ENIGMATIC CROCODYLIFORMS FROM THE EARLY MIOCENE OF CUBA

Transcription

1 Journal of Vertebrate Paleontology 34(5): , September 2014 Ó 2014 by the Society of Vertebrate Paleontology ARTICLE ENIGMATIC CROCODYLIFORMS FROM THE EARLY MIOCENE OF CUBA CHRISTOPHER A. BROCHU*,1 and OSVALDO JIM ENEZ-V AZQUEZ 2 1 Department of Earth and Environmental Sciences, University of Iowa, Iowa City, Iowa 52242, U.S.A., chris-brochu@uiowa.edu; 2 Gabinete de Arqueologıa, Oficina del Historiador de la Ciudad de La Habana, Cuba ABSTRACT Early Miocene deposits from the Domo de Zaza locality, in the south-central Cuban province of Sancti Spiritus, preserve crocodyliform remains, including compressed serrated teeth closely resembling those of South American sebecids. Fragmentary cranial and mandibular material is more difficult to assess. Referral to any other post-paleogene crocodyliform known from the Western Hemisphere can be ruled out, and phylogenetic analyses are unable to pinpoint its relationships. Similarities can be found with planocraniids, including ventrally oriented and mediolaterally expanded orbital surfaces, but the morphology of the quadrate is inconsistent with a planocraniid affinity. A sebecid in the Miocene of Cuba would be congruent with evidence from other vertebrates suggesting extensive dispersal between the Greater Antilles and South America during the Neogene, and it would be the first Neogene record of the group outside South America. The other crocodyliform may indicate the presence of an endemic West Indian lineage not closely related to any contemporaneous group. It is also consistent with extant Crocodylus arriving in the Neotropics within the past 5 10 million years. INTRODUCTION Domo de Zaza is a hill in southeastern Sancti Spiritus Province of Cuba. It was transected during construction of a dam in the 1970 s, exposing a sequence of fine- to coarse-grained terrigenous deposits with infrequent limestone intercalations. These are referable to the early Miocene (Burdigalian) Lagunitas Formation, which formed in a series of lagoonal, fluvial, and terrestrial depositional settings (Kantchev et al., 1976; MacPhee and Iturralde-Vinent, 1994). Marine invertebrate fossils and correlation with global sea level patterns suggest an age of Ma (MacPhee et al., 2003). This locality preserves one of the very few Neogene nonmarine vertebrate faunas in the West Indies. Fossils collected since the 1990 s include sirenians, platyrrhine monkeys, capromyid rodents, megalonychid sloths, and pelomedusoid turtles (MacPhee and Iturralde-Vinent, 1995 a, 1995b; White and Mac- Phee, 2001; MacPhee et al., 2003; MacPhee and Meldrum, 2006). These lineages suggest faunal exchange with South America, and the locality is central to our efforts to understand the biogeographic history of the West Indies over the past 20 million years. Crocodyliforms are known from this locality (MacPhee and Iturralde-Vinent, 1994). They, too, might play an important biogeographic role. Two species occur naturally in the region today the American (Crocodylus acutus) and Cuban (C. rhombifer) crocodiles. The spectacled caiman, Caiman crocodilus, isa 20th century invasive (Velasco and Ayarzag uena, 2010). The two crocodiles are part of a Neotropical radiation that, based on fossil and molecular evidence, dates back only to the late Miocene or Pliocene (Brochu, 2000; Hekkala et al., 2011; Meredith et al., 2011; Oaks, 2011; Scheyer et al., 2013). Neotropical crocodiles are recently enough diverged that hybridization is commonly observed (Dever et al., 2002; Ray et al., 2004; Rodriguez et al., 2008, 2011; Weaver et al., 2008; Milian-Garcia et al., 2011). They thus presumably dispersed into the region at some point within the past 5 10 million years (Brochu, 2001; Meredith et al., 2011). *Corresponding author. Given their ubiquity elsewhere when climatic and environmental factors were adequate, we can assume there were crocodyliforms in the West Indies before Crocodylus arrived. But who were they? The Cenozoic record of West Indian crocodyliforms is sparse and discontinuous. A tomistomine and a ziphodont form are known from the middle Eocene of Jamaica (Berg, 1969; Domning and Clark, 1993; Velez-Juarbe and Brochu, in press), and a gryposuchine gharial is known from the Oligocene of Puerto Rico (Velez-Juarbe et al., 2007). By the Quaternary, fossil crocodyliforms are closely related or referable to one of the species of Crocodylus currently living in the region (Brown, 1913; Varona, 1966, 1984; Pregill, 1982; Olson et al., 1990; Morgan, 1993; Morgan et al., 1993; Franz et al., 1995; Steadman et al., 2007; Velez-Juarbe and Miller, 2007; Morgan and Albury, 2013). Other than fragments from the early Miocene of Puerto Rico (Brochu et al., 2007) and the undescribed specimens from Domo de Zaza, nothing is known from the Neogene.Here, we describe the crocodyliform fossils from Domo de Zaza. They include isolated serrated teeth, suggesting that crocodyliforms were among the primary land predators in Cuba during the early Miocene. They also include cranial remains not clearly referable to any other known lineage, living or extinct, suggesting the presence of a derived endemic clade in the West Indian Neogene that was later replaced by Crocodylus. Institutional Abbreviations AMNH, American Museum of Natural History, New York, U.S.A.; FMNH, Field Museum of Natural History, Chicago, U.S.A.; MNHNCu, Museo Nacional de Historia Natural, Havana, Cuba; TMM, Texas Memorial Museum, Austin, U.S.A. SYSTEMATIC PALEONTOLOGY CROCODYLOMORPHA Walker, 1970 CROCODYLIFORMES Hay, 1930 Gen. et sp. indet. (Fig. 1) Referred Specimen MNHNCU P3035, two tooth crowns. 1094

