Content uploaded by Spencer G. Lucas
Author content
All content in this area was uploaded by Spencer G. Lucas on Nov 01, 2018
Content may be subject to copyright.
565
Lucas, S.G. and Sullivan, R.M., eds., 2018, Fossil Record 6. New Mexico Museum of Natural History and Science Bulletin 79.
DESCRIPTION OF A JUVENILE SPECIMEN OF THE LATE TRIASSIC AMPHIBIAN
APACHESAURUS GREGORII: DEVELOPMENTAL AND RELATIVE GROWTH
LARRY F. RINEHART and SPENCER G. LUCAS
New Mexico Museum of Natural History, 1801 Mountain Road NW, Albuquerque, NM;
-email: larry.rinehart@earthlink.net
Abstract—We describe a juvenile specimen of Apachesaurus gregorii Hunt, 1993, a terrestrially-
adapted metoposaurid amphibian from the Upper Triassic of Arizona, New Mexico and Texas.
This important specimen consists of an incomplete skull of 83.9 mm length, incomplete mandibles,
shoulder girdle elements, some forelimb bones, and an incomplete vertebral column. Comparisons to
adult A. gregorii specimens and to adult and juvenile specimens of other metoposaur species reveal
much about the ontogenetic growth and ecology of the species. A preliminary assessment indicates
that the Apachesaurus skull grows isometrically, except for the orbits and the posterior width of the
cultriform process of the parasphenoid. Depending on location, lateral line canals grow in negative to
moderately positive allometry, but never reach the full development seen in other genera. Pleurocentra
grow relatively more robust throughout life, and some shape changes in the interclavicle are noted.
This pattern of growth differs substantially from that of the other metoposaurids for which comparable
data are available and thus is an ontogenetic trajectory distinctive of Apachesaurus.
INTRODUCTION
Juvenile fossil specimens of extinct taxa are important in
understanding ontogeny and relative growth. The allometric
growth characteristics of a sample of adult specimens can be
projected forward or backward to predict the shape of a younger
or older individual if simple allometry is assumed, However,
without juvenile specimens it cannot be known with certainty
whether the allometry is simple, exhibiting a single rate of
change, or complex, wherein the relative growth rate changes
throughout ontogeny (e.g., Rinehart and Lucas, 2001).
Metoposaurids were the common Late Triassic
temnospondyl amphibians across much of Pangea (e.g., Hunt,
1993; Lucas, 2018). There are hundreds of preserved specimens
of adult metoposaurs (e.g., Rotten Hill, Texas: Lucas et al, 2016;
Argana, Morocco: Dutuit, 1976; Lamy, New Mexico: Rinehart
et al., 2009; Lucas et al., 2010; Krasiejow, Poland: Sulej, 2007),
but small, juvenile metoposaur specimens are exceptionally
rare (e.g., Hunt, 1989; Morales, 1993; Zanno, et al., 2002). No
doubt, this is partially because of taphonomic “size biasing”
(Lyman, 1994) wherein the remains of small animals are more
thoroughly destroyed by scavengers and predators, more easily
weathered and fragmented, and, even after fossilization, because
of their high surface-area-to-volume ratio, are more subject to
dissolution.
In addition to the taphonomic factors, Rinehart et al.
(2008) and Rinehart and Lucas (2009, 2016) used survivorship
data, age distribution, limb bone allometry, and lateral line
development to conclude that many juvenile metoposaurs
and other temnospondyls (Rinehart and Lucas, 2016; Lucas,
et al., 2016) were terrestrial, thus removing them from the
aquatic environments that promote rapid burial and subsequent
fossilization. Thus, this additional biological factor further
contributes to their rarity in the fossil record.
Apachesaurus is unique among the metoposaurids in
that it shows numerous terrestrial adaptations. These include
elongate, well-developed intercentra, robust rib facets, reduced
lateral line canals, and the morphology of the acetabulum,
which suggests precise femur articulation (Hunt, 1993). Here,
we describe an important juvenile specimen of Apachesaurus
gregorii that has been illustrated and/or partially described
previously by Davidow-Henry (1989, cursory description only),
Hunt (1993, g. 13) and Spielmann and Lucas (2012, g. 18).
Since then, additional preparation has revealed more of the skull
morphology, the shoulder girdle, some vertebrae, particularly
the atlas-axis complex, and some limb bones that were not
available to earlier workers.
In this publication, all tables are consolidated in Appendix
1.
ABBREVIATIONS
Institutional: MNA, Museum of Northern Arizona,
Flagstaff, AZ; MNHN, Museum National D’Histoire Naturelle,
Paris; MU, University of Missouri Museum of Paleontology,
Columbia, MO; NMMNH, New Mexico Museum of Natural
History and Science, Albuquerque, NM; UCMP, University of
California Museum of Paleontology, University of California,
Berkeley, CA; YPM, Yale Peabody Museum, Yale University,
New Haven, Connecticut; ZPAL, Institute of Paleobiology,
Polish Academy of Sciences, Warsaw.
Anatomical: Ang, angular; C, covered; Clav, clavicle; Dnt,
dentary; EPt, ectopterygoid; F, frontal; Fg, fragment; J, jugal; IO,
infraorbital lateral line canal; Lac, lacrimal; M, mandible; Ms,
missing; Mx, maxilla; N, external naris; Na, nasal; O, orbit; OB,
orbit border; Pal, palatine; Par, parietal; PO, postorbital lateral
line canal; PoF, postfrontal; PoO, postorbital; PoP, postparietal;
PrF, prefrontal; Pt, pterygoid; QJ, quadratojugal; Ra, radius; SO,
supraorbital lateral line canal; Sq, squamosal; St, supratemporal;
Tab, tabular; Ul, ulna; V, vomer.
Metrics: AFL, adductor fossa length; AFW, adductor fossa
width; ChL, choana length; ChW, choana width; GW, greatest
width; IOW, interorbital width; INW internarial width; NL,
narial length; NW, narial width; OL, orbit length; OW, orbit
width; PVL, palatal vacuity length; PVW, palatal vacuity width.
PROVENANCE
According to the UCMP locality database records, the
Apachesaurus gregorii specimen (UCMP 171591) described
here is from the Inadvertent Hills Locality (UCMP locality
V82250) in Petried Forest National Park, Arizona, and was
collected by R.A. Long and a UCMP eld party in 1982. UCMP
V82250 in Apache County, Arizona is a productive vertebrate
and invertebrate fossil locality located in mudstone just below
a white sandstone bench in the Painted Desert Member of the
566
Petried Forest Formation of the Chinle Group. In addition
to unionid bivalves and abundant gastropods, remains of
the aetosaur Typothorax coccinarum, a phytosaur, a small
undescribed reptile, a theropod dinosaur limb, and amphibians
were collected there (UCMP locality search, 2018). For
additional stratigraphic information on this locality see Heckert
and Lucas (2002) and Parker (2006).
SYSTEMATIC PALEONTOLOGY
Amphibia Gray, 1825
Temnospondyli Zittel, 1887-90
Metoposauridae Watson, 1919
Apachesaurus Hunt, 1993
Apachesaurus gregorii Hunt, 1993
Revised diagnosis: Small metoposaurid distinguished
from other metoposaurid genera by having a lacrimal positioned
laterally that does not enter the orbit; jugal and prefrontal
contact each other between the lacrimal and the orbit border;
pineal foramen located 2/3 of the way posterior along the length
of the parietals; otic notch wide and shallow; ridge-and-groove
ornamentation of the skull roof poorly developed; and elongate
vertebral centra with well-developed rib facets (Hunt, 1993;
Schoch and Milner, 2000).
Referred specimen: UCMP 171591, skull, lower jaws and
incomplete postcranium.
Horizon and locality: Painted Desert Member of Petried
Forest Formation, Chinle Group; UCMP locality V82250 in the
Petried Forest National Park, Arizona.
