CES on ae iy ba 4 es Sor Sar asl Fons wach We Fe SE Mig Th ONAL ruts ke Al Regt) v= a
# . if “ener as : if Raiaaie tH IRA APOE “ milage sn2e Tors pices: ‘ x ry . . Sab ote® a? a ee SS ee
mapas y . r 7 : EAN Ih aig n ¢ lh prt oe ee I EOL ee T= Ee Sa
5 ee Saul i A 6 en ert Ms TES
Seale oka! athe dl rg ee
- Neuen ray” 8 Rr Ae pert eer
. Se er
Se ae ee ae at SS ee pr etree

Se : . Wet won are :
= ett Lounger SA = Sr preersetnr ateg aed
pts yesiee

Eade To ded Des S nes hehe Ade ET ETRE EDGE p REREAD
» eam chsega Y Frat eS SacTeiscs otis i Pi So 3 ’ > Singh toe ‘ ‘ Me eit ods
5 ee es nee z Pe Twp ers a ENE S :
sty ithe dace de® a ee TE ees
ALP iad RRs NE ~ . :
eo So Te TL Fag Nh BOER SEE tO gh MOT Bh et OH
esata et
Ps Bath ted
¥ ye ees ~ S she: é 2 ’ ere ee
ye tee TN OLN Date Sikes Wntrndey.** 22, ta thes ti Bapitast oes a s . ~ . ° 6 ‘ . - =‘ lag alge ew a oe
oie ton Wat mt hata a posers : . 5 < = "ud sie 5 : Danae AWA EG I he
ESET OR ENT. : i é res pe 4 ‘ : <a
alte TO ee “
aac Paded : - Ms
oe wi . tiie ¢ " ~ = - :
Se ad > - ; Loe a : ein :
"Teen aeo ‘ I nt ’ Ped eS ? - . : ; . . Ea er
aanser : 3 etn : Near - 2 : : 3 a Ee hr gation
seer. ¥ : ee ee . Z a ra : ts SES — p - 1 ngs salle ‘ a” . lia ioe ald i eda
— pone! ms : <p - sem “a c . ~ % 5 Ho SIM 2S Ar ON eon al nly Ia Yaa ys 5 :
Sentehibts ts : - Es inti intaderehchaieiiecuiaedh ‘ats Teco kee . . : . . vars
7 aon Wie Mg tether a
om atin ct Tw we % me Pare eo
Setting Pay me tag Bene pa a et TS le pe) RE Le etn et A Rt Fe op ee ee
ors oe aS ee et OE > er a ee i ee ae tO ae
. Beh ote ts ae BA shag Pagans MF Re UNIS en Ralhe = ene ee ee ee
Ss VRE Cr ee Sees rgee a ee ee ee ee ee
a ~ ; aes ee tee aes le ee en eel ode —_ > Paes
mag seinen . = “ : - news = . ra ) oo am Rm wa aR Pe Oe ge ihe eNO <a t- ek Ae
ings Arve = r ; ; ~ * % . : : . = . OE eg be Sete ~ 4
$ “ PRD 4 ~ Na te eee An AE .

were

sede

(Or

ROYAL ONTARIO MUSEUM

Digitized by the Internet Archive
in 2011 with funding from
University of Toronto

_http://www.archive.org/details/latepleistocenefOOedmu

Pee SCTEN CES CONTRIBUTION 63

Ae SOR DON) EDMUND

A Late Pleistocene Fauna

from the Santa Elena Peninsula,
Ecuador

Sorel ON TAR DO MUSE UM e) oU NEM ER SET Y.: OF “TORONTO

MUSEUM LIBRARIES

SM.

Contribution No. 63
LIFE SCIENCES
ROYAL ONTARIO MUSEUM

UNIVERSITY OF TORONTO

\. GORDON EDMUNDA Late Pleistocene
Fauna from the Santa
Elena Peninsula,
Ecuador

A. G. EDMUND is Associate Curator in charge of the Department of Vertebrate
Palaeontology Royal Ontario Museum, University of Toronto.

PRICE $1.00
© The Governors of the University of Toronto, 1965

PRINTED AT THE UNIVERSITY OF TORONTO PRESS

Our knowledge of the Pleistocene fauna of the coastal regions of Ecuador
is based largely on the material from the tierra brea found between Santa
Elena and Salinas. This has yielded a great quantity of well-preserved
bone which has been described by Spillmann (1931, 1948) and later by
Hoffstetter (1952).

In 1958, the author was shown a new locality containing terrestrial
vertebrates of Upper Pleistocene age in an apparently estuarine deposit, in
association with marine fossils such as mollusks, echinoderm spines and
shark teeth. With the cooperation of Mr. S. Marchant, Chief Geologist of
the Anglo-Ecuadorian Oilfields Ltd., a small collection was made, suffi-
cient to identify the fauna as being a late Pleistocene assemblage. The
locality, shown in Fig. 1, is on the bank of a tributary of the dry Rio
Engabao, about half way between the villages of Ancon and Atahualpa,
approximately 80° 49’ W. longitude, 2° 18’ S. latitude.

:
ae

Bahia de Santa Elena

SALINAS

W SANTA ELENA

ew enme-

EO Ssh
oe he ray

IN a Ne

Lowi ATAHUALPA
Persie

GUAYAQUIL Sa

CG

yp ree Golfo de

Guayaquil

Figure 1. Map of the Santa Elena Peninsula, Ecuador.

The author returned to the same area in 1961. On this occasion, with
the assistance of Mr. Neville Lindner, who succeeded Mr. Marchant,
bones were found at three sites in the same tributary system, and a repre-
sentative collection was secured. Since the localities are close together and
of very similar lithology and horizontal level, they are considered to be
contemporaneous. Some of the bones were incomplete in situ, and all were
soft and fractured so that the larger elements had to be collected in plaster
bandages. There was, however, little or no evidence of water wear, and
the bones were remarkably undistorted.