2 BROCHU AND JIM ENEZ-V AZQUEZ CUBAN MIOCENE CROCODYLIFORMS 1095 FIGURE 1. MNHNCU P3035, Crocodyliformes, gen. et sp. indet., serrated tooth crowns, lower Miocene (Burdigalian) Lagunitas Formation, Domo de Zaza, Cuba. One tooth is shown in labial (A) and mesial (B) views, the other in lingual (C) and distal (D) views. Scale bar equals 1 cm. Occurrence Domo de Zaza, southeastern Sancti Spiritus Province, Cuba; Lagunitas Formation, lower Miocene (Burdigalian). Description Two flattened, serrated crowns lacking roots were recovered from the locality (Fig. 1). Each is slightly concave lingually, with smooth, unserrated lingual and labial surfaces. The mesial margin is strongly concave and the distal margin is almost linear, imparting a D -shape to the teeth in lingual or labial view. The narrow mesial and distal surfaces bear serrated carinae. These are true serrations formed by clefts in the enamel and not pseudoserrations formed from the intersection of enamel ridges with a carina (Prasad and de Lapparent de Broin, 2002). The mesial carina on the specimen in Figure 1 A and B is not preserved, but the distal carina bears approximately 4.3 denticles/mm. Conversely, the mesial carina is better preserved on the other tooth (Fig. 1D) and bears approximately 5 denticles/ mm. NEOSUCHIA Benton and Clark, 1988 EUSUCHIA Huxley, 1875 Gen. et sp. indet. (Figs. 2, 3) Referred Specimens MNHNCU P3051, frontal (Fig. 2 A, B); MNHNCU P3016, cranial fragment (possibly parietal; Fig. 2 C); MNHNCU P3110, left quadrate, squamosal, and exoccipital (Fig. 2D F); MNHNCU P3026, right quadrate ramus with portions of quadratojugal and squamosal (Fig. 2G J); MNHNCU P3031, partial basioccipital and right exoccipital (Fig. 2 K); MNHNCU P3040, two osteoderms (Fig. 3). The cranial remains probably pertain to a single individual. Occurrence Domo de Zaza, southeastern Sancti Spiritus Province, Cuba; Lagunitas Formation, lower Miocene (Burdigalian). Description The main body of the frontal (MNHNCU P3051; Fig. 2 A, B) is nearly complete. Its dorsal surface is planar at the center and modestly expanded dorsally along the orbital margins. The sutural surface for the prefrontal in dorsal view intersects the incomplete frontal anterior process at an approximately 50 angle, and it emerged from the orbit at an acute angle. The sutural contact for the parietal indicates an anteriorly concave suture that did not intersect the supratemporal fenestrae. In ventral view, the orbital surfaces are mediolaterally wide, semicircular structures lateral to the groove for the olfactory tract. A triangular bone associated with these remains (MNHNCU P3016; Fig. 2 C) may be the anterior part of the parietal. What we interpret as the anterior surface is the sutural contact for the frontal. It conforms with the posterior surface of the frontal except for a V -shaped midsagittal indentation that would accommodate a triangular posterior process of the frontal. No such structure is evident on the frontal, but the sutural surface along the posterior margin is damaged at the midline, suggesting that a process might have broken away. Shallow depressions posterolaterally are consistent with the anteromedial sections of the supratemporal fenestrae; if so, the supratemporal fenestrae would have been relatively large. We also considered the possibility that MNHNCU P3016 is the posterior part of the parietal. The V -shaped depression would then be for an anterior process of the supraoccipital exposed dorsally on the skull table. Although this cannot be completely discounted, we think it unlikely. In virtually all other crocodyliforms, the sutural surface between the parietal and supraoccipital is oriented ventrally or posteroventrally, not directly posteriorly, and the posterodorsal surface of the parietal is smooth where it forms the dorsal margin of the posttemporal fenestra. The dorsal surface of the left squamosal (MNHNCU P3110; Fig. 2D) is complete, preserving at least part of the sutural surface with the postorbital anteriorly and the parietal posteromedially. Its dorsal surface is planar and pitted, forming the lateral and posterior margins of the unconstricted supratemporal fenestra. The lateral squamosal margin of the skull table is broadly convex in dorsal view and bears a shallow groove with parallel dorsal and ventral borders on its lateral surface. The squamosal forms the roof of the otic recess and the anterior surface of the paroccipital process. The right quadrate ramus (MNHNCU P3026) is more complete than its left counterpart (MNHNCU P3110) and preserves part of the quadratojugal and its sutural surface for the jugal (Fig. 2 H, J). The foramen a ereum is located on the dorsomedial surface. A modest muscle attachment tuberosity can be seen ventrally. Both the squamosal and quadrate enclose the external otic aperture (Fig. 2E, I). On both sides, the aperture is a triangular opening with a concave ventral margin. The squamosal-quadrate suture intersects the aperture along its posterodorsal margin, not at its posteroventral corner. At first, the opposite appears to be true on the left side in Figure 2E, but the apparent suture is a crack, and the suture can be seen dorsal to it. There is insufficient