DESCRIPTION
Some previous preparation of UCMP 171591 had been
undertaken, but considerable calcite and hematite concretion
still covered much of the fossil. With permission from the
UCMP, additional preparation was carried out at NMMNH. The
additional preparation exposed more of the skull and mandibles,
especially on the palate side, a complete clavicle, an interclavicle,
a paired radius and ulna, and a few vertebral intercentra.
Skull and Jaws
Measurements of the skull of UCMP 171591 are listed in
Table 1. Estimates were made to compensate for broken and
displaced elements.
Comparative size of the specimen—The juvenile skull
(UCMP 171591) is 49.6% of the length of the holotype skull
(UCMP 63845) and 35% of the estimated length of the largest
known specimen (UCMP 63846), a left postorbital fragment.
Our estimate of the total midline length of UCMP 63846 is 238
mm. This estimate was made by applying a scaling factor based
on the greatest width of the specimen compared to the GW of
the holotype.
The juvenile skull is 38% of the estimated length of another
large specimen, YPM 4201. YPM 4201 is a right postorbital
skull fragment that we estimate to have had an original midline
length of 221 mm. Hunt (1993) estimated its original length at
“under 200 mm” by an unknown method, an approximate 9%-
10% discrepancy. We used the aforementioned GW scaling
method. As a point of interest, the holotype skull is 71% of the
estimated midline length of the largest known skull.
Skull roof—The skull is incomplete, but the entire left
side except the premaxillary area is present, so the skull roof
morphology is clear (Figs. 1A, 2A). The sutures are clear and
easily discerned (Fig. 1B). The right orbital and preorbital areas
are missing, except for the posterior orbit border. The skull is
split at or near the midline suture, and the right side is displaced
several mm under the left side. The posterior portion of the
skull table is separated by a fracture extending from the parietal,
anterior to the pineal foramen, across the supratemporal and
squamosal to the lateral margin. This posterior fragment is tilted
anteriorly and slightly displaced under the anterior part of the
skull table. The pineal foramen is visible just posterior to this
fracture.
The posterior border of the external naris is present
and is enclosed by the maxilla and nasal. By projecting an
appropriately-sized external naris and narial border, we estimate
that 6 mm of length are missing from the snout tip. The lateral
skull margin is complete except for a short segment of the jugal
and maxilla about half way between the orbit and the posterior
skull margin (Figs. 1A, 2A). The posterior skull margin is
complete, including both otic notches. A single occipital condyle
is present but is separated from the skull.
Palate—The palate is approximately as complete as the
skull roof, but much of it is covered by the lower jaws, the left
clavicle, a radius and ulna pair, and various bone fragments
(Fig. 1C, 2B). Parts of the vomer, palatine, and ectopterygoid
are present, but the sutures between them are indistinct (Fig. 1D,
2B). Two palatine tusks, occupying the typical dual emplacement
pits, are present on the left side, and they are accompanied by
a row of teeth extending posteriorly along the palatine parallel
to the maxilla. About ve teeth are preserved in this row, but it
is obviously incomplete. Anteriorly, the vomer and palatine are
both broken and missing large pieces, so none of the choana
border is preserved. Additionally, no vomer tusks are visible,
even though they are large and prominent on the holotype adult
skull (UCMP 63845), and none of the marginal toothrow is
exposed. Small portions of the ventral skull roof are exposed,
and visible sutures allow the identication of numerous skull
roof bones (Fig. 1C-D).
The anterior parasphenoid is covered, but its posterior and
central cultriform process is well exposed. The anterior portion
of the process is covered by the distal ulna (Fig. 1C-D).
Mandible—Both left and right mandibles are incomplete
and are pressed against the skull, so no teeth are visible in either
the maxillae or dentaries (Figs. 1C-D, 2B). The more complete
left mandible is nearly in articulation but is missing short
segments near its center and at its proximal end. The sutures are
indistinct. Ornamentation is heavily impressed on the lateral and
ventral surfaces, including the angular, dentary, and parts of the
splenial.
Specimen UCMP 171591 includes a fragment of a left
mandible (Fig. 3) that does not belong to the principal juvenile
skull and jaws (Figs. 1-2). It is from a somewhat larger, though
probably still juvenile, animal and includes the posteriormost
dentary, lateral half of the adductor fossa, and anterior part of
the glenoid.
Lateral line canals—The lateral line canals are essentially
undeveloped in UCMP 171591. The supraorbital canal and the
infraorbital canal are missing altogether. A single short segment
of the postorbital canal may be seen just anterior to the point
where it exits the skull roof moving posteriorly (lateral to the
otic notch, position I in Fig. 4D). This short segment is 1.3 mm
wide and very shallow.
Vertebral Column
The vertebral column is incomplete and disarticulated. A
total of 20 intercentra are present, including the atlas and axis,
~ three cervicals, ~ nine dorsals, and ~ six caudals (Figs. 5-6).
Measurements of the atlas/axis and some of the better-preserved
dorsal intercentra are tabulated (Tables 2-3).
Preservation varies from pristine to incomplete and
severely weathered. Heavy calcareous and hematitic deposits
hindered the preparation of some of the fragile intercentra. No
evidence of pleurocentra is seen in any of the vertebral elements,
including facets to receive them on the intercentra, and no neural
arch material is preserved.
Both the atlas and axis are relatively shorter in length
(compared to their diameters) than any of the subsequent
567
FIGURE 1. Apachesaurus gregorii, incomplete juvenile skull with associated incomplete mandibles, forearm elements, and left
clavicle (UCMP 171591). A, skull in dorsal view. B, bone map of the more complete left half of the skull in dorsal view. C, skull
plus incomplete mandibles, forearm elements, and a left clavicle in ventral view. D, bone map of skull in ventral view.
568
FIGURE 2. Apachesaurus gregorii, incomplete juvenile skull with associated incomplete mandibles, forearm elements, and left
clavicle (UCMP 171591), stereo pairs. A, dorsal view. B, ventral view.
vertebrae. Both are amphicoelous, but not notochordal. All
subsequent vertebrae are notochordal.
Atlas—The atlas (Fig. 5A) is sub-rectangular in anterior or
posterior view and nearly rectangular in lateral view. It is twice
as wide as its length and is slightly greater in height than length.
The dorsal surface shows a wide, shallow sulcus that forms
the oor of the neural canal. Ventrally, the atlas shows lateral
excavations that each occupy about 1/3 of its width. They are
bordered anteriorly by a narrow ridge that represents the ventral
border of the occipital condyle facets. The remaining central 1/3
of the ventral surface is slightly raised to form a low, wide ridge
between the lateral excavations.
Anteriorly, the atlas shows two circular, concave facets for
articulation to the occipital condyles. These facets face slightly
anterodorsally at an angle of approximately 15 degrees from
vertical and are separated by a narrow sulcus. The posterior
surface of the atlas is concave, with a deeper, circular pit near
the center.
Axis—Anteriorly, the axis (Fig. 5B) is sub-rectangular to
match the shape of the posterior atlas; posteriorly, it is essentially
circular. The anterior surface is wider than the posterior surface,
and both are concave. Anteriorly, it is slightly less than twice
as wide as its length. No rib facets are evident. The lateral and
ventral surfaces are anteroposteriorly excavated and, especially
on the ventral surface, show small foramina.
Cervical—Cervical intercentra are generally smaller than
the dorsals and are deeply amphicoelous with a central notochord
hole penetrating the element on its axis. The rib facets face
anterolaterally and posterolaterally and are of distinctly different
sizes; the posterior facet being much larger. A short, narrow
lateral ridge, much thinner than that of the dorsal vertebrae
described below, connects the anterior and posterior rib facets.
Dorsal—The dorsal intercentra (Fig. 6A) show deep funnel-
shaped sculpting of both the anterior and posterior surfaces with
a central notochord penetration that connects the two “funnels”
near the center of the spool-shaped element. The dorsal surface
has a central, anteroposteriorly-oriented sulcus forming the oor
of the neural canal. Low, anteroposterior ridges on either side
of the neural canal sulcus provide articulation points for the
pedicles of the neural arch.