Study of the bones was greatly facilitated by having for comparison two
collections from the famous brea pits in the area between Salinas and
Santa Elena. One, in the Escuela Politecnica Nacional (E.P.N.) in Quito,
formed the basis of much of Hoffstetter’s 1952 monograph. Thanks to the
kindness of Prof. Gustavo Orcés, the author was able to measure and
photograph a large part of this collection. The second reference collection
is in the Royal Ontario Museum (R.O.M.) in Toronto, Canada, and was
made in the vicinity of La Libertad (at or near Spillmann’s “Coralito”
locality) by the author in 1961. The geology of these vertebrate deposits
has recently been studied by Dr. R. R. H. Lemon of the Royal Ontario
Museum as part of a survey of the stratigraphic history of the Santa Elena
Peninsula. Earlier studies were made by Spillmann (1948 ef al.) and by
Hofistetter (1952), the latter giving a good bibliography of the subject.

The geology of the Santa Elena Peninsula is complex, and has given rise
to conflicting opinions. The area is characterized superficially by numerous
flat erosional surfaces, locally known as tablazos, found at different eleva-
tions. The correlation of these surfaces, some of which are faulted and
raised to different levels, has been a source of dispute among the various
geologists who have examined the area.

The Santa Elena Peninsula is underlain by a sequence of late Cretaceous
and Early to Middle Tertiary sediments of various types including conglom-
erates, sandstones, greywackes and shales. These are complexly faulted
so that the precise structure is still in some doubt despite the accumulation
of much data from well logs. References to the subsurface work are sum-
marized in Marchant (1961).

The Cretaceous and Tertiary rocks are overlain by several wave-cut
terraces (tablazos) marking erosional surfaces developed at intervals dur-
ing the Pleistocene when the land surface was considerably lower with
reference to sea level than at the present time. The deposits associated
with the marine terraces range from normal shelly beach sediments to fine-
grained silts and poorly sorted sands of estuarine type. These merge land-
wards into downwash deposits, alluvial fans and fluvial gravels of
Pleistocene to Recent age derived from the Cerros de Estancia.

The sediments containing the vertebrate remains described here are of
estuarine type and were probably laid down by a large stream ancestral to
the modern Rio Engabao, but with considerably greater run-off than the
present stream. The vertebrate localities probably were near or at the
mouth of the old river, and at or very close to sea level. Conditions very
similar to those postulated for the Pleistocene Rio Engabao can be seen at
the present time at Palmar, about eighteen miles to the north, where a
small stream reaches the coast and spreads out into a brackish water
swamp with mangroves.

The vertebrate remains found in the brea pits between Santa Elena and
Salinas, although occurring in sands and silts soaked by pitch, were prob-
ably buried in a similar estuarine environment. In this case it is likely that
a stream ancestral to the modern Rio Grande emerged in this general area
and that there were estuarine swamps or salt flats at, or very close to, the
sea level of that time. It seems probable that the introduction of the

4

petroleum into the deposit is a relatively recent phenomenon. The present
elevation of the marine terrace at the brea pits (Spillmann’s “Coralito”
site) is approximately twenty-five feet above sea level, less than at the Rio
Engabao vertebrate site. Thus it may be younger, but the precise ages of
both localities await the results of radiocarbon determinations.

A detailed report on the Pleistocene geology of the Santa Elena Penin-
sula is being prepared by Dr. R. R. H. Lemon, who kindly supplied the
geological notes summarized here. The author also wishes to acknowledge
the very generous assistance of the management and staff of the Anglo-
Ecuadorian Oilfields Ltd. In particular, I wish to thank Mr. S. Marchant
and Mr. Neville Lindner, Chief Geologists, for providing information and
making the innumerable arrangements which made possible the field work
in the area. Dr. C. S. Churcher, Research Associate of the Royal Ontario
Museum, provided information on the genera Odocoileus and Smilodon
and supplied helpful criticisms of the manuscript. Thanks are also due to
Professor Gustavo Orcés of the Escuela Politecnica Nacional in Quito for
making that rich collection available for study. The turtle material was
identified by Dr. Walter Auffenberg of the Florida State Museum, Gaines-
ville, Florida.

CLASS MAMMALIA
Family MEGATHERIIDAE
Eremotherium cf. carolinense
The most abundant material from the Rio Engabao site is that of a large
megathere. The definitions of the genera and species of megatheres are
not on a firm basis, and are currently being studied by the writer. It is
generally recognized, however, that the genus Megatherium, as typified by
M. americanum, did not extend into the northern part of South America.
The northern megatheres belong to a distinct genus, Eremotherium, known
from Brazil, Venezuela, Colombia, and Ecuador, as well as Central
America. Several species have been proposed, E. carolinense Spillmann
and E. rusconii (Schaub) being the best known.

Hoffstetter (1952) described a smaller megathere from the coast of
Ecuador under the name Eremotherium elenense. This has subsequently
been collected in Peru by the writer, and appears to be very similar to other
specimens from Chile and Bolivia. Because of its significantly smaller size,
and other features, we can safely exclude this group from consideration
here.

The material from Rio Engabao was compared with specimens of
Megatherium americanum trom Argentina, and with several populations
of the genus Eremotherium from Ecuador, Colombia, Venezuela, Brazil
and Panama. On the basis of the elements in the Rio Engabao fauna, it is
possible to rule out the presence of Megatherium, and to assign the material
to Eremotherium. Because of the great individual variation inherent in the
megatheres, and because of the lack of more diagnostic elements in the
Rio Engabao fauna, it is difficult to assign a specific name to the material.
However, the tibia appears to indicate a much closer affinity to FE. caro-
linense than to other species. Also, a well-preserved partial skull of
E. carolinense in the Peabody Museum, Yale University (Y.P.M. 14159)

5

was collected in what appears to have been a similar type of deposit not
very distant from the Rio Engabao site. Hoffstetter (1952) gives measure-
ments of and illustrates a restored cast of this specimen (E.P.N., V.1508).
For these reasons, the material from this site will be assigned to FE. caro-
linense.

Many specimens of Eremotherium are now known from sites across the
northern part of South America and through Central America into the
south-eastern United States. These have often been given different specific,
subgeneric and generic names, some of which will certainly prove to be
synonymous. A review of the group is given in Hoffstetter (1952) but
numerous other specimens have come to light since that date.