preservation to tell whether a preotic siphonium was present. The morphology of the quadrate condyle is difficult to assess. The right quadrate bears a hemicylindrical condylar surface that appears to be of uniform depth in posterior view and posteriorly expanded in dorsal view (Figs. 2D, G, 4 A, B), but whether the medial hemicondyle was ventrally reflected (the ancestral condition for Crocodylia) depends on how one orients the ramus (Fig. 4). If oriented such that the skull table is horizontal, the long axis of the lateral hemicondyle would intersect the transverse plane at an angle of approximately 30. This would imply a dorsally expanded medial hemicondyle, but we cannot know whether the skull table was horizontal (although that would be the default expectation), nor do we know if the ramus was twisted postmortem. The right lateral braincase wall is badly preserved on P3031, and sutures cannot be traced. There evidently was a sulcus anteriorly that would have separated the braincase wall from the basisphenoid rostrum. The exoccipital forms the posterior surface of the paroccipital process and the dorsolateral margin of the foramen magnum. On

3 1096 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 34, NO. 5, 2014 FIGURE 2. Eusuchia, gen. et sp. indet., lower Miocene (Burdigalian) Lagunitas Formation, Domo de Zaza, Cuba. MNHNCU P3051, frontal in dorsal (A) and ventral (B) views. OHCP P3016,?parietal, dorsal view (C). OHCP P3110, left squamosal, quadrate, and exoccipital in dorsal (D), lateral (E), and posterior (F) views. MNHNCU P3026, right quadrate, quadratojugal, and squamosal in dorsal (G), ventral (H), lateral (I), and anterodorsal (J) views. MNHNCU P3031, basioccipital and right exoccipital in posterior view (K). Abbreviations: bo, basioccipital; cqc, cranioquadrate canal; eo, exoccipital; eoa, external otic aperture; fae, foramen a ereum; fvf, fossa for vagus foramen; leu, position of lateral Eustachian opening; lhc, lateral hemicondyle; lsg, lateral squamosal groove; meu, position of median Eustachian opening; mhc, medial hemicondyle; o, orbit; oc, occipital condyle; q, quadrate; qj, quadratojugal; sc, sagittal crest; sj, sutural surface for jugal; sn, sutural surface for nasal; spa, sutural surface for parietal; spo, sutural surface for postorbital; spr, sutural surface for prefrontal; sq, squamosal; sqj, sutural surface for quadratojugal; stf, supratemporal fenestra. Scale bar equals 1 cm. the left side, one can observe a completely enclosed cranioquadrate canal bound by the exoccipital and quadrate and, medially, a fossa for the vagus foramen (Fig. 2 F). Foramina for cranial nerve XII and the carotid foramen are not preserved. On the right side, a triangular descending process is preserved passing along the lateral surface of the basioccipital (Fig. 2 K). The process terminates dorsal to the basioccipital tubera. The basioccipital (Fig. 2 K) bears the occipital condyle. Its posterior surface ventral to the occipital condyle bears a thin sagittal crest and expanded tubera along the ventraloateral margin. A ventrolateral concavity on the left indicates the position of the lateral Eustachian opening, and a large median Eustachian opening is partly preserved at the ventral tip. Part of the basisphenoid is preserved on the anterior surface of the basioccipital, revealing an anteroposteriorly thin, sheet-like posterior lamina. Two osteoderms from the locality, MNHNCU P3040 (Fig. 3), are consistent with a crocodylian of the size indicated by the cranial remains and might be referable to the same individual. Both have pitted dorsal surfaces. One of them is square with a modest midline keel and smooth, flattened anterior imbrication surface. The other also has an imbrication zone, but has an oval outline and a more robust keel.

4 BROCHU AND JIM ENEZ-V AZQUEZ CUBAN MIOCENE CROCODYLIFORMS 1097 FIGURE 3. MNHNCU P3040,?Eusuchia, gen. et sp. indet., lower Miocene (Burdigalian) Lagunitas Formation, Domo de Zaza, Cuba, osteoderms (A, B), dorsal view. Scale bar equals 1 cm. DISCUSSION Three ziphodont crocodyliform groups are known from the Cenozoic Sebecidae, Planocraniidae (formerly Pristichampsinae; Brochu, 2013), and Quinkana. Given the presence of groups with a South American origin (e.g., platyrrhines, capromyids, pelomedusoids) at Domo de Zaza, one might suspect that ziphodont crocodyliforms from the locality would be sebecids, which are known from the Paleocene middle Miocene of South America (e.g., Simpson, 1937; Langston, 1965; Gasparini et al., 1993; Paolillo and Linares, 2007; Molnar, 2010; Scheyer and Moreno- Bernal, 2010; Pol and Powell, 2011). If so, the Domo de Zaza form would be the only known occurrence of the group outside South America during the Neogene. A biogeographic case can be made for Planocraniidae, which is known from North America and Eurasia (Langston, 1975; Li, 1984; Rossmann, 1998); but this would be a substantial stratigraphic range extension for the group, which last unambiguously FIGURE 4. Right crocodyliform quadrates in posterior view showing morphology of condylar surface. That of MNHNCU P3026 is shown with the condyle aligned as it might be if the Domo de Zaza form has the plesiomorphic condition for Crocodylia (A) and with the dorsal surface of the skull table parallel with the transverse plane (B). C, FMNH 23505, Gavialis gangeticus. D, TMM m-7484, Alligator mississippiensis. E, FMNH 17157, Crocodylus niloticus. F, FMNH