A New Small Short-Snouted Dyrosaurid (Crocodylomorpha, Mesoeucrocodylia) From The Paleocene of Northeastern Colombia
A New Small Short-Snouted Dyrosaurid (Crocodylomorpha, Mesoeucrocodylia) From The Paleocene of Northeastern Colombia
A New Small Short-Snouted Dyrosaurid (Crocodylomorpha, Mesoeucrocodylia) From The Paleocene of Northeastern Colombia
To cite this article: Alexander K. Hastings , Jonathan I. Bloch , Edwin A. Cadena & Carlos A.
Jaramillo (2010) A new small short-snouted dyrosaurid (Crocodylomorpha, Mesoeucrocodylia) from
the Paleocene of northeastern Colombia, Journal of Vertebrate Paleontology, 30:1, 139-162, DOI:
10.1080/02724630903409204
ARTICLE
ABSTRACT—The fossil record of dyrosaurid crocodyliforms spans the Late Cretaceous to Middle Eocene of Africa, Asia,
Europe, North America, and South America. Prior to this study, specimens from South America have been limited to a few
fossils with only two taxa diagnosed. We describe a nearly complete skull and unassociated mandible of a new dyrosaurid,
Cerrejonisuchus improcerus gen. et sp. nov., from the Paleocene Cerrejón Formation of northeastern Colombia. The skull
of C. improcerus has relatively elongate supratemporal fenestrae and well-developed occipital tuberosities, both diagnostic
characteristics of Dyrosauridae. The rostrum of adult C. improcerus comprises 54–59% of the length of the skull, making it
the shortest snout of any known dyrosaurid. A cladistic analysis using 82 cranial and mandibular characters for all species
of Dyrosauridae known from crania yielded two most-parsimonious cladograms with C. improcerus as the sister taxon to
a clade including Arambourgisuchus, Dyrosaurus, Hyposaurus, Congosaurus, Rhabdognathus, Atlantosuchus, and Guarin-
isuchus. Only Chenanisuchus, Sokotosuchus, and Phosphatosaurus, all known only from Africa, are more primitive within
Dyrosauridae. Chenanisuchus from the Paleocene of Morocco, the only other known short-snouted dyrosaurid, is not closely
related to C. improcerus and a short-snouted condition appears to have evolved independently at least twice within Dyrosauri-
dae. Our analysis supports an African origin of Dyrosauridae with dispersals to the New World by the Late Cretaceous or
earliest Paleocene. The presence of C. improcerus, together with undescribed taxa from the Cerrejón Formation, suggests a
radiation of dyrosaurid crocodyliforms, possibly following the K-P boundary, in tropical South America.
139
FIGURE 1. World map of locations of known crocodyliform fossil material during the Paleocene. Circles represent locations ascertained as Pale-
ocene in age, squares represent locations that have been contested as possibly Late Cretaceous in age, and the star represents the new locality in
northeastern Colombia. Map of the Paleocene by Scotese, 2001.
four teeth and associated postcranials including a humerus, ulna, serving most cranial elements from the terminus of the
left femur, fibula, tibia, left and right pubi, 17 vertebrae, one rib, premaxillae to the occipital condyle, with many features pre-
and eight osteoderms; UF/IGM 32, a complete snout and partial served in three dimensions (Figs. 3 and 4). The posterior por-
orbital region. tion of the palatines, both pterygoids, both ectopterygoids, the
Diagnosis—Shorter snout, approximately 54–59% of the dor- internal choanae, prootic bones, and quadratojugals are missing
sal skull length, than that of all other known dyrosaurids. in UF/IGM 29 (Fig. 5). The lateral margin of the right quadrate
Also differs from all other dyrosaurids in having approxi- is missing, as is the posterior portion of the jugal.
matly 11 teeth in each maxillary, 8 of which are anterior
to the orbits. Further differs from all known dyrosaurids ex-
Description
cept Chenanisuchus in having a wide interfenestral bar that
is square-shaped in cross-section. Further differs from all General—Unless otherwise noted, all descriptions and mea-
other dyrosaurids except Phosphatosaurus (and possibly Aram- surements are taken from the holotype (UF/IGM 29; Figs. 3–5).
bourgisuchus) in having a reduced fourth premaxillary tooth. The snout is relatively narrow and comprises about 54–59% of
Further differs from Phosphatosaurus and Sokotosuchus in the dorsal skull length, measured from the tip of the snout to the
lacking a ‘festooned’ lateral margin of the snout in dorsal posterior margin of the parietal (Table 1). The lateral margins of
view. Further differs from Hyposaurus, Rhabdognathus, Atlanto- the snout remain parallel in dorsal view from the external nares
suchus, and Guarinisuchus in having a mediolaterally straight to anterior to the orbits (Fig. 3). Ornamentation consists of shal-
posterodorsal margin of the parietal. Further differs from low pits consistent on the dorsal surfaces (Fig. 3) as well as the
Chenanisuchus and Sokotosuchus in having well-developed oc- preserved lateral surfaces of the specimen (Fig. 4A), and is con-
cipital tuberosities. Further differs from Chenanisuchus in hav- sistent across cranial bone sutures.
ing ornamentation continuous across dorsal and lateral surfaces Cranial Openings—The external nares are mediolaterally oval,
with no interruption across sutures and orbits medially and dor- oriented dorsally, and surrounded exclusively by the premaxil-
sally placed, most closely approximating the orbit position of lae (Fig. 3). The external nares are located extremely anteriorly,
Dyrosaurus. Further differs from Hyposaurus rogersii in having with only a thin wall (6.57 mm) of the premaxillae around the
teeth with straight, rather than twisted anterior carinae. anterior edge. These conditions are consistent across all spec-
imens and are not likely the result of deformation. The ante-
Preservation rior lateral margins are angled toward the midline of the snout.
The lateral margins of the external nares are approximately level
Although the type specimen (UF/IGM 29) is slightly with the posterior and anterior margins, and do not slope ven-
dorsoventrally compressed, it is a nearly complete skull, pre- trally or dorsally (Figs. 3, 4A). Two low ventral ridges extend
142 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 1, 2010
FIGURE 3. Skull of Cerrejinosuchus improcerus, UF/IGM 29, from the Cerrejón coal mine of northeastern Colombia, middle–late Paleocene, in
dorsal view. A, photograph; B, sketch. Abbreviations: bo, basioccipital; en, external nares; eo, exoccipital; f, frontal; ifb, interfenestral bar; j, jugal;
l, lacrimal; lsp, laterosphenoid; m, maxilla; n, nasal; or, orbit; ot, occipital tuberosity; p, parietal; pm3, third premaxillary tooth; pm, premaxilla; po,
postorbital; prf, prefrontal; q, quadrate; sq, squamosal; stf, supratemporal fenestra; vt, ventral tubercle. Dotted line represents a suture which was not
clear. Scale bar equals 10 cm.
posteriorly from the anterior margin of the external nares, par- gins of the orbits are comprised of the lacrimal, with the jugal only
alleling what would have been the midline of the snout (Fig. 3). participating in the posterolateral corners. The medial margins of
At the posterior end of the external nares are two small tuber- the orbits are entirely comprised of the frontal bone. The poste-
cles on the ventral surface of the external nares projecting from rior margins of the orbits are slightly more than half comprised
the posterior wall (Fig. 3). These tubercles are also present, but of the postorbitals that have prominent anterolateral processes.
partly eroded, in UF/IGM 31 and 32 (Figs. 6 and 7), and therefore Less than half of the remaining posterior orbital margins are
the feature seems representative of the species. formed by the frontal (Fig. 3). The separation between the orbits
Due to anteroventral flattening, the incisive foramen has been by the frontal is approximately one-third the width of an orbit.
displaced laterally from the midline of the external nares (Fig. 5). Flattening has distorted much of the infratemporal fenestrae
The incisive foramen is entirely surrounded by the premaxillae. of the holotype (Fig. 4A). However, UF/IGM 31 has preserved
The incisive foramen is visible in dorsal view through the exter- the left infratemporal fenestra (Fig. 6). This fenestra is elongate,
nal nares as a small pit between the low anterior ridges (Fig. 3). anteroposteriorly longer than it is dorsoventrally wide. The pos-
In ventral view, the foramen has a triangular anteriorly directed torbitals constitute the anterodorsal portions of the fenestrae.
process placed medially over the foramen, between the pm1 teeth Along the ventral margins of the infratemporal fenestrae the
(Fig. 5). The entire foramen is level with the pm2 alveoli. jugals rise dorsally at a point nearly even with the postorbital-
The orbits of UF/IGM 29 are oval in outline and oriented an- squamosal joint and approximately halfway along the margin
terodorsally, near the midline (Figs. 3, 4A). The prefrontals par- (Figs. 4A, 6). The posterodorsal corner in UF/IGM 31 is bound
ticipate in the anteromedial corners of the orbits. The lateral mar- by the quadratojugal (Fig. 6).
HASTINGS ET AL.—NEW SHORT-SNOUTED DYROSAURID FROM COLOMBIA 143
FIGURE 4. Skull of Cerrejinosuchus improcerus, UF/IGM 29, from the Cerrejón coal mine of northeastern Colombia, middle–late Paleocene. A,
UF/IGM 29 in lateral view; B, UF/IGM 29 in occipital view; C, UF/IGM 29 sketch of occipital view. Abbreviations: bo, basioccipital; car, carotid
foramen; eo, exoccipital; fm, foramen magnum; fv, foramen vagi; oc, occipital condyle; ot, occipital tuberosity; p, parietal; prp, paroccipital process;
ptf, posttemporal fenestra; q, quadrate; so, supraocciptal; sq, squamosal. Both scale bars equal 10 cm.
TABLE 1. Snout proportions of all members of Dyrosauridae The supratemporal fenestrae are 51.8% longer than they
with material complete enough for skull length estimation. No- are wide and narrow anteriorly to 63.7% of the width across
tably absent is Rhabdognathus aslerensis, which lacks most of
the front of the snout needed to estimate dorsal skull length
the posterior fenestrae (Fig. 3). The anterior and posterior
as well as preorbital skull length. Congosaurus bequaerti is only margins of the supratemporal fenestrae are rounded, making the
known from the anterior portion of the skull (Jouve and Schwarz, fenestrae longitudinally ovular with a straight medial margin,
2004); however, dorsal skull length and body length were esti- and lateral margins that taper anteriorly toward the midline.
mated by Jouve et al. (2008). Abbreviations: DL, dorsal skull The fenestrae are bound anteromedially by the frontal. The
length; PreoL, preorbital skull length; R, ratio of preorbital skull postorbitals constitute the lateral halves of the anterior margins
length to dorsal skull length; TBL, estimated total body length
using method by Sereno et al. (2001). Citations are marked as well as the anterior halves of the lateral margins. The parietal
by numbered superscripts: 1Jouve et al., 2005a; 2an estimation forms approximately the medial three-quarters of the posterior
from Jouve et al., 2008; 3an estimation from figure 2 in Bar- margins of the supratemporal fenestrae. The posterior halves of
bosa et al., 2008; 4Jouve et al., 2005b; 5Jouve et al., 2006; 6Jouve the lateral margins, as well as the remaining one-quarter of the
et al., 2008. posterior margins, of the fenestrae are bound by the squamosal.
