Kogiids are known by two living species, the pygmy and dwarf sperm whale (Kogia breviceps and K. sima). Both are relatively rare, and as their names suggest, they are closely related to the sperm whale, all being characterized by the presence of a spermaceti organ. However, this organ is much reduced in kogiids and may have become functionally different. Here we describe a fossil kogiid from the late Miocene of Panama and we explore the evolutionary history of the group with special attention to this evolutionary reduction. The fossil consists of cranial material from the late Tortonian (~7.5 Ma) Piña facies of the Chagres Formation in Panama. Detailed comparison with other fossil and extant kogiids and the results of a phylogenetic analysis place the Panamanian kogiid, herein named Nanokogia isthmia gen. et sp. nov., as a taxon most closely related to Praekogia cedrosensis from the Messinian (~6 Ma) of Baja California and to Kogia spp. Furthermore our results show that reduction of the spermaceti organ has occurred iteratively in kogiids, once in Thalassocetus antwerpiensis in the early-middle Miocene, and more recently in Kogia spp. Additionally, we estimate the divergence between extant species of Kogia at around the late Pliocene, later than previously predicted by molecular estimates. Finally, comparison of Nanokogia with the coeval Scaphokogia cochlearis from Peru shows that these two species display a greater morphological disparity between them than that observed between the extant members of the group. We hypothesize that this reflects differences in feeding ecologies of the two species, with Nanokogia being more similar to extant Kogia. Nanokogia shows that kogiids have been part of the Neotropical marine mammal communities at least since the late Miocene, and gives us insight into the evolutionary history and origins of one of the rarest groups of living whales.
Citation: Velez-Juarbe J, Wood AR, De Gracia C, Hendy AJW (2015) Evolutionary Patterns among Living and Fossil Kogiid Sperm Whales: Evidence from the Neogene of Central America. PLoS ONE 10(4): e0123909. https://doi.org/10.1371/journal.pone.0123909
Academic Editor: Laurent Viriot, Team 'Evo-Devo of Vertebrate Dentition', FRANCE
Received: September 25, 2014; Accepted: February 23, 2015; Published: April 29, 2015
Copyright: © 2015 Velez-Juarbe et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: Funding for this project was provided by National Science Foundation Partnerships for International Research and Education grant #0966884 to JVJ, ARW, CDG, and AJWH; National Science Foundation Earth Sciences Postdoctoral Fellowship grant #1249920 to JVJ; and Secretaria Nacional de Ciencia, Tecnología en Innovación grant #APY-NI10-016A to CDG. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Kogiidae is a family of toothed whales represented by two extant species, the pygmy (Kogia breviceps (Blainville )) and the dwarf (K. sima (Owen )) sperm whales, and having a worldwide distribution from temperate to tropical regions [3–4]. However, they are some of the rarest whales, and hence little is known of their life history . This is also true of their fossil record, which is relatively sparse, with a total of five species described so far: Thalassocetus antwerpiensis Abel , from the early-middle Miocene of Belgium ; Scaphokogia cochlearis Muizon , from the late Miocene (Tortonian) of Peru [9–10]; Praekogia cedrosensis Barnes , from the late Miocene (Messinian) of Baja California; Aprixokogia kelloggi Whitmore and Kaltenbach , from the early Pliocene (Zanclean) of North Carolina; and Kogia pusilla (Pilleri ), from the late Pliocene (Piacenzian) of Italy . In addition to these, there are a number of additional Neogene records of kogiids based on family-level diagnostic elements such as isolated earbones (e.g. ), showing that the group was already widespread throughout the Neogene in subtropical to temperate regions. Nonetheless, we still know very little about their ancient diversity and distribution due to the scarcity of cranial material described, as well as the limited taxonomic information provided by other elements.
Here we describe the first fossil kogiid from the Central American and Caribbean region, based on cranial material recovered from the late Miocene (latest Tortonian) Chagres Formation [16–17] on the Caribbean coast of Panama (Fig 1). The Chagres kogiid represents a new taxon closely related to Praekogia cedrosensis and Kogia spp. The other known Neotropic fossil kogiid is Scaphokogia cochlearis from similar-aged (Tortonian) deposits in Peru. These two occurrences show that kogiids have been established in the Neotropics at least since that time. However, the new Panamanian taxon and Scaphokogia cochlearis display much greater morphological disparity between them, relative to what is seen in extant Kogia spp., clearly showing that we are far from fully understanding deep-time diversity in kogiids. The new taxon from Panama is added to an increasing list of fossil marine mammals from the region, and it is a step further towards a better understanding of deep-time diversity of Neotropical marine mammals.
A, map of Central America showing the localities. B, map of central Panama, showing the distribution of the Chagres Formation as well as the localities of fossil cetaceans mentioned in the text. C, chronostratigraphic and lithostratigraphic relationships of the Chagres Formation (modified from Hendy et al. ).
Materials and Methods
Measurements of the skulls and mandible (Tables 1 and 2) follow those outlined by Perrin , while the morphological terminology follows Mead and Fordyce . The chronostratigraphy used here follows that of Cohen et al. . For the phylogenetic analysis we followed Lambert et al.  and throughout the text we have cross-referenced the morphology of specimens with the corresponding character states for ease of comparison and clarity, e.g. (c. 6) refers to state 0 of character 6.
Permits for fieldwork in Piña and other parts of Panama were obtained from the Dirección de Recursos Minerales de Panamá. Fieldwork only involved geological and paleontological sampling and collecting; it did not involve endangered or protected species. The specific coordinates of the field sites are provided in the Results section. Export of specimens was possible through a permit issued by the Ministerio de Comercio e Industria de la República de Panamá to the Smithsonian Tropical Research Institute.
We compared the Panamanian material with specimens from the mammalogy collections at the Natural History Museum of Los Angeles County (LACM), Florida Museum of Natural History (UF) and National Museum of Natural History (USNM) for the following taxa: Kogia breviceps (LACM 27082, LACM 95745, UF 13562, UF 14213, UF 14214, UF 17532, UF 18702, UF 18704, UF 19128, UF 25545; USNM 504902, USNM 504921), and Kogia sima (LACM 47142, LACM 95817, UF 18705, UF 18706, UF 24629, UF 25573, UF 25575–25578). We also made comparisons with fossil taxa from the following collections: Muséum National d’Histoire Naturelle (MNHN) for Scaphokogia cochlearis (MNHN PPI 229); Museo de Historia Natural de la Universidad Nacional Mayor de San Marcos (MUSM) for Livyatan melvillei (MUSM 1676); Natural History Museum of Los Angeles County and National Museum of Natural History for Aprixokogia kelloggi (LACM 117744 [cast] and USNM 187015), and Museum of Paleontology, University of California (UCMP) for Praekogia cedrosensis (UCMP 315229).
The electronic edition of this article conforms to the requirements of the amended International Code of Zoological Nomenclature, and hence the new names contained herein are available under that Code from the electronic edition of this article. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix "http://zoobank.org/". The LSID for this publication is: urn:lsid:zoobank.org:pub:ADA83315-CCAC-4E6C-B7D2-90BC77D2F044. The electronic edition of this work was published in a journal with an ISSN, and has been archived and is available from the following digital repositories: PubMed Central, LOCKSS.
Mammalia Linnaeus, 1758 
Cetacea Brisson, 1762 
Odontoceti Flower, 1867 
Pan-Physeteroidea new clade name
Physeteroidea Gray, 1821 
Kogiidae Gill, 1871 
Nanokogia gen. nov. urn:lsid:zoobank.org:act:D2BEAD8F-D882-44F3-85EE-6BE6E601AEF7
The name ‘Nanokogia’ is formed by the combination of ‘nano’, from the Latin “nanus” which means dwarf, in reference to the small size of the skull and estimated total length of the body, relative to most other kogiids, with ‘Kogia’, fem., which is the genus name of the extant members of the group and a widely used suffix for other fossil taxa of the group.
Late Miocene (latest Tortonian-early Messinian ) of Panama.
Same as that for the type species until other species are described.
Nanokogia isthmia sp. nov. urn:lsid:zoobank.org:act:25FB8526-74BA-4A03-A80A-8DECD9DB2234
UF 280000, nearly complete adult skull and mandible; missing teeth, ear bones, and a portion of the left side of the cranium, and part of the right horizontal ramus. Collected by J. Velez-Juarbe, December 12, 2013.
The specific name ‘isthmia’, derives from the Latin ‘isthmus’ in reference to the Isthmus of Panama.