Material for comparison with the Rio Engabao megathere represents
several populations. The well-known tar-seeps of coastal Ecuador, des-
cribed by Hoffstetter (1952), lie near the town of La Libertad, between
Salinas and Santa Elena, in the Santa Elena Peninsula. Collections from
these seeps are in the Escuela Politecnica Nacional (E.P.N.) in Quito,
Ecuador, and in the Royal Ontario Museum (R.O.M.) in Toronto,
Canada. A Panamanian population is represented by the extensive collec-
tions from the Province of Herrera, Panama (Gazin, 1956) deposited in
the Museo Nacional de Panama (M.N.P.) and the United States National
Museum (U.S.N.M.) and has been assigned to E. rusconii (Schaub). The
Venezuelan material does not represent a distinct population, but rather
scattered occurrences which seem to represent the same species at about
the same geological age. These occurrences include the type of
Eremotherium rusconii, the material from San Miguel in the museum of
the Collegio de La Salle in Barquisimeto, both from the Estado de Lara,
and the Muaco fauna (Estado de Falcon) collected by Royo y Gomez
and Cruxent and now in the Universidad Central, Caracas. The Colombian
material is best represented by an articulated skeleton in the Museo de
Servicio Nacional de Geologia, Bogota (M.G.N. 2000), excavated by
Stirton and Royo y Gomez in Huila province, and referred to E. rusconii
(de Porta, 1961). The Brazilian specimens have been referred to various
species (rusconil, laurillardi, lundi, etc.) but all may prove to be rusconii.
As with the Venezuelan specimens, the Brazilian material is from numerous
localities of uncertain geological ages. Most of it is in the Museu Nacional
in Rio de Janeiro (M.N.R.J.).

Material assigned to Eremotherium carolinense includes the following:

An Isolated Tooth, R.O.M. 3895, of the usual rectilinear pattern with
two sharp transverse ridges. Comparison with cranial material from the
Santa Elena Peninsula confirmed it as a typical second upper tooth. The
size agrees well with that of similar teeth from Venezuela, Colombia and
Panama. Antero-posterior length is 43.4 mm., width at anterior ridge is
49.1 mm., and width at posterior ridge is 49.2 mm.

A First Cervical Vertebra (atlas), R.O.M. 3896, recovered in frag-
mentary condition, lacking both wings. There is little difference in the form
and measurements of this bone between any of the known large megatheres.
The essential measurements fall into the ranges of the populations from the
Santa Elena Peninsula and from Panama. Fifteen measurements were com-
pared, five of the more important being shown in Table I. The Rio

6

usuoads
ovqesuy Ory
9688 “W'O'ed

v8T g 6L1
8g 4g vS
98 fi O8
961 G UL
6ST Gg eri
x N THA
lopenoy

‘epnsuluad eus[y eJUeS WOT]
ISUBUYJOADI WNLAAY OMA

CRT a SOI C1Z sasAydodesAz4sog ssoioe YyIpI\\
o¢ € OF 9¢ sasXydodesAzaig Jo YIPIM\
$6 Ve CR FOL sasAydodesAzaig jo yysuaT]
9Z1 € CIT 681 _ VYslay 3s97vaI5)
Crt ¢ OOT CRT sasAydodesAzaig ssoioe YIpIA\
xX N AN ETN

rueueg WO]
WUUOISNA WNLABYJOWILA

(S¥]}B) LIG9IIOA [VOTAIIO JSIY 9Y} JO S}JUBWINsvoU dATVeIedWIOD “| ATAV]

Engabao atlas appears to belong to a mature individual, but since it lacks
the wings one cannot determine the degree of fusion of the epiphyses.

The Sixth Cervical Vertebra, R.O.M. 4304, is almost completely
preserved. The morphology and dimensions are almost exactly the same
as those of R.O.M. 4311, a specimen of E. carolinense from the brea pits
near La Libertad, Ecuador. Table II gives the essential measurements.

TABLE II. Measurements of the sixth cervical vertebra of Eremotherium

Rio Engabao Eremotherium carolinense
Specimen R.O.M. 4311, Santa

R.O.M. 4804 Elena Peninsula

Greatest height 190 204

Width across transverse processes Zou ——

Height of centrum 72 *68

Length of centrum 73 HZ

Greatest width of neural canal 73 (5

Width across postzygapophyses 138 147

*Estimated measurement.

Other axial material definitely referrable to Eremotherium includes the
following:

4307 Cervical vertebra III, centrum

4309 Mid-thoracic vertebra, centrum

4308 Caudal vertebra, about No. 12, centrum

4306 Proximal caudal vertebra with part of the neural arch

3899 Sternal element; agrees in size and form with Eremotherium

specimens from the Ecuadorian brea pits.

Figure 2. Medial view, right lunar of Eremotherium, R.O.M. 3898.

8

O€
oS
L9
bP
89
9G
Is
SS
611

uautoeds
ovqesuy Oy

WOd

62
cS
62
[ks
16
06
Icl

PIGUo]O.) “000%
ND'IN Wuoosns
MNLAAY OWI

LE
8¢
GL
LY
OS
Ig
SOT
06
SEI

xX

7 0g cP 9¢ ! 8Z cP UOTYINIAAV WOPaUND JO YIP! \\
re CG e9 RC kt CG £9 UOT E[NoMse WdOflgunog jo 1Y4s1I9}
7 19 eg 08 n e) €8 UOHYINIATY wWoOslouN Jo IYysIaH
F fae 1G 6F i IF 1c uolyRE|NIyIe wiNnusew JO YyIpIA\
a red) CR OS ) py £8 uoTqE[|NoyIe wnuseU JO IYSIOPY
r 0g ce 6Z ! Ve 08 UOHYINIAAe proydvos JO YIP! \\
r cy CIT C6 D C8 TOI uoNYjNoAe proydeos jo yi3ue7]
7 PL 16 cy ! 6L 68 UONZ[NAIe [eIPes JO YIP! AA
7 val 6F1 LEI L Wal OST UOTY|NAIL [eIpes JO YSU]
N MA “XIN x N UA = “XP
JOpenoy ePUeUued wot}

‘e[nsuluead eusa[y eJURSG WOT]

ISUIUIJOADI WNLABYJOMWIAT

WUOISNA WNLABYJOWILA

souoq Ivun] JO sjUaWaINsvaw aATVeIedwoy “[]] AAV,

4310 Occipital condyle, damaged, 93 mm. long.

THE ANTERIOR LIMB is represented by a right lunar, fourth meta-
carpal, and phalanx 2, digit IV.