The posterior walls of the supratemporal fenestrae are weakly
R (PreoL/ sloped anteroventrally from the skull roof and are largely visible
DL PreoL DL) × TBL in dorsal view (Fig. 3). This feature may have been exaggerated
(cm) (cm) 100 (m) slightly from dorsoventral flattening, but a significant amount
of exposure is evident even in the less flattened specimens. The
Cerrejonisuchus improcerus 31.4 17.2 54.78 1.72–2.22
(UF/IGM 31) interfenestral bar is very wide, 24.60 mm at its widest and 16.71
Cerrejonisuchus improcerus 25.8 15.3 59.30 1.22–1.79 mm at its narrowest. The cross-section of the interfenestral bar is
(UF/IGM 29) rectangular in shape, and not T-shaped.
Chenanisuchus lateroculi1 57.6 36.5 63.37 3.57–4.24 The temporal canals have largely been lost. The right parietal-
Congosaurus bequaerti2 63 41 65.08 3.97–4.66 squamosal suture is incomplete and slightly separated with a sub-
Guarinisuchus muniz i3 47.1 30.8 65.41 2.79–3.43 tle widening of a crack (0.40 mm) at roughly the expected level
Hyposaurus rogersii1 42.9 28.2 65.73 2.48–3.11 of the temporal canal. This depression likely represents the cav-
Sokotosuchus ianwilsoni1 60.1 39.7 66.06 3.75–4.44 ity that was the temporal canal and is dorsolaterally bound by the
Phosphatosaurus 107 72 67.29 7.22–8.05 squamosal and medioventrally bound by the parietal.
gavialoides1 The posttemporal fenestrae are located dorsal to the occipi-
Arambourgisuchus 100 71.5 71.50 6.71–7.51 tal tuberosities and are oriented mediolaterally, and parallel to
khouribgaensis4
the skull roof (Fig. 4). The posttemporal fenestrae are typically
Dyrosaurus phosphaticus1 104 75 72.12 7.00–7.82
dorsoventrally short among dyrosaurids (Jouve, 2005), but due to
Dyrosaurus maghribensis5 89–97 — 73.00 5.89–7.28
taphonomic dorsoventral distortion, this characteristic has been
Rhabdognathus keiniensis1 73.1 53.4 73.05 4.72–5.44
further exaggerated. The fenestrae are bound dorsolaterally by
Atlantosuchus coupatez i6 104.5 83 79.43 7.04–7.86
144 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 1, 2010
FIGURE 5. Skull of Cerrejonisuchus improcerus, UF/IGM 29, from the Cerrejón coal mine of northeastern Colombia, middle–late Paleocene, in
ventral view. A, photograph; B, sketch. Abbreviations: bo, basioccipital; bsp, basisphenoid; car, carotid foramen; eo, exoccipital; fv, foramen vagi; if,
incisive foramen; j, jugal; m, maxilla; m1–9, first through ninth maxillary alveoli/teeth; mef, medial eustachian foramen; pal, palatine; pm, premaxillary;
pm1–4, first through fourth premaxillary alveloi/teeth; q, quadrate; sof, suborbital fenestra. Dotted lines represent features/sutures that were not clear.
Scale bar equals 10 cm.
the squamosals and dorsomedially by the parietal. The ventral of the medial margin, near the m9 alveolus (Fig. 5). The fenes-
surface is formed by the occipital tuberosites of the exoccipitals. tra reaches anteriorly to the level of the m8 alveolus. In addi-
The supraoccipital is not complete, but evidently had little to no tion, one-third of the anteromedial margin of the left suborbital
participation to the posttemporal fenestrae. fenestra is also present, where the palatine has been flattened
Much of the palatines have been crushed and/or lost, and both so that it is flush dorsally against the braincase. Posteriorly the
ectopterygoids are absent, thus little can be discerned of the sub- preservation of the left suborbital fenestra ceases with the pos-
orbital fenestrae (Fig. 5). A small portion of the medial and an- teromost preservation of the left maxilla, which is at the level of
terolateral margins of the right suborbital fenestrae remain. The the m11 alveolus. The suborbital fenestrae are narrow and slen-
preservation of the right suborbital fenestra includes the antero- der in shape and anteriorly acute. The suborbital fenestrae are
lateral margin formed by the maxilla, and ceases anteriorly at the very posteriorly placed, and their anterior margins do not reach
anteromedial corner, comprising about two-thirds of the anterior the level of the orbit.
margin of the fenestra. This margin is round and smooth, with a The foramen magnum has been crushed dorsoventrally, only
slight protrusion of the palatine into the fenestra in the middle leaving a shallow depression (Fig. 4B). The foramen is bound
HASTINGS ET AL.—NEW SHORT-SNOUTED DYROSAURID FROM COLOMBIA 145
FIGURE 6. Referred skull of Cerrejonisuchus improcerus, UF/IGM 31, from the Cerrejón coal mine of northeastern Colombia, middle–late Pale-
ocene, in dorsal view. A, UF/IGM 31 photograph; B, UF/IGM 31 sketch. Abbreviations: bo, basioccipital; en, external nares; eo, exoccipital; f, frontal;
ifb, interfenestral bar; itf, infratemporal foramen; j, jugal; l, lacrimal; m, maxilla; n, nasal; or, orbit; ot, occipital tuberosity; p, parietal; pm3, third
premaxillary tooth; pm, premaxilla; po, postorbital; prf, prefrontal; q, quadrate; qj, quadratojugal; sq, squamosal; stf, supratemporal fenestra; vert,
vertebra vt, ventral tubercle. Dotted lines represent sutures that were not clear. Specimen has a partial thoracic vertebra fused to the ventral side of
the skull as a result of deformation, visible also in dorsal view. Scale bar equals 10 cm.
laterally and dorsally by the exocciptals, comprising three- still adjacent to, their respective alveoli (Fig. 5). The premax-
quarters of the margin. The ventral margin is formed by the illae do not expand laterally with respect to the maxillae. The
basioccipital. premaxilla-maxilla suture is level with the pm3 alveoli on the
Premaxillae—The premaxillae are well preserved in UF/IGM dorsal and ventral surfaces (Figs. 4, 5). The posterodorsal process
29. The anterior end of the snout has been dorsoventrally extends posteriorly to the interalveolar space between the m1
compressed with slight warping toward the left side (Figs. 3, 5). and m2, and penetrates between the maxillae and nasal (Fig. 3).
This has resulted in five premaxillary teeth displaced from, but Four alveoli are present on each premaxilla. The left pm1 tooth
146 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 1, 2010
is notably shorter in length from the right (8.44 mm vs. 15.43 margins of the supratemporal fenestrae, with the posteromost
mm), likely due to different stages in replacement. The teeth of alveolus being just anterior to this level. In addition, the maxillae
pm1 are placed very close together (2.11 mm apart). A deep gap contribute posteriorly to the anteroventral portion of the lower
is present between the pm1 and pm2 alveoli, large enough for temporal bar (Fig. 4A). The alveoloar walls project ventrally,
lower dentition to occlude. These deep notches are visible on forming raised walls above the interalveolar space. A deep
the dorsal surface as slight indentations along the anteromost sulcus is located just anterior to the m1 alveolus (Fig. 5). Of
margin of the snout (Fig. 3). The pm3 alveoli are notably raised the preserved dentition, the m3 alveoli are by far the largest.
from the adjacent alveoli and much more robust (Fig. 4A). Interalveolar spaces are fairly consistent, and typically less
Measurements of alveolar size and spacing are provided in Table than one-third the length of the adjacent alveoli. Deep sulci
2. The premaxillae have shallow pitted ornamentation, with are present between the m4 and m5 alveoli, deeper than those
anterior ornamentation directed toward the midline. between the pm4 and m1. Smaller, yet still deep, occlusal pits
Maxillae—The maxillae of UF/IGM 29 comprise most of are found between the m3 and m4 alveoli as well as between
the snout. The left maxilla has 11 preserved alveoli, the right the m5 and m6 alveoli. The maxillary alveoli are mostly cir-
maxilla has 9 variably preserved alveoli (Fig. 5). The terminal cular in outline, but especially posteriorly they have a slight
end of the left maxilla represents the end of this bone poste- tendency toward an anteroposterior ovate shape. See Table 2
riorly where it would have fused with the ectopterygoid that for alveolar size and spacing. A small ridge is visible in ventral
was not preserved. An indentation is present just posterior view along the suture between the left and right maxillae
to the premaxilla-maxilla suture along the lateral snout mar- (Fig. 5). The maxillae are dorsally and laterally ornamented
gin, visible in both dorsal and ventral views, particularly on (Figs. 3, 4A).
the right side (Figs. 3, 5). The maxillae have smooth, even Nasal—The nasals are fused in UF/IGM 29 (Fig. 3) and
margins that gradually expand posterolaterally (Fig. 3). The UF/IGM 31 (Fig. 6), but are clearly unfused in UF/IGM 32 (Fig.
maxillae are dorsally separated by the nasal (Fig. 3). The two 7). As UF/IGM 29 and 31 are similarly sized, and UF/IGM 32 is
maxillae contact each other posteriorly until the level of the slightly smaller, this character may be related to ontogeny. The
interalveolar space between m4 and m5 where the maxillae are single nasal of UF/IGM 29 gradually widens from its anterior
medially separated by the palatines and extend posterolaterally. point to 12.99 mm at its posteromost contact with the premaxilla.