UF 273554, adult skull, missing rostrum and part of the right side of the cranium; Chagres Formation, Piña 2 (UF locality YPA087), about 100 m northeast of mouth of Quebrada La Toba, Piña, Colón Province, Panama (09.29338°N, 80.03772°W) (Fig 1B); collected by C. De Gracia, July 5, 2012.
Small kogiids, with an estimated body length of ~1.95–2.16 m (based on equation for Physeteroidea from Pyenson and Sponberg ), similar in size to Kogia sima . Recognized as kogiid based on: bizygomatic width of less than 40 cm (c. 8); presence of a sagittal crest; c. 14); external nares greatly asymmetric; c. 18) and located at the level of the supraorbital processes; absence of nasals (c. 19); and right maxilla reaching the sagittal plane of the skull on the posterior wall of the supracranial basin. Differs from all other kogiids by the following combination of characters: absence of upper teeth, shared with Scaphokogia cochlearis and Kogia spp., unknown in Praekogia cedrosensis and Thalassocetus antwerpiensis; antorbital notches form a narrow slit (c. 9), shared with Scaphokogia, Praekogia, and Kogia; postorbital process overhanging the zygomatic process, shared with Thalassocetus antwerpiensis and Kogia; presphenoid not covered ventrally by the vomer, shared with Aprixokogia kelloggi, Scaphokogia, and Kogia; left premaxilla not reaching the sagittal facial crest, shared with Aprixokogia, Scaphokogia and Kogia. Shares with Praekogia and Kogia: antorbital notches within the supracranial basin (c. 10); long postglenoid process of the squamosal (c. 28); and, a wide notch on the squamosal for the enlarged posterior process of the tympanic (c. 29). Shares with Aprixokogia an elongated temporal fossa (c. 26) in contrast to the anteroposteriorly-shortened fossa of Kogia, or the rounded fossa in Praekogia. Shares with Praekogia relatively mediolaterally-thin maxillary crests; and wide but relatively shallow supracranial basin, shared with Praekogia and Thalassocetus, unlike the deeper basin seen in Aprixokogia and Kogia, or the broad, rounded basin of Scaphokogia. Shares with Kogia the presence of a narrow, rounded notch between the hamular process and the medial lamina of the pterygoid (absent in Aprixokogia, unknown in Thalassocetus, Scaphokogia and Praekogia). Differs further from Kogia sima and K. breviceps in having a relatively longer rostrum, similar to K. pusilla. Diagnosed by the following autapomorphies: in lateral view the posterodorsal corner of the lacrimal + jugal is not deeply wedged between the frontal and maxilla (c. 23); the lateral edge of the frontal portion of the right premaxilla is convex, not forming a crest; and the nuchal crest and posterolateral edges of the supracranial basin overhang the occipital surface of the cranium in dorsal, lateral and posterior views, but not to the extreme seen in Scaphokogia.
Description of the skull is based on both the holotype (UF 280000) and the referred specimen (UF 273554) (Figs 2–9, 10–14 and S1–S2 Figs). Because in the temporal and occipital regions the relationships between the different bones are not clear we treat these as separate subdivisions below in the description instead of the individual bones. The holotype is a nearly complete skull, missing parts of the basicranium and parts of the posterolateral surface of the skull. The referred specimen is missing the rostrum, the right supraorbital process and the right half of the supracranial basin. The skull of Nanokogia is small (c. 8) (Table 1), asymmetric, with a short rostrum (c. 1) that tapers distally, not gradually as in Kogia, and has a marked constriction at about the middle of its length (Figs 2 and 3). The distal end of the rostrum is squared off, not pointed as in Kogia, nor cylindrical as in Scaphokogia, nor rounded as in Aprixokogia. Sutures are fused on both skulls, indicating that they belonged to adult individuals.
Abbreviations: adif, anterior dorsal infraorbital foramen; an, antorbital notch; en, external nares; et, ethmoid; fr, frontal; la+j, lacrimal + jugal; lmc, lateral maxillary crest; mrg, mesorostral groove; mx, maxilla; pmx, premaxilla; pdif, posterior dorsal infraorbital foramen; scf, supracranial fossa; sfc, sagittal facial crest; sq, squamosal; vo, vomer. Gray shaded areas indicate sediment; diagonal lines denote broken surfaces. Gray shaded areas indicate sediment; diagonal lines denote broken surfaces.
Abbreviations: an, antorbital notch; bo/bs, basioccipital/basisphenoid; boc, basioccipital crest; cb.e, cerebral endocast; eam, external auditory meatus; fg, frontal groove; in, internal nares; jn, jugular notch; la+j, lacrimal + jugal; ma, mandible; mx, maxilla; npp, notch for posterior process of tympanic; oc, occipital condyles; pl, palatine; prs, presphenoid; ptdl, dorsal lamina of pterygoid; ptha, pterygoid hamulus; ptml, medial lamina of pterygoid; ptsf, pterygoid sinus fossa; tsr, tympanosquamosal recess; vif, ventral infraorbital foramen; vo, vomer. Gray shaded areas indicate sediment; diagonal lines denote broken surfaces.
Abbreviations: amc, anterior meatal crest; as, alisphenoid; bo, basioccipital; boc, basioccipital crest; dif, dorsal infraorbital foramen; eam, external auditory meatus; en, external nares; fg, frontal groove; fo, foramen ovale; fp, falciform process of squamosal; fr, frontal; jf, jugular foramen; jn, jugular notch; la+j, lacrimal + jugal; lmc, lateral maxillary crest; mx, maxilla; oc, occipital condyles; pmx, premaxilla; pop, postorbital process; prp, preorbital process; prs, presphenoid; ptdl, dorsal lamina of pterygoid; ptml, medial lamina of pterygoid; ptsf, pterygoid sinus fossa; scb, supracranial basin; so, supraoccipital; sq, squamosal; tsr, tympanosquamosal recess; vif, ventral infraorbital foramen; vo, vomer. Gray shaded areas indicate sediment; diagonal lines denote broken surfaces.
Abbreviations: amc, anterior meatal crest; as, alisphenoid; bo/bs, basioccipital/basisphenoid; boc, basioccipital crest; eo, exoccipital; fg, frontal groove; fm, foramen magnum; fo, foramen ovale; fp, falciform process of squamosal; fr, frontal; jf, jugular foramen; la+j, lacrimal + jugal; mx, maxilla; oc, occipital condyle; of, optic foramen; pgp, postglenoid process; pop, postorbital process; prs, presphenoid; ptdl, dorsal lamina of pterygoid; ptml, medial lamina of pterygoid; ptsf, pterygoid sinus fossa; sop, supraorbital process; sp, spiny process; sq, squamosal; tsr, tympanosquamosal recess; vif, ventral infraorbital foramen; vo, vomer; zp, zygomatic process. Gray shaded areas indicate sediment; diagonal lines denote broken surfaces.
Abbreviations: eo, exoccipital; fg, frontal groove; fr, frontal; la+j, lacrimal + jugal; ma, mandible; mx, maxilla; lmc, lateral maxillary crest; npp, notch for posterior process of tympanic; of, optic foramen; pa, parietal; pgp, postglenoid process; pmx, premaxilla; ptdl, dorsal lamina of pterygoid; ptha, pterygoid hamulus; ptml, medial lamina of pterygoid; ptn, pterygoid notch; ptsf, pterygoid sinus fossa; scb, supracranial basin; sfc, sagittal facial crest; sq, squamosal; tc, temporal crest; zp, zygomatic process. Gray shaded areas indicate sediment; diagonal lines denote broken surfaces.
Abbreviations: boc, basioccipital crest; eam, external auditory meatus; eo, exoccipital; fr, frontal; lmc, lateral maxillary crest; mx, maxilla; npp, notch for posterior process of tympanic; oc, occipital condyle; of, optic foramen; pa, parietal; pmx, premaxilla; pop, postorbital process; prp, preorbital process; prs, presphenoid; ptdl, dorsal lamina of pterygoid; ptml, medial lamina of pterygoid; scb, supracranial basin; smc, supramastoid crest; sq, squamosal; tc, temporal crest; vo, vomer; zp, zygomatic process. Gray shaded areas indicate sediment; diagonal lines denote broken surfaces.
Abbreviations: boc, basioccipital crest; eo, exoccipital; fm, foramen magnum; jn, jugular notch; oc, occipital condyles; so, supraoccipital; sop, supraorbital process of frontal; tc, temporal crest. Gray shaded areas indicate sediment; diagonal lines denote broken surfaces.