The Lunar, R.O.M. 3898, is perfectly preserved and closely resembles
the specimens of Eremotherium from Colombia, Panama and Ecuador. As
shown in Fig. 2, the main contact with the scaphoid is distinctly separated
from the distal articulations by a broad and rather deep groove, the floor
of which contains several deeply penetrating canals. There are two secon-
dary articulations with the scaphoid, a larger dorsal facet, 21 x 28 mm.
and a narrow ventral one about 6 & 25 mm. Both are continuous with the
facet for the magnum, but are clearly defined from it. These secondary
articulations are common in all megatheres, and are quite variable in size
and configuration.

The articulations of the lunar are identical in all Pleistocene megatheres,
with only minor proportional differences. Measurements of comparable
material are shown in Table III. These indicate that the lunar from the
Rio Engabao site is from a rather small individual, but that it belongs with
the populations of Eremotherium from Colombia, Ecuador, and Panama.

Metacarpal IV, R.O.M. 3171 (Fig. 3) is perfectly preserved, except
for some damage to the proximal articulations. Unlike the lunar, the fourth

re

1 RS
‘a 4.

aes
ag eH 0 2 4 6 8 10 Cm.
eS a a ras TE eae

c 5S

Figure 3. Lateral view, left metacarpal IV of Eremotherium, R.O.M. 3171.

metacarpal of Eremotherium is distinguishable from that of Megatherium.
The general form, length, and relationships of the articulations are the
same in both genera. Measurements of the bone from the Rio Engabao
site were compared with those of several specimens of M. americanum. In
the latter, the bone is generally of lighter construction, and the articulations,
especially those for the third metacarpal and unciform, are smaller.

The Rio Engabao metacarpal was compared with those from other
populations of Eremotherium (Table IV), and the proportions were found
to be in excellent agreement. The Rio Engabao bone obviously belongs to
a large individual, but it is not beyond the range expected for this highly

10

ve 96 Sa 30 FT y 00% JE!
ani ont ie! cel Vv G I Sf
VG 1G a GG Vv 1G ¥G
OS LG 1g LG Vv GG OE
a6 oS 8V GS Vv US ag
ke 96 =? G& V 6G 9€
9& SE igs GE v SS 8E
ISG STG COG CLG «OV SIG G8G
uaunoeds eIQUIO]OD) (qse_) xX N “UlT “XP
ovqesuy Ory 000 “N'D'IN MUOISNA me
Pale aN, Orel CUM OOST Mawel = MANA ROUN A AGT, Jopenoy
jo adAy, ‘e[NSUIUdG PUI{H VIULS UOT]

ISUIULIOLDI WNLLAYJOW AAT

"yy Sua] JSso}eaI3 JO 94 se passaidxy,

90 1 Co oon” Cet
Gt ci ae val
LG v 6G O&
6G SI 9G GG
Og 9 LY oS
GE L OE VE
ve L OE 66
L8G CLG OGE
x N ELIA eel

eureurg WOl]
WUOISNA WNLAIYJOWILA

YIPIM [RWIXOIg
1ysIoy [PWIXO1g

O1Ve

YIPEM
1319

yoory A] [edieorqayy ‘oney

y}oory A [edivorjayy
JO 1YyS10y [PUIA-OS1OC]
4Jooe} WIOFIOUN
jO 1YySIey [e41UIA-OSIO(]
«ty Stoy [ISIC]
xYIPIM [PUNTXOIg
x YSIoy [PUNTXOIg
y1Sus] Js9} vaIt)

A] [edaeorjour jo suonsodoid pur sjuaweinsvaul sAtyereduoy “\] ATAV

Il

0 2 4Cm.

Figure 4. Lateral view, right phalanx 2, digit IV of Eremotherium, R.O.M. 3170.

variable group. It seems impossible however, to distinguish between the
various populations of Eremotherium from the measurements in Table IV.

Phalanx 2, Digit IV, R.O.M. 3170, is illustrated in Fig. 4. It is
assigned to Eremotherium on the basis of its general form, and on the
measurements in Table V. As with the corresponding metacarpal, it repre-
sents a large individual, but not one too large to be part of either of the
populations for which comparative measurements are available. Morpho-
logically it is indistinguishable from them.

TABLE V. Comparative measurements of phalanx 2, digit IV, manus, of Eremotherium

E. carolinense,
Santa Elena

E. rusconii, Peninsula,
Panama Ecuador R.O.M. 3170

Mean of 2 Mean of 4 Rio Engabao

specimens . specimens specimen
Proximal height 80 74 83
Proximal width 79 67 19
Length 52 51 5D
Proximal articular width 65 62 63
Proximal articular height (2 68 (Gh
Distal articular width 62 55 68
Distal articular height, lateral 58 63 69
Distal articular height, medial 50 49 62

The corresponding phalanx in Megatherium americanum (based on a
single specimen in Museo de La Plata, MLP 2-190) is of about the same
length, (52 mm.) but is about twenty percent smaller in most other
dimensions. This, too, agrees with the lighter construction of the fourth
metacarpal in Megatherium.

THE POSTERIOR LIMB is represented by a right tibia, fibula and a
fifth metatarsal.

The fused tibia and fibula, R.O.M. 4312, is well-preserved, but lacks
most of the fibula (Fig. 5). The bone is from a mature individual, as
shown by the well-closed epiphyseal sutures. However, it is not especially

12

Figure 5. Tibia of Eremotherium, R.O.M. 4312. A—anterior view. B—posterior view.

