with the robustness of cranial and mandibular parts in these two specimens. General cranial features Displaying a complete rostrum and well-preserved facial region as in MUSM 1401, the cranium of MUSM 2527 is slightly more damaged on the roof of the temporal fossae, and its temporal crests are lost (Figs 5 - 11; Appendix 2). Compared to MUSM 1401 (Appendix 3), it has better-preserved orbits and ventral region in general, including well-defined maxillary and premaxillary alveoli, and complete hamular processes and jugals. Its in situ periotics as well as one detached tympanic bulla, one malleus, and teeth, are other parts not preserved in MUSM 1401 (and any other Chilcacetus spp. specimen with regards to the ear bones). Apart from a somewhat longer rostrum, cranial dimensions of MUSM 2527 are remarkably similar to MUSM 1401 (Table 1). The robust rostrum makes 73 % of the condylobasal length, a proportion that is slightly higher than in MUSM 1401, and the postorbital width is greater than in the two specimens of Chilcacetus cavirhinus. As in other Chilcacetus spp. specimens, the temporal fossa is anteroposteriorly longer than high, and its posterior part was not fully dorsally covered by the frontal and maxilla. The vertex is only slightly shifted to the left side compared to the sagittal plane, indicating a minor degree of asymmetry. Premaxilla As for Chilcacetus cavirhinus, the premaxilla is alone on the anterior tip of the rostrum for a relatively short distance, about 50 mm (Fig. 6 A). In MUSM 2527, this premaxillary portion held 4 to 5 teeth, and the corresponding alveoli have an anteroposterior diameter of 9 mm and interalveolar spaces of 2.5 mm (Fig. 7). The premaxilla-maxilla suture is distinct until its tip on the lateral surface of the rostrum. More posteriorly, this suture is not located in a broad lateral groove, though a narrow sulcus occupies the intermediate third of the rostrum, starting from a foramen that is located along the suture. As in all other Chilcacetus spp. specimens, the dorsolateral surface of the premaxilla is marked along the anterior half of the rostrum by numerous shallow and sinuous sulci, leaving from a main longitudinal sulcus that start at the tip of the rostrum (Fig. 5). Right and left premaxillae contact each other above the mesorostral groove from a point at mid rostrum length, before diverging 200 mm anterior to the antorbital notch level. The premaxillary foramen is located at the level of this notch. The anteromedial sulcus is moderately long (about 80 mm) and the posteromedial sulcus is deep and broad; these two sulci define a roughly flat, dorsomedially sloping prenarial triangle. Each premaxillary sac fossa is transversely and anteroposteriorly concave. The left premaxillary sac fossa is only slightly wider than the right, while the opposite condition occurs in MUSM 1401, so no clear signal can be detected for the asymmetry of this region in Chilcacetus ullujayensis n. sp., as in C. cavirhinus MNHN. F. PRU 11. No prominence is seen on the posterolateral corner of the thick ascending process of the premaxilla. The premaxilla contacts the anterolateral and lateral margin of the nasal for most of the length of the latter. The presence of a contact between premaxilla and frontal could not be assessed due to uncertainties about the lateral extent of the nasal-frontal suture on both sides of the vertex. Maxilla Five and four dorsal infraorbital foramina are observed in the area anteromedial to the antorbital notch of MUSM 2527, on the right and left sides, respectively (Fig. 5). This count does not differ much from MUSM 1401. At least one additional foramen is found on the left side, 35 mm posteromedial to the notch. Only the right posteriormost foramen is preserved, level to the posterior edge of the bony nares and 18 mm lateral to the premaxilla-maxilla suture. Better preserved in MUSM 2527 than in MUSM 1401, the antorbital notches are deep and wide, slightly more anterolaterally open than in Chilcacetus cavirhinus MNHN. F. PRU 11. As in the latter, the right notch is narrower than the left, due to a more anteriorly projected antorbital process, reaching a level 12 mm anterior to the preorbital process of the frontal. The dorsal surface of the right antorbital process makes an oblique, prominent blade as in MNHN. F. PRU 11. Along the vertex, the raised medial wall of the maxilla is distinctly shorter than in MUSM 1401 and, to an even greater extent, than in the two specimens of C. cavirhinus, corresponding to an anteroposteriorly shorter posterior part of the vertex. With a total count of 42 to 43 upper alveoli per row, MUSM 2527 had 37 to 39 maxillary teeth per row (Fig. 7), a count that is higher than in C. cavirhinus MNHN. F. PRU 11 (34). The alveolar groove ends 60 mm anterior to the level of the antorbital notch, and the last alveoli are shifted slightly medially compared to the lateral margin of the rostrum. Most maxillary alveoli have an anteroposterior diameter of 9 to 10 mm, with a slight decrease in posteriormost alveoli. Interalveolar spaces range from 2.5 mm anteriorly to 7 mm at mid rostrum length, 11 mm at 3 / 4 of rostrum length, and 3 - 4 mm between posteriormost alveoli. 170 mm from the rostrum tip, the vomerine trough widens and deepens backwards with a V-shaped cross section, until 420 mm from tip, where it starts shallowing. A first pair of palatine foramina is located 500 mm from rostrum tip, each followed anteriorly by a well-defined sulcus. A second pair of smaller foramina is observed 15 mm more posteriorly, and a third pair is along the anteromedial border of the palatines, at 70 - 80 mm from the level of the antorbital notches. Palatine and pterygoid The anteromedial part of the palatine-maxilla suture is visible on both sides of the deeply keeled rostrum base, until a level 25 mm posterior to the last pair of palatine foramina (Fig. 7). The pterygoid is much shorter anteriorly, not reaching the level of the antorbital notch. The anteriorly short pterygoid sinus fossa is laterally defined by a low lateral lamina of the pterygoid. This lamina is directed posterolaterally, but remains low before disappearing anterior to the path for the mandibular nerve V 3. This lamina does not contact the falciform process of the squamosal, a condition already proposed for Chilcacetus cavirhinus MNHN. F. PRU 11 and, for example, also observed in Squaloziphius emlongi Muizon, 