Capitular rib facets face anterolaterally and posterolaterally
from a position lateral to the anterior and posterior surfaces of
all dorsal intercentra. The anterior facets are somewhat smaller
than the posterior, but less so than in the cervicals. A narrow,
variously-developed, lateral ridge connects the anterior and
posterior facets. A small dorsolateral excavation is present
569
FIGURE 3. Apachesaurus gregorii mandible fragment.
Cataloged under UCMP 171591, but not part of the principal
skull and jaws. Posterior part of the dentary through part of the
glenoid is included. A, lateral view with sutures delineated; and
B, medial view, most sutures unclear.
FIGURE 4. Skull and shoulder girdle measurement protocol:
A, skull roof landmarks. A: snout tip at midline suture; B:
anterior border of external nares; C: anterior border of the
orbits; D: anterior border of the pineal foramen; E: posterior
border of the skull at midline suture; H: posteriormost extent
of the occipital condyles; GW: greatest width of the skull. A-C
= preorbital length. C-E = postorbital length; B, interclavicle
width; C, clavicle length and width; and D, generalized lateral
line designations and measurement points.
above the ridge, and a wider, but shallower, excavation is present
ventral to it.
Caudal—The caudal intercentra are deeply amphicoelous
and notochordal (Fig. 6C-D). Two retain incomplete haemal
arches that extend posteroventrally from the intercentrum
whereas four more do not have haemal arches but are judged
to be caudals by their small diameters and reduced or absent
rib facets. The haemal arches articulate to the intercentra by
means of wide, rounded pedicles that contact most of the ventral
surface.
One large, apparently basal, caudal shows no rib facets or
haemal arch articulations, and is a simple spool with smooth,
anteroposterior excavation around its lateral and ventral
surfaces. The smaller, apparently distal, caudals are essentially
simple spools with no apparent lateral rib facets, dorsal sulcus,
or ventral haemal arches; these are difcult to orient.
Shoulder Girdle
Clavicles—The clavicles (Fig. 7A-C) are somewhat
“bullet-shaped” in ventral aspect, being less triangular and more
parallel-sided with a somewhat more rounded than pointed
anterior tip. On both the left and right clavicles, the posterodorsal
horn-like ends of the ascending processes, which contain the
scapulocoracoids and where the cleithral facets are located, are
broken and missing. Otherwise, although somewhat fractured,
the clavicles are essentially complete.
Clavicle length was measured from the anterior tip of
the posteriormost extent of the ventral side, not including
the posterior extension of the ascending process. Width was
measured at the widest point, near the posterior end (Fig. 4C).
The left clavicle is adhered to the underside of the skull by
calcite and hematite concretion and is visible in ventral aspect.
The anterior tip covers the posterior left postfrontal, and the
posterior end extends past the left posterior margin of the skull
(Fig. 1C-D). The anterior 1/3 of the clavicle is separated from
the posterior portion by a narrow crack with slight sinistral
offset, making the exact length of the bone difcult to measure.
A careful estimate yields an original length of approximately
44 mm. The clavicle appears to be slightly laterally compressed
and shows a width, as preserved, of 17 mm. The right clavicle is
free of the skull, is essentially complete, and appears relatively
undeformed. It measures 42.6 mm long and is 19.7 mm wide.
The ventral sides of the clavicles bear the typical reticulate
and ridge-and-groove ornamentation of metoposaur dermal
bones. The reticulate pattern is restricted to the posterolateral
area of the clavicles, whereas the ridge-and-groove pattern
radiates medially and anteriorly from this area to the edges of
the bone. On the better-preserved right clavicle, the reticulate
pattern extends medially from its lateral edge 45% of the clavicle
width, and, from the posterior edge, it extends anteriorly 22%
of the clavicle length, excluding the posterior extension of the
ascending process (Fig. 7A).
Interclavicle—A nearly complete interclavicle is preserved
but is broken into two pieces. The posterior ~ 2/3 of the element
(Fig. 7B-C) have been prepared, the anterior 1/3 of it is cemented
to the dorsal side of the right clavicle. Due to the concretionary
matrix and the fact that it lies atly against the clavicle, it was
decided not to prepare it.
The interclavicle is perfectly tted to the clavicles that
almost surely belong to the skull of UCMP 171591 (Figs. 1, 2,
7A, C), and is judged to be part of that individual. The ventral
side was prepared, but because it is extremely thin and fragile,
a layer of hard calcareous matrix was left on the dorsal side to
strengthen and protect the specimen.
The specimen is broken from the right side, approximately
one cm anterior to the posterior extent of the clavicular facet,
across to the left side, just posterior to the clavicular facet. The
anterior 1/3 is cemented to the dorsal side of the right clavicle
570
FIGURE 5. Atlas and axis of Apachesaurus gregorii. A, juvenile atlas (UCMP 171591) in ve views; B, juvenile axis (UCMP
171591) in ve views; C, adult atlas (NMMNH P-22487) in ve views.
571
FIGURE 6. Apachesaurus gregorii dorsal and caudal intercentra of: A, juvenile (UCMP 171591) in ve views; B, adult (NMMNH
P-22486) in ve views. Some photos ipped horizontally to present the best-preserved side; C, juvenile caudal with incomplete
haemal arch (UCMP 171591) in ve views. Some photos ipped horizontally to present the best-preserved side; D, juvenile caudal
with incomplete haemal arch (UCMP 171591) in left lateral view.
572
573
North Apache Canyon at Quarry 2 (source: UCMP specimen tag
accompanying UCMP 63845) (also see Spielmann and Lucas,
2012).
Colbert and Imbrie (1956) showed that skull growth
trajectories may vary between different populations of the
same metoposaur species. Subsequently, Rinehart and Lucas
(2017) used an allometric method to show that sometimes these
intraspecic, population-based variations are of the same order
of magnitude as generic differences, concluding that growth
trajectories alone should not be used to erect new taxa.
The juvenile specimen, being from eastern Arizona, and
the adult specimen, being from eastern New Mexico, cannot
be considered to belong to the same population. They are also
of different ages; the Arizona skull is Revueltian (early-mid
Norian), the New Mexico skull is Apachean (late Norian-
Rhaetian) (Lucas, 2018). Thus, we consider the allometric
analyses here to be a rst approximation only.
Comparison to Adult Apachesaurus gregorii and Other
Metoposaurids
Skull shape—Overall, the shapes of the juvenile and the
adult Apachesaurus skulls are very similar (Figs. 1, 2, 8). The
greatest width is 80% of skull length in the juvenile and 77%
in the adult. The taper of the skull is estimated by comparing
greatest width to the preorbital width (at anterior orbit borders).
In the juvenile, the preorbital width is 40% of greatest width,
and in the adult, 42%. These small differences seem to be well
within the range of individual variation and measurement error.
Little or no overall shape change during growth is implied by
these simple comparisons.
A more comprehensive analysis was undertaken, although
due to the rarity of Apachesaurus skulls, it only involves two
skulls, the adult and the juvenile. Here, we used the standard
allometric method; the natural log transforms of various skull
metrics are plotted against the natural log transform of skull
midline length (Fig. 9). The slopes (coefcient of x in the linear
curve t equations) of the data lines are then observed. Slopes >1
indicate that a shape change occurs throughout growth because
the metric under study shows positive allometry whereby it
grows at a higher rate than skull length. If the slope =1, isometry
is indicated, by which the study metric grows at the same rate
as the skull length, and geometric similarity is maintained
throughout growth. Slopes <1 indicate negative allometry
whereby shape change occurs because the study metric grows at
a lower rate than skull length (e.g., Huxley, 1932; Gould, 1966).
Simple allometry has been assumed in this analysis because it
has been shown to be the case in other species of metoposaurs,
e.g., Metoposaurus diagnosticus, Koskinonodon perfectum,
and Dutuitosaurus ouazzoui (Lucas et al., 2016, g. 74 and
associated text; authors’ unpublished data from Rinehart and
Lucas, 2017).