LIBRARY

ROYAL ONTARIO MUSEUM ia

igs GE ete Vv G& GE ve
= GS OS Vv 6G G& G&
1G GG 1G v OG 1G &G
9G 8G LG Vv xe 6G SG
LV 1S LV v Sv 6F iS
8S = 8¢ & og 09 G9
er Ogg 16¢ v O8¢ 86S cZS
usuttoeds PIquojOZ) 4 N “ULI “XP Bx
ovqesuy Or = 0006 “N"D"N —
cley WOU PMO Jlopenoy

‘eynsuluag eua]y PUG WOT]

ISUIUTIJOADI WNLAIYJOWIAGT

9 O€ LE

5 O€ VE
Sj 1G VG
F LG 6G

[ SP a

7 8¢ F9

9 GI¢ Fo9
N SCLIN Sey

vueued Wo]
UUOISNA WHLAIY OWI

JIL] II] YIOI] IS1v|
JO Yy Sug] 10119}sod-o19]uy
YIPIM AR[NVAe [eRIsIqG

JIL] [LIOW] |LIPItu
YIPIM asJOASURI |

JOO] |LIOW] |eIpsw
YySuy9] 1O1Ia}sod-o19] Uy
YIPIM TLPNI VIE [LWIXOIg
YPM [RUTXOIg

"UU UL (25007) Y2sueg] [e107

Y}8uyg] [v0] JO sadvjUsoIed sv passaidxa avIqh Jo s}UsWAINSvITY “[.\ ATAV],

14

rugose, and therefore not of particularly advanced age. The morphology of
the tibia is quite constant in all of the large megatheres studied, and no
obvious differences were noted between those of Megatherium and Ere-
motherium. The description of Owen (1859, p. 813) applies equally well
to both. Measurements of the tibiae of Megatherium americanum agreed
fairly closely with those of the several specimens of Eremotherium, except
that the proximal width and proximal articular width are both significantly
greater in the former. The proportional measurements given in Table VI
show closest agreement between the Rio Engabao specimen and the fauna
from the Ecuadorian brea pits. The proportional measurements for the
proximal width, proximal articular width, and length of the median femoral
facet all fall well within the range of the specimens of EF. carolinense, but
are completely or almost completely out of the range for the Panamanian
material. This is one of the few instances where there appears to be a valid
distinction between these faunas, as based on elements in the Rio Engabao
fauna. Thus, the tibia may be assigned to an individual from the same
population as the specimens of E. carolinense from the brea pits of the
Santa Elena area.

Figure 6. Medial view, right metatarsal V of Eremotherium, R.O.M. 3892.

The Fifth Metatarsal is perfectly preserved, and represents a thoroughly
mature individual, as shown in Fig. 6. The Rio Engabao metatarsal was
compared in detail with seventeen others representing Eremotherium popu-
lations from Panama, Colombia, and the Santa Elena Peninsula, Ecuador,
as well as Megatherium americanum from Argentina. The measurements
and proportions of all of these were remarkably similar, and no valid dis-
tinction between populations could be detected on the basis of elementary
statistics. The measurements of the Rio Engabao specimen are given in
Table VII.

The Rio Engabao megathere material can confidently be assigned to
the genus Eremotherium on the basis of several elements. Reference to
the species E. carolinense is likely on the basis of measurements of the
tibia, and on the geographic proximity of the site to that of the type
material.

IS

TABLE VII. Measurements of metatarsal V,
Eremotherium sp. R.O.M. 3892

Length 221
Proximal height 93
Distal height 70
Distal width 69
Length of cuboid articulation 71
Height of cuboid articulation 68
Length of Metatarsal IV articulation 59
Height of Metatarsal IV articulation 65
Height of distal articulation 27
Width of distal articulation 20

Family DASY PODIDAE
Pampatherium cf. occidentale

This species is represented by a single scute (R.O.M. 3891) from one of
the fixed shields. I have assigned it to Hoffstetter’s species purely on
geographical grounds, since it is impossible to be more precise on the basis
of an isolated scute.

The specimen measures 45 mm. long, 39 mm. wide, and 10.8 mm. in
maximum thickness. It is apparently from close to the margin of the
shield, since the central raised portion is displaced well toward one end.
The surface pattern is typical of the genus, agreeing in detail with the
numerous scutes from the brea pits of Talara, Peru and the Santa Elena
Peninsula of Ecuador, as well as with those from North America.

The generic name Pampatherium Ameghino 1875 has been adopted
in preference to Chlamytherium Lund 1839 or Holmesina Simpson 1930.
Chlamytherium, which appears to have been a misprint for Chlamydo-
therium, was used in 1838 by Bronn for a glyptodont, and thus becomes a
nomen nudum, As pointed out by Paula Couto (1954) the valid name is
Pampatherium Ameghino 1875.

Holmesina was proposed by Simpson 1930 for the chlamytheres of
North America. The great similarity between the North American and
Ecuadorian forms was recognized by Hoffstetter (1953), and the Ecua-
dorian species C. occidentale transferred to Holmesina. The writer has
recently examined most of the available specimens, and found little evi-
dence to support the generic difference between classical Chlamytherium
and the species referred to Holmesina. Until a more comprehensive study
has been made of new material now on hand, the genus Pampatherium is
used for the Ecuadorian material, and the Rio Engabao scute is assigned
to the species occidentale on the basis of size, form and geographical
location.

Family EQUIDAE
Equus (Amerhippus) cf. santae-elenae
Two specimens were recovered, a single tooth, R.O.M. 3169, and part of
a proximal phalanx, R.O.M. 3894.

The tooth is a badly worn lower third molar, shown in Fig. 7. It
resembles Hoffstetter’s species Equus (Amerhippus) santae-elenae, but
there are minor differences. The measurements are given in Table VIII.

16

Figure 7. Lower third molar, Equus (Amerhippus) santae elenae, R.O.M. 3169.

TaBLeE VIII. Measurements of Equius molar from Rio Engabao compared with speci-
mens of FE. (A.) Santae-elenae

E. (A.) santae-elenae

Rio Engabao

Usual range* R.O.M. Specimen
Two very old d490F R.O.M.
Max. Min. specimens* ‘Talara, Peru 3169
Antero Posterior-length 31.3 29 .6 32.6-35.5 36.0 36.1
Width 15e8 13.4 16.0 15.5 ; 16.0

*From Hoffstetter 1952.