Abbreviations: al, tooth alveoli; ar, alveolar row; gn, gnathion; mf, mandibular fossa; ms, mandibular symphysis. Diagonal lines denote broken surfaces.
Strict consensus tree resulting from three most parsimonious trees, 95 steps long, with CI = 0.589 and RI = 0.723. Arcs indicate stem-based taxa, while circles denote node-based clades; numbers below nodes indicate decay indices/bootstrap values; for definition of clades see text. Abbreviations: Aq., Aquitanian; Bar., Bartonian; Burd., Burdigalian; Chatt., Chattian; La., Langhian; M., Messinian; P., Piacenzian; Plio., Pliocene; Pleist., Pleistocene; Priab., Priabonian; Rupel., Rupelian; S., Serravallian; Tort., Tortonian; Z, Zanclean. (Time scale based on Cohen et al. .)
Aprixokogia kelloggi (USNM 187015), 11A, Scaphokogia cochlearis (MNHN PPI 229), 11B, Praekogia cedrosensis (UCMP 315229), 11C, Nanokogia isthmia gen. et sp. nov. (based on UF 280000 and 273554), 11D, Kogia sima (LACM 47142), 11E, and, K. breviceps (LACM 95745), 11F. Each bone is color-coded for ease of comparison. Red dashed lines denote the extent of the supracranial/premaxillary fossa. Areas in white are not preserved and have been reconstructed with plaster on the specimens. Illustration of S. cochlearis modified from Muizon ; all other illustrations based on the specimens listed.
Aprixokogia kelloggi (USNM 187015), 12A, Scaphokogia cochlearis (MNHN PPI 229), 12B, Praekogia cedrosensis (UCMP 315229), 12C, Nanokogia isthmia gen. et sp. nov. (UF 273554), 12D, and (UF 280000), 12E, Kogia sima (LACM 47142), 12F, and K. breviceps (LACM 95745), 12G. Each bone is color-coded for ease of comparison. Areas in white are reconstructed, light gray areas are covered with sediment; diagonal lines denote broken surfaces. Illustrations based on direct observations of the specimens listed.
Aprixokogia kelloggi (left side, reversed, USNM 187015), 13A, Scaphokogia cochlearis (MNHN PPI 229), 13B, Praekogia cedrosensis (UCMP 315229), 13C, Nanokogia isthmia gen. et sp. nov. (based on UF 280000 and 273554), 13D, Kogia sima (LACM 47142), 13E, and K. breviceps (LACM 95745), 13F. Each bone is color-coded for ease of comparison. Areas in white are reconstructed; diagonal lines denote broken surfaces. Illustrations based on direct observations of the specimens listed.
Aprixokogia kelloggi (USNM 187015), 14A, Scaphokogia cochlearis (MNHN PPI 229), 14B, Praekogia cedrosensis (UCMP 315229), 14C, Nanokogia isthmia gen. et sp. nov. (UF 280000), 14D, Kogia sima (LACM 47142), 14E, and K. breviceps (LACM 95745), 14F; each bone is color-coded for ease of comparison. Red dashed lines denote the extent of the supracranial/premaxillary fossa. Areas in white are reconstructed, light gray areas are covered with sediment; diagonal lines denote broken surfaces.
The lateral margins of the premaxillae are nearly parallel throughout the length of the rostrum and seem to have reached the anterior tip (c. 2). The dorsal surface along the rostrum is flat to slightly convex. There are no teeth in the premaxillae, as in Kogia and Scaphokogia (c. 7). There are no foramina on the left premaxilla (c. 15), and none seems to be present on the right one, but this could be due to the poor preservation of the surface. Posteriorly, the left premaxilla curves posteromedially as a tongue-like projection, forming the anterior, lateral and posterior margins of the left naris, thus resembling Aprixokogia and Kogia spp.; in contrast, the left premaxilla of Praekogia cedrosensis is expanded posteriorly, reaching and forming part of the sagittal facial crest [11–12] (Figs 2, 11, 14 and S2 Fig). The right premaxilla forms the anterior, lateral and posterior margins of the right external naris. The external nares are greatly asymmetric as in all pan-physeteroids (c. 18), and the nasals are absent as in all known kogiids (c. 19). The frontal portion of the right premaxilla expands laterally, medially and posteriorly where it joins the left maxilla, together forming the sagittal facial crest (c. 12, 14). The dorsal surface of the right premaxilla is flat to convex, thus lacking the central premaxillary fossa seen in Thalassocetus and Kogia (Lambert :figs 16–17) (Figs 11, 14 and S2 Fig). The lateral edge of the right premaxilla is convex at midlevel of the sagittal facial crest, differing from the concave margin of Scaphokogia and Praekogia, or the more flange-like margin observed in Aprixokogia, Thalassocetus, or Kogia spp. The posterior half of the sagittal facial crest seems to be composed only of the right premaxilla as in Praekogia and differing from Kogia where both maxilla and premaxilla form the crest (Figs 2, 11, 14 and S2 Fig). The sagittal facial crest overhangs the left side of the supracranial basin and tapers posteriorly (c. 13), nearly reaching the nuchal crest, similar to the condition in Praekogia cedrosensis (Figs 11, 14 and S2 Fig). Overall, the outline of the sagittal facial crest of Nanokogia is very similar to what is observed in Praekogia and Kogia. However, it differs in that the lateral expansion of the right premaxilla at the beginning of the crest is convex, not concave as in Praekogia, nor does it overhang laterally over the right side of the supracranial basin as in Kogia (Figs 11, 14 and S2 Fig).
On the rostrum, the maxilla reaches the distal tip, together with the premaxilla and the vomer (c. 2). When viewed dorsally, the lateral borders of the rostral portion of the maxilla taper anteriorly with a marked constriction at about mid-length, its width diminishing anteriorly and becoming narrower in dorsal view than the premaxilla anteriorly (c. 4, 5). Along the rostrum, the dorsal surface of the maxilla is flat to convex, especially in the area anteromedial and medial to the antorbital notches. The antorbital notches are deep, narrow slits (c. 9) entering the supracranial basin (c. 10), as in Praekogia and Kogia. Posterior to the antorbital notches, the maxillae are greatly expanded over the cranial roof, covering nearly all of the frontals, and forming the supracranial basin (c. 3, 21) (Figs 2, 11, 14 and S2 Fig). The outline of the supracranial basin is oval, similar to that of Praekogia and contrasting with the more rounded outline of Kogia. The lateral maxillary crests are high, with rounded to sharp, medially recurved dorsal margins that slightly overhang the basin, similar to Praekogia and differing from the more inflated crests of Kogia and the thinner crests of Aprixokogia. The lateral maxillary crests of Nanokogia reach a maximum height and width of about 20 mm at the level of the antorbital notches; posteriorly the crests diminish in height and thickness (reaching a minimum of ~ 5 mm). Posterolaterally the maxillary crests overhang the temporal fossa and occipital region (Figs 2, 4, 6 and 7). Medially, the left maxilla joins the right premaxilla and forms part of the anterior half of the sagittal facial crest as in Praekogia  (Figs 2, 11C, 11D, 14C, 14D and S2C–S2D Fig); this condition contrasts with that of Kogia where the maxilla forms part of the sagittal facial crest throughout its length  (Figs 11E, 11F, 14E, 14F and S2 Fig). On the surface of the supracranial basin, posterolateral to the right external naris, there is a large, kidney-shaped fossa, herein termed supracranial fossa (= premaxillary fossa of Barnes ). The deepest part of this fossa is towards its anteromedial edge, where it reaches a maximum depth of about 2 cm; posteriorly it becomes shallower. A similar fossa is present in Praekogia cedrosensis and Aprixokogia kelloggi. However, in Praekogia the fossa is floored by both the premaxilla and maxilla, whereas in Nanokogia, as in Aprixokogia, it seems to be formed solely by the maxilla. The supracranial fossa of Aprixokogia differs from Nanokogia and Praekogia in that it extends further posteromedially (Figs 11, 14 and S2 Fig). On the right maxilla of Nanokogia there are at least three small, dorsal infraorbital foramina (c. 11). The single anterior dorsal infraorbital foramen is located anteromedial to the antorbital notch, is rounded, and has a diameter of about 5 mm. There are also at least two posterior dorsal infraorbital foramina. The largest of these is located posteromedial to the antorbital notch, is rounded, and has a diameter of about 6 mm. The second foramen is located near the anterolateral edge of the supracranial fossa, has a smaller diameter (<5 mm), and continues posterolaterally as a shallow groove along the edge of the fossa. A similar foramen is also observed in Praekogia cedrosensis  (Figs 2 and 11C). The left maxilla has at least two posterior dorsal infraorbital foramina (Figs 2 and 14D). Both foramina are located posterolateral to the external nares; the anteriormost is the smallest (<5 mm in diameter) and opens anteromedially whereas the second is larger (~7 mm in diameter) and opens dorsally. On the left maxilla of the referred specimen, UF 273554, a single dorsal infraorbital foramen is preserved. It is located posterodorsal to the left external naris, opens laterally, and is anteroposteriorly longer than dorsoventrally wide (~1.5 cm long by 0.5 cm wide (Fig 4A and 4B).