The preorbital/postorbital length ratio is known to change
allometrically throughout growth in some populations of
Koskinonodon (e.g., Popo Agie population of Wyoming and
Lamy population of New Mexico; Colbert and Imbrie, 1956),
and in weak allometry in others (Rotten Hill population; Lucas
et al., 2016). This allometry manifests in the relative forward
or rearward migration of the orbits as well as in changes in the
relative proportions of the snout and skull table throughout
growth. In Apachesaurus, the preorbital length/postorbital
and cannot be seen. Thus, the posterior portion of the clavicular
facet and the point of maximum width is retained on the right
side but is missing on the left (Fig. 7B).
The ventral side of the interclavicle shows the usual central
reticulate pattern of ornamentation surrounded by ridge-and-
groove ornamentation that extends radially outward almost to
the edges, but the texture, particularly the reticulate portion,
is lightly impressed and somewhat faint (Fig. 7B). The ridge-
and-groove pattern ends approximately 1 mm from the lateral
and ventral edges of the interclavicle, and the remainder of the
surface is smooth with rounded edges. The central reticulate
area is ~ 11 mm in diameter; approximately 1/3 of the width of
the interclavicle.
The width of the interclavicle, estimated by doubling
the measurement from the center of the reticulate area to the
lateralmost tip of the thick, ornamented portion (not including
the clavicular facet) (Fig. 4B) is 34 mm, and the maximum
width, including the clavicular facet, is 43 mm.
A rounded, sub-triangular process extends posteriorly
from the posterior margin of the interclavicle (Fig. 7B-C). The
ornamentation is faint in this area but appears to be more of the
reticulate type. Somewhat subjective measurements indicate that
the process is approximately 5 mm wide where it contacts the
body of the interclavicle at its base and extends approximately 3
mm posteriorly to its apex.
A fragment of an interclavicle that does not t the shoulder
girdle of the principal part of the specimen is present under the
specimen number UCMP 171591 (Fig. 7D). It is approximately
twice as thick as the interclavicle associated with the skull and
has larger, deeper ornamentation. It is obviously from a larger, but
still juvenile animal. It appears to be from the anterior half of the
right side and shows the typical ridge-and-groove ornamentation
of the thicker, ornamented portion of the interclavicle plus a short
segment of the thin, anterolateral clavicular facet. Although it is
a small fragment of the original bone, it is well preserved, and
its identication is without question. This fragment represents
another amphibian in the deposit; probably, but not certainly,
Apachesaurus.
Limb Bones
A radius and ulna lie parallel to each other and are adhered
to the underside of the skull (Fig. 1C-D, 2B). Both are weathered
and somewhat attened. Their poor preservation precludes
identication of detailed features, but their position and shape
suggest that they are from the right forelimb.
Radius—The radius in incomplete. The distal end enters
the left orbit of the skull where it is apparently broken and
missing part of its length. As preserved, it is 17 mm long.
Ulna—The ulna is complete and measures 24.4 mm long.
The olecranon process is easily identied and establishes the
proximal end. The distal half of the ulna lies across the anterior
cultriform process of the parasphenoid, and its distal end is near
the posteromedial orbit border.
COMPARATIVE GROWTH AND ALLOMETRY
In this section, unless otherwise stated, juvenile refers to
UCMP 171591 and adult refers to UCMP 63845, the holotype of
Apachesaurus Hunt 1993 (Fig. 8). UCMP 63845 is from UCMP
locality V6148 in the Redonda Formation of Quay County, New
Mexico. It was collected by Cosgriff and Lefer in 1961 in
FIGURE 7 (facing page). Shoulder girdle of Apachesaurus gregorii. A, right clavicle of juvenile Apachesaurus gregorii (UCMP
171591); B, incomplete interclavicle of juvenile Apachesaurus gregorii (UCMP 171591); C, articulated clavicle and interclavicle of
juvenile Apachesaurus gregorii (UCMP 171591); D, amphibian (larger juvenile of Apachesaurus gregorii?) interclavicle fragment
catalogued under UCMP 171591, but not part of the principal specimen; E, the shoulder girdle of an adult Apachesaurus gregorii
(UCMP 63849) in ventral view, showing the nearly complete interclavicle with incomplete clavicles in articulation. The left clavicle
(right side of photo) shows its full width and the relationship of the reticulate and ridge and groove ornamentation. 5 cm scale bar; F,
Apachesaurus gregorii adult interclavicle (uncataloged specimen, YPM 6649. 1958/31). 2 cm scale bar; G, Apachesaurus gregorii
adult clavicle (uncataloged specimen, YPM 6649. 1958/31). 2 cm scale bar.
574
FIGURE 8. Apachesaurus gregorii Hunt 1993, holotype (UCMP 63845): A, skull in dorsal view (UCMP 63845); B, skull in ventral
view (UCMP 63845); C, skull in occipital view (UCMP 63845); D, right mandible in lateral (labial) view (UCMP 63847); E, right
mandible in medial (lingual) view (UCMP 63847); F, right mandible in dorsal (occlusal) view (UCMP 63847); G, bone map of the
holotype skull in dorsal view after Hunt (1993). Lateral line canals are stippled. Upper scale bar for A, B, C; lower scale bar for D,
E, F.
575
length ratio equals 0.46 in the juvenile and 0.47 in the adult.
This essentially identical preorbital /postorbital length ratio
implies isometric growth of these two regions and shows no
relative migration of the orbits in Apachesaurus. The allometric
analysis conrms that both the preorbital and postorbital skull
grow in isometry; their allometric constants are 1.02 and 0.99,
respectively (Fig. 9A).
Length to the pineal foramen also grows in isometry with
an allometric constant of 1.06. This indicates that the relative
position of the pineal foramen with respect to skull length
remains constant throughout growth. The position of the pineal
foramen relative to the parietals (at 2/3 of the way posteriorly
along the length of the parietals) is a character of Apachesaurus
noted in the adult skull by Hunt (1993). This feature is identical
in the juvenile.
Skull widths measured at the anterior orbits, the pineal
foramen, and at the greatest width (across the squamosals) show
allometric constants of 0.95, 1.01, and 0.97, all essentially =1
(Fig. 9A). Thus, approximately isometric growth of skull width
at these locations is indicated.
Orbit lengths and widths grow in negative allometry;
the orbits decrease in size relative to skull length throughout
growth, as indicated by the allometric constants of 0.78 for
length and 0.87 for width (Fig. 9B). Orbit width is 64% of orbit
length in the juvenile and 68% in the adult, indicating that the
elliptical orbit shape changes little during ontogenetic growth.
By comparison, in Koskinonodon (Rotten Hill population),
orbit length grows in positive allometry with respect to skull
length (allometric constant = 1.94) (Lucas et al., 2016, g. 74B).
This relatively strong allometry is accounted for by numerous
specimens (N=13) and is a high condence result. In contrast
to Apachesaurus, the orbits of Koskinonodon grow relatively
longer as it matures.
Interorbit width, with an allometric constant of 1.09, may
show weak positive allometry; meaning that the orbits move
slightly relatively farther apart throughout growth (Fig. 9B). This
slight allometry is similar to that of Koskinonodon (the Popo
Agie population sample at MU) with an allometric constant of
1.07, though somewhat different from Dutuitosaurus ouazzoui
with an allometric constant of 0.97 (unpublished data associated
with Rinehart and Lucas, 2017). Nonetheless, all are very close
to isometry.