No other species of horse was found to resemble the Rio Engabao speci-
men. The geographically closest species, E. andium is considerably smaller
and not morphologically similar. Also, it frequented higher altitudes, and
thus is unlikely to be found in a coastal fauna. The measurements of
R.O.M. 3169 indicate reasonably good agreement with specimens of E.
(A.) santae-elenae from both Peru and Ecuador.

The cusp pattern, however, is unusual in one detail. The entoconid
(Hoffstetter’s terminology) is less well developed in the Rio Engabao
specimen than in either the specimens illustrated by Hoffstetter (1952) or
those from Talara, Peru. In both of these latter groups of specimens the
enamel of the entoconid is folded inward and posteriorly for about 3 milli-
metres. This fold is completely lacking in R.O.M. 3169. Possibly this is a
feature of advanced age, since the crown is very badly worn. The most
nearly comparable tooth from Talara, Peru, and identified as E. (A.)
santae-elenae, has at least an additional one centimetre of its crown height
remaining. However, Hoffstetter illustrates the lower dentition of a 15
year old individual of the same species with a well marked fold on the
entoconid. The crown height of this tooth is unknown.

Specimen R.O.M. 3894 is a fragment of the distal end of a proximal
phalanx of the right manus. It agrees well with specimens of E. (A.)
santae-elenae from Talara, Peru. Unfortunately, however, it is of no value
in determining the specific identity of the animal.

The horse specimens from Rio Engabao can be confidently assigned to
Equus (Amerhippus), and at least tentatively to the species santae-elenae.

Family CERVIDAE

Odocoileus cf. salinae

Two specimens, a right astragalus and an antler tine, indicate the presence
of a deer similar to that found in the brea pits of Peru and Ecuador. Since
no more diagnostic material was recovered, it is impossible to assign con-

Lf

fidently a specific name to the Rio Engabao cervid. However, the astragalus
was compared with the figures of Frick (1937), Hoffstetter (1952) and
with the original material from Peru described by Churcher (1962) and it
agrees well in all respects with Odocoileus salinae Frick.

Ex ahead
BUS fat

Figure 8. Right Astragalus of Odocoileus cf. salinae, R.O.M. 3168. A—dorsal view.
B—medial view.

TABLE IX. Measurements of the Odocoileus astragalus from Rio Engabao compared
with those from Talara, Peru and the Santa Elena Peninsula

alara, Peru Santa Elena Peninsula

Churcher (1962) Hoffstetter (1952) Rio Engabao
c = R.O.M.
Min. Max. N xX Min: sViax. . IN See 3168
Proximo-distal
length ol. 2 af.07 20’ 33.2 3087 “3bi8: 3 S— 34.4
Distal width ZONOs 227.9% NOT 216 RS SiiZona two + 22Ke

The astragalus, R.O.M. 3168, shown in Fig. 8, is from a mature
individual and well preserved. The measurements given in Table IX show
that it lies well within the range of O. salinae. Churcher points out that
Hoffstetter’s material from the brea pits of the Santa Elena Peninsula repre-
sents larger individuals than that from Peru. The Rio Engabao astragalus
is at the upper end of the range of measurements for the Peruvian material,
but is well within the range of that from Ecuador. Although it is based on
very slight evidence, one might conclude that the Rio Engabao specimen
belongs to the population of slightly larger individuals characteristic of the
Santa Elena region of Ecuador.

The antler fragment, R.O.M. 3892, is part of the tip of the main, or
terminal tine, 54 mm. long. At the proximal end it is somewhat triangular
in section and measures about 11.6 mm. in greatest thickness. This is
about one-third larger than a comparable fragment from Peru, but this
measurement is extremely variable with sex and age.

Family MUSTELIDAE

cf. Lutra sp.

A small carnivore is represented by an isolated vertebral centrum (R.O.M.
3172), lacking all of the neural arch, spines, etc. It measures 21.4 mm.
long, and about 15 mm. across the central faces. On comparison with
modern material it proved to be quite similar to the eleventh thoracic
vertebra of Lutra, both in size and form. If it is assignable to this genus
it represents a large form, but no larger than some present-day species in
South America. The fragment, however, is insufficient to do more than
indicate the probable occurrence of an otter in the Rio Engabao fauna.

18

Family FELIDAE
Smilodon neogaeus (Lund)

The sole evidence for this species is a perfect specimen of the proximal
phalanx of digit 1, R.O.M. 3890. However, this bone (Fig. 9) is quite
distinctive and agrees with the large series from Talara, Peru. The measure-
ments given in Table X, fall well within the range for S. neogaeus. The
measurements for the Talara material are supplied by Dr. C. S. Churcher,
whose revision of the genus Smilodon is nearing completion.

Cm.- 0

Figure 9. Proximal phalanx, digit 1 of manus of Smilodon neogaeus, R.O.M. 3890.
A—dorsal view. B—medial view.

TABLE X. Measurements of the proximal phalanx, digit I, manus of Smilodon

Specimens from Talara, Peru
— R.O.M. 3890

Maximum Minimum Ie Rio Engabao
Total length 39.6 34.7 37.3 36.5
Proximal width 27.4 23.2 24.8 24.1
Proximal height 21 19 18.9 202 19.0
Distal width 26 .0 21.8 23.4 2220)
Distal height MG a2 et Ko 7 14.5

*N = 25.

Panthera atrox was also present in North-Western South America, as
revealed by a fine skull and other elements collected by the writer at Talara,
Peru, in 1958. The proximal phalanx of this animal is compared with that
of Smilodon in Merriam & Stock (1932), and there is no doubt that they
are morphologically distinguishable. The Rio Engabao specimen clearly
does not belong to Panthera atrox.

Hoffstetter (1952), had very little Smilodon material and did not
attempt to assign a specific name to the material from the Ecuadorian brea
pits. The writer was, however, fortunate in recovering material of excellent
diagnostic value from that site in 1961, and there is no question that it also
belongs to S. neogaeus.