Posteriorly, the maxillae extend to the level of the posterior The nasal contacts the premaxillae dorsally at the level of the pm3
HASTINGS ET AL.—NEW SHORT-SNOUTED DYROSAURID FROM COLOMBIA 147
alveolus. Anteriorly, the nasal comes to a point near the narial Parietal—The interfenestral bar is slightly crushed at the con-
opening, with only a small space (20.46 mm) along which the pre- tact between the frontal and parietal, with the posterior margin of
maxillae fuse together before the posterior margin of the exter- the frontal being uplifted onto the anterior margin of the parietal
nal nares, such that the nasal does not participate in the posterior (Fig. 3). The parietal portion of the interfenestral bar is equally
margin of the external nares (Fig. 4). The width is relatively con- as thick as that portion of the frontal, and comprises 72% of the
stant; from this point of contact with the premaxilla (13.03 mm length of the interfenestral bar. The interfenestral bar is roughly
wide) until the first point of contact with the lacrimal (15.70 mm square in cross-section, with slight overhang on the dorsal sur-
wide), a 17% increase in width. The area where the nasal joins face. The thickness of the interfenestral bar is 16.71 mm at the
the frontal, lacrimal, and prefrontal is distorted, with the poste- closest point to the frontal that gradually thickens to 24.60 mm
rior end of the nasal overlapping the anterior end of the frontal at its base at the posterior margin of the supratemporal fenes-
bone. The posterior end of the nasal terminates at the level trae. The anteroventral portion of the parietal within the right
of the m5 alveoli. Due to lack of preservation, whether or not supratemporal fenestra has been pushed up relative to the rest
the nasal bifurcates into two posterior processes cannot be dis- of the parietal, overlaying its otherwise smoothly dipping sur-
cerned in UF/IGM 29. The posterior nasals are better preserved face and obscuring its suture with the laterosphenoid (Fig. 3).
in UF/IGM 31 and 32 and do not bifurcate (Fig. 7). The nasal The parietal overhangs anterodorsally onto the posterior mar-
does not contact the orbit in either UF/IGM 29 or UF/IGM 32. gin of the supratemporal fenestra. The overhang originates at
The ornamentation of the nasal is relatively shallow and uniform. the parietal-squamosal suture and deepens toward the midline,
Prefrontals—The size disparity between the prefrontals of where it merges with the very slight overhang of the interfen-
UF/IGM 29 and 31 is 47% (∼16.7 mm anteroposterior length estral bar. This is less pronounced than the overhang in the an-
in UF/IGM 29 and 34.8 mm in UF/IGM 31) despite only a teromedial corners of the supratemporal fenestrae by the frontal
17.8% difference in skull length (see Table 1). The prefrontals of bone. The parietal-quadrate suture is parallel with the skull roof,
UF/IGM 32 are incomplete, but appear more similar to the large and is not visible in dorsal view. The parietal-squamosal suture
condition seen in UF/IGM 31. In all specimens, both prefrontals extends laterally from the point of the posterior margin of the
have a straight contact with the frontal, parallel to the midline. supratemporal fenestra, and passes through the location of the
Lacrimals—The left lacrimal of UF/IGM 29 is more complete presumed temporal canal, slightly lateral to the mid-width of
than the right and bears a very prominent long and narrow ridge the supratemproal fenestra, and continues to extend ventrolat-
ascending from its contact with the left jugal. A portion of this erally. The right parietal-squamosal suture is relatively straight,
long, narrow ridge is present on the right side, but neither its without a zigzag pattern, but the left suture is too distorted to
contact with the jugal, prefrontal, or nasal is preserved. The be certain. The posterior portion of the parietal comprising the
prominent lateral ridge composing the lateral margin of the skull roof contacts the left and right squamosals and forms the
orbit widens posterolaterally. The lacrimal constitutes most of posterior margin of the supratemporal fenestrae is broad, thick,
the lateral margin of the orbit, with the jugal comprising only and flat anteroposteriorly and transversally straight (Fig. 3). The
the posteromost portion. The lacrimal expands anteriorly and parietal is dorsoventrally thin, slopes gently posteroventrally, and
curves medially toward the midline, constituting much of the is slightly exposed in occipital view (Fig. 4B). A small process
semicircular anterior margin of the orbit. The lacrimal extends projects ventrally, visible in posterior view, which has a raised
anteriorly to the level of the m4 alveolus. The left lacrimal’s midline and depression on either side such that it forms a thin
length of contact with the prefrontal, 14.65 mm, is greater than ridge projecting ventrally from the skull roof (Fig. 4B). The
that with the nasal, 10.72 mm. The ornamentation of the lacrimal parietal-supraoccipital suture is crescentic, and not ‘W-shaped.’
is limited to the lateral surface. The parietal is ornamented evenly across its dorsal surface, in-
Frontal—The single cruciform frontal of UF/IGM 29 is nearly cluding across the suture with the frontal on the interfenestral
complete, except where the frontal-nasal suture was not pre- bar (Fig. 3).
served (Fig. 3), and whether or not the frontal penetrates the Postorbitals—The right postorbital has better preservation
nasal is unclear. The frontal reaches farther than the prefrontals than the left and is essentially complete (Fig. 3). The postorbital
anteriorly (Figs. 3, 6). The frontal participates in the anterior constitutes the anterolateral margin of the supratemporal fen-
28% of the thick interfenestral bar. In cross-section, the anterior estra, and the posterior margin of the orbit (Fig. 3). The right
projection of the frontal is roughly triangular, with the point di- postorbital has a strong anterolateral ventrally directed process
rected ventrally, and is roughly as wide as the interfenestral bar. that contacts the jugal and marks the posterolateral corner of the
The frontal contributes to the margins of both orbits and both orbit. The postorbital-frontal contact occurs roughly at the mid-
supratemporal fenestrae. The bone comprises the entire medial width of the anterior dorsal margin of the supratemporal fenes-
margin of the orbit as well as the medial half of the posterior tra. The posterior extensions of the postorbitals extend to about
margin (Fig. 3). The suture with the postorbital is at the midline one-third the length of the supratemporal fenestrae from the pos-
of the orbit. The participation of the frontal to the posteromedial terior edge (Fig. 3). As the infratemporal fenestrae are incom-
corner of the orbit is a smooth, semicircular curve. The parietal- plete, the amount of participation of the postorbitals can only be
frontal contact occurs along the dorsal surface of the interfen- estimated, but is likely less than half of the dorsal margin (Figs.
estral bar and extends anterolaterally and ventrally through the 4A, 6), and whether or not it contacts the quadratojugal is uncer-
supratemporal fenestra. The contact with the laterosphenoid is tain. The postorbital bars have been crushed beneath the frontal
located anterior to the dorsal frontal-parietal suture. The an- and postorbitals, and little can be discerned of their thickness or
teromedial margin of the supratemporal fenestra formed by the relative participations of the postorbital and jugal. The postor-
frontal is slightly overhung. However, the interfenestral bar is bitals have similar ornamentation to the rest of the dorsal sur-
barely overhung. On the ventral surface of the anterior por- face of the skull and ornamentation is continuous across sutures
tion of the frontal is a shallow midline groove, visible near the (Fig. 3).
midline, indicating the pathway for the olfactory tract (Brochu Squamosals—The squamosal-parietal contact occurs at the
et al., 2002). The ornamentation of the frontal is shal- mid-width of the posterior margin of the supratemporal fenes-
lowly pitted, present at sutures, and oriented anteroposteriorly tra (Fig. 3). The squamosal-postorbital suture occurs roughly in
(Fig. 3). Two faint shallow grooves can be discerned extending the middle of the supraoccipital fenestrae. The squamosal nar-
just anterior to the supratemporal fenestral margin in an antero- rows slightly anteriorly toward the suture with the postorbital,
medial direction, ceasing anteriorly about level with the postero- and is thick posteriorly, along the posttemporal bar (Fig. 3).
medial corners of the orbits. The squamosal contacts the postorbital ventrally, participating in
148
TABLE 2. Alveolus dimensions and spacing for the upper dentition of the holotype, UF/IGM 29, and the lower dentition of the referred mandible, UF/IGM 30. Alveolus width measured
transversally at anteroposterior midpoint. Alveolus length measured from anteromost to posteromost points along alveolar rim. Width between left and right alveoli measured from left medial
anteroposterior midpoint to the right medial anteroposterior midpoint. Interalveolar length measured from posterior transverse midpoint to anterior transverse midpoint of adjacent alveolus.
Dashes are used when preservation is not sufficient for measurement. All measurements in mm.
Upper Dentition
Alveolus Position pm1 pm2 pm3 pm4 m1 m2 m3 m4 m5 m6 m7 m8 m9 m10 m11
Left Alveolus Width 4.81 5.59 6.36 — 4.80 3.38 9.71 2.76 — 6.06 6.75 5.58 5.59 2.36 4.56
Right Alveolus Width 5.89 4.80 10.21 1.91 6.82 4.39 9.25 2.13 4.78 5.47 6.92 10.34 5.28 — —
Left Alveolus Length 7.00 5.80 10.90 — 6.31 6.54 10.06 4.85 — 8.58 8.75 6.96 5.35 6.93 5.60
Right Alveolus Length 6.89 5.24 8.35 2.34 4.05 4.09 8.74 3.10 6.61 7.01 9.83 9.40 5.71 — —
Width between Left and 1.91 26.39 25.99 — 23.71 21.19 23.51 32.46 — 38.44 44.56 52.67 67.32 — —
Right alveoli
Interalveolar Position pm1–2 pm2–3 pm3–4 pm4–m1 m1–2 m2–3 m3–4 m4–5 m5–6 m6–7 m7–8 m8–9 m9–10 m10–11
Left Interalveolar Length 15.78 11.17 — — 10.09 12.51 8.09 — — 7.91 7.13 6.50 6.75 4.51
Right Interalveolar Length 11.34 11.97 10.80 17.16 12.08 13.82 7.58 14.33 7.09 6.38 3.19 3.4 — —
Lower Dentition
Alveolus Position d1 d2 d3 d4 d5 d6 d7 d8 d9 d10 d11 d12 d13
Left Alveolus Width — — 2.35 4.12 — — 4.80 7.10 2.67 3.42 4.17 4.20 —
Right Alveolus Width 5.49 — 3.91 5.06 — 2.04 — — — — 4.32 4.11 3.54
Left Alveolus Length — — 1.78 6.90 — — 4.49 8.03 3.31 2.84 3.94 5.07 3.52
Right Alveolus Length — 2.82 3.37 8.76 — 2.83 — — — — 4.83 4.62 4.23
Width between Left and 11.16 — 15.98 12.82 — — 14.90 17.17 — — 33.56 40.53 —
Right alveoli
Interalveolar Position d1–2 d2–3 d3–4 d4–5 d5–6 d6–7 d7–8 d8–9 d9–10 d10–11 d11–12 d12–13
Left Interalveolar Length — — 5.51 — — — 3.48 3.38 8.37 7.94 5.57 3.94
Right Interalveolar Length 3.87 12.00 3.88 — — — — — — — 5.31 3.92
HASTINGS ET AL.—NEW SHORT-SNOUTED DYROSAURID FROM COLOMBIA 149
the posterodorsal margin of the infratemporal fenestra (Fig. 4A). is arched dorsally, creating a smooth, concave outline in lateral
The lateral margin of the squamosal bears a faint groove that view.