PR399, Boverisuchus vorax. G, TMM , Sebecus icaeorhinus (cast of AMNH 3160). Abbreviations: dn, dorsal notch between medial and lateral hemicondyles diagnostic for Alligatoroidea; lhc, lateral hemicondyle; mhc, medial hemicondyle. Scale bar for A and B equals 1 cm; other specimens not shown to scale, but all show mature morphology. appears in the middle Eocene (Berg, 1966; Antunes, 1986; Busbey, 1986; Westgate, 1989). The mekosuchine crocodyloid Quinkana is known from the Neogene, but has been found only in Australasia (e.g., Molnar, 1981; Megirian, 1994; Willis and Mackness, 1996). Serrated teeth have been found in the Eocene of Jamaica (Velez-Juarbe and Brochu, in press). The other vertebrates from that locality suggest a North American affinity for the fauna, which would imply that the Jamaican ziphodont form could have been a planocraniid. It is possible that the Cuban and Jamaican ziphodont forms were closely related, but this is very difficult to test based on teeth alone. Taxonomic assessment should be based on morphology rather than place and time. Unfortunately, crocodyliform teeth are notoriously uninformative. We can really only say that one of the primary terrestrial predators in the early Miocene of central Cuba was a ziphodont crocodyliform. The teeth resemble those of at least some sebecids in one respect visually, the denticles on the carinae look finer than in most derived planocraniids or Quinkana (C.A.B., pers. observ.) but this is difficult to quantify, and denticle concentrations vary among sebecids (Langston, 1965; Legasa et al., 1993). Hence, although a sebecid affinity seems more likely than the alternatives, a referral must await discovery of more complete material. The skull and osteoderms are from a substantially smaller animal than the teeth. If all of this material comes from a single species (and there is no evidence that it does), we can rule out a referral to Sebecidae. The enclosed cranioquadrate canal and thin, vertically oriented posterior basisphenoid lamina show that the Domo de Zaza form is a derived neosuchian and most likely a crocodylian. Like other notosuchians, sebecids have broad posterior basisphenoid laminae (Colbert, 1946; Paolillo and Linares, 2007; Molnar, 2010; Turner and Sertich, 2010). Sebecids also have a ventrally oriented quadrate ramus with a prominent crest on the dorsal surface, imparting a nearly triangular appearance to the quadrate condyle in posterior view (Fig. 4G). The trochlea separating the hemicondyles is very broad and oriented ventrolaterally (Colbert, 1946; Molnar, 2010). The sebecid quadrate condyle has a similar morphology in derived planocraniids (Fig. 4 F; Langston, 1975; Brochu, 2013). The quadrate of the Domo de Zaza skull is open to multiple interpretations (see below), but none of them would be consistent with a sebecid. We can exclude most other crocodyliforms known from the Mio-Pliocene of the Western Hemisphere. The descending rami of the exoccipitals do not extend far enough ventrally to contribute to the basioccipital tubera, as they would in a caimanine; and although the supratemporal fenestra is incompletely preserved, it suggests the absence of the tightly constricted margin seen in caimanines (Norell, 1988; Brochu, 1999, 2011; Hastings et al., 2013; Bona et al., 2013). In any case, the quadrate foramen a ereum is not shifted dorsally, as it would be in an alligatoroid (Norell et al., 1994; Brochu, 1999). The basioccipital of the Domo de Zaza form bears an acute sagittal crest and has a triangular ventral tip, the lateral Eustachian foramina are dorsolateral (and not lateral) to the median opening, and the basisphenoid was anteroposteriorly thin in ventral view (Fig. 2 K) features that argue strongly against affinity with derived gavialoids, including gryposuchines (Langston and Gasparini, 1997; Brochu, 2004; Velez-Juarbe et al., 2007; Riff and Aguilera, 2008). The dorsolateral position of the lateral Eustachian opening is also inconsistent with the condition in Crocodylus (Brochu, 2000). The lateral squamosal groove of the Domo de Zaza form has parallel dorsal and ventral margins. In gavialoids and most tomistomines (including the form found in the region during the Neogene, Thecachampsa), the groove flares anteriorly (e.g., Velez-Juarbe et al., 2007). There are similarities between this material and planocraniids. At least one planocraniid, Boverisuchus vorax, shares a

5 1098 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 34, NO. 5, 2014 squamosal-quadrate suture extending dorsally along a linear posterodorsal margin of the otic aperture with the Domo de Zaza form. The very broad orbital surfaces on the frontal (Fig. 2B) are also seen in derived planocraniids (e.g., Boverisuchus, Planocrania datangensis; Brochu, 2013). These similarities are less informative than they first appear. The shape of the otic aperture is plesiomorphic for Crocodylia; in alligatoroids and crocodyloids, the posterodorsal margin is concave. The dorsal extension of the squamosal-quadrate suture adjacent to the aperture is also characteristic of alligatoroids (Brochu, 1999). The combination of plesiomorphic aperture shape and derived sutural arrangement is also found in the close crocodylian relative Allodaposuchus (C.A.B., pers. observ.). And although planocraniids and the Domo de Zaza form both have very wide orbital surfaces, the surfaces in planocraniids are also dorsoventrally deep (Brochu, 2013). In