CLASS. REPTILILA
The only reptiles in the fauna are chelonians, represented by fragments of
shells and a few limb bones. These have been examined by Dr. Walter
Auffenberg. One tortoise is present, tentatively assigned to Geochelone
(Testudo) cf. G. (T) gallardoi on the basis of parts of the plastron and
carapace. Also present is Geoemyda sp., as well as some unidentifiable
turtle scraps.

19

CONCLUSIONS

The list of terrestrial fauna is as follows:
Class Mammalia

Order Carnivora

Smilodon neogaeus

cf. Lutra sp.

Order Perissodactyla
Equus (Amerhippus ) cf. santae-elenae

Order Artiodactyla
Odocoileus cf. salinae

Order Edentata
Eremotherium cf. carolinense
Pampatherium cf. occidentale

Class Reptilia
Order Chelonia
Geochelone cf. gallardoi
Geoemyda sp.
Turtle indet.

The scattered and ill-sorted nature of the vertebrate remains, with an
admixture of marine material, argues strongly for the estuarine environ-
ment of the deposit. Probably there was a wide muddy delta with numerous
small islands and shallow water-courses at or near sea level.

The comparatively rich vegetation of such an area no doubt attracted
a wide variety of herbivores with their attendant carnivores. On the death
of these animals, the skeletons probably were shifted about as they fell
apart, and were rapidly buried in the mud.

The fauna is a selection of common late Pleistocene vertebrates, and
is especially similar to that from the better-known brea pits near La
Libertad. No recognizable specimens of Mastodon or Scelidodon were
found, but since the number of specimens recovered was relatively small,
they could have been excluded from the sample on the basis of chance
alone. The fauna is quite distinctively a coastal one, with none of the
highland mastodons or horses, or Glossotherium (Oreomylodon) wegneri.

The age of the fauna cannot be determined with any certainty, but it
is obviously fairly late, since there are no remains of Macrauchenia, toxo-
donts or other animals which persisted until fairly late in the Pleistocene
elsewhere. On the other hand there are remains of Smilodon and
Odocoileus, both relatively recent immigrants, thus indicating a late
Pleistocene age.

Many faunal collections of Pleistocene age are now known from the
northern part of South America, but these have received scant attention.
Only a few have been published, and little attempt has been made to
correlate them. It is hoped that a series of papers, similar to the present
publication, will permit description of the fauna and stratigraphy of these
occurrences with a view to understanding better the palaeozoogeography of
this interesting area.

20

REFERENCES

CHURCHER, C. S., 1962
Odocoileus salinae and Mazama sp. from the Talara tar seeps,
Peru. Royal Ontario Mus., Life Sci. Div., Contrib. 57, 1-27.
PRICK, C., 1937
Horned Ruminants of North America. Bull. Amer. Mus. Nat. Hist.
69: 205-214.
HOFFSTETTER, R., 1952

Les Mammifeéres Pléistocénes de la République de I’ Equateur. Soc.
Géol. France. Mém. 66: 1-391, 8 pls.
1953
Sur la présence d’un Tatou géant du genre Holmesina dans le
Pléistocéne de l’Equateur (Amérique du Sud). Soc. Géol. France.
C.R. Somm. 6: 101—2.
MARCHANT, S., 1961
A photogeological analysis of the structure of the western Guayas
Province, Ecuador: with discussion of the stratigraphy and Tablazo
formation, derived from surface mapping. Quart. J. Geol. Soc.
London 466; 215-232.
MERRIAM, J. C. AND STOCK, C., 1932
The Felidae of Rancho La Brea. Carnegie Inst. Wash., Publ. 422:
1-231, 42 pls.
PAULA COUTO, CARLOS DE, 1954
Megaterios Intertropicais do Pleistoceno. Anais da Academia
Brasileira de Ciencias 26, 3: 447-463.
SIMPSON, G. G., 1930
Holmesina septentrionalis, extinct giant armadillo of Florida. Am.
Mus. Novit. 442: 1-10.
SPILLMANN, F., 1931
Die Saugetiere Ecuadors im Wandel der Zeit. t.I: 1-112. Quito,
Univ. Central.
1948
Beitrage zur Kenntnis einer neuen gravigraden Reisensteppentieres
(Eremotherium carolinense gen. et. sp. nov.), seines lebensraumes
und seiner lebensweise. Paleobiologica, 8: 231-279.

21

' « *

Se ed ei

7

: \S
an

4 Te ne hts my hu
We e % sr sh ‘i Lan we ae y
AN 4

me ;
a” A F

"ay
oe” ear?

ye
va ey ‘ uM AY 1,8
i U oh) i Re
a vi wih
1 Wee Pi

ie

evar ys)
d i i a

Yea oe
‘ i

ait)

' iG i “ }

he

4
oe

TA ne ar ee iviet
ra eit im
Les By. Aa } EAT ris

i, ws
eae ere ="

=. T,

PS yw ee PF toy
ot,

eee ee

q

sates UN Se es es
JS SEN eee 5
SO Uwe es & ;

ithe oud Soe oe ee

ae -e ee rs : i en tect SEAT NON A SS ie “YP
a een

Rote. mee ee tk ee
Larne TS te :
Se? ee MS ere.