The juvenile Apachesaurus otic notch measures 11.5 mm
wide and 4.5 mm deep (the R and S metrics of Gregory, 1980,
g. 7.1), representing, respectively, 13.7% and 5.7% of skull
FIGURE 9. Skull and orbital allometry of Apachesaurus. A, rates of growth with respect to skull midline length for the preorbital
and postorbital lengths, and for the width at the anterior orbits, width at the pineal foramen, and greatest width. Exact measurement
landmarks are shown in Fig. 4. All allometric constants (coefcient of x in the curve t equations) are approximately one; B, rates of
growth with respect to skull midline length for the orbits. Orbit length and width show negative allometry with allometric constants
of 0.78 and 0.87, respectively. Interorbit width shows slight positive allometry (allometric constant = 1.09).
length. The adult otic notch measures 24 mm wide and 6.5 mm
deep, 14% and 3.8% of skull width, respectively. The width of
the otic notch apparently grows in isometry with skull length,
but it appears to be relatively deeper in the juvenile. Additional
detailed otic notch measurements are available in Gregory
(1980).
Bone layout—Spielmann and Lucas (2012) provided a
cranial osteology of Apachesaurus gregorii and noted only
minor differences between it and other metoposaurids, except
for the position of the lacrimal as indicated by Hunt (1993).
Bone maps of the juvenile and adult skulls appear essentially
identical (Figs. 1B, 8G).
Parasphenoid—Hunt (1993) noted that in the adult
skull of Apachesaurus gregorii the cultriform process of the
parasphenoid is broad anteriorly, but narrower posteriorly than
it is in other metoposaurids (Fig. 8B), so that the skull width
divided by the minimum width of the cultriform process is >
17 (< 5.9% of skull width). This feature is listed as part of the
diagnosis for the genus Apachesaurus (Hunt, 1993). However, in
the juvenile specimen the cultriform process is wide posteriorly
(Figs. 1C-D, 2B), so skull width divided by minimum cultriform
process width equals 9.5 (10.5% of skull width).
In comparison to other adult metoposaurs, in the holotype
specimen of Koskinonodon perfectum (MU 537), the minimum
width of the cultriform process is 9.5% of skull width (authors’
unpublished data). In well-preserved Dutuitosaurus specimens
(e.g., MNHN XIII/14/66) the minimum cultriform process width
is 10% of skull width (Dutuit, 1976, pl. VIII). In Metoposaurus
diagnosticus, the cultriform process is of variable width; wide
along its entire length in ZPAL AbIII/894, being 12% of skull
width, whereas in ZPAL AbIII/1674 it is 7.5% of skull width
at its center but 13% of skull width posteriorly (Sulej, 2007,
gs. 8B, 11B). Other Metoposaurus diagnosticus skulls show
intermediate development of the cultriform process width.
Thus, the Apachesaurus juvenile posterior cultriform
process width is similar, percentage-wise, to that of adults of
other genera, but it grows in negative allometry. Throughout
growth, the posterior cultriform process of Apachesaurus
becomes relatively narrower until, per Hunt (1993), it is
proportionately narrower than in any other genus.
Lateral line canals—In general, the lateral line canals
on the postorbital skull are better developed at all ages. The
lateral line canals of the juvenile skull (UCMP 171591) (83.9
mm length) are completely undeveloped except for a very small
portion of the postorbital canal that measures 1.3 mm in width.
576
In the holotype skull (UCMP 63845) (169 mm length), the
supraorbital and anterior parts of the infraorbital canals appear
as a series of connected pits. The canals are better-developed on
the left side of the skull. The postorbital canal and the posterior
part of the infraorbital canal are slightly more developed into a
round-bottomed, smoother groove, but still retain some of the
connected-pit morphology.
In a large, but incomplete, adult skull of Apachesaurus
gregorii (UCMP 63846) (estimated 238 mm length) comprising
the left postorbital area, the lateral line canals are yet more fully
developed into deeper, wider grooves with rounder, smoother
bottoms; very little of the connected-pit structure is retained in
this large skull.
Lateral line width and depth were measured at various
landmarks (Fig. 4D) for four Apachesaurus gregorii skulls (two
very incomplete) and ve Koskinonodon perfectum skulls (Table
4). Lateral line canal width versus skull length and lateral line
width allometry at position I (Fig. 4D) were plotted (Fig. 10E-
F). The allometry plot shows a linear curve t line slope of ~one
(0.96), indicating approximately isometric growth of the lateral
line canal width in the posterior part of the postorbital canal; the
canal width grows at the same rate as skull length. Allometric
constants for other landmark positions are tabulated (Table 5).
A similar analysis of adult Koskinonodon skulls of various
sizes yielded an allometric constant of 1.3 at position I,
indicating moderately strong positive allometry; the canal width
grows at a higher rate than the skull length in this more aquatic
species. Allometric constants for Koskinonodon at all landmarks
are tabulated (Table 5).
The data show considerable scatter, but some generalizations
seem apparent (Table 5). In Apachesaurus (analysis limited to
the posterolateral portion of the skull because the lateral line
canals are essentially undeveloped in the juvenile) width grows
in negative allometry and depth grows in positive allometry.
In Koskinonodon, the opposite is true; most widths grow in
positive allometry (seven of the nine measured positions), and
the remaining two are near isometry, whereas depths generally
grow in less positive to somewhat negative allometry.
Note that the lateral line canals of Koskinonodon perfectum
are much larger in both width and depth than those of
Apachesaurus gregorii (Table 4). The lateral lines are an aquatic
sensory system but given that the lateral line sensory organs are
very small, it is unclear why a larger animal or a more aquatic
animal such as Koskinonodon would need a wider canal, but these
data suggest that this may be so. One possibility is that because
the skull ornamentation of Koskinonodon is more coarse and
disruptive to water ow over the surface, wider, deeper lateral
line canals were necessary to provide enough space between the
ornamentation and the lateral line organs for laminar surface
ow to be reestablished. These data also raise the question as to
whether the lateral line canals of Apachesaurus are vestigial, as
might be expected in a more terrestrial animal.
Mandible—The juvenile mandible (Figs. 1C-D, 2B),
similar to that of the adult (Fig. 8D-F), has heavy ornamentation
on the lateral and ventral sides. Neither of the juvenile mandibles
is complete, so an allometric assessment is not possible, but the
juvenile mandible appears to be more robustly constructed,
relative to skull size.
Vertebral column
Atlas—Juvenile and adult atlantes show numerous
differences (Table 2; Fig. 5A, C): during growth, the neural
canal sulcus deepens and widens; the neural arch ossies and
fuses to the intercentrum; the posterior face becomes more fully
rounded (less subrectangular); the wide, raised central portion
of the ventral surface becomes a wide sulcus with a small,
central keel-like structure; and lateral excavations become more
pronounced. Some of these differences may be due to increasing
ossication throughout growth.
Dorsal—The juvenile dorsal intercentra are somewhat
attened cylindrical spools, showing slightly less height than
width both anteriorly and posteriorly (Table 3; Fig. 6A, 10A).
Both height and width grow in weak positive allometry (Fig.
10B). The width grows in slightly more positive allometry
than the height (Fig. 10B), thus presenting an increasingly
horizontally-oriented oval aspect anteriorly and posteriorly (i.e.,
they become more dorsoventrally attened). The relative size of
the well-developed rib facets increases throughout growth.
Both juvenile and adult intercentra are amphicoelous, but
the juveniles have a deep, funnel-like depression in both the
anterior and posterior surfaces, whereas the adults are more
weakly amphicoelous. In some large dorsal intercentra (e.g.,
NMMNH P-17062), the anterior and posterior ends are barely
concave, and the notochord perforation has sharp, rather than
rounded, gradual edges.
Caudal—The sample of caudal vertebrae available for
study is small. Within this restriction, there appears to be a weak
tendency for the caudals to grow from a somewhat waisted,
slightly hourglass shape (Fig. 6C-D) to a more straight-sided
cylinder shape. The haemal arches become relatively more
robust and more rmly fused to the ventral intercentrum (e.g.,
NMMNH P-22486, containing both adult dorsal and caudal
intercentra).
Similar to the dorsal intercentra, the juvenile caudals are
strongly amphicoelous and become more weakly amphicoelous
throughout growth.