the Domo de Zaza form, the surfaces are not as deep and face ventrally (Fig. 2B). Additional referrals can be proposed or rejected depending on how the quadrate condyle of MNHNCU P3026 is restored. If one looks only at the condyle as preserved in posterior view (Fig. 4 A), it strongly suggests the plesiomorphic condition for Crocodylia: medial and lateral hemicondyles of approximately the same dorsoventral thickness, but with the medial hemicondyle oriented ventromedially relative to the lateral hemicondyle. This is the condition found in gavialoids (Fig. 4 C), Borealosuchus, and the basal-most planocraniids (Brochu et al., 2007; Brochu, 2013). There is no evidence on MNHNCU P3026 for the prominent notch separating the hemicondyles typical of alligatoroids and the basal-most planocraniids (Fig. 4D; Brochu, 1999, 2013). The problem is the rest of the specimen. If one orients the dorsal surface of the squamosal parallel to the transverse plane, the lateral hemicondyle appears to be oriented ventrolaterally relative to its medial counterpart (Fig. 4B). This would elevate the dorsal margin of the medial hemicondyle relative to the lateral. This resembles the crocodyloid condition, in which the medial hemicondyle is dorsoventrally expanded (Fig. 4E; Brochu et al., 2007). It could also be interpreted as resembling the quadrate of Boverisuchus, in which the central part of the condyle is dorsoventrally deeper than either the lateral hemicondyle (Langston, 1975; Fig. 4 F). The similarities, however, are imperfect. In crocodyloids, the medial hemicondyle is expanded both dorsally and ventrally. The outline of the condyle as a whole thus approximates a figure-eight pattern. In Boverisuchus, the central part of the condyle is dorsoventrally deeper than either medial or lateral hemicondyles, and the overall outline is rhomboid. The ventromedial surface on MNHNCU P3026 is also elevated, and the medial hemicondyle does not appear to be expanded ventrally. It appears as though the condyle outline would resemble that of a gavialoid, but with a different orientation relative to the transverse plane. Early Miocene crocodyliforms are known from the lower Miocene Cibao Formation of Puerto Rico, including AMNH a partial dentary, frontal, and squamosal possibly referable to one species (Brochu et al., 2007). Although even more fragmentarily known than the Domo de Zaza form, we can rule out conspecificity between the two samples. The frontals in particular have very different shapes, with the orbital margins of the Domo de Zaza form being anterolaterally oriented and expanded, giving the main corpus a triangular shape (Fig. 2 A, B); the Cibao Formation frontal has the laterally oriented orbital margins found in most crocodylians (Fig. 5). The frontalprefrontal suture emerged from the orbit at a low angle in the Domo de Zaza form, but was almost perpendicular to the orbit in the Cibao form. The Cibao specimen is from a larger individual, but differences between AMNH and MNHNCU FIGURE 5. AMNH 24494,?Neosuchia, gen. et sp. indet., lower Miocene Cibao Formation, Puerto Rico, frontal in dorsal (A) and ventral (B) views. Abbreviations: o, orbit; spr, sutural surface for prefrontal. Scale bar equals 1 cm. P3051 are outside the range of ontogenetic variation for any living crocodylian (C.A.B., pers. observ.). The osteoderms might be informative if they are referable to the Domo de Zaza form, but only if the square element (Fig. 3 C, D) is from one of the paramedian rows. Gavialoids and tomistomines (including Thecachampsa) have rectangular paramedian osteoderms (Erickson and Sawyer, 1996; Brochu, 2004). Those of gavialoids also lack midline keels and have prominent anterolateral convexities. But this particular osteoderm might be from one of the lateral (accessory) rows, making it consistent with virtually any crocodylian with keeled osteoderms. Although we can say what the Domo de Zaza is not, we cannot say what it actually is. There is nothing preserved in the available material to tell us which crocodylian group is its closest relative. Indeed, declaring it a member of Crocodylia is somewhat less than robust. When added to the matrix of Brochu et al. (2012; Appendix 1) and subjected to a maximum parsimony analysis using TNT 1.1 (Goloboff et al., 2008) following the same analytical protocols, multiple positions close to the root of Crocodylia, and even outside Crocodylia itself, are equally optimal (Fig. 6). In the analysis leading to the tree in Figure 6, the condition of the quadrate condyle was treated as unknown. If we instead treat the condition in the Domo de Zaza form as plesiomorphic for Crocodylia, sister-group relationships with Planocraniidae or Brevirostres are one step longer than optimal, and between Alligatoroidea or Crocodyloidea two steps. If, on the other hand, we give it the same character state assigned to crocodyloids, only a single optimal position exists as the sister taxon to Brevirostres. Arguments could be made for or against either of these codings; indeed, the Domo de Zaza quadrate condyle may be unlike that of any other eusuchian. This highlights not only the incompleteness of the Domo de Zaza form, but the distance yet to be traveled to resolve basal relationships within Crocodylia.