does not flare anteriorly (Fig. 4A). The squamosal terminates Quadrates—The quadrates have been crushed dorsoventrally
posterodorsally in a short posteriorly directed process that is such that they are flat and parallel to the skull roof. The right
roughly level with the posterior edge of the occipital tuberosity quadrate is very well preserved, reaching as far anteriorly as
and the base of the occipital condyle (Fig. 3). The squamosal does the level of the postorbital-squamosal contact, and lies immedi-
not extend ventrally in occipital view, and its visibility in this view ately ventral to the squamosal throughout its length (Fig. 5). The
is limited to the dorsal-most portion (Fig. 4B). Ornamentation is quadrates are posteroventrally oriented and the cranioquadrate
pronounced near the parietal, but otherwise consistent across the canals are not preserved (Figs. 4A, 5). The condylar portion of
dorsal surface of the squamosal (Fig. 3). the left quadrate is lacking the lateral portion, leaving only the
Jugals—The anterolateral portion of the jugal of UF/IGM 29 medial-most one-third. The quadrate condyle is better preserved
is all that remains of the bone on the left side, from its contact in UF/IGM 31, which shows a partly eroded participation of the
with the lacrimal to the anterior three-quarters of its participa- quadratojugal to the condyle’s articulation with the lower jaw
tion in the infratemporal fenestrae (Fig. 4A). Less is preserved of (Fig. 6). The quadrates lack a crest at mid-width, which does not
the right jugal, only from the presumed contact with the lacrimal seem to be an artifact of preservation (Fig. 5).
to nearly the anterior half of the infratemporal fenestra (Fig. 3). Quadratojugals—The left quadratojugal is preserved in
The jugal of UF/IGM 29 only participates minimally in the pos- UF/IGM 31 only (Fig. 6), and comprises the posterodorsal mar-
terolateral corner of the orbit. The point where the maxillary- gin of the infratemporal fenestrae. The quadratojugal-quadrate
jugal and maxillary-lacrimal sutures meet is at the posterolateral suture is somewhat obscured, but includes the quadratojugal in
corner of the orbit (Fig. 3). The anteromost extent of the jugal the articulation with the lower jaw.
is roughly level with the posterior one-third of the orbit. The Palatines—The palatines of UF/IGM 29 have been flattened
lacrimal-jugal contact is slanted anteroventrally-posterodorsally. against the skull, and most of the posterior portions are missing,
There is no evidence of foraminae along the jugals, thus it seems but clearly expand posterolaterally. The anterior portions form
unlikely that UF/IGM 29 had a well-developed anterior jugal a sharp V between the maxillaries that reach anteriorly to the
siphonial foramen (as discussed for Rhabdognathus in Brochu interalveolar space between the m6 and m7 alveoli (Fig. 5). The
et al., 2002). palatine forms the medial margin of the suborbital fenestra as
Supraoccipital—The supraoccipital is recessed between the well as its anteromedial corner.
occipital tuberosities and only slightly contributes to the me- Laterosphenoids—Little identifiable material remains of the
dial portion of the tuberosities (Fig. 4B). The supraoccip- laterosphenoids. A small portion of them is visible in dorsal view
ital is not V-shaped, but instead more crescentic, bowing in the supratemporal fenestrae ventral and lateral to the parietal
inward toward the braincase. The supraoccipital is dorsally and frontal (Fig. 3). Much of the laterosphenoids in ventral view
bordered by the parietal and medioventrally by the exoc- are obscured by the palatines crushed upon them (Fig. 5). A small
cipital. The exoccipital separates the supraoccipital from the anterior portion is exposed, in which the sutures of the parietal
foramen magnum. The supraoccipital does not contribute to the and frontal are parallel with the skull roof. The laterosphenoid
posttemporal fenestrae or contact the squamosals. does not contact the squamosal.
Exoccipitals—The exoccipitals are mostly preserved, yet com- Basisphenoid—The ventral portion of the basisphenoid that
pressed dorsoventrally (Fig. 4B). The bones are present along includes the basisphenoid rostrum is missing (Fig. 5). The pos-
both sides of the occipital condyle, but displaced on the right terior branch of the medial eustachian foramen is visible in ven-
side, and too distorted to make accurate measurements. Exoc- tral view roughly 2 mm anterior to the ventral-most point of the
cipital participation to the width of the occipital condyle may be basisphenoid that surrounds it entirely. The medial eustachian
estimated from the better-preserved left side to be about 36.8% foramen has a diameter of 1.41 mm.
dorsally and 26.3% ventrally. The exoccipital participates to the Mandible—A fairly complete mandible, UF/IGM 30
dorsal three-quarters of the foramen magnum. Dorsally, the ex- (Fig. 8), was recovered approximately 740 m from the site
occipital comprises the majority of the occipital tuberosity, which where UF/IGM 29 was discovered, within the same strati-
is well developed, rounded, and directed posteriorly (Fig. 3, 4B). graphic level. Based on its size, proportions, and number of
The exoccipital extends laterally to constitute the ventral por- teeth corresponding to the upper dentition, it is referred to
tion of the paraoccipital process. Much of the basioccipital tubera Cerrejonisuchus improcerus gen. et sp. nov. UF/IGM 30 likely
have been crushed and lost, and participation of the exoccipitals represents a second individual due to the distance between fossil
cannot be estimated (Fig. 5). The left foramen vagi (X–XI), that localities and that the mandible is slightly smaller than would
houses the vagus (X) and accessory nerves (XI) as well as the be expected for UF/IGM 29. The preservation of the mandible
jugular vein (Brochu et al., 2002) is preserved, as well as the right is very similar to that of UF/IGM 29; it has been dorsoventrally
carotid foramen (Fig. 5). The foramen vagi is located along the flattened toward the left lateral direction. Due to dorsoventral
lateral margin of the base of the ocipital condyle and is directed flattening, the height and width of the symphysis and splenials is
ventroposteriorly. The exoccipital forms the entire ventral sur- too compressed for meaningful description or comparison.
face of the posttemporal fenestra. The specimen includes most of both dentaries and much of
Basioccipital—The basioccipital forms the bulk of the occipital the splenials, but is missing the articular, angular, and surangu-
condyle and extends ventrally, forming the basioccipital tubera, lar bones (Fig. 8). The d3–13 alveoli are present on the left den-
which have mostly been lost in UF/IGM 29 (Figs. 4B, 5). The tary, and the d2–13 alveoli on the right dentary. The d1 alveoli
basioccipital tubera are oriented posterolaterally relative to the on both dentaries are damaged; however, only a small part of the
midline, forming a V-shape in ventral view (Fig. 5). The tubera anteromost mandible is missing. The anteromost end of the right
are of indeterminate shape in occipital view (Fig. 4B). There is dentary suggests that the d1 alveolus for this tooth would have
a relatively wide, flat rugosity extending along the anteroventral faced anterodorsally, not directly dorsal. The anterior portion of
midline of the basioccipital that is more ventral than the lateral the mandible is not expanded, but rather continuous along most
edges. The ventral surface of the occipital condyle lacks any dis- of the length of the symphysis (Fig. 8). The tooth of the d2 alve-
tinct grooves paralleling this rugosity (Figs. 4B, 5). Due to poor olus on the right dentary is flattened, but reduced relative to the
preservation in this region, it is difficult to determine how much space where the d1 alveolus would have existed. Measurements
the basioccipital extends ventrally from the occipital condyle. The of the alveolar diameter and interalveolar spaces of UF/IGM 30
area between the occipital condyle and the basioccipital tubera are compiled in Table 2. The d4 alveoli are the largest in the lower
150 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 1, 2010
jaw and have dorsally higher alveolar walls. The dentition corre- d6–8 teeth, a misplaced tooth that may be referred to the right
sponds to the upper jaw of UF/IGM 29, such that the enlarged d4 d9 alveolus, worn left d10–13, and worn right d11–12 (Fig. 8).
teeth could occlude in the large interalveolar spaces between the The maxillary dentition of UF/IGM 29 is relatively homodont.
p4 and m1 teeth. The large diastema between the d2 and d3 teeth The only significant deviations in size are the enlarged m3
would correspond well with the enlarged pm3 teeth. The m3 teeth alveoli and the reduced m4 alveoli (Fig. 5). Their general shape
would occlude well in the large interalveolar space between the is conical and labiolingually compressed. The preserved teeth
d6 and d7 alveoli. The d7 alveoli are only slightly reduced relative have a relatively rounded apex. The teeth are defined labially
to d8. The spreading angle of the dentaries measured at the labial and lingually by strongly developed anterior and posterior
midpoint of the left and right d13 alveoli to the anterior termi- carinae. Striae are not discernable on either surface of the teeth.
nus of the splenials is approximately 40.7◦ . Ornamentation along Crenulations are present on at least the lingual side of the crown
the lateral and ventral surfaces is weak. The splenials fuse along of the left m7 tooth, but preservation for the other teeth is too
the symphysis and taper anteriorly to a wedge between the two poor to assess this feature on the other maxillary teeth. The teeth
dentaries The splenials end anteriorly at the interalveolar spaces do not increase or decrease in size posteriorly based on alveolar
between d6 and d7 (Fig. 8). The splenials extend posteriorly be- diameter, but without posterior teeth, it is difficult to be certain.
yond the last tooth at d13, as seen on the left dentary. The level The premaxillary tooth roots are more narrow and comparable
at which the symphysis ends posteriorly is estimated to be level to the apex width. In general the premaxillary teeth are thinner
with the d9 teeth. The left d13 alveolus supports the final tooth and longer than those of the maxillary. The maxillary tooth roots
in the lower dentition. The anterior alveoli are oriented dorsally are wide relative to the crowns. The teeth do not curve distally.
and slightly anteriorly, grading to more dorsally directed alveoli No preserved teeth possessed a ‘twisted’ carina.
and teeth toward the posterior end of the jaw.