Intercentrum allometry—Anterior and posterior width,
anterior and posterior height, and notochord perforation diameter
(the minimum diameter, measured as deeply as possible in the
vertebral body) of the dorsal intercentra of both juvenile and
adult Apachesaurus were tabulated (Table 3) and plotted (Fig.
10A). The relative growth with respect to length (allometry) of
these metrics is plotted (Fig. 10B). The allometric constants (the
coefcient of x in the curve t equations, Fig. 10B) of the heights
and widths are all greater than one (average 1.14), showing
positive allometry. Thus, both height and width (vertical and
horizontal diameter) grow at a higher rate than length, and the
intercentra become more robust throughout growth. Thinking of
the intercentra as short, cantilevered beams, and given that beam
strength increases as the square of the diameter (e.g., Marks,
1951, Gieck and Gieck, 1997) this represents a signicant
relative strength increase in the vertebral column throughout
life.
Notochord perforation diameter, on the other hand, grows
with an allometric constant of 0.23 (Fig 10B). Thus, the
notochord perforation diameter grows at a much lower rate than
intercentrum length. Although both Hunt (1993) and this study
note that the notochord perforation becomes relatively smaller
throughout growth, it appears that in even the largest adult
intercentra, the notochord perforation goes completely through
the element.
Comparison to Koskinonodon perfectum—Growth
and allometry plots of Koskinonodon perfectum intercentrum
metrics (Table 6; Fig. 10C-D) show that the height and width
of its intercentra grow in very slight positive allometry (near
isometry) wherein the allometric constants of height and width
average 1.06. Thus, Koskinonodon shows less relative strength
increase in its intercentra than Apachesaurus. This is probably
because Koskinonodon has been shown to become more fully
aquatic (probably obligatorily so) throughout life (Romer, 1939;
Rinehart et al., 2008; Rinehart and Lucas, 2009; Lucas et al.,
2016), whereas Apacheasurus is assessed to be more terrestrial
(Hunt, 1993).
Shoulder girdle
Clavicles—The width/length ratio of the juvenile clavicle
577
FIGURE 10. Intercentrum and lateral line canal metrics and allometry. Intercentrum measurements do not include rib facets, only the
spool-shaped vertebral body. A, Apachesaurus gregorii, intercentrum anterior and posterior width, anterior and posterior height, and
notochord perforation diameter versus length; and B, A. gregorii, intercentrum anterior and posterior width, anterior and posterior
height, and notochord perforation diameter allometry. For comparision: C, Koskinonodon perfectum, intercentrum mediolateral
width and height versus length. D, K. perfectum, intercentrum mediolateral width and height allometry; E, A. gregorii lateral line
width in position I, just lateral to the otic notch where the canal exits the skull roof, versus skulll length; and F, A. gregorii lateral
line width allometry at position I.
578
Koskinonodon (e.g., Lucas et al., 2016, gs. 27, 29, 31, 33).
Some of the ornamentation pits in the posterior nasal-prefrontal-
anterior frontal region of the juvenile skull are slightly elongated.
This is true in the adults also, but with no further development.
SUMMARY
Unlike several other species of metoposaurids, the
juvenile skull of Apachesaurus varies little from the adult in
its proportions, although it must be noted that few skulls as
small as UCMP 171591 are available for other species. Our
rst-approximation allometry study shows essentially isometric
growth in the preorbital and postorbital skull lengths (Fig. 9A).
Likewise, the skull width, measured at the anterior orbits, pineal
foramen, and greatest width all grow isometrically with respect
to skull length.
The signicant exceptions to isometry appear in the orbital
length and width and the posterior width of the cultriform
process of the parasphenoid. The positive allometry of the
posterior width of the cultriform process is especially interesting,
as it is a dening character of Apachesaurus (Hunt, 1993) that
is substantially different in the juvenile. The width of the otic
notch of Apachesaurus appears to grow isometrically, but it
appears relatively deeper in the juvenile.
Lateral line canal widths grow in negative allometry to
near isometry (Table 5; Fig. 10F); depths of the canals grow
in weak to moderate positive allometry (Table 5). They are
essentially undeveloped in the juvenile and grow wider and
deeper throughout life. Even so, they never reach the relatively
deep, wide, smooth-bottomed canal development seen in other
genera (e.g., Koskinonodon, Dutuitosaurus, Metoposaurus). We
also nd that although some of the larger, more aquatic, adult
metoposaurids (e.g., Koskinonodon), show highly developed
lateral line canals, the canals are also undeveloped in their
juveniles (Fig. 11A-B).
The vertebral column of Apachesaurus grows allometrically
(Fig. 10B), in contrast to the approximately isometric growth
seen in Koskinonodon (Fig. 10D). The anterior and posterior
widths and heights of the intercentra all grow in weak to moderate
positive allometry; the robustness of the intercentra increases
throughout growth. Conversely, the notochord perforation
grows in strong negative allometry, becoming relatively much
smaller throughout growth.
Within the shoulder girdle, the clavicles are difcult to assess
because of poorly preserved adult elements (Fig. 7E-G). Given
the material at hand, the clavicles appear to grow isometrically,
including the proportions of the reticulate and ridge-and-groove
ornamentation. The interclavicle, however, does show shape
change throughout growth. Most signicantly, a posterocentral
process that is present in the juvenile (Fig. 7B) is missing in the
adult (Fig. 7E-F). The proportions of reticulate and ridge-and-
groove ornamentation change. The reticulate pattern increases
from 32% to 50% of interclavicle width.
To a rst-approximation, we nd that Apachesaurus has
a unique ontogenetic trajectory among metoposaurs, by which
the skull grows in overall near isometry. Some skull details,
including the posterior width of the cultriform process, the orbit
dimensions, and at least some of the lateral line canals, show
allometric growth. These results will, no doubt, be rened as
more Apachesaurus material is discovered and added to the
analysis.
ACKNOWLEDGMENTS
The authors thank Patricia Holroyd for the loan of UCMP
171591. We thank Andrew Heckert for recognizing the
importance of the specimen and arranging its loan to NMMNH.
Permission from UCMP to further prepare the specimen is
appreciated. Andrew Heckert and Adrian Hunt reviewed this
work and improved it with their comments and additions.
(UCMP 171591) equals 0.46, whereas in an adult clavicle (YPM
6649. 1958/31) this ratio equals 0.39. Given the condition of the
YPM clavicle, in which the edges are broken and it is uncertain
if the longest and widest parts are preserved, this difference is
probably within measurement error.
The left clavicle of adult Apachesaurus gregorii (UCMP
63849, an approximately complete interclavicle with incomplete
left and right clavicles) (Fig. 7E) shows ~ 42% of its width
covered by the reticulate ornamentation, and the right adult
clavicle (YPM 6649. 1958/31) (Fig. 7G) shows 45%-50%;
almost exactly as seen in the juvenile specimen, which is 45%
(UCMP 171591). Additionally, the juvenile clavicle (UCMP
171591) shows 22% of its length covered by a reticulate pattern,
whereas the adult clavicle (YPM 6649. 1958/31) shows ~ 24%
of its length covered by reticulate sculpture.
This comparison shows that there is, within measurement
error, an approximately isometric relationship of the clavicle
length and width and the reticulate and ridge-and-groove
ornamentation patterns of the clavicles throughout ontogeny.
The condition of the juvenile clavicle is excellent, but a more
certain assessment of relative growth awaits the discovery of
better adult material.
Interclavicle—The most striking difference between the
adult (UCMP 63849, YPM 6649. 1958/31) (Fig. 7E-F) and
juvenile (UCMP 171591) (Fig. 7B-C) interclavicles is the
shape of the posterior lobe. The central, posterior process of
the juvenile is lost through ontogenetic growth, so the posterior
border of the interclavicle becomes smoothly rounded.
The reticulate patterned portion of the juvenile interclavicle
represents ~ 32% of its width (UCMP 171591) (Fig. 7B). This
increases to ~ 50% in the adults (UCMP 63849, YPM 6649.