The palatal surface of the maxilla is flat to gently convex. There are no maxillary teeth and only faint indications of a vestigial upper alveolar groove (Figs 3 and 12E; c. 6). The ventrolateral edges are flange-like (maxillary flange of Mead and Fordyce ) with their ventral surfaces transversely concave anteromedial to the antorbital notches. The anteriormost extension of the pterygoid sinus is represented by a shallow, oval concave fossa located anteromedial to the ventral infraorbital foramen (Figs 3 and 6), similar to the condition observed in Kogia spp. . The infraorbital foramen is located far anteromedial to the frontal groove (Figs 3–5), being at the level of the middle of the supraorbital process of the frontal; the foramen seems to be bounded dorsally and medially by the maxilla, laterally and ventrally by the lacrimal, and posteriorly by the frontal (Figs 3, 4C, 4D and 5).
The palatines are convex, located medial to the maxillae, but their borders are not well defined (Fig 3).
Lacrimal + Jugal.
The lacrimal and jugal are fused as in all physeteroids for which these elements are known (c. 22); they are large (~ 5 cm long) and with a triangular outline in lateral view (Fig 5). The ventral tip of this bone is rounded (diameter of ~20 mm) and ends in a blunt tip. In lateral view the posterodorsal corner of the lacrimal + jugal of Nanokogia is not deeply wedged between the frontal and maxilla, in contrast to Scaphokogia, Praekogia and Kogia (c. 23) (Muizon :fig. 35; Fig 13). In ventral view the lacrimal contacts the anteromedial surface of the frontal, and its posterior end tapers and curves posteromedially (Fig 4C and 4D).
With the exception of the supraorbital processes, nearly all of the dorsal surfaces of the frontals are covered by the maxillae (Figs 2, 4, 11D, 14D and S2 Fig). In lateral view the frontal-maxilla suture forms an angle of about 30° relative to the coronal plane (Fig 5; c. 25), similar to most other fossil kogiids but differing from Kogia where the angle is greater than 35° (Fig 13; c. 25). The lateral margin of the supraorbital process is oriented parasagittally in its anterior portion, whereas the posterior part, including the postorbital process, is projected posterolateroventrally. The lateral surface between the pre- and postorbital process is laterally concave (Fig 4A and 4B). The preorbital process is at about the same dorsoventral level as the dorsolateral margin of the base of the rostrum (Figs 6 and 13D; c. 24). The postorbital process slightly overhangs the zygomatic process of the squamosal and is separated from the anterodorsal margin of the zygomatic process by about 1 cm (Figs 6 and 13D). The overhang of the postorbital processes is also seen in Thalassocetus and more extremely in Kogia, whereas there is no overhang in Aprixokogia and Praekogia (Fig 13). The anterior margin of the preorbital process is rounded and blunt in contrast to the sharper ventral margin of the postorbital process. On the ventral surface of the supraorbital process there is a long (~5.5 cm) frontal groove (Figs 3–5). The groove is oriented anterolaterally at a greater angle than that observed in Kogia spp. and leads proximally to the optic foramen (~1 cm in diameter) (Figs 5–7).
The temporal crest is sharp along its posteroventral border; it curves anterodorsally towards the supraorbital processes as a lower, much less prominent crest (Figs 6 and 7). Posterodorsally, the frontal (and/or parietal) and maxilla form a shelf that laterally overhangs the temporal wall and nuchal crest and posteriorly the occipital region to a greater degree than what is observed in Kogia spp. and other fossil kogiids (Figs 7 and 12). The temporal fossa is similar to that in Aprixokogia kelloggi in being anteroposteriorly elongated in outline in lateral view (c. 26), contrasting with the anteroposteriorly-shortened fossa seen in Kogia spp. (Fig 13).
In dorsal view, the rostral part of the vomer seems to have reached the anterior end of the rostrum and forms the floor and lateral walls of the mesorostral groove (Fig 2). Ventrally, it is exposed along the anterior half of the rostrum as a narrow sliver (Fig 3); however, the sutures are not well preserved. Posteriorly, the vomer is divided into a pair of processes that cover the lateral, but not the ventral, surfaces of the presphenoid (Figs 3, 4C and 4D). This is similar to the condition observed in Physeter, Aprixokogia, Scaphokogia, and Kogia spp.  but differs from the condition observed in Praekogia cedrosensis, where the vomer covers the ventral surface of the bone .
Ethmoid and Presphenoid.
The ethmoid and presphenoid form the bony septum that, along with the vomer, form the medial walls of the internal and external nares. The external nares are greatly asymmetric, with the left naris being more than twice as wide as the right one (c. 18). Posteriorly, the presphenoid widens and contacts the basisphenoid/basioccipital (Figs 3, 4C and 4D).
The pterygoid is long. The hamuli of the right and left pterygoids meet along a midline suture; when viewed ventrally they have a triangular outline. There is a narrow notch with a rounded terminus between the hamular process and the medial lamina as in Kogia. The medial lamina (= vaginal processes of Whitmore and Kaltenbach ) forms the lateral and posteroventral walls of the internal nares. The pterygoid extends posteriorly as a thin (<5 mm wide) lamina whose ventral edge is inflected medially. Anteriorly, left and right medial laminae are nearly straight and parallel, diverging posterolaterally near the level of the presphenoid-basisphenoid suture, to eventually meet the basioccipital crest (Figs 3, 4C and 4D). The dorsal lamina of the pterygoid is anteroposteriorly shorter (~4 cm long) than the medial lamina ~10 cm), and are oriented perpendicular to each other. The dorsal laminae floor the anteromedial and medial surface of the optic foramen as in Kogia spp. (Figs 3, 4C, 4D, 5, 7C and 7D).
The alisphenoid forms the posteromedial edge of the frontal groove. The ventral surface of the bone is concave, and together with the medial lamina of the pterygoid it forms part of the pterygoid sinus fossa (Fig 4C and 4D). Near the posteromedial border of the bone is the foramen ovale, which is oriented laterally. The foramen is about 7 mm long anteroposteriorly long and 5 mm wide dorsoventrally, and seems to be roofed by the squamosal.
The zygomatic process is triangular in cross section (c. 27 ). The glenoid fossa is gently concave and oriented anteromedially. The postglenoid process forms the anterior meatal crest and projects farther ventrally than the paroccipital process of the exoccipital in the holotype (c. 28; Fig 6). Medially, the anterior meatal crest, terminates in a short spiny process (Figs 4C, 4D and 5). Posterior to the postglenoid process there is a transverse notch representing the vestigial external auditory meatus (Figs 3, 4C, 4D, 7A and 7B). Medial to the postglenoid process and mandibular fossa is the tympanosquamosal recess (Figs 4 and 5); this area forms a fossa which likely housed the middle ear air sinus as in other kogiids in which this feature is preserved . Further along the posteromedial border of the squamosal is the short (~7 mm high) and ventrally-oriented falciform process (c. 32; Fig 5). The squamosal fossa is mediolaterally concave and slopes anteriorly. Along the dorsolateral border of the zygomatic process, the supramastoid crest is sharp, and joins the temporal crest (= lambdoid crest of Whitmore and Kaltenbach ) posteromedially (Fig 7A and 7B). The posteroventral margin of the squamosal is posteriorly concave, with the exoccipital being exposed posterolaterally to a greater degree than is observed in P. cedrosensis (Fig 7). The posteroventral surface of the squamosal and the anteroventral surface of the exoccipital form an anteroventrally-oriented, deep, round notch (c. 29; Figs 3, 6, 7 and 13), more similar to what is observed in Praekogia cedrosensis than to Kogia spp. In Kogia spp. this notch accommodates the enlarged posterior process of the tympanic (not preserved here), and it is inferred that both Nanokogia and Praekogia had tympanics with enlarged posterior processes.