Dentition—Seven teeth were found in association with the Comparison
maxilla, five of which were essentially preserved in place: the left
and right m4, as well as the left m6–8 (Fig. 5). The premaxilla The elongate supratemporal fenestrae, posteriorly directed
has preserved the right and left pm1 teeth, right and left pm3 occipital tuberosities formed by the exoccipitals, strong par-
teeth, and fragments of the right and left pm4. The posterior ticipation of the exoccipitals to the occipital condyle, and
maxillary dentition is not preserved. UF/IGM 30 possesses an the reduction of the seventh mandibular tooth and close
extremeley crushed and distorted right d2 tooth, very worn left placement to the eighth are major qualifying characters that unite
HASTINGS ET AL.—NEW SHORT-SNOUTED DYROSAURID FROM COLOMBIA 151
Dyrosauridae (Jouve et al., 2006). The presence of all these char- gavialoides (Buffetaut, 1978, 1979), Hyposaurus rogersii (Den-
acters in all preserved specimens indicates that Cerrejonisuchus ton et al., 1997), and Congosaurus bequaerti (Jouve and Schwarz,
improcerus belongs to the family Dyrosauridae. 2004). Rhabdognathus aslerensis (Brochu et al., 2002), Atlanto-
Snout proportions vary among Dyrosauridae. UF/IGM 29 and suchus coupatez i (Jouve et al., 2008), and Sokotosuchus ian-
31 possess the shortest snout in proportion to the dorsal skull wilsoni (Buffetaut, 1979) have entirely unfused nasal bones,
length of all known dyrosaurids (Table 1), at 59.30% and 54.78%, whereas Chenanisuchus lateroculi has an anteriorly fused nasal
respectively. Chenanisuchus lateroculi (Jouve et al., 2005a) has bone. The anterior portion of the preserved Rhabdognathus
the second shortest snout within Dyrosauridae, at 63.37%. keiniensis is fused; however, the posterior portion is unpreserved
The incisive foramen is described for Dyrosaurus maghribensis (Jouve, 2007).
as being heart-shaped and the posterior border as being com- Posteromedially, the frontal participates in the thick in-
prised of the maxillae, as opposed to V-shaped and entirely sur- terfenestral bar, contrasting the thin bar of A. khouribgaen-
rounded by the premaxillae, as in C. improcerus (Figs. 3, 5). The sis (Jouve et al., 2005b). The frontal participates 28.5% to
orbits of C. improcerus are placed much closer to the midline than the interfenestral bar in UF/IGM 29, as opposed to 20%
those of Chenanisuchus lateroculi and the interorbital space is in A. khouribgaensis (Jouve et al., 2005b) and Atlantosuchus
relatively small, more like the condition seen in D. phosphaticus. coupatez i (Jouve et al., 2008), 25% of D. phosphaticus (Jouve,
The orbits are more oval in shape than the circular shape of D. 2005) and Rhabdognathus keiniensis (Jouve, 2007), around
phosphaticus (Jouve, 2005) and Guarinisuchus muniz i (Barbosa 29% for Phosphatosaurus gavialoides (Buffetaut, 1978), 33%
et al., 2008). The supratemporal fenestrae of C. improcerus (Fig. of Chenanisuchus lateroculi (Jouve et al., 2005a) and Rhab-
3) are similar to all other known dyrosaurids in being much dognathus aslerensis (Brochu et al., 2002), and roughly 50%
larger than the orbits (Buffetaut, 1979) and longer than they are in Hyposaurus rogersii (Troxell, 1925). UF/IGM 29 has a
wide (Jouve et al., 2005b). The separation of the supratemporal wide interfenestral bar, 16.71 mm at its minimum width, very
fenestrae of C. improcerus is wide and thick, unlike the thin similar to that of C. lateroculi with a value of 17 mm at its mini-
“sagittal crest” of Rhabdognathus aslerensis (Brochu et al., 2002) mum width (Jouve et al., 2005a).
and Guarinisuchus muniz i (Barbosa et al., 2008). The cross- The frontoparietal suture of C. improcerus is much straighter
section of the interfenestral bar is much more consistent with the (Fig. 3) than the zigzag pattern described for D. phosphaticus
square shape of Chenanisuchus lateroculi (Jouve et al., 2005a) (Jouve, 2005). C. improcerus lacks the V-shaped groove on the
than with the ‘T-shape’ of Dyrosaurus (Jouve, 2005; Jouve et al., dorsal surface (see Fig. 3) as in Rhabdognathus aslerensis (Brochu
2006) and Atlantosuchus coupatez i (Jouve et al., 2008), or the et al., 2002; Jouve, 2007), and also lacks the deep sulcus and dor-
smooth and narrow condition of Phosphatosaurus gavialoides sal bulges at the posterior end when in occipital view (Fig. 3B).
(Buffetaut, 1978; Moody and Buffetaut, 1981), or the ‘V-shape’ C. improcerus also lacks the strong indentation of the posterior
of Rhabdognathus keiniensis (Jouve, 2007). The posterior walls margin (see Fig. 3) of the parietal as in Rhabdognathus (Jouve,
of the supratemporal fenestrae are visible in dorsal view in C. 2007), Hyposaurus rogersii (Denton et al., 1997), Atlantosuchus
improcerus as well as in Rhabdognathus (Jouve, 2007), but not coupatez i (Jouve et al., 2008), and Guarinisuchus muniz i (Bar-
in Dyrosaurus (Jouve, 2005). The lateral arches of the supratem- bosa et al., 2008). C. improcerus does not have the W-shaped
poral fenestrae are robust, not thin as in Rhabdognathus (Jouve, parietal-supraoccipital suture, seen in occipital view of D. phos-
2007). phaticus (Jouve, 2005), making it more similar to the genus Rhab-
Cerrejonisuchus improcerus, like other dyrosaurids, possesses dognathus (Jouve, 2007), which also lacks this character.
four premaxillary alveoli. The pm4 alveolus of UF/IGM 29 The supraoccipital of C. improcerus does not display a
is greatly reduced, whereas in Dyrosaurus and C. bequaerti V-shape (Fig. 4B) as in Dyrosaurus (Jouve, 2005; Jouve
the pm1, pm2, and pm4 alveoli have similar diameters (Jouve et al., 2006), Chenanisuchus lateroculi (Jouve et al., 2005a), and
and Schwarz, 2004; Jouve, 2005; Jouve et al., 2006). Jouve Arambourgisuchus khouribgaensis (Jouve et al., 2005b).
et al. (2005b) inferred that A. khouribgaensis also has a reduced The occipital tuberosities are mostly comprised of the exoccipi-
pm4 tooth. tals, but medially of the supraoccipital in C. improcerus (Fig. 4B),
In dyrosaurids such as Phosphatosaurus gavialoides and are posteriorly directed as in all dyrosaurids (Jouve et al.,
(Buffetaut, 1978 and 1979) and Sokotosuchus ianwilsoni 2006). The occipital tuberosities of C. improcerus seem at least as
(Buffetaut, 1979), the lateral margin of the snout is described well developed as the ‘moderately developed’ state of D. phos-
as ‘festooned,’ referring to the undulations in the maxillae and phaticus (Jouve, 2005); they are much more developed than in S.
premaxillae corresponding to alveoli and interalveolar spaces. ianwilsoni and C. lateroculi (Jouve et al., 2005a), and most similar
This character is entirely absent in C. improcerus (Fig. 3); the to the length and shape of R. keiniensis and R. aslerensis (Jouve,
margins are very smooth, as in Chenanisuchus lateroculi (Jouve 2007).
et al., 2005a), Dyrosaurus phosphaticus (Jouve, 2005), Hy- The lateral exoccipital forms the ventral portion of the
posaurus rogersii (Denton et al., 1997), and Arambourgisuchus paraoccipital process of C. improcerus (Figs. 3, 4B), as in Dy-
khouribgaensis (Jouve et al., 2005b). The total maxillary tooth rosaurus phosphaticus (Jouve, 2005) and Dyrosaurus maghriben-
count of C. improcerus is 11 teeth per maxilla. This number is sis (Jouve et al., 2006) and possibly other dyrosaurids as
more similar to estimates of 12 for Hyposaurus rogersii (Denton well. The ventral exoccipital contributes one-third to the
et al., 1997), 13 for Chenanisuchus lateroculi (Jouve et al., 2005a), basioccipital tuber in D. phosphaticus (Jouve, 2005), and
and 13–14 for Sokotosuchus ianwilsoni (Buffetaut, 1979) and less than one-fourth in Chenanisuchus lateroculi (Jouve et
Guarinisuchus muniz i (Barbosa et al., 2008), than 15 for Phos- al., 2005a). The participation to the basiocciptial tuber of
phatosaurus gavialoides (Buffetaut, 1978), 16 for Congosaurus the better-preserved left exoccipital of UF/IGM 29 is 26%,
bequaerti (Jouve and Schwarz, 2004), 17 for Arambourgisuchus more consistent with the condition of C. lateroculi (Jouve
khouribgaensis (Jouve et al., 2005b), 19–22 for Dyrosaurus et al., 2005a). Similar participation to the basioccipital tu-
(Jouve et al., 2006), or 23–24 of Rhabdognathus keiniensis bera is also noted in Rhabdognathus aslerensis (Brochu
(Jouve, 2007). et al., 2002). Both C. improcerus and D. phosphaticus (Jouve,
The nasals are entirely fused in UF/IGM 29 and 31, but un- 2005) exhibit an arched interspace between the occipital tuberos-
fused in the smaller UF/IGM 32 (Fig. 7). The two conditions for ity and the basioccipital tubera.
nasal fusing suggest the suture fuses with ontogeny in C. impro- No crest is visible on the ventral face of the quadrate of C.
cerus. The fused condition is consistent with other dyrosaurids improcerus, which does not seem to be an artifact of preservation
such as Dyrosaurus phosphaticus (Jouve, 2005), Phosphatosaurus (Fig. 5). The lack of crest is dissimilar to A. khouribgaensis (Jouve
152 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 1, 2010
et al., 2005b), Rhabdognathus aslerensis (Brochu et al., 2002), and ing lacked striations on at least the labial surfaces (Jouve et al.,
Rhabdognathus keiniensis (Jouve, 2007), which all possess a crest 2005a), as in C. improcerus. Hyposaurus rogersii has striations
on the ventral side of the quadrate. on the lingual surface (Denton et al., 1997) and Hyposaurus der-
Dyrosaurus phosphaticus, like C. improcerus (Fig. 5), has a bianus possesses longitudinal ridges on the lingual face and a
concavity at the middle of the medial margin of the suborbital smooth condition on the exterior of its mandibular teeth (Cope,
fenestra, as a result of a restriction along the mid-length of the 1886). Arambourgisuchus khouribgaensis (Jouve et al., 2005b)
palatine (Jouve, 2005). was described as having a somewhat intermediate form of vari-
The anterior dentition of the dentary of C. improcerus is ably present and either absent or weak superficial striae. The
similar to that of Arambourgisuchus khouribgaensis (Jouve et teeth of C. improcerus do not curve distally, as in Dyrosaurus
al., 2005b), Hyposarus rogersii (Troxell, 1925), and Dyrosaurus maghribensis (Jouve et al., 2006), Atlantosuchus coupatez i (Jouve
(Jouve, 2005; Jouve et al., 2006) in that the d2 alveoli are et al., 2008), Guarinisuchus muniz i (Barbosa et al., 2008), and
reduced in size and the interalveolar spaces between the d2 Phosphatosaurus gavialoides (Buffetaut, 1978), nor are they
and d3 are greater than between d1 and d2, as well as d3 slightly recurved as in Sokotosuchus ianwilsoni (Buffetaut, 1979)
being placed very close to an enlarged d4 (Fig. 8). Aram- and H. derbianus (Cope, 1886). The ‘twisted’ carinae characteris-
bourgisuchus khouribgaensis also has the raised d4 alveolar tic of H. rogersii (Denton et al., 1997) was not noticed in any of
wall found in C. improcerus (Jouve et al., 2005b). The ante- the preserved teeth.