1991 and Waipatia maerewhenua Fordyce, 1994 (Muizon 1991; Fordyce 1994). The medial wall of the pterygoid thickens both transversely and dorsoventrally in posterior direction, more gradually than in C. cavirhinus MNHN. F. PRU 11, making a robust hamular process with a ventrolateral edge that is keeled anteriorly and more rounded posteriorly (Figs 6 - 8). Right and left processes are separated by a deep V-shaped valley. The main body of the hamular process is distinguished from its long, finger-like posteromedial projection by a broad notch that is ventral to the choana. Right and left projections do not contact each other (8 mm apart), and are 9.5 mm broad and 25 mm long. They are broadly similar, though shorter, than in ‘ Argyrocetus ’ joaquinensis, Simocetus rayi Fordyce, 2002, and Olympicetus thalassodon Velez-Juarbe, 2023 (Kellogg 1932; Fordyce 2002; Velez-Juarbe 2023), similar but more distant to each other than in Ziphiodelphis abeli (Pilleri 1985; plates 39, 40), longer than in C. cavirhinus, and more posteriorly directed than in S. emlongi. The medial lamina of the pterygoid thickens gradually towards the posterior region of the basioccipital basin, ending 60 mm from the posterior tip of the basioccipital crest. Nasal Though the anteromedial margin of each nasal is abraded in MUSM 2527, considering the better-preserved anterolateral margins the anterior outline of the nasals did most likely not differ much from MUSM 1401, displaying a smoothly rounded outline and a short projection above the bony nares (Figs 5; 6; 11 B). The dorsal surface of the joined nasals is transversely convex. The nasal-frontal sutures are difficult to follow laterally, but they indicate that each nasal is transversely wider than medially long and that medially each nasal is shorter than the corresponding frontal. As in MUSM 1401, each nasal sends a posterolateral projection, whose full extent cannot be assessed here. There is no indication of interdigitated nasal-frontal sutures in both MUSM 2527 and MUSM 1401, a possible difference with Chilcacetus cavirhinus. Frontal On the vertex, the anteromedial part of the frontals is slightly prominent, as in C. cavirhinus MNHN. F. PRU 11, reaching dorsally the same height as the highest preserved medial portions of nasals. From there, the dorsal surface of the frontals slopes posteroventrally to a narrow depression just anterior to the nuchal crest. In lateral view (Fig. 6), the preorbital process is moderately thickened, as in Chilcacetus cavirhinus. The orbit roof is highly arched, with a maximum depth of 21 mm on the right side. Better preserved on the left side, the postorbital process is robust and roughly vertical. Lacrimojugal complex The lacrimojugal complex makes the ventrolateral wall of the antorbital notch, as a thin oblique plate along the anteroventral margin of the preorbital process of the frontal (Figs 7; 8), with no major difference with C. cavirhinus MNHN. F. PRU 11. The base of the styliform process of the jugal extends at least 15 mm more anteromedially than the bottom of the antorbital notch. The styliform process remains thin (transverse thickness 3 mm) for most of its length, only expanding a short distance before the contact with the zygomatic process of the squamosal (better seen on the left side). Supraoccipital The nuchal crest is thin anteromedially, but it thickens considerably towards the dorsolateral corners of the occipital shield (Figs 5; 11 A), as in C. cavirhinus MNHN. F. PRU 11 and Macrodelphinus kelloggi, reaching a maximum thickness of 15 mm, measured on the left side, which is greater than in MUSM 1401. In posterior view, most of the outline of the nuchal crest is horizontal, as in MUSM 1401 and C. cavirhinus MNHN. F. PRU 11. The occipital shield is moderately transversely concave and its dorsal region is slightly dorsoventrally concave, while more ventrally a broad sagittal groove is present. The ventrolateral parts of the occipital shield being lost, the cranial endocast is visible, with the outer surface better preserved on the left side. Exoccipital The exoccipital does not reach as far laterally as the squamosal (Fig. 11 A). The region of the paroccipital process is too damaged on both sides to allow for a description of this area, as well as of the details of the narrow jugular notch. The occipital condyles are slightly mediolaterally narrower and dorsoventrally higher than in Chilcacetus cavirhinus MNHN. F. PRU 11, with a transversely narrower foramen magnum. The condylar neck is well defined (Figs 6; 7), and the dorsal, condyloid fossa is broad and deep (Fig. 11 A). Alisphenoid The anterior portion of the ventral exposure of the alisphenoid is excavated by the broadest part of the pterygoid sinus fossa, the latter being laterally margined by a moderately developed subtemporal crest (Figs 7; 8). Posterior to the sinus fossa, the path for the mandibular nerve V 3 extends obliquely from the foramen ovale. Squamosal The zygomatic process of the squamosal is long and robust, pointing anterodorsally (Fig. 6). Its preserved position compared to the postorbital process of the frontal on both sides suggests either a long, oblique contact between the two bones, or a near contact. The dorsal margin of the zygomatic process is moderately convex in lateral view, and this margin thickens transversely posteriorly. The contact facet with the jugal is best seen on the left side, as a 23 mm long concave portion of the ventral margin, defined posteroventrally by a low tuberosity. From this point, the concave ventrolateral margin of the zygomatic process remains thin until the base of the robust postglenoid process. Not preserved in C. cavirhinus MNHN. F. PRU 11 and MUSM 1401, the latter is anteroposteriorly long (29 mm from the anterior meatal crest to its anterior margin) and transversely thick in MUSM 2527, similar to the condition in eurhinodelphinids, Eoplatanista spp., Amphidelphis bakersfieldensis n. comb., ‘ Argyrocetus ’ joaquinensis, Caolodelphis milleri Godfrey & Lambert, 2023, and Perditicetus yaconensis Nelson & Uhen, 2020, but anteroposteriorly shorter than in squaloziphiids and possible relatives Crisocetus lydekkeri Gaetán, Paolucci & Buono, 2023, Dolgopolis kinchikafiforo Viglino, Gaetán, Cuitiño & Buono, 2021, Squaloziphius emlongi, and Yaquinacetus meadi, as well as in Enigmatocetus posidoni Godfrey & Lambert, 2023 (Muizon 1991; Lambert et al. 2019; Nelson & Uhen 2020; Viglino et al. 2021; Gaetán et al. 2023; Godfrey & Lambert 2023); it reaches ventrally the same level as the basioccipital crest and the exoccipital. One main sternocephalicus fossa is best seen on the right side, not extending much along the zygomatic process (ending before the level of the anterior margin of the postglenoid process). As in C. cavirhinus MNHN. F. PRU 11, the tympanosquamosal recess is not clearly separated from the mandibular fossa (Figs 7; 8). The recess excavates the base of the postglenoid process’ medial surface and extends anteriorly until the anterior limit of the mandibular fossa, thus not reaching farther along the zygomatic process. Probably complete on both sides of MUSM 2527, the falciform process is pointing ventrally just anterior to the periotic, but extends distally (in the ventral direction) less than in C. cavirhinus MNHN. F. PRU 11. The falciform process lowers abruptly until the posterior margin of the path for the mandibular nerve V 3, where it disappears (Figs 8; 9). Basioccipital Better preserved on the left side of MUSM 2527, each basioccipital crest is thickened along its posteromedial surface (Figs 7; 8), but not to the extent of Caolodelphis milleri. Right and left crests draw an angle of 35 ° in ventral view. A curved ridge runs from the base of each basioccipital crest towards the medial region of the basioccipital basin. Periotic Both periotics of the holotype (MUSM 2527) are preserved in situ (Figs 7 - 9), with the anterior process closely appressed to the corresponding falciform process of the squamosal and with the posterior process displaying a long dorsal contact with the exoccipital and squamosal. Anatomical details are thus only available for the ventral and ventrolateral regions. The anterior process is short (Table 1) and bears a shallow but transversely broad (4.5 to 5 mm wide) anterior bullar facet. This condition differs from the narrower facet seen in an interpretive drawing of the now lost periotic of the holotype of Chilcacetus cavirhinus MNHN. F. PRU 11 (Lambert et al. 2015). Part of a well-defined parabullary sulcus can be seen on the medial surface of the anterior process of the left periotic (Fig. 9), along the falciform process of the squamosal. It is much closer to the anterior bullar facet than in Yaquinacetus meadi, and is more similar to Perditicetus yaconensis (Nelson & Uhen 2020), though the latter shares with Y. meadi an anterior bullar facet that is shifted distinctly medially compared to the lateral margin of the anterior process (condition unknown in MUSM 2527, the lateral margin being hidden by the falciform process). A medium-size accessory ossicle is preserved in situ on both sides, differing from the right periotic of MNHN. F. PRU 11. The anteromedial outline of the pars cochlearis is rounded, while the posteromedial corner is marked by a prominence just medial to the fenestra rotunda. The lateral tuberosity is large, with a ventromedial surface that is broader than the mallear fossa, extending far laterodorsally and anteriorly defining a deep and narrow hiatus epitympanicus. Though being partly hidden by the posterior process of the tympanic bulla on the left side and by sediment and bone fragments on the left right side, the posterior process is tentatively interpreted as long (estimated length 35 mm), nearly reaching the posterior margin of the exoccipital in a posterolateral direction. It retains a considerable breadth for most of its visible / preserved portion, a condition that is reminiscent of the outline proposed in ventral view for MNHN. F. PRU 11. A transverse break surface tentatively observed in ventral and medial view on the left periotic of P. yaconensis (Nelson & Uhen 2020: fig. 7) may indicate a similarly elongated posterior process. The exposed proximal surface of the right posterior bullar facet is concave along an oblique axis. Tympanic bulla In ventral view, the finely preserved left tympanic of MUSM 2527 is olive-shaped, with the inner posterior prominence being slightly shorter than the outer prominence, and the latter being markedly transversely broader (Fig. 10 A-E). The ventral surface of the inner prominence is moderately keeled, and the keel extends anteriorly for 21 mm. The median furrow is wide but shallow posteriorly, vanishing anteriorly before mid-length of the tympanic. The anterior spine is incipient, barely thickened, associated with an inconspicuous anterolateral notch. In medial view the posterior part of the involucrum is 14 mm thick. After 5 mm the dorsal surface slopes gradually down to a slight indentation, followed by a second plateau before a more abrupt descent towards the anterior opening of the tympanic cavity. The medial surface of the tympanic is crossed at about mid-height by an anteroposteriorly directed ridge, presumably marking the ventral limit of the peribullary sinus. In dorsal view the sigmoid process is slightly obliquely directed, hiding the conical process in lateral view. The lateral furrow is narrow and nearly vertical. The anterodorsal crest is moderately convex in lateral view. The elliptical foramen is 2.5 mm wide and at least 4 mm high. Preserved in situ in the basicranium, the posterior process of the left tympanic lacks its thin distal part (Figs 7; 8), indicating that this process was long and sub rectangular in outline, as proposed for the posterior process of the periotic. Except for the somewhat more swollen outer posterior prominence, the ventral and medial views of this tympanic are highly similar to these views in Chilcacetus cavirhinus MNHN. F. PRU 11. Strong similarities are also noted with Yaquinacetus meadi and eurhinodelphinids, though these generally display a deeper and anteriorly longer median furrow (see Muizon 1988 a; Lambert 2005 a, b). An even shallower, nearly indistinct median furrow is seen in Eoplatanista spp. (Muizon 1988 a). Malleus The left malleus of USNM 2527 displays an articular head that makes 70 % of the height of the bone in posteromedial view, corresponding to a markedly shortened tubercule (Fig. 10 F), as seen in eurhinodelphinids, Inticetus Lambert, Muizon, Malinverno, Di Celma, Urbina & Bianucci, 2017, Eoplatanista, Yaquinacetus, physeteroids, and ziphiids (Muizon 1988 a; Lambert 2005 b; Bianucci et al. 2010; Lambert et al. 2018, 2019). On the tubercule, which is longer than in physeteroids and most ziphiids, the manubrium is