The surface of the occipital region is convex and smooth, forming an angle of about 60° with the long axis of the skull (c. 30, 31; Figs 6 and 8). The foramen magnum is roughly circular, and the occipital condyles are dorsoventrally elongated with rugose articular surfaces. The dorsal condyloid fossae are shallow as in Kogia and the condyles are separated ventrally by a shallow intercondylar notch. Relative to the occipital condyles, the rostrum is oriented anteroventrally (Fig 13). The paroccipital processes are oriented posterolaterally and are posteriorly concave. The jugular notch is a ~5 mm deep incisure, located between the paroccipital process of the exoccipital and the basioccipital crest. The broken left squamosal and occipital region of the holotype reveal a structure that is interpreted as the cerebral endocast (Fig 3).
Only the mandible of UF 280000 is preserved, although missing parts of the horizontal ramus posterior to the mandibular fossa on the left side and most of the horizontal ramus posterior to the symphysis on the right (Fig 9). The mandibular symphysis is long and fused, with its ventral border straight and keeled (Fig 9, S1 Fig). The alveolar row extends posteriorly to near the anterior margin of the mandibular fossa, and the alveoli are round and oriented dorsolaterally. Posterior to the mandibular symphysis, the lateral alveolar margin overhangs the external surface of the bone. The left ramus preserves alveoli for about 14 small round teeth (c. 37, 39, 40). Posterior to the symphysis, the medial surface of the horizontal ramus is convex, becoming flatter towards the mandibular foramen. The mandibular fossa is dorsoventrally broad as in other odontocetes, with a V-shaped apex forming its anterior margin. In lateral view, the gnathion marks an abrupt change in the orientation of the ventral surface and the end of the symphysis; the latter is relatively long compared to Kogia breviceps and Kogia sima (S1 Fig). Ventrally, towards the proximal end of the symphysis there is a pair of oval depressions, which would correspond to the attachment sites of the geniohyoid muscles in Kogia . In dorsal view the mandibular rami gradually diverge posterolaterally. The preserved portion of the mandible is about 121 mm long, the symphysis is 101 mm long, and the tooth alveoli are around 10 mm in diameter.
In order to determine the relationships between Nanokogia and other fossil and extant kogiids, we performed a phylogenetic analysis using the character-state matrix for physeteroids of Lambert et al. . The analysis includes two outgroup and 16 ingroup taxa. In contrast to Lambert et al.  we treated all characters as unordered, we re-scored some characters for Livyatan melvillei based on personal observations of MUSM 1676, Thalassocetus antwerpiensis based on the description by Lambert , and Praekogia cedrosensis based on personal observations of UCMP 315229 (formerly University of California Riverside 15299), and we treated each species of Kogia as a separate taxonomic unit (S4 Text). We analyzed the matrix using PAUP* , by doing a heuristic search using the tree bisection-reconnection (TBR) algorithm. Statistical support analyses were done by searching for successive longer trees to calculate decay indices and 1000 bootstrap replicates.
The phylogenetic analysis resulted in three most parsimonious trees 95 steps long with consistency index (CI) = 0.589 and retention index (RI) = 0.723. The overall topology of the strict consensus tree is identical to the one shown in Lambert et al. (:Fig 2), with the only difference being the polytomy between Nanokogia and Praekogia (Fig 10). We use the phylogenetic definitions proposed by Lambert  for Physeteridae and Kogiidae. Physeteridae Gray  is phylogenetically defined as the group that includes all physeteroids more closely related to Physeter macrocephalus Linnaeus  than to Kogia breviceps (Blainville ) and K. sima (Owen ). Kogiidae Gill  is defined as the group that includes all physeteroids more closely related to Kogia breviceps (Blainville ) and K. sima (Owen ) than to Physeter macrocephalus Linnaeus . In addition, we phylogenetically define Physeteroidea Gray  as the crown group composed of the last common ancestor of Physeter macrocephalus Linnaeus , Kogia breviceps (Blainville ), and K. sima (Owen ). Pan-Physeteroidea is defined as the panstem that includes crown Physeteroidea. Finally, Kogia Gray  is defined as the crown group composed of the last common ancestor of Kogia breviceps (Blainville ) and K. sima (Owen ), and all its descendants.
Additionally, we performed an analysis of the matrix using the same parameters outlined above but setting certain characters as ordered, following Lambert et al. . This analysis resulted in 73 most parsimonious trees 102 steps long with consistency index (CI) = 0.569 and retention index (RI) = 0.723. The resulting strict consensus tree (S3 Fig) included an unresolved polytomy between crownward Pan-Physeteroidea, Physeteridae and Kogiidae, with the Acrophyseter + Zygophyseter + Brygmophyseter clade fully collapsed, and Physeterula no longer being within Physeteridae. Kogiidae remained largely stable, only with a polytomy between Praekogia and Nanokogia.
Comparison and Relationships
Nanokogia isthmia is a derived kogiid most closely related to P. cedrosensis and Kogia spp. (Fig 10). Among these, the overall morphology of Nanokogia resembles Praekogia cedrosensis much more than any of the other taxa, especially in the shape of the supracranial basin and its small size (Figs 11–14 and S2 Fig). In contrast, the differences in cranial morphology between Nanokogia and Scaphokogia, both from Tortonian-age deposits, are striking; a greater degree of cranial disparity is observed between them compared to the disparity observed between Kogia breviceps and K. sima (Figs 11–14 and S2 Fig). In fact, both extant species of Kogia are so similar that, until 1966, they were considered to represent a single species .
The rostrum of Nanokogia resembles more closely that of Kogia spp. by having a triangular outline in dorsal and ventral views, contrasting with the more rounded outline of the rostrum of Aprixokogia and the cylindrical outline of Scaphokogia (Figs 11–14 and S2 Fig). However, the dorsal surface of the rostrum of Nanokogia is not as concave as that of Kogia or Physeter, resembling in that respect the more flattened to shallowly convex dorsal surface of the rostrum of Aprixokogia, while the markedly convex rostrum of Scaphokogia is unique among kogiids (Fig 14, S2 Fig). In lateral view, the height of the rostrum is fairly similar in Nanokogia, Aprixokogia and Kogia, while in Scaphokogia it is notably greater, giving it a cylindrical outline (Figs 11–14 and S2 Fig). The concave dorsal surface of the rostrum of Kogia, Physeter and Livyatan  can be considered as an extension of the supracranial basin, and accommodates the hypertrophied soft tissue structures of the forehead, namely the melon in the former, and junk and spermaceti organ in the latter two. These structures project anterodorsally beyond the limits of the rostrum (Cranford et al. :Fig 8; Cranford :Fig 3). The flat to concave dorsal surface of the rostrum of Nanokogia, Aprixokogia and Scaphokogia suggests that the associated soft tissue was less hypertrophied anteriorly in these taxa.
The basicranium of Nanokogia isthmia shows a unique combination of morphological features that sets it apart from other kogiids. Nanokogia differs from P. cedrosensis in that the presphenoid is not covered ventrally by the vomer, a characteristic which Nanokogia shares with Physeter macrocephalus, Aprixokogia kelloggi, Scaphokogia cochlearis, and Kogia spp. [12, 29] (Fig 12). However, it does share with Praekogia and Kogia spp. the presence of a wide notch in the squamosal for the posterior process of the tympanic bulla; this notch is much shallower in Aprixokogia, and absent in Thalassocetus [7, 12] (Fig 13). Nanokogia resembles Aprixokogia in having a temporal fossa that is longer than high; thus it differs from the rounded fossa of Praekogia, and from Thalassocetus and Kogia spp. where the temporal fossa is higher than long.
The pterygoid of Nanokogia resembles that of Kogia in the presence of a notch between the hamular process and the medial lamina (Fig 13D–13F); this feature is either absent or not preserved in other kogiids. However, in lateral view, the pterygoid medial lamina and other features of the basicranium of Kogia are oriented more anteroventrally than in Aprixokogia, Nanokogia, or, probably, Scaphokogia. Based on the shape of the basioccipital crest, Praekogia seems to be similar to Kogia (Fig 13). Because of this reorientation, the basicranium of Kogia looks anteroposteriorly shortened or recurved in comparison to other kogiids (Fig 13: i.e., notice the more vertical or diagonal orientation of the long axis of the pterygoid in Kogia). This results in the postorbital process of the frontal overhanging the posterior half of the zygomatic process, foreshortening of the temporal fossa, and reposition of the glenoid fossa to a more ventral position relative to the rostrum (Fig 13E and 13F), likely having an effect on the shape of the mandible as well.