rior orientation of the d1 alveolus of C. improcerus is sim-
ilar to that of Hyposaurus derbianus (Cope, 1886) and un- PHYLOGENETIC RELATIONSHIPS
like that of Dyrosaurus phosphaticus, which has a more dorsal
orientation. The anterior portion of the C. improcerus mandible Five phylogenetic analyses at the genus level of the fam-
is not expanded, and thus not spatulate as in some Hyposaurus ily Dyrosauridae have been published previously (Buffetaut,
(Jouve, 2007), but more similar to H. derbianus (Cope, 1886). 1978; Jouve, 2005; Jouve et al., 2005b; Barbosa et al., 2008; Jouve
The mandibular symphysis ends posteriorly in UF/IGM 30 et al., 2008). Buffetaut (1978) included Phosphatosaurus, Rhab-
at approximately the level of the d9 alveolus (Fig. 8), which is dognathus, Dyrosaurus, and Hyposaurus in a manual cladogram.
dissimilar to the d12 of Hyposaurus rogersii reported by Jouve The most basal of which was Phosphatosaurus, with Hyposaurus
(2007). This condition of the mandibular symphysis contrasts and Dyrosaurus being more derived, and Rhabdognathus falling
more strongly with that of other dyrosaurids, which have longer in between. Jouve (2005) conducted an analysis of 12 characters
mandibles, such as the d14 alveolus in Guarinisuchus muniz i across five genera of Dyrosauridae. This analysis was expanded
(Barbosa et al., 2008), the d15 in Hyposaurus sp. (Jouve, 2007), by Jouve et al. (2005b) to 30 characters and nine members of
and Hyposaurus derbianus (Cope, 1886), the d16 in Congosaurus Dyrosauridae. Jouve et al. (2008) further expanded this analy-
bequaerti (Jouve and Schwarz, 2004) and Arambourgisuchus sis to 43 characters and 11 species of Dyrosauridae. These stud-
khouribgaensis (Jouve et al., 2005b), the d17 in Dyrosaurus ies found Sokotosuchus and Phosphatosaurus to be basal gen-
(Jouve, 2005), and beyond the d19 alveolus in Rhabdognathus era of the family, supporting the subfamily Phosphatosaurinae,
(Jouve, 2007). which at the time it was named included only these two genera
The splenials end anteriorly at the interalveolar space be- (Buffetaut, 1980). When Chenanisuchus was included, it was
tween the d6 and d7 alveoli in C. improcerus (Fig. 8), sim- basal to all other members of Dyrosauridae (Jouve et al., 2005b;
ilar to that of Hyposaurus rogersii for which they end at Jouve, 2008). Barbosa et al. (2008) modified the 30 charac-
the d7 alveolus (Jouve, 2007). This splenial-alveolus relation- ters of Jouve et al. (2005b) and added their new taxon, Guar-
ship contrasts that of other dyrosaurids with more poste- inisuchus muniz i. This study also found a basal placement of
rior positions such as the interalveolar space between the Chenanisuchus, Sokotosuchus, and Phosphatosaurus, and placed
d9 and d10 alveoli in Congosaurus bequaerti (Jouve and Guarinisuchus as sister to a clade uniting Arambourgisuchus and
Schwarz, 2004) and Phosphatosaurus gavialoides (Buffetaut, Rhabdognathus.
1978), the interalveolar space between d10 and d11 in Arambour- New cranial specimens of Cerrejonisuchus improcerus allow
gisuchus khouribgaensis (Jouve et al., 2005b) and Hyposaurus for a reevaluation of the phylogenetic affinities of South Ameri-
derbianus (Cope, 1886), and d12 in Dyrosaurus (Buffetaut, can dyrosaurids. To help resolve the relationship of C. improcerus
1978). to other dyrosaurids, we performed a cladistic analysis of 82 cra-
The overall dentary tooth estimation for H. rogersii is as nial and mandibular characters (and one postcranial) for 13 dy-
high as 17–20 (Buffetaut, 1980) but as low as 15 (Denton rosaurid species and three outgroup taxa (Appendices 1 and 2).
et al., 1997), somewhat comparable to the 13 of C. improcerus. Early Cretaceous Sarcosuchus imperator (Sereno et al., 2001) and
The size of the d7 alveolus is much more comparable to the d8, Elosuchus cherifiensis (de Lapparent de Broin, 2002) from Africa
contrary to the very reduced d7 alveolus relative to the d8 char- and Late Cretaceous Terminonaris robusta (Wu et al., 2001) from
acter traditionally used as a synapomorphy of Dyrosauridae. C. North America were used as outgroup taxa. The cladistic analysis
improcerus lacks the ‘festooned’ appearance (Fig. 8) of H. roger- was rooted with Sarcosuchus imperator because it is the geolog-
sii along the lateral margins of the dentaries. Due to the flattening ically oldest mesoeucrocodylian thought to be closely related to
of UF/IGM 30, the diagnostic characters of width as compared to dyrosaurids but clearly not in the ingroup (Sereno et al., 2001;
height of the lower jaw cannot be discerned. Jouve et al., 2005b), and because its cranium has been thor-
The apices of the crowns of the preserved teeth of UF/IGM oughly described (Sereno et al., 2001). We included only mem-
29 are more rounded than acute, as in C. bequaerti (Jouve bers of Dyrosauridae known from at least somewhat complete
and Schwarz, 2004). The teeth of C. improcerus possess well- cranial material in this analysis. The known dyrosaurid species
developed anterior and posterior carinae as in C. bequaerti that were not included due to lack of cranial material are the
(Jouve and Schwarz, 2004), D. phosphaticus (Jouve, 2005), D. Cretaceous- or Paleocene-aged Hyposaurus derbianus of Brazil
maghribensis (Jouve et al., 2006), and H. derbianus (Cope, 1886), (Cope, 1885, 1886), the Paleocene dyrosaurids Hyposaurus pau-
defining the labial and lingual surfaces. Unlike Congosaurus be- cidens from Morocco (Arambourg, 1952; Jouve, 2007) and Con-
quaerti (Jouve and Schwarz, 2004), Dyrosaurus (Jouve, 2005; gosaurus compressus from Mali (Jouve, 2007), and the Eocene
Jouve et al., 2006), and Atlantosuchus coupatez i (Jouve et al., dyrosaurid Rhabdognathus acutirostris from Tunisia (Buffetaut,
2008), C. improcerus lacks noticeable striations on either face 1978). All characters except 26, 29, 30, 32, and 53 (continuous,
of the teeth, but instead possesses apical lingual crenulations on multistate characters; e.g., location of the posteromost alveolus
the left m7 tooth. Chenanisuchus lateroculi was noted as hav- relative to the orbit/supratemporal fenestrae: character 30) were
HASTINGS ET AL.—NEW SHORT-SNOUTED DYROSAURID FROM COLOMBIA 153
Despite the lack of sufficient cranial material, species not in- South American dyrosaurids are now known from the Pale-
cluded in the phylogenetic analysis have biogeographic relevance ocene of Bolivia, the Early Paleocene of Brazil, and the middle
to the current study (Fig. 10). If Hyposaurus derbianus is at least or late Paleocene of northeastern Colombia, as well as possibly
considered a valid member of the genus Hyposaurus, its pres- from the Late Cretaceous in Bolivia and Brazil. The most com-
ence in Brazil (Cope, 1886) implies a radiation of the genus Hy- plete material prior to this study is from the Pernambuco region
posaurus into the New World during the Late Cretaceous. Fur- of Brazil (Cope, 1885, 1886; Barbosa et al., 2008), Hyposaurus
thermore, the presence of Hyposaurus paucidens in Morocco, as derbianus and G. muniz i. The genus Hyposaurus is best known
well as other indeterminate Hyposaurus material (Jouve, 2007), from Late Cretaceous and Paleocene deposits of eastern North
implies a form of Hyposaurus remained in Africa into the Pa- America, all referred to the species Hyposaurus rogersii (Denton
leocene. Neither Congosaurus compressus nor Rhabdognathus et al., 1997). Hyposaurus is also known from much more fragmen-
acutirostris would imply drastically different paleobiogeography tary remains from the Paleogene of Africa (Jouve, 2007).
because both genera are already known from Africa (Jouve, Considering the temporal placement of the New World taxa
2007). (see Fig. 10), we suggest three independent dispersals from Africa
The monophlyly of Dyrosauridae is supported by 12 unam- to the New World. Presence of C. improcerus together with a
biguous synapomorphies in the current study. Presence of all number of undescribed taxa from the Paleocene of northeast-
basal dyrosaurid taxa, together with two of three outgroups, in ern Colombia suggests that there was a radiation of dyrosaurid
Africa strongly supports previous paleobiogeographic hypothe- crocodyliforms, possibly following the K-P boundary, in tropical
ses for an African origin of Dyrosauridae (Jouve et al., 2005b, South America. Barbosa et al. (2008) propose a dispersal from
2008). Africa to Brazil with continued immigration into North America.
HASTINGS ET AL.—NEW SHORT-SNOUTED DYROSAURID FROM COLOMBIA 155
Hastings, A., and J. Bloch. 2008. New long-snouted dyrosaurid (Crocody- Storrs, G. W. 1986. A dyrosaurid crocodile (Crocodylia: Mesosuchia)
lomorpha, Mesoeucrocodylia) from the Paleocene of Northeastern from the Paleocene of Pakistan. Postilla Number 197:1–16.
Colombia. Journal of Vertebrate Paleontology 28:89A. Swinton, W. E. 1930. On the fossil Reptilia from Sokoto Province. Geo-
Hay, O. P. 1930. Second bibliography and catalogue of the fossil Verte- logical Survey of Nigeria 13:1–62.
brata of North America. Vol. 2. Carnegie Institute of Washington Swinton, W. E. 1950. On Congosaurus bequaerti Dollo. Annales du
Publication 2:1–1074. Musée du Congo Belge 13:9–56.
Head, J., J. Bloch, A. Hastings, J. Bourque, E. Cadena, F. Herrera, Swofford, D. 2003. PAUP∗ 4.0b10. Sinauer Associates, Sunderland, Mas-
P. Polly, and C. Jaramillo. 2009. Giant boid snake from the Pale- sachusetts.
ocene neotropics reveals hotter past equatorial temperatures. Na- Troxell, E. L. 1925. Hyposaurus, a marine crocodilian. American Journal
ture 457:715–718. of Science 8:489–514.
Jaramillo, C., A. Pardo-Trujillo, M. Rueda, G. Harrington, G. Bay- Vickaryous, M. K., and B. K. Hall. 2008. Development of the dermal
ona, V. Torres, and G. Mora. 2007. Palynology of the Upper Pale- skeleton in Alligator mississippiensis (Archosauria, Crocodylia) with
ocene Cerrejón Formation, Northern Colombia. Palynology 31:153– comments on the homology of Osteoderms. Journal of Morphology
189. 269:398–422.