approximately at the same height as the low muscular process; the manubrium is considerably less prominent than in eurhinodelphinids and Inticetus, closer to Eoplatanista and Yaquinacetus. The anterior facet for the incus does not project outside the articular head, drawing a general outline in posteromedial view that is more regularly rounded than in Eoplatanista italica MGP 26166 (Muizon 1988 a: fig. 12). Mandible As the mandible of Chilcacetus ullujayensis n. sp. is poorly preserved in MUSM 1401 (see Lambert et al. 2015), our description is mostly based on MUSM 2527 (Fig. 12). The latter only preserves a short (110 mm) posterior part of the symphyseal region. The anterior cross section is semi-circular to subtriangular, with no indication of a lateral groove, as in other Chilcacetus spp. specimens. While the two dentaries remain attached to each other, the suture line is visible dorsally, ventrally, and in cross section, indicating that no strong ankylosis occurred along the symphysis. At the level of the posterior end of the symphysis, alveoli have an anteroposterior diameter of 10 mm, with interalveolar spaces of 8 to 10 mm. This is in the range of MUSM 1401, with alveoli larger than in Chilcacetus cavirhinus MNHN. F. PRU 11 and MUSM 4692. Eight alveoli are posterior to the symphysis; the last alveolus is 135 mm posterior to the end of the symphysis and 255 mm anterior to the posterior end of the mandibular condyle. There is no precoronoid crest and the dorsal margin of the coronoid crest flattens and widens in the posterior direction, with thin longitudinal ridges along the medial and lateral edges of the process. Neither of the coronoid processes are complete in MUSM 2527. The angular process draws an angle of about 90 °. The mandibular condyle is laterally convex and slightly medially concave, with a dorsoventral height that is slightly greater than in C. cavirhinus MNHN. F. PRU 11. The condyle is in line with most of the alveolar groove, the last alveolus being roughly at the same dorsoventral level as the upper edge of the condyle. The mandibular foramen has a regularly rounded anterior outline, similar to C. cavirhinus MNHN. F. PRU 11, but ending farther anteriorly, 150 mm from the mandibular condyle. Teeth Four teeth were found detached along the skull of MUSM 2527 (Fig. 13). Total length ranges between 27 and 35 mm, with a crown length between 8.7 and 10.9 mm, a maximum diameter at the base of the crown between 4.5 and 5.7 mm, and a maximum diameter of the root between 7.2 and 9.2 mm. These dimensions are close to those of the posterior upper teeth of Chilcacetus cavirhinus MNHN. F. PRU 11. Each crown is moderately curved lingually, is slightly labio-lingually flattened, and bears barely discernible mesial and distal carinae, lacking any accessory denticle, as observed in C. cavirhinus. The enamel is thin, with a nearly smooth surface only marked by very low apicobasal ridges. The crown apex is complete, with no apical wear visible on any of the four teeth. No clear occlusion facet could be identified along the prepared surfaces (but see below for one in situ tooth). The root is somewhat inflated in all teeth, with a slight to moderate curvature along a plane perpendicular to the curvature of the crown. The surface of the root along the root-crown boundary is smoothened, possibly corresponding to the part emerging from the alveolus, where some degree of hydrowear or acid corrosion may have occurred (see Marx et al. 2023). A couple of anterior maxillary teeth are preserved in situ. One of them displays a long but shallow occlusal surface along the distal surface. Humerus Previously figured to highlight shark bite marks (Bianucci et al. 2018: fig. 13 b, c), the complete left humerus of MUSM 2527 (Fig. 14) has a total length of 149 mm, a maximum diameter of the humeral head of 62 mm, a minimum mediolateral width of the diaphysis of 36.7 mm, an anteroposterior length of the distal epiphysis of 60 mm, and a mediolateral width of the distal epiphysis of 38 mm. The large, posterolaterally facing head is prominent, with a well-defined neck. The attachment area for the supraspinatus on the greater tubercle reaches proximally as far as the head. The lesser tubercle extends slightly more proximally, but not to the extent seen in Amphidelphis bakersfieldensis n. comb. YPM 13406 (see below), the humerus YPM 13437 attributed to Macrodelphinus kelloggi, and Xiphiacetus bossi (Kellogg, 1925) USNM 8842 (Kellogg 1925: pl. 12), where the lesser tubercle extends farther proximomedially (better seen in posterior view). The diaphysis is rectilinear in lateral view, while in anterior / posterior view its long axis is not as curved as in A. bakersfieldensis n. comb., being more similar to M. kelloggi. The narrow (9 mm), proximodistally elongated (21 mm), and well-defined fossa for the infraspinatus is located 15 mm anterodistal to the head. The moderately anteriorly thickened, 20 mm mediolaterally wide deltoid tuberosity is located at about mid-length of the bone, as in A. bakersfieldensis n. comb., M. kelloggi, and Eurhinodelphis cocheteuxi IRSNB M. 299 (Abel 1931: pl. 28, fig. 12: collection number 3403). The tuberosity is located more distally in X. bossi. A distinct constriction marks the posterior margin of the diaphysis just before the distal epiphysis. Separated by a rectilinear ridge, the radial and ulnar facets are 31 and 28 mm long, respectively. The latter is followed by a distinct, 10 mm long facet for the olecranon. Thoracic vertebrae Three partial anterior thoracic vertebrae, corresponding to bicipital ribs, were found associated with MUSM 2527 (Fig. 15). They are listed from Ta to Tc based on the increase of centrum length, ranging from 48 to 58 mm (Table 2), but marked differences in the height of the transverse process, especially between Ta and Tc, suggest that intermediate thoracics could be missing. The three thoracics have a centrum that is wider than long, and while Ta has anterior and posterior heights that are roughly equal to its length, Tb and Tc are proportionally longer. Only Ta has moderately well-preserved transverse processes, projecting anterolateroventrally and each bearing a robust facet for the rib tuberculum. The articulation facets for the rib capitula on the anterior epiphysis are located slightly below mid centrum height; facets on posterior epiphysis are just above mid height. Only the pedicles, base of the neural spine, and base of transverse processes are preserved on Tb, defining a pentagonal neural canal. The posterior epiphysis of this vertebra bears facets for ribs on the dorsolateral corner of the epiphysis that are much more developed than the barely visible facets on the anterior epiphysis. On Tc, in addition to the centrum only part of the left pedicle is preserved. The facets for ribs are similarly located on the dorsolateral corner of the posterior epiphysis. Ribs As for the humerus, one of the ribs of MUSM 2527 was figured in Bianucci et al. (2018: fig. 13 a) to illustrate shark bite marks. In the same paper, a transverse thin section of a rib from that specimen was also figured, showing sediment infilling in cancellous bone (Bianucci et al. 2018: fig. 11 e).	