Nanokogia is the only fossil kogiid for which an associated mandible is known. The mandible of Nanokogia differs from that of Kogia breviceps and K. sima in having a straight ventral border of the horizontal ramus (Fig 9 and S1 Fig). In addition, the orientation of the mandible posterior to the gnathion differs markedly between Nanokogia and Kogia: the orientation in the former is posterodorsal (S1A Fig), while in the latter it is posteroventral (S1B–S1C Fig), with the mandibular condyle located far ventrally with respect to the alveolar row. These differences are most likely the result of the reorientation of the basicranium of Kogia as mentioned above.
The supracranial basin also shows marked differences among kogiids (Figs 11–14 and S2 Fig). In nearly all taxa, the supracranial basin is anterodorsally oriented as the posterior edge of the basin is elevated, although not as much as in pan-physeteroids and physeterids (e.g. Acrophyseter deinodon  and Physeter macrocephalus ); the one exception being Scaphokogia whose basin seems to be oriented dorsally (Fig 13). The sagittal facial crests of Nanokogia, Aprixokogia, Scaphokogia, and Kogia spp. differs from that in Praekogia in that the left premaxilla does not reach and form part of the sagittal facial crest (Figs 11, 14 and S2 Fig). The sagittal facial crest of Aprixokogia differs from that of other kogiids in that it does not taper posteriorly and does not reach the posterior margin of the supracranial basin (Fig 14). The crest is tapering and proportionately longer in Thalassocetus, Nanokogia, Praekogia and Kogia, but is much reduced in Scaphokogia. Both Nanokogia and Praekogia share with Aprixokogia the presence of a supracranial fossa mostly confined to the right side of the supracranial basin, although the extent of the fossa differs slightly (see Description) (Figs 11, 14 and S2 Fig). In Scaphokogia cochlearis this fossa is developed to an extreme, with the sagittal facial crest displaced towards the left side of the supracranial basin, resulting in a much larger fossa  (Figs 11, 14 and S2 Fig). The supracranial fossa of these taxa seems to be homologous to the central premaxillary fossa of Kogia spp. and Thalassocetus ([7, 14]; but see  for a different interpretation). In Kogia, the central premaxillary fossa corresponds to the nasofrontal sac and the relatively small spermaceti organ [34, 38–39]. Muizon  and Bianucci and Landini  interpreted the enlarged fossa of Scaphokogia as indicative of an enlarged spermaceti organ. Considering these two structures, the supracranial fossa and central premaxillary fossa, as homologous, we hypothesize that Aprixokogia, Praekogia, and Nanokogia also possessed a spermaceti organ proportionately larger than that of Thalassocetus and Kogia spp., which will also be the most parsimonious explanation, according to our phylogenetic analysis (see below).
The results of our phylogenetic analysis place Aprixokogia kelloggi as the most basal kogiid (Fig 10). This, together with the presence of an enlarged supracranial basin in pan-physeteroids (e.g. Acrophyseter deinodon , Zygophyster varolai ), indicates that an enlarged spermaceti organ is most likely the plesiomorphic condition for crown physeteroids. This suggests that reduction of the spermaceti organ has evolved iteratively within kogiids: once in Thalassocetus antwerpiensis, and secondly in Kogia spp. In Physeter macrocephalus the spermaceti organ is related mainly to sound generation [39–40], but has also been considered as used for interspecific aggression , among other functions. In Kogia it seems that reduction of the organ has resulted in changes in its functionality relative to Physeter, and it has become part of a complex sound-generating system . It has been hypothesized that strong sexual selection has influenced the hypertrophy of the spermaceti organ and nose of Physeter , so it could be argued that the smaller organ of Kogia is the result of less intense sexual selection. Furthermore, cetaceans that display strong sexual selection tend to have larger pelvic bones , which in Kogia seem to be extremely reduced or absent . However, we would need more data from living kogiids as well as fossils with associated cranial and postcranial material in order to test this further.
Our results also differ from previous analyses regarding previously published divergence estimates between Kogia breviceps and K. sima. McGowen et al.  estimated that the average time of divergence between Kogia breviceps and K. sima was around 9.33 Ma (based on a 4.03–15.38 Ma range). Instead, our results, which place Kogia pusilla from the late Pliocene of Italy  as the fossil taxon most closely related to extant Kogia, coupled with the occurrence of Kogia-like periotics in the early Pliocene Yorktown Formation of North Carolina (; JVJ pers. obs.), imply that the divergence between extant Kogia occurred sometime after the early Pliocene, and that it was the result of a much more recent speciation event (Fig 14).
The depositional environment of the Chagres Formation has been a subject of debate, with a variety of paleobathymetric depths estimated for its three distinct members [16–17]. The Piña facies, from which Nanokogia was collected, has been widely described as upper bathyal following Collins et al. (:Fig 1). De Gracia et al.  used fossil osteichthyan and chondrichthyan occurrences to characterize the Piña facies as reflecting both neritic and mesopelagic open-ocean settings, giving a paleobathymetric estimate of 100–700 m. More recently, Hendy et al. , using the mollusk assemblage, estimated the depositional depth of the Piña facies to be typical of the outer continental shelf, around 100–150 m of water depth. The fish assemblage, which includes billfishes  and abundant myctophids , among others, is suggestive of areas of coastal upwelling . This upwelling system likely supported a large diversity of fishes and pelagic invertebrates, such as squids.
Kogia breviceps and K. sima feed mainly on squids (teutophagy), but are also known to consume myctophids [49–50], feeding mainly on prey within the epi- and mesopelagic zones [5, 51]. Because of the morphological similarities between the rostrum and other features of the skull of Nanokogia and that of Kogia, coupled with other paleontological and geological evidence (i.e. high abundance of myctophids, presence of teuthid statoliths, and paleobathymetry estimates for the Piña facies), we hypothesize that Nanokogia fed mainly on diel-migrating fishes (e.g. myctophids) and squids; while the relatively short rostrum is indicative of suction feeding . However, Nanokogia may have had different sound-generating capabilities, as its spermaceti organ was likely larger than that of Kogia (see Comparison and Relationships). Furthermore, when we compare Nanokogia with the other known Tortonian kogiid, Scaphokogia cochlearis, it is evident that there is a greater morphological disparity of the rostrum between these coeval taxa than what is observed between modern species (Figs 11–13 and S2 Fig). The disparity between Nanokogia and Scaphokogia is probably related to different feeding or foraging specializations in the latter. Unfortunately there are no extant analogs for the unique cranial morphology of Scaphokogia, making it difficult to infer its paleoecology.
Fossil Marine Mammals of Central America
There are few other reports of fossil marine mammals from Central America. These include a balaenopterid from the lower Pliocene of Nicaragua , odontocetes from the mid- and upper Miocene of Costa Rica [54–55], and sirenians and odontocetes from the lower through upper Miocene of Panama [56–58]. Based on isolated teeth, several odontocete families were tentatively identified from the Miocene of Costa Rica [54–55]. However, because isolated teeth are poorly diagnostic and highly convergent amongst very different groups, these identifications must be viewed with caution. Previously described marine mammals from Panama consist of isolated postcranial elements (ribs and vertebrae) representing unknown sirenians, mysticetes and odontocetes, from other coeval or older formations [56–57]. So far the only other odontocete reported from the Chagres Fm. is an isolated tooth of a pan-physeteroid. The tooth was described by Vigil and Laurito  as a physeterid, although it should be more properly referred to as a pan-physeteroid due to the presence of an enameled crown. Additional cetacean material from the Piña facies, under study by one of us (JVJ), includes an undescribed pan-physeteroid with Scaldicetus-type teeth (teeth with large bulbous roots and small enameled crowns; Smithsonian Tropical Research Institute [STRI] 34111), an inioid (USNM 546125), and remains of small delphinoids (STRI 37037 and STRI 37039), hinting at a greater diversity of late Miocene Central American marine mammals than currently known. This would be consistent with the diversity seen in other contemporaneous or nearly contemporaneous sites in North America (e.g. Agricola  and Isla Cedros [60–61] faunas in Florida and Baja California, respectively) and South America (e.g. Montemar  and Cerro Ballena  faunas in Peru and Chile, respectively).