Jouve, S. 2004. Etude des Crocodyliformes fini Crétacé-Paléogène Walker, A. D. 1970. A revision of the Jurassic Reptile Hallopus victor
du Bassin des Oulad Abdoun (Maroc) et comparaison avec (Marsh), with remarks on the classification of crocodiles. Philosoph-
les faunes africaines contemporaines: systématique, phylogénie et ical Transactions of the Royal Society of London Series B, Biologi-
paléobiogéographie. Unpublished Ph.D. dissertation, Muséum Na- cal Sciences 257:323–372.
tional d’Histoire Naturelle, Paris, 651 pp. Whetstone, K. N., and P. J. Wybrow. 1983. A ‘‘cursorial’’ crocodilian
Jouve, S. 2005. A new description of Dyrosaurus phosphaticus (Thomas, from the Triassic of Lesotho (Basutoland), southern Africa. Occa-
1893) (Mesoeucrocodylia: Dyrosauridae) from the Lower Eocene sional Papers of the University of Kansas Museum of Natural His-
of North Africa. Canadian Journal of Earth Sciences 42:323– tory 106:1–37.
337. Wu, X.-C., A. P. Russell, and S. L. Cumbaa. 2001. Terminonaris (Ar-
Jouve, S. 2007. Taxonomic Revision of the Dyrosaurid Assem- chosauria: Crocodyliformes): new material from Saskatchewan,
blage (Crocodyliformes: Mesoeucrocodylia) from the Paleocene Canada, and comments on its phylogenetic relationships. Journal of
of the Iullemmeden Basin, West Africa. Journal of Paleontology Vertebrate Paleontology 21:492–514.
81:163–175.
Jouve, S., and D. Schwarz. 2004. Congosaurus bequaerti, a Paleocene Submitted December 11, 2008; accepted March 16, 2009.
dyrosaurid (Crocodyliformes; Mesoeucrocodylia) from Landana
(Angola). Bulletin de l’Institut Royal des Sciences Naturelles de APPENDIX 1. Description of characters used in phylogenetic
Belgique, Sciences de la Terre 74:129–146. analyses. Outgroup consists of Sarcosuchus imperator, Termi-
Jouve, S., B. Bouya, and M. Amaghzaz. 2005a. A short-snouted nonaris robusta, and Elosuchus cherifiensis. Ordered characters
dyrosaurid (Crocodyliformes, Mesoeucrocodylia) from the are labeled as such; all other characters are unordered. Except
Palaeocene of Morocco. Palaeontology 48:359–369. where otherwise stated, characters are original to this publica-
Jouve, S., B. Bouya, M. Amaghzaz. 2008. A long-snouted dyrosaurid
(Crocodyliformes, Mesoeucrocodylia) from the Paleocene of Mo-
tion. Characters that have been modified from other publications
rocco: phylogenetic and palaeobiogeographic implications. Palaeon- are cited with the character number in bold after the citation.
tology 51:281–294.
Jouve, S., M. Iarochène, B. Bouya, and M. Amaghzaz. 2005b. A new General
crocodyliform dyrosaurid from the Palaeocene of Morocco and
a phylogenetic analysis of Dyrosauridae. Acta Palaeontologica (1) Length of snout: greater than 68% (0); lesser than 68% (1)
Polonica 50:581–594. of total skull length, measured from tip of snout to posterior
Jouve, S., M. Iarochène, B. Bouya, and M. Amaghzaz. 2006. A new margin of parietal. Modified from Jouve, 2005 5.
species of Dyrosaurus (Crocodylomorpha, Dyrosauridae) from the
early Eocene of Morocco: phylogenetic implications. Zoological Cranial Openings
Journal of the Linnean Society 148;603–656.
Langston, W. L. 1995. Dyrosaurs (Crocodilia, Mesosuchia) from the Pa- (2) Medial tubercles of external nares on the posterior margin:
leocene Umm Himar Formation, Kingdom of Saudi Arabia. U.S. absent (0); dorsal (1); ventral (2).
Geological Survey Bulletin 2093-F:F1–F36. (3) Thickness of anterior margin of external nares: less than
Maddison, W. P., and D. R. Maddison. 2008. Mesquite: a modu- half anteroposterior length (0); greater than half anteropos-
lar system for evolutionary analysis. Version 2.5. Available at
http://mesquiteproject.org. Accessed October 8, 2008.
terior length (1).
Marshall, L.G. and C. de Muizon. 1988. The dawn of the age of Mammals (4) Orbits positioned: medially (0); laterally (1).
in South America. National Geographic Research 4:23–55. (5) Lateral margin of the orbit comprised dominantly by:
Moody, R. T. J., and E. Buffetaut. 1981. Notes on the systematic and lacrimal (0); jugal (1).
palaeoecology of the crocodiles and turtles of the Metlaoui Phos- (6) Supratemporal fenestrae length and width: subequal or
phates (Eocene) of Southern Tunisia. Tertiary Research 3:125– wider than long (0); longer than wide (1). Modified from:
140. Clark, 1994 68; Ortega et al., 2000 41; Wu et al., 2001 67;
Ortega, F., Z. B. Gasparini, A. D. Buscalioni, J. O. Calvo. 2000. A new Sereno et al., 2003 4; Jouve et al., 2005b 15; Jouve et al.,
species of Araripesuchus (Crocodylomorpha, Mesoeucrocodylia) 2006 12; Jouve et al., 2008 15.
from the Lower Cretaceous of Patagonia (Argentina). Journal of
Vertebrate Paleontology 20:57–76.
(7) Interfenestral bar: wide (0); narrow (1). Modified from:
Parris, D. C. 1986. Biostratigraphy of the fossil crocodile Hyposaurus Clark, 1994 33; Wu et al., 2001 33; Sereno et al., 2001 26;
Owen from New Jersey. New Jersey State Museum Investigation Jouve, 2005 10; Jouve et al., 2005b 18; Jouve et al., 2006 90;
Number 4, 16 pp. Jouve et al., 2008 18.
Pomel, A. 1894. Découverte de champsosauriens dans les gisements (8) Interfenestral bar: ornamented (0); unornamented (1).
de phosphorite du suessonien de l’Algérie. Comptes Rendu de Modified from: Jouve et al., 2005b 30; Jouve et al., 2008 30.
l’Académie des Sciences 118:1309–1310. (9) Cross section of interfenestral bar: square-shaped (0);
Scotese, C. R. 2001. Digital Paleogeographic Map Archive on CD-ROM. shaped as an upright T (1); upside-down V-shaped, in the
PALEOMAP Project, Arlington, Texas. form of a sagittal crest (2). Following Figure 8 of Jouve et
Schwarz, D., E. Frey, and T. Martin. 2006. The postcranial skeleton of
the Hyposaurinae (Dyrosauridae; Crocodyliformes). Palaeontology
al., 2005a.
49:695–718. (10) Posterior walls of supratemporal fenestrae: almost vertical
Sereno, P.C., H. C. E. Larsson, C. A. Sidor, and B. Gado. 2001. The giant and barely visible in dorsal view (0); dorsally inclined and
Crocodyliform Sarcosuchus from the Cretaceous of Africa. Science largely visible in dorsal view (1). Modified from: Jouve et
294;1516–1519 al., 2005b 10; Jouve, 2005 6; Jouve et al., 2008 10.
158 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 30, NO. 1, 2010
Jugal Laterosphenoids
(49) Ventral portion of postorbital bar: medially displaced with (63) Laterosphenoid sutures with parietal: parallel to skull table
ridge separating postorbital bar from lateral surface of ju- (0); descends posteriorly relative to the skull table (1).
gal (0); medially displaced from lateral surface of jugal with
no ridge (1); nearly flush with lateral surface of jugal (2).
Mandible
Modified from: Ortega et al., 2000 34; Jouve et al., 2006 72.
(50) Postorbital bar: nearly vertical (0); slightly anterodorsally (64) Size of second mandibular alveoli relative to first mandibu-
inclined (1); strongly anterodorsally inclined (2). From lar alveoli: similar in size (0); reduced in size relative to both
Jouve et al., 2008 35. adjacent alveoli (1).
(65) Interalveolar space between second and third mandibular
Supraoccipital alveoli relative to that of first and second: roughly equal in
proportion (0); greater (1).
(51) Supraoccipital participation to posttemporal fenestrae: (66) Third mandibular alveoli placement: interalveolar space is
present (0); absent (1). roughly equal to that between second and third mandibular
alveoli (0); close to mandibular fourth (1).
Exoccipitals (67) Fourth alveolus in proportion to other mandibular alveoli:
(52) Exoccipital participation to occipital condyle: slight (0); enlarged (0); similar in size (1).
large (1). Modified from: Jouve, 2004 96; Jouve et al., 2005b (68) Fourth alveolar wall: level with adjacent alveoli (0); raised
5; Jouve et al., 2006 104; Jouve et al., 2008 5. relative to adjacent alveoli (1).
(53) Occipital tuberosities: absent (0); small and reduced (1); (69) First alveolus: directed entirely dorsal (0); slants anteriorly
strongly developed (2); Ordered. Modified from: Jouve, as well as dorsally (1).
2005 1; Jouve et al., 2005b 3; Jouve et al., 2008 3. (70) Anterior portion of mandible: typically exhibits expansion
(0); exhibits straight margins without significant lateral ex-
Basisphenoid pansion (1). Modified from Sereno et al., 2001 72.
(71) Point at which mandibular symphysis ends posteriorly: an-
(54) Basisphenoid rostrum: short (0); extremely long anteriorly terior to anterior 3/4 alveoli (0); posterior to anterior 3/4
(1). From: Jouve, 2005 2; Jouve et al., 2005b 7; Jouve et al., alveoli (1).
2008 7. (72) Occlusal pits for maxillary occlusion: minimal (0); prolific
(1).
Basioccipital
(73) Seventh alveolus: comparable in size to adjacent alveoli (0);
(55) Single wide ventral rugosity of basioccipital medial to ba- reduced in size and close to eighth alveolus (1). Modified
sioccipital tubera: absent (0); present (1). from: Jouve, 2004 153; Jouve, 2005 3; Jouve et al., 2005b 8;
(56) Basioccipital tubera in ventral view: oblong (0); V- Jouve et al., 2006 164; Jouve et al., 2008 8.
shaped/tear-drop shaped (1). (74) Symphysis: wider than high (0); about as wide as high (1).
(57) Paired grooves along ventral surface, extending from base From: Jouve et al., 2005b 17; Jouve et al., 2008 17.
of occipital condyle to basioccipital tubera: absent (0); (75) Retroarticular process: short (0); extremely long and pos-
present (1). terodorsally curved (1). From: Jouve et al., 2005b 1; Jouve
(58) Lateral eustachian foramen: level with medial eustachian et al., 2008 1.
foramen (0); dorsolateral to medial eustachian foramen, (76) Posteromedial wing of retroarticular process: dorsally situ-
largely hidden from occipital view by the basioccipital (1). ated or at mid-height (0); ventral on the retroarticular pro-
Modified from Jouve et al., 2006 145. cess (1). From: Jouve et al., 2005b 2; Jouve et al., 2008 2.