1958/31) (Fig. 7E-F).
Limb bones—There are no adult Apacheasurus limb bones
available for comparison to the juvenile. This is unfortunate
because the relative growth of the limb bones of the presumably
more terrestrial Apachesaurus should be quite different from
that of the increasingly aquatic taxon Koskinonodon.
A comparison of the midshaft diameter/length ratio of
adult Koskinonodon ulnae (e.g., Lucas et al., 2016, g. 66) to
a juvenile Apachesaurus ulna (UCMP 171591) shows that the
adult Koskinonodon ratio is ~ 0.15, whereas the juvenile is
relatively more robust with a ratio of ~ 0.2.
We consider the radius of UCMP 171591 to be too
incomplete for meaningful comparison.
Juvenile Koskinonodon perfectum
A juvenile preorbital metoposaur skull (MNA V8415),
from MNA locality 1393 (= NMMNH locality 4127), the Blue
Hills locality, low in the Blue Mesa Member of the Petried
Forest Formation, east-central Arizona ( Zanno et al., 2002), is
referable to Koskinonodon perfectum because the lacrimal enters
the anterolateral orbit border (Hunt, 1993) (Fig. 11). Zanno et al.
(2002) provided a thorough description of this fossil.
One of the most signicant aspects of this specimen is
that the lateral line canals, at least in the preorbital portion,
are undeveloped, as they are in the Apachesaurus juvenile. We
believe that the feature Zanno et al. (2002) interpreted as a lateral
line canal crossing the lacrimal is actually a short fracture with
a small amount of vertical offset (Fig. 11A, and especially Fig.
11B, the stereo pair). This is an interesting observation in that
the presumably fully aquatic adult Koskinonodon show highly
developed lateral line canals, but their juveniles show none (at
least in the preorbital skull).
In contrast to the poorly-developed skull ornamentation
of the Apachesaurus juvenile (Figs. 1-2), the Koskinonodon
juvenile shows deeply-sculpted reticulate ornamentation.
Essentially no ridge-and-groove ornamentation is present in the
juvenile preorbital skull (Fig. 11A-B), as is also true of adult
579
FIGURE 11. Koskinonodon perfectum, right preorbital skull of a juvenile (MNA V8415) described by Morales (1993) and Zanno,
et al. (2002). A, dorsal view; B, ventral view; C, stereo pair, dorsal; and D, stereo pair, ventral.
580
REFERENCES
Colbert, E.H., and Imbrie, J., 1956, Triassic metoposaurid
amphibians: Bulletin of the American Museum of Natural History,
v. 110, p. 403-452.
Davidow-Henry, B., 1987, New metoposaurs from the southwestern
United States and their phylogenetic relationships [ M.S. thesis]:
Lubbock, Texas Tech University, 76 p.
Davidow-Henry, B, 1989, Small metoposaurid amphibians from the
Triassic of western North America and their signicance, in Lucas,
S. G. and Hunt, A.P., eds., Dawn of the age of dinosaurs in the
American Southwest: Albuquerque, New Mexico Museum of
Natural History, p. 278-292.
Dutuit, J. M., 1976, Introduction à l’etude paléontologique du Trias
continental marocain. Description des premiers stegocephales
recueillis dan le couloir d’Argana (Atlas occidental): Memoires
du Museum National d’Histoire Naturelle, no. C36, p. 1-253.
Gieck, K., and Gieck, R., 1997, Engineering formulas, 7th ed. McGraw-
Hill, Inc., New York.
Gould, S.J., 1966, Allometry and size in ontogeny and phylogeny:
Biological Reviews, v. 41, p. 587-640.
Gray, J.E., 1825, A synopsis of the genera of reptiles and Amphibia
with a description of some new species: Annals of Philosophy
(Thompson), (N. S.), v. 10, p. 193-217.
Gregory, J.T., 1980, The otic notch of metoposaurid labyrinthodonts;
in Jacobs, L. L., ed, Aspects of vertebrate history: Essays in honor
of Edwin Harris Colbert: Flagstaff, Museum of Northern Arizona
Press, p. 125-136.
Heckert, A.B., and Lucas, S.G., 2002, Revised Upper Triassic
stratigraphy of the Petried Forest National Park, Arizona, U.S.A.,
New Mexico Museum of Natural History and Science Bulletin 21,
P. 1-36.
Hunt, A.P., 1989, Comments on the taxonomy of North American
metoposaurs and a preliminary phylogenetic analysis of the family
Metoposauridae, in Lucas, S. G. and Hunt, A.P., eds., Dawn of the
age of dinosaurs in the American Southwest: Albuquerque, New
Mexico Museum of Natural History, p. 293-300.
Hunt, A.P., 1993, Revision of the Metoposauridae (Amphibia:
Temnospondyli) and description of a new genus from western
North America: Bulletin of the Museum of Northern Arizona, v.
59, p. 67-97.
Huxley, J.S., 1932, Problems of relative growth: Baltimore, The Johns
Hopkins University Press, 276 p.
Lucas, S. G., 2018, Late Triasic terrestrial tetrapods: Biostratigraphy,
biochronology and biotic events; in Tanner, L. H., ed., The Late
Triassic world. Springer Topics in Geobiology, v. 46, p. 351-405.
Lucas, S.G., Rinehart, L.F., Krainer, K., Spielmann, J.A., and Heckert,
A.B., 2010, Taphonomy of the Lamy amphibian quarry: A Late
Triassic bonebed in New Mexico, U.S.A.: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 298, p. 388-398.
Lucas, S.G., Rinehart, L.F., Heckert, A.B., Hunt, A.P., and Spielmann,
J.A., 2016, Rotten Hill: A Late Triassic bonebed in the Texas
Panhandle, USA: New Mexico Museum of Natural History and
Science, Bulletin 72, 97 p.
Lyman, R.L., 1994, Vertebrate taphonomy. Cambridge, Cambridge
University Press, 524 p.
Marks, L.S., 1951, Mechanical engineers’ handbook. New York,
McGraw-Hill Book Company, Inc., 2236 p.
Morales, M., 1993, A small metoposaurid partial skull from the Chinle
Formation near St. Johns, Arizona: New Mexico Museum of
Natural History and Science, Bulletin 3, p.353.
Parker, W.G., 2006, The stratigraphic distribution of fossil localities
in Petried Forest National Park, Arizona: Museum of Northern
Arizona Bulletin 62, p. 46-61.
Rinehart L.F., and Lucas, S.G., 2001, A statistical analysis of a growth
series of the Permian nectridean Diplocaulus magnicornis
showing two-stage ontogeny: Journal of Vertebrate Paleontology,
v. 21, p. 803-806.
Rinehart, L.F., and Lucas, S.G., 2009, Limb allometry and lateral
line groove development indicates terrestrial-to-aquatic lifestyle
transition in Metoposauridae (Amphibia: Temnospondyli):
Geological Society of America, Abstracts with Programs, v. 41,
no. 7, p. 263.
Rinehart, L.F., and Lucas, S.G., 2016, Eocyclotosaurus appetolatus,
a Middle Triassic amphibian: Osteology, life history, and
paleobiology: New Mexico Museum of Natural History and
Science, Bulletin 70, 118 p.
Rinehart, L.F., and Lucas, S.G., 2017, Late Triassic metoposaurid skull
allometry: Comparison of the Lamy, New Mexico population to four
other populations, New Mexico Geological Society Proceedings
Volume, April 7, 2017, Socorro, NM, p. 59. https://nmgs.nmt.edu/
meeting/2017/NMGS_Spring_Meeting_Program_2017.pdf
Rinehart , L.F., Lucas, S.G., and Heckert, A.B., 2009, Lateral line
development as an indicator of terrestriality in metoposaurid
amphibians: Journal of Vertebrate Paleontology v. 29, p.
171A-172A.