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62729FF8EC9099EB8FE697D4C.taxon	description	urn: lsid: zoobank. org: act: 888073 F 3 - FD 61 - 422 F-BE 20 - 36 B 7 FCDF 7991	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62729FF8EC9099EB8FE697D4C.taxon	type_taxon	TYPE SPECIES. — Amphidelphis bakersfieldensis (Wilson, 1935) n. comb.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62729FF8EC9099EB8FE697D4C.taxon	diagnosis	DIAGNOSIS. — Same as for the only species, Amphidelphis bakersfieldensis n. comb.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62729FF8EC9099EB8FE697D4C.taxon	etymology	ETYMOLOGY. — From ancient Greek “ ἀμΦΊ ” (amphi, from both sides), in reference to the distribution of this taxon in both the Northern and Southern hemispheres (California and Peru), and “ delphis ”, the Latin word for dolphin.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62729FF80C8CB98D9FB397E99.taxon	description	(Figs 16 - 20)	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62729FF80C8CB98D9FB397E99.taxon	materials_examined	TYPE MATERIAL. — Holotype. United States • 1 specimen (partial cranium lacking the anterior part of the rostrum, most of the occipital shield, the right side of the basicranium, and the ear bones, with two associated cervical vertebrae and the left humerus); California, Kern County, southwest of Woody; undifferentiated sediments from Freeman Silt and Jewett Sand; Aquitanian to early Burdigalian (Early Miocene); YPM 13406. TYPE LOCALITY. — Approximately eight km southwest of the small town of Woody, Kern County, California, United States, the same locality as the holotype of Macrodelphinus kelloggi, and a short distance from the type locality of the allodelphinid Allodelphis pratti Wilson, 1935 (Wilson 1935; Kimura & Barnes 2016). TYPE HORIZON AND AGE. — Undifferentiated sediments from the Freeman Silt and Jewett Sand, 50 feet (about 15 m) lower in the section than the holotype of Allodelphis pratti, Woody Local Fauna, Early Miocene (Aquitanian to early Burdigalian), Saucesian benthic foraminifera stage, Arikareean North American Land Mammal age (Wilson 1935; Barnes 1976; Shimada et al. 2014; Kimura & Barnes 2016). NEWLY REFERRED SPECIMEN FROM THE EAST PISCO BASIN. — MUSM 4691, partial cranium lacking the occipital shield and basicranium, with associated partial right mandible. This specimen was found at Roca Negra, a locality close to Zamaca (Fig. 1 B). Geographic coordinates: 14 ° 38 ’ 59.5 ” S, 75 ° 38 ’ 51.8 ” W. This cranium was reported in the geological map of Zamaca provided by Di Celma et al. (2019) with the field number ZM 36 and referred there to Odontoceti indet. It was collected 25.9 m above the contact with the underlying Otuma Formation in the Ct 1 a facies association (sandstones and conglomerate beds) of the Ct 1 allomember of the Lower Miocene Chilcatay Formation (Fig. 1 E). The age of Ct 1 a in the Roca Negra-Zamaca areas is constrained between 19.1 and 18.3 Ma (early Burdigalian), based on a Strontium Isotope Stratigraphy analysis of oyster shells collected in the underlying Ct 1 c facies association and shark teeth in the overlying Ct 2 a facies association, respectively (Bosio et al. 2020 a, 2022).	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62729FF80C8CB98D9FB397E99.taxon	diagnosis	DIAGNOSIS. — The differential diagnosis focuses primarily on differences with taxa that were found to be closely related to Amphidelphis in our phylogenetic analysis and comparison (other members of the Chilcacetus clade, Eoplatanistidae, Eurhinodelphinidae, and Squaloziphiidae). Amphidelphis bakersfieldensis n. comb. is a small (bizygomatic width estimated at 174 mm in the holotype), longirostrine, and homodont dolphin species differing from Argyrocetus patagonicus in its smaller size, in the rostrum being proportionally considerably shorter (ratio between preorbital width and rostrum length estimated at 0.5), and lacking an extended premaxillary portion, in the dorsal opening of the mesorostral groove being narrower than the premaxilla at rostrum base, in the premaxillary foramen being roughly at the level of the antorbital notch, and in the absence of ankylosis for the symphysis of the mandibles; from ‘ Argyrocetus ’ joaquinensis in its smaller size, in the dorsal opening of the mesorostral groove being narrower than the premaxilla at rostrum base, in the presence of more than one dorsal infraorbital foramen at rostrum base, in the proportionally shorter and wider nasals, and in the nasals partly overhanging the bony nares; from Chilcacetus in its smaller size, in the rostrum being proportionally shorter, in possessing a deep sulcus anterior to the main dorsal infraorbital foramen at rostrum base, and in the palatines not being separated anteromedially for a long distance at rostrum base; from Perditicetus in its smaller size, in the premaxillary foramen being roughly at the level of the antorbital notch, and in the zygomatic process of the squamosal being dorsoventrally more slender; from Caolodelphis in its smaller size, the frontals not being separated anteromedially on the vertex, and the basioccipital crests being transversely thinner; from Macrodelphinus in its much smaller size, in the rostrum being proportionally shorter and lacking an extended premaxillary portion, in the premaxillary foramen being roughly at the level of the antorbital notch, and in the exposure of the frontals on the vertex being shorter and narrower. It differs from Crisocetus, Dolgopolis, Squaloziphius, and Yaquinacetus in the postglenoid process of the squamosal being significantly shorter anteroposteriorly, and from Dolgopolis, Squaloziphius, and Yaquinacetus in the dorsal opening of the mesorostral groove being more gradual anterior to the bony nares. It differs from most members of other longirostrine to hyper-longirostrine homodont extinct families (including Eoplatanistidae and Eurhinodelphinidae) in the absence of a deep lateral groove along most of the rostrum and in the absence of ankylosis for the symphysis of the mandibles. It further differs from Eurhinodelphinidae in lacking an extended edentulous anterior premaxillary portion of the rostrum and in the nasals partly overhanging the bony nares. It further differs from Eoplatanistidae in the premaxillary foramen being roughly at the level of the antorbital notch, in the thinner and flatter antorbital process, in the acute anterior margin of the nasal partly overhanging the bony nares, and in the less anteriorly projected supraoccipital shield.