NGS a Ke

RR pk ola ee ene
oe ee Oe RT * I os,
- OT TES Bet mE MR
~ = saga ee
oy T RQSS pC EES OLE 0 gp
LA IA Fen ee:

on ren me
flit tae ae

RS ORD,

mone Mee ONS

Set ny Seen ea

— ve Rae Sea ee ee
. et i kt a -,
oe Seta ANOS Pag .
a - NF OS ar re oS aad ee ~ =, 3, ~ ae n
SPL, RCS Soe syn wap eae ee ee TS Se aS i Som eaters,

- pate’ amy orate Sans att nhdeeeneiion, Saeeneieisesienaaies satethae oe
Lele. Sota DI Pe RE NE eee OP Se eee ae
A a ade lo oo, 1 eb ee ae el EO ee ies

Pre nhi tite ey
IPE My tome e & POPP yea ae gh we nt 7 > v . = : Yet Hpi ae wie sarang chat a mae
CaP NSS RPE Pam ees mes Hap Soe a - ee 7 2 4 > ' ~ : 2 : = awe
PUP Oe ue ye ph o _ . i

NOt May Pe HLA IONI ST Oe .
¥ = * “< 1 aed te ae
a «te fas phan odin oe PRR 0 Re on yy ee =
ea aria te inh glee ns tt ee ie ae ee
Asa aa Ci re es ee

FPN See eet,
SENT Ae re Loe
FS LORS EGS I a er athe we pace

> EI OP Ay OO pene
ew. ore vee é me 7 ~~ >. ~ oS eH wee er oe "y —— ~ ta See ie ree
EE WET SL a © » <a . » 4 ° oute-g - 2 : v ow ak i el ‘ SO A a A Mere

PE TE GEST Le OS eA
SP POE ERTS" 6 A SE OG
STE NEAL ce PRE CERN Be ey.
NS RR tt RON ALIEN SEN SNE ie ca,
FAS TORE NS PY Bk ERT Jy Pe PANO mE
OS PL IE FO POT ee Dee = PENNS A RE) AR TET ESS BO oye,

] oor eee

1 : pe gw tgti"e 1° Ng OSH Ragen” ante — a 3 AN
° ‘ ithited<3.< na - s TO Oe aoa aia peeehies A nee ee ee rr en II ES CL IO GOR LE ee Si A
Re tame 3s rh taliis th x rar, pt emi Pia =} ES OUP Ne eer Mrbithitibie hia hae ae ee Ne ee ee Sterne,
ewe MST aX Ra ac wale haga emaye - : rie on pretties nae CI EAN GND GRE ON em cag eee tes Perl een? fee Leen SON tie ene
hak eet ~ i ni : . POL Le ICO IO RS Site OR GRP OOP Seles AES eee a
. er oF Aes ey SeTeee . < “ oF rs. ain 3 we
SPA scien co 54 tie ws pt ee oa toe in aint li ae : < - $ * ‘s 2 “0 : SS net Cig eg te hapa ~ Sy! We, es
Ae Sap r . eae my pon vx S ae he e ae c Gare ip tea oes re Lr go ene 1 ae P : —- t ” og ET at iS et, eden tate Ee penne yong PSHE 2 ey py ‘ =
ONS) oy Ordnginimget wh cat mon F = eins ae 3 a * y ~ ee : e 5% - Dacca anal * soe. r, es “te ye" en 1 LO Pm tae oe on ee =e = ae se im oe onerom
EEO er er OUP, ares ee, . Ores Sa eita tt at at oe tel ad ‘ “ . ee ee ee Ate eS Se wagt tat Sit gm core, pean em a " goer
Beye me ae, : " Pit FS 4 no a mer yn Flt Dome Sit = baat . Kitt 53 a a ne ae ere oe 2 eee v ee | =
bay Pane. Sette A aes _ f Sahota tae we ae ee ie aa ee NR PAE A ey re 3 nessa
mss ay reyes <4 erg Semmes,
Sacro aso rows ae (oe ote tamer Sia et Serre .
ai ee POT « a gO On ONCE OF -
a er a, *

SOY OPS RD
Sh aaa ty, ips kat Th ee)
RW a yr eet
SEP I RTO oar tek ree oe

ondtpey pitt talitca oacterme nie sonatas otal

. pi - = “J : Sete babcta ee
atin teeta yg n. ntet es ~ igt any . d ¢ TET 8 * rt a tee eee a, PO ELE EA

IOS ORE Mey

BP Pty RO = ne oe : se ERE SO, OR ae ntay AENEAN 1S GR Ps nO UAT, aA RES, een poe tpet ty A, aoa
ES BOLE OF PLLA DINE WED Nes AeA L Ot Arn gee gee EAT wee G “ DAF wana od Py : POCA): Beets hat yd SN
NS I RS RU ety Se eon en bs ie pes N 2 =~ TF NS i TNS TH motte oe Se,
ano oe med

a eT ek le FOLIO MBN EPS. OF ON PO
OO 7D HS A ae ar 8 oy
= By Er PS Peet ye Per PAP 8 Se heeed to oy:

Paiabin taht iad ith baka 1h Son Sabet dened dete ithe tee eee a
OY SOT Ae ry pm emer s TAL ot ST siethlenhnteeetated bottel ake sae
PE PROP LO AEET IE. Pn NOS a8 EG ee eg oe IN

Tey

5 = ~ : oR Re A yey ee f “\eure Sy DP en er, hes ear PP Ie WEP SSE OS ole
clef I NINE SS OEE eS ae es oe ee
Se ; 2 ERE fon wegrse emer ct

Seta. Re ee ir OR EA UT NER OD
pe Ss ne Me oat Be eeth fi no he = a

ere a0 Pie,
nue a"
co met y

* a

ee alt tee a ee ee

tasese: 4,
rommede way
+ wea enero Reet yey eka epe me

Aa Rt 8 0 genet y agige 20a 9s Saray ees we eye . : en :
Sugeno wn ey aaeet Epoastee en i ae Kees
iM pate 3 : etreene nego

5 fel Saat ta if

rey .
pt + POE PRET NEO Re Bere Dae Al a a Neath enpni tomer) gaan b iret
SCT a ash’ One wrt a er ey OE aT, LTO OOS Creep Seng fw ee da) é
See Neng cc AEE RRO tyre Hm in, Cl Cree Peg ene ht) LE ANT OOO aS EM ot WIE A Sets PAG me apogee
pth IkA metereiaeh te ke he ee Os tS pare Ry Darter 8 Oe De reels een eraees Me ree ;
PEI “OP eee CE ee rE er tee <. oe eo Ay8 9 et ee ae Pe a Qe COE NTI Of ERE rel ape
I AORN LR eR EE NIN Rete SER ey long at yen NT Nat REY 6 Nene ee ihe 4 ges Set Pon toi Se
Rts Pate Leet ea abi Grain att neh ict Gr bE Tae Bk nn, Oe cs Cae nn i ei ei i ae ud a ee ee oo