(59) Ventral part of basioccipital: vertical, largely visible in oc- (77) Coronoid: present (0); absent (1). From: Jouve et al., 2005b
cipital view (0); strongly inclined, weakly visible in occipital 6; Jouve et al., 2008 6.
view (1). From: Jouve et al., 2005b 13; Jouve et al., 2008 13. (78) External mandibular fenestra: absent or reduced to a thin
slot (0); wide (1). Modified from: Jouve et al., 2005b 20;
Jouve et al., 2008 20.
Quadratojugals (79) Posterior process of angular relative to posterior extremity
(60) Extension of quadratojugal to cranial condyle for articula- of the retroarticular process: reaches process (0); remains
tion with jaw: does not extend (0); extends largely (1). From: far anterior (1). From: Jouve et al., 2008 39.
Jouve et al., 2005b 19; Jouve et al., 2008 19.
Dentition
Palatines (80) Teeth: robust with very wide alveoli (0); thin and long teeth
(1). From: Jouve, 2005 11; Jouve et al., 2005b 21; Jouve et
(61) Extension of the palatines to choanae: anterior margin or al., 2008 21.
not (0); antermedial margin (1). From: Jouve et al., 2005b 4; (81) Distance between teeth from middle of maxilla or mandibu-
Jouve et al., 2008 4. lar symphysis: nearly constant (0) much longer than in ante-
rior portion of maxilla or mandibular symphysis (1). From
Jouve et al., 2008 43.
Pterygoids
Postcrania
(62) Anteromost point of posterior margin of pterygoidian wing:
far anterior to medial eustachian foramen (0) about level (82) Lateral margin of medial dorsal osteoderms: linear (0);
with medial eustachian foramen (1). Modified from: Jouve, serrated and curved, sutured with the lateral osteoderm
2005 8; Jouve et al., 2005b 14; Jouve et al., 2008 14. (sends an acute process laterally) (1). From Jouve et al.,
2008 37.
160
APPENDIX 2. Character-taxon matrix used in phylogenetic analyses of Dyrosauridae. Outgroup consists of Sarcosuchus imperator, Elosuchus cherifiensis, and Terminonaris robusta. See
Appendix 1 for character descriptions.
0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8
Sarcosuchus imperator 0 0 0 0 0 0 0 0 ? 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Terminonaris robusta 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 1 0 0 0 0 1
Elosuchus cherifiensis 0 ? 0 0 1 0 0 0 ? 0 0 1 1 0 0 0 0 1 0 0 ? 0 0 1 ? 0 0 1
Arambourgisuchus 0 ? ? 0 0/1 1 1 1 2 1 ? 0 0 0 1 1 1 1 ? 1 1 1 1 0 0 1 1 1
khouribgaensis
Hyposaurus rogersii 1 ? 0 0 1 1 1 1 ? 1 1 ? ? 0 1 ? ? ? ? 1 1 ? 1 0 1 2 1 ?
Rhabdognathus 0 ? ? 0 ? 1 1 1 2 1 1 ? ? ? ? ? ? ? ? ? ? ? 2 0 ? 0 1 1
keiniensis
Rhabdognathus ? ? ? 0 1 1 1 1 2 1 1 ? ? ? ? ? ? ? ? ? ? ? 1/2 0 1 ? 1 1
aslerensis
Dyrosaurus 0 1 0 0 1 1 1 0 1 1 1 0 0 0 1 1 0 1 1 1 1 1 2 0 0 0 1 1
phosphaticus
Dyrosaurus 0 ? ? 0 1 1 1 0 1 1 1 0 0 0 1 1 0 1 1 1 1 1 2 0 0 0 1 1
maghribensis
Chenanisuchus 1 ? 1 1 1 1 0 0 0 0 1 ? 0 1 1 ? ? ? 1 ? 1 0 0 0 ? 2 1 ?
lateroculi
Sokotosuchus 1 1 1 0 ? 1 1 1 ? 0 1 ? ? 0 1 1 ? 1 1 1 1 0 0 1 ? 2 1 ?
ianwilsoni
Phosphatosaurus 1 1 1 0 1 1 1 1 ? 1 1 1 1 0 1 1 1 1 1 1 1 0 0 1 1 1 1 1
gavialoides
Congosaurus bequaerti 1 1 1 0 1 ? ? ? ? ? ? 0 0 1 1 1 0 1 1 ? 1 1 2 0 0 1 ? ?
Atlantosuchus 0 ? ? 0 0 1 1 1 1 ? 1 ? ? ? ? ? ? ? ? ? ? 1 2 0 ? ? ? ?
coupatez i
Guarinisuchus muniz i 0 1 0 0 0 1 1 1 2 1 1 0 ? 0 1 1 0 1 1 1 1 ? 2 0 1 2 1 1
Cerrejonisuchus
improcerus 1 2 0 0 0 1 0 0 0 1 0 0 1 1 1 1 1 1 1 1 1 0 1 0 0 2 1 1
gen. et sp. nov.
2 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5
9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4
Sarcosuchus imperator 0 0 0 0 0 0 0 0 0 0 0 ? ? 0 0 ? 0 0 0 0 0 0 ? 0 0 0
Terminonaris robusta 0 1 1 0 0 0 0 1 1 1 1 0 0 0 0 ? 0 0 0 0 0 0 ? 0 0 ?
Elosuchus cherifiensis 0 0 0 0 1 0 0 1 1 0 0 ? ? 0 1 0 0 0 0 0 0 0 0 0 0 0
Arambourgisuchus 1 2 1 2 0 1 1 0 ? 0 0 1 1 0 0 1 1 0 1 1 1 ? 0 1 2 ?
khouribgaensis
Hyposaurus rogersii ? ? 1 ? ? ? ? ? ? 1 ? ? ? 1 ? ? 0 0 1 0 1 2 ? 1 2 ?
Rhabdognathus ? ? 1 1/2 ? ? ? ? ? 1 1 1 1 1 1 1 2 1 1 1 ? 2 ? 1 2 ?
keiniensis
Rhabdognathus 1 2 ? 0/1 ? ? 0 1 ? 1 1 ? ? 1 1 1 2 1 1 1 2 2 1 1 2 1
aslerensis
Dyrosaurus 1 1 1 2 0 0 0 1 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 2 1
phosphaticus
Dyrosaurus 1 1 1 2 0 0 0 1 0 0 0 1 ? 0 0 0 2 0 1 0 1 1 0 1 2 1
maghribensis
Chenanisuchus ? ? 1 1 1 ? 0 1 1 0 1 1 0 0 1 1 0 0 0 0 0 1 ? 1 1 ?
lateroculi
Sokotosuchus ? ? 0 0 1 ? ? ? ? ? ? ? ? 0 ? ? 0 0 0 0 ? 1 ? ? 1 ?
ianwilsoni
Phosphatosaurus 1 2 0 2 1 1 ? 0 ? 0 0 ? ? 0 ? ? 0 0 ? 0 ? 1 ? ? ? ?
gavialoides
Congosaurus bequaerti ? ? 1 2 0 ? 1 ? ? 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Atlantosuchus ? ? 1 0 0 ? 1 1 0 0 0 1 ? 1 1 0 2 1 ? 1 ? ? ? 1 2 ?
coupatez i
Guarinisuchus muniz i 1 2 1 0/1 ? 1 0 1 0 0 ? ? ? 1 ? ? 2 ? 1 1 ? 2 ? 1 2 ?
Cerrejonisuchus
improcerus 2 2 1 2 1 1 ? 0 0 1 0 0 1 0 0 1 2 1 1 0 1 ? 1 1 2 ?
gen. et sp. nov.
(Continued on next page)
161
162
APPENDIX 2. Character-taxon matrix used in phylogenetic analyses of Dyrosauridae. Outgroup consists of Sarcosuchus imperator, Elosuchus cherifiensis, and Terminonaris robusta. See Ap-
pendix 1 for character descriptions. (Continued)
5 5 5 5 5 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 8 8 8
5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2
Sarcosuchus imperator ? ? ? 0 0 0 0 0 ? 0 0 0 0 0 0 0 ? 0 0 0 0 0 0 0 0 0 0 0
Terminonaris robusta 0 ? 0 0 0 0 0 1 0 0 1 0 1 0 1 0 0 1 0 0 ? ? ? 0 0 1 0 0
Elosuchus cherifiensis 0 0 0 1 0 0 1 1 ? 0 0 0 0 ? 1 0 0 0 0 0 0 ? ? 1 0 0 1 0
Arambourgisuchus 1 1 ? ? 0 1 1 1 0 1 1 1 0 1 ? 1 ? 1 1 1 ? ? ? ? ? 1 0 ?
khouribgaensis
Hyposaurus rogersii ? ? 0 ? 1 1 1 0 ? 1 1 1 0 ? 1 1 1 0 1 1 1 1 1 1 0 1 1 1
Rhabdognathus 1 1 ? 1 1 1 ? ? 0 ? ? ? ? ? ? ? ? ? ? 1 ? ? ? ? ? 1 1 ?
keiniensis
Rhabdognathus 1 1 1 1 1 1 1 1 ? ? ? ? ? ? ? ? ? ? ? 1 ? ? ? ? ? 1 1 ?
aslerensis
Dyrosaurus 1 1 1 1 0 1 1 0 1 1 1 1 0 1 1 1 ? ? 1 1 1 1 ? 0 ? 1 0 0
phosphaticus
Dyrosaurus 1 1 ? 1 0 1 1 1 1 1 1 1 0 ? 1 ? 1 ? 1 0 1 1 1 0 1 1 0 0
maghribensis
Chenanisuchus 1 ? ? ? 0 1 ? ? 0 ? ? ? ? ? ? ? ? ? ? 0 1 1 ? ? ? 1 0 ?
lateroculi
Sokotosuchus ? ? ? ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 ? ? ? ? ? 0 0 ?
ianwilsoni
Phosphatosaurus ? ? ? ? 0 ? 1 1 ? ? ? ? ? ? ? ? ? ? ? 1 ? ? ? ? ? 0 0 ?
gavialoides
Congosaurus bequaerti ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 ? ? 1 ? 1 1 1 0 0 1 1 1
Atlantosuchus ? ? ? ? ? 1 ? ? ? 0 1 1 1 0 1 1 ? ? 1 1 1 ? ? ? 1 1 1 ?
coupatez i
Guarinisuchus muniz i ? 1 ? ? 1 1 1 ? ? ? ? ? ? ? ? ? ? 1 ? 0 ? ? ? ? ? 1 ? ?
Cerrejonisuchus.
improcerus 1 1 0 1 ? 1 ? ? 0 1 1 1 0 1 1 1 0 0 1 ? ? ? ? ? ? 1 ? ?
gen. et sp. nov.