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62729FF80C8CB98D9FB397E99.taxon	discussion	REMARK ON THE TAXONOMY AND SYSTEMATICS OF AcROdeLPHIS ABEL, 1900 AND ARGYROceTUS Originally placed in the genus Acrodelphis as Acrodelphis bakersfieldensis n. comb. by Wilson (1935), this species was later referred to the genus Argyrocetus by Barnes (1976), together with the type species Argyrocetus patagonicus and the other Californian species ‘ Argyrocetus ’ joaquinensis. Later, due to the fragmentary state of the type specimens and the lack of synapomorphies for this genus, the referral of the two Californian species to this genus was questioned (Lambert 2005 a; Lambert et al. 2015). The description of the new specimen MUSM 4691 allows for the identification of major differences with A. patagonicus, further supporting the removal of ‘ Argyrocetus ’ bakersfieldensis from the genus Argyrocetus. The original genus, Acrodelphis, has been previously identified by Muizon (1988 a) as a junior synonym of Champsodelphis Gervais, 1848, and the latter has been regarded in the same work as an incertae sedis, restricted to the type specimen (a partial mandible) of its type species, Champsodelphis macrogenius (Fischer, 1829). A new genus name is thus proposed here for ‘ Argyrocetus ’ bakersfieldensis.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62729FF80C8CB98D9FB397E99.taxon	discussion	COMMENTS Despite a few minor differences with the holotype YPM 13406, including a slightly wider dorsal opening of the mesorostral groove, a relatively broader dorsal exposure of the maxilla at rostrum base, and a slightly more anterior premaxillary foramen, MUSM 4691 is morphologically similar enough to YPM 13406 to allow for its attribution to the same species, Amphidelphis bakersfieldensis n. comb. Differences between these two individuals may be explained by intraspecific variation; YPM 13406 represents for example a relatively young animal, as indicated by unfused epiphyses for two cervical vertebrae (Fig. 20 E, F; Galatius & Kinze 2003) The attribution of MUSM 4691 to A. bakersfieldensis n. comb. provides information on cranial regions that are not preserved in the holotype. The anterior portion of the rostrum is especially informative for the comparison with Argyrocetus patagonicus and eurhinodelphinids.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62727FFBACB749B9FFDC27FFE.taxon	description	(Figs 21 - 25)	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62727FFBACB749B9FFDC27FFE.taxon	materials_examined	REFERRED SPECIMENS FROM THE EAST PISCO BASIN. — MUSM 602, partial cranium lacking the anterior part of the rostrum, part of the roof of the temporal fossae, ear bones, and teeth, with the ventral surface of the basicranium and palate damaged. This specimen was collected in 1993 by M. Urbina in an undetermined level of the Lower Miocene Chilcatay Formation from the locality of Zamaca (Fig. 1 B). MUSM 4961, partial cranium lacking most of the rostrum, ear bones, and teeth, with superficial bone being either damaged or covered with a thin layer of hard sediment in many regions of the neurocranium. This specimen was collected at an unknown date by M. Urbina in an undetermined level of the Chilcatay Formation from the locality of Roca Negra (Fig. 1 B; see Lambert et al. 2018). Considering the above-mentioned dates obtained through 40 Ar / 39 Ar radiometric analyses and Strontium Isotope Stratigraphy, and in agreement with silicoflagellate and diatom biostratigraphy, the age of MUSM 602 can be restricted between 19.3 and 17.95 Ma and the age of MUSM 4961 between 19.3 and 18.3 Ma (both Burdigalian) (Lambert et al. 2018; Di Celma et al. 2019; Bosio et al. 2020 b).	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62727FFBACB749B9FFDC27FFE.taxon	description	DESCRIPTION OF REFERRED SPECIMENS Most of the description is based on the better-preserved cranium MUSM 602 (Figs 21 - 24), except for the few parts that are only preserved in MUSM 4961 (Fig. 25; Appendix 5), that differ between the two specimens, or that display shared differences with Amphidelphis bakersfieldensis n. comb. Both specimens have similar cranial dimensions, with MUSM 4961 being only slightly smaller. Except for the vertex of the cranium that is transversely broader than in YPM 13406, no marked differences are noted with specimens of A. bakersfieldensis n. comb. Premaxilla Though the rostrum of MUSM 602 appears to be slightly transversely crushed in its anterior portion, right and left premaxillae most likely contacted each other dorsomedially for at least 100 mm, before a gradual divergence of the medial margins towards the bony nares (Fig. 21). In both specimens, and to an even greater extent in MUSM 4961 (Fig. 25 A), the resulting dorsal opening of the mesorostral groove is narrower than in MUSM 4691 and YPM 13406 (and also ‘ Argyrocetus ’ joaquinensis, as well as squaloziphiids) and is not spindle-shaped. In MUSM 602, the premaxillary foramen is 15 mm posterior to the antorbital notch on the right side and 20 mm posterior on the left side. As in Amphidelphis bakersfieldensis n. comb., the anteromedial sulcus is long (110 mm on the right side and 120 mm on the left). At least in MUSM 602, the prenarial triangle is longer than in A. bakersfieldensis n. comb. The posteromedial sulcus is poorly defined, and the posterolateral sulcus is narrow, becoming nearly indiscernible before the anterior margin of the bony nares. The premaxillary sac fossae are generally less dorsally elevated in MUSM 602 (but