6 BROCHU AND JIM ENEZ-V AZQUEZ CUBAN MIOCENE CROCODYLIFORMS 1099 FIGURE 6. Strict consensus of 2230 equally optimal trees (tree length D 646 steps, consistency index D 0.355, retention index D 0.812) based on a phylogenetic analysis of 179 morphological characters and 96 ingroup taxa. Details within clades are shown in Brochu et al. (2012). Dots show positions, at different levels of optimality, for the Domo de Zaza cranial material. The incompleteness of this material, along with a conflicting combination of plesiomorphic and derived character states, limits what can be said. We can, however, draw two conclusions. First, crocodyliforms were among the terrestrial predators on Cuba during the early Miocene. Second, at least one crocodyliform unrelated to any contemporary mainland lineage lived in the Greater Antilles at the time. Further work at this locality may yet shed more light on these enigmatic forms, their role in the ecosystem of a Neogene island, and their bearing on the historical biogeography of the Neotropics. ACKNOWLEDGMENTS We are indebted to the late J. Thorbjarnarson, who organized the workshop on Cuban crocodile biology and evolution that brought the authors together to work on these fossils. His impact on the field of crocodylian studies is incalculable. M. E. L. Gold took the photos used in Figure 5. We are grateful to C. Mehling, A. Resetar, S. C. Sagebiel, R. Ramos-Targarona, and R. Rodrıguez-Soberon for access to collections and facilities, and to R. MacPhee, G. Morgan, N. Albury, J. Daza-Vaca, J. Velez-Juarbe, Y. Milian-Garcıa, and the UI Paleo group for discussion and insight. Careful reviews by J. Velez-Juarbe and T. Scheyer improved the manuscript. This work was financially supported by NSF DEB and (to C.A. B.) and, through J. Thorbjarnarson, the Wildlife Conservation Society. LITERATURE CITED Antunes, M. T Iberosuchus et Pristichampsus, crocodiliens de l Eocene donnees complementaires, discussion, distribution, stratigraphique. Ci^encias da Terra 8: Benton, M. J., and J. M. Clark Archosaur phylogeny and the relationships of the Crocodylia; pp in M. J. Benton (ed.), The Phylogeny and Classification of the Tetrapods, Volume 1. Clarendon Press, Oxford, U.K. Berg, D. E Die Krokodile, insbesondere Asiatosuchus und aff. Sebecus?, aus dem Eoz an von Messel bei Darmstadt/Hessen. Abhandlungen des Hessischen Landesamtes f ur Bodenforschung 52: Berg, D. E Charactosuchus kugleri, eine neue Krokodilart aus dem Eoz an von Jamaica. Eclogae Geologicae Helvetiae 62: Bona, P., D. Riff, and Z. B. Gasparini Late Miocene crocodylians from northeast Argentina: new approaches about the austral components of the Neogene South American crocodylian fauna. Earth and Environmental Science Transactions of the Royal Society of Edinburgh 103: Brochu, C. A Phylogeny, systematics, and historical biogeography of Alligatoroidea. Society of Vertebrate Paleontology Memoir 6: Brochu, C. A Phylogenetic relationships and divergence timing of Crocodylus based on morphology and the fossil record. Copeia 2000: Brochu, C. A Congruence between physiology, phylogenetics, and the fossil record on crocodylian historical biogeography; pp in G. Grigg, F. Seebacher, and C. E. Franklin (eds.), Crocodilian Biology and Evolution. Surrey Beatty and Sons, Chipping Norton, New South Wales, Australia. Brochu, C. A A new Late Cretaceous gavialoid crocodylian from eastern North America and the phylogenetic relationships of thoracosaurs. Journal of Vertebrate Paleontology 24: Brochu, C. A Phylogenetic relationships of Necrosuchus ionensis Simpson, 1937 and the early history of caimanines. Zoological Journal of the Linnean Society 163:S228 S256. Brochu, C. A Phylogenetic relationships of Paleogene ziphodont eusuchians and the status of Pristichampsus Gervais, Earth and Environmental Science Transactions of the Royal Society of Edinburgh 103: Brochu, C. A., A. Nieves-Rivera, J. Velez-Juarbe, J. D. Daza-Vaca, and H. Santos Tertiary crocodylians from Puerto Rico: evidence for Late Tertiary endemic crocodylians in the West Indies? Geobios 40: Brochu, C. A., D. C. Parris, B. S. Grandstaff, R. Denton, and W. B. Gallagher A new species of Borealosuchus (Crocodyliformes: Eusuchia) from the Late Cretaceous early Paleocene of New Jersey. Journal of Vertebrate Paleontology 32: Brown, B Some Cuban fossils. American Museum Journal 13: Busbey, A. B Pristichampsus cf. P. vorax (Eusuchia, Pristichampsinae) from the Uintan of West Texas. Journal of Vertebrate Paleontology 6: Colbert, E. H Sebecus, representative of a peculiar suborder of fossil Crocodilia from Patagonia. Bulletin of the American Museum of Natural History 87: Dever, J. A., R. E. Strauss, T. R. Rainwater, and S. M. McMurry Genetic diversity, population subdivision, and gene flow in Morelet s Crocodile (Crocodylus moreletii) from Belize, Central America. Copeia 2002: Domning, D. P., and J. M. Clark Jamaican Tertiary marine Vertebrata; pp in R. M. Wright and E. Robinson (eds.), Biostratigraphy of Jamaica. Geological Society of America, Boulder, Colorado. Erickson, B. R., and G. T. Sawyer The estuarine crocodile Gavialosuchus carolinensis n. sp. (Crocodylia: Eusuchia) from the late Oligocene of South Carolina, North America. Monographs of the Science Museum of Minnesota (Paleontology) 3:1 47. Franz, R., G. S. Morgan, N. Albury, and S. D. Buckner Fossil skeleton of a Cuban crocodile (Crocodylus rhombifer) from a blue hole on Abaco, Bahamas. Caribbean Journal of Science 31: Gasparini, Z., M. Fernandez, and J. Powell New Tertiary sebecosuchians (Crocodylomorpha) from South America: phylogenetic implications. Historical Biology 7:1 19.