Rinehart, L.F., Lucas, S.G., Heckert, A.B., and Hunt, A.P., 2008,
Preliminary analysis of growth and age structure of Buettneria
(Amphibia: Metoposauridae) assemblages from the Upper
Triassic of west Texas and New Mexico: New Mexico Geological
Society Spring New Mexico Geological Society Proceedings
Volume, April 18, 2008, Socorro, NM, p. 43. https://nmgs.nmt.
edu/meeting/2008/NMGS_2008_SPRING_PROGRAM.pdf
Romer, A.S., 1939, An amphibian graveyard: The Scientic Monthly,
v. 49, p. 337-339.
Schoch, R.R., and Milner, A.R., 2000, Stereospondyli: Handbuch der
Paläoherpetologie, München, Verlag Dr. Friedrich Pfel, 203 p.
Spielmann, J.A., and Lucas, S.G., 2012, Tetrapod fauna of the Upper
Triassic Redonda Formation, east-central New Mexico: The
characteristic assemblage of the Apachean land-vertebrate
faunachron: New Mexico Museum of Natural History and Science,
Bulletin 55, 119 p.
Sulej, T., 2007, Osteology, variability, and evolution of Metoposaurus,
a temnospondyl from the Late Triassic of Poland: Palaeontologia
Polonica, v. 64, p. 29-139.
UCMP locality search, 2018, https://ucmpdb.berkeley.edu/loc.html
Watson, D.M.S., 1919, The structure, evolution, and origin of the
Amphibia, The “orders” Rhachitomi and Stereospondyli:
Philosophical Transactions of the Royal Society of London B, v.
209, p. 1-72.
Zanno, L.E., Heckert, A.B., Krzyzanowski, S.E., Lucas, S.G., 2002,
Diminutive metoposaur skulls from the Upper Triassic Blue Hills
(Adamanian; Latest Carnian) of Arizona: New Mexico Museum of
Natural History and Science, Bulletin 21, p. 121-125.
Zittel, K, von, 1887-90, Handbuch der Paläontologie, I Abt,
Paläozoologie, v. 3, Vertebrata (Pisces, Amphibia, Reptilia):
München & Leipzig, Oldenbourg, 1890 p.
581
APPENDIX 1
Tabulated data. Abbreviations listed near the beginning of the paper.
TABLE 1. Metrics (mm) of the juvenile Apachesaurus gregorii skull (UCMP 171591) and the adult holotype skull (UCMP 63845).
UCMP 171591 is missing an estimated 6 mm of the anterior snout, UCMP 63845 is missing an estimated 14 mm of anterior snout.
These estimates are included in measurements below.
SKULL ROOF Length Preorbit Postorbit Width
A-B A-C A-D A-E A-H C-E at B at C at D GW
UCMP 171591
juvenile 26.4 62.1 83.9 57.5 45.2 61.8 66
UCMP 63845
adult (holotype) 54 130 169 181 115 88 125 130
Features:
NL NW INW OL OW IOW PFL PFW
UCMP 171591
juvenile 5.2 7 13 8.3 23 ~2.1 ~1.8
UCMP 63845
adult (holotype) 22.5 15.3 49.2 4.6 4.6
PALATE Length Features
A-PB A-PC A-PD A-PE ChL ChW IChW PVL PVW AFL AFW
UCMP 171591
juvenile
UCMP 63845
adult (holotype) 25 43 111 16.4 8.4 38 89 38 51 ~36
TABLE 2. Atlas and axis measurements (mm) of juvenile Apachesaurus gregorii (UCMP 171591) and, for comparison, adult atlas
(NMMNH P-22487).
Length Ant width Post width Height Width ant facets
UCMP 171591 atlas 4.5 9 9 5.7 4.4L / 4.3R
axis 4.6 8.7 6.9 7.1
NMMNH P-22487 atlas 17 30.8 25.5 14.9 14.3L / 13.7R
582
TABLE 3. Apachesaurus gregorii selected dorsal intercentrum measurements (mm) of juvenile (UCMP 171591) and adult (all
NMMNH) specimens. N = 11. Width measurements do not include rib facets, only the spool-shaped vertebral body.
Length Ant W Pst W Noto dia Pst H Ant H
UCMP 171591 6.33 6.8 7.04 1.4 5.85 6.02
6.47 7.16 1.8
5.87 6.6 6.4 1.6 5.94 5.97
NMMNH P-17062 16.71 24.7 23.9 1.96 19.45
15.8 18.15 17.1 2.06 13.6 13.6
12.5 18.6 18.24 2.34 17.1 17.9
15.23 17.9 16.6 2.07 18 15
NMMNH P-22486 11.9 13.8 13.7 2
NMMNH P-4753 13.6 17.9 18.2 1.5 15.4 15.6
NMMNH P-4855 8.2 9.5 9.8 1.8 7.9 7.9
8.2 8.3 8.5 1.78 7.8 7.7
TABLE 4. Measurements (mm) of lateral line widths and depths at various skull landmarks (Fig. 4D) for Apachesaurus gregorii
and, for comparison, Koskinonodon perfectum. Note that skull lengths for UCMP 63846 and YPM 4201 are estimated by scaling
incomplete skulls by greatest width in comparison to UCMP 63845, the holotype. No accounting for possible allometry is included
in these estimates. YPM 4201 measured from scaled photographs; all others measured directly.
Taxon, specimen skull a a b b c c d d e e f f g g h h i i
Apachesaurus L W D W D W D W D W D W D W D W D W D
UCMP 171591 83.9 faint 1.3 faint
UCMP 63845 type 169 2.5 1 2.6 0.87 2.4 0.74 2 0.79 2.9 0.8 2.7 0.7 2.8 1
UCMP 63846 ~238 3.5 1.2 3.3 1.1 3.6 1
YPM 4201 ~221 2.3 2.5 3.2
Koskinonodon
UCMP 27039 608 11.1 4.3 10.3 3.6 10.5 3.6 11 1.8 14 4.3 11.5 2.5 18.8 4.7 20 5.5 17 5.4
NMMNH P-36228 373 8.4 3.6 8.8 2.8 9.6 3.7 8.4 2.4 10.4 3 7.9 2.5 10.2 2.9
NMMNH P-36229 169 2.8 0.8 3.3 1 3.3 1.2 3.9 0.9 2.9 1.1 1.7 0.8 5.4 1.4 3.5 1.2 3.3 1.5
NMMNH P-61420 522 8.4 3.2 9.2 3.2 9.5 2.4 13.3 1.3 12.5 3.1 10.7 3 9.3 2.5 15.3 3.8 15 1.7
NMMNH P-61419 339 6.4 2 6.3 2.5 5 1.2 5.9 1.6 4.9 1.3 4.6 2.6 8.2 3.2 5.5 2.7
583
TABLE 5. Allometric constants for relative growth of lateral
line width and depth at various landmarks in the skulls of
Apachesaurus gregorii and Koskinonodon perfectum. Constants
preceded by ~ are based on only two skulls.
Koskinonodon Apachesaurus
width depth width depth
a 1.03 1.29
b 0.9 1
c 0.94 0.82
d 0.93 0.45
e 1.28 1.06
f 1.56 0.96
g 0.8 0.76 ~0.55 ~1.18
h 1.38 1.12 ~0.59 ~1.3
i 1.3 0.67 0.96
TABLE 6. Koskinonodon perfectum dorsal intercentrum
measurements (mm) from the Lamy, NM locality. N = 11. Rib
facets not included in width measurements.
Specimen Length Med-Lat
width Height H/L
NMMNH 30.8 58.4 53.7 1.74
uncataloged 31.9 48.6 1.52
21 42 36.1 1.72
36.3 75.2 56.7 1.56
28.1 51.5 1.83
24.9 48.4 37.4 1.5
26.1 54.2 46.9 1.8
37.1 83.8 71.3 1.92
26.9 43 41 1.52
46.9 90.2 69.4 1.49
43 88.6 2.06
Average: 1.697