Herein we described Nanokogia isthmia, from late Miocene (Tortonian) deposits on the Caribbean coast of Panama, and analyzed disparity and temporal patterns among kogiids, including the living dwarf and pygmy sperm whales. A phylogenetic analysis places Nanokogia in a polytomy with Praekogia cedrosensis, more closely related to Kogia spp., than to other taxa. However, Nanokogia shows a unique combination of morphological features that set it apart from these and other kogiids. Based on morphological features that Nanokogia shares with Kogia, as well as other lines of evidence from the fossil record, we hypothesized that the Panamanian kogiid fed mainly on diel-migrating fishes and squids by suction feeding. Based on our phylogenetic analysis, we concluded that an enlarged spermaceti organ is the plesiomorphic condition for kogiids, and that reduction of the organ has occurred iteratively, at least twice within the clade. Furthermore, the crown clade Kogia seems to have originated at or after the early Pliocene, which is later than previously published molecular estimates. Accordingly, we suggest that future studies of cetacean phylogeny take more into account the use of fossils for calibrations of their trees at all possible taxonomic levels.
In addition to Nanokogia isthmia, other cetaceans from the Piña facies of the Chagres Fm. include physeteroids, inioids, and small delphinoids, indicating that the upper Miocene of Panama may hold a diverse array of marine mammals similar to other contemporaneous sites in North and South America. Finally, the presence of kogiids in the Neotropics shows that the group has been present in the region at least since the late Miocene, and highlights the importance of research in this region in order to further understand the evolutionary history of marine mammals.
S1 Dataset. List of characters and states, and matrix used in the phylogenetic analysis.
S1 Fig. Lateral views of kogiid mandibles.
Lateral views of mandibles of Nanokogia isthmia gen. et sp. nov. (UF 280000) S1A, Kogia sima (right side, reversed, LACM 47142), S1B, and Kogia breviceps (LACM 95745), S1C. Dashed lines denote the ventral curvature of the symphysis. Abbreviations: gn, gnathion. Scale applies to all specimens.
S2 Fig. Anterodorsal views of kogiid skulls.
Anterodorsal views of skulls of Aprixokogia kelloggi (LACM 117744 [cast of USNM 187015]), S2A, Scaphokogia cochlearis (MNHN PPI 229), S2B, Praekogia cedrosensis (UCMP 315229), S2C, Nanokogia isthmia gen. et sp. nov. (UF 280000), S2D, Kogia sima (LACM 47142), S2E, and Kogia breviceps (LACM 95745).
We thank D. J. Bohaska and C. W. Potter (USNM), C. Argot, G. Billet, and C. de Muizon (MNHN), R. C. Hulbert, Jr. and C. L. McCaffery (FLMNH) for providing access to specimens under their care, and D. Janiger (LACM) for help locating specimens. We thank L. M. Chiappe (LACM) and N. D. Pyenson (USNM) for comments on an earlier version of this manuscript, and L. G. Barnes and J. Dines (LACM) for discussions on fossil and living physeteroids. Special thanks to J. F. Parham (CSUF) for help with phylogenetic analysis and discussions on nomenclature, and C. Pimiento (UF) for suggestions of names for the new taxon, and both for comments on an earlier version of this manuscript. We extend our gratitude to J. Crowell-Davis, Z. Leisure, E. Stiles and S. Widlansky for assistance in the field in Panama; and S. Moran for the picture of UF 280000 in anterodorsal view. Finally we thank the Dirección de Recursos Minerales de Panamá for collecting permits. This work greatly benefited from detailed reviews by O. Lambert (IRSNB) and D. P. Domning (HU). Parts of this work were done while JVJ, ARW and AJWH were Post Doctoral Associates at the FLMNH. This is University of Florida Contribution to Paleobiology number 666.
Conceived and designed the experiments: JVJ. Performed the experiments: JVJ. Analyzed the data: JVJ ARW CDG AJWH. Wrote the paper: JVJ ARW CDG AJWH.
- 1. Blainville H de. Sur les cachalots. Ann Fr Étrang d’Anat Physiol. 1838; 2: 335–337.
- 2. Owen R. On some Indian Cetacea collected by Walter Elliot, Esq. Trans Zool Soc London. 1866; 6:17–47.
- 3. Willis PM, Baird RW. Status of the dwarf sperm whale, Kogia simus, with special reference to Canada. Can Field Nat. 1998; 112: 114–125.
- 4. Bloodworth BE, Odell DK. Kogia breviceps (Cetacea: Kogiidae). Mammalian Species. 2008; 819: 1–12.
- 5. McAlpine D. Pygmy and dwarf sperm whales. In: Perrin WF, Würsig B, Thewissen JGM, editors. Encyclopedia of marine mammals. San Diego: Academic Press; 2002. pp. 1007–1009.
- 6. Abel O. Les odontocètes du Boldérien (Miocène supérieur) des environs d’Anvers. Mém Mus Royal Hist Nat Belgique. 1905; 3: 1–155.
- 7. Lambert O. Sperm whales from the Miocene of the North Sea: a re-appraisal. Bull Inst R Sc N B-S. 2008; 78: 277–316.
- 8. de Muizon C. Les vertébrés fossiles de la Formation Pisco (Pérou). Troisième partie: Les odontocètes (Cetacea, Mammalia) du Miocène. Travaux de l’Institut Français d’Etudes Andines. 1988; 78: 1–244. pmid:17356899
- 9. Ehret DJ, MacFadden BJ, Jones DS, DeVries TJ, Foster DA, Salas-Gismondi R. Origin of the white shark Carcharodon (Lamniformes: Lamnidae) based on recalibration of the upper Neogene Pisco Formation of Peru. Palaeontology. 2012; 55: 1139–1153.
- 10. Lambert O, de Muizon C. A new long-snouted species of the Miocene pontoporiid dolphin Brachydelphis and a review of the Mio-Pliocene marine mammal levels in the Sacaco Basin, Peru. J Vertebr Paleontol. 2013; 33: 709–721.
- 11. Barnes LG. Praekogia cedrosensis, a new genus and species of fossil pygmy sperm whale from Isla Cedros, Baja California, Mexico. Contrib Sci, Nat Hist Mus Los Angeles County. 1973; 247: 1–20.
- 12. Whitmore FC Jr, Kaltenbach JA. Neogene Cetacea of the Lee Creek Phosphate Mine, North Carolina; In: Ray CE, Bohaska DJ, Koretsky IA, Ward LW, Barnes LG, editors. Geology and Paleontology of the Lee Creek Mine, North Carolina, IV. Virginia Mus Nat Hist S Pub 14; 2008. pp. 181–269.
- 13. Pilleri G. The Cetacea of the Italian Pliocene with a descriptive catalogue of the species in the Florence Museum of Paleontology. Vammala: Vammalan Kirjapaino Oy; 1987.
- 14. Bianucci G, Landini W. Kogia pusilla from the middle Pliocene of Tuscany (Italy) and a phylogenetic analysis of the family Kogiidae (Odontoceti, Cetacea). Riv Ital Paleontol S. 1999; 105: 445–453.
- 15. Fitzgerald EMG. Pliocene marine mammals from the Whalers Bluff Formation of Portland, Victoria, Australia. Mem Mus Victoria. 2005; 62: 67–89.
- 16. Collins LS, Coates AG, Berggren WA, Aubry MP, Zhang J. The late Miocene Panama isthmian strait. Geology. 1996; 24: 687–690.
- 17. Hendy A JW, Jones D, De Gracia C, Velez-Juarbe J. Paleoecology of the Chagres Formation (latest Miocene) of Panama: reinterpreting the paleoenvironment of a vertebrate-rich marine fauna. J Geol. In review.
- 18. Perrin WF. Variation of spotted and spinner porpoise (genus Stenella) in the eastern Pacific and Hawaii. Bull Scripps Inst Oceanog Univ Calif. 1975; 21: 1–206.
- 19. Mead JG, and Fordyce RE. The therian skull: a lexicon with emphasis on the odontocetes. Sm C Zool. 2009; 627: 1–248.
- 20. Cohen KM, Finney SC, Gibbard PL, Fan J-X. The ICS international chronostratigraphic chart. Episodes. 2013; 36: 199–204.
- 21. Lambert O, Bianucci G, Post K, de Muizon C, Salas-Gismondi R, Urbina M, et al. The giant bite of a new raptorial sperm whale from the Miocene epoch of Peru. Nature. 2010; 466: 105–108. pmid:20596020
- 22. Linnaeus C. Systema naturae per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. Tomus 1, Editio decima, reformata. Stockholm: Laurentii Salvii; 1758.
- 23. Brisson MJ. Regnum animale in Classes IX distributum, sive synopsis methodica sistens generalem animalium distributionem in Classes IX, et duarum primarum Classium, Quadrupedum scilicet & Cetaceorum, particulare divisionem in Ordines, Sectiones, Genera, et Species. Paris: T. Haak; 1762.