7 1100 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 34, NO. 5, 2014 Goloboff, P. A., J. S. Farris, and K. C. Nixon T.N.T. Tree Analysis Using New Technology Willi Hennig Society, Tucuman, Argentina. Hastings, A. K., J. I. Bloch, C. A. Jaramillo, A. F. Rincon, and B. J. Mac- Fadden Systematics and biogeography of crocodylians from the Miocene of Panama. Journal of Vertebrate Paleontology 33: Hay, O. P Second bibliography and catalogue of the fossil Vertebrata of North America (Volume 2). Carnegie Institute of Washington Publication 390: Hekkala, E., M. H. Shirley, G. Amato, J. D. Austin, S. Charter, J. Thorbjarnarson, K. A. Vliet, M. L. Houck, R. DeSalle, and M. J. Blum An ancient icon reveals new mysteries: mummy DNA resurrects a cryptic species within the Nile crocodile. Molecular Ecology 20: Huxley, T. H On Stagonolepis robertsoni, and on the evolution of the Crocodilia. Quarterly Journal of the Geological Society 31: Kantchev, I Resultados del levantimiento geologico y las investigaciones cientıficas en la provincia de Las Villas. Academia de Ciencias de Cuba, Havana, Cuba, 1480 pp. Langston, W Fossil crocodilians from Colombia and the Cenozoic history of the Crocodilia in South America. University of California Publications in Geological Sciences 52: Langston, W Ziphodont crocodiles: Pristichampsus vorax (Troxell), new combination, from the Eocene of North America. Fieldiana: Geology 33: Langston, W., and Z. Gasparini Crocodilians, Gryposuchus, and the South American gavials; pp in R. F. Kay, R. H. Madden, R. L. Cifelli, and J. J. Flynn (eds.), Vertebrate Paleontology in the Neotropics: The Miocene Fauna of La Venta, Colombia. Smithsonian Institution, Washington, D.C. Legasa, O., A. D. Buscalioni, and Z. Gasparini The serrated teeth of Sebecus and the iberoccitanian crocodile, a morphological and ultrastructural comparison. Studia Geologica Salmanticensia 29: Li, J A new species of Planocrania from Hengdong, Hunan. Vertebrata PalAsiatica 22: MacPhee, R. D. E., and M. A. Iturralde-Vinent First Tertiary land mammal from Greater Antilles: an early Miocene sloth (Xenarthra, Megalonychidae) from Cuba. American Museum Novitates 3094: MacPhee, R. D. E., and M. A. Iturralde-Vinent a. Earliest monkey from Greater Antilles. Journal of Human Evolution 28: MacPhee, R. D. E., and M. A. Iturralde-Vinent. 1995b. Origin of the Greater Antillean land mammal fauna, 1: new Tertiary fossils from Cuba and Puerto Rico. American Museum Novitates 3141:1 31. MacPhee, R. D. E., and J. Meldrum Postcranial remains of the extinct monkeys of the Greater Antilles, with evidence for semiterrestriality in Paralouatta. American Museum Novitates 3516:1 65. MacPhee, R. D. E., M. A. Iturralde-Vinent, and E. S. Gaffney Domo de Zaza, an Early Miocene vertebrate locality in south-central Cuba, with notes on the tectonic evolution of Puerto Rico and the Mona Passage. American Museum Novitates 3394:1 42. Megirian, D A new species of Quinkana (Eusuchia: Crocodylidae) from the Miocene Camfield Beds of northern Australia. The Beagle 11: Meredith, R. W., E. R. Hekkala, G. Amato, and J. Gatesy A phylogenetic hypothesis for Crocodylus (Crocodylia) based on mitochondrial DNA: evidence for a trans-atlantic voyage from Africa to the New World. Molecular Phylogenetics and Evolution 60: Milian-Garcıa, Y., M. Venegas-Anaya, R. Frias-Soler, A. J. Crawford, R. Ramos-Targarona, R. Rodrıguez-Soberon, M. Alonso-Tabet, J. Thorbjarnarson, O. I. Sanjur, G. Espinosa-Lopez, and E. Bermingham Evolutionary history of Cuban crocodiles Crocodylus rhombifer and Crocodylus acutus inferred from multilocus markers. Journal of Experimental Zoology 315: Molnar, R. E Pleistocene ziphodont crocodilians of Queensland. Records of the Australian Museum 33: Molnar, R. E A new reconstruction of the skull of Sebecus icaeorhinus (Crocodyliformes: Sebecosuchia) from the Eocene of Argentina. Uberlandia 1: Morgan, G. S Quaternary land vertebrates of Jamaica; pp in R. M. Wright and E. Robinson (eds.), Biostratigraphy of Jamaica. Geological Society of America, Boulder, Colorado. Morgan, G. S., and N. A. Albury The Cuban crocodile (Crocodylus rhombifer) from late Quaternary fossil deposits in the Bahamas and Cayman Islands. Bulletin of the Florida Museum of Natural History 52: Morgan, G. S., F. Richard, and R. I. Crombie The Cuban crocodile, Crocodylus rhombifer, from Late Quaternary fossil deposits on Grand Cayman. Caribbean Journal of Science 29: Norell, M. A Cladistic approaches to paleobiology as applied to the phylogeny of alligatorids. Ph.D. dissertation, Yale University, New Haven, Connecticut, 272 pp. Norell, M. A., J. M. Clark, and J. H. Hutchison The Late Cretaceous alligatoroid Brachychampsa montana (Crocodylia): new material and putative relationships. American Museum Novitates 3116:1 26. Oaks, J. R A time-calibrated species tree of Crocodylia reveals a recent radiation of the true crocodiles. Evolution 65: Olson, S. L., G. K. Pregill, and W. B. Hilgartner Studies on fossil and extant vertebrates from San Salvador (Watlings) Island, Bahamas (West Indies). Smithsonian Contributions to Zoology 508:1 15. Paolillo, A., and O. J. Linares Nuevos cocodrilos Sebecosuchia del Cenozoico Suramericano (Mesosuchia: Crocodylia). Paleobiologia Neotropical 3:1 25. Pol, D., and J. E. Powell A new sebecid mesoeucrocodylian from the Rio Loro Formation (Palaeocene) of north-western Argentina. Zoological Journal of the Linnean Society 163:S7 S36 Prasad, G. V. R., and F. de Lapparent de Broin Late Cretaceous crocodile remains from Naskal (India): comparisons and biogeographic affinities. Annales de Paleontologie 88: Pregill, G. K Fossil amphibians and reptiles from New Providence Island, Bahamas. Smithsonian Contributions to Paleobiology 