not in MUSM 4961) than in A. bakersfieldensis n. comb.; they are transversely concave, with an anterior portion that is also longitudinally concave while the thicker posterior portion has a roughly flat surface. At the level of the anterior margin of the bony nares, the lateral margin of each premaxilla is laterally concave, with an abrupt posteromedial turn at mid length of the bony nares. Better preserved on the right side of MUSM 602, the ascending process of the premaxilla is narrower than in MUSM 4691 and YPM 13406, with parallel lateral and medial margins, a condition that is reminiscent of Squaloziphius emlongi and Yaquinacetus meadi. Posteromedially, each premaxilla broadly contacts the corresponding nasal. Due to the complex and not fully resolved outline of the nasal-frontal suture, the presence of a contact between premaxilla and frontal on the vertex cannot be assessed. Maxilla Along the rostrum base, the lateral margin of the maxilla is incomplete in MUSM 602. In MUSM 4961, this margin is directed more posterolaterally than in MUSM 4691 and YPM 13406, suggesting that both MUSM 602 and MUSM 4961 had a broader rostrum base compared to A. bakersfieldensis n. comb. In both MUSM 602 and 4961, the lateralmost portion of the dorsal surface of the maxilla is moderately thickened just anteromedial to the antorbital notch (maxillary flange). In both MUSM 602 and MUSM 4961, the antorbital notch is much more laterally open and shallow than in MUSM 4691 and YPM 13406; differing from the latter (and also Chilcacetus spp., Macrodelphinus kelloggi, Perditicetus yaconensis, Squaloziphius emlongi, and Yaquinacetus meadi), the antorbital region does not extend anterior to the bottom of the antorbital notch, in a way more similar to ‘ Argyrocetus ’ joaquinensis. The dorsal surface of the maxilla is flat in this area. In MUSM 602, four dorsal infraorbital foramina are present in the right antorbital notch region and three in the left region. Along the vertex, the dorsomedial margin of the maxilla overhangs the more ventral part of the bone. Posteriorly, the maxilla similarly raises abruptly to contribute to the acute nuchal crest. Roughly rectilinear along the vertex, the nuchal crest turns moderately posterolaterally towards the posteriormost part of the roof of the temporal fossa. Roughly complete in MUSM 4961, the posterolateral corner of the maxilla is more medial than the postorbital process in dorsal view, indicating that the posterior part of the temporal fossa remained dorsally open to an extent similar to YPM 13406 (at least for the right side of the latter). In ventral view, a few small maxillary alveoli are visible in the posterior portion of the alveolar groove (Fig. 23 A). The first posterior alveolus is 50 mm anterior to the antorbital notch, but it cannot be excluded that a few more posterior alveoli have been abraded. The mean diameter of the preserved alveoli is 3 mm and interalveolar septa are short, 1 - 2 mm long, a condition that does not differ substantially from Amphidelphis bakersfieldensis n. comb. Vomer In MUSM 602, the vomer is exposed in the vomerine trough from a level 62 mm anterior to the antorbital notch (Fig. 23 A). Lacrimojugal complex Better preserved on the right side of MUSM 602, the lacrimojugal complex sends a thin plate laterally, anteroventral to the preorbital process of the frontal (Fig. 23). A similar plate is instead directed anterolaterally in Amphidelphis bakersfieldensis n. comb. This plate thickens dorsoventrally in its lateralmost portion, up to 7 mm. Palatine / pterygoid The pointed anterior end of the joined palatines is 23 mm anterior to the level of the antorbital notch in MUSM 602. The palatine-maxilla suture gradually diverges posterolaterally for about 20 mm before abruptly turning towards the ventral infraorbital foramen. No parts of the pterygoid could be identified on any of the two specimens, except for part of the medial lamina along the pharyngeal crest. The pterygoid sinus fossa is anteriorly short, with an anterior margin 25 mm posterior to the level of the antorbital notch, a feature that is similarly seen in Amphidelphis bakersfieldensis n. comb., Chilcacetus spp., Macrodelphinus kelloggi, Perditicetus yaconensis, and Yaquinacetus meadi; the fossa is somewhat anteriorly longer in Squaloziphius emlongi and, to a greater extent, Dolgopolis kinchikafiforo. Nasal As in Amphidelphis bakersfieldensis n. comb., the top of the cranium is made by the nasals, with a dorsal surface sloping posteroventrally (Figs 21; 22; 24 A). Their anterodorsal portion is damaged in MUSM 602; the break surface suggests that each nasal originally projected anterodorsally, as seen in YPM 13406, overhanging the bony nares for more than 4 mm, in a way similar to Chilcacetus spp. The outline of the nasal-frontal suture is complex and difficult to interpret, due to a series of deep grooves, as in YPM 13406, but also Chilcacetus cavirhinus and Yaquinacetus meadi. The nasals were either as long or longer than the frontals on the vertex. As preserved, the left nasal is wider than long. Frontal As in YPM 13406, the dorsal surface of the frontals on the vertex of MUSM 602 and MUSM 4961 is irregular, marked by a series of meandering grooves, pits, and crests, a feature also seen for example in Chilcacetus cavirhinus and Macrodelphinus kelloggi. As in Amphidelphis bakersfieldensis n. comb., the preorbital process of the frontal is dorsally exposed. In lateral view (Fig. 22), it is only moderately thickened and anteroventrally margined by the lacrimal. Better preserved on the left side of MUSM 602 and right side of MUSM 4961, the postorbital process is robust and directed ventrally. Except for the right side of MUSM 602, the posterior surface of the postorbital process roughly contacts the similarly robust anterior margin of the zygomatic process of the squamosal. A slight degree of post-mortem deformation may have either separated the two bones on the right side of MUSM 602 or, less likely, brought these closer to each other on the left side of the latter and in MUSM 4961. In ventral view, the infratemporal crest is distinct from the tip of the postorbital process to the posterior wall of the frontal groove, as seen in YPM 13406. Supraoccipital The dorsomedial part of the occipital shield is anteroporsteriorly concave, this depressed region being anteriorly defined by the high posterior wall of the nuchal crest (Figs 21; 22; 24 A). From the nuchal crest, the surface of the shield leaves in a nearly horizontal posterior direction, as in YPM 13406, before becoming markedly anteroposteriorly convex. A broad, longitudinal groove separates two posteriorly swollen regions corresponding to the right and left brain hemispheres. Only partly preserved in both specimens, the temporal crest was originally short but distinct, and directed posterolateroventrally. Exoccipital The occipital condyles protrude posterior to the rest of the exoccipital, with a well-defined condylar neck and moderately excavated dorsal condyloid fossae (Figs 22; 24 B). The posterior surface of the exoccipital slopes more posteroventrally in MUSM 602 compared to MUSM 4961 and YPM 13406, but this may be the result of some post-mortem deformation. The jugular notch is too incomplete on the two specimens to be compared to the deep notch observed in Amphidelphis bakersfieldensis n. comb. YPM 13406. Basioccipital Only the dorsal part of the basioccipital crests is preserved in both specimens, defining a basioccipital basin that is narrow anteriorly (38 mm at about the posterior tip of the pterygoids in MUSM 602), with a gradual posterior widening (distance between the two crests reaching 45 mm at their posterior end). The medial surface of each crest is marked by a ridge that turns anterodorsally and dorsomedially. Squamosal The outline of the zygomatic process of the squamosal varies from one side to the other and from one specimen to the other (Figs 22; 25 B). It is more slender, anteroposterioly elongated on the right side of MUSM 602 and markedly dorsoventrally thicker (22 mm) and shorter on the right side of MUSM 4961, with intermediary conditions on the left side of MUSM 602 and left side of MUSM 4961. The preserved right zygomatic process of Amphidelphis bakersfieldensis n. comb. YPM 13406 (Fig. 18 A) is more similar to the left side of MUSM 4961. The anterior margin of the process is anteriorly concave in lateral view, with a shallow transverse groove at mid-height possibly corresponding to the surface of contact with the postorbital process of the frontal. The shape of the dorsal margin of the zygomatic process varies also markedly, from roughly rectilinear on the right side of MUSM 602 and left side of MUSM 4961 to concavo-convex on the other sides of these two specimens, with a specially dorsally bulging posterior part on the left side of MUSM 4961. Such a variation may be related to a highly variable development of the sternocephalicus fossa, which is much deeper and anteriorly extended on the right side of MUSM 4961 and, probably, the left side of MUSM 602. Such an extent of morphological variation between the two sides of a single specimen should urge to remain cautious with the use of zygomatic process characters to diagnose new taxa. On the right side of MUSM 602, the sternocephalicus fossa is subdivided in two fossae by a thick oblique crest, a condition that is similarly observed in YPM 13406. Only preserved in MUSM 4961 (Fig. 25 B), the postglenoid process is ventrally long, reaching much farther than the posttympanic process, and slightly anteroposteriorly longer (more than 19.5 mm) than transversely thick (18 mm) on the left side. This process appears only slightly anteroposteriorly longer than in YPM 13406 and does not differ much from ‘ Argyrocetus ’ joaquinensis, Caolodelphis milleri, Chilcacetus ullujayensis n. sp., and Perditicetus yaconensis. Based on the preserved proximal part, similar proportions of the postglenoid process can be proposed for MUSM 602. As for the corresponding zygomatic process, the right postglenoid process of MUSM 4961 displays unusual proportions, forming an oblique plate that is 10 mm thick and 22 mm long. It cannot be excluded that such a highly asymmetric condition is pathological. In ventral view, the tip of the ventral surface of the zygomatic process displays a circular articulation surface for the jugal, with an anteroposterior diameter of 12 mm in MUSM 602 (Fig. 23). The mandibular fossa is broad (24 mm on the right side) and medially margined by a shallow tympanosquamosal recess. Separated from the mandibular fossa by a well-defined step, the latter extends anterolaterally for a short distance beyond the level of the anterior margin of the squamosal fossa, in a way similar to YPM 13406. The falciform process is not adequately preserved in any of the two specimens. The squamosal fossa is shorter in MUSM 4961 compared to MUSM 602, lacking the anterior part with an anteroposteriorly convex floor seen in the latter.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
276487C62727FFBACB749B9FFDC27FFE.taxon	discussion	COMMENTS These two crania are highly similar in both their dimensions and main cranial features. Several differences were noted above, and those mostly focus on the squamosal (zygomatic process, sternocephalicus fossa, and postglenoid process). Because strong differences are also noted between the left and right sides of one specimen for this region, their diagnostic value can be reasonably questioned. Therefore, it is proposed that MUSM 602 and MUSM 4961 belong to the same species. As they share several differences with Amphidelphis bakersfieldensis n. comb. MUSM 4691 and YPM 13406, including the outline of the rostrum base in dorsal view, the shape of the antorbital notch, and the extent of the dorsal opening of the mesorostral groove at rostrum base, it cannot be excluded that they correspond to another, closely related species. Furthermore, the lack of more anterior sections of the rostrum prevents from investigating the relative anterior extent of the premaxilla and maxilla, a key character to test for eurhinodelphinid affinities. Pending the discovery of more complete and better-preserved specimens to provide diagnostic characters, it is proposed to identify these two specimens as Odontoceti aff. A. bakersfieldensis n. comb.	en	Lambert, Olivier, Muizon, Christian de, Bennion, Rebecca F., Urbina, Mario, Bianucci, Giovanni (2025): New data on archaic homodont odontocetes from the Early Miocene of Peru reveal a second species of Chilcacetus Lambert, Muizon & Bianucci, 2015 and a Southern Hemisphere record for a northeastern Pacific species. Geodiversitas 47 (9): 369-408, DOI: 10.5252/geodiversitas2025v47a9, URL: https://sciencepress.mnhn.fr/sites/default/files/articles/pdf/geodiversitas2025v47a9.pdf