- 24. Flower WH. Description of the skeleton of Inia geoffrensis and of the skull of Pontoporia blainvilii, with remarks on the systematic position of these animals in the order Cetacea. Trans Zool Soc London. 1867; 6: 87–116.
- 25. Gray JE. On the natural arrangement of vertebrose animals. London Med Rep. 1821; 15: 296–310.
- 26. Gill T. The sperm whales, giant and pygmy. Am Nat. 1871; 4: 725–743.
- 27. Pyenson ND, Sponberg SN. Reconstructing body size in extinct crown Cetacea (Neoceti) using allometry, phylogenetic methods and tests from the fossil record. J Mamm Evol. 2011; 18: 269–288.
- 28. Caldwell DK, Caldwell MC. Pygmy sperm whale Kogia breviceps (de Blainville, 1838): Dwarf sperm whale Kogia simus Owen, 1866. In: Ridgway SH, Harrison R, editors. Handbook of Marine Mammals, vol. 4: River dolphins and the larger toothed whales. London: Academic Press; 1989. pp. 235–260.
- 29. Schulte HvonW. The skull of Kogia breviceps Blainv. Bull Am Mus Nat Hist. 1917; 37: 361–404.
- 30. Fraser FC, Purves PE. Hearing in cetaceans. Evolution of the accessory air sacs and the structure and function of the outer and middle ear in recent cetaceans. Bull British Mus Nat Hist, Zool. 1960; 7: 1–140.
- 31. Schulte HvonW, Smith MDeF. The external characters, skeletal muscles, and peripheral nerves of Kogia breviceps (Blainville). Bull Am Mus Nat Hist. 1918; 38: 7–72.
- 32. Swofford DL. PAUP* v.40b10. Sinauer Associates, Sunderland; 2002.
- 33. Gray JE. On the cetaceous animals. In: Richardson J, Gray JW, editors. The zoology of the voyage of H. M. S. Erebus and Terror under the command of Captain Sir James Clark Ross, R. N., F. R. S., during the years 1839 to 1843. Vol. 1, pt. 3 (Mammals). London: Longman, Brown, Green and Longmans; 1846. pp. 13–53.
- 34. Cranford TW, Amundin M, Norris KS. Functional morphology and homology in the Odontocete nasal complex: implications for sound generation. J Morphol. 1996; 228: 223–285. pmid:8622183
- 35. Cranford TW. The sperm whale’s nose: sexual selection on a grand scale? Mar Mamm Sci. 1999; 15: 1133–1157.
- 36. Lambert O, Bianucci G, de Muizon C. A new stem-sperm whale (Cetacea, Odontoceti, Physeteroidea) from the latest Miocene of Peru. C R Palevol. 2008; 7: 361–369.
- 37. Huggenberger S, André M, Oelschläger HHA. The nose of the sperm whale: overviews of functional design, structural homologies and evolution. J Mar Biol Assoc UK. 2014.
- 38. Kernan JD Jr, Schulte HvonW. Memoranda upon the anatomy of the respiratory tract, foregut, and thoracic viscera of a fetal Kogia breviceps. B Am Mus Nat Hist. 1918; 38: 231–266.
- 39. Clarke MR. Production and control of sound by the small sperm whales, Kogia breviceps and K. sima and their implications for other Cetacea. J Mar Biol Assoc UK. 2003; 83: 241–263.
- 40. Bianucci G, Landini W. Killer sperm whale: a new basal physeteroid (Mammalia, Cetacea) from the Late Miocene of Italy. Zool J Linn Soc-Lond. 2006; 148: 103–131.
- 41. Carrier DR, Deban SM, Otterstrom J. The face that sank the Essex: potential function of the spermaceti organ in aggression. J Exp Biol. 2002; 205: 1755–1763. pmid:12042334
- 42. Dines JP, Otárola-Castillo E, Ralph P, Alas J, Smith AD, Dean MD. Sexual selection targets cetacean pelvic bones. Evolution. 2014; 68: 3296–3306. pmid:25186496
- 43. Benham WB. On the anatomy of Cogia breviceps. Proc Zool Soc London. 1901; 71: 107–135.
- 44. McGowen MR, Spaulding M, Gatesy J. Divergence date estimation and a comprehensive molecular tree of extant cetaceans. Mol Phylogenet Evol. 2009; 53: 891–906. pmid:19699809
- 45. Luo Z, Marsh K. Petrosal (periotic) and inner ear of a Pliocene kogiine whale (Kogiinae, Odontoceti): implications on relationships and hearing evolution of toothed whales. J Vertebr Paleontol. 1996; 16: 328–348.
- 46. De Gracia C, Carrillo-Briceño J, Schwarzhans W, Jaramillo C. An exceptional marine fossil fish assemblage reveals a highly productive deep-water environment in the Central American Seaway during the late Miocene. Geol Soc Am Abs. 2012; 44: 164.
- 47. Fierstine HL. A new marlin, Makaira panamensis, from the Late Miocene of Panama. Copeia. 1978; 1978: 1–11.
- 48. Schwarzhans W, Aguilera O. Otoliths of the Myctophidae from the Neogene of tropical America. Palaeo Ichthyologica. 2013; 13: 83–150.
- 49. Fitch JE, Brownell RL Jr. Fish otoliths in cetacean stomachs and their importance in interpreting feeding habits. J Fish Board Can. 1968; 25: 2561–2574.
- 50. Staudinger MD, McAlarney RJ, McLellan WA, Pabst DA. Foraging ecology and niche overlap in pygmy (Kogia breviceps) and dwarf (Kogia sima) sperm whales from waters of the U.S. mid-Atlantic coast. Mar Mamm Sci. 2014; 30: 626–655.
- 51. Lindberg DR, Pyenson ND. Things that go bump in the night: evolutionary interactions between cephalopods and cetaceans in the Tertiary. Lethaia. 2007; 40: 335–343.
- 52. Werth AJ. Mandibular and dental variation and the evolution of suction feeding in Odontoceti. J Mamm. 2006; 87: 579–588.
- 53. Lucas SG, McLeod SA, Barnes LG, Alvarado GE, García R, Espinoza E. A baleen whale from the Pliocene of Nicaragua. Rev Geol Am Cen. 2009; 41: 14–24.
- 54. Laurito CA, Valerio AL, Hernández AC, Ovares E. Primer registro de un cetáceo fossil (Mammalia, Cetacea, Odontoceti, Squalodontidae) en la Formación Río Banana, Mioceno medio de Costa Rica, América Central. Rev Geol Am Cen. 2011; 44: 153–156.
- 55. Valerio AL, Laurito CA. Cetáceos fósiles (Mammalia, Odontoceti, Eurhinodelphinoidea, Inioidea, Physeterioidea) de la Formación Curré, Mioceno superior (Hemphilliano temprano tardío) de Costa Rica. Rev Geol Am Cen. 2012; 46: 151–160.
- 56. Uhen MD, Coates AG, Jaramillo CA, Montes C, Pimiento C, Rincon A, et al. Marine mammals from the Miocene of Panama. J S Am Earth Sci. 2010; 30: 167–175.
- 57. Velez-Juarbe J, Wood A, Ridgwell N, Bloch J, MacFadden B. Partial skeleton of a toothed whale (Odontoceti, Cetacea) from the mid to late Miocene Gatun Formation, Panama. J Vertebr Paleontol Progr Abstr. 2013: 231.
- 58. Vigil DI, Laurito CA. Nuevos restos de un Odontoceti fósil (Mammalia: Cetacea, Physeteroidea) para el Mioceno tardío de Panamá, América Central. Rev Geol Am Cen. 2014; 50: 213–217.
- 59. Morgan GS. Miocene and Pliocene marine mammal faunas from the Bone Valley Formation of Central Florida. Proc San Diego Soc Nat Hist. 1994; 29: 239–268.
- 60. Barnes LG. Fossil odontocetes (Mammalia: Cetacea) from the Almejas Formation, Isla Cedros, Mexico. PaleoBios. 1984; 42: 1–46.
- 61. Barnes LG. Miocene and Pliocene Albireonidae (Cetacea, Odontoceti), rare and unusual fossil dolphins from the eastern North Pacific Ocean. Nat Hist Mus Los Angeles County Sci Ser. 2008; 41: 99–152.
- 62. Pyenson ND, Gutstein CS, Parham JF, Le Roux JP, Carreño Chavarría C, Metallo A, et al. Repeated mass strandings of Miocene marine mammals from the Atacama Region of Chile point to sudden death at sea. Proc R Soc B. 2014; 281: 20133316.s pmid:24573855