48: Ray, D. A., J. A. Dever, S. G. Platt, T. R. Rainwater, A. G. Finger, S. T. McMurry, M. A. Batzer, B. Barr, P. J. Stafford, J. McKnight, and L. D. Densmore Low levels of nucleotide diversity in Crocodylus moreletii and evidence of hybridization with C. acutus. Conservation Genetics 5: Riff, D., and O. A. Aguilera The world s largest gharials Gryposuchus: description of G. croizati n. sp. (Crocodylia, Gavialidae) from the upper Miocene Urumaco Formation, Venezuela. Pal aontologische Zeitschrift 82: Rodriguez, D., J. R. Cede~no-Vazquez, M. J. Forstner, and L. D. Densmore Hybridization between Crocodylus acutus and Crocodylus moreletii in the Yucatan Peninsula: II. Evidence from microsatellites. Journal of Experimental Zoology 309: Rodriguez, D., M. J. Forstner, P. E. Moler, J. A. Wasilewski, M. S. Cherkiss, and L. D. Densmore Effect of human-mediated migration and hybridization on the recovery of the American crocodile in Florida (USA). Conservation Genetics 12: Rossmann, T Studien an k anozoischen Krokodilen: 2. Taxonomische Revision der Familie Pristichampsidae Efimov (Crocodilia: Eusuchia). Neues Jahrbuch f ur Geologie und Pal aontologie, Abhandlungen 210: Scheyer, T. M., and J. W. Moreno-Bernal Fossil crocodylians from Venezuela in the context of South American faunas; pp in M. R. Sanchez-Villagra, O. A. Aguilera, and A. A. Carlini (eds.), Urumaco and Venezuelan Paleontology: The Fossil Record of the Northern Neotropics. Indiana University Press, Bloomington, Indiana. Scheyer, T. M., O. A. Aguilera, M. Delfino, D. C. Fortier, A. A. Carlini, R. Sanchez, J. D. Carrillo-Brice~no, L. Quiroz, and M. R. Sanchez- Villagra Crocodylian diversity peak and extinction in the late Cenozoic of the northern Neotropics. Nature Communications 4:1907. Simpson, G. G New reptiles from the Eocene of South America. American Museum Novitates 927:1 3. Steadman, D. W., R. Franz, G. S. Morgan, N. A. Albury, B. Kakuk, K. Broad, S. E. Franz, K. Tinker, M. P. Paterman, T. A. Lott, D. M. Jarzen, and D. L. Dilcher Exceptionally well preserved late Quaternary plant and vertebrate fossils from a blue hole on Abaco, the Bahamas. Proceedings of the National Academy of Sciences of the United States of America 104: Turner, A. H., andj. J. W. Sertich Phylogenetichistoryof Simosuchus clarki (Crocodyliformes: Neosuchia) from the Late Cretaceous of Madagascar.SocietyofVertebratePaleontologyMemoir10:

8 BROCHU AND JIM ENEZ-V AZQUEZ CUBAN MIOCENE CROCODYLIFORMS 1101 Varona, L. S Notas sobre los crocodilidos de Cuba y descripcion de una nueva especie des Pleistoceno. Poeyana (ser. A) 16:1 34. Varona, L. S Los crocodrilos fosiles de Cuba (Reptilia: Crocodylidae). Caribbean Journal of Science 20: Velasco, A., and J. Ayarzag uena Spectacled caiman Caiman crocodilus; pp in S. C. Manolis and C. Stevenson (eds.), Crocodiles: Status Survey and Conservation Action Plan, 3rd edition. Crocodile Specialist Group, Darwin, Northern Territory, Australia. Velez-Juarbe, J., and C. A. Brochu. In press. Eocene crocodyliforms from Seven Rivers, Jamaica: implications for Neotropical crocodyliform biogeography and the status of Charactosuchus Langston 1965; in R. W. Portell and D. P. Domning (eds.), The Eocene Fossil Site of Seven Rivers, Jamaica: Geology, Paleontology, and Evolutionary and Biogeographic Implications. Springer Verlag, Dordrecht, The Netherlands. Velez-Juarbe, J., and T. E. Miller First report of a Quaternary crocodylian from a cave deposit in northern Puerto Rico. Caribbean Journal of Science 43: Velez-Juarbe, J., C. A. Brochu, and H. Santos A gharial from the Oligocene of Puerto Rico: transoceanic dispersal in the history of a nonmarine reptile. Proceedings of the Royal Society of London B 274: Walker, A. D A revision of the Jurassic reptile Hallopus victor (Marsh), with remarks on the classification of crocodiles. Philosophical Transactions of the Royal Society of London, Series B 257: Weaver, J. P., D. Rodriguez, M. Venegas-Anaya, J. R. Cedeno-Vazquez, M. J. Forstner, and L. D. Densmore Genetic characterization of captive Cuban crocodiles (Crocodylus rhombifer) and evidence of hybridization with the American crocodile (Crocodylus acutus). Journal of Experimental Zoology 309A: Westgate, J. W Lower vertebrates from an estuarine facies of the middle Eocene Laredo Formation (Claiborne Group), Webb County, Texas. Journal of Vertebrate Paleontology 9: White, J. L., and R. D. E. MacPhee The sloths of the West Indies: a systematic and phylogenetic review; pp in C. A. Woods and F. E. Sergile (eds.), Biogeography of the West Indies: Patterns and Perspectives. CRC Press, Boca Raton, Florida. Willis, P. M. A., and B. S. Mackness Quinkana babarra, a new species of ziphodont mekosuchine crocodile from the early Pliocene Bluff Downs Local Fauna, northern Australia with a revision of the genus. Proceedings of the Linnean Society of New South Wales 116: Submitted July 30, 2013; revisions received September 26, 2013; accepted October 9, Handling editor: Randall Irmis. APPENDIX 1. Character-state codings for the Domo de Zaza crocodyliform used in the phylogenetic analysis in this study. Characters are based on Brochu et al. (2012). Codings are based on skull material only; teeth and osteoderms were disregarded. Typographical errors were noted by Scheyer et al. (2013) for Alligator thomsoni (character 97; should have been state 0 rather than 9) and two gryposuchine gavialoids (Piscogavialis jugaliperforatus and Gryposuchus colombianus, character 156; should not have been state 2 in either species). Scheyer et al. (2013) assigned state 1 to the gryposuchines for this character, but the correct state is actually 0.??????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????0?????????000?00?0?1